You are on page 1of 614

Preface

The present volume consistsof chapters by participants in the Language and Space . In most casesthe chapters conferenceheld in Tucson, Arizona , 16- 19 March 1994 have beenwritten to reflect the numerous interactions at the conference , and for that reason we hope the book is more than just a compilation of isolated papers. The conferencewas truly interdisciplinary , including such domains as neurophysiology, , and linguistics. Neural , psychology, anthropology , cognitive science neuropsychology es mechanisms , and cultural factors were all grist for the , developmental process mill , as were semantics , syntax, and cognitive maps. The conferencehad its beginnings in a seemingly innocent conversation in 1990 betweentwo new colleaguesat the University of Arizona (Bloom and Peterson ), who MAP .) assumed of them confusions. One of left right wondered about the genesis ( that theseconfusions reflecteda languageproblem; the other (P. B.) was quite certain that they reflected a visual perceptual problem. Curiously, it was the perception researcherwho saw this issueas being mainly linguistic and the languageresearcher who saw it as mainly perceptual. In true academic form they decided that the best way to arrive at an answer would be to hold a seminar on the topic , which they did the very next year. Their seminar on languageand spacewas attended by graduate students , postdoctoral fellows, and many faculty membersfrom a variety of departments . Rather than answering the question that led to its inception, the seminar ? What aspectsof spacecan we raised other questions: How do we represent space ? And what role doesculture play in talk about? How do we learn to talk about space all thesematters? One seminar could not explore all of theseissuesin any depth; an enlarged group of interestedcolleagues(the four coeditors) felt that perhaps several workshops might . The Cognitive NeuroscienceProgram at the University of Arizona , in collaboration with the Cognitive ScienceProgram and the PsychologyDepartment, sponsored . Although two one-day workshops on the relations between space and language : other rise to still questions How does stimulating and helpful, the workshops gave

-

VIII

Preface

? How many kinds of spatial representations are there? the brain represent space ? Should What happensto spatial representationsafter various kinds of brain damage to closed restricted be and between relations of the tests language space experimental classlinguistic elementsor must the role of open-classelementsbe consideredas well? Given the scopeof thesequestion, we decidedto invite investigators from a variety . of disciplines to a major scientific conference , and Language and Spacetook shape . We do not imagine that the The conferencewas judged by all to be a great success chaptersin this book provide final answersto any of the questionswe first raised, but we are confident that they add much to the discussionand demonstrate the importance . We expectthat increasedattention of the relations betweenspaceand language will be given to this fascinating subject in the years ahead and hope that our conference , and this book , have made a significant contribution to its understanding. Meetings cannot be held without the efforts of a considerablenumber of people, . Our thanks to Pauline Smalley for all work and the support of many funding sources she did in organizing the conferenceand making sure participants got to the right place at the right time and to Wendy Wilkins , of Arizona State University, for her . We gratefully acknowledgethe gracious help both before and during the conference ' : McDonnell Pew Cognitive NeuroscienceProgram support of the conferences sponsors , the Flinn Foundation Cognitive Neuroscience Program, and the Cognitive ScienceProgram and Department of Psychology at the University of Arizona . We , which greatly thank the participants for their intellectual energy and enthusiasm ' of the MIT Pierce thank we . contributed to the conferences success Finally , Amy Pressfor her help with this volume. Editors Bloom and Petersontosseda coin one eveningover margaritas to determine whose name would go first.

Chapter of the Linguistic-Spatial Interface The Architecture ~ Ray Jackendoff

1.1

Introduction

? More specifically How do we talk about what we see , how does the mind / brain encodespatial information (visual or otherwise), how does it encodelinguistic information , and how does it communicate betweenthe two? This chapter lays out some of the boundary conditions for a satisfactory answerto thesequestionsand illustrates the approach with somesampleproblems. The skeleton of an answer appears in figure 1.1. At the language end, speech perception converts auditory information into linguistic information , and speech production converts linguistic information into motor instructions to the vocal tract. Linguistic information includes at least somesort of phonetic/phonological encoding es of visual perception convert retinal information . ! At the visual end, the process of speech . into visual information , which includes at least some sort of retinotopic mapping. The connection betweenlanguageand vision is symbolized by the central it is clear there cannot be a direct relation double-headedarrow in figure 1.1. Because betweena retinotopic map and a phonological encoding, the solution to our problem lies in elaborating the structure of this double-headedarrow.

1.2 Representational Modularity
The overall hypothesisunder which I will elaborate figure 1.1 might be termed Representational , chapter I ) . , chapter 12; Jackendoff 1992 Modularity (Jackendoff 1987 formats distinct in information encodes mind brain that the The generalidea is many / " " for or languagesof the mind. There is a module of mind/ brain responsible each of these formats. For example, phonological structure and syntactic structure are distinct levels of encoding, with distinct and only partly commensurateprimitives and principles of combination. RepresentationalModularity therefore posits that the architecture . Each of the mind / brain devotesseparatemodules to thesetwo encodings

Ray Jackendoff

auditory signals ---.........

...- eye 4 ~ visual information information linguistic ~ motor signals ~ - - - -_ J C \ - - - "' -- Y - - - ---- I ~ ~ - -yVISION LANGUAGE

Figure 1.1 Coarse sketch of the relation betweenlanguageand vision.

of thesemodules is domain-specific (phonology and syntax, respectively ); and (with " in Fodor ' s " certain caveatsto follow shortly) each is informationally encapsulated . Representational modules differ from Fodorian modules in that they ) sense ( 1983 are individuated by the representationsthey processrather than by their function as faculties for input or output ; that is, they are at the scale of individual levels of , rather than being entire faculties such as languageperception. representation A conceptual difficulty with Fodorian Modularity is that it leavesunansweredhow ; modules communicate with each other and how they communicate with Fodor s ' central, nonmodular cognitive core. In particular , Fodor s languageperception module ' " derives " shallow representations - some form of syntactic structure; Fodor s " " " central faculty of " belief fixation operatesin terms of the languageof thought , a " " nonlinguistic encoding. But Fodor doesnot tell us how shallow representations are " " converted to the languageof thought, as they must be if linguistic communication is to affect belief fixation . In effect, the language module is so domain-specific and informationally encapsulatedthat nothing can get out of it to serve cognitive purposes .2 And without a theory of intermodular communication, it is impossible to approach the problem we are dealing with here, namely, how the languageand vision modules manageto interact with each other. es this difficulty by positing, in The theory of RepresentationalModularity address . modulesproposed above, a systemof interfacemodules addition to the representation An interface module communicatesbetweentwo levels of encoding, say Ll and L2 , by carrying a partial translation of information in Ll form into information in L2 : the phonologyform. An interfacemodule, like a Fodorian module, is domain-specific to-syntax interface module, for instance , knows only about phonology and syntax, -purpose audition . Such a module is also in not about visual perception or general : the phonology-to -syntax module dumbly takes whatever formationally encapsulated phonological inputs are available in the phonology representationmodule, translates the appropriate parts of them into (partial) syntactic structures, and delivers them to the syntax representation module, with no help or interference from , say, beliefs about the social context. In short, the communication among languagesof the mind es as well.3 is mediated by modular process

-Spatial The Architecture of the Linguistic Interface
g-p

auditory ............ ........- phonology ~ .. motor
eye ~ retinotopic
. ~

~ syntax

4

~

imagistic

/ ,haptic *,,action localization .... auditory
.. ~

audition ,smell ,emotion ,... / , * structure / :..~ conceptual tresentation spatial rep ;

1.2 . Figure less sketch of coarse Slightly

the relation between language and vision .

The levelsof representationI will be working with here, and the interfaces among them, are sketchedin figure 1.2. Each label in figure 1.2 standsfor a level of representation served by a representation module. The arrows stand for interface modules. Double-headedarrows can be thought of either as interface modules that processbi directionally or as pairs of complementary unidirectional modules (the correct choice is an empirical question) . For instance , the phonology-syntax interface functions from left to right in speechperception and from right to left in speechproduction . " " Figure 1.2 expands the linguistic representation of figure 1.1 into three levels involved with language : the familiar levelsof phonology and syntax, plus conceptual structure, a central level of representation that interfaces with many other faculties. " " Similarly, visual representation in figure 1.1 is expandedinto levelsof retinotopic, ' imagistic, and spatial representation , corresponding roughly to Marr s ( 1982 ) primal sketch, 21 0 sketch, and 3 D model, respectively ; the last of theseagain is a central representationthat interfaceswith other faculties. In this picture, the effect of Fodor ian faculty -sized modules emergesthrough the linkup of a seriesof representation and interface modules; communication among Fodorian faculties is accomplishedby interface modules of exactly the same general character as the interface modules within faculties. The crucial interface for our purposeshere is that betweenthe most central levels of the linguistic and visual faculties, conceptual structure and spatial representation . Beforeexamining this interface, we have to discusstwo things: ( I ) the generalcharacter of interfaces betweenrepresentations(section 1.3); and (2) the general character of conceptual structure and spatial representationthemselves (sections 1.4 and 1.5) .

1.3 Character of Interface Mappings
To say that an interface module " translates " between two representations is , strictly speaking , inaccurate . In order to be more precise, let us focus for a moment on the

Ray Jackendotr

interface between phonology and syntax, the two best-understood levels of mental . representation It is obvious that there cannot be a complete translation betweenphonology and syntax. Many details of phonology, most notably the segmentalcontent of words, , many details of syntax, for instance the play no role at all in syntax. Conversely elaborate layering of specifiersand of arguments and adjuncts, are not reflected in phonology. In fact, a complete, information -preserving translation betweenthe two representationswould be pointless; it would in effect make them notational variants - which they clearly are not. The relation between phonology and syntax is actually something more like a partial homomorphism. The two representationsshare the notion of word (and perhaps .4 But ), and they share the linear order of words and morphemes morpheme segmentaland stressinformation in phonology has no direct counterpart in syntax; and syntactic category (N , V , PP, etc.) and case , number, gender , and person features 5 have no direct phonological counterparts. Moreover, syntactic and phonological constituent structures often fail to match. A classicexampleis given in ( I ) . ( I ) Phonological: [ Thisis the cat] [that ate the rat] [that ate the cheese ] Syntactic: [ Thisis [the cat [that ate [the rat [that ate [the cheese ]]]]]] The phonological bracketing, a flat tripartite structure, contrasts with the relentless , English articles cliticize phoright -embeddedsyntactic structure. At a smaller scale nologically to the following word , resulting in bracketing mismatches such as (2) . (2) Phonological: [the [ big]] [ house ] Syntactic: [the [ big [ house ]] Thus, in general, the phonology-syntax interface module createsonly partial correbetweenthesetwo levels . spondences A similar situation obtains with the interface between auditory information and phonological structure. The complex mappingbetweenwaveforms and phonetic segmentation in a sense the relative order of information : a particular auditory preserves cue may provide evidencefor a number of adjacent phonetic segments , and a particular be a number of phonetic segmentmay signaledby , but the adjacent auditory cues " bands" of the in an stream overlapping correspondenceprogress through speech orderly linear fashion. On the other hand, boundaries betweenwords, omnipresentin phonological structure, are not reliably detectable in the auditory signal; contrari -

The Architecture of the Linguistic - Spatial Interface

wise, the auditory signal contains information about the formant frequenciesof the ' speakers voice that are invisible to phonology. So again the interface module takes only certain information from each representation into account in establishing a betweenthem. correspondence These examples show that each level of representation has its own proprietary information , and that an interface module communicates only certain aspects of this information to the next level up- or downstream. Representational modules, : precisely to the extent that they then, are not entirely informationally encapsulated receiveinformation through interface modules, they are influenced by other parts of the mind.6 In addition to general principles of mapping, such as order preservation, an interface module can also make use of specialized learned mappings. The clearest instances of suchmappings are lexical items. For instance , the lexical item cat stipulates that the phonological structure / kret/ can be mapped simultaneously into a syntactic ' noun and into a conceptual structure that encodesthe word s meaning. In other words, the theory of Representational Modularity leads us to regard the lexicon as a learned component of the interface modules within the language faculty (see Jackendoff forthcoming) .

Structure 1.4 Conceptual
Let us now turn to the crucial modules for the connection of language and spatial cognition : conceptual structure (CS) and spatial representation (SR) . The idea that these two levels share the work of cognition is in a sensea more abstract version . To use the terms of Mandler (chapter 9, of Paivio' s ( 1971 ) dual coding hypothesis this volume), Tversky (chapter 12, this volume), and Johnson- Laird (chapter II , this " " volume), CS encodes" propositional representations , and SR is the locus of image " " " . schema or mental model representations in , 1990 ) is an encoding of Conceptual structure, as developed Jackendoff ( 1983 linguistic meaning that is independent of the particular languagewhose meaning it . It is an " algebraic" representation encodes , in the sensethat conceptual structures are built up out of discrete primitive features and functions. Although CS supports " " formal rules of inference , in , it is not propositional in the standard logical sense that ( I ) propositional truth and falsity are not the only issueit is designedto address , and (2) unlike propositions of standard truth -conditional logic, its expressionsrefer not to the real world or to possibleworlds, but rather to the world as we conceptualize it . Conceptual structure is also not entirely digital , in that some conceptual features and some interactions among features have continuous (i.e., analog) characteristics effectsto be formulated. that permit stereotypeand family resemblance

Ray Jackendoff

The theory of conceptualstructure differs from most approach es to model-theoretic ' " " semanticsas well as from Fodor s ( 1975 ) Languageof Thought , in that it takes for " " grant~ that lexical items have decompositions ( lexical conceptual structures, or LCSs) made up of features and functions of the primitive vocabulary. Here the approach concurs with the main traditions in lexical semantics(Miller and JohnsonLaird 1976 ; Lehrer and Kittay 1992 ; Pinker 1989 ; Pustejovsky 1995 , to cite only a few . parochial examples ) As the mental encoding of meaning, conceptual structure must include all the . A sample : nonsensorydistinctions of meaning made by natural language I . CS must contain pointers to all the sensorymodalities, so that sensoryencodings and correlated (seenext section may be accessed ). 2. CS must contain the distinction betweentokens and types, so that the concept of an individual (say a particular dog) can be distinguished from the concept of the type to which that individual belongs (all dogs, or dogs of its breed, or dogs that it lives with , or all animals) . 3. CS must contain the encoding of quantification and quantifier scope . 4. CS must be able to abstract actions (say running) away from the individual performing the action (say Harry or Harriet running) . 5. CS must encodetaxonomic relations (e.g., a bird is a kind of animal) . 6. CS must encodesocial predicatessuch as " is uncle of ," " is a friend of ," " is fair ," and " is obligated to." " " 7. CS must encode modal predicates , such as the distinction between is flying, " " can " and " " isn' t ' " can t fly . flying , fly , I leaveit to my readersto convince themselves that none of theseaspectsof meaning can be representedin sensoryencodings without using special annotations (such as , or footnotes); CS is, at the very least, the systematicform in which pointers, legends such annotations are couched. For a first approximation, the interface between CS and syntax preservesembedding relations among constituents. That is, if a syntactic constituent X express es the CS constituent X ' , and if another syntactic constituentY express es the CS constituent Y' , and if X contains Y, then, as a rule, X ' contains Y' . Moreover, a verb (or other argument-taking item) in syntax corresponds to a function in CS, and the subject and object of the verb normally correspond to CS argumentsof the function . Hence much of the overall structure of syntax corresponds to CS structure. (Some instancesin which relative embeddingis not preservedappear in Levin and Rapoport 1988and Jackendoff 1990 , chapter 10.) Unlike syntax, though, CS has no notion of linear order: it must be indifferent as to whether it is expressedsyntactically in , say, English, where the verb precedes

-SpatialInterface of the Linguistic TheArchitecture

7

the direct object, or Japanese , where the verb follows the direct object. Rather, the 7 embeddingin CS is purely relational. At the same time, there are aspectsof CS to which syntax is indifferent. Most prominently , other than argument structure, much of the conceptual material bundled up inside a lexical item is invisible to syntax, just as phonological features are. As far as syntax is concerned , the meaningsof cat and dog (which have no argument structure) are identical, as are the meanings of eat and drink (which have the same argument structure) : the syntactic reflexes of differences in lexical meaning are . extremely coarse In addition , some bits of material in CS are absent from syntactic realization ), is (3) . , given by Talmy ( 1978 altogether. A good example (3) The light flashed until dawn. . But this repetition is The interpretation of (3) contains the notion of repeatedflashes not coded in the verbflash : Thelight flashed normally denotesonly a single flash. Nor is the repetition encodedin until dawn, because , Bill slept until dawndoes , for instance of notion . Rather, the not imply repeatedacts of sleeping (a) repetition arisesbecause ; (b) the light until dawn givesthe temporal bound of an otherwise unbounded process make these c to and bounded therefore event and a is ; ) ( temporally point flashed " coercion" 1991 Jackendoff or construal of a ; (Pustejovsky compatible, principle This notion of . in time out 1991 by repetition ) interprets the flashing as stretched repetition, then, appearsin the CS of (3) but not in the LCS of any of its words. The upshot is that the correspondencebetween syntax and CS is much like the correspondencebetweensyntax and phonology. Certain parts of the two structures are in fairly regular correspondenceand are communicated by the interface module, but many parts of each are invisible to the other. Even though CS is universal, languagescan differ in their overall semantic patterns . First , languagescan have different strategiesin how , in at least three respects bundle , Talmy up conceptual elementsinto lexical items. For example they typically ) documents how English builds verbs of motion primarily by bundling up ( 1980 motion with accompanying manner, while Romance languagesbundle up motion primarily with path of motion , and Atsugewi bundles up motion primarily with the type of object or substanceundergoing motion . Levinson (chapter 4, this volume) shows how the Guugu Yimithirr lexicon restricts the choice of spatial frames of referenceto cardinal directions (see section 1.8) . These strategies of lexical choice . ( This is affect the overall grain of semanticnotions available in a particular language of course in addition to differencesin meaning among individual lexical items across , such as the differences among prepositions discussed by Bowerman, languages chapter 10, this volume.)

8

T RayJackendot

Second , languagescan differ in what elementsof conceptual structure they require the speakerto expressin syntax. For example , French and Japanese require speakers to differentiate their social relation to their addressee always , a factor largely absent from English. Finnish and Hungarian require speakersto expressthe multiplicity (or , using iterative aspect , a factor absent from English, as seenin repetition) of events (3) . On the other hand, English requiresspeakersto expressthe multiplicity of objects . by using the plural suffix, a requirement absent in Chinese Third , languages can differ in the specialsyntactic constructions they useto express particular conceptual notions. Examples in English are the tag question (They shoot ' " " ' horses , don t they?), the One more construction (One more beer and I m leaving ) " " (Culicover 1972 ), and the The more . . . , the more construction ( The more you drink , the worseyou feel ). These all convey special nuancesthat go beyond lexical mean Ing . " level -specific" semantic 1 have argued (Jackendoff 1983 ) that there is no language of representation intervening between syntax and conceptual structure. Language differences in semantics of the sort listed are in localized the interface specific just between syntactic and conceptual structures. 1 part company here with Bierwisch ( 1986 ), Partee ( 1993 ), and to a certain extent Pinker ( 1989 ) . Within my approach, a , in part becausethe syntax- CS interface module separatesemanticlevel is unnecessary has enough richnessin it to capture the relevant differences ; 1 suspectthat these other theories have not considered closely enough the properties of the interface. However, the issuesare at this point far from resolved . The main point , on which Bierwisch, Pinker, and 1agree(I am unclear about Partee ), is that there is alanguageindependent and universal level of CS, whether directly interfacing with syntax or mediated by an intervening level.

1.5 SpatialRepresentation
- the encoding of objects and their configurations For the theory of spatial representation - we are on far shakier ground. The best articulated (partial) in space theory ' of spatial representation I know of is Marr ' s ( 1982 ) 3-D model, with Biedermans " " ( 1987 ) geonic constructions as a particular variant. Here are some criteria that a spatial representation(SR) must satisfy. I . SR must encode the shapeof objects in a form suitable for recognizing an object at different distancesand from different perspectives , that is, it must solve the classic 8 of . problem object constancy 2. SR must be capable of encoding spatial knowledge of parts of objects that cannot be seen , for instance , the hollownessof a balloon.

The Architecture

of the Linguistic - Spatial Interface

3. SR must be capableof encoding the degrees of freedom in objects that canchange their shape , for instance , human and animal bodies. 4. SR must be capable of encoding shapevariations among objects of similar visual , making explicit the range of shape variations characteristic of type, for example different cups. That is, it must support visual object categorizationas well as visual object identification. 5. SR must be suitable for encoding the full spatial layout of a sceneand formediating " among alternative perspectives( What would this scene look like from over " there? ), so that it can be used to support reaching, navigating, and giving instructions (Tversky, chapter 12, this volume) . 6. SR must be independentof spatial modality , so that haptic information , information from auditory localization, and felt body position (proprioception) can all be brought into registration with one another. It is important to know by looking at an object where you expect to find it when you reach for it and what it should feel like when you handle it . , criteria 5 and 6 go beyond the Marr and Biederman theories of Strictly speaking . But there is nothing in principle to prevent thesetheories from serving object shape as a component of a fuller theory of spatial understanding, rather than strictly as theories of high-level visual shape recognition. By the time visual information is converted into shapeinformation , its strictly visual character is lost- it is no longer ' - nor , as Marr stress es retinotopic , for example , is it confined to the observers point 9 ofview . SR contrasts with CS in that it is geometric (or even quasi-topological) in character , rather than algebraic. But on the other hand, it is not " imagistic" - it is not to be " " thought of as encoding statuesin the head. An image is restricted to a particular point of view, whereasSR is not . An image is restricted to a particular instance of a ' category (recall Berkeley s objection to imagesas the vehicle of thought : how can an image of a particular triangle stand for all possible triangles?! O ), whereasSR is not. An image cannot representthe unseenparts of an object- its back and inside, and ' s view other the parts of it occluded from the observer by objects- whereasSR does. An image is restricted to the visual modality , whereas SR can equally well encode information receivedhaptically or through proprioception. Nevertheless , even though SRs are not themselves imagistic, it makessenseto think of them as encoding image schemas : abstract representationsfrom which a variety of imagescan be generated . Figure 1.2 postulates a separatemodule of imagistic (or pictorial ) representation one level toward the eye from SR. This correspondsroughly to Marr ' s 2t -O sketch. It is specifically visual; it encodeswhat is consciouslypresent in the field of vision or visual imagery (Jackendoff 1987 , chapter 14 ) . The visual imagistic representation is

Ray JackendofT

restricted to a particular point of view at anyone time; it doesnot representthe backs and insides of objects explicitly . At the sametime, it is not a retinotopic representation becauseit is normalized for eye movementsand incorporates information from both eyesinto a single field, including stereopsis . (There is doubtlessa parallel imagistic representationfor the haptic faculty , encoding the way objects feel, but I am not aware of any researchon it .) It is perhapsuseful to think of the imagistic representationas " perceptual" and SR as " cognitive" ; the two are related through an interface of the general sort found in the languagefaculty : they sharecertain aspects , but each has certain aspectsinvisible to the other. Each can drive the other through the interface: in visual perception, an imagistic representation gives rise to a spatial representation that encodesone' s ; in visual imagery, SRs give rise to imagistic representations understanding of the visual scene . In other words, the relation of images to image schemas(SRs) in the to thoughts. Image schemas are present theory is much like the relation of sentences not skeletal images , but rather structures in a more abstract and more central form of representation . 11 This layout of the visual and spatial levels of representation is of course highly , I have not addressedthe well-known division of visual oversimplified. For instance " and the " where " " labor between the what system , which deal, roughly system ' , with object identification and object location respectively (O Keefe and speaking Nadel 1978 rod 1994 ; Ungerleider and Mishkin 1982 ; Farah et al. 1988 ; Jeanne ; Landau and Jackendoff 1993 ). My assumption, perhaps unduly optimistic , is that such division of labor can be captured in the present approach by further articulation of the visual-spatial modules in figure 1.2 into smaller modules and their interfaces , much as figure 1.2 is a further articulation of figure 1.1.

1.6 Interface between CS andSR We comeat last to the mappingbetween CS and SR, the crucial link between the 12 visualsystem and the linguisticsystem . What do these two levels share , suchthat it
is possiblefor an interface module to communicate betweenthem? The most basic unit they share is the notion of a physical object, which appearsas a geometrical unit in SR and as a fundamental algebraic constituent type in CS. 13In addition , the Marr -Biedermantheory of object shapeproposesthat object shapesare decomposedinto geometric parts in SR. This relation maps straightforwardly into the part -whole relation , a basic function in CS that of course generalizesfar beyond object parts. The notions of place (or location) and path (or trajectory) playa basic role in CS ; Jackendoff 1983 (Talmy 1983 ; Langacker 1986 ); they are invoked, for instance , in

the entry for dog might look something like (4) . and because locations and paths can be given obvious geometric counterparts. Individual . More specula tively. The difficulty with an image of a prototype is that it is computa: it does not meet the demands of object shape identification tionally nonefficacious laid out as criteria 1. SR encodes all the details of object shapes . On the other hand. .Biederman theory does not contain placesand paths becausethey arise only in encoding the behavior of objects in the full spatial field. (4) Id ~gl + N . and conceptual structure. but that theseare invisible to SR. + count .g.) The notion of physical motion is also central to CS. Becausethesesentences visual input . These German shepherd features do not lend themselves at all to the geometric sort of algebraic coding found in CS. Jackendoff 1990 ). ( Learned and stored) lexical entries for physical object words can contain a spatial representation of the object in question. syntactic. representations Our discussionof interfacesin previous sectionsleadsus to expect someaspectsof each representationto be invisible to the other. the category of dogs). it is a good bet that these constituents are shared between CS and SR. the notion of force appearsprominently in CS (Talmy 1985 . the shapeof violin or a butter knife or a ' s ears. and obviously it must be representedin spatial cognition so that we can track moving objects. and taxonomic relations (a bird is a kind of animal). A more abstract spatial representation .4 noted that CS encodesthe token versustype distinction (a particular dog vs. 14(The Marr . Rosch and Mervis " 1975 . in addition to their phonological. quantificational relations. and to the extent that we have the impression of directly perceiving forces in the visual field (Michotte 1954 ). . individual matchings can be learned and stored. What might someof theseaspectsbe? Section 1. + sing.4 in the previous section .The Architecture of the Linguistic -Spatial Interface locational sentences such as The book is lying on tile table (place ) and The arrow flew can be checked against through tile llir past my llead (path) . Type of Carnivore Function: (often) Type of Pet SR: [3-D model wi motion affordances ] : Auditory [sound of barking] Phono: Syntax: CS: In (4) the SR takes the place of what in many approaches (e. Type of Animal . Putnam 1975 ) has been informally called an image of a prototypical instance " of the category. .V . In addition to generalmappings betweenconstituent types in CS and SR.. For instance . . they are absolutely natural to (at least the spirit of ) SR. an aspect of visual cognition not addressedby thesetheories. these too might well be shared between the two 1S . for instance .

is a specialist .3a. The structures that make this a " lexical item" rather than just a " concept " : the simply representan additional modality into which this concept extends . we can state a criterion of : all other if a certain kind of distinction is encodedin SR.CONCEPr Figure1. it may have an SR as well.3 Two waysto viewtheintegrationof spatialstructures . details of shape are not duplicated in CS.the concept of a dog. Phonological and syntactic structures can then be viewed as further structures tacked onto to this knowledge to make it linguistically . the mind has to expressible have a way to unify multimodal representationsand store them as units (that is. meetsthesecriteria much better. The idea. much of which could be shared by a nonlinguistic organism.3b. and that each particular kind of information belongs in a particular module. With or without language . is that the " meaning" of a word goes beyond the features and functions available in CS. But an alternative suggested by figure 1. (4) represents just such a unit . What is left is the nonlinguistic " " knowledge one has of dogs. For the general case . we must raise the ? The question of exactly what information is on either side of it . it is therefore a more satisfactory candidate for encoding one' s knowledgeof what the object looks like. of course is that each module .) Such an approach might be seen as threatening the linguistic integrity of lexical items: as . in particular permit ting detailed shape information in a lexical SR. . a lexical entry should encode(pointers to ) other sensorycharacteristicsas well. For instance . How do we decide overall premise behind RepresentationalModularity . it breaks out of the purely linguistic system view of entries like (4) places them in a different light . to establish long-term memory " binding " in the neuroscience sense ). then. One way to view (4) + + CS Syntax IPhonology I+SA LANGUAGE ? ? ? b. As suggestedby the inclusion of " auditory structure" in (4). economy things being equal . Suppose one deletes the phonological and syntactic structures from (4) . linguistic modality Having establishedgeneral properties of the CS.SR interface. and taxonomic relations are not duplicated in SR. (A word must have a lexical CS.Ray Jackendoff a. Anotherway to view (4) LANGUAGE + Syntax IPhonology I+[~ !:~ ~ ~ . as suggestedin figure 1. into lexicalentries along the lines of a Marr 3-D model.

just ( Note applies only to syntactic and not lexical differences As pointed out in section 1. The count-massdistinction has repercussions in the marking of grammatical number and in the choice of possible determiners (count nouns usemany and few. that the information encodedby CS and SR is entirely incommensurate Let us call this the criterion of interfacing. if a semanticdistinction is communicatedto syntax. Rather. so that it makes a syntactic difference . Now if a particular semanticdistinction parallels to the semanticsof spatial concepts appearsin nonspatial domains as well as in the spatial domain. multiple individuals (a herd of cows). . semantically . the count-massdistinction appearsin abstract domains. Jackendoff 1976 . dog and cat look exactly the sameto syntax.) According to the criterion of economy. Therefore. Furthermore. all other things are not equal. it cannot be encoded in SR alone.thesehave radically different and spatial behavior over time (Marr and Biederman appearances . and substances (milk ). I take this maximal segregation to be the default assumption. Thus we do not expect . Lakoff and Johnson 1980 . that distinction must be present in CS and not SR. I will call this the criterion of nonspatialabstraction. which by definition pertains only to spatial cognition . they must share at least the notions mentioned at the beginning of this section.The Architecture of the Linguistic -Spatial Interface it should not also be encodedin CS. as a baseline . As is well known : Talmy 1978 . What evidencewould help decidewhether a certain kind of information is in CS as well as SR? One line of argument comesfrom interaction with syntax. have little or nothing to say about what substances look like. * muchthreat but the threat is grammatically a count noun (many threatsf ).for instance . differences But all elseis not equal. massnouns usemuch and little . A secondline of argument concernsnonspatial domains of CS.7 A SimpleCase : TheCount A familiar example will make thesecriteria clearer. for example ) . of course. SR obviously must make a distinction betweensingle individuals (a cow). The two modules must share enough structure that they can communicate with each other.) Let us call this the criterion of grammatical effect. that this criterion . Consider the count-massdistinction . Hence the criterion of grammatical effect suggests that the count-massdistinction is encodedin CS also. 1983 (Gruber 1965 . 1986 the semantics of Langacker ). Recall that CS is by hypothesis the form of central representation that most directly interacts with syntactic structure. and vice versa . many nonspatial conceptual domains show strong . Of course .4. SR should be the only level that encodes these . For example . -Mag Distinction 1. similarities between spatial and nonspatial domains must be captured in the algebraic structure of CS. all else being equal.

An example is the contrast between noodles (count) and spaghetti (mass ). unlike standard parts such as a handleor a leg. it is qualitatively different from that of objects. Jackendoff 1991 . (In Italian . and one can refer to a single spa ghetto. and ends behavegrammatically like parts of the object.RayJackendoff very similar adviceis a massnoun (much advicej* many advices ). the criterion of nonspatial extensionapplies to this case .its top. it is clear that SR and syntax alone cannot make sense of the discrepancy . 1993 .syntax interface. Hencethere must be a mismatch somewherebetweenSR and syntax. Rather. A single one of these objects can be described as a singular noodle .g. and . I . there is no reasonto believethat they havedifferent lexical SRs.. 16 . 1. In addition . respectively Whichever solution is chosen . Hinrichs 1985 Dowty 1979 . the count-mass distinction is closely interwoven with features of temporal event structure such as the event-processdistinction ( Verkuyl 1972 . Rather. then syntax follows suit by treating the concepts in question as grammatically count or mass .down axis determines top and bottom .8 Axes and Framesof Reference We now turn to a more complex casewith a different outcome. front . Thus the criteria of nonspatial extension and grammatical effect both apply again to argue for the count-massdistinction being encodedin CS. the mismatch might be betweenCS and SR. A further piece of evidencecomes from lexical discrepanciesin the grammar of count and mass nouns.the anything only place to put it is in CS. Bloom 1994 ) is to treat them as alike in CS as well and to localize the mismatch somewherein the CS. I will call them collectively the " axial " vocabulary. In this scenario .) Becausenoodlesand spaghetti pick out similar entities in the world . sides . Becausethe distinction ' t " look between threats and advice cannot be encoded spatially.it doesn like " . And distinctions of temporal event structure have a multitude of grammatical reflexes. the front -back axis determines front and back. but . The up. they are regions of the object (or its boundary) determined by their relation to the object' s axes . Alternatively . The " axial parts" of an object. bottom. Pustejovsky 1991 ) . spaghettiis a plural count noun.nouns that pick out essentially the same sorts of entities in the world . CS is necessary as an intermediary betweenthem. Three subsetsof the vocabulary invoke the spatial axesof an object. CS has the option of encoding a collection of smallish objects (or even largish objects such asfurniture ) as either an aggregateor a substance . they have no distinctive shape . That is. A standard strategy (e. To the extent that eventshave a spatial appearance . . but the massnoun forcesone to usethe phrasal form stick (or strand) of spaghetti . back.

Bierwisch and Lang 1989 ).The Architecture of the Linguistic -Spatial Interface a complex set of criteria distinguishing horizontal axes detennines sides and ends . which in (many of these appear as special casesin Miller and Johnson. and practically every chapter in this volume) that the axial vocabulary is always used in the context of an assumedframe of reference . the choice of frame of referenceis often ambiguous. . In the motion frame. Certain spatial prepositions. among ) A . For instance axes . and a deictic or observer more complex. Miller and Johnson. 3. For instance . The " dimensional adjectives . thick. it is possibleto distinguish at least eight different available frames of reference . the dimension of greatestextensioncan determine its length 1. the axial vocabulary too is ambiguous. next to. For instance . " 2. the front of an otherwise symmetrical double-ended tram is the end facing toward its current direction of motion (figure 1.Laird 1976 .4b) . left of. . Notice that many of the axial " are prepositions morphologically related to nouns that denote axial parts. in front of X denotes a object s axes out into the surrounding space region of space in proximity to the projection of X' s front -back axis beyond the . in front of. ) By referenceonly to the region subtendedby X . behind .Laird 1976 turn cites Bierwisch 1967 and Fillmore 1971 others . Landau boundary of X in the frontward direction (Miller and Johnson-Laird 1976 and Jackendoff 1993 Landau 8 this volume . A specialcaseconcernsanimals. Landau and Jackendoff 1993 (Miller and Johnson-Laird 1976 ). chapter . The geometric frame usesthe geometry of the object itself to determine the . sometimeswith referenceto tertiary axes the horizontality or verticality of these axes (Bierwisch 1967 . whosefront is intrinsically marked by the position of the eyes . . contrast inside X makes . secondary and . . and out a detennined .to -bottom axis dividing the symmetrical halvesand a side-to-side axis passingfrom one half to the other (figure 1. pick region by extending the reference ' . width. Teller 1969 . Actually the situation is centeredframe. not to any of its axes . such as above . and depth refer to dimensions of objects measuredalong principal . (figure ) Symmetrical geometry often implies a top. wide nalizations height. length. . and deep and their nomi high. near X denotesa " region in proximity to X in any direction at all. and becausethe frame determines the axesin tenDSof which the axial vocabulary receives its denotation. .4a . 2. Moreover. long. It has been frequently noted (for instance . thickness . alongside right of. Olson and Bialystok 1983 . below .4c) . The literature usually invokes two frames of reference : an intrinsic or object -centeredframe. Four intrinsic frames all make referenceto properties of the object: I . Viewing a frame of referenceas a way of determining the axes of an object. thefront of a moving object is determined by the direction of motion .

1 ~~ ~~ ~ ~~ ~ f":' -- \t'( . For instance . For instance ...4d has the wheelslowermost. The canon ical orientation frame designatesas the top (or bottom ) of an object the part which in the object' s normal orientation is uppermost (or lowermost). even though it is pointing obliquely upward in this picture. the part of a house where the public enters is .. the canonical orientation of the car in figure 1. so the part the wheels are attached to is the canonical bottom . f~ Two intrinsic framesdependon functional properties of the object. Intrinsic parts of an object can also be picked out according to the canonical encounterframe. WI.. RayJackendoff w. even if it does not happen to be at the moment.

5 . t " (. .- . such as Hausa. The gravitational frame is determined by the direction of gravity . Figure1. in as .The Architecture of the Linguistic -Spatial Interface l'r :J1 1 . 0 . fr8 'l\~ . mirroring observer frame.4e) . imposing axes on the object based on the cardinal directions north . for instance is on top of the car. chapter 4. . The figure on the page inherits this axis.) than the the wall of a different building may be the Four environmentalframes project axesonto object basedon properties of the environment: 1. regardlessof . the front is the side that the public normally faces . . so that the front from the inside front from the outside. south. and therefore its width is measured in the samedirection. In this frame. another object. this volume) . east.5b page figure geometric pictures page figure an intrinsic side-to -side axis that determines its width . The contextual frame is available when the object is viewed in relation to . For instance has .5c. the hat in figure 1. or a similar system(Levinson.5a the orientation of the object. whose own axesare imposed on the first object. figure facing the observer " . in some languages _ . 2. The geographical frame is the horizontal counterpart of the gravitational frame. . 4. The a which is drawn on a 1. regardlessof orientation . _ ~f ~ . The observerframe may be projected onto the object from a real or hypothetical es the front of the object as the side . frames reference Environmental functionally the front (figure 1. 3. and west. (Inside a building such as a theater. Alternatively . This frame establish observer " We might call this the orientation 1.

right to 3. that is. In the geographic frame. chapter 3. let us take a moment to look at how frames of reference are used in giving route directions (Levelt. Geographic frame: From I . straight/ forward to 4. are entirely free of direct or indirect environmental influence.Ray Jackendoff 's the front of the object is the side facing the same way as the observer " front . The route from circle I to circle 5 can be describedin two different ways: " " (5) a. right to 3.4d) are derived from gravitational axesin an imagined normal orientation of the object. 2 3 . though. Similarly . right to 4. this volume. at least ) has its own unambiguousvocabulary. chapter 12 . the traveler maintains a constant orientation .4e) are derived from a ' hypothetical observers position in the canonical encounter." 4 1 5 . In the precedingexamples alone. go up/ forward to 2. One of the reasons the axial vocabulary has attracted so much attention in the literature is its multiple ambiguity among frames of reference .. " b. as in figure 1. Only the geographical frame (in English. for instance . The problem is highlighted by the step from 3 to 4. thi ~ volume).- r 1 Figure1. axes in the canonical encounter frame (figure 1. go up/ forward to 2. down to 5. " Observer frame: From I .6. ' so that up always means up on the page .5d.. three different usesof front appear. the geometric and motion frames. So in fact only two of the eight frames.o. We might call this the orientation -preservingobserver frame. Before going on.6 ' Oneof Levelt s " maps . Why should this be? And what does it tell us about the distribution of information betweenCS and SR? This will be the subject of the next section. The proper way to think of this seems to be to keep track of hypothetical traveler' s " " orientation . right to 5. Consider a simple case of Levelt' s diagrams such as figure 1.. Tversky. which is describedas " right " in " " (5a) and straight in ( 5b) . the traveler s axes are set contextually by the page (frame B3) ." It should be further noted that axesin the canonical orientation frame (figure 1.

. For example . and asked to mark on them their length. was not the only influence on the choice of axes Details in the shapeof the Narasimhan figure also exerted an influence. In Narasimhan s experiment. one pictures oneself traveling through the diagram. And of course geographicalaxes(frame 8 I ) are available as well if the cardinal directions are known. or some combination of the three. straight or forward . while height -based contextual blases the subject toward environmental axes (gravitational or page ) . ' . The linguistic input . " of Axial Vocabulary 1. they did not insist by any meanson orthogonal axes . as in Tversky s examples ' . width. they cannot simply call up intrinsic axesstored in long-term memory as part of the canonical representation of a familiar object.. Intuitively .8b has a flattish surface near the (contextual) bottom . and height depend ' . Some subjects (8% ) apparently interpreted this surfaceas a basethat had beenrotated from its canonical orientation . In But of course linguistic information is also involved in the subjects responses the choice of to mark influences that the is asked the dimension .9 LexicalEncoding Narasimhan ( 1993 ) reports an experiment that has revealing implications for the semantics " of the axial vocabulary. subject particular . Bierwisch and Lang 1989 . a third possibility . often ' only in rectilinear orientations. subjects tended to mark its length as an oblique. and its height as an environmental vertical. width . the subjects have to compute axesof novel shapeson-line. If they are replaced by ' landmarks that do have intrinsic axes . " " " " and the samedirection from 3 to 4 is " .8a. revealedsubjects on choice of axes . which have no intrinsic orientation . confronted with a shapesuch as figure 1. Subjectswere shown irregular shapes( Narasimhan " of the sort in figure 1. whose longest dimension is oblique to the contextual vertical. using the motion frame (A2 ) .7. " " From this the solution follows immediately: forward is determined by the ob' s last move that is server . however.The Architecture of the Linguistic -Spatial Interface " frame where the direction from 2 to 3 is observer The puzzling case is the ' ~ . they drew the height of the figure as an axis orthogonal to this base . Sometimessubjects even marked these ! axeson the very samefigure.g. The circles. that of setting the traveler s axescontextually by the landmarks (frame 83 emerges again) . basedon visual input . that . Levelt 1984 ) has dealt only with rectilinear figures or familiar objects. Becauselength. Previous experimental This experiment is unusual in its use of irregular shapes research on axial vocabulary with which I am familiar (e. play no role in determining the frame. right as Levelt and Tversky point out . figure 1. . figures ) height. responses judgments about axis placement . as might be expectedfrom the work of Bierwisch and Lang ( 1989 ) Length the subject in favor of intrinsic geometric axes (longest dimension). blases axis. Thus.

Maximum T Observer ' s line of sight .Ray Jackendoff No base Flat base Tilted base Up -down axis Maximum (vertical ) Up -down axis Vertical .

[ ::f: secondary ]. the lexical items influence SR indirectly. The fact that the subjectsactually draw in axesshowsthat the computation of axes must involve SR. width.SR interface would then map features [ ::f: observer ] . as shown in the figure. The CS hypothesis .The Architecture of the Linguistic -Spatial Interface is. For . I propose that the SR hypothesis is closer to correct.e. and height specify the axes and frames of referencethey can pick out? There are two possibilities: I . We see . as a " canonical vertical. we know that lexical items may contain elements of SR such as the shapeof a dog. width. then. when subjectsjudge the axesof Narasimhan figures. The SR hypothesis . the approach advocated by Bierwisch and Lang. This would allow the axesand reference frames to be unspecified(or largely so) in the CS of thesewords. that people use SR to pick out axesand frames of referencein novel figures. Hence it is possiblethat the lexical entries of length. the hypothesis of Representational subjects responses does not allow us to just leave it at that. ' the frames of reference could be specified by CS features such as [ ::f: contextual] . and height also contain SR components that specify axesand frames of reference directly in the geometric format of SR. when subjectsjudge the axes of Narasimhan figures.) 2. According to this story. Alternatively.. the shapepresentedthree different choicesfor its axis system . According to this . people freely switch frames of referencein visuomotor tasks. ' response How does the linguistic input get to SR so that it can influence the subjects ? That is. The angle and positioning of a drawn axis is continuously variable. in the latter . However. The first argument comes ' from the criterion of economy. we normally adopt an egocentric (or observer example ) frame for reaching but an environmental frame for navigating. I believe . that linguistic and visual input interact intimately in determining ' in this experiment. [ ::f: vertical]. We must also ask at what level Modularity of representation (i. at what levels of representation do the words length. and Narasimhan s experiment ) demonstrates confirms. The obvious choicesare CS and SR. via these general interpretations of the dimensional features of CS. in a way expected in the geometric SR but not expected in the algebraic feature complexes of CS. (This is. General correspondences into the geometry of SR. and height by features in CS such as [ ::f: maximal] . in which module) this interaction takes place. In addition . width. in the CS. the SR of the lexical hypothesis items interacts directly with SR from visual input . As a result of this extra possibility. Marr ( 1982 ." Nothing in the linguistic input created this new possibility : it had to be computed on-line from the visual input . The axes could be specified in the lexical entries of length. we seeourselvesmoving through a .

But that is about as far as it goes which axis a word refers to . If axesand frames of reference to encodethem in CS. ranks. numbers. A cognitive system with more than one dimension etc. two miles wide (with dimensional adjectives ). it is necessary seem to be few grammatical there in this domain. Talmy 1978 . the axial vocabulary will contain this feature. But can be shown to have grammatical effects . This means from systemfor thesedistinctions at best duplicates information in SR it cannot take the place of information in SR. unlike the count-mass system . is not) . not an environment rushing past. It is well known that analoguesof spatial axes occur in other semantic fields. adjectives such pragmatically plausible casesas * eighty degreeshot and * twelvepounds heavy are ungrammatical. and that axial vocabulary generalizes to these domains (Gruber 1965 . consider the use of axis systems Turning to the criterion of nonspatial extension and frames of referencein nonspatial domains. Other than dimensional . of higher beauty. as in three incheslong. weights. Similarly . sevenblocks up the street (with axial prepositions) . many prepositions do not occur with measurephrases (* ten inchesnear the box). axes and frames of referencecannot be eliminated that a CS feature SR. and again the axial vocabulary is not employed. The only thing specialabout the syntax English axial vocabulary is that dimensional adjectivesand axial prepositions can be precededby measurephrases . 19Thus the criterion of grammatical effect dictates at most that CS needsonly a feature that distinguishes axesof objects from other sorts of object parts. but languagedoes not express of axial in sort differences color using any vocabulary. Thus whether a word pertains to an axis does seemto make a grammatical difference . Theseare SR functions. In English. the only English adjective that can occur with a measurephrase is old. ) . andfour feet behind the wall. Distinguishing axes from each on grammatical other and frames of referencefrom each other appearsunnecessary . Consequently . and comparative adjectives .Ray Jackendoff 17 stationary environment. and those that do are for the most part axial (though away. lower temperature (higher number Is there a referenceframe? One' s first impulse is to say that the referenceframe is gravitational . of the effects . at least in English.perhaps becausewe speak of the temperature rising and falling and of rising in the ranks of the army. Next consider the criterion of grammatical effect. and becauserise and fall in the spatial domain . 18 as in a mile awayfrom the house . the axis is almost always up/ down . . temperatures .) . more less ) ( / beautiful/salty/exciting/ is the familiar three-dimensional color space . etc. lower rank. Jackendoff 1976 I of are Lakoff 1987 only one-dimensional. But all other axis systems know for example . much lesswhich frame of referencethe axis is computed in . Langacker 1986 . my mood is up. not CS functions. . Kinship systemsmight be another multidimensional case . when a nonspatial axis is invoked. grounds . No grammatical effectsseemto depend on .

A notion of frame of referencealso appears in social cognition . plus some special machinery for time and perhaps for social point of view. For instance . but in fact we find it very difficult . impoverished In short. Perhapsmultiple axes and frames of referenceare available in CS. All we need for most purposesis the distinction betweenthe vertical and other axes . this notion is quite limited : it is analogous to adopting an observer referenceframe for a different (real or hypothetical) observer . If one insists on a " functional " view. Or perhapsthe nature of the real world does not lend itself to such thinking outside of the spatial domain. say the relation of income to educational level to age . a one-dimensional systemthat goesfront to back. Moreover. But on secondthought. where beforemeans prior to . Our tentative conclusion is that most of this detail is encoded only in the SR component of the axial vocabulary. Hence the criterion of nonspatial extension also gives us scant reasonto encodein CS all the spatial distinctions among three-dimensional axesand frames of reference . but we do not recruit them for nonspatial conceptsbecausewe have no need for them in our nonspatial thought . when it so clearly . where " " before means subsequentto . for ? example About the only exception to the use of the vertical axis in nonspatial domains is time. not in the CS component. But one also speaksof the hard times before us.2O Time is also exceptional in that it doesdisplay referenceframe distinctions. so such conceptscannot be usedsensibly . one speaksof the times " " " " beforenow. why do we inevitably reduce social status to a linear ranking.-Spatial TheArchitecture of theLinguistic Interface pertain most specifically to the gravitational frame. we really wouldn ' t know how to distinguish among reference frames in these spaces ." A skeptic committed to the CS hypothesis might raise a " functional " argument against this conclusion. But compared to spatial frames of reference . as though the observer is facing the future. Thus again an apparent parallel proves to be relatively . there is no parallel to any of the other seven varieties of reference frames. I would urge quite a different argument. it thus parallels such lexical SR componentsas the shapeof a dog. in the social domain there is no notion of axis that is built from these frames of reference . What would it mean to distinguish an intrinsic upward from a gravitational upward. Certainly nothing outside the spatial domain calls for the richnessof detail neededfor the spatial axial vocabulary. For a more ecologically plausible case . Let me call this the " Mostly SR hypothesis . It would often be extremely useful for us to be able to think in terms of detailed variation of two or three nonspatial variables. as though the observer(or the front of an event) is facing the past. very little of the organization of spatial axes and frames of referenceis recruited for nonspatial concepts . where we speak of adopting another' s point of view in evaluating their knowledge or attitudes.

so that we can multidimensional spatial intuitions to the variation in question. that the proper way to encodethe relevant distinctions is in terms of a set of discrete(or digital ) annotations to the geometry of SR. This is more or lesswhat would be predicted by the Mostly SR hypothesis That is. and SR. it must be . We have examinedcases where the choice betweenCS and SR comesout in encoded different ways. This shows that the issueis not a simple prejudged matter . But in the absenceof a serioustheory of SR. the crucial question has proved to be at which level or levels of representationit is to be .Ray Jackendoff involves many interacting factors? The best way we have of thinking multidimensionally is to translate the variablesin question into a Cartesian graph. and spatial to take the issue of how theselevels communicate with and seriously cognition . It might turn out . encodeit . we have to turn to SR to help us . know how to continue this line of research 1. one level further removed from syntactic structure. it is hard to . This makes it difficult to decide among (or even to formulate) competing hypothesesin any more than sketchy fashion. other than the small bit of work by Marr and Biederman even in gestation. Compared to the richnessof phonological and syntactic theory. it is useful to think in terms of Representational Modularity . one another.one would certainly want to check the data out we cross linguistically before making a stronger claim. evaluated for each case For the moment. for instance . abstract conceptual thought. it would be hard to distinguish an SR encoding of thesedistinctions from a CS encoding.10 FinalThoughts To sort out empirical issuesin the relation of languageto spatial cognition . This conclusion is tentative in part becauseof the small amount of linguistic evidence adduced for it thus far . This time we conclude that most of the relevant distinctions are not encodedin CS. .we can our apply . we are at the mercy of the limitations of theory. The caseof axesand frames of referencethus comesout differently from the case of the count-massdistinction . however. the " functional " argument can be turned around and used as evidencefor the Mostly SR hypothesis . But it is also tentative because do not have enough formal theory of SR to know how it encodesaxesand frames of reference . It is hoped that the present volume will spur theorists to remedy the situation. but only in SR. is hardly infancy. the theory of CS is in its . This suggests that CS is actually relatively poor seeit as a path or a region in space in its ability to encodemultidimensional variation . This forces us to distinguish the levels of representationinvolved in language . In looking at any particular phenomenon within this framework . In such a case .

or else(2) interfacesare simply an integrated part representations of larger modules and need not themselvesbe modular. this volume) . Lynn Nadel. came from participants in surrounding the ideasin this chapter. The first pass of syntactic parsing has to be part of the mapping from phonological to syntactic structure. Further important suggestions the Conferenceon Spaceand Language sponsoredby the Cognitive Anthropology Research Group at the Max Planck Institute for Psycholinguisticsin Nijmegen in December1993and of course from the participants in the Arizona workshop responsiblefor the presentvolume. I do not find any support for theseinterpretations specifiedprocessaccomplish in the text. becauseinterface modules as conceived here are too small to be Fodorian modules (they are not input -output faculties). where the relation betweenlinear order in phonology and syntax is unclear. and Emile van der Zee for extensivediscussion . for more detailed discussion . Various colleagueshave offered interpretations of Fodor in which some further vaguely es the conversion. This is an oversimplification. so that word boundaries can be imposed. Fodorian modularity can also solve the problem of communication among modules by adopting the idea of interface modules.13849 to Brandeis University. Of course . Fodor ' s claims about informational encapsulation are largely built around evidence that semantic es of lexical retrieval /pragmatic information does not immediately affect the process and syntactic parsing in speechperception.~ tic-Spatial The Architecture of the I . 5. becauseof the existenceof languagesthat make use of the visual/gestural modalities. 3. express 6.ingul Interface Acknowledgments I am grateful to Barbara Landau. by a Keck Foundation grant to the Brandeis University Center for . See Jackendoff 1987 . Caveatsare necessary concerning nonconcatenativemorphology such as reduplication and Semitic inflection . along lines proposed here. This evidenceis also consistent with Representational Modularity . but the phonology itself has no knowledge of what syntactic features theseaffixes . Fodor s discussionshows that this first pass usesno semantic information . syntactic featuresare frequently realized phonologically as affixeswith segmental content. in person and in correspondence . Bhuvana Narasimhan. 2. . SeeEmmorey (chapter 5. Manfred Bierwisch. chapters 6 and 12. This researchwas supported in part by National ScienceFoundation grant IRI -92. and by a fellowship to the author from the John Simon Guggenheim Complex Systems Foundation. To be sure. there are two possibilities: either ( I ) the scaleof modularity has to be reducedfrom faculties to . but also in part an empirical one. Paul Bloom . I take the choice betweenthese two . . However. so that candidate semantic interpretations can subsequentlybe formulated and tested . this first pass uses no semantic information either. to say the least. possibilities to reflect in part a merely rhetorical difference 4. The first pass of lexical retrieval has to be part of the mapping from ' auditory signal to phonological structure. Notes I .

A different sort of example . II . placesand paths as independententities may be a higher-level cognitive (nonperceptual ) aspect of spatial understanding. 13. Paul Bloom has asked ( personalcommunication) why I would considerforce but not . arguing that objects are always seenfrom some point of view or other. An imagistic representation must representthe road being specifically on one side or the other. concerned 12. interpret object-centered as meaning that the encoding of the object is independent of point of view. + bounded. it is not so obvious that places and paths are encoded in imagistic representation becausewe do not literally see them except when dotted lines are drawn in cartoons. As a corollary . That is. whoseperspective with respectto the viewer changesduring rotation . SR must support the generation of mentally rotated objects.a fact noted by Kosslyn ( 1980 ).inherent structure] distinguishes objects (or accomplishments ) from process from groups of individuals . Jackendoff 1991decomposes the notion of object into the more primitive feature complex [material. It is unclear to me at the moment what relationship this notion of image schemabears to that of Mandler ( 1992and chapter 9. I leaveopen for future researchthe question of whether the presentconception can help sharpen the issueswith which Mandler is . and also closedevents es. 14. as also argued by Talmy (chapter 6. On the other hand. ' " 9. this volume). viewed from any arbitrary vantage point . although there is certainly a family resemblance . must include the possibility of the road being on either side of the river. Marr . Although fundamental.Ray Jackendoff 7. It is as though syntactic structure is a way of converting embedding structure into linear order. such a type is not necessarilyprimitive . does not deal with this further step of reinjecting the object into the scene . . offered by Christopher Habel at the Nijmegen spaceconference " " (seeacknowledgments ) : the image schema for along. It is surely significant that syntax sharesembeddingwith CS and linear order with phonol ogy. ( ). . but also substances from aggregates and homogeneousprocess es from repeatedevents . chapter 10. Mandler ' s formulation derivesfrom work such as that of Lakoff ( 1987 ) and Langacker ( 1986 ). who is not concerned with spatial layout but only with identifying the object. The feature [material] is shared by substances and aggregrates . this volume) .at the very least the ' s. The feature [ + bounded] distinguishes objects from substances . However I " " observer . and in which no explicit connection is made to researchin visual perception. This section is derived in part from the discussionin Jackendoff 1987 . This neutrality permits the appearanceof the object to be computed as necessaryto fit the object into the visual sceneas a whole. it distin them all from situations events and states times and various sorts of abstract . spaces guishes entities. as in the road is along the river. 15. This is particularly crucial in rotation on an axis parallel to the picture plane becausedifferent parts of the object are visible at different times during rotation . 10.inherent structure] . in which the notion of level of representation is not well developed . say " " anger to be encoded in SR becausewe have the impression of directly perceiving anger as . so that structured meaningscan be expressed as a linear speechstream. 8. This may be another part of SR that is invisible to imagistic representation . But I seesuch a step as altogether within the spirit of his approach. The feature [ . Somecolleagues have objectedto Marr s characterizingthe 3-D sketchas " object-centered .

observerand the referenceobject. Cognitive Anthropology ). 3. In a " personmorphic" frame for right and is the visual field divided into two halves . . this frame requires the genitive casefor the referenceobject. however. for instance ten poundsheavier ( than X ) . back. cognitive consequences 18. with the division line running through the left . problematic caseslike scissorsand trousers 17. would not yield any generalizationsin terms of sharedcomponents 16. 94(2). B. which goes from right to left. ( 1987 ).down axis is used for the the conception of time as an opposition of past and future. Some semanticuniversalsof German adjectivals of .namely. This leavesopen the possibility of CS. Here they are licensednot by the adjective itself. ( I 994b . References Bickel in spatialdeixisand the typologyof reference . . the of distinguishing referenceframes. this frame requires the ablative casefor the referenceobject. present enterprise how this distinction is encoded so that further inferencescan be drawn from it .direction of force well. Review . however. and right. therefore .fifty timesasfunny ( as X ) . 20. For a recent discussion of the psychophysics and neuropsychology of the distinction . A number of people have pointed another nonvertical axis system .36 Language . Max PlanckInstitute for Working paperno. and an up. between environmental motion and self-motion . The difference is that physical force has clear geometric components to encodeother spatial and often contact betweenobjects. frames . CognitiveAnthropologyResearch Group. the Nepalese languageBelhare is a counterexampleto the generalization about time going front to back: a transverseaxis is used for measuring time. six timesmore beautiful ( than X ) . . B. the political spectrum. According to the description of Bickel 1994b . left .147 . Mappingoperations . . I. and culture . 115 Psychological Bierwisch . as. M. Bickel 1994a . I leave it for future researchto ascertain how widespreadsuch grammatical distinctions are and to what extent they might require a weakeningof my hypothesis . Nijmegen Group. points out that the NepaleselanguageBelhare makes distinctions of grammatical casebasedon frame of reference . Foundations . Unpublishedmanuscript Research . Thus force seems a natural extensionof the family of spatial concepts . Max PlanckInstitutefor Psycholinguistics Biederman : A theoryof humanimageunderstanding . . anger has no such geometrical characteristics . threefeet shorter ( than X ) . ( 1967 ). . its parameters belong to the domain of emotions and interpersonal relations. I leavethe issueopen. cognition ). Nijmegen Psycholinguistics Bickel : Where to orient . By contrast.syntax discrepanciesin the more grammatically . ( 1994a ). 31 . does not appear to addressthe issue crucial the of . 19. but by the comparative morpheme.which are independentlynecessary entities suchas trajectories and orientations. In a " physiomorphic" frame for right and left. 1.~ tic-Spatial The Architecture of the IJmgul Interface . .components . Extending SR to anger. Spatial operationson deixis oneselfin Belhare (revisedversion . see Wertheim 1994 and its commentaries to Wertheim. the referenceobject projects four quadrants whosecentersare focal front . Measure phrasesalso occur in English adjective phrasesas specifiersof the comparatives moref-er than and as . . Recognition by.

. ( 1993 ) . and dangerous ) . Amsterdam: Elsevier / North-Holland . Word meaningand Montague grammar. Cognition. Parts and boundaries. Amsterdam: North Holland .236. ( 1987 . PhiD . Chicago: University of Chicago Press Lakoff . ( 1975 . ). G . Lingua. Consciousness Jackendoff. Cambridge. cognition . ( 1971 ) Santa Cruz lectureson deixis. ( 1983 . Fillmore . MA : MIT Press . OM -sentences .329. ) . 765. D . M . ( 1983 . Cambridge. G. ( 1980 Lakoff . R. E. 20. On the nature of semantic fonn in natural language H. MA : MIT Press Jackendoff. 17. Foundationsof Language ) . and Calvanio. Semanticsand cognition. M . Women things. R. ( 1991 ). Bierwisch. . A compositional semanticsfor Aktionsarten and NP referencein English. and Jackendoff. R. MA : MIT Press . Languages of the mind. and Johnson .. 16. M . Jackendoff. Linguistic Inquiry. R. (Eds. ). 199 unspecifiableinterpretations. ( 1987 and the computationalmind. .fire ..201. Levine. Klix and ) . Behavioraland Brain Sciences . Massachusetts . Bloomington : Indiana University Linguistics Club. Hammond .. S. Jackendoff. ( 1985 ) . R. Cambridge. R.. Chicago: University of ChicagoPress " " " " Landau. diss. M . What and where in spatial languageand spatial . Image and mind. 439. Cambridge. 9 45. 92. Toward an explanatory semanticrepresentation 89. Human memoryand cognitivecapabilities : Mechanisms andperformances . 1976 . Press Jeanne rod .462. J. Cambridge. MA : MIT . In F. Reprinted in Gruber .D . ( 1965 Institute of Technology ). B. ( 1979 Farah. Ohio State University . Metaphorswelive by. Cognitive Psychology imagery: Dissociable systemsof representation .. Jackendoff. The architecture of the language faculty . 41.. ( 1986 . Possiblenames nominals. MA : MIT Press Jackendoff. E. and Lang. ) . MA : Harvard University Press Fodor . Studiesin lexical relations. MA : MIT Press Gruber . MA : Harvard University Press Kosslyn. Culicover.238. R. P. ( 1994 : The role of syntax-semanticsmappings in the acquisition of ) . ( 1994 ) . J. (forthcoming). Ph. ) The languageof thought. diss. ( 1990 .) ( 1989 . ( 1988 ) . ) Modularity of mind. ( 1980 ). 7. Fodor .150 .. 297. J.). Cambridge. Hagendorf (Eds. ( 1992 .. 217. Dowty . Behavioral and Brain Sciences . K .T RayJackendot Bierwisch. The representing brain: Neural correlates of motor intention and . Dordrecht: Reidel. C. ) . Berlin: Springer. Ray ( 1976 . Lexical structures in syntax and semantics . P. Hinrichs . Visual and spatial mental . Semanticstructures Jackendoff. ) . . Cambridge. 187 imagery . Dimensionaladjectives Bloom. 8. R. M . ) . ( 1972 : On the derivation of sentenceswith systematically ) .784. D . Cambridge.

Frames . Hinsdale Olson . van de Grind. ( 1988 .. Knowledge and Language . andcontrasts . In E. ( 1993 properties ). In A. vol.. UniversityPress . Cognition .. In K. Reprint . and Winston . ( 1978 .604 . limitations in talking about space Levelt . Rinehart Paivio . ( 1989 ). Cognitive . Stanford . How to build a baby . J. Learnability . and Rapoport ).). and height . Family resemblances . Someperceptual . Conceptual Mandler . van Doom. Psychological Review : 2. ( 1975 . 1. Spatialframesof reference . Limits in perception W. The syntaxof eventstructure Pustejovsky . B. Spatialcognition . SanFrancisco Marr. In Papers Levin . C. T. ( 1995 ). . . width . : A synopsis . ( 1975 ). Books . G. ( 1978 ). : Freeman . R. Cambridge lexicon . D.605 . Machinery .). NewYork: Association processing . 587 . 41. . W. primitives . Foundations Langacker . The relation of grammarto cognition Talrny for Computing in naturallanguage issues Theoretical . 573 Psychology categories . Vision Universitaires . Vol. Minneapolis : Universityof Minnesota . UniversityPress .81. Oxford: Oxford as a cognitive O' Keefe map . P.The Architecture of the Linguistic .). J. Imageryand verbal .fourth from the twenty . es . Lexicaland conceptual . . Themeaning . . of Linguistics Department . S. 7. CA: Stanford . and (Ed. Koenderink (Eds . New York: Holt. and J. NJ: Erlbaum . 99. 2. Louvain : Publications dela causalite Michotte . Utrecht : Coronet . J. . Semantic .fields. Gunderson of meaning Putnam . Dordrecht : structure . Language . . J.. ( 1984 ). andJohnson . D. Cambridge structure : Theacquisition andcognition Pinker . in the useof length Narasimhan . UniversityPress . Hinsdale . Erlbaum 1979 Hinsdale NJ: . Abraham andsemantic structures Partee . Thegenerative Pustejovsky " " .). . and Kittay. Language . : 275 289 the . ( 1983 . and Nadel ). ( 1976 . A. MA: Harvard ). Press . MA: MIT Press . (Eds ).. Universityof Chicago Chicago Linguistics Society Chicago regional meeting of . 2. ( 1982 ). and Bialystok ). B. Cambridge andperception Miner. In D. 2d ed. La perception . ( 1954 ). H. L. B.30 (Eds Kluwer. ( 1993 ). . MA: MIT Press . . A.Spatial Interface . Reuland andW. The hippo campus . BostonUniversity manuscript Unpublished . ( 1991 ). L. A. of argument . Waltz (Ed. E. E. de Louvain -Laird. 7. Lexicalsubordination ..) ( 1992 Lehrer . NJ: Erlbaum .193 . and Mervis ). Vol. ( 1992 ). 47. mind . 131 knowledge in the internal structureof : Studies Rosch . E. ( 1971 process ). ( 1986 grammar of cognitive ).

Shopen (Ed. A theoryof aspectuality UniversityPress Wertheim . On thecompositional . Lexicalization patterns: Semantic structure in lexical forms. (Eds of visual natureof theaspects . ( 1993 Verkuyl ). andMishkin. Motion perceptionduring selfmotion: The direct versusinferential . ( 1985 Talmy ). M.). . ( 1982 Ungerleider ) Two corticalvisualsystems . Also in Cognitive Science . In D. A. . L . NewYork: Plenum Press . Ingle .100 ' s work on German Teller and extension of Manfred Bierwisch . H. structures . Cambridge MA: MIT Press . Behavioral revisited andBrainSciences . research . 49. In H. ( 1994 ). Acredolo(Eds . andapplication in language and thought . ( 1980 ) . Goodale . Dordrecht : Reidel . Chicago : Universityof Chicago . Department Regional Meetingof theChicago LinguisticSociety of Linguistics . Foundations . ( 1983 ). Languagetypology and syntactic description . Cambridge : Cambridge . . 185 217 adjectivals of Language . 293 controversy .' Ray Jackendoft Talmy. L. In Papers . Pick and L. L. and R. New York : Cambridge University Press . ( 1969 ). . Spatial . How language Talmy space orientation : Theory . 5.). Mansfield .). Forcedynamics from the Twenty -first . 3. Analysis behavior . M. P..311 . In T . L. H. 17 . . ( 1972 Verkuyl ). vol. . 12( 1988 ). Somediscussion .

character of natural language same reason . and all the rest.unlike pictures. temperature.Chapter How 2 Space Gets into Language ? Much Manfred Bierwisch 2. The contrast obviously dependson the meaning of enter versusadmire. and the want. furthermore . as in ( la ) and ( lb ). The conclusion that syntactic elementsand relations do not accommodatespatial information seemsto be confronted with certain objections. Comparing ( la ) with (2). we observethat completely identical structurescan express : both spatial and clearly nonspatial situations. pursuing the following question: Which components of natural languageaccommodatespatial information . blueprints. even though linguistic utterances do not exhibit color . rather than iconic. that identical or at least very similar spatial eventscan be expressed by meansof rather different syntactic constructions: ' (2) We went into Saint Peter s Cathedral.1 Introduction We can talk about spatial aspectsof our environment with any degreeof precision we . though. kinship .do not exhibit spatial structure in any relevant way. The present chapter will be concerned with certain how space gets into language aspectsof this problem. This apparent paradox is . depending on its syntactic position : . b. respectively ' ( I ) a. and how? Looking first at syntax. We admired Saint Peter' s Cathedral. We entered Saint Peter s Cathedral. kinship. Thus the PP at the end has a temporal meaning in (3a) but a spatial one in (3b). raisesthe by no meanstrivial question where and The apparent paradox nevertheless . temperature. For the simply due to the symbolic. we notice. even though linguistic expressions like . we can talk about color . and so on. maps.

Similarly . . as is evident from pairs like those in (4) : (4) a. Consider (6) as a casein point : (6) He entered the church. that there is a simple and clear distinction between spatial and nonspatial vocabulary. ( bridge ) directional It is. in particular the field of lexical semantics . This appears to be a matter of the syntactic . I will take up this issue in section 2. Er schwammunter DernSteg (He swam under the bridge.7. As a matter of fact. not the contrast betweenIml and Inl . most words that undoubtedly have a spatial interpretation may alternatively carry a nonspatial reading under certain conditions. At the end b. shesigned (3) a. respect ( ). the crucial distinction can be reducedto a component. Er schwammunter den Steg . determining different interpretations for the PP in (3) . this does not hold for the so-called notional or content cases . each associatedwith a specificcaserequirement (seeBierwisch 1988fordiscussion ) in languageswith rich morphology . but rather that betweendative and accusative that is relevant here. She signedthe letter at the end. differing with respectto place and goal: .Manfred Bierwisch the letter. Shesignedthe letter with this intention. In the presentcase systematicdifferencebetweena locative and a directional reading of the preposition unter. however. the contrast betweenplace and goal in (5) is ultimately a matter of the two different readingsof unter. including all major lexical categories . With this intention . assignthe contrast betweenspatial and nonspatial interpretation to the position as such. however. of course.) location b. Another problem shows up in caseslike ( 5). . Hencethe main area to be explored with respectto our central question is thesemantic component. Further illustrations could be multiplied at will . (5) a. Pending further clarification . This does not mean. As already mentioned with to I it is the word meaning of enter that carries the spatial aspect . Whereascasecan thus be shown to be related to spaceonly as an indirect effect. What we observein (3) and (4) is rather the effect the different syntactic structure has on the compositional semanticsof adjuncts (the details of which are still not really understood). One cannot. she signedthe letter. In any case . b. we will nevertheless conclude that phrasestructure does not reflect spatial information per se. however. He swam under the . syntax and morphology as such do not reflect spatial information .

. we must not only account for the spatial information that enter projects in cases like ( Ia ) and one reading of (6). (6) can also have an interpretation under which it means he becamea priest. one might consider the (speciesspecific ) languagecapacity as emerging from brain structures that allow for the discrete . With this overall orientation by the syntactic and morphological operations of I -language in mind . but also for the switch to the nonspatial interpretation in the second reading of (6) .compositionality. that is. The verb to enter thus has a spatial or nonspatial interpretation depending on the reading of the object it combines " " with . that constitute the interfaceswith the respectiveextralinguistic domains. Given PF and LF as interface levels . the internal or I -languagein the senseof ). that is. bypassing the . except where it will be useful to compare how it relates to A -P with the far more complex conceptual phenomenathat concern us. This is an instanceof what Pustejovsky( 1991 ) calls co. Following the terminology of Chomsky ( 1993 ). a compositional structure where one constituent determinesan interpretation of the other that is not fixed outside the combinatorial process . we have to investigatehow lexical items relate to space and eventually project theserelations by meansof compositional principles. where church refers to an institution and enter denotesa changein social relations. which " " " " Chomsky calls phonetic fonn (PF) and logical form (LF ). detennined by I -languageand interpreted in terms of APand C-I .I -language.2 LexicalSemantics andConceptual Structure Let me begin by placing lexical and compositional semanticsin the more general perspectiveof linguistic knowledge.. Because there is apparently no direct relation that connects spatial infonnation to sound structure. In other words. we arrive at the . and C-I . Assuming universal grammar (UG ) to be the formal characterization of this capacity. 2.+ C-I A -P comprises the systemsof articulation and perception. correspondencebetweenthem is established . I -languageis to linguistic expression be construed as a computational systemthat detenninesa systematiccorrespondence betweentwo different domains of mental organization: (7) A -P +. the systemsby which experienceis conceptually organized and intentionally related to the external and internal environment.? How Much SpaceGets into Language Besidesthe spatial interpretation corresponding to that of ( Ia ). recursive mapping between two representational systemsof different structure and origin . respectively the . I will have correspondenceestablishedby the computational system of I -language nothing to say about PF. I -language provides two representational systems . To conclude these preliminary considerations . which underlies the properties of external or E-languageof setsof Chomsky ( 1986 . in order to answer our central question.

and 3.Manfred Bierwisch schema from theconditions . two general remarks about UG and the organization of I -languagemust be made. This view would be mandatory if in fact UG were not restricted to acoustic signalsbut allowed also for systemslike sign language . First .! As to specification I . Primes that are recruited from and interpreted by C-I .SYNTAX . the notion of conditions or constraints to construct scopeof the present discussion primes of I -language seemsto be indispensableif we addresstype 2. and 2. for each of the major components of I -language . emerging language capacity the types of representation and the operations available for I -languageare given in advance . As for type 3. 2.I y I -language ~ va This schemais meant as a rough orientation .+.. A general format of the type of representationsand operations of the component. A way to recruit the primitive elementsby which representationsand operations of the component are specified . Primes that are recruited from and interpreted by A -P.lPF + . Alternatively . and if semanticrepresentationsare to go beyond a highly restricted core of lexical items. whereI -language followinggeneral emerges specified by UG throughthe interactionwith the systems of APand C-I: ( 8) A .+. Although the details of this issuego beyond the . one prosodic categories might think of them as being recruited from the auditory input and articulatory patterns by means of certain constraints within UG . varying only to . are basedon universally fixed options in UG . the primes in terms of which I -languageinterfaceswith C-I . from the conditions of the as such. Primes that function within I -languagewithout external interpretation . I will return to theseissuesbelow. phonetic features and . which must comprise the featuresspecifying syntactic and morphologi cal categories . three types of primes are to be distinguished: I . leaving crucial questionsto be clarified.. Before I turn to details of the relation between I -language and C-I . In other words . The most parsimonious assumption is that specification 2 is fixed across languages . the primes of PF. universal grammar (UG ) must provide two specifications : I . which provides not the repertoire of these features but rather some sort of recipe to make them up . thesemust be determined directly by the conditions on syntactic and morphological representationsand operations falling under type 2.C . It is usually assumedthat type I . namely..+ LFJ+..P +.

and so forth .How Much SpaceGets into Language ? . 1986 ). I will adopt this view. The general format that VG determinesfor lexical items is (9) : (9) [PF (le). and GF (le) can representinformation of any complexity in accordancewith the two requirementsnoted above.) The basic elementsof LF are lexical items. or rather their semantic component. GF (le) representssyntactic and morphological properties of Ie. I will call the set of lexical items. rather. whatever internal . head movement. there is a remarkable lack of agreementamong otherwise closely related approaches. LF (le)]. According to the conceptualframework of Chomsky ( 1981 . only syntactic representation to which independent . 1993 . in number. syntax. C-I . or more LF technically. and the syntactic features associated with them. which must be learned on the basis of individual experience . it cuts acrossall of them. or other process discussed .2. With regard to the crucial question how C. LF (le) consistsof primes of LF specifiedby Ie. This the extent to which they can be affected by intrusion from the interface levels in fact be the case for which conditions syntactic might morphological categoriesby take up conceptual content. arguing that there are . the primes of LF .I relates to I -language . are to be identified with word meanings . varying from languageto . (9) also indicates the basic format of linguistic expressionsin general . with the part of lexical items. In any case structure should be assignedto the semanticsof lexical items is essentiallya matter of . together with the general conditions " LS of I . LS is not a to which they must conform . not structurally reflectedin I -language In contrast to this view. (In Chomsky 1993 is fact the level of . by Hale and Keyser ( 1993 ) . GF (le). following Katz ( 1972 ) and others. person. assignslexical items a rich and systematicinternal structure. the " lexical system ( ) language of I . where PF (le) determinesa representationof Ie at PF. In other words. for example . LF is a level of syntactic in whose status lies its the interface with conceptual representation particular forming LF in structure. for example . if we assumethat PF (le). Second . morphology . alongside phonology. I will have more to say about the organization of lexical entries at the end of section 2. systematic conditions apply . but which are language determined by VG with respectto their general format in accordancewith specifications 1 and 2. which is claimed to be linguistically relevant. the computation determined by I -languagedoes not in general proceedin terms of primitive elementsbut to a large extent in terms of chunks of them fixed in lexical items. LF (le). which according to type 2 above connect I -language to C-I . Lexical items are idiosyncratic configurations. Jackendoff ( 1983and subsequentwork ). combining information of all components of I -language . and separatecomponent language semantics . including complex items originating es of " sublexical syntax" as from incorporation .

In ( 12a ) leave is (most likely) interpreted as a physical movement and institute as while the time adverbial a year ago of ( 12b ) turns leave into a change in place.+. He left the institute a year ago. which among other things must fix the truth and satisfaction conditions. As this point is crucial with respect to our central question . and GF (/e) a specification of morphological and syntactic properties ( 11) A .lPF + .+ SF) +. What I want to show is . world knowledge. which is basedon the semanticform of lexical items exhibiting a systematic . -independentand must be assignedto C-I . b. respectively . Supposenow that we call the representationalsystembasedon semantic primes the " semantic form " SF of I . He left the institute an hour ago.. which is based on phonetic ( ) language ). Before I take up some controversial issuesthat are related to these assumptions . SF(/e) a configuration of SF. On the other hand. linguistically relevant internal structure. which are language in ( 12 To begin with the secondpoint ..language ~ va at LF according The systemof SYNTAX is now to include the information represented 2 to (8) .SYNTAX . affiliation and institute into a social institution .C . the interpretation of an expressione is detennined by its semanticform SF(e).+. GF (/e). assuming theseare the elementsidentified in type 2 that connect I -language to C-I . On the one hand.parallel to PF. I will clarify the problem by meansof someexamples twofold .I . and hence primes. the conceptual interpretation c of a linguistic expressione is determined by the semantic form of e and the conceptual knowledge underlying C. the conceptual interpretation of e. The basic idea behind the organization of knowledge suggestedin ( II ) is that I -language needsto be distinguished from the various mental systemsthat bear on . compare the sentences ): ( 12 ) a..P +. The two interpretations of leave the . depends in crucial respectson commonsensebeliefs. I call the basic elementsof this structure " semantic primes.structural phenomenadirectly involved in I -languagethat turn on the internal structure " of lexical items. We will consequentlyreplace schema(9) of lexical items by ( 10 the overall schema(8) by ( II ) : ( 10 ) [PF (/e). I will briefly illustrate their empirical motivation . More specifically A -P and C-I . SF(/e)] with PF (/e) a configuration of PF. and situational aspects ..I y I .

The most . depending on the conceptual content of the resulting state s. For a systematicexposition of this framework in general in BECOME of DO and the and for 1988 . of course world but with or otherwise. Bierwisch ( 1983 . This has . In ( 14 the proposition [BECOME [ NEG [x AT f ))] . a certain about knowledge first now to the point concerning the internal structure of SF(le). again due to encyclopedic along. see Dowty particular ). seeBierwisch being at y . for example ) and Dolling ( 1995 knowledge interact. ( 14 ) specifies ' the complex condition that x brings about a change of state that results in x s not . knowledge about changesof location or institutional affiliation and their temporal frames. and so forth and variables like x . In short. I will Turning illustrate the issue by looking more closely at leave . John lost his money by gambling at cards. More . y. or an institutional affiliation . that the choice between the locational and the striking point of ( 12 social interpretation is determined by the contrast betweenyear and hour. z. Notice incidentally. point . see ) is. AT . however. ]] [ [ ) Correspondingly ( or a change in social position . interpretation ( all the elements that noted at this be 1979 . . as in ( 12a be interpreted by a spatial movement BECOMEs can x DO 12b . .compositionality as already illustrated by sentence (6) above. ( 13 b. To begin with . John lost his money by speculatingat the stock market. and where For extensive discussion of these phenomena . . DO is a relation between an individual x and a proposition p with the specifically " " ). as in [x AT y] might be a spatial relation . where BECOME defines a transition into a state s characterizedby the condition that x be not at y . c. SF consists of functors and arguments that combine by functional Generally speaking . pis conceptual interpretation that could be paraphrasedby xperforms p . providing at the same time an . ( 14 outline of the format to be assumed for SF representations ) indicates the slightly simplified semanticform of leaveas it appearsin ( 12 ): ( 14 ) [x DO [BECOME [ NEG [x A Ty ]]]] .? into Language Gets HowMuchSpace institute are casesofco . while in ( 13c domain. that his moneyin ( 13a ) refers to the coins or notes John is carrying ) it is likely to refer to all his wealth. The basic elementsof SF in the sensementioned in type 2 above application DO BECOME like are constants . encyclopedic linguistic ) . ) ( . the physical or abstract interpretation of lose and moneyin ( 13 ) : adverbial in the different world on adjuncts knowledge coming through depends ) a. In a similar vein. whether linguistic nothing to do . Thus with differing conceptual interpretations highly abstract and hencecompatible ). with the meaning of theseitems as such. It should ) are showing up in ( 14 . John lost his money through a hole in his pocket.

to useearlier terminology) . by y by assignedto again: ( 16 ) a. while under the restitutive reading ( 15 ) states only that John brings about of his not being at the institute .Manfred Bierwisch But why should the lexical meaning of leavebe representedin the manner of ( 14 ). Two points are to be emphasized . [AGAIN [x DO [BECOME [ NEG [x A Ty ]]]]] b. ( 15 ) statesthat John leavesthe institute for (at least) the second time. conceptual conditions. if the conceptual interpretation must account for more specificdetails anyway? This brings us to the linguistic motivation of the internal structure stipulated for SF(Ie) . or more specifically . it could not be represented . the relationship between variables in SF(le) and the syntactic argument structure (or subcategorizatio . These two interpretations can be indicated by ( 16a . where x must be interpreted ). the variabley of ( 14 ) can be left without a in case it must be determined value which . furthermore. LEA VE would fail to x y] provide representation[ The optionality of the object of leaveon which ( 17 ) relies brings in . An even remotely adequateanswer to this question would go far beyond the scopeof this chapter. Second . First . interpreted by contextual syntactically conditions providing a kind of neutral origo . which servesas a kind of default. although for different reasons . The secondphenomenon to be mentioned concernsthe intransitive use of leaveas in ( 17 ): ) John left a year ago. the intimate relationship between SF(le) and GF (le). henceI can only indicate the type . rather than simply as [x LEAVE y]. which obtained before. Consider first ( 15 of motivation that bearson ( 14 ). ) by pointing out two phenomena which is ambiguous betweena repetitive and a restitutive reading: ) John left the institute again. if leave extralinguistic. as one ) would be a more component into the syntactic information GF (le). ( 17 Two observationsare relevant here. [x DO [BECOME [AGAIN [ NEG [x AT y]]]]] For discussionof intricate details left out here. ( 15 Under the repetitive reading. rather than . such that ( 18 : complete lexical entry for leave . Once again. the ambiguity of ( 15 ) carries over to both the physical and the institutional interpretation . it is determinedby linguistic. the global the relevant structure. the state [x AT y] under this condition is almost automatically restricted to the locative interpretation . optionally or obligato rily interpreted by syntactic constituents. First . seevon Stechow ( 1995 ) . Supposewe include a specification of the SF variable. however. Second were to be characterizedby the unanalyzedlexical meaning [x LEA VE y] . respectively ) and ( 16b AGAIN is a shorthand for the SF to be and the institute and where John .

The problem is for the relation of languageand space . / " " + + + + + A CS PS SS . syntactic structure (SS . and chapter 1.3 The architecture sketchedin ( 11 ): ) is thus to be replacedby something like ( 19 ) Audition ( 19 Articulation Vision . CS is held to be external to I -language representation must obviously be identified with C-I (or perhaps a designated level of C-I ) . for example . etc. This proposal is interface systemsor correspondence " connectedto what he calls " representationalmodularity . Technically. For details of this aspect 2. or argument positions for that matter. including LF .very close in spirit to the SF information SF(Ie) of lexical items. Bierwisch ( 1988 ). see . suggestingthat autonomy of mental modules is a property of representationalsystemslike phonological structure ). x and y can in fact be considered as syntactic lambda operators.3 Remarkson Modularity of Knowledgeand Representation The main reason to distinguish SF from syntactic representations . is the linguistically relevant internal structure of lexical items connectedto the conceptual . this volume) about conceptual structure (CS). however . 1987 . amounts semantically to functional . respectively constituents. conceptual structure (CS. PF (le) GF (le) SF(Ie) Here x and y specify the obligatory subject position and the optional object position of leave . he explicitly aside claims that conceptual structure (CS.details . . abstracting over the pertinent variables.). which has consequences " this. ~ ~ ~"V ~ ~! J Ix . but also articulation . The compositional structure claimed for interpretation of linguistic expressions SF is very much in line with the proposals of Jackendoff ( 1983 . such that assigningtheta roles.? How Much SpaceGets into Language ( 18) / Ieave / ~ . rather than complex faculties like I -language this sort are then connected to each other by interface or correspondencesystems . (PS). identifying the semantic variables to be bound by the corresponding . Hence CS . and hence LCS) is an extralinguistic level of .~ DO [ BE CO M [ NEG [x AT y ]]] ! Ey . with one important difference ./ "" uditiol1 l J y Locomotion I -language Jackendoffproposesa principled distinction betweensystemsor modules of representation ). and supporting the levels of representation indicated by the labels in ( 19 rules representedby the arrows. Autonomous modules of vision. application . In other words. Although what Jackendoff calls " lexical conceptual structure (LCS) is.

but merely on the appropriate characterization of I -language . of course . ' I -language . the phenomenadiscussedabove in connection with the interpretation of leave enter . argue that the problem is not conceptual. rules relating PS and SS . But this would spoil the modular autonomy of CS and its extralinguistic status. First . SF(le) point . to the initial claim schematizedin (7). and so on could not reason ably be explained without accounting for their fairly abstract linguistic structure and the specific distinctions that depend on factual knowledge. The bulk of correspondence . in Jackendoff s conception. Let me return . the subsystems within I -language . or rather LCS. but also acrosslanguageand other mental systems I do not think this is the right solution . there seem to be substantial generalizations that crucially depend on the linguistic nature of SF(le). there seems to be a systematicdistinction betweenlinguistic and extralinguistic factors determining conceptual and referential interpretation . on the one hand. While this is a plausible way to look at lexical items. that linguistic knowledge is not a neatly separated system of mental organization. . are lexical items. Third . which terminological. in this regard. as it is their very nature to mediate betweenmodules. . I -language is altogether a highly specific interface mediating two independent systems of computation and representation . does not? To put it differently . the nature of correspondencerules in general remains rather elusive. namely that I -language (based on UG ) simply determines a systematiccorrespondencebetween the domains APand C-I . on the other. they must belong to the core of linguistic knowledge based on the principles of UG . though. and the correspondence rules connecting them to their adjacent levels but not CS. then. either CS (and hence LCS) is not included in I -language or lexical items belong to the system of rules included in linguistic knowledge. How can lexical items as part of the correspondencerules belong to I -language . To some extent.by definition . Although one might argue that this is just a consequence of actual fact. Second . or even GF (le) .cannot be autonomous.) In this respect is no lesspart of I -languagethan PF(le). SS .which. comprisesPS.4 correspondence One might . In this view. it createsa conceptual problem. the principles of argument-structure being a major casein . but they appear also to depend on quite different principles of mental organization. it seemsto me this conclusion can and in fact must be avoided. . In other words. and SS and CS. turning cannot in be schematized as 22 the lexical system not only cuts across simply ( ). for at least three reasons . but not both. If thesedistinctions are not captured by two levelsof representation SF and C I in my terminology then two aspectsof CS must be distinguished in somewhat similar ways. (This is a contention clearly sharedby Jackendoff. if SF(le). institute.

determining systemsof rules or principles can rely on the same system of representation . rather than levelsof representation subsystems The notion of level of representation need by no meanscoincide with that of an autonomous module. nor must a level of representation belong to only one module of knowledge. But neither must one module of knowledge be restricted to one level of representation . and so on. say 12121942 or your bank account. first of all . It is the aim of this chapter to make this C-I with the language . respectively SF(le) to its conceptual interpretation . different aspectsof actual representations might best be illustrated by examplesfrom different nonlinguistic domains. exploiting the same representational resources .? How Much SpaceGets into Language Under this perspective . that the difficulties concerning the status of CS are largely due to the notion of representational modularity . different constituents. Instead. respectively . The same sequence case is the representational system consisting of sequences . view more precisewith respectto the subdomain of C-I representingspace Notice . the componentsof PF (le) and SF(le) as such provide the interface of APand -internal computation . Each of theseinterpretations belongs to a different subsetof . Instead of speculating about the nature of these intermodular systems(are ?). PF and SF are theoretical constructs sorting out those aspects of APand C-I that are recruited by UG in order to compute the correspondencein question. which is intended to overcome the ' ) concept of modularity . Such interpretations of the same representation are based on different rules or systemsof knowledge. where no simple correlation betweenlevelsand I want to indicate that . each of which is construed as an autonomous module. Notice that certain operations on the representation would have the same effect for each of the interpretations. A simple of digits. there is no systemof knowledgewithout representations to which it applies. I will not go here through the intricate details of subsystems and levelsof syntactic representation . or PF (le) to articulation for that matter. of tacit knowledge. . as illustrated in (20a) and (20b). your office phone number . each subsetdefinesdifferent neighbors . becausethey affect the shared properties of the representationalsystem . might happen to be your birth date. different . however. Hence PF (le) and SF(le) representstructural conditions projected into configurations . in a more general sense modules obtains. To be sure. Rather. There are no correspondencerules connecting in APand C-I . subject to different restrictions. What I have in mind . I suggestwe go back to the they supposedto be encapsulatedand impenetrable notion of modularity first proposed by Chomsky ( 1980 ). For none of them is the fact that the sequences number is divisible by 29 relevant. Replacing the inadequaciesencountered by Fodor s ( 1983 overall languagemodule by a number of representationalsystems . characterizing systemsand . while others would have different effectson alternative : recruitings. Jackendoff is forced to posit interface systems as well.

The notes exhibit simultaneously a position within the tonal systemand. This seemsto me correct. knowledge of languageis essentially basedon tacit knowledge. It might be objected that the representationsconsideredabove are not really identical under their different interpretations. within the Latin alphabet. In this respect . but it does not change the fact that we are dealing with annotations imposed on configurations of the same representational system . while the examples rely on rules and exhibiting an E. . Obviously. the relevant elements annotated in some way with respect to the rules of different systemsof knowledge. and so on. albeit under different interpretations.and an I -aspect elementsthat are more or lessexplicitly defined. of " their " names . digits and notes are comparable to language . Third .are crucial with respect to the way in which different modules of knowledge are . from which the additional interpretation recruits designated components . different rules apply to the two interpretations. the tonal and the graphemic interpretation of the representation . Again. and year. 6 imposing its own constraints. month . according to extensions . This case is closer to what I want to elucidate than the different interpretation of digits . especiallyif we try to identify the information contained in their I -representation : digits representingdatesare grouped according to . Second . notes. telephone numbers.must be construed as other words. Although s all notes have alphabetic names . In day. it is the internal structures and the pertinent knowledge they are based on that we are interestedin. Both of the and indication of affiliation aspects identity representationalsystem specific . First . the conceptual considerationscarry over to I -languageas well as other mental systems . Moreover. However. and so forth . because . In other words. even though theseillustrations are given in terms of external representations . the artificial character of the twofold interpretation in our examples by no means excludes the existence of the same structural relationship with respectto systemsof implicit knowledge. the more complete interpretation (in this casethe tonal one) determinesthe full representation .digits. not all letters are representableby notes. the apply simultaneously two interpretations rely on different cutouts of the shared representation .

- A-P I -language y - . but rather on two types of knowledge that participate . which include various aspectssuch as effectsof the particular ' speakers voice. so to speak . . These considerations lead to what interfaced by a given representational system ' " " " might be called modularity of knowledge.. the notion of interface... In other words. dependingon which modules are at issue that establish es an interface betweenAPand C I . which does not exclude the levels of C I further that APor .. PF and SF are the in one and the samerepresentationalsystem interfaces of I -languagewith APand C-I . then the levelsof PF and SF are each determined by (at least two modules of knowledge. as we will see representation support possibility below. I -languagemust be interfaced . imposing conditions on. each other. ignored by I -language subsystems interest here.v.. respectively . possibly adding annotations in the sensementioned above. Furthermore. if this is so.. using configurations . Second ) . but are. The moral should be obvious. in contrast to Jackendoff s representational " modularity .? How Much SpaceGets into Language . or recruiting elementsof ..SYNTAX . Turning finally it as the designatedaspectof C 1 to which I languageis directly related.+ SFJ +. First . I -languageas a whole is a system concept. One might. Theseare determined by their own . +. This sort of interface is not basedon rules that map with APand C-I . I want to schematizethe view proposed here by a slight modification of (8) : ( 22) . linguistic representation various possibilities concerning ( 1) how SF components recruit elementsor configurations of C-I .or correspondencefor that matter. . (2) what annotations of SF must be assumed . We will turn to thesequestionsin the sectionsbelow. recognize (le) as ( ) perspective specifically Looking APIt is based on on the linguistic aspectimposed temporal patterns determined by articulation and perception.+ lPF + . distinguish different aspects of one representation by setting up different . emotional state. . To conclude this section.- C-I .I ' . respectively one representationonto another. it must .+ . and so on. but some comments seemto be indicated. While this may be helpful for descriptive purposes levels of representation . with languagecapacity basedon UG providing the requisite knowledge. not obscurethe sharedelementsand properties of the representationalsystem PF we PF this at le under more . we will now recognize which is of to SF le primary ( ). of course . and (3) how rules and principles of C 1will contribute to the interface representationwithout being reflected in I -language . . Y . This leaves open of C-I as elements of its own.is a relative .

substances .) The dots in (22) indicate the (largely unknown) internal organization of C-I . it is involved primarily in visual perception and kinesthetic information . such that more completely Loc (x . plasticity and rigidity . or shadows . on the one hand. t) identifies the place of x at time t . All these systems provide nonspatial information as well. 2. including regions. I will therefore assume . presupposingstandard assumptionsabout time intervals. auditory perception. and metrical structure. with a topological and metrical structure imposed on this orthogonal dimensions set. and contrasting " " with physical. SR should meet the following conditions: According to generalconsiderations I . I -spacein this sense must be assumedto control and draw on information from a variety of sources . SR should thus be construed as an interface representationin the sense . SR is based on a (potentially infinite ) set of locations. (Parallel considerationsapply to PF. and the conceptual systemC-I . two further conditions are determined for locations: . 2. Locations can be occupied by spatial entities (physical objects and their derivates like holes. Spatial properties of physical entities are thus related to the structure imposed on the set of locations. and events ). Loc (x ) must be taken as time-dependent . although the aspect concerned need not be identical. but it also integrates information from the vestibular system . that is. In general . and so forth . . external or Espace. on the other. 3. corresponding to I -language . Vision integrates color and texture. that I -spaceselects information from different sourcesand integratesit in a particular system of spatial representation(SR) . that SF is governed by conditions of I -language as well as those of C.4 TheConceptualization of Space What interests us is the internal representationof spaceand the knowledge underlying " it . Before looking more closely at the status of SR and its role for the relation betweenI -spaceand I -language .I .Manfred Bierwisch The main point is. among other things.) 4. comparable to the way in which PF reconciles articulation and audition with I -language . topological structure. related according to three . this volume). and haptic information . As a first approximation. (Motion can thus be identified by a sequence of placesassignedto the samex by Loc (x . I will provisionally indicate the format and content to be assumedfor SR. following Jackendoff (chapter I . haptic and kinesthetic information distinguish. which we might call " I -space . of course . as mediating just discussed betweendifferent perceptual and motoric modalities. In addition to dimensionality. such that Loc (x ) is a function that assignsany spatial entity x its place or location. to which we turn now. t ) .

let me clarify the point by the following simplified example: (23) a. and auditory perception. B LEFT -OF C & B OVER A (24) a. x OVER y correspondsLoc (x ) + Loc (y) (23a) is a pictorial representationof a situation for which (23b) gives three possible indicated in (24).4 abovecan serve . while conceptual structure must provide a representation strictly bound to spatial experience for experienceof all domains. information not bound to sensory domains in direct ways. C correspondsto ~ . however. social relation . b.the . the type of representation at SR is depictive of or analogous to what it representsin crucial . orientation with respectto a designatedorigo and/ or observerand intrinsic conditions of objects (canonical position or motion ) .? How Much SpaceGets into Language a. the dimenDepending on how physical objects are perceived and conceptualized . SR level of conceptual structure (CS) is assumedto be domain specific . and so on. but also emotion. which we have in fact assumedfor SR in conditions 1 and 2 . provided the correspondences propositional representations " conceptual lexicon apply. that is. i. All that respect " in the is neededfor a representationalsystemto be depictive is a " functional space senseexplained in Kosslyn ( 1983 ). algebraic. or even zero dimensions of this would have to be made precise in a serious theory of SR. nondepictive. A correspondsto O . goals of action. The provisional outline given by conditions 1. as a basisfor the following remarks. All sionality of their locations can be reduced to two . representingproperties and distinctions that are . together with the principles that relate the functional structure" underlying (23a) to the compositional structure of the representationsin . b. propositional . while CS is abstract. it should still clearly be distinguishedfrom the for at least two interrelated reasons . taste. Notice that although SR is transmodal in the sense already mentioned and must be considered as one of the main subsystemsthat contribute to the conceptual and intentional organization of experience . A OVER B & C RIGHT -OF B iii . Second . A OVER B & B LEFT -OF C ii . orientation of the dimensions . . Becausethe distinction betweenthe depictive nature of SR and the propositional character of CS is crucial for the further discussion . smell. 0 D~ b. including not only color . B correspondsto D . one. First . marking especially a directed vertical dimension (basedon gravitation ). that is.

3. not necessarilyan autonomous module of knowledge. which supports an analogical relation to conditions of Espace. where it could optionally be added but only in necessarilyexplicit manner (e. to 4. and so on. allowing for essentially different operations based on its depictive character.. Presupposingan intuitive understanding of the correspondencein question. Finally . with objects representedin terms of 3 0 models in the senseof Marr ( 1981 ). even though in necessarilyimplicit way. and so on in (23b). the circle. this volume) and by Byrne and Johnson Laird ( 1989 ).establishedby (24a). nor are there explicit elementsin (23b) representingthe properties of the objects in (23a). SeeJackendoff ( 1990 I . who demonstrate interesting differences between inferences basedon this type of representation . by adding coded units of measurement ). while no such anaphoric repetition showsup in (23a). (23b) allows for various alternative representations correspondingequivalently to the unique representation in (23a). this volume) for further discussion . 2. LEfT OF . I will point out the following essential differencesbetweenthe format of (23a) and (23b) : I . that is. Whereasthere is an explicit correspondencebetweenunits representingobjects in (23a) and (23b). for example . It differs from CS by formal properties like I .Manfred Bierwisch (23b) . representations type ( ) Returning to SR. and chapter II . it is not indicated in (23b). for the samereason . Given the variety of sourcesit . 4. it seemsin fact plausible to assumethat SR draws on different systemsof integrates mental organization.Laird ( 1983 . The different distance between the objects is necessarilyindicated in (23a). which could be made precisein various ways. the square . visual perception providing the most fundamental as well as the . and chapter ) . while no such connectivesmay appear in (23a). (23b) requires logical connectivesrelating the elementarypropositions. or configurations of geons as proposed by Biederman ( 1987 . by different distances between the objects. the properties of (23a) are essentially those of mental models in the sense discussedby Johnson. it seemsto be a plausible that it constitutes a conjecture pictorial representation in the senseof (23a). as opposedto inferencesbasedon propositional of 23b . According to the view proposed in the previous section. SR might rather be considered as one aspect of a representational system shared by different modalities. while (23a) would allow for variations that need not show up in (23b)..there are no explicit units in (23a) representing the relational concepts OVER .g. The next point to be noted is that SR as construed here is a level of representation . In general . Additional properties or relations specified for an object in (23b) require a repeated " " representationof the object in question.

the SR aspect of the representational system is subject to or participates in operations ) like imaging or mental rotation of objects. This conclusion is that C-I comprises at least two different levels of representation conclusion should not be surprising.as of places . for instance Obviously. which are argued by Kosslyn et ale( 1985 . The obvious . objects interacting relevant shapecharacteristics . The main purpose of the outline given above is to indicate the sort of CS information that SR is to account for .are sequences volume) points out further aspectsand requirements to be added. On the other hand. but also modality -specific This leavesus with the question of how SR relatesto the overall systemC-I and the level of conceptual structure in particular . . the correspondencebetweenSR and CS must provide the SR rendering of the following specifications for spatial conditions: I . and 4. relations of objects with respectto the location of other objects. changesof place in time. what I have metaphorically called the " " conceptual lexicon (24) correspondsin a way to the lexical entries.4 are not independentof eachother. and in a less metaphorical vein. paths. Shapeof objects. This leavesopen whether. Jackendoff (chapter I . and to what extent. SR be two interlocked aspectsof the samelevel of representation to the extent to which it is to be identified with Johnsonmust belong to C-I . this already mentioned. Size of objects. without trying to actually specify the format of representations . it has in fact a straightforward parallel in 1. specifications1. albeit of a different character. to be not only depictive. because Laird ' s system of mental models. let alone the principles or rules by which the relevant knowledge is organized. Paths of (moving) objects. More generally. To carry this analogy one step further . which I need not repeat here. that is.? How Much SpaceGets into Language most differentiated contribution . and so forth . Shape is to some extent determined by size and place of parts of an object. . . parts with respectto their conceptually relevant axesor dimensions(3 D models with the of characteristics metrical that is 2. where PF and SF also constitute two essentiallydifferent representational language systemswithin the sameoverall mental capacity. proportional metrical characteristicsof objects and their ). that is. then SR and CS cannot . it supports logical operations similar in effect to those basedon the propositional -level CS. the pertinent 3-D model determinesthe representationof a given concept on the level of SR. Place of objects. that is. 3. If the comments on the propositional character of CS and the depictive nature of SR are correct. . Just as PF (le) indicates how the corresponding SF(Ie) is to be spelled out at the level of PF.

for example . More complex analogiesin the expressionof spatial. . namely . loudness . This observation. temporal. the taxonomic classification. and (2) by virtue of this common ground. in turn . My secondcomment has two parts. We are dealing with . Thus the commonsensetheory about cats will include conditions about the characteristic behavior. the informal rendering of SR in conditions 1. by Gruber ( 1976 ) and by Jackendoff ( 1983 ). further domains. events . possessional . oriented spacewith eventsbeing mapped onto intervals just like . . I will assume . places . Hierarchies of different sorts. along with access to the shapeas specifiedin SR. the transmodal aspect in question clearly must exceed I -space (in the experience sense assumedthus far ). one might say. which immediately raises the question of how this generalizedcharacter of basic spatial structures is to be explained.objects being mapped onto locations. First . such as social. and so on or in fact I because the corresponding ontology . is important for two reasons : ( I ) in spite of its domain specificity. necessarily holds also for SR. are construed in spatial terms.come easily to mind. that is. Time appearsnecessarily be conceptualizedas one-dimensional. to external conditions that constitute real. I will return to this problem in the next section. that 3-D models spell out properties in SR that general conceptual knowledge combines with nonspatial knowledge about specific types of physical objects.4 at the beginning of this section freely refers to objects. and even time do not rely on the same sourcesof primary . is entrenchedin someway in I -space More specifically. geometrical space . the sortal and type structure of concepts . spacein the literal sense as mentioned above. evaluative. Becausetaxonomies.temperature . taxonomic. ( I ) I want to affirm that spatial representation as discussedthus far respondsto properties and relations of physical objects.Manfred Bierwlsch I will conclude this sketch of the status of I -spacewith two comments that bear on the way spatial information is conceptually structured and eventually related to SF. it also participates in a general framework that underpins the interface with general conceptual structure. color . The conclusion from this observation is this. and hence to I -language . tonal scales . and so forth of cats. SR shareswith general conceptual organization basic onto logical structures. . and so on. relations. This leads to (2) the observation that spatial structures are to extensivelyemployed in many other conceptualdomains. it is worth noting that commonsenseontology . basedon spatial perception of various sorts. SR not only provides entities in terms of which intended reference in C-I can be established and interpreted. The basic conditions of I -spaceas listed at the beginning of this section seemto be available as a general framework underlying different domains of experience . functioning as an organizing structure of generalconceptual knowledge. properties. for example . relations have been discussed . social relations. suppose legitimately.

this volume) remarks. which would be of alternative 2. given the present state of understanding of conceptual structure. whether imported to I -spaceaccording to alternative 1. I will not deal with explicit a natural consequence transfer but will argue that implicit transfer is a major reason for the observation . the structure emerges in the primary domain. to correspond common sense with to to. dimensionality and orientation require appropriate structures of other domains. according to alternative 2. actual three-dimensional spaceis the prevailing. originate as an intrinsic condition of I -spaceand are projected to other domains on demand. The notion of color spaceor property spaceis based structure on primary experience on this sort of explicit transfer. or exported from it according to alternative 2. Whereas 1space is basically three-dimensional. without explicit stipulation . whose conceptualization follows these patterns automatically . I tentatively : ( 1) I -spaceis assumethat alternative 2 is correct for the following two reasons not only a privileged instantiation of spatial structure but is also the richest and most detailed instantiation of spatial structure. These structural considerations might be supplementedby onto genetic and phylogenetic considerations . which I will not pursue here. with its type and sortal distinctions. The boundary betweenexplicit and implicit transfer need not be clear in advanceand might in fact vary to some extent. (2) While size and place carry over to the other domains with scalar and topological properties. or 2. I will consider as implicit transfer the dimensionality and orientation of domains like time or social hierarchies . constitute a general schema of conceptual knowledge imposed on different domains according to their respectiveconditions. In what follows . explicit transfer shows up in cases where dimensionality is used as a secondary organization. dominant instantiation of an abstract structure that exists in a senseindependent of this instantiation as a result of experience . This is similar to what has been said earlier respect ontology . rather than as an abstract potential that happens to be completely instantiated in I -spaceonly . According to alternative 1. I will thus consider the full structure of I -spaceas intrinsic to this domain due to its specific input . and phylogeneticrespects impact in structural . or rather of conceptual structure in general . In contrast. Orientation with respectto frame of referenceis accordingly reduced to only one dimension. It might be useful to distinguish two types of transfer of spatial structure. as Jackendoff (chapter 1. other domains are usually of reduced dimensionality . onto genetic to make. imposing an additional . compared to other domains. but it is a difficult choice . The choice between these alternatives has clear empirical . shape has only very restricted analogy in other domains. In any case .? How Much SpaceGets into Language Basic structures of spatial organization must therefore either 1. that is.

that there is no clear distinction between spatial and nonspatial terms . and chapter I . ) Reyle ( supposethat CS exhibits a fairly rich sortal structure common sense provided by ontology . two very general assumptionswill be sufficient in the presentcontext. furthermore. Second I will . as discussed I will refrain from speculationsabout the primitive elementsof CS. for example . their format and organization are . CS is modeled as tightly as possible in accordancewith the structure of linguistic expressions to be interpreted in CS. with two exceptions : ( I ) the primes of SF must be compatible with basic or complex units of CS. that CS is propositional in nature. a particularly explicit version being Kamp and assumptions can be made precise 1993 . looking more closely at the way in which spatial information is accommodatedin CS. CS dependson various other systemsrelating it to domain-specific information . however. On the one hand. CS is basedon functor -argument-structure.noted at the outset . this volume) . Note . on the other hand. that CS must not be identified with encyclopedicknowledge in . in the sense generalagreement above and discussedin more detail. There seemsto be indicated . 1990 . enter . or leave are not restricted to space because of the implicit transfer of the framework on which they are based. by ( ) and by Jackendoff .5 Types of Space Relatednea in Conceptual Structure Let us assume . Hence CS does not rely on sequential ordering of elements but only on nesting according to the functor -argument structure. As to the general format of CS. by in . with functional application being the main (and perhaps only) type of combinatorial operation. it is made to comply with requirements of logical inferencesbased on situations and texts. including SR. I will return to exception 2 shortly . There are various ways in which these . The two main sourcesrelied on in specifying ( 1983 CS are languageand logic. if the assumptions about SF and its embedding in CS are correct. 2. . for example Fodor 1975 . Both assumptionsshould allow CS to be interfaced with the semanticform (SF) of linguistic expressions earlier. and (2) CS must accommodateinformation from various domains. Notice first of all that assumptionsabout the properties of CS can only be justified by indirect evidencebecause . that the conceptual-intentional system(C I ) provides a level of representation(CS) by which information of different modules is integrated. Although commonsensetheories by which experienceis organized and explained general must have access to representationsof CS. by definition . specificationsof 3-0 modelsas basicelementsthat feature in CS representations example . to conclude the foregoing discussion . possibly treating for . namely . First . The relations expressed.

although inherently spatial. Moravcsik 1981 ) that commonsensetheories are organized by explanatory factors according to Aristotelian categorieslike structure. It has been suggested . 2 . size . Pending further clarification . informational structures (like arguments or and social institutions . observation 3 separatesconceptual units that specifypurely spatial conditions for whatever entities fall within their range from conditions that inextricably involve additional conceptual information . function . three observationsseemto me warranted: I . as Landau (chapter 8.. 3. including in particular shapeand sizeof objects and portions of substance This observation distinguishes " aspatial" conceptual entities. substance . I will I simply assumethat C determinesrelevant aspectsof CS on the basis of principles that organize experience . Also . this volume) vary in the course of onto genetic development . but rather a gradual distinction . plant. while animal. Turning next to the way in which CS and commonsenseknowledge integrate I -space . which.g. Encyclopedic knowledge mayor may not determine particular definitional or characteristic spatial properties within the limits set by ( I ) . Although these aspatial entities are invested with spatial characteristics by the physical objects implementing them.mental states . it should be clear enough that . It remains to be seenhow this conjecture can be made explicit in the formal nature of commonsense knowledge. tool. Thus square . snake . Dog.furniture exemplify conceptsof the secondtype. table. Pustejovsky 1991 (e. songs . showing that young children initially tend to invest conceptsin general with argues spatial information . Thus the concept of vehicle is spatially far less specific than that of cat or flute . from those subject to spatial characterization. This observation simply notes that spatial entities are divided into those whose . While observations I and 2 distinguish conceptual entities with respect to their participation in spatial conceptualization.HowMuchSpace Gets into Language ? to be distinguished from bare representational aspectsof CS. or pencil expressconcepts of the first type. Actually observation 2 does not set up a strictly binary. Conceptual units may specify spatial properties or relations without involving any nonspatial properties of entities they can refer to. dependingon the specificity of shape . and positional information . are not characterizedby particular spatial information . a poem as a conceptual entity is to be from the letters that distinguished printed representit . but it still contains spatial conditions absent in the concepts of machineor musical instrument . and those without specifications typical or essentialproperties involve spatial characteristics of this sort. Commonsense . for example . poems ). ontology -requiresphysical entities to exhibit spatial characteristics . child. and relation. . even though theseare not aspatial. the specifity of spatial properties seemsto .

SeeKeil ( 1987 ) typical or characteristic.Manfred Bierwisch .further systematic conceptual knowledge.in addition to shapeand size information . social institutions like parliament or informational structures like novel or sonata . black. for example . without being definitional in the strict sense for relevant discussion . even if they are essentiallyspatial. margin. Extrinsically spatial c. but neednot fix it . Strict Example live. using linguistic expressionsonly as a convenient way of indication . Intrinsic d. instrument horse . the distinction between (25b) and (25c) is hence possibly to be replaced by various steps according to the specificity of spatial information .d). duration animal. sky red. Strictly spatial . observation 2 separates (25d) " " from (25a c) . hour. circle. and so on. tool blood. margin. It should be borne in mind that we are talking here about conceptual units. as already mentioned. animal. For the time being. Expressions for aspatial concepts .2. hour. man. Extrinsically spatial refers to conceptsthat require spatial properties but do not specify them. Thus lexical items expressingstrictly spatial concepts are extensively used to refer to " typical implementations" like corner. Aspatial b. " intrinsically spatial" indicates the specificationof (someof ) these properties. top (in one reading) expressstrictly or exclusively spatial conceptswhile edge or dog cup include. on the other hand. Extrinsic c. The different spatial character of conceptsdiscussedthus far can be schematically summarizedas follows: Example fear . The main point is that concepts can involve more or less specificspatial information . Intrinsically spatial . color and time being casesin point : (26) Type of color -relatedness a. It should be noted that intrinsically spatial properties might be . zebra . are used to refer to spatial . height liquid. colorlessness (25) Type of concept a. No relation b. Theseare problems objects where they are located or represented of conceptual shift of the sort mentioned in section 2. square . which must be analyzed in their own right . height square Observation 1 distinguishesbetween(25a) and (25b. which might be of various sorts. we ignore variability in the interpretation of lexical items. robot. violind . As already mentioned. It is worth noting that the samedistinctions (with similar provisos) apply to other domains of conceptual organization.

not about the nouns. The point at issueis merely that the observations . wet. and (2) it is essentiallyrich conceptsthat constitute commonsensetheories: . concepts ' " with a fairly rich array of different conditions. however. slanting Relation . relations. that we are talking about concepts . once again. about function or lessspecificshapeand size information with knowledge . No relation b. them. es actually a three place relation . lion . Intrinsic d. beat hour.3 noted above are not an isolated phenomenonof space Thus far I have illustrated the distinctions in question with respect to objects of different sorts. basedon highly restricted conditions of only one or two domains. with respects rich conceptsbeing subject to further elaboration. Aspatial Extrinsic Intrinsic Strict Property clever . no sharp boundary here. combine more . c. 1. sober . and functions. There is. near. Let square " " and " me call thesetwo kinds " rich concepts . and so on that might be gradually extended on the basis of additional . write close . long.3. which would have to be clarified with respect to the particular domains in question. Strict Example number . however. behavior. in much the same way to other onto logical types. The main below. water. there are. beginning . on the one hand. inauguration . show . for the sakeof discussion spareconcepts . . famous colored . substance . commettee death. First . as shown in the appendix practice. rather than a property . open upright. squeezed under . while spareconceptsare just what they are. pierce. express point should be clear. Thus difficulties must be observed further long. there are relatively spare concepts such as near. of course . but the differenceis relevant in two : ( I ) spareconceptsmight in fact enter into conditions of rich concepts . experience stand . b. The distinctions discussedthus far are directly related to two additional observations important in the present context. such as properties. travel fear . (28) gives a sample illustration : (28) a. duration There are numerous problems in detail.? How Much SpaceGets into Language (27) Type of time-relatedness a. place Notice. explanation. On the other hand. verbs. during acknowledge kill . In addition to distinctions blurred by this .Pustejovskys ( 1991 ) qualia structure " for . . Conceptsof different types are subject to the distinctions related to observations 1. broken. Extrinsic c. Concepts example integrated into theories of commonsense artifacts like car or elevator but also of natural kinds like dog or raven . The observations apply. prepositions expressing adjectives . solid striped.

the distinction of rich and spare try) is explained by circle. Thus. record and circle. table. for example which relies. representedin SR by 3 D models. encapsulatedin configurations of SR. Becausewe have further assumedthat CS is the interface of C-1 with I -language . The moral of all of this with respect to our initial question would thus be something like the following . An object schemaspecifiesthe conditions that explicit representationscould extract from encapsulatedshape informa - . steepmust be able to extract the relevant long strictly spatial dimensional and orientational information from the encapsulatedshaperepresentation of man.e. Contrasting. that is. tall . . In contrast. and even if the usual problems with borderline cases A major problem to be faced in this connection is the fact that in CS strictly spatial (i. long table. they do not . far . enters the pertinent which meansthat neither the internal structure of 3-D models nor conceptsimplicitly . defining. but can be integrated into intrinsically spatial ones . For almost trivial reasons between extrinsic and with the distinction but is not identical relates to ) ( concepts . Specificationsrepresentedin SR can be " " " " relied on in CS in two ways. intrinsically spatial conceptsrepresentspatial information implicitly . however. the or information of short . opposed strictly spatial concepts not vice versa. and (2) explicitly representedby means of conceptual primes that directly recruit elementsof SR. One might take this as a corollary conditions spatial : of the classification illustrated in (25) in the following sense Strictly spatial conceptsrepresentspatial information explicitly in terms of conceptual primes. shape part of detail sound storage (in varying degrees ). it follows that SF has to SR. but rather the shape information as a whole. for instance . or steeproof. we notice that circle is explain anything . there are essentialprovisos to be made. on knowledgeexplaining in record information of the . I will return to this point below. while nothing (beyond mere geome . can be overcome . Related to this is the secondobservation. for the complex conceptsexpressed by short man. or roof A useful proposal to overcomethis problem is the notion of object schematadevelopedin Lang ( 1989 ). even if the notion of explicit and implicit representation can be made formally precise . and so on must explicitly representthe relevant spatial conceptslike behind in terms of conceptual primitives . strictly representations . " " " the properties reconstructing them like " four -legged or long-necked enter CS . which I will call explicit and implicit .Manfred Bierwisch although spare concepts like in or long can feature in explanations. Strictly spatial concepts to as intrinsic spatial concepts . Although I take this moral two typesof access to be basically correct as a kind of guideline. CS extracts information from SR in two ways: ( I ) encapsulated in SR configurations that are only treated holistically. so to speak .. explicit) conceptsmust appropriately combine with implicit spatial information . Detailed shape information . an open set of primes in terms of conditions in SR.

In other words. How is I -spacereflectedin CS? ? 2. which are . An object schemamakes 3-D models respond to explicitly . it is not just a simplification of the model. The latter would include only explicit concepts . For details seeBierwisch and Lang ( 1989 A final distinction emergingfrom the observationsabout I -spaceand C-I should be of the implicit transfer imposing basic structures of I -space noted. which we noted above. ) and Lang ( 1989 long instrumentshow. As a consequence on other domains. Notice that there are default schemataalso for extrinsically . in particular . dimensionality. being interpreted by structures of SR that transfer to other domains. such that the conditions on the format of SF representationsoutlined in section 2. the central question we posed at the outset boils down to two related questions: 1. How are spatial aspectsof CS taken up in SF We have already dealt with question 1. A partial answer to question 2 is implied by the assumption that SF and CS. we found a fairly rich typology of different elementsand configurations thereof in CS. it seems plausible to assumethat explicitly spatial concepts like in. Pursuing now question 2 in more detail. but rather its rendering in terms of primes of the strictly spatial sort. as combinations like ). but systemsof knowledge. strictly spatial only if interpreted in I space Not surprisingly.2 would carry over to the format of CS. that SF can be thought of as embeddedin CS.? Gets intoLanguage HowMuchSpace tion . unlessspecific additional requirementsare motivated by independent . so to speak spatial concepts spatial conceptsthat do not provide a specified3-D model. Even though an object schemais lessspecificthan a 3-D model.6 BasicSpatialTenns Assuming that the relation of spatial cognition and conceptual structure is to be construed along the lines sketched thus far . canonical orientation and subordination of axes relative to eachother. I would like to stressthat the observations from which this typology derives . depending only on the way in which SR as a representational systemrelatesto I -spaceas well as other cognitive domains. we are led to a distinction between elementsof CS that are exclusively interpreted in SR and elementsthat are neutral in this respect . I will stick to the assumption made earlier. : Outlineof a Program 2. neednot be construed as disjoint representationalsystems modules of to different rather as ways to recruit pertinent configurations according knowledge. although determined by distinct and autonomous . length. are not stipulated conditions but about the architecture of subsystemsof of basic assumptions simply consequences C-I and their internal organization. and around do in fact relate to I -space and other domains to which the pertinent structures are transferred.

for example . Thus my aim in assumingthese criteria is not to justify the delimitation they define. First we define the systemof basic spatial terms ( BST. native. while hexagonaland squeeze . under . adjectives Case ). statementswith respectto encapsulated information of intrinsically spatial concepts . or even trivial . practical reasons It is immediately obvious that the two criteria . and perhaps classifiers and inflections for (verbs. Given theseconsiderations . Such additional requirementsmight relate. and . respectively heuristic notion with no systematic impact beyond its role in setting up a research strategy. for short) of a given language . BSTs cannot. An additional CS are recruited for lexicalization language point concerning further grammaticalization in terms of morphological categorieswill be taken up in section 2. GF (le). discussed are not . violating Thus short. side . morphologically simple). and (2) understanding object schemata the lexicalization of strictly spatial conceptswould be a necessary precondition in any case . lead to various systematicallyrelated subsystems I . prepositions. core of the lexical systemof a given language 2. 2 . SF(le)] that belong to the basic (i. BSTs identify strictly spatial units in the sense above. .. by question in I . Becausespace is a far and is similar in spirit . Linguistically . but rather to rely on them for . . be restricted to adjectives basic color terms can. .evidenceconcerning the nature of CS. BSTs belong to different syntactic and morphological categories . Basic spatial terms can be characterizedby the following criteria: I . The notion of then we look at the properties they exhibit with respectto question 2 basic spatial terms has beenborrowed from Berlin and Kay ' s ( 1969 ) basiccolor terms . of motion etc. ) place. Conceptually. Hence one might relax or change the criteria should this be indicated in order to arrive at relevant generalizationsor insights.e. . and has in fact been followed implicitly by a great deal of researchin this area. It should be emphasizedthat BST is a purely criterion I and criterion 2. I will restrict the issueof lexicalization to strictly spatial conceptsfor two reasons : ( I ) to go beyond obvious. shape . would by far exceedthe limits of this chapter. including the intervening effectsof . the following researchstrategy seemsto be promising. In their semantic form [SF(le)]. though different in certain respects more complex than color .7. nouns. change . even in their rather provisional of BSTs: form . for . as . lie are BSTs. BSTs are interpreted by different aspects of space (size . size . BSTs are lexical items [ pF(le). to commonsenseontology and the sortal systemit induces example 2 is which elementsof the main issue raised With theseprerequisites .

The boat is long. . by which UG organizes BSTs with respectto their SF. This includes in particular the question whether universal grammar provides an a priori system of potential conceptual distinctions that can be relied on in the SF of BSTs. as well as their syntactic and morphological properties. : respectively (29) a. systematicor incidental. a researchstrategy taking BSTs as a starting point is oriented toward (at least) three aims. dimensional aspect of the speaking entity it applies to and assignsit a quantitative value. high. The boat is ten feet longer than the truck. Generally . or simply as different ways to idiosyncratically exploit the range of possibilities provided by principles of C-I and UG . for short) like long. The boat is twenty feet long and five feet wide. Identification of the conceptual repertoire available to BSTs. specifying somehow opposite quantitative values with respect to the same dimension. either strict or preferential. Identification of systematicoptions that distinguish languageswith respectto the repertoire and the patterns they rely on. possibly complex. and low. however. of course . As a preliminary illustration . The boat is ten feet shorter than the truck . a measurephrase like tenfeet can naturally be combined only with the " positive" DA . d. the relation betweenlinguistic ( 1) and conceptual 2 whether ( ) subsystems . b. where it combines with the positive as well as the negative DA . Theseand a wide range of .or whether the distinctions made in SF are abstracted from actual experienceand its conceptualization. Ultimately . Characteristically.parallel to what is generally assumedfor PF primes. The opposite direction of quantification specified by antonymous DAs creates rather intriguing consequences . as can be seenin (30) : (30) a. all of which are related to our central question: .exceptfor the comparative. Identification of basic patterns. * The boat is ten feet short and three feet narrow. The boat is short. DAs come in antonymous pairs like long and short. . Thus the sentencesin (29) state that the maximal dimension of the boat is above or below a certain norm or average .How Much SpaceGets into Language ? Of particular interest is. This problem might be couched in terms of parametersallowing for a restricted number of options. tall . a DA picks out a particular. b. In other words. c. the interpretation of which combines conditions on shape and size . I will have a look at the reason ably well understood structure of dimensional adjectives (DAs .hencethe deviancy of (30b). short.

the elementsmaking up the SF of theseitems have an obvious interpretation in terms of the structural conditions provided by SR. marked by the grammatical feature Deg that selectsmeasurephrasesand other degree . Three points can be made on the basis of this fairly incomplete illustration . x and j are operators binding semantic variables ). the semantic form of dimensional adjectives . the entries (31) and (32) immediately account for the fact that long and short apply not only to spatial . the entry for leave to syntactic arguments where the optional degreecomplement is morphologically . The scalar value thus determined must amount to the sum or differenceof v and y . Third . indicating the norm or averageof the class C which x belongs to..58 ManfredBierwisch other phenomena discussedin Bierwisch ( 1989 ) can be accounted for . has a nontrivial compositional structure in the sense introduced in section 2. the following entries for long and short can be given: (31) jlongj Adj x (j ) [[QUANT [MAX x ]] = [ v + y]] I Deg (32) jshortj Adj x (j ) [[QUANT [MAX x ]] = [v .2.y]] I Deg As in ( 18 . long and short are identical except for the different complements functor + as opposed to . where the choice of the value for v is subject to rather general semanticconditions responsiblefor the phenomenaillustrated by (29) and (30) . which then is mapped onto an appropriate scaleby the operator QUANT . from which crucial aspectsof the linguistic behavior of these items can be derived. providing one type of BSTs. while in (30) v must be specifiedas the initial point 0 of the scale selectedby QUANT . The common functor MAX picks up the maximal dimension of the argument x . Using the notational conventions illustrated in ( 18 ). While x and yare bound to argument positions to be filled in by syntactic constituents the DA combineswith . Second . even though this interpretation is anything but trivial . v is left unspecifiedin the positive and made available for a syntactically explicit phrase by the comparative morpheme. One option for the choice of the variable v is Nc. Semantically. Especially the way in which MAX and other dimensional operators like VERT or SEC for the vertical or secondary dimension of x are to be interpreted follows intricate conditions spelledout in detail in Lang ( 1989 ) . if DAs are assumedto involve three elements : ( 1) an object x evaluated with respect to a spe cified dimension. (2) a value v to be compared with . First . It accounts for the so-called contrastive reading that shows up in (29). and (3) adifferencey by which x either exceeds or falls short of v.

Person point . and tense are obvious casesin . syntactic computation of English and many other languages concord. their content being restricted to their role within the . The details of inflection .2. type 3 as indicated in section 2. Before taking up this problem with respectto spatial properties. is only indirect and does . there cannot be any doubt computational systemof I -language do tense or that . It is equally clear there must be some kind of an operator in SF related to [ + Plural] that imposesa condition on individual variables turning their interpretation into a multiplicity of individuals . person have semanticpurport in someway. The problem to be consideredbriefly in this section concernsthe relation between elementsof the morphosyntactic structure of I -language and spatial interpretation . gender conceptual interpretation . such as the pluralia tantum thesetwo aspectsbecomes . for example not dependon their spatial interpretation as such. morphological and syntactic primes. looking at further conditions for basic patterns and their variation . On the other hand. for example . there are categoriesof I -language that clearly enter . due to the projection of spatial conditions to other domains in the sensediscussedabove. The relation between clear in casesof conflict . a long interval. [ :t Plural] is clearly a feature that enters the morpho number as a paradigm case . As part of the interface. it is clear enough that theseare strictly formal conditions or operations. strictly morphological and syntactic relations and operations such as agreement of to conditions related that are but concord. I will place this initial specifiedas a consequence illustration of BSTs in a wider perspectivein the appendix. obviously . such as a long trip . their impact on the computational conceptual interpretation of linguistic expressions structure of I language .7 Grammaticalizatio The elementsand configurations consideredthus far are supposedto be part of the . As rationale for this question. they determine directly the semantic form of I -language .How Much SpaceGets into ! Ian"guage entities in the narrower sensebut to all elementsfor which a maximal dimension is defined. differ from phonetic featuresand semanticcomponentsby the lack of any extralinguistic interpretation . although the details once again need not concern us. a short visit. and categorial selection . The problem to be clarified is the need to reconcile two apparently incompatible claims. number. I will briefly consider the status of grammatical categorieswith semanticimpact more generally. and agreementthat depend on this feature need not concern us here. and so on. The way out of this apparent dilemma can be seen by looking more closely at . Note that the choice of the scaleand its units determined by QUANT must be appropriately of the interpretation of MAX . Dof Space 2. via argument positions. On the one hand.

the feature [ + Plural] of glasses cannot be responsiblefor the set reference in (33a). His glasses " " Obviously. [ . * Who does not talk to each other? (Eve and Paul. a configuration of primes of SF. Their glasses . Paul. where " glasses refers to a set of objects in ( 33a (33b) : were collected by the waiter. Thus the suffix / . More specifically. which relates morphological categoriesto configurations in PF. with someelementin SF. which resembles respectsthe phonological realization of [ :t: Plural] and other morphological categories . however. where who must allow for set interpretation . Who was invited? (Eve. as " each other" : antecedent the required by plural provide (34) a.) b. the presenceor absenceof the semanticset operator.s/ is the default realization of [ + Plural] for English Nouns. Notice . were sitting on his nose b. temporal reference More generally. I am not going to make serious proposals as to the formal nature of R at the moment. albeit more complex. as it must be lacking in (33b) . such as in [ + Plural].) Further types of dissociation betweenmorphological number and semantic individual / set interpretation could easily be added. for morphological categories .Plural] to the lack of this operator. but to a single object in of (33). The simplest assumption would be to associatea morphological category. if there is a morphological category M to which Cisrelated by certain rules or conditions R.that is. just as different from [ + Plural] as SET is. and Max were invited. but not identical to . Similar. The feature [ :t: Plural] is related to . whereas a consequence . such as SET. The potential suppression specificallymarked cases of the autonomous character of the morphological category. in some How this relation is to be captured is a nontrivial problem. but does not 34 ( ). but is. [ + Plural] in the default causeis related to the operator SET. The conclusion to be drawn from these observations is obvious. the following terminological convention to be useful: seems A semanticcondition . in a way that will be suspended be then would the association of .is grammaticalized. or tenseand its relation to . of course. that both the phonological realization and the semanticinterpretation of the default case might be instrumental in fixing the morphological category in acquisition as well as . Another type of conflict is illustrated by " " shown by (34a ). The conditions R should be considered as the semantic counterpart to inflectional morphology . accounts might be given for categories in languagechange its relation to sex and animateness and like gender .Manfred Bierwisch " ). (33) a.

I will briefly look at the grammaticalization of spatial componentsin the sense specifiedin the above convention. ahaz . but rather straight and simple realization of spatial information by thesecategories a more or lesssystematicmapping. even though it is committed to a different theoretical . Budapest ' in ' Budapest -re c. and (2) classifier systems . In agglutinative languageslike Hungarian.ban in the house . where classifiers are obligatory with numerals for syntactic reasons . seeHjelmslev ( 1935 Classifier systemsare subject to similar variations with respect to differentiation and grammatical systematization . A characteristic example is Chinese . whose transparencywill vary. We must expect in general not a . thin ob_ yi tiao lie one CL street ' one street' liang tiao he two CL river ' two rivers' . Budapest ' ' to Budapest Even though things are far less transparent in more elaborate systems . For an extensivestudy of complex casesystems(including Lak and Tabassarian ) that is relevant under this perspective . framework . it is sufficiently clear that place information can be grammaticalizedby inflectional categories . sufficiently rich distinctions of so-called notional cases . Instead of pursuing these speculations . part 1) . That notional casesare related to spatial information about location is uncontroversial and has beenthe motivation for the localistic theory of casementioned earlier. there is no clear boundary separating . Two candidates are of primary interest in this respect : ( I ) casesystemsincluding . The semantic information related to locational and directional postpositions from cases caseslargely matchesthe schemaof the corresponding prepositions discussed in the appendix. as shown in simple caseslike (35) : ' ' (35) a. respectively corresponding to location and shape . liDO(longish. and related to shape in caseslike (36) : iects) (36) a. the housein -ben b. depending on how entrenchedthe morphological categoriesare in autonomous computational relations like concord and agreement .37.? HowMuchSpaceGets into Language its actual realization indicates the conceptual purport of the formal category inquestion .

To cover of case . in Langacker ( 1987 case the. Spatial structure is thereby turned into a completely generalsystemof formal distinctions that makesthe explanation either vacuous or circular. if any. for Tzeltal positional adjectivesdiscussedin the appendix. in Jakobson is less in Hjelmslev ( 1935 37). for dimensional adjectivesof English or German. The most orientation and coherence along . but also for positional verbs like lie.4. the way in which caseis . more general considerations concerning the range and limits of these phenomena . in fact.Bierwisch Manfred b.lo cationist to I will restrict . contact. in to considerations myself syntax general spatial structural to the related of the properties varying range phenomena theory. The first position takes There are. a slightly rigorous proposal developed ( 1936 ) . categories I will conclude thesesketchy remarks on the grammaticalization of spacewith two . zhang(planar objects) liang zhang xiangpian two CL photograph ' ' two photographs san zhang zhuozi table three CL ' three tables' c. A tradition directly relevant is the locationist theory of case . two opposite positions in this respect the as structure computational structure of I -language immediately supporting spatial and the categoriesof syntax and morphology . casesare to be explained in terms of spatial concepts like distance lines is given these account ambitious . Even more crucially . Even though the details need clarification . which must be available. or stand. more recent proposals of ) extend so-called cognitive grammar as put forward . according to which not only notional but also structural . kuai (three-dimensional objects) yi kuai zhuan one CL brick ' one brick ' san kuai feizao three CL soap ' ' three cakesof soap The SF conditions to which theseclassifiersare related are not particular 3-D models but rather abstract object schemata of the sort mentioned above. sit. it should be obvious that shape information can correspond to grammatical . albeit in different modes of specification. While thesetheories are concernedwith caseonly. for example . an extremely abstract construal of spacemust be assumedthat has little . among others. connection to spatial cognition as sketched in section 2.

shape may . and morphological categoriesin general . definiteness discussed . spatial distinctions as representedin SF can correspond to elements of grammatical form . Jackendoff argues that axial systemsand the pertinent frames of encoding of space referenceare representedin spatial representation but generally not in conceptual structure. furthermore . That comparable effects are missing for practically all spatial distinctions. which has obvious consequences tactic categories in English. It is basedon the following consideration. is a purely directional in. in de. even though it is directional. The conclusion to be drawn here has already been stated. but are clearly to be distinguished from them. it does not by itself expresslocation or direction. and sizemay correspond . A clear indication for the conceptual encoding of a given distinction is the effect it has on grammatical structure. gender. Whether and which spatial distinctions are taken up explicitly by elementsof semantic form and whether these correspond.? How Much SpaceGets into Language related to spatial conditions is notoriously opaque and indirect. I agree with representedin conceptual structure. is advocated by Jackendoff (chapter I . The claim. provisos just . gender. but that do not in general representthose distinctions directly . and person. Comparing two options with regard to the . is a matter of languageparticular variation . tense . respectively . at least in English. as should be expected . to effects in computational aspectsof I -language . English keepsmost of them . tense are elements of the computational structure that may correspond to conceptual distinctions. the dative/ accusativecontrast of /e (into the /e (in the school) versus in die Schu German for example . and so on. On the other hand. which is in a way the opposite of the first one. Jackendoff for morphosynnotes the count-mass distinction . nor that they are excludedfrom conceptual . Schu form of locative and connected to the semantic formal condition school). Cases . the mass distinction Given or the count . In other words. as mentioned above (seeappendix for illustration ) . applies to spatial structure in general. As a casein point . But this does not mean that they are excluded within the limits of lexical semantics from grammatical effectsin other languages . is then taken as an indication that they are not . presumably. and semanticrepresentationsof English expressions . In many languages caseis involved in the distinction betweenplace and direction . like number. location may correspond to notional case correspond to classifiers to degreeand constructions like comparative. this volume) . equative. number. This is " borne out by the fact that " zur Schule (to the school) requires the dative. but only in spatial representation of the pertinent Jackendoff in assumingthat grammatical effectsindicate the presence distinctions in conceptual structure. But it seemsto me that the conclusion is the opposite becausethe major spatial patterns are no less accessiblefor grammatical effectsthan conceptual distinctions related to person. The secondposition .

3. where strictly spatial conceptsare to be identified as configurations that interpret elementsof SF by exclusively spatial conditions on objects and situations. the axesand frames of referenceof their location . Conceptual representation of spatial structure provides. are neverthelesscomponents of the formal structure of I language . size . The distinction between these two types of categories varies for obvious reasons . based 5. Thus tense . Spatial information " visible" in I -language is thus restricted to strictly spatial . which map semanticform onto phonetic form . and number are usually more transparent than (abstract) case or infinite categoriesof verbs. recruiting configurations of it by basic components of semantic form . among other things. 4. and location of objects and the situations in which they are involved will be combined with other aspectsof commonsenseknowledge. 2. seemto fall into two types: syntactic categories . however transparent their correspondencemight be to elements of the interfaces of I -language with other mental systems .Manfred Bierwiscb 2. Categoriesof the combinatorial system . and morphological categories computational conditions of I -language in SF to in or less more (or PF configurations transparent ways may correspond for that matter) . Spatial cognition or I -spacecan be considered a representational domain within the overall systemof C-I of conceptual and intentional structure integrating various perceptual and motoric modalities. depending on the systematicity of the correspondencein question. which . Linguistic knowledge or I -languageinterfaces with conceptual structure. more abstract schemataspecifying the dimensionality of objects and situations. person. Representationsof I -spacemust be integrated into propositional representations of conceptual structure. all other spatial information being supplied conceptsand their combinatorial effects by representations of ~ -I and the commonsense knowledge on which they are . which serve the exclusively . With all the provisos required by the wide range of unsolved or even untouched problems. where in particular shape . the question raised initially might be answeredas follows: I -spaceis accommodatedby semanticform in terms of primitives interpreted by .8 Conclusion The overall view of how language accommodatesspace that emergesfrom these considerationsmight be summarizedas follows: I . and metrical scales with respectto which sizeis determined. strictly spatial concepts . The computational categoriesof I -language .

realized thedefiniteness ( ) yields ) ( Combining by operator : 38 saturated 37 is of in (39 by ( ) position ( ) ). I will consider a general schema that covers a wide range of phenomenashowing up within the systemof locative propositions. + Obj. the lexical entry for the follows: as be stated in can preposition (37) / in / [ . . Locative Prepositiol W To begin with . we assumea compositional structure. c. respec x entities two relates in general y . the manin the garden . In other words. . By meansof the notational conventions introduced in ( 18 ) and (31) above. . respectively . or by .] ) fin the gardenl [ pP (39 eitherby the head The remainingargument positioni. SF(le) of in (andin fact of prepositions andthe relatum and . asin (40a ) and (40b PPaspredicate . . The manis in the garden . which specifiesthe interior of its argument. . .] . b. wherethe objectargument i.V .N . identifyingthe theme ) that is to be checked by a complement specified tively. now motivate by a number of comments Icture Intuitively Stn a IM I Argument Variables . . based on Bierwisch ( 1988 According to this analysis . of this PP is to be saturated ofa the modifiedby thePP subject copulathat takesthe ). The relatumy is syntactically : a complement of that(38 . Suppose case for objective representation such ) is a simplified DEF Ui[GARDEN ] Ui [DP. asin (4Oc ): (40) a. including the relation LOC semanticform of in is composedof a number of elements and the functor INT . .i (j ) [x [LOC (INT y])] I [ + Obj] ). [DEF Ui ] ]]]] ] : [x [LOC [INT Ui . instead of a simple relation IN . a two dimensional . . whose conceptual GARDEN abbreviates . Themanis waitingin thegarden .~ How Much SpaceGets into Language Appendix In what follows .6 by looking somewhat more closely at locative prepositions and dimensional adjectives . . I will illustrate the types of questions that arise with respect to the program sketched in section 2. the ) and Wunderlich ( 1991 . amongother things thePP with 38 37 the .] I ) Ithe garden (38 interpretation of the noungarden the SF constants . relating to place and shape. which I will . DEF indicates includes objectschema .

indicated by bracketing y in (37) . Knowledge about fishes . Both " water" and " mouth " are associatedwith a three-dimensional object schemain (43a) and (44a) but conceptualizedas belonging to a two-dimensional surfacein (43b) and (44b) . SemanticPrimes The variablesx and y in (37) are related by the constants LOC and INT . The interpretation of in can thus be stated more preciselyas follows: (42) a. First . There is a hole in the purse. b. . There are somecoins in the purse. (41) He is not in today . I will take up the consequences of this point shortly .Manfred Bierwisch The main point to be noted here is the way in which the saturation of argument positions imposes conditions on the variables provided by the lexical SF(/e) of in. where y is left as a free variable in SF(/e) and will be specified by default conditions applying in C-I without conditions from SF. The boat is in the water. Relevantconditions include in particular the dimensionality of the object schemaconceptually imposed on x and y . the bINTy interior of y Three commentsare to be made with respectto this analysis. b. x LO Cp identifies the condition that the location of x be (improperly) included in p identifies a location determined by the boundaries of y. and pipes supports the different construal of both INT and LOC . additional conditions applying to x and y will affect how LOC and INT are interpreted in C-I . The fish is in the water. boats. Somewhatdifferent factors apply to the following cases : (45) a. Thus the actual location of the theme in (43b) would rather be expressed by underif it were identical to that in (43a) : (43) a. b. A similar casein point is the following contrast: (44) a. A final remark on the argument positions of in concerns the optionality of its object. Both are explicitly spatial in the sense that they identify conceptualcomponents that representsimple (possibly primitive ) spatial conditions. It accounts for the intransitive use in cases like (41). alongside with further conceptual knowledge. He has a strawberry in his mouth. He has a pipe in his mouth. fruits . that is.

in is a BST only if it relatesto I -space not if it relates (in equally literal fashion) to time or institutions . contrast INT with a prime ON with roughly the following property : ON y identifies a location that has direct contact with (the designatedside of ). Examples like those in (48) indicate. b. although the way in which conceptual knowledge creates the differences in question is by no means a . The different interpretations illustrated by (42). while in (43) and (44) the inclusion determined by LOC differs accordingly. I will thus continue to use BST without additional comment. that the notion of BST crucially dependson how implicit transfer of spatial structures . Thus from (45a) and (46) the conclusion (47a) derives into inferences not follow from (45b) and (46) : . c. to which further variants could easily be added. mouth. but is construed. in (45b) the conditions . but I do not think that this terminological issuecreatesseriousproblems. There is a hole in my briefcase I do not think that water. Notice that in (45) it is only the coming from hole enforce the substanceschema interpretation of INT that varies. the specification of the theme and/ or the relatum provides the conditions on which LOC and INT are interpreted. purse are lexically ambiguous. It seemsto me an important observation that in under this construal of BST is not an exclusivelyspatial term. Again .? Getsinto Language How Much Space 67 In (45a) purse relies on the object schema of a container. but (47b) does . but does not intersect with . Thus English and German. And third . there is no reason to assumethat trivial issue . The differencesresulting from theme or relatum may enter . the conditions identified by LOC and INT are subject to implicit transfer : to domains other than I -space (48) a. . There are somecoins in my briefcase . for example . In one possibleinterpretation . He came in November. are due to conditions of I -spaceand conceptual knowledge not reflectedin the lexical SF(/e) of in. Second . however. it must be left aside here. severalstepsin the calculation . He lost his position in the bank. y . (47) a. The argument applies only in this case dreadings in linguistics e. the range of I -spaceconditions identified by INT dependson the distinctions a given language happens to representexplicitly in SF by distinct primes. In any case in is ambiguous between(37) and someother lexical SF(/e) .(47). (46) The purse is in my briefcase . b.

i [[x LOC [PROX yll : [x PARALLEL (50) jalongj [ . The ball Der Ball rolite unter den Tisch. and around are more complex. . Thus Wunderlich ( 1991 ) claims that . in order to keep to the limits of BST. They were in the school. where F is a variable ranging over functors that specify locations determined by y : (49) [x LOC [Fy]] Not only do in and on fit into (49). What is nevertheless of interest is the systematicityof variation theselexical exhibit. If this is correct. explicit spatial interpretation . under at over and several .assumingthat SF(Ie) of on is [x LOC [ONy ]]. . . using pertinent replace F. all of which must Cmight be a configuration of basic elements have a direct. specifyingFby INT and ON . across . the nail in the table and the nail on the table. . Hence the variation in patterns of according to general principles of I -language lexical representationsI will briefly look at are fully detennined by the basic elements involved. They went into the school.N . the generalschemaof locative prepositions is of y. The ball rolled under the table. for example . respectively (51) instead of (49) : (51) [[x LOC [Fyll : [xCyll where C is a condition on x and y .] [MAX y]]] PROXy and MAX y detennine the proximal environment and the maximal extension . but also constants to other near. whether schema(49) covers the full range of conditions that locative prepositions can impose. . for example . . as illustrated in (62) : (y) . Sle gingen in die Schule Sle waren in der Schule was under the table. along. as shown for English and German examplesin (52) : (52) a. however. .V . Another systematicaspectof locative prepositions concernstheir relation to directional counterparts. introducing an additional condition . Der Ball war unter DernTisch. as exemplified in (50). b. . .Manfred Bierwisch This yields the different interpretations of . representations The first point to be noted is the obvious generalization about locative prepositions . respectively . such that the surface of the table could provide the location identified by Spanish INT .whereasin Spanish el clavo en la mesawould apply to both casesbecausethere is no in/ on contrast in . prepositions. all of which instantiate schema(49). It is not obvious. . The Pattern of Locative PrepositiO18 I have assumedthroughout that the categorization inherent in the primes of SF determines the compositional structure of SF .

] identifies a transition whose final state is specifiedby [ . LOC and F of (49) are realized by separateitems with roughly the entries in (56). ] The relevant observation in the presentcontext is the systematicstatus of CHANGE in lexical structure. the way components of basic schema(49) are realized by separateformatives. . the generalpatterns of prepositions have beenconsidered as the frame by which lexical knowledge of a given language is organized.to in onto.ui . The general schemaof a standard directional preposition would then be (53) : (53) CHANGE [[x LOC [Fy]] : [xCy]] where CHANGE [ . behind can be used as locative or directional prepositions. notational of the to be checkedby the object preposition Using in phonology. yielding (57) : . . . or it is locative . in Korean . and Latin is CHANGE with appropriate morphological case . In other words. the relation in question can be expressedas in (54) for German in: (54) lint Dir ] [ . assigns + oblique case and does not contain CHANGE . . which identifies the top or surface of its argument and provides the complement of the locative element e.ta . the directional preposition identifies a path whose end is specified by Semantically the corresponding locative preposition.N .(ui ). the occurrence of CHANGE is connected to . largely related to accusative introduced devices . which . that is. over. as can be seen in (55) . one of which concerns " what might be called " lexical packaging. Typological Variation Thus far . into.i [ < CHANGE ) [x LOC [INT yll ] I Obl ] [ .oblique case and contains the CHANGE component . Crosslinguistic comparison revealsvariations of a different sort. In languages like Russian .? How Much SpaceGets into Language . IX y .' ' The relatum ch aeksang(optionally marked for genitive) functions as complement of the noun ui. taken from Wunderlich ( 1991 ): ' .IX This means that in is either directional . for example . Let CHANGE p be an operator that turns the proposition p into the terminal state of changeor path . assigns .e iss kkotpyong i (55) Ch aeksang there Pres be Nom Gen top Loc vase desk ' There is a vaseon the desk. . . A straightforward alternative is found .V . German. Besidesmere optionality in caseslike under .

Nujul boch ta te k ib -down gourd -bowl Loc the water -jar upside 'The -jar . .Manfred Bierwisch (56 ) a. stant . lei [ . that indicatemainly positionaland shape adjectives like sit.V ] y i [z LOC [ ENVy]] ENV abbreviates an indicationof any (proximal . to yield(58a applies ): ' ' x [[UPRIGHT CYLINDRIC x] : (60 ) waxalta ch uj te [ + N . ltal [ . but ratherin termsof positional .] I [L] ' -OF [ DESK i [z LOC [ TOP (57 ) ch aeksang (ui)wui-e [ .' down on the water gourdis upside Waxalandnujulbelongto about250positionals . L] -OF x] X [ TOP I ) (Gen N i [zLOC [ N ]] b. + V .] PLANK ]]]] [x LOC [ENV [ WOOD . Waxal ta ch uj te -jar . Loc wood the water upright plank ' Thewater is ' on the jar standing plank.V . . includingthe featureL of the noun wui. derivingfrom some70 roots representing andpositionalcharacteristics Brown 1994 for discussion shape (see ). additionalspecification doesnot come . . .] ]]] The details ad hoc . completelyunspecific general locativeparticle as ta.V.N . 51 form : Levinson 1990 ( ) givesexamples specifications ( ) ' ' ' te k ib (58 ) a. . .somewhat information .] b. lie in English ably more differentiated variety of . A highly provisionalindicationof waxaland the only locativeprepositionta would look like (59 ): x [ UPRIGHTCYLINDRIC x] ) a. . The PP ta ch'uj te' in ) environment asan adjunctwith the predicate waxalas shownin (60 ) combines ). however . . . are somewhat . by nominaltermsidentifyingpartsor aspects of the relatum .N. Like Korean Tzeltal has a . Iwaxall (59 [ + N . which then (77a to the NP te k 'ib. . . . but the main point shouldbe clearenough : e and wui combineto createa structurethat is closelyrelatedto the SF of Englishon or Germanauf A differenttype of packaging for locativeconstructions is found in Tzeltal and other Mayan languages . . but with a remark !. realized . ' b. . Iwuil [ + N . . . . + V . .N .

an realized in ( 63 ) : by appropriate measure phrases .seems to be perspicuous . although already Variables that an and Argument adjectives ) x to Structure in English optional As already mentioned dimensional object ( or event are syntactically complement of the . than v under Bill feet too long for this garage .place morphology : relations or the prepositions construction becomes the variable visible comparative accessible syntactic specification ( 64) a. Dime _ Here Based onal Adjectives briefly add some points to the analysis of DAs sketched in section 2 . ( 61 ) express es the fact two . minutes ( 63) a. I will not go into further typo logical variations related to the way in which general principles of semantic form accommodate locational information in basic spatial terms of different languages. The c . but rather will take a look at issues that arise with respect to terms encoding aspects of explicit shape information .6 .the type of packaging of SF material . to prepositions . variable particular conditions . symphony A particular and the point that distinguish that es DAs from to locative Ps concerns earlier the variable to this v particular . His field speech is 60 yards was only long fifteen and 30 yards long wide . b . The car is two In a way . The stick is long is enough twice as to touch long as the ceiling the sonata . . The car is just as long as the garage . that explicitly specify the and for this garage syntactic are complements .long desk b . John is two feet taller than Bill . relating DA that specifies adegreey . the relevant issue.place predicates . . Due relations when to two . The c . b . as in ( 62 ) . a six . as mentioned . This make apply three . of / ong given I will on the analysis in ( 31 ) and repeated here as ( 61 ) : ( 61) /long / Adj x ( j ) [ [QUANT [MAX x ]] = [ v + y]] I Deg I will some keep of the to the same sort points have of comments been taken given up with above respect .foot . DAs conditions variable like too are semantically .place in fact it . or more complex expressions as ( 62) a. rather than .How Much SpaceGets into Language ? Although various details are in need of clarification .

The Pattern of DimensionalAdjectives The characteristicproperties of D As show up more clearly if we look at the general schemaof their SF. heavyand light are gravitational . = . MAX . while QUANT . of which only MAX has a specifically spatial interpretation . and = and + have the usual arithmetical interpretation with respect to scalar values . the shapeand the size information in contained long and short are defined by MAX . An apparently simple modification is shown by languageslike Russian. Where long/ short pick out the maximal dimension. measure phrasescan only be combined with the respectivenouns. and + identify quasi-arithmetical operations underlying quantitative . As ball would not allow sculpture for a schemaof this sort. 10 m long could not come out 10 m dlinnij . I will now indicate someof the possibilities to modify the schemaitself in various ways.SemanticPrimes The variables x . = . and v are related in (61) by meansof the four constants QUANT . and by QUANT . and + or . which do not allow measurephraseswith DAs . identifying the maximal dimension with respect to the shape of y . by constructions like . and so forth . In other words. Hence semantically. scalarevaluations quite generally. which marks the position for different dimensional components . [QUANT Y ] is a function that maps arbitrary dimensions Y on an appropriate abstract scale . on the one hand. More specifically . Also . like fast and slow. Typological Variation Thus far . = . we have consideredvariation within schema(65) . It might be noted that the interpretation of the different dimensional constants requires the projection of an appropriate object schemaon the term providing the value for x : a tall sculptureinducesa schemawhosemaximal dimension is vertical for . high/ low pick out the actually vertical axis by means of VERT .. that is. the quantitative conditions may carry over to various other domains: old and young are strictly temporal . For details of this mechanismseeLang ( 1989 ). a tall ball is deviant. a term qualifying movement. As a matter of fact. and so forth . and + . y. the constants replacing the variable DIM in (65) turn an adjective into a spatial term like tall or thin. shapeand sizeare interlocked in ways that differ remarkably from their interpretation in SR. which automatically accounts for the fact that they usually come in antonymous pairs as already noted: (65) [[QUANT [DIM y]] = [v :t x ]] The secondpoint of variability in (65) besidesthe :t alternation is indicated by DIM . which does not provide this condition by itself. on the other. long is a spatial term only insofar as MAX determinesdimensional conditions that rely on shapeand size of objects or events . and tall combines both MAX and VERT . a temporal term like young or late. .

strictly spatial specificationsof objects to which they apply . Things seemto be a bit more complicated. First . dlinnej. . there is remarkably systematicvariation among different languageswith respectto both the choice of basic distinctions recruited for lexicalization and the different types of packaging according to more . elementsplaya theoretically relevant role for the linguistic representationof space with respectto the linguistic properties Second . the analysis of basic spatial terms. though: measure with comparativesare possible . Chinese . among the entries of the core lexical system of I -language . although only in termsof prepositional phrases phrases translates into the adjectival construction na 2 m with na. promises to give us a more detailed much into . In general could be illustrated only by two types of cases . Russian. Their subsystemof items that are strictly spatial in the sense semanticform [SF(/e)] consistsexclusivelyof primes that are explicitly interpreted in . Where the twenty-odd DAs of most Indo -European languages rely on object schemata in a rather abstract and indirect way. for example . 2 m longer. Finally . Here. exemplifiedby . a similar preferencedistinguishes German and Russian. not only the degreeargument position is dropped. Tzeltal positional adjectives but the whole quantificational component. This suggeststhat Russian DAs do not have a syntactic argument position for degreecomplements . as indicated provisionally in (59a ) . their basic elementsbeing components of a representational aspectdetermined by VG . preserving otherwise schema84. and Korean have been isolated in Lang ( 1995 ). it attests a different strategy to recruit conditions on shape and position of objects. while Object schematain Chinese seemto be based on proportion of dimensions Korean takes observer orientation as prominent . Even though the delimitation of this subsystemis terms of conditions of I -space subject to intervening factors. We have already seena much more radical variation of schema(65). there is a illustrated in section 2. This is not merely a matter of quantity . and argument structure of theseterms must be structure of the SF more generally. rather subtle. Although organizing principles and actual details of Tzeltal positional adjectives remain to be explored. corresponding to length ofmeters . there are characteristic consequences of theseitems. I cannot go into the details of this matter. the positional adjectivesof Tzeltal include fairly specific . as shown by the appearanceof degreephrases . even though it general patterns. Hence the compositional assumedto belong to I language . its .5. Let me summarizethe main points of this rather provisional sketch of basic spatial terms. of how ) spacegets ( language understanding . suchas implicit or explicit transfer of interpretation .? Getsinto Language How Much Space 73 dlinna 10 metrov. retaining only [ DIM x ]. then. but supplying it with a much more detailed system of specifications . but clear distinctions determining alternativesin DAs of German. rather.

Akademie Bierwisch mantis cheund konzeptuelle lexikalischer . Alban Berg in his Lyrische Suite introduces a theme that consists of the notes es ( = e ftat)-c h ( = b)-e-g. and Kay . RuzickaandW. Although it is still compatible with the possibility of parametric variation regarding the way options provided by specification 2 are exploited in individual languages .i.components . Thus. A very special " interface representation systemof numbering ' s famous used in G Odel of the proof incompletenessof arithmetic . This doesnot necessarily . and Dieter Wunderlich. Notes I . even if I do not agreewith him in certain respects . 115 PsychologicalReview .6.Manfred Bierwisch Acknowledgments The presentchapter benefitsfrom discussionsat various occasions .. B. I will examine more concrete possibilities along these lines in section 2. in order to honor Schonberg . Berkeley Biedennann. ) . one stating properties of the other. 2. ( 1987 : A theory of human image understanding. ) . Besidesthe membersof the Max Planck ResearchGroup on Structural Grammar . " " 5. although longing properly language recognizesthe need for correspondencerules connecting it to articulation and perception. . It should be noted that Jackendoff considers the structure (. P. I am indebted to the participants of the project on Spatial and Temporal Referenceat the Max Planck Institute for Psycholinguistics . " " 3. Ewald Lang. Basic color terms.by. Se Reprasentation In R. PF) phonological .). References Berlin. StephenLevinson. it should be clear that they must be construed as computational systemswith their own specific representationalproperties. One might in fact consider LF a systematiccategorization imposed on SF. this sort of parametric variation should be considered as bound to lexical information . . whose stimulating proposals are visible throughout the paper. Even though Chomsky ( 1993 ) refersto APand C-I occasionallyas perfonnance systems . M.e. I . stipulating LF in imply a proliferation of levelsof representations addition to SF. Recognition. and thus ultimately to the choice of primitives in the senseof specification I . 94. representing all and only the letters in Schonbergcorresponding to the German rendering of notes. Particular debts are due to Ray Jackendoff. as beto I he . . ( 1983 ). where numbers are given two mutually exclusivesystematicinterpretations.147 . further discussions included Dieter Gasde . Motsch(cds .. including the minimalist program proposed in Chomsky ( 1993 ) . just as PF must be subject to certain aspectsof syntactic structure. This view is in line with fundamental developmentsin recent linguistic theory. " in the intended senseis the 6. ( 1969 : University of Call fomi a Press . 4. Untersuchungen zur Semantik : StudioGrammatico XXlI .100 -Verlag 61 . Berlin . Paul Kiparsky . Einheiten .

( 1936 ). Consciousness . W. I . J. ( 1976 ). Bierwisch. . In R. The semanticsof gradation . Journalof Memoryand Laird Johnson and R. ( 1993 ).). Ontologicaldomains Dolling. R. 785 .) . N. Byrne . Lang (Eds. Keyser .further and further. .muchdeeper Bierwisch . On argument . J. ( 1990 ). ( 1988 ) .52. K. ( 1994 ). Dordrecht andbinding ongovernment . Russian : Gesamtbe Casuslehre : Beitragzur allgemeinen Berlin . Cambridge . Heidelberg . Semantics . und Lexikon: Rudolf Ruzicka zum 65. NewYork: Mouton. International sorts . . E. M . 471 interpretation of static : The semantics Brown .). : 1931 and Slavicgrammarstudies . adjectives . Arhus: Universitetsforlaget descas . Lectures Chomsky . Motsch.65. ( 1995 . Berlin: Akademie-Verlag.. A. . .).103 . ( 1980 ). Syntax. Spatialreasoning . . J. Cambridge structures Jackendoff . and systematic ambiguity . origin . Essays .). anduse . . Studies and the lexicalexpression structure Hale ofsyntac . MA: MIT Press Jackendoff . ( 1983 ). Amsterdam in lexicalrelations Gruber . 11 Writings deutungen . N.1981 . NewYork: Praeger : Its nature . Semantik. Bierwisch. La categorie Hjelmslev . R. ( 1975 of thought ). Essays from Building20.37 . M. Contribution to the generaltheory of case . Dordrecht andMontague grammar ). ( 1993 ). and Lang ). ( 1935 ).71. S. M . R. MA: MIT Press . A minimalistprogramfor linguistictheory Chomsky : Theview in honorof SyvianBromberger in linguistics . Zimmennann (Eds. . 743 790 32 . Themodularity . P. ( 1989 ). Rules Chomsky : Foris. J. Studies Journalof Human . MA: MIT Press Fodor. 23. Bierwisch and E. Thelanguage .? How Much SpaceGets into Language Bierwisch. In Hale and Keyser(Eds tic relations ~ 109 . Selected Kasus der russischen . . ( 1986 of language ). Cambridge mind andthecomputational . MA: MIT Press . 43. NewYork: Springer .514 . . L. 59 Russiancases . On the grammar ofloca1 prepositions. ( 1989 longer . Knowledge Chomsky . NewYork: ColumbiaUniversityPress andrepresentations . ( 1981 ). D. : Reidel .. Cambridge andcognition Jackendoff . Linguistics descriptions . Somewhat andconceptual structure : Grammatical . N. Geburtstag 1. MA : MIT Press . ( 1987 ). NewYork: Springer . . Hale and S. . In M . Vol. 53 from Building . . : North Holland . MA: MIT Press Cambridge . ( 1979 . Language .261. ( 1983 of mind ). semantic ). . NewYork: Cromwell Fodor. Dimensional and Lang (Eds In Bierwisch adjectives . In M . In K. A. . . P. Heidelberg. andKeyser : in honorof SylvianBromberger in linguistics . S. N. and . R. (Eds . .807 -Computer . ( 1989 ) . M. location of . Dimensional : Grammatical structure and conceptual interpretation.). I . R. J. Cambridge 20 Theview . (Originalversion . Wordmeaning Dowty. 575 28 564 . 71. N. J. Semantic of the : Generalmeanings Jakobson . The INS and ONS of Tzeltallocativeexpressions . Jakobson . J.

Dordrecht : Kluwer. J. D. ). C. 114 . How do prepositionalphrases . Mentalmodels of language . 81 of language papers from 1991Konstanz Conference Amsterdam : Benjarnins . Lexicaldecomposition . 5.. ). 66 . NewYork: Norton. ( 1983 Kosslyn ). 263 .. J. 53 ). ( 1991 ).24. A. 1. ( 1981 . . inference . . Language .117 in the organization : Selected . .441 .. In M. Neisser ). FAS in Linguistics . . NewYork: Springer : A first look at universalfeatures and typological .417 . R. M. J. . ( 1990 ). How do wordsgettheirmeanings . E. and the Lexicon . In U. Nounsandverbs Langacker Levinson . Nijmegen : Max PlanckInstitutefor Psycholinguistics . W. Holtzmann . Cambridge . Thegenerative Linguistics von Stechow in syntax . UniversityPress to logic . ). Basicdimensionterms Lang . S. MA : Harvard . In E. SanFrancisco : Freeman . Conceptual (Ed. S. J.. 63.-Laird. fit into compositionalsyntax and Wunderlich . ( 1972 . ( 1989 ).341 brain patients .100 -Papers variation . Vision Moravcsik ?Journalof Philosophy . 591 . Farah . (Eds . ( 1985 Kosslyn ). C. ( 1995 ). E. U. 175 development UniversityPress ' in theminds machine . . H. ( 1983 : Towards a cognitive Johnson science . ( 1991 . 17 . Bierwisch . . Ghosts . 311 Psychology of dimensional of spatialobjects . ( 1981 .). P. 78 . NewYork: Harperand Row. ( 1993 ). A computa tional analysis of mentalimagegeneration : Evidence from functionaldissociations in split . ( 1995 ). Semantic theory Keil. Heidelberg . J. Space . E. ). F. and Reyle .200 andconceptual . 409 Pustejovsky ).94. S. Dimensional : Grammatical structure andconceptual adjectives interpretation . Cambridge and consciousness : CambridgeUniversity Press . Urs et al. The semantics Lang designation and E.).). D. Cambridge : Cambridge . N. M. Fromdiscourse Kamp Katz. 29. Computational . November Marr. lexicon . J. and Gazzaniga . Figureandgroundin Mayanspatialdescription conference Time . ( 1987 and category structure . M. MS .621 ? Linguistics semantics . M. Journalof Experimental : General . . . ( 1987 . Concepts development . D. Thelexicon . . Paper delivered to the . Lang (Eds .

The message macroplanning kind . it has to be tuned to the target languageand to the momentary informational . . . chapter 1. we decide in some way or another on what to express intention choice of content will eventually make our communicative recognizable to our audience or interlocutor . Dutch . M. paramount importance for spatial discourse In an effort to cope with the alarming complexities of conceptual preparation. There is also microplanning. to construe distance from speakerin a different way than English use of here . The result of conceptual preparation is technically termed a message ). to someextent.Chapter 3 Taking and Ellipsis in Spatial Descriptions Perspective Willem J. this situation . have a bipartite system . In particular . presenteda figure in my book Speaking( 1989 must be in some kind of propositional It is intended to expressthe claim that messages or " algebraic" format (cf. eventually express language But there is more to conceptual preparation than considering what to say. namelyperspectivetaking. I ) that is reproduced here as figure 3. Ideally. there are languagessuch as Spanishor Japanese in a tripartite way: proximal -medial-distal. Depending on the communicative . they must be composed out of lexical concepts . Levelt 3. in that is . . This chapter is about an aspect of microplanning that is of needsof the addressee . Lexical conceptsdiffer from languageto language be specificto the target language will in and therefore another be nonlexical A lexical concept in one language may . it is the conceptual entity the speakerwill (or a string of messages formulate. this volume) to be suitable for formulation . such as English or of use distal. or has to be of a particular . ' . Other languages . An concepts for which there are words or morphemes in the speakers language immediatecorollary of this notion is that conceptualpreparation will . proximal aqui-ahi-alli requires Spanish -there.1. that is. Jackendoff.1 Thinkingfor Speaking There exists happy agreementamong students of languageproduction that speaking normally involves a stageof conceptual preparation. To give one spatial example (from to be expressed need a slightly different message that treat deictic proximity Levelt 1989 ).

. But to formulate " " . information . Levelt 1989 ). and generatepredications about theseentities that accurately capture their ." . and so on experiences . for instance . we attend to entities that are relevant to the communicative task at hand. in particular spatial. Focusing on some portion of the field with respectto which the referent' s spatial . I will call this portion the referent.sem rep I Cn ~ .." disposition is to be expressed 3. kinesthetic.essarilyrequires an act of abstraction. Although this term will in the presentchapter be restricted to its original spatial domain. This processof abstracting from the visual scenefor spatial relations within the scene " " I will call speaking perspectivetaking . Spatially relating the referent to the relatum (or expressingthe referent' s path or orientation ) in terms of what I will call a " perspectivesystem . 2. we cast them in propositional form . Thinking for speaking is always involved when we expressnonpropositional.. it is easily and fruitfully generalizedto other domains of discourse(cf. I will call this portion the " relatum. When talking about a visual scene .1 depicts the notion that when we talk about our spatial. Focusing on some portion of the scenewhose spatial disposition (place . 3.1 The mind harbors multiple representationalsystemsthat can mutually interact. musical. " " Slobin (1987 . " " orientation ) is to be expressed ( Talmy 1983 ) .2 Perspective Tsking Perspectivetaking as a processof abstracting spatial relations for expressionin language typically involves the following operations: I . path. M. which is an elegant ) has usefully called this thinking for speaking synonym for microplanning . Levelt Figure 3. This ne" . propositional any representation linguistically requires its translation into a semantic code (reproduced from Levelt 1989 ). 1 nA J Y ' \ D I ~ FO ( p re ) me Willem J. Figure 3.

This description is valid in the intrinsic perspectivesystem . Here the speakerintroduces the chair as the relatum and then express es the spatial disposition of the ball (to the right of the chair) . you can take another perspectivesystem ) . a speaker . The perspectivesystemin terms of which the relating is done is the deictic system . FigurE Let me exemplify this by meansof figure 3. Two things are worth noticing now. That is why the ball is to the right of the chair in this system . the ball is the referent. it is only a less preferred one. tend to select smaller and more Speakers foregrounded objects as referentsand larger or more backgroundedentities as relata. depending on the choice of referent. The ball in figure 3. Here they tend to follow the Gestalt organization of the scene( Levelt 1989 . A chair has a front and a back. Hence. Here the referent' s location ' is expressed in terms the relatum s intrinsic axes .2 is at the chair' s left side. . One way of describing this sceneis ( I ) I seea chair and a ball to the right of it . that is.centeredrelative system . no matter from which viewpoint the speakeris observing the scene . a left and a right side. ! When you focus on the relatum (the chair).Perspective Taking and Ellipsis in Spatial Descriptions 3.2 This spatial array can be described in myriad ways . This is an equally valid description of the scene .2. First . as in (2) : (2) I seea ball and a chair to the left of it . Second You can also describethe sceneas (3) : (3) I seea chair and a ball to its left. your gazemust turn to your right in order to focus on the referent (the ball ) . you can swap relatum and referent. Still another perspectivesystem allows for the description in (4) : . and perspective . relatum.

Brown and Levinson . . The implication of thesetwo observations is that perspectiveis linguistically free. Finally . Tzeltal usesa mix system system of absolute and intrinsic perspectives . In the strongest non-Whorfian . and the absolute systems . 1993 of Yimithirr are exclusive users of an absolute perspective ) Speakers Guugu are exclusive users of an intrinsic . This description is valid if indeed ball and chair are aligned on a north -south dimension ' .nd English usesall three systems there are personal style differencesbetween speakersof the same language . and these will affect a culture s or a speakers choice. In other words. as linguistic description of that scene Levinson and Brown have demonstrated ( Levinson 1992a . This freedom of perspectivetaking does not mean. The secondissuegoesback to figure 3. how hard or easythey are to align betweeninterlocutors . Mopan speakers . but rather to a fixed bearing." For instance . In particular . however. There is no biologically determined one-to-one mapping of spatial relations in a visual sceneto semantic relations in a . Each perspectivesystem has its specific advantages ' ' and disadvantagesin language use . it is neither relative to the speakers nor to the relatum' s coordinate system . This is termed an absolutesystem .A (or semantic ) representation for the purpose of expressingit in language crucially important question now is whether the spatial representationsthemselves are already " tuned to language . After all . There is no unique way of perspectivetaking . The first one is pragmatics. A speaker is not " at the mercy" of a spatial representation in thinking for speaking . And cultures have taken different options here. the intrinsic . the deictic. . On the other ( 1992b ) presents experimental evidence that this is indeed the case hand. Levelt 1982b found that on the same task some use a deictic . a speakerof Guugu Yimithirr . the same speaker may prefer one system for one purpose and another system for another purpose as Tversky ( 1991 ) and Herrmann and Grabowski ( 1994 ) have shown. may exclusively perspective developedthe habit of representing any spatial state of affairs in an oriented way.1 and to " thinking for speaking .b. that the choice of a perspectivesystem is arbitrary . M." I defined ' perspectivetaking as a speakers mapping of a spatial representationonto a propositional . I will ask how suitable thesesystemsare for spatial reasoning . Levinson The speakershould then have rememberedthe scene . I argued above that perspectiveis free.Willem J.a . Similarly. ' s absolute orientation . whether for languageor not. ( ) speakersconsistently system whereas others consistently use an intrinsic perspective system . and to what extent the systemsare mutually interactive. there is a pragmatics of perspectivesystems . In the rest of this chapter I will addresstwo issues . any spatial scenemay becomethe topic of discourseat a different place and time. who uses absolute well have . Levelt (4) I seea chair and a ball north of it . I will compare some advantagesand disadvantagesin using the three systemsintroduced above.

The ape can be on the right side ( to the for the intrinsic system violate converseness " " " right ) of the bear at the sametime the bear is on the right side ( to the right ) of the . or being involved in spatial planning discourseall require the ability to infer spatial layouts from linguistic description. Intril Bic. 3. I will specifically ask whether ellipsis is generated from a perspectivized or from a perspective .3. here 0 marks the position where a second occurrence of right is elided ) . 3. pole by instance . where west is the converseof east. Tversky Spatial reasoning abounds in daily life (cf.free representation . 1991 . which necessarily implies that the tree (referent) is to the left of the ball (relatum) .Taking and Ellipsis in Spatial Descriptions Perspective case. And the potential for spatial inference . assembly equipment ) Following instructions. One way of how speakers operate when they produce spatial ellipsis sorting this out is to study (such as in go right to blue and then 0 to purple . For holds converseness R then the other one by . spatial . This is demonstratedin figure 3. spatial representations will be language . I will analyzed some essentiallogical properties of the deictic and intrinsic systems . and Absolute Perspective Of many aspects that may be relevant for the use of perspective systems I will discuss the following three : ( I ) their inferential potential . system that it is about noon somewherein the Northern Hemispherewith the sun shining.1 Inferential Potential .3. Perspectivesystems CoDverseness An attractive logical property is converseness . summarizethem here and extend the analysis to the absolute system . such as front -back. the ball (referent) is to the right of the tree (relatum) . and it is perspective taking that maps them onto language specific semantic representations . Byrne and Johnson Laird 1989 search instructions road directions .independent . if object A is above object B. If the latter turned out to be the case. holds for the deictic system and for most cases2of the absolute Converseness . Assuming not for the intrinsic system but . If the two -place relation expressed by one pole is called R and the -1 if R (A . and ( 3) their mutual support or interference . north -south. A ) . B) ~ R 1(B.3 Some Properties of Deictic . ' also holds for the (three-place) deictic relation. Using this absolute bearing.free spatial representations . that would plead for the existence of perspective . In Levelt ( 1984 is crucially dependenton the perspectivesystembeing used )I . But it is easy to " . the tree must be west of the ball . ( 2) their ease of coordination between interlocutors . From the speakers Converseness point of view. the shadowsof the tree and ball indicate that the ball is east of the tree. B will be below A . aboveusually (though not always) involve directional opposites below.

. ~ ~ lreq ~ ~ * * ~ G .~ ~ ~ @ @ [ . ~~ ~ i .

gaze ) . This is the casefor the bear. and transitivity . see perspectivesystemthey employ in spatial discourse(for references but not . in following route And spatial reasoningabounds in everydaydiscourse . An example of nonagreement turned up in an experiment where I asked subjectsto describecolored dot patterns in such a way that other subjectswould be able to draw them from the tape-recordeddescriptions. and B is to the right of C. C ). This is the casefor the absolute and deictic systems . 51) .3. Levelt 1989 .Takingand Ellipsisin SpatialDescriptions Perspective 83 ape. Usually there is tacit agreement about the system used always. Converseness Theseare seriousdrawbacks of the intrinsic system . and so directions.4. and ape scene systemcannot rely on transitivity . A typical deictic description of this pattern that most subjectsused deictic perspective is the following : Begin with a yellow dot . cow. . B) and R(B. up. . which is a major the relation between referent and relatum in the intrinsic system . It essentiallytells you where the gazemoves(seeLevelt 1982b centered moves the Hurwitz 1984 up. Henceone cannot create a chain of inference . are very desirable properties if you want to make inferencesfrom spatial premises . It is therefore impossible to infer the relation betweenrelatum and referent from . An exampleof such a pattern is presentedin figure 3. It turned out . one cannot reliably conclude that A is to the right of C. and so forth . drawback for spatial reasoning Tra. Subjectswere instructed to start at the arrow. right . for on. and ball " " " the transitivity of " east of in the absolute system and of to the right of in the . but not for the intrinsic system scene depicts This state of affairs is demonstratedin figure 3. using the previous referent as a relatum for the next one. For the intrinsic system it is easy to construct a case that violates deictic system .5. for instance . it follows that R (A . . Shepardand . The flag. moves toward and away from the subject were typically expressedby vertical dimension . For the pattern in figure 3. This is characteristic for deictic perspective . Although the dot pattern was always flat on the table in front of the subject. And left of it is a black one. down ) . Then one step down there is a red dot.itivity Transitivity holds if from R (A . From A is to the right of B. The user of an intrinsic transitivity . I will shortly discussfurther drawbacksof the intrinsic system spatial reasoning 3. tree. Then right to a blue dot and from there further right to a purple dot . Then one step up is a greendot and further up is a brown dot . becauseit is viewerterms (up.5.2 Coordination betweenInterlocutors It is more the exception than the rule that interlocutors make explicit referenceto the and discussion . right . C ) . in jointly planning furniture arrangementsor equipment assembly .

~ ~ : ~ ~ 8uU 8uU . IUD aa R . t8 ~ ~ ai : ' IUD .i @ @ ~ [ * JJtI m MO MO lI .

~ ~ . You start at a yellow point . down.5 Pattern used in a spatial description task. Notice that all terms would have beendifferent if the pattern had been turned by 90 degrees . but derives from the intrinsic directions of the pattern itself. . There are no vertical dimension terms here. the ones useddominantly in intrinsic descriptions. This is a typical intrinsic 3 description. From there turn right to red and again right to a black dot. ~ right andEllipsis in Spatial Perspective Taking Descriptions right right t Figure 3. on the inside.. The nodes were colored ( here replaced by color names ) . and left. But other subjects used the intrinsic system . Thesedirectional terms in the description are depicted at the exterior side of the pattern . They described the scene as if they were moving through it or leading you through it . the directional terms . On the outside of the arcs are the dominant directional tenDS used in deictic descriptions . Now turn right to a blue dot and from there straight to a purple dot . Then go straight to a greendot and straight again to brown.' ' . The description is not viewer-centered .

5 from its intrinsicdescription A subject ~ ~ - - - ) tjj - . Most reproductions are like the one in figure 3.5 would still be valid if the pattern were turned by 90 degrees depicts the directional terms used in this intrinsic description. of the patternin figure3. subjects drawings tended to be incorrect. When I gave the deictic descriptions to subjects for drawing.Willem J. The interior of figure 3. What has happenedhere is obvious. The incongruent term straight is interpreted as " . In this example the listener tacitly assumesone perspectivesystem where the speaker has in fact used a different one.6. a deictic perspectiveand forces the intrinsic description The listener tacitly assumes into this deictic Procrustean bed. they usually reproduced ' the pattern correctly." This then is a caseof .6 ' s reconstruction . Levelt . M . which is a typical example. / hearercoordination . and systematically so. . failing speaker up Coordination failures can be of different kinds. But when I presentedthe intrinsic description. Our deictic and intrinsic systemsare subject to this confusion becausemany of the subjectended drawinghere (black dot) - t subjectbegan drawinghere (yellow dot) Figure3.

The felicity of speaker / hearer coordination in the intrinsic system is.~in Spatial Descriptions andEllipsl Perspective Taking dimensional tenDs are the same or similar in the two systems . not only awareness of requires recognition its localization. not a strictly visual one. therefore.b) . and theseuses functional ones ~ left right V left right r front L ~ front 0 ? ? -v V l' CD front Figure 3. Second . that is. but also its orientation . the intrinsic system of the relatum on the part of the listener. spatial discourse This contrasts with the intrinsic and absolute systems . derived from our characteristic usesof theseobjects. a major problem in coordination is that the systemderives 's ' from the speaker viewpoint. coordination in the crucially dependent intrinsic systemis only possible if the relatum is oriented. First . The interlocutors must keep track of their partners' viewpoints throughout .7 The alignment of an object' s left . and right side does not dependon its spatial. Any object that does not have an intrinsic front is excluded as a basefor the front / back and left/right dimensions . There is no visual feature that characterizesboth the front of a chair and the front of a desk (see figure 3. on the shared image of the relatum.7a. the speakers position and orientation in the . there is continuous switching back and forth in conversation betweenthe coordinate systemsof the interlocutors . which are speaker . front . frontness is an interpretative category (Miller and Johnson. however. The intrinsic .Laird 1976 ) . The utterance the ball is to the right of the chair can only ' effectively localize the ball for the interlocutor if not only the chair s position is known . properties. But also within the sameperspectivesystemcoordination failure can arise. These properties are . In a perceptual scene . but on its functional . therefore. requires that the interlocutor is aware of independent system the relatum' s orientation . And because scene the viewpoints are never fully shared . . For the deictic system .

seeFriederici and Levelt 1990for someexperimental results in outer space ) . and right is not fixed. Depending on the communicative task at hand. which tend to use a more functional approach. the alignment of an object s front . it may even be undetermined or ambiguous (as is the casefor the church in figure 3. left . for thesedescriptions or absence of vertical dimension terms . ' Herrmann and Grabowski 1994 ) . additional problems arise. In addition . What about speaker / hearer coordination in terms of an absolute ? Here. for which we have a designatedsensorysystem(and even this one can nowadays be tampered with . Not all intrinsic systemsshare all of theseproblems. a speakers choice of perspectiveis often given away by the terminology typical for that perspective . So evenan absolute systemis not without its drawbacks in spatial communication. Surprisingly. Levelt can be complex.Willem J. So far we discussed someof the coordination problems in utilizing the deictic or the intrinsic system .7a and 3. for instance on system what is north . perspective there are more subtle differences . The only absolute dimension that is entirely unproblematic dwellings (Levinson 1992b is verticality.7c).3. When a speakeruses terms such as north or east the chosen cannot be deictic or intrinsic .7b). I have mentioned the presence of vertical dimension terms in deictic directions in a horizontal plane and their total absencein intrinsic directions (the relevant data are to be found in Levelt 1982b ) . dead reckoning will be required if successful spatial communication is to take place in the dark . Even if such a main direction is indicated in the landscapeas a tilt or a coastline. the subjects in my experiment completely system is being used . One important factor is the choice of a default solution. Theseproblems do not arise for the deictic and absolute systems . Still worse. Hence. 3. What we experienceas the front side of a church from the outside ' (figure 3. presence givesaway which perspective . M . orientation -determining parts of objects than English or Dutch . and this has to be shared between interlocutors. Levinson ( 1992a ) was able to show that speakersof Tzeltal are much more vision-bound in deriving the intrinsic . This mechanism failed in the network description task in figure partners speech 3. interlocutors tend to opt for the same solution (Taylor and Tversky 1996 . or must at least be aware of the system used by their in . in the fog. Various factors can contribute to the establishmentof agreement . but dependenton its characteristic use(compare the alignments for chair and desk in figures 3. Interlocutors must agree on a system . farther away from one' s village. the use of intrinsic perspective always requires detailed interpretation of the relatum' s shape . Still . A first problem already appearedin the previous section. And . or inside unfamiliar ) .3 Interaction betweenPerspective Systems When languageusershave access to more than a singleperspectivesystem .7c) is its rear or back from the inside. the interlocutors must agree on absolute orientation .6.

4 The ball can now be abovethe chair with respectto one. ) or the German who utters Der Stuhl ist zu ihrer Linken (Ehrich 1982 . they varied widely in the " " 6 objects depicted and in backgrounds. Deictic abovealone. respectively ) . whether or not it is aligned with deictic above (scenes e and g. The basis for verticality is different in the three systemsunder consideration. The fastest responses .8 and asked them to name the spatial relation between the referent and the relatum. two . absolute perspectiveis quite dominant here responses " above" cases scenes a in absolute perspective dare of absolute ( ) . more or less .8 shows the percentageof above for eachconfiguration . In any perceptual situation thesethree basesof verticality mayor may not coincide. But further work by the same authors (Carlson. / speaker or all three of thesebases . More generally.I Descriptions andEllipsis in Spatial Perspective Taking ignored this distinctive information when they drew patterns such as in figure 3. Two problems that arise with multiple perspectives are alignment and preemption . (scene / ) is insufficient to release" above" responses . The three systemsmutually facilitate or interfere. Although the sceneswere formally the ones in figure 3. Figure 3. in which reaction times of judgments were measuredfor the same kind of scenes . This we know from the work by Carlsonand Irwin 1993 who Radvansky ( ). In the absolute systemverticality is determined by the direction of gravity . one perspective may gain (almost) full dominance.8. I had introduced that principle to . Clearly. The eight possibilities that arise are depicted in figure 3. Let us consider situations where there is a ball as referent and a chair as relatum and there is an observer . ' There are still other linguistic cues . the deictic dimension does not seem to contribute much in any combination. Different perspectives mayor may not be aligned in a particular situation . In the deictic systemit is probably determined by the direction of your retinal meridian (Friederici and Levelt 1990 ) . But in the absence above .8.8. the reaction times roughly follow the are for abovein absolute perspective judgment data in figure 3. .6.Radvansky and Irwin 1994 ). however. put subjectsin the positions depicted in figure 3. In addition . When you say The chair is on Peter s left. I am not familiar with any empirical study about the effectiveness ) 's of such linguistic cuesin transmitting the speaker perspectiveto the listener. This is most easily demonstrated from the use of preempting the other perspectives vertical dimension terms. depending on their alignment. such as in A is above / below B. followed by intrinsic and then deictic aboveresponses These findings throw a new light on a discussion of my " principle of canonical orientation " (Levelt 1984 ) by Garnham ( 1989 ) . intrinsic above keeps having some force. and so is the Frenchman who saysla chaiseest a definitely using the intrinsic system la gauchede ma soeur (Hill 1982 . you are . In the intrinsic systemit is determined by the top/ bottom dimension of the relatum. and if they are not aligned. showed that all three relevant systems contribute to the reaction times.5 The appropriatenessof saying the ball is above the chair varies dramatically for the depicted speakerin the eight scenes .

Levelt WOJj PU ( :+ II ( ~ ( ~ w ( + + .g :3 ( s ~ ( ".Willem J. M.' a1n : : ~ ( + + + - ~ .

@ @ @ Figure 3." "The ball is to the left of the chair. but not in ( b) .9 According to the principle of canonical orientation .Perspective in Spatial andEllipsis Descriptions Taking "The ball is in front of the chair. . It can be intrinsically infront of the chair in (d) and (f ). the ball can be intrinsically to the left of the chair in (a) and (c). but not in (e) ." .

however. The perceptual frame of the ball is the visual sceneas a whole. given the perceptual frame. Here the left /right dimension of the chair (the relatum) is not in canonical position . horizontal plane with respectto the perceptual frame. which agreeswith intuitions of many native speakersof English to whom I showed the scene(the formal experiment has never beendone. frame. " Why does the principle refer to the perceptual frame of orientation of the referent " and not " " . and this is because the front / back dimension of the relatum (the chair) is not in a canonical. as it derives from vestibular and visual environmental cues. In particular. According to the principle of canonical orientation this is a possibledescription in ' a ( ) . The chair is in canonical position with respectto this perceptual frame. just to the perceptual frame of orientation ? In figure 3. the description is virtually impossible in (b) . and in particular its vertical direction . virtually unacceptablefor (e). b. that is. and c. That dimension is in canonical orientation to the relatum' s perceptual frame. The perceptual frame for the chair ' s orientation is in this case the normal gravitational field.9 it is indeed impossible to distinguish betweenthesetwo. The description refers to the relatum s intrinsic left /right dimension. determines whether some dimension of the relatum (the chair) is in canonical position . its front / back dimension is. though) . Because the principle is directly relevant to the presentdiscussion of alignment and preemption. Levelt account for certain cases where the intrinsic systemis " immobilized" when it conflicts with the deictic system . Although in (/ ) the chair is not in canonical position. the chair' s left/right dimension has a canonical direction . Here the chair is not in canonical position either. More generally. it is not in a horizontal plane. This description is fine for (d ) . refer to the intrinsic description the ball is to the left of the chair. in the left -hand side of the figure. Hence the description is again possibleaccording to the principle. it is for many native speakersof English acceptableto say the ball is to the left of the chair in caseof (c) . M . Casesa.Willem J. However. I cite it here from the original paper: The principle of canonical orientation is easily demonstratedfrom figure 3. a referent' s perceptual frame of orientation will normally be the experienced vertical . but the chair' s left /right dimension is. and . Its orientation .9. it lays in a plane that is horizontal in the perceptual frame. It is. it is in a horizontal plane of the perceptual . Hence the principle of canonical orientation is satisfiedin this case The state of affairs is similar for the intrinsic description the ball is in front of the chair. Finally and surprisingly.

. If this is a subject' s experience . . . . This is becausethe intrinsic left /right dimension in which the fly is spatially related to John' s nose is canonically oriented with respect to the perceptual frame. . then it is appropriate to say there is a fly aboveJohn's head . becauseit is so close to it . cf.7 Fly 3 ' ' is further away from John s head and does not naturally take John s head as its .Perspective Taking and Ellipsis in Spatial Descriptions fly 2 . let us consider the figure in some more detail. will be the samefor referent and relatum. . according to the principle . Hence the perceptual frame of the referent can be different from the larger perceptual frame in which the relatum I 1 . But there are exceptionsin which a dominant visual Gestalt adopts the function of perceptual frame for the referent. To show this. In that paper I argued that it is not impossible in this caseto say about fly 2 in the ' picture: there is afly aboveJohn s head even though the top/ bottom dimension of ' John s head is not in canonical orientation .10. . " . the principle of canonical orientation predicts that it is appropriate to say. Notice that in these three casesJohn' s head itself has the bed and its normal gravitational orientation as its perceptual frame. And fly 2 may similarly take John' s face as its perceptual frame. Here John' s face is a quite dominant background pattern which may becomethe perceptual frame of orientation for the fly .10 to the of canonicalorientation According . In that case .Radvansky and Irwin ( 1993 . Figure3. figure 3. which is reprinted here from Levelt ( 1984 ). This can happen in the sceneof figure 3. fly I can be intrinsicallyto the left of ' s nose and principle 's head John . Hence it is less appropriate here to say it is " above" perceptual frame of reference ' John s head. . but not fly 3. It is in a plane perpendicular to the top/ bottom dimension of the face.8g) now confirm that this can indeed be the case . r fly 3 . beginning at the location of fly 1. John fly 2. ' . And this is in agreementwith the principle . fly ? " - . I . . The experimental findings by Carlson. . there is afly to the left of John's nose . canbeabove from Levelt 1984 (reproduced ).

there can be a hierarchy of frames. the " framework vertical constraint. If in a scenecanonical orientation does not hold . in part . according to Garnham. however. but it does not. and it is not neces sanly the casethat the referent and the relatum share a frame.9. M . in agreementwith the the experimentsinvolved casessuch as the principle of canonical orientation because one just discussedfor fly 2 in figure 3. above / below derives in this case from the normal gravitational framework. We have seenthat systemsdiffer in inferential potential and pragmatic significance . For instance acrossa street.8 system In this section I have discussed various properties of perspectivesystemsthat are of . In other words. be traced back to a vagueness of the term canonicalposition.Willem J. The only tenable " " interpretation of canonical position is a weaker one: With this further specification." which says that " no spatial description may conflict with the meaningsof aboveand belowdefined by the framework in which the related objects are located. Garnham' s critique of my 1984formulation of the principle can. This is obviously " false." But the results by Carlson. That allowed him to ignore the distinction betweenthe referent' s and the relatum' s perceptual frame and to formulate a really simple principle. The findings are. if a vehicle is parked ) correctly pointed out. even if the perceptual frame for the bollard is given ' by the street (whoseright side is opposite to the vehicle s right side) .10. Garnham ( 1989 ) challenged the principle of canonical orientation . which should make this description impossible according to his constraint.8 contradict this because .10. a bollard [traffic post] to the intrinsic right of the vehicle can still be describedas to its right " (p. Hence there is a conflict betweenthe meaning of abovein this framework and the description the ball is above the chair. 59).Radvansky and Irwin ( 1993 ) for scenese and g in figure 3. the intrinsic system is evaded by the standard average European (SAE) language user. then. it is preempted by the deictic or by the absolute . as Garnham ( 1989 . It does not positively exclude the following strict interpretation : the dimension on which the intrinsic location is made should coincide with the samedimension in the perceptual frame. Levelt is embedded . he rejected those with agreedwith the intuitions concerning the scenes respectto figure 3. Although he in figure 3. the principle of canonical orientation seems to be in agreementwith intuition and with experimental data.

conceptsfor which there . The issuein this section is whether spatial ellipsis originates beforeor after perspective taking . or rather.andEllipsis in Spatial Descriptions Taking Perspective in their demands on coordination between interlocutors.11represents as left or as right . er. the process of perspective taking in terms of some perspective system a a onto propositional or semanticone.4 Ellipsisin SpatialExpressions . it is neither intrinsic nor deictic. of surface ellipsis (roughly following Hankamer and Sag 1976on " . concepts speaker message ' . " and " surface " anaphora ) deep Compare the following two descriptionsfrom our data.5. and I will return to it below. the same referent/ relatum relation is expressed 3.. When we talk about Perspectivetaking is one aspect of our thinking for speaking about we create spatial properties of entities or predications spatial configurations. This assumption is itself perspective mayor may not be correct. on thinking for speaking here. The latter is the maps spatial representation 's which consists of lexical . in a particular lexical concept " " the latter case . one " unit from the yellow dot we place a blue dot . that is. are words in the speakers target language This state of affairs is well exemplified in figure 3. in that respect elliptic " " Full deictic: Right to yellow . Right to blue. does the speaker decide not to mention a particular feature of the spatial representation . and the outcome dependson quite abstract properties of the rivaling systems . One. concurring systemsare not totally dormant in the speakers ' mind. " Elliptic deictic: From pink we go right one unit and place a yellow dot. dependenton the speakers perspectives . These predications usually relate the entity to some relatum . plus the move that precedesit . We also have seenthat if ' one systemis dominant . of canonical orientation of the implication principle 3. It is on the perspectivesystem being used . Depending on the perspective one critical detail (circled) of this example 3. The first . Finished.12. that is. the secondone is description is nonelliptic with respectto the directional expression . Figure taken. In short.11 express es that the choice of lexical concept (and ultimately of lexical item) depends . In other words. Figure in two systematicallydifferent ways. as is the . Both relate to the encircled trajectory in the left pattern of figure 3. The samepattern is expressed ' . . referents in the scene . does the speakerdecidenot to express ? In the first case we will speak of " deep ellipsis" . It is that the to be clear on the spatial representation underlying assumption important -free. Their rivalry appearsfrom the kind and speedof a subject s spatial judgments.

elliptic expressionis that it contains no spatial term that relates the blue dot to the (previous) yellow one. Levett intrinsi deictic perspe perspectiv taking taking ~ ~ lexical lexical concep concept RIGHT LEFT ~ ~ SELECT LEXICAL ) C . . thus activating the lexical concept RIGHT a secondtime. ~ ~ word word " " l" eft r"ight representationfrom a The crucial feature of the latter . M . This is deep ellipsis. The first one is that the speaker in scanning the spatial configuration recognizesthat the new visual direction is the sameas the previous one. two possibilities. Before getting into perspectivetaking.Willem J. How does the speakercreate this ellipsis? There are. This repeatedactivation of the concept then leads to the decision not to formulate the lexical concept a secondtime. the speakerdecides not to prepare that direction for expressionagain. essentially . The second possibility is that the speakerdoesapply deictic perspectiveto the secondmove.

Here is an instance of a full intrinsic description of the sametrajectory : Full intrinsic : " Then to the right to a yellow node and straight to a blue node. straight) be ? deleted t t that is. Does ellipsis occur in intrinsic descriptions of this kind? If . This is surfaceellipsis. In case of deep ellipsis this should be possible .12 Deictic and intrinsic descriptions for two patterns. the condition for surface ellipsis is not met for the intrinsic speaker . Can the last spatial tenD (right. In other words. it has to be formulated in speech . the present intrinsic one will scan the spatial sceneand recognizethat the new direction is the sameas the previous one and the speakermay decide not to prepare it again for expression . obligatory to usea directional expression This state of affairs can now be exploited to test empirically whether spatial ellipsis is deep or surfaceellipsis.13. Just as the previous deictic speaker . whereasthe direction of the secondmove is mapped onto STRAIGHT . In the intrinsic system the direction of the first move is mapped onto the lexical concept RIGHT . intuitions waver here." Can the samestate of affairs be describedelliptically ? This should produce something like: Then to the right to a yellow nodeand to a blue node . But in case of surface ellipsis the intrinsic speaker has a problem. it is . Because the latter is not a repetition of the former . not to repeat the word right .PerspectiveTaking and Ellipsis in Spatial Descriptions right right Figure 3. it is optional to mention the direction. The alternatives can now be distinguished by observing what happensin descriptions from an intrinsic perspective . The answer is not obvious. Thesetwo alternatives are depicted in figure 3.

Levelt MODEL I " Surface ellipsis " ( ellipsis is perspective . Is it reiterating a lexical concept or a spatial direction that ? matters -+ .independent) move next ~ new the direction of is the direction as the the same move ? move the of preceding no + use of directional is obligatory expression yes + directional use of is optional expression Figure3.Willem J. M . the same directional term to be used ? next yes - use of directional expression is obligatory of directional use is optional expression MODEL2 " Deep ellipsis " (ellipsis is perspective.e. is the same ( lexical ) concept to be expressed .13 Surface ellipsisversusdeepellipsis.dependent ) move ~ given perspective . i .

. if we find ellipsis in such cases . A normal full intrinsic description of this trajectory (plus the previous one) is Full intrinsic : Then right to green. For each critical move I determined whether a directional term would be obligatory or optional (i. it is also obligatory under the deep model and vice versa. Surface ellipsis is impossible here because" right " is not a repetition of the previous directional term (" up" ) .9 and hence with 44 x 14 = 616 pattern descriptions to be checked . It should be noticed that the two models make the same predictions with respectto deictic descriptions. Deep ellipsis is impossible because the trajectory direction is different from the previous one. And we can create an alternative case where surface ellipsis is possible for intrinsic descriptions. we will have to reject both models.PerspectiveTaking and Ellipsis in Spatial Descriptions so. .e.12.12.b) we had asked 53 subjectsto describe 53 colored dot patterns. In this set I found a total of 43 casesof ellipsis. And from the greencircle we go right to a black circle. This left me with 31 consistent deictic subjects and 13 consistent intrinsic ones. But this is not so for the intrinsic descriptions. The table presentsthe actual occurrence of ellipsis for these four caseswithin each model. It should be clear that neither deep nor surface ellipsis is possible in a deictic description of this pattern. In an experiment reported in Levelt ( 1982a . I removed all subjectswho did not have a consistent perspectiveover these 14 critical patterns. Take this full deictic description from our data: Full deictic: From white we go up to a greencircle. Hence. we have an argument for deep ellipsis. An exampleconcernsthe encircled trajectory in the right pattern of figure 3. Hence there are four casesper model. producing " Then right to green. I will call the circled moves " because in thesepatterns " critical moves the surface and deep models make predictions about them that differ critically for deictic and intrinsic descriptions in the way . 10 Theseare presentedin table 3. And then right to black. And to black" or some similar expression ? That is an empirical issue . Is surface ellipsis possible here. The table presentspredictions and results under both models of ellipsis.14. if use of a directional term is obligatory under the surface model.12 ) . among them those in figure 3. elidible) under the model in deictic and in intrinsic descriptions (such as I did above for the critical moves of the patterns in figure 3.1. I checked all 53 subjects to detennine whether made they elliptic descriptions for any of these 14 critical trajectories. a ' subject s 14 pattern descriptions should either be all deictic or all intrinsic . they given in figure 3. Among the test patterns there were 14 that contained such critical just described moves are . but not deep ellipsis.

-Q i 1 Figure 3.. The critical in two cases rotated ) (though . M .-.--.-~ 3 ~ ~ o- t t1 t €~ -oo o.-~ ~ x t ~ o.14 " " Fourteen test patterns containing critical moves . including the two example patterns of or the other examplepattern as a substructure either the one includes test Each pattern figure 3. Levett 1 re f - toe t ~ t ~ ~1 ty"(D t ~ o -. moves .. are circled.100 Willem J.12.-o~~ ::::9 .

the same level of representation representation (seefigure 3. These findings show that the decision to skip mentioning a direction is really an 's . decideon someorder of description because speechis a linear medium of expression 1989 are nonlinguistic The principles governing these linearization strategies(Levelt 1981 . the subject is shown an array of two objects A and B on a table.. But this contradicts the convincing experimental findings reported by perspectivized Brown and Levinson ( 1993 ) and by Levinson (chapter 4. there is only one deictic 11All other deictic and all intrinsic violates it . where A is (deictica11y ) left of B (henceAB ) . namely. This is. perspective -dependent in ) spatial representations nonlinguistic spatial matching tasks. and then asked to . the deep model is in good shape our sample .11 where linearization decisionsare taken. speaker linguistic perspective system not on a semantic basedon a visual or imagistic representation .. Where it prescribesobligatory use of a directional term. that model is in If a model says" obligatory .Takingand Ellipsisin SpatialDescriptions Perspective Table3. Then the subject is turned around 1800to another table with two arrays of the sameobjects. this volume). How do the two models fare? It is immediately obvious from the table that the surface model is out. for a total of 19. we must . that descriptionsrespectthe description deep model.1 of Ellipsis underSurface and Deepmodels Distributionof Elliptical Descriptions Model Description is Directional tenD is obligatory optional Total Surfaceellipsis deictic intrinsic Total Deep ellipsis deictic intrinsic Total I 24 25 18 0 18 19 24 43 1 24 25 0 18 18 I 42 43 occur.14 or the layout of our living quarters). spatial representations are apparently not perspective ." but ellipsis does nevertheless trouble. ) ) in character. casesof ellipsis) and one among the deictic descriptions. That is almost half . If ellipsis runs on a -free spatial representation . there are no less then 18 violations among the intrinsic descriptions (i. (lexical-conceptual) ) . It precedes the speaker application ofa perspective early step in thinking for speaking ' is irrelevant here. they relate exclusively to the image (and in fact nonsemantic itself. A -B and BA .e. its speakersuseoriented(i. For instance . which show that when a languageusesabsoluteperspective . The decision is s the . probably . When we describe2-D or 3-D spatial patterns (such as the patterns in figure 3. But these very clear results on ellipsis create a paradox. In contrast.e.

Levelt indicate whic~ of the two arrays is identical to the one the subject saw before. the corresponding lexical concepts will be the sameas well (e. in addition . paradox remains. One could then ' s show that absolute and deictic perspective that Brown and Levinson findings argue are " Whorfian . The two models can only be distinguished and the surface model under deictic perspective 's when the speaker is intrinsic (cf. if two arcs have the samespatial direction or orientation . The problem is. After all .1. the above ellipsis data would be explained as well.. however. why intrinsic perspectiveshould be non-Whorfian . it seems " is linguistically free. where A is deictically to the right -=+ is the absolutedirection of the vector AB . a property of the spatial representationitself. Hence spatial representationsare perspectivizedalready. human interest. in the sense in tasks even follow the dominant perspectiveof the language . What the subjectapparently preserves A native English or Dutch subject. columns 1 and 4. that information about intrinsic position will be crucial for an intrinsic spatial description. If . or both east ). As and the . speakersof Mopan . absolute perspectivebehaveslike deictic perspective perspective is absolute. There are preferences . both north . The " " absolute subjectinvariably choosesthe BA array. for sure. What we pick out from a scene in terms of entities and spatial representations relations to be expressedin language is not subject to fixed laws. perspectives ellipsis predictions are different for different perspectives can be seenfrom table 3. is the fact noted in the introduction that perspective More important . will profit from registering the position of foregrounded objects relative to background objects that have intrinsic orientation . ellipsis data of the kind analyzedherecan only distinguish between 's the deep and surface models if the speaker perspectiveis intrinsic . to me. the samepredictions result from the deep . How to solve this paradox? One point to note is that the above ellipsis data and Brown and Levinson' s ( 1993 ) data on oriented spatial representationsinvolve different . the deep and surface models of ellipsis make the samepredictions. violations perspective could show that neither model is correct. In other words. and so on. and this chapter has discussedvarious reasonsfor choosing one perspectiverather than another. that is.g. If a speaker . If at some later time the sceneis talked about from memory. . dependingon communica- . There is no " hard-wired mapping from spatial to semantic . In this respect deictic under only perspective 's . of course. M . But if we discard the option of excluding intrinsic perspective from " Whorfianness" the . nonlinguistic " " 12 where there is no thinking for speaking taking place. columns 2 and 5). exclusiveusersof intrinsic perspective .102 Willem J. of B. Similarly." that is. we can go for one perspectiveor but they are no more than preferences another if our culture leaves us the choice. the intrinsic systemis not Whorfian in the samesense . typically produces the deictic response that they (A -B) . following Gestalt properties of the scene .

rather than A is in B' s neighborhood) are prerogatives of the speakerthat are not thwarted by the limited choice of perspective . that can be formulated. We decide on what to say. we decideon what to say first . relata. Any speakercan attend to it and make it the ground for ellipsis. As all other speakers . or message . 3. And we linearize the information to be expressed that is. In -relevant spatial featuresdoes not preempt or other words. and their spatial relations for expression Microplanning involves. we translate the information to be expressedin some kind of " propositional " format . That one arc in figure 3. we can say that macroplanning language involves selectingreferents . In microplanning. A more likely state of affairs is this. A is north of B. such as the speakerregister in memory spatial features that are perspective absolute orientation of the scene . spatial relations to be expressed . In particular . the pattern of linearization chosen when the sceneis complex. Hence. In macroplanning we elaborate our communicative intention . from Levelt ( 1989 ).. When we apply thesenotions to spatial discourse . This would be impossible if the spatial representationdictated its own semantics . and even the decision to expressabsolute perspectiveat all (e. . the addition of perspective suppressthe registration of other spatial properties that can be referred to or used in discourse . . " and so forth . absolute perspective . The choice of referents . among other things. It is correct to say that Guugu Yimithirr speakerscan choosefrom only one.speaker of any culture. what to say next. relata. selecting information whose expressioncan be effective in revealing our intentions to a partner in speech . .12 is acontinuation of another arc is a spatial feature in its own right that is available to a . creating a semantic must . applying some perspectivesystemthat will map spatial directions/ relations onto lexical concepts .5 Conclusion This chapter openedby recalling. This does not mean. Brown and Levinson' s ( 1993 ) important Whorflan findings cannot mean that spatial and semantic representationshave a " hard-wired" isomorphia. however. A culture' s dominant perspectivemakes a speaker attend to spatial properties that are relevant to that perspectivebecauseit will facilitate . but that doesnot obliterate their freedom in expressingspatial configurations in language . In particular . they can in what deem relevant and in that are express language they ways communicatively effective.g.in Spatial Descriptions Perspective Taking and Ellipsis 103 tive intention and situation. the Guugu Yimithirr can attend to various aspects of their spatial representations . that an ellipsis decision must make referenceto such features . this message representation consist of lexical concepts that is for which there are in words the . theseattentional blasesmake the (later) discourseabout the scene -specific . the distinction betweenmacroplanning and microplanning . or " thinking for speaking . concepts target . .

this volume). M. cultural agreementon perspectivesystems . The maiQdistinction is betweenrelative. intrinsic . It has considered the advantagesand disadvantages of deictic. chapter I . in particular about the pragmatics of different perspective systems . requirementsof the communicative task at hand. What is it in our biological roots that makesthe choice so limited? Notes I . intrinsic . 2. The relative systemsare three-place relations betweenreferent. the intrinsic and absolute systemsare two-place relations betweenreferent and relatum. It has also considered how a / hearer coordination in spatial discourse speakerdeals with situations in which perspectivesystemsare not aligned. performing well-defined restricted map. and absolute systems . Still . I argued representationmight be perspective that this is paradoxical only if the mapping from spatial representationsonto semantic " " representationsis hard-wired.104 WillemJ. and absolute systems for spatial reasoning and for speaker . spannedby two converseterms.and microplanning . The discussedrecent findings by Levinson and Brown strongly suggestthat such is indeed the case . to the question whether this perspectival thinking is just for speaking or more generally permeatesour spatial thinking . there is no law that the speaker must acknowledge orientednessof a spatial representation (if it exists ) when deciding on what to expressexplicitly and what implicitly . In particular . speakershave great freedom in both macro. The three perspectivesystemsdiscussedhere are relative egocentric ( = deictic). where the traverse direction in the absolute systemis not polarized. The ping operations interfacing requirements are too specific for these perspective systemsto be totally arbitrary . intrinsic allocentric. There are no strict laws that govern the choice of relatum and referent. and so on. it is not my intention to imply that anything goes in thinking for speaking . Brown and Levinson ( 1993 ) present the caseof Tenejapan . I then presentedexperimental ' s decision data on spatial ellipsis showing that perspectiveis irrelevant for a speaker to elide a spatial direction term. " " Thinking for speaking led. But much more challenging is the dawning insight from anthropological work that there are only a few such systemsaround. contrary to the Whorfian findings. relatum. that dictate how to linearize information . There are only (often strong) preferenceshere that derive from Gestalt factors. in some Whorfian way. that is. But this is not so. and each has an egocentric and an allocentric variant. there is just one . Levelt The chapter has been largely about microplanning. and baseentity (" me" in the deictic system ). easeof coordination between interlocutors. " " Perspectivesystemsare interfaces between our spatial and semantic modules (in Jackendoff' s sense . and so forth . and absolute allocentric. that is. as a matter of course. Having speculatedthat the underlying spatial -free. I am in full agreementwith Levinson' s taxonomy of frames of reference(here called " perspective " systems ) in chapter 4 of this volume.

The new path is straight. StephenLevinson ( personal communication). I am ignoring a further variable. a 900 angle between the relevant bases . 7. but there also should be no coordination that can be interpreted as one directional tenD having scopeover two constituents. What is intrinsic top/ bottom in SAE is " longestdimension" or the " modal axis" of an object in Tzeltal. 5. But Buhler' s caseis not strong unobjectionable as long as we do not confound the two systems for this network.Radvanskyand Irwin do not discussitem-specificeffects . is applicable and holds for this system(seebelow in text). Herrmann and Grabowski 1994 ). the fonner . each fonning a base for a (different) deictic system . 10. Another point to keep in mind is that the always agree between subject and item analyses " " on the experimental procedure may invite the development of perspectivestrategies part of " " . 9. namely. who can use their intrinsic system when the relatum' s critical dimension is not in canonical orientation . 3. but not the latter . of course . that their statistical findings . It is not essential in the route-type description that " I " (the speaker in his imagination) make the moves and turns. I have excluded all caseswhere subjects mention a line on which the nodes are located. though. But the Tzeltal intrinsic systemdiffers substantially from the standard average European (SAE) intrinsic system (see Levinson 1 992a ) . 6. the notion of converseness is not applicable. Carlson. and occasionally the employment of an unusual perspective subjects .8. " Deixis am Phantasma . as in From pink right successively yellow and blue or A road turns right from pink and meets first yellow and then blue. If there were a ball rolling through the pattern.PerspectiveTaking and Ellipsis in Spatial Descriptions 105 tenD meaning " traverse . the listener' s viewpoint/ orientation . My criterion for ellipsis was a strict one. however. Speakerscan and often do expressspatial relations from the interlocutors perspective . a real one and imaginary one. as in for you. has a connotation of verticality . Another case studied by Carlson. There would be two speakersthen. ' t have deictic the directional tenDSwould be just the same . There should. right. It is the case ." Obviously. be no directional tenD. Hence it is the intrinsic orientation of the current path that is taken as the relatum. The notion of transitivity . Barbara Tversky ( personal communication) has correctly pointed out that Buhler ( 1934 ) " would treat this caseas a derived fonD of deixis. This is . What perspective the speaker in fact does in this description is to use the last directed path as the relatum for the subsequentpath. the ball is to the left of the chair. 8. has presentedevidence that the principle does not hold for speakersof Tzeltal. Conditions for this usagehave been studied by Herrmann and his colleagues(cf. or left from the current one. Carison-Radvansky and Irwin included severalscenes that were fonnally of the sametype as scene(g) in figure 3.9 with fly 2. no reasonwhy this should also hold in other cultures. however. where the speaker imaginesbeing somewhere(for instancein the network) . Here I am considering only one case of nonalignment. But a ball doesn . These numbers differ from those reported in Levelt ( 1982b ) becausethe present selection criterion is a different one. . There is. among them the one in figure 3. for instance .Radvansky and Irwin ( 1993 ) is 1800 nonalignment. although it is likely that the type of relatum used is not irrelevant. 4.

( 1989 ). ( 1990 ). . Working paperno.14. andSag .. 253 .).30. and action . Perception andPsychophysics . J. Klein (Eds linguisticstudies . London . J. Deep Hill .106 Willem J. 33. The speaker problem of the . . M. Weissenborn . K. .. ( 1989 . Cognition . ( 1982b ). I prefer to seeit as a mistake or omission. J. and Irwin. It goesas follows. Petersand . A. Levelt II . Chichester Speech topics -Laird. and Irwin. W. ( 1993 ). M. A. In R. M. .268 and W. 28. front/ back . 391. Hereand there : Cross on deixisand and W. . 47. ( 1981 ). 646 Ehrich . DE . ( 1982 ). ( 1982a ). Linguistic Inquiry . place . Dordrecht : Reidel . Max PlanckInstitutefor Psycholinguistics Buhler : Die Darstel /ungsfunktion . J.).). Saarinen . . and Johnson ). 31 . Linearizationin describingspatial networks . In R. place . 24. . Klein . R. 219 (Eds . The structureof living space descriptions .671 . Linguisticand nonlinguisticcodingof spatialarrays in Mayancognition . M.). andquestions . A. A unified theory of the meaningof somespatial relational terms . A. Jena . .266 factorsandmentalrepresentations . Klein (Eds . .220 E. P. andaction . 45 Hankamer andsurface . figure 3. ( 1934 ).315 . W. L. M. Spatialreasoning Byrne . J. Cognitive styles . Jarvellaand W. Process es .244 ? Cognition : Whereis above . B95 . Journalof Memoryand . Wiley Friederici in weightlessness : Perceptual . The discussion that follows in the text is much inspired by discussions with Stephen Levinson. Klein (Eds . J.. D. A major derSprache : Fischer . 199 (Eds Levelt in theuseof spatialdirectionterms . Philosophical Levelt s linearization Transaction . S. J. P.60 . Spatialreference . Amsterdam demonstration : Benjamins . 564 575 Language -Radvansky Carlson in vision and . ( 1976 anaphora. and Levinson . Chichester : . Jarvellaand W. ( 1993 ). Journalof MemoryandLanguage termassignment . J. . . In S. 251 Chichester : Wiley. Up/down study of Hausaand English In J. Speech in deixisandrelatedtopics : Studies . and Levelt . N. DE . CognitiveAnthropologyResearch Explorations . 223 language -Radvansky Carlson frameactivationduringspatial . W. Jarvella . The case occurs in a deictic description of the fourth pattern down the first column in . I. 1982 . 305 RoyalSociety Levelt .. ( 1982 . 13 ' . From there left to a pink node This obviously violates both models of ellipsis. Framesof reference . Sprachtheorie in translationin R. W. References Brown : .). 12... part on deixisfrom this work appeared : Studies in deixisandrelated : Wiley. 46. beliefs .. 9.249 . C. left/right: A contrastive ). J. L. ). place . Nijmegen Group. Andfrom there to a green node .42. A. V. . 7. Garnham . Reference . andaction . M. ( 1994 ).426. Speech : Studies in deixisand relatedtopics .

( 1984 ) . G. Michaelis. Vision . 27. The psychologyof learning Tversky.. ( 1992a ). A . Speaking : Tzeltalbody part tenninol Levinson . andJ. ( 1983 .). Journal of Memory Taylor . Limits of perception W. and linguisticdescription . S. 12 .146 . ( 1984 ). Press . 435. Acredolo (Eds. Cambridge. M. Group . A . and application. research . B. ( I 992b) . Cognitive Anthropology Research . Max Planck Institute for Psycholinguistics .Perspective Taking and Ellipsis in Spatial Descriptions 107 . Workingpaperno.. ( 1991 . MA : Miller . H . Nijmegen of spatial Levinson.). vol . Language and cognition : The cognitive consequences description in Guugu Yimithirr . W. (Eds. AskeN . P. Upward direction . 13. C.358 . In G. R. J.. W. L . Berkeley: Berkeley Linguistics Society . and Hurwitz . 161 . S. Koenderink of Maarten (Eds . . ( 1996 and Language(in press ). B.193 . Working paper no. 18. of left and right turns in maps. D . J. Cognition. and Johnson-Laird . ( 1976 ) . Nijmegen. S. Languageand perception .). In H . In J. Thinking for speaking of the Thirteenth Annual Meeting. Spatial mental models. Some perceptual in honour : Essays . L . Harvard University Press . ( 1989 ). H . CognitiveAnthropologyResearch Group. vandeGrind.).: MIT Press .444. In A. Max PlanckInstitutefor Psycholinguistics . and H . Spatial ) . How languagestructures space Talmy . M. ( 1987 ) . ) . Pick and L . New York : Plenum Press ). ogyand objectdescription . R. Beery. Filip Slobin. Perspectivein spatial descriptions. C. Berkeley Linguistics Society: Proceedings . New York : Academic Press in researchand theory. limitationson talkingaboutspace Levelt . 109 and motivation: Advances . 323 . vanDoom. . Bower (Ed. shape . MA . Cambridge to articulation : Fromintention Levelt . orientation: Theory. Utrecht : VNU Science A. Bouman . and discrimination Shepard . and Tversky. mental rotation . N .

. suggest slight of the existing distinctions. it is also a question of which frames of referencethey employ. I will offer only a twisting trail . The trail begins with some quite surprising cross cultural and cross linguistic data.3. for . some languagesdo not employ our apparently fundamental spatial notions of left/right/front / back at all . which leads inevitably on into intellectual swamps " conversewith one another . and even unconsciousgesture have adopted the same and modalities all these capacities systemsthat drive . Then we can. languagesmake use of different frames of referencefor spatial description. frame of reference Thesefindings. Second frame of referenceacross modalities or inner representation systems .4 Chapter - : Cross Framesof Reference and Molyneux' s Question UnguisticEvidence C. describedin section 4. bring some of the distinctions made in other modalities into line with the distinctions made in the study of . information exchangingspatial To preview the data. yet instead of vistas. instead they may. first . it seems . specifying locations in terms of north/ .2. In short. -modal tendency for the same frame of referenceto be employed in there is a cross language tasks.issuesabout how our " inner languages and minefields . inference tasks. The choice of a frame of referencein linguistic ) correlates with preferencesfor the same frame coding (as required by the language of referencein nonlinguistic coding over a whole range of nonverbal tasks. we " " reformation a and must clarify the notion frame of reference in language .1 WhatThisis AUAbout The title of this chapter invokes a vast intellectual panorama. This is not merely a matter of different use of the same set of frames of reference(although that also occurs). For example . we must ask whether it even makes senseto talk of the same ! . employ a cardinal direction system example south/ east / westor the like. This suggests tasks. Levinson Stephen 4. First . recall and recognition memory tasks. imagistic reasoning that the underlying representation . There is a secondsurprising finding . prompt a seriesof theoretical ruminations " " in section 4.

i. The results . instead. and so on. are not intrinsic ) . nonverbal memory. T7 . Cross . Clark 1973 Miller and Johnson-Laird 197 ( ..3 but can be thought of as labeling distinct . modal transfer of spatial information .3.e. In English we achieve this either by utilizing features or axes of the " " ground (as in the boy is at the front of the truck ) or by utilizing anglesderived from ' " the viewer s body coordinates (as in the boy is to the left of the tree" ) .e. for example . something like our cardinal directions north . and are also independentof properties of the ground object. 1 : Evidence 4.000 people in the Indian community of . This has interesting implications for what is known as frame of reference " the 's " question about how and to what extent there is cross Molyneux question. ( Brown and Levinson 1993a .. distinct . kinds of coordinate systems At first sight. The first " the second a " relative frame of solution I shall call an " intrinsic frame of reference . with someindication of the range of subtypes language in a Mayan language .lt. . " reference (becausethe description is relative to the viewpoint . south. and indeed on close consideration (see . plane by utilizing fixed bearings east. the only natural solutions for a bipedal creature with particular bodily asymmetrieson our planet. but the particular dialect described is spoken by at least 15. A tentative typology of the three major frames of reference in .from the other side of the tree the boy will be seento be to the right of the tree) .b. we turn to the question Why does the same frame of reference tend to get employed across ? It turns out that modalities or at least across distinct inner representation systems information in one frame of referencecannot easily be converted into another. Finally . so that some sensecan be made of the idea of sameframe of reference language acrosslanguage . Some languagesuse neither of thesesolutions. I will therefore refer to the relevant population as Tenejapans Tenejapa with of an conducted here are a Penelope Brown part ongoing project. Levinson " " .1 . as found Here I wish to introduce one such absolute system . mental imagery. will be found in section 4. and west.110 Stephell C. Somc languagesusejust the first solution. The notion " frame of " will be reference explicated in section 4. But they are not. . i. are not relative. Tzeltal is a Mayan languagewidely spoken in Chiapas. . Levinson and Brown 1994 ). reported ... they solve the problem of finding angleson the horizontal . Mexico. Spatial descriptions utilizing such a solution can be said to be in an " absolute" frame of reference becausethe ( anglesare not relative to a point of view.ModalTra Mferof Frame of Reference from Tenejapan To describewhere something (let us dub it the " figure" ) is with respectto something else (let us call it the " ground" ) we need some way of specifying angles on the horizontal . these solutions seem inevitable.

Second acrossis labeled identically in both directions (eastand west). And speakersof the languagepoint to a specificdirection for down.. this is the basic way to describethe relative locations of all objects separatedin spaceon whatever scale . but it is of limited utility for spatial description becauseit is usually only employed to describe objects in strict contiguity . although spatial description is required. testsof this ability to keep track of directions (in effect. Figure 4. to dead reckon). even . and acrossrecurs in a number of distinct lexical systemsin the language . First . Thus if one wanted to pick out one of two cups on a table.. linguistic specificationslike our to the left. it is transparently derived from a topographic feature : Tenejapa is a large mountainous tract . behindare not " available in the language . and they will continue to point to the same compassbearing when transported outside their territory . and so forth . there are therefore certain ambiguities in the interpretation of the relevant words. specializedconcrete nominals of different roots that describe . descending (going north ). Thus for objects separatedin space . say. but the essentialpoints to grasp are the following . edgesalong the relevant directions. thus there is no way to encodeEnglish locutions like pass " " " " the cup to the left . the particular direction can be specifiedperiphrastically. show that Tenejapans .2 Third . and traversing (going east or west) . however. the useof the systempresupposes a good sense of direction . together with its insistent use in spatial description. It applies to objectson the horizontal as well as on slopes . and uphill . Hencedownhill hascome to mean (approximately) north . This is in essence it has certain peculiarities. First . or take the first right Turn. make the three-way distinction an important feature of language use. by referring to landmarks. with many ridges and crosscut ting valleys . the coordinate system is deficient. for example . the answer might be " ascending (going south). one might ask for . which neverthelessexhibits an overall tendency to fall in altitude toward the north -northwest. Thus there are relevant abstract nominals that describe directions.1 may help to make the systemclear. down. one might . the boy is in front of the tree. to the right.1 Tzeltal Absolute Linguistic Frame of Reference Tzeltal has an elaborate intrinsic system(seeBrown 1991 . if one wanted to describewhere a boy was hiding behind a tree. in that the orthogonal designatessouth. The three-way semanticdistinction betweenup. There are many other interesting features of this system ( Brown and Levinson 1993a ). and motion verbs that designate ascending (i. infront . Third . say. Second .2. Levinson 1994 ). the systemis a true fixed-bearing system . This linguistic ramification .egoing south). if one wanted to ask where designate " someonewas going. Despite this.Frames of Referenceand Molyneux ' s Question 4. the uphill one. the north (downhill ) side of the tree. another system of a cardinal direction system .

w . .. Tzeltal /downhill system Tenejapan uphill .1 Flame .dIG . Levinson a . " "The bottleis uphill of thechair. .112 StephenC. . -I ...-Ji'oI . IiI8 * Ir .8p6IlI dIG 4.1 . ~ ~ ~ at II.

The ' aim is to find tasks where subjects will reveal which frame of reference . or objectively to the north (seefigure 4. A male subject. The simple underlying design behind all the experimentsreported herecan be illustrated as follows. and after a delay. my colleagues and I have devised experimental means for revealing the underlying nonlinguistic coding of spatial arrays for memory (seeBaayen and Danziger 1994 ) . the Tzeltal linguistic system does not provide familiar viewer-centered locutions like " turn to the left " or " in front of the tree." All such directions and locations can be adequatelycoded in terms of antecedentlyfixed. or relative. the subject Left .1 Memory and Inference As part of a larger comparative project. In short. cae : TASK Choose arrow same as stimulus z 1 r RELATIVE ABSOLUTE ~ ~ .2.2. seesan array on a table (table I ) : an arrow pointing to his right . Here we concentrate on the absolute versus relative coding of arrays.2.2) .2 Use of an Absolute Frame of Reference in NonverbalTasks 4.Frames of Referenceand Molyneux ' s Question '~ fo z 1 Table STIMU r 113 2 Table r .2 . 4. work on an Australian language(Haviland 1993 . responses intrinsic . say. a seriesof experimentswere run in Tenejapa to ascertainwhether nonlinguistic coding might follow the pattern of the linguistic coding of spatial arrays. Levinson 1992b Following ) where such a linguistic system demonstrably has far -reaching cognitive consequences . absolute bearings . has been employed during the task. absolute. The array is then removed. do indeed maintain the correct bearingsof various locations as they move in the environment. Underlying design of the experiments without visual access to the environment. Figure 4.

one pointing to his right and one to his left . and as closely matched in content as could be achievedacrosslanguages Recall Memory Method The design was intended to deflect attention from memorizing direction towards memorizing order of objects in an array. Dutch like English relies heavily of course on a right/left/front / back system goes -centered coordinates for the description of most spatial arrays. one to the north and one to the south. They were run on at least twenty five Tenejapansubjects(dependingon the experiment) of mixed age and sex / sex . although the prime motive was to 3 tap recall memory for direction. Instructions (in Dutch or Tzeltal) were of the or another of the frames of reference " kind " Point to the pattern you saw before. or the wording used in training sessions . asking the subject to place an arrow so that it is the sameas the one on table I ) and various kinds of inference(as sketchedbelow) . horse. So the of speaker hypothesisentertainedin all the experimentsis the following simple Whorfian conjecture : the coding of spatial arrays. then it is clear that he coded the original array without regard to his bodily coordinates. Using the samemethod. which have rotated with him. it is clear that he coded the first arrow in terms of his own bodily coordinates. . On the other hand. . sheep ) familiar in both cultures.that is. that is. As far as the distinction coding linguistic composition . cow. He is then askedto identify the arrow like the one he saw before. and a Dutch comparison group of at least thirty -nine subjectsof similar age relative between absolute and . we expect Dutch subjectsto solve all the nonlinguistic tasks utilizing a relative frame of reference . We will describeherejust three such experimentsin outline form (seeBrown and Levinson 1993b for further details and further experiments ) . two arrows. The stimuli consistedof two identical sets of four model animals (pig. we expectTenejapan subjectsto solve the nonlinguistic tasks . From the set of four . provides a dominant relative frame of reference . but with respectto somefixed bearing or environmental feature. becauseTzeltal offers only an absolute frame of reference for the relevant arrays. recall memory (by. Here there are. as much devoid of spatial information as possible . Ifhe chooses the one pointing to his right (and incidentally to the south). pointing north (and to his left ).that is. Clearly it is crucial that the instructions utilizing an absolute frame of reference for the experiments . like English. the conceptual representationsinvolvedin a range of nonverbal tasks should employ the same frame of reference that is dominant in the languageused in verbal tasks for the same sort of arrays. . we can explore a range of different psychological faculties: recognition memory (as just sketched ). say. do not suggestone . If he choosesthe other arrow . Because Dutch . Levinson is rotated 180degreesto face another table (table 2) ." " Remake the array just as it was." " Rememberjust how it is. for example .114 StephenC.

except that Tenejapan subjectsare lessconsistent than Dutch ones. Results Ninety -five percent of Dutch subjectswere consistent relative coders on at least four out of five trials . first with correction for misorders on table I . with the stimulus always presented on table I . then after a three . then without correction under rotation on table 2.' Frames of Reference and Molyneux s Question 115 three were aligned in random order . Five main trials then proceeded .3. inconsistenciesbetween codings over trials were representedby indices in the interval. The data are displayed in the graph of figure 4. The subject was asked to identify the card most similar to the stimulus. One card was usedas a stimulus in a particular orientation . The subject saw the stimulus on table I . or to the " school" -like nature of task performed by largely unschooled subjects less dominant. where 0 representsa consistentrelative responsepattern and 100a consistent absolute pattern. and after a delay the subject was rotated and led over to table 2. As the graph makes clear. the data have been analyzed in the ' following way. while 75% of Tzeltal subjects were consistent absolute . Each subject s performance was assignedan index on a scalefrom 0 to 100 . Responses " were coded as " absolute if the direction of the recalled line of animals " " preservedthe fixed bearingsof the stimulus array . . where subjectsfrom each population have beengrouped by 20-point intervals on the index. The eight trials .4) . and as relative if the recalledline preservedegocentricleft or right direction. but is available that of reference frame interferencefrom an egocentric Only two Tenejapan subjects were consistent relative coders (on 4 out of 5 trials) . The remainder failed to recall direction so consistently. on table 2. coders by the samemeasure For the purposes of comparison across tasks. This may be due to various factors: the unfamiliarity of the situation and the tasks. RecognitionMemory Method Five identical cards were prepared. and the responserequired under rotation . and with delay. The result appears to This pattern is essentially repeated across the experiments confirm the hypothesis that the frame of referencedominant in the language is the frame of referencemost available to solve nonlinguistic tasks. on each there was a small green circle and a large yellow circle. the subject saw this card on table I . . all heading in ( a randomly assigned) lateral direction on table I . the curves for the two populations are approximately mirror images . Subjects were trained to memorize the array before it was removed " " . The other four were arrayed on table 2 in a number of patterns so that each card was distinct by orientation (seefigure 4. like this simple recall task. which was then removed .4 The trials wereconducted as follows.quarters of a minute delay to rebuild it exactly as it was .

The other two cards servedas controls. where it was made clear that sameness of type rather than token identity was being requested . whilethe subjects consistently .3 Figure 2 0 20 40 ~ Dutch (n.. Results We find the same basicpatternof resultsasin the previoustask . as shown in figure4.116 StephenC. Training was conducted first on table I . Levinson 4. Onceagain the Dutch are relativecoders . were coded as indicated in figure 4... " small circle toward me" ) was selected . the response was coded as a relative response . . while the card which maintained the fixed bearings of the circles (" small circle north " ) was coded as an absolute response ...27) 60 80 100 Estimated absolute tendency (%) Animals recall task: direction. Tenejapan (n.37 ) .5.3: if the card which maintained orientation from an egocentricpoint of view (e.g. to indicate a basic comprehensionof the task.

and piloted in Tamilnadu by Pederson( 1994 ). over 80% were absolute coders. as opposed to intrinsic . . but there is also here an additional factor becausethis experiment -west) testedfor memory on both the transverseand sagittal (or north -south and east . This makes it possible to devise a task where. " from two spatial arrays or nonverbal " premises . this volume) . e Il Jference Levelt ( 1984 ) observed that relative. The following task was designed by Eric Pedersonand relative frame of reference BernadetteSchmitt. transitive and converseinferences relations . The greater inconsistency of Tenejapan subjects may be due to the same factors mentioned above. of the Tenejapan subjects who perTenejapans are less consistent fonned consistently over 6 or more of 8 trials. Nevertheless .4 " " " " Chips recognitionl task: absolute versus relative solutions. Levinson ( 1992a ) noted support spatial that absolute spatial relations also support transitive and converse inferences(see also Levelt. ~ ca ~ . 2 table table 1 E3 ADS ~ REL ~ (. chapter 3. while the other codesboth east " .Frames of Referenceand Molyneux ' s Question 117 Figure 4. This was indeed the case . a third spatial array. or nonverbal " conclusion" can be drawn by transitive inference utilizing either an absolute or a . the linguistic absolute axesare asymmetric: one axis has axes distinct labels for the two half lines north and south. inconsistencyon the east Trasiti . If there was some effect of this and west identically (" across ) linguistic coding on the conceptual coding for this nonlinguistic task. one might expect more errors or -west axis. As mentioned above.

118 4.6) . where they are given just the blue cone A and asked to place the red cylinder C in a location consistent with the previous " " nonverbal " premises ..39) . Chips recognition 100 \ 80 60 40 20 0 10 60 80 20 40 0 StephenC.24) Estimated absolute tendency (%) " " . Levinson "'-""-h Dutch (n. if a female subject. a blue Design Subjects seethe first nonverbal premise on table 1. sees( premise 1 ) . say. point of view as. in the seconddiagram in figure 4. for example cone A and a yellow cube B arranged in a predetermined orientation .6 illustrates one such array from the perspectiveof the viewer. The top diagram in figure 4." For example .5 Figure task ... Tenejapan (n. for example subjectsare rotated again and led back to table 1. Finally .. Then " " . a red cylinder C and the yellow subjectsare rotated and seethe second premise cube B in a predeterminedorientation on table 2 (the array appearing from an egocentric .

inference Transitive C : red : t Solution ) blue A ( Relative Table 1 .the visual arrays.6 Figure .Frames of Reference and Molyneux ' s Question 119 Table 1 6 . blue A yellow B ' First premise ' 2 Table yellow B ' ' Second premise EJ red C 1 J Table ~ red C Absolute Solution ( ) blue A 4.

6.e. and the red cylinder farther south of the yellow cube on table 2.rather than exact locations on a particular table ..was the relevant thing to remember . when given the blue cone. this amounts to the reversearrangementfrom that . but more inconsistently.6) ! To seewhy this is the case . randomized across the transverse and sagittal axes ( i . view of the consider figure 4. If instead the subject utilizes " " fixed bearings or absolute coordinates. but in the other half of the trials the relative (involving notions of relative distance one in just the sameway. are of for the The reasons presumably greater inconsistency Tenejapan performance the sameas in the previous experiment: unfamiliarity with any such procedure or test -west axis lacking situation and the possibleeffectsof the weak Absolute axis (the east . with the red cylinder to the left of the blue cone (seethe " last diagram labeled " absolute solution in figure 4. and in half the trials . When transitive inferences were achieved on table I . It should be self-evident from the top two diagrams in figure 4. ' If the situation. this result can only be expectedif the subjectcodesthe arrays in terms of egocentricor relative coordinates which rotate with her. than the absolute more solution was complex Method Three objects distinct in shape and color were employed .7. In this case the absolute inference is somewhat more complex than a simple transitive inference ). thus the conclusion must be that the red cylinder is south of the blue cone. Results The resultsare given in the graph in figure 4. then ( premise 2 ) the red cylinder to the right of the yellow cube. south of the yellow cube on table I . representing the arrays seen sequentially. we have the same Dutch : subjectsare consistently pattern of resultsas in the prior memory experiments relative coders . she may be expectedto place the red cylinder C to the right of the blue cone A . no correction was given unless the placement of the conclusion was on the orthogonal axis to the stimulus arrays . the arrays were either in a line across or along the line of vision ) . There were then ten trials . the blue cone will be. which givesa bird s eye experimental subjectdoesnot usebodily coordinates that rotate with her. However. Training was conducted on table I .120 StephenC. produced under a coding using relative coordinates. why the " " " " third array (labeled the relative solution ) is one natural nonverbal conclusion from the first two visual arrays. Levinson " " the yellow cube to the right of the blue cone.in fact the reversearrangement . 90% were absolutecoders(just two out of25 subjectsbeing relative coders ). As the diagram makesclear. where it was made clear that the positions of each object relative to the other object . we can expect a different conclusion .8 Essentially . say. subjects were introduced to the rotation between the first and second premises . Of the Tenejapanswho produced consistentresultson at least 7 out of 10trials . and Tenejapan subjects strongly tend to absolute coding.

' Frames of Reference and Molyneux s Question

1 Table TASK : PllCeC " ~ A ,--ca '~ ~ ~ . ~ ' ; O / M ~ ' 1 ~I ~ ~ 1 Table 1 Table C . , '" A A ( c3 C(:---r~~
RELATIVE Response
ABSOLUTE Response
Figure 4.7 - bird ' s- eyeview of experimental situation. Transitive inference

~ III I

1 ~ Table Sub -~ " ' ~ " ~ / ~ 1 B IJ ~ -ca {:: A
~.. N

100 80 ~ 60 : c ' i 5 t c 40 ~ 20 0 10 80 40 60 20 0
122
StephenC . Levinson
--....... Dutch ( n - 39 )

..... Tenejapan (n- 25)

Estimated absolute (%) tendency

4.8 FiIUre Transitive inferencetask

distinct linguistic labels for the half lines) . Once again, Tenejapansmade most errors -west axis. or performed most inconsistently, on the east DiSC IIS S ;OIl The results from these three experiments , together with others unreported here (seeBrown and Levinson 1993b ), all tend in the samedirection. While Dutch subjectsutilize a relative conceptual coding ( presumably in terms of notions like left , right , in front , behind ) to solve these nonverbal tasks, Tenejapan subjects use an absolute . This is of coursein line with the coding predominantly coding system . The samepattern built into the semanticsof spatial description in the two languages holds across different psychological faculties: the ability to recall spatial arrays, to

' Frames of Referenceand Molyneux s Question

123

recognize those one has seen before, and to make inferences from spatial arrays. Further experiments of different kinds, exploring recall over different arrays and inferencesof different kinds, all seemto show that this is a robust pattern of results. The relative inconsistencyof Tenejapanperformance might simply be due to unfamiliar materials and proceduresin this largely illiterate , peasantcommunity . But as accumulatedon one absoluteaxis inparticular above, errors or inconsistencies suggested . However, becausethe experimentswere all run on one set of fixed bearings , the error pattern could have been due equally to a strong versus weak egocentric axis -west (and in fact it is known that the left -right axis- here coinciding with the east half the subjects than the front back axis . Therefore axis is less robust conceptually ) . were recalled and the experiments rerun on the orthogonal absolute bearings do indeed accumulate The results showed unequivocally that errors and inconsistencies -west absolute axis (although there also appears to be some interference on the east from egocentric axes ) . This is interesting becauseit shows that Tenejapan subjectsare not simply using an ad hoc system of local landmarks, or some fixed; rather, the conceptual primitives bearing systemtotally independentof the language used to code the nonverbal arrays seem to inherit the particular properties of the semanticsof the relevant linguistic distinctions. This raises the skeptical thought that perhaps subjectsare simply using linguistic mnemonics to solve the nonverbal tasks. However, an effective delay of at least three-quarters of a minute betweenlosing sight of the stimulus and responding on table 2 would have required constant subvocal rehearsalfor the mnemonic to remain available in short-term memory. Moreover, there is no particular reasonwhy subjects should converge on a linguistic rather than a nonlinguistic mnemonic (like crossing - which the fingers on the relevant hand, or using a kinesthetic memory of a gesture would yield uniform relative results ) . But above all , two other experimental results suggest the inadequacy of an account in terms of a conscious strategy of direct linguistic coding. 4.2.2.2 Visual Recall and Gesture The first of these further experimentsconcerns the recall of complex arrays. Subjectssaw an array of betweentwo and five objects on table I , and had to rebuild the array under rotation on table 2. Up to five of these , a model of a chair, a truck , a tree, a , for example objects had complex asymmetries . The majority of Tenejapan subjectsrebuilt the horse leaning to one side, or a shoe arrays preserving the absolute bearings of the axes of the objects. This amounts to mental rotation of the visual array (or of the viewer) on table I so that it is reconstructed on table 2 as it would look like from the other side. Tenejapansprove to be exceptionally good at this, preservingthe metric distancesand preciseanglesbetween objects. It is far from clear that this could be achievedeven in principle by a linguistic

124

StephenC. Levinson

coding: the precise angular orientation of each object and the metric distances between objects must surely be coded visually and must be rebuilt under visual control of the hands. This ability argues for a complex interaction between visual memory and a conceptual coding in terms of fixed bearings : an array that is visually distinct may be conceptually identical, and an array visually identical may be conceptually distinct (unlike with a systemof relative coding, where what is to the left side of the " that an visual field can be describedas to the left) . Thus being able to " see array is conceptually identical to another in absolute terms may routinely involve mental rotation of the visual image. That a particular conceptual or linguistic system may exerciseand thus enhanceabilities of mental rotation has already beendemonstrated for American Sign Language(ASL ) by Emmorey (chapter 5, this volume) . Tenejapans appear to be able to memorize a visual image of an array tagged , as it were, with the relevant fixed bearings . There is another line of evidencethat suggests that the Tenejapan absolute coding of spatial arrays is not achievedby conscious artificial use of linguistic mnemonics . , To show this, one would wish for some repetitive, unconscious nonverbal spatial behavior that can be inspected for the underlying frame of referencethat drives it . There is indeedjust such a form of behavior, namely, unreflective spontaneousgesture . Natural Tenejapan conversation can be inspectedto see accompanying speech whether, when places or directions are referred to , gesturespreservethe egocentric coordinates appropriate to the protagonist whose actions are being described , or whether the fixed bearings of those locations are preservedin the gestures . Preliminary work by PenelopeBrown showsthat such fixed bearingsare indeed preservedin spontaneousTenejapan gestures A pilot experiment seemsto confirm this. In the experiment, a male subject, say, facing north , seesa cartoon on a small portable monitor with lateral action from east to west. The subject is then moved to another room where he retells the story as best he can to another native speakerwho has not seenthe cartoon. In one condition , the subject retells the story facing north ; in another condition the subject retells the story facing south. Preliminary results show that at least somesubjectsunder rotation systematicallypreservethe fixed bearing of the observed action (from east to west) in their gestures , rather than the direction coded in terms of left or right . (Incidentally, the reversefinding has been established for American English by McCullough 1993 ) . Becausesubjects had no idea that the experimenter was interested in gesture , we can be sure that the gestures record unreflective conceptualization of the directions. Although the gesturesof course accompany , gesturespreserving the fixed bearings of the stimulus often occur speech without explicit mention of the cardinal directions, suggestingthat the gesturesreflect an underlying spatial model, at least partially independentof language .

Frames of Referenceand Molyneux ' s Question

125

4.2.3 Conclusion from theTenejapan Studies
Putting all these results together, we are led to the conclusion that the frame of referencedominant in the language , whether relative or absolute, comes to bias the choice of frame of referencein various kinds of nonlinguistic conceptual representations . This correlation holds across a number of " modalities" or distinct mental : over codings for recall and recognition memory, over representations representations for spatial inference , over recall apparently involving manipulations of visual and over whatever kind of kinesthetic representation systemdrives gesture . , images These findings look robust and general similar observations have been ; previously made for an Aboriginal Australian community that usesabsolute linguistic spatial -cultural survey over a ; Levinson 1992b description (Haviland 1993 ), and a cross dozen non-Western communities shows a strong correlation of the dominant frame of referencein the linguistic systemand frames of referenceutilized in nonlinguistic tasks (seeBaayenand Danziger 1994 ).
4.3 Frames of Reference aerna Modalities

Thus far , we have seenthat ( I ) not all languagesuse the samepredominant frame of referenceand (2) there is a tendency for the frame of referencepredominant in a particular languageto remain the predomina~t frame of referenceacrossmodalities, as displayed by its use in nonverbal tasks of various kinds, unconsciousgesture , and so on. The results seemfirm ; they appear to be replicable acrossspeechcommunities, but the more one thinks about the implications of thesefindings, the more peculiar they seem to be. First , the trend of current theory hardly prepares us for such Whorfian results: the general assumption is rather of a universal set of semantic primes (conceptual primitives involved in language ), on the one hand, and the identity or homomorphism of universal conceptual structure and semantic structure, on the other. Second , ideas about modularity of mind make it seemunlikely that such -modal effectscould occur. Third , the very idea of the sameframe of reference cross acrossdifferent modalities, or different internal representationsystemsspecializedto different sensorymodalities, seems incoherent. In order to make sense of the results, I shall in this section attempt to show that the notion " same frame of referenceacross modalities" is, after all , perfectly coherent, and indeed already adumbrated across the disciplines that study the various mod" across alities. This requires a lightning review of the notion " frame of reference the relevant disciplines (section 4.3.1 and 4.3.2); it also requires a reformation of the linguistic distinctions normally made (section 4.3.3). With that under our belts, we can then face up to the peculiarity , from the point of view of ideas about the

126

StephenC. Levinson

modularity of mind , of this cross modal adoption of the same frame of reference some intrinsic 4.4 . Here properties of the different frames of reference may ) ( section offer the decisive clue: if there is to be any cross- modal transfer of spatial information , we may have no choice but to fixate predominantly on just one frame of reference.

" of Reference 4.3.1 " SpatialFrames
" The notion of " frames of reference is crucial to the study of spatial cognition across all the modalities and all the disciplinesthat study them. The idea is as old as the hills: medieval theoriesof space , were deeply preoccupiedby the puzzle raised , for example in the river. If we think about the location moored the boat case of by Aristotle , the of an object as the place that it occupies , and the place as containing the object, then the puzzle is that if we adopt the river as frame of reference , the boat is moving, but - 201 for a if we adopt the bank as frame, then it is stationary (seeSorabji 1988 , 187 discussionof this problem, which dominated medieval discussionsof space ). " " But the phrase frame of reference and its modern interpretation originate, like . How , for so much elseworthwhile , from Gestalt theories of perception in the 1920s acrossthe the moon skims as when of motion illusions account for do we , , example clouds, except by invoking a notion of a constant perceptual window against which motion (or the perceived vertical, say) is to be judged? The Gestalt notion can be summarized as " a unit or organization of units that collectively serve to identify a coordinatesystemwith respect to which certain properties of objects, including the " 6 , 404; emphasismine) . phenomenalself, are gauged (Rock 1992 In what follows , I will emphasizethat distinctions betweenframes of referenceare , essentiallydistinctions betweenunderlying coordinate systemsand not , for example .7 In a recent review, between the objects that may invoke them. Not all will agree ) ranging over the philosophical and psychologiphilosophersBrewer and Pears( 1993 cal literature , conclude that frames of referencecome down to the selectionof reference - when I go from one room to another, do on my nose objects. Take the glasses "" they change their location or not? It dependson the frame of reference nose or s room. This emphasison the ground or relatum or referenceobject9 severelyunderplays the importance of coordinate systemsin distinguishing frames of reference , as I 10 of I : can of reference shall show below. Humans usemultiple frames happily say the ' " ' sameassemblage (ego looking at car from side, car s front to ego s left ): the ball is " in front of the car" and " the ball is to the left of the car, without thinking that the ball has changed its place. In fact, much of the psychological literature is concerned with ambiguities of this kind . I will therefore insist on the emphasison coordinate " " systemsrather than on the objects or units on which such coordinates may have their origin .

Frames of Referenceand Molyneux ' s Question

127

" acroa Modalities andthe Disciplines that StudyThem 4.3.2 " Frames of Reference
" across different If we are to make senseof the notion " same frame of reference modalities, or inner representationsystems , it will be essentialto seehow the various distinctions betweenthe frames of referenceproposed by different disciplines can be ultimately brought into line. This is no trivial undertaking, becausethere are a host of such distinctions, and eachof them has beenvariously construed, both within and acrossthe many disciplines (such as philosophy, the brain sciences , psychology, and " frames of reference " A serious review . that the notion linguistics) explicitly employ of thesedifferent conceptionswould take us very far afield. On the other hand, some sketch is essential , and I will briefly survey the various distinctions in table 4.1, with . ll somedifferent construals distinguished by the letters a, b, C " " ' " " . Newton s distinction betweenabsolute First , then, relative versus absolute space and relative spacehas played an important role in ideas about frames of refer-

4.1 Table Frames of Reference : Some Distinctions in theLiterature Spatial
" ven8 " absolute ": " Relative , brain sciences , linguistics ) (philosophy a. Spaceas relations betweenobjects versusabstract void b. Egocentric versusallocentric c. Directions: Relations betweenobjects versusfixed bearings " " " " Egocentric ven8 aUocentric and behavioralpsychology , brain sciences ) (developmental a. Body-centeredversusenvironment-centered(Note many ego centers : retina, shoulder, etc.) b. Subjective(subject- centered ) versusobjective " versus" " " " " Viewer- centered " " object- centered or 2} -0 sketch ven8 3- D models ) ( vision theory, imagery debatein psychology " ven8 " orientation " Orientation- bound -free" , imagery debatein psychology ) ( visualperception " " Deictic" ven8 " intril Ltic ) (linguistics a. Speaker - centric versusnon-speaker - centric versusthing b. Centered on speakeror addressee c. Ternary versusbinary spatial relations " versus" " " " Viewer-centered " -centered object-centered vers18 environment (psycholinguistics ) = " gazetour " versus" body tour " perspectives " " versus" route = ?" survey perspective perspective

128

StephenC. Levinson

ence , in part through the celebrated correspondencebetween his champion Clarke and Leibniz, who held a strictly relative view. 12 For Newton , absolute spaceis an abstract, infinite , immovable, three-dimensional box with origin at the center of the universe , while relative spaceis conceivedof as specifiedby relations betweenobjects. " , Newton claimed, we are inclined to relative notions: Relative space Psychologically is some moveable dimension or measure of the absolute spaces , which our senses instead of absolute . . and so determine by its position to bodies. placesand motions, " we use relative ones (quoted in Jammer 1954 , 97 98) . Despite fundamental differences thinkers in philosophy and psychology in philosophical position , most succeeding - spaceanchored to have assumedthe psychological primacy of relative space the places occupied by physical objects and their relations to one another- in our mental life. A notable exception is Kant , who cameto believethat notions of absolute , space are a fundamental intuition , although grounded in our physical experience that is, in the useof our body to define the egocentriccoordinates through which we ' deal with space(Kant 1768 ) . O Keefe and ; seealso Van Cleve and Frederick 1991 ' Nadel ( 1978 ; seealso O Keefe 1993and chapter 7, this volume) have tried to preserve this Kantian view as essentialto the understanding of the neural implementation of our spatial capacities , but by and large psychologists have considered notions of " absolute" irrelevant to theories of the naive spatial reasoning underlying language space , 380) . (Absolute notions of ; Miller and Johnson- Laird 1976 (seeClark 1973 space may, however, be related to cognitive maps of the environment discussed " " under the rubric of allocentric frames of referencebelow.) Early on, the distinction betweenrelative and absolute spaceacquired certain additional associations , relative space became associatedwith egocentric ; for example coordinate systems , and absolute space with non-egocentric ones (despite Kant 13 1768 ), so that this distinction is often confused with the egocentric versus allo centric distinction (discussedbelow) . Another interpretation of the relative versus absolute distinction , in relating relativistic spaceto egocentric space , goeson to emphasize in the different ways coordinate systemsare constructed relative versusabsolute : " Ordinary languagesare designedto deal with relativistic spatial conceptions ; with space relative to the objects that occupy it . Relativistic space provides space three orthogonal coordinates, just as Newtonian space does , but no fixed units of to extend without coordinates need , nor is there any for angle or distanceare involved " limit in any direction (Miller and Johnson Laird 1976 , 380; emphasismine) . Thus a to the relativistic " space is directions or cardinal of fixed , , opposed bearings system " whether , which Miller and Johnson-Laird egocentric or object-centered concept, 1973 like Clark other authors and ) and Svorou , ), Herskovits ( 1986 , 395) ( ( 1976 many , 213), have assumedto constitute the conceptual core of human spatial thinking ( 1994 . But because , some languagesuse as a conceptual basis coordi , as we have seen

Frames of Referenceand Molyneux ' s Question

129

nate systemswith fixed angles (and coordinates of indefinite extent), we need to " " recognizethat thesesystemsmay be appropriately called absolute coordinate systems . Hence I have opposed relative and absolute frames of referencein language (seesection 4.3.3). " versus Let us turn to the next distinction in table 4.1, namely, " egocentric " allocentric." The distinction is of course between coordinate systemswith origins within the subjectivebody frame of the organism, versuscoordinate systems centered elsewhere(often unspecified , ) . The distinction is often invoked in the brain sciences where there is a large literature concerning frames of reference(see , for example , the ) . This emphasizesthe plethora of different egocentric compendium in Paillard 1991 coordinate systemsrequired to drive all the different motor systemsfrom saccades to arm movements(see for Stein 1992 or the control of the head a as platform , , ), example for our inertial guidanceand visual systems(again seepapersin Paillard 1991 ). In addition , there is a general acceptance(Paillard 1991 , 471) of the need for a ' distinction (following Tolman 1948 ; O Keefe and Nadel 1978 ) between egocentric ' and allocentric systems . O Keefe and Nadel' s demonstration that something like Tolman ' s mental maps are to be found in the hippocampal cells is well known. 14 O' Keefe' s recent ( 1993 ) work is an attempt to relate a particular mapping systemto the neuronal structures and process es. The claim is that the rat can use egocentric measurementsof distance and direction toward a set of landmarks to compute a non-egocentric abstract central origo (the " centroid" ) and a fixed angle or " slope." Then it can keep track of its position in terms of distancefrom centroid and direction from slope. This is a " mental map" constructed through the rat ' s exploration of the environment, which gives it fixed bearings (the slope), but just for this environment. Whether this strictly meetsthe criteria for an objective, " absolute," allocentric system 15 has been questioned (Campbell 1993 , 76- 82). We certainly need to be able to " " distinguish mental maps of different sorts: egocentric strip maps (Tolman 1948 ), allocentric landmark-based maps with relative angles and distances between landmarks (more Leibnizian), and allocentric maps basedon fixed bearings (more Newtonian 16 , this is the sort of thing neurophysiologistshave in mind ) . But in any case " and " allocentric" frames of reference when they oppose" egocentric .17 Another area of work where the opposition has beenusedis in the study of human . For example , Acredolo ( 1988 ) showsthat , as Piagetargued, conceptualdevelopment infants have indeed only egocentric frames of referencein which to record spatial memories ; but contrary to Piaget (Piaget and Inhelder 1956 ), this phase lasts only for perhaps the first six months. Thereafter, they acquire the ability to compensate for their own rotation , so that by sixteen months they can identify , say, a window in one wall as the relevant stimulus even when entering the room (with two identical windows) from the other side. This can be thought of as the acquisition of a

130

StephenC. Levinson

" " " " non-egocentric . ls Pick , absolute or geographic orientation or frame of reference , 35) points out , however, that such apparently allocentric behavior can be mimicked ( 1993 ' by egocentric mental operations, and indeed this is suggestedby Acredolo s , 165 ( 1988 ) observation that children learn to do such tasks by adopting the visual " if " strategy you want to find it , keep your eyeson it (as you move) . These lines of work identify the egocentric versusallocentric distinction with the . opposition between body-centered and environment-centered frames of reference But as philosophers point out (see , for example , Campbell 1993 ), ego is not just any old body, and there is indeed another way to construe the distinction as one between . The egocentricframe of referencewould subjectiveand objective frames of reference then bind together various body centeredcoordinate systems with an agentivesubjective , distinct zones of spatial interaction (reach, being, complete with body schema " " peripheral vs. central vision, etc.). For example , phenomenalike phantom limbs or proprioceptive illusions argue for the essentiallysubjectivenature of egocentriccoordinate . systems " versus" " The next distinction on our list , " viewer-centered , comes object-centered ' from the theory of vision, as reconstructed by Marr ( 1982 ) . In Marr s well-known a of vision should take us from retinal image to visual conceptualization, theory object recognition, and that , he claimed, entails a transfer from a viewer-centered " frame of reference , with incremental processing up to what he called the 2! -D " sketch, to an object centered frame of reference , a true 3-D model or structural 19Because . we can an even when foreshortenedor viewed description recognize object in differing lighting conditions, we must extract someabstract representationof it in terms of its volumetric properties to match this token to our mental inventory of such types. Although recent developments have challenged the role of the 3-D model within a modular theory of vision,2O there can be little doubt that at someconceptual level such an object-centeredframe of referenceexists. This is further demonstrated by work on visual imagery, which seemsto show that , presented with aviewer centered perspective view of a novel object, we can mentally rotate it to obtain different perspectival " views" of it , for example , to compare it to a prototype and Metzler 1971 1980 1991 ; Kosslyn (Shepard ; Tye , 83- 86). Thus at somelevel, the visual or ancillary systemsseem to employ two distinct reference frames, viewercenteredand object-centered . This distinction between viewer-centeredand object-centeredframes of reference relatesrather clearly to the linguistic distinction betweendeictic and intrinsic perspectives discussedbelow. The deictic perspectiveis viewer-centered , whereasthe intrinsic perspectiveseemsto use (at least in part) the same axial extraction that would be neededto compute the volumetric properties of objects for visual recognition (see Landau and Jackendoff 1993 ; Jackendoff, chapter 1, this volume; Landau, chapter 8,

's and Molyneux of Reference Frames Question

131

this volume; Levinson 1994 ) . This parallel will be further reinforced by the reformation in section 4.3.3. distinctions the of suggested linguistic " " " " This brings us to the orientation bound versus orientation -free frames of reference .21The visual imagery and mental rotation literature might be thought to have . After all , visual imagery would seem to be little to say about frames of reference thus 2 D and at most necessarilyin a viewer-centeredframe of reference necessarily ! to a 3-D description) . But recently there have (evenif mental rotations indicate access beenattempts to understandthe relation betweentwo kinds of shaperecognition: one where shapesare recognized without regard to orientation (thus with no response of orientation from a familiar related stimulus), curve latency associatedwith degrees are where and another shapes recognizedby apparent analog rotation to the familiar related stimulus. The Shepard and Metzler ( 1971 ) paradigm suggestedthat only where handednessinformation is present (as where enantiomorphs have to be discriminated ) would mental rotation be involved, which implicitly amounts to some distinction betweenobject-centered and viewer-centeredframes of reference ; that is discrimination of enantiomorphs dependson an orientation bound perspective , while 22 the recognition of simpler shapesmay be orientation free. But some recent controversies seemto show that things are not as simple as this (Tarr and Pinker 1989 ; in fact tasks that rotation 1985 . and Just Cohen and Kubovy 1993 ) argue ) Carpenter ( can be solved using four different strategies , someorientation -bound and someorientation -free.23 Similarly , Takano ( 1989 ) suggeststhat there are four types of spatial information involved, classifiable by crossing elementary(simple) versusconjunctive (partitionable) forms with the distinction betweenorientation boundand orientation for rotation mental forms should bound require free . He insists that only orientation ) claim that such a view makes the recognition. However, Cohen and Kubovy ( 1993 wrong predictions becausehandednessidentification can be achieved without the . In fact, I believe that despite these mental rotation latency curves in special cases the recent controversies , original assumption that only objects lacking handedness can be recognized without mental rotation - must be basically correct for logical .24In any case reasonsthat have beenclear for centuries , it is clear from this literature that the study of visual recognition and mental rotation utilizes distinctions in frames of referencethat can be put into correspondencewith those that emerge from , for . Absolute and relative frames of referencein language , the study of language example (to be firmed up below) are both orientation -bound, while the intrinsic frame is orientation -free (Danziger 1994 ). " " " " have , long distinguished deictic versus intrinsic frames of reference Linguists " of the is in front like the sentence of a of the rather obvious ambiguities because boy " see for house ; Clark 1973 ) . It has also , 168 ; Fillmore 1971 , Leech 1969 ( , example beenknown for a while that linguistic acquisition of thesetwo readingsof terms like

132

StephenC. Levinson

in front , behind , to the side of is in the reverse direction from the developmental sequenceegocentric to allocentric (Pick 1993 ) : intrinsic notions come resolutely earlier than deictic ones (Johnston and Slobin 1978 ) . Sometimesa third term, extrinsic , is opposed , to denote, for example , the contribution of gravity to the interpretation of words like aboveor on. But unfortunately the term deictic breedsconfusions. In fact there have been at least three distinct interpretations of the deictic versus intrinsic contrast, as listed in table 4.1: ( 1) speaker -centric versusnon-speaker -centric (Levelt 1989 ); (2) centered on any of the speechparticipants versus not so centered ( Levinson 1983 ); (3) ternary versus binary spatial relations (implicit in Levelt 1984 and chapter 3, this volume; to be adopted here ) . These issueswill be taken up in section4.3.3, wherewe will ask what distinctions in framesof referenceare grammaticalized or lexicalized in different languages . Let us turn now to the various distinctions suggestedin the psychology of language . Miller and Johnson- Laird ( 1976 ), drawing on earlier linguistic work , explored the opposition betweendeictic and intrinsic interpretations of such utterancesas " the cat is in front of the truck " ; the logical properties of thesetwo frames of reference , and their interaction , have beenfurther clarified by Levelt ( 1984 , 1989 , and chapter 3, this volume) . Carlson- Radvansky and Irwin ( 1993 , 224) summarize the general assumption in psycholinguisticsas follows: Threedistinctclasses of reference frames existfor representing the spatialrelationships among -centered -centered in theworld. . . viewer centered , object , andenvironment objects frames frames . In a viewer framesof reference -centered frame in a retinocentric , objectsare represented , 's -centriccoordinatesystem head -centricor body basedon the perceiver of the perspective world. In an object -centered frame arecodedwith respect to their intrinsicaxes . In an , objects -centered environment frame with respect to salientfeatures of the , objectsare represented environment . In order to talk about space , suchas gravity or prominentvisual landmarks , verticalandhorizontalcoordinate axes mustbeoriented with respect to oneof these reference " and " to the left of " can be framesso that linguisticspatialtermssuchas " above . assigned added ) (Emphasis Notice that in this formulation frames of referenceinhere in spatial perception and : above may simply be semantically general over cognition rather than in language the different frames of reference , not ambiguous (Carlson- Radvansky and Irwin 25 , 242) . Thus deictic, intrinsic , and extrinsic are merely alternative labels for ( 1993 the linguistic interpretations corresponding, respectively , to viewer-centered , objectcentered . , and environment-centeredframes of reference There are other oppositions that psycholinguists employ, although in most cases they map onto the sametriadic distinction . One particular set of distinctions, between different kinds of surveyor route description, is worth unraveling becauseit has - 144 causedconfusion. Levelt ( 1989 , 139 ) points out that when a subject describesa

whereasLevelt ( 1989 chosenin the very processof mapping from is of reference that a frame freely argued to or perception spatial representation language(seealso Logan and Sadier. Language is flexible and it is an instrument of communication. Tversky ( 1991 Levelt' s intrinsic to call 12 this volume in and . This is exactly what one might expect. whether at a peripheral (sensory ' ) level or at a central (conceptual) level. this volume) has . seealso Taylor and . Typically .27 The view that frames of reference in linguistic descriptions are adopted in the mapping from spatial representationor perception to languageseemsto suggestthat make no useof frames of reference the perceptionsor spatial representationsthemselves : there has to be some coordinate system . . as a precondition for coordinating perception and language . but seeLevelt. On the latter conception. ( ) body or driving tour. 224) make the assumption that a frame of referencemust be adopted within somespatial representationsystem . they are different ways of conceiving the samepercept in linguistic semantics order to talk about it . and found two strategies( I ) : a gaze tour perspective . not merely terminological but results from the failure in the literature to distinguish coordinate systemsfrom their origins or centers(seesection 4. conceptual or linguistic. Thus Tversky s deictic a surveyperspective ! This confusion is. where a pathway . we seemto pattern into units that can be describedin a linear sequence window small a D 2 D or 3 . assign front . the linearization of speech requires that we chunk the . as it were. left. effectively an intrinsic perspective to used of the the direction and is found through the array . path and so on from anyone point (or location of the window in describing time) .Radvansky and Irwin ( 1993 .such frames of referenceapply . description complex " " . is that framesof reference ) seemto establish at the perceptual or spatial conceptual level do not necessarilydetermine frames of referenceat the linguistic level.thus it naturally allows us.3. this volume) . ) opts Tversky chapter Tversky press " " " " his deictic perspective perspectivea deictic frame of reference or route description and ' s " intrinsic " or " ' " " " 26 is Levelt . there seemsto Finally .'s and Molyneux Frames of Reference Question 133 " " complex visual pattern. arrays . I believe nondeictic perspective . at which level perceptual. in psycholinguistic discussionsabout frames of reference be someunclarity . What Levelt s results (chapter 3. frames of referencein languageare peculiar to the nature of the linear. effectively the adoption of a fixed deictic or viewer-centeredperspective 2 a and . chapter 13. propositional representation system that underlies . this volume ) has examinedthe description of 2 D arrays. Levelt of motion through units or chunks of the array (chapter 3. But this of course is not the case involved in any spatial representationof any intricacy . traversing configurations through represent into a description is converted static of the that is the array. that is. chapter 3. or sometimesovert disagreement .3) . Because can be thought of as egocentric both perspectives . . for . this ' volume) or Friederici and Levelt s ( 1990 . Thus Carlson.

Another is that somelanguages frames of referenceto replace so-called topological notions like in. Nevertheless . whosetroubling flexibility has led to various confusions. the ability to cast a description . Levinson ' . deictic versus intrinsic usages of projective prepositions. in this chapter we are concernedwith making sense commonalities.134 StephenC. prepositions like in vs. It should already be clear that there are . . many. We should seizeon these " is related to " environment-centered . especiallybecause " across modalities and . or under. we need to distinguish in discussionsof frames of referencebetween at least three levels: ( I ) perceptual. that on the appropriate construals. usesuch . Miller and . in Croalinguistic Perspective 4. The analysis of spatial terms 28 in familiar European languages remains deeply confused . There is much further pertinent literature in all the branches of psychology and brain science . behind). like many Australian ones ). " " One major upset is the recent finding that many languagesuse an absolute frame of reference(as illustrated in the caseof Tzeltal) where European languageswould " use a " relative or viewpoint-centered one (see . it is essentialto deal with linguistic framesof reference .4) . before proposing an alignment of thesedistinctions acrossthe board. of the " same frame of reference representational systems However. confusingly different classifications not to mention many unclarities and many deep confusions in the thinking behind them. perhaps predictions languages .3.. . for example Herskovits 1986 . but we must leave off here. there are someobvious common basesto the distinctions we have " " reviewed . b. and psycholinguists Clark 1973 . and (3) linguistic. Levinson I 992a . Vandeloise 1991 Johnson. Thus the various alleged universals should be taken with a great pinch of salt (in fact.g. (2) conceptual. They talk happily of topological vs. A third is that familiar spatial notions like left and right and even sometimes front and . But the truth is lesscomforting . and within strict limits (seebelow) to convert betweenthem. egocentric " " " " " " " correspondsto viewer-centered and 2. projective spatial relators (e. many of them can be directly jettisoned) . while intrinsic " " " " " " .Laird 1976 ) . -0 sketch to deictic frame. Bierwisch 1967 . for example Haviland 1993 . It is clear for example . and different construals of the sameterms. and we needto consider the possibility that we may utilize distinct frames of referenceat each level (but see section 4.3 Linguistic Framesof Reference will give the impression that the linguists literature of the linguistic inspection Cursory have their house in order. to take the other person s perspective example in one frame or another implies an underlying conceptual ability to handle multiple frames. Confident of all a third back are missing from many. Lyons 1977 . absolute maps onto object-centered or 3-D model frames of reference " and so forth . Further . and those in other languagesalmost entirely unexplored. In any case . and so on (see . on.

Lyons 1977 .55) organizes and summarizes the standard assumptionsin a useful way: we can cross . origin on speaker ) Relatum: Speaker .) . egocentricvs. linguistic. Coordinates: Deictic (i.Laird 1976 (see . It will follow that the traditional distinction deictic versus .e. giving us a good universal solution to one axis. This yields. for example .. 48. Suppose then we call the usage " deictic" just in case the coordinates are centered on the " " . In what follows I shall confine myself to linguistic descriptions of static arrays. and languagesdiverge widely in their solutions to the basic problem of how to specify anglesor directions on the horizontal . allocentric) . and I shall excludethe so-called topological notions. for example ." " relative" and " absolute. . In particular. three main frames of referenceemergefrom thesenew findings as solutions Essentially to the problem of description of horizontal spatial oppositions. Miller and Johnson.29Moreover. We may start by noting the difficulties we get into by trying to make the distinction between deictic and intrinsic . neurophysiological. and psychological literatures. for one very good reason . Clark 1973 . All this requires some intrinsic collapses explanation. Linguistic framesof referencecannot be defined with respect to the origin of the coordinate system (in contrast to . we shall focus on what we seemto needin the way of coordinate systemsand associated referencepoints to set up a cross linguistic typology of the relevant frames of reference . Indeed. and (b) whether the relatum (ground) is the speakeror not. Thesedevelopmentscall for some preliminary typology of the frames of reference that are systematicallydistinguished in the grammar or lexicon of different languages (with the caveat that we still know little about only a few of them) . Levelt ( 1989 . for example . They are appropriately named " intrinsic .classify linguistic usesaccording to (a) whether they presumethat the coordinates are centeredon the speaker(deictic) or not (intrinsic). intrinsic otherwise.theseare not opposedterms. but . the linguistic frames of referencepotentially crosscutmany of the distinctions in the philosophical. This is not whimsy: the perceptual cuesfor the vertical may not always coincide." even though theseterms may have a somewhatdifferent interpretation from someof the construals reviewedin the section above. for which a new partial typology concerning the coding of concepts related to in and on is available (Bowerman and Pedersonin prep. But they overwhelmingly converge the two horizontal coordinates are up for grabs: there simply is no corresponding force like gravity on the horizontal . the following classification of speaker : examples ( I ) The ball is in front of me.3oConsequentlythere is no simple solution to the description of horizontal spatial patterns. 690) . I shall focus on distinctions on the horizontal plane.Frames of Referenceand Molyneux ' s Question 135 and assumptionscan be found in the literature that no such languagescould occur .

" 31Thus the : the notion " in front of " is here a logical structure of ( I ). origin not on speaker ) Relatum: Chair C. with a viewpoint (acting as the origin of the coordinate ). (2) and (5) have a different logical . it is the locus of the origin of the coordinates that is relevant to the traditional opposition deictic versusintrinsic . Coordinates: Intri . In contrast. Coordinates: Deictic (i. not speaker) Relaturn: Addressee (5) The ball is to the right of the lamp." inclusive of first and secondpersons .e. Coordinates: Intri .. . but just the sameholds for " the ball is in front of me/you. otherwise we would group (2) and (3) as both sharing a nondeictic relatum. (2) and (5) should be classedtogether: they have the sameconceptual structure. Similarly. where the projected angle is found by referenceto an intrinsic or inherent facet of the ground object. We might therefore be tempted simply to alter the designations . ( I ) and (4) belong together: the is the same . The ball is in front of the chair presumes(on the relevant reading) an intrinsic front and usesthat facet to define a searchdomain for the ball .speakerand addressee . with the samecoordinate systems . and label ( I ). " " " " (4). Clearly. ic (i. respectively(moreover. a relatum distinct from the viewpoint .again the origin system alternatesover speakeror addressee . in a normal " " construal of " deictic. there interpretation of the expressions arejust different origins. as far as frames of referenceare concerned . Second " " ( I ) and (4) is in fact shared with (3) . from your point of view.136 (2) The ball is in front of the tree. with arguments constituted by the figure (referent) and the ground (relatum). Levinson Stephen Clearly. The problem comeswhen we pursue this classi fication further : (4) The ball is in front of you . Coordinates: Intri _ ic (origin on addressee ) Relatum: Lamp Here the distinction deictic versusintrinsic is self-evidently not the right classification... (3). in another grouping. and (4) is the same binary spatial relation . the conceptual structure of the coordinate systemsin ( ) versus(2) and (5) .e. (2). origin on speaker ) Relatum: Tree ' (3) The ball is in front of the chair (at the chair s front ) . it would conftate the distinct conceptual structures of our groupings ( I ) and 4 . both are deictic origins ) . But this would produce a further confusion. and a referent. First . ic (origin on addressee. and (5) all deictic as opposed to (3) intrinsic .

in contradistinction to " absolute" and " intrinsic " descriptions ). ic Origin : Addressee Relatum: Addressee ' (2 ) The ball is in front of the tree Coordinates: Relative Origin : Speaker Relatum: Tree ' (5 ) The ball is to the right of the lamp. a figure. thesetwo kinds of spatial relation have quite different logical properties. the proposed classification is ' ( I ) The ball is in front of me Coordinates: Intri . Let us dub the binary relations intrinsic . ic Origin : Speaker Relatum: Speaker ' ' (3 ) The ball is in front of the chair (at the chair s front ) Coordinates: Inm. all distinct. while use of the relative system entails that they are distinct (the relation is . this volume). In fact. ic Origin : Chair Relatum: Chair ' (4 ) The ball is in front of you Coordinates: Inm. and ground. and chapter 3. the ball is in front of the tree. from your point of view Coordinates: Relative Origin : Addressee Relatum: Lamp ' (6 ) John noticed the ball to the right of the lamp For John. .Frames of Referenceand Molyneux ' s Question 137 structure: " in front of " is here a ternary relation . but the ternary " " relations relative (becausethe descriptions are always relative to a viewpoint. presuming a viewpoint V (the origin 32 of the coordinate system ). betweenFand G ). as demonstrated elsewhere . . To summarize then. Coordinates: Reladve Origin : Third person (John) Relatum: Lamp (or Tree) Note that useof the intrinsic systemof coordinates entails that relatum (ground) and origin are constituted by the sameobject (the spatial relation is binary . but only when distinguished and by Levelt ( 1984 " " in this grouped way. .

9.36 for the description of all systems Let us first define a set of primitives necessary The application of some of the primitives is sketchedin figure 4. G. three canonical exemplarsfrom each of our three main types of system . origins. instead. and origins as deictic and non deictic (or .2. reference 4. and (4 ) together.3. that is. sometimes expressionswill combine two frames. the use clas Combinations of theseprimitives yield a large family of systemswhich may be : ( I ) intrinsic frame of reference . whether the origin is speaker(or addressee ). This is obvious in the caseof the grouping of ( I ). let us proceed . Note . say. Becauseframes of reference are coordinate systems . 3 absolute and . it also clear that although the viewpoint in relative usesis normally speaker in (6' ) . say that eventhe most closely related languages(and probably systems in the details of the underlying coordinate systems will differ them within even dialects ) .3. too . and so on. Let us review them together. egocentric vs. alternatively. and viewpoint V ) . that whether the center is deictic. intrinsic side in two languageswill differ considerably in the way in which sideis in fact determined. namely.138 StephenC. absolute frames of . can you define the frames of referenceif you no longer employ the feature of deicticity to distinguish them? I will expendconsiderable effort in that direction in section4. Thus the student of languagecan expect that expressionsglossed as. we need to oppose coordinate systemsas intrinsic versusrelative.3. (2) relative sified in the following tripartite scheme . there are exactly three frames of referencegrammaticalized or lexicalized in language (often. How . but 34 often each frame will have distinct lexemesassociatedwith it ) . is simply irrelevant to this classifi' ' ' cation. It is -centric.1 The Three Linguistic Framesof Reference As far as we know . on the other. allocentric). cannot be distinguished according to their characteristic. in passing will illustrate which we of we need the primitives in table 4. I expect a measure of resistanceto this reformation of the distinctions. and so on. the critic will argue.centric party may easily be addressee Hence deictic and intrinsic are not opposed . the preferential interpretation of ambiguous lexemes presumptive origins of the coordinates. how wide and how distant . frames of reference . frame of reference ( ) frame of reference .2. a searchdomain it specifies . which illustrates . if only " " because the malapropism deictic frame of reference has become a well-worn phrase. it follows that in language but variable .3. With that caveat. Each of thesethree es a whole family of related but distinct semantic frames of reference encompass 3SIt is true to . Minimally . Levinson ternary. on the one hand. and according to a suitably catholic construal. the and their geometry. (3 ). lexemesare ambiguous over two of 33 these frames of reference . illustrated as third on a centered or even . But first we must compare thesetwo systems with the third systemof coordinates in natural language . betweenF.

X G ~ F 139 RELATIVE "He 's to theleftofthehouse .Frames of Referenceand Molyneux ' s Question INTRINSIC "He's In front of the house. examplesof the three linguistic frames of reference e .He's north of the house.1 ~ : : B i~ E G ABSOLUTE .9 Figure Canonical ." L ." ~ ~ 4." X .

for example . reflection . But in some languages . it is (apart from top and bottom. One primary coordinate systemC can be mapped from origin X to secondaryorigin X2. though widely used in the literature .140 Table 4. for example . 2. by the following transformations: . where the coordinates are determined by the inherent " features or facets of the object to be used as the ground or relatum. Levinson 1. in Tzeltal the assignment of sides utilizes a closely based on shape volumetric analysis very similar to the object-centeredanalysis proposed by Marr . by specification of two or more axes . 3. Anchoring system A = Anchor point . so that thefront (see of a TV is the side we attend to . . " " .Laird 1976 . or more often a combination of these algorithm by The procedure varies fundamentally across languages . . and so forth . Points F = figure or referent with center point at volumetric center Fc.2 Inventory of Primitives StephenC. it is much more . rotation . In English. Coordinates may be polar. with volumetric center Gc. X2 = secondaryorigin A = anchor point . G = ground or relatum. . to fix labeled coordinates L = designatedlandmark 4. For example . this frame of referenceinvolves an object" centeredcoordinate system . yielding parallel lines acrossenvironment in eachdirection Slope = fixed-bearing system Intrinsic Frame of Reference Informally . and specialarrangementsfor humans and animals) largely functional . the sketch in Miller and Johnson. by rotation from a fixed x -axis. translation . in landmark systemsA = L . or rectangular. sidedness " inherent features" . . is misleading: such phrase " facets " as we shall call them have to be . with G or V. (and possibly a combination ) to yield a secondarycoordinate systemC2. Coordinates a. while thefront of a car is the facet that canonically lies in the direction of motion . 403). such labels may or may not form a fixed armature or template of oppositions. conceptually assignedaccording to some or learned on a case -case basis . The . and with a surrounding region R V = viewpoint X = origin of the coordinate system . b. Systemof labeled angles -specific Labeled arcs are specifiedby coordinates around origin (language ).

outwards for a determined distance. and so forth . how far the cat is from the truck ) can vary greatly. In addition . as it were. But whatever the procedure in a particular language the conceptual properties of the object: its shape . in that case from the volumetric determines a direction the locus of back. and function and canonical orientation is largely 37 irrelevant (seeLevinson 1994 ) . characteristic motion and use . as it were. . for example.g. More often ' . is relevant. finding front does not predict .. or line.4OIn either case . Having found . this may be used to anchor a readymade 39 . and at the foot of . " The cat is to the frame of reference(unlike the relative one) it makesno sense " front and to the left of the truck . the chair ) . for example . within or on which the figure object can be found (as in the " statue in front of the town hall ) . and so on. sides the opposed sides . thefront . it relies primarily on objects. The relation R does not support transitive inferences . so that in the intrinsic to say. by specific shapes . Systemswith fixed armatures of contrastive expressionsgenerally require the angles projected to be mutually exclusive (nonoverlapping). The origin X of the coordinate system C is always on (the volumetric center of ) G. an object s penumbra. the notion of a region. related perhaps . but nevertheless center of the object through thefront . and " and in " heads" " feet " " horns " " roots " etc. G ) assertsthat F lies in a searchdomain extending from G on the basis of an angle or line projected from the center of G. where R typically namesa part of G. each object having parts languages determined. An intrinsic relation R (F. there may be no such fixed armature. others allow the projection of enormous search domains ( in " front of the church lie the mountains. The cat is in front of .Frames of Referenceand Molyneux ' s Question ( 1982 ) in the theory of vision. with arguments F and G. lefts. . and F may be a part of G. . Systemsutilizing single parts make no such constraints " " (cf. which can be used for spatial description. we can use the designatedfacet to extract an angle. " backs. through an anchor point A (usually the named facet R). Alternatively . In many languagesthe morphology makes it clear that human or animal body (and occasionally plant) parts provide a prototype for " " " " " " : hence we talk about the " front . The geometrical properties of such intrinsic coordinate systemsvary crosslinguis tically . in other systemof oppositionsfront / back. sides . nor converseinferences(seebelow) . canonical orientation . running far off to the horizon ) . of other " . radiating out " from the ground object. The attribution of such facets provides the basis for a coordinate systemin one of two ways. ) many languages rights 38 . Some languages require figure and ground to be in contact. the metric extent of the searchdomain designated(e. F and G may be any objects whatsoever (including ego).41 to its scale More exactly An intrinsic spatial relation R is a binary spatial relation . . or " visually continuous.

we call it the " anchor point . In that casewe have polar coordinates.g. Given A . back. seemsgenerally to be basedon the planesthrough the human body.. for example . e Frame of Reference This is roughly equivalent to the various notions of viewer-centeredframe of referencementioned above (e. clockwise rotation in 900stepswill yield side. Bill kicked the ball to the left of the goal. with quadrants counted clockwise from front to right . for example which occlusion plays a role in the definition of behind . Such a system of coordinates can be thought of as centered on the main axis of the body and anchored by one of the body parts (e.142 StephenC. The coordinate system . but it is not quite the same . A. English " The ball is to the left of the tree" is of this kind of course. translation (movement without rota - . in addition ." But coordinates need not be polar . but perhaps innocent enough. for example .. and so on. Levinson Coordinates mayor may not come in fixed armatures. and a figure and ground distinct from V. centered on viewer V. and left (Herskovits 1986 ) . Calling it " deictic " however is " " .take. we may derive a line BGc(or an arc with angle determined by the width of B). The relative frame of referencepresupposes a " viewpoint " V (given by the location of a perceiver in any sensorymodality). we call A once again the " anchor point . Because A determines the line. or indeed part of a fixed set of oppositions.g. we know which facet back is. Although the position of the body of viewer V may be one criterion for anchoring the coordinates. it thus offers a triangulation of three points and utilizes coordinates fixed on V to assigndirections to figure and ground . Because A fixes the coordinates. there can be little doubt that the deictic uses of this systemare basic (prototypical ). The mapping involves a transformation which may be 1800rotation ." Relati. with no necessary implications about the locations of other intrinsic parts. . the direction of gaze may be another. they tend to be polar. Languages . Here there is a set of four labeled oppositions." Nevertheless . side. When they do . for example . a secondary set of coordinates is usually derived by mapping (all or some of ) the coordinates on V onto the relatum (ground object) G.thus " at the entrance to the church" designatesa searcharea on that line (or in that arc). Marr ' s " 21-0 sketch. given that facet A is thefront of a building . Becausethe perceptual basis is not necessarilyvisual. back. front . calling this frame of reference " is " viewer-centered potentially misleading. with one privileged facet. the extent to may differ in the weight given to the two criteria . given that facet B is the entranceof a church and Gcits volumetric center. back/front and left/right set of half lines. giving us an up/ down. But this set of coordinates on V is only the basis for a full relative system . chest ) . back." or the psycholinguists " deictic" ). and there is no doubt that relative systemsare closely hooked into visual criteria. potentially pernicious becausethe viewer need not be ego " and neednot be a participant in the speech event. conceptually prior . and so on.

also more or lessguaranteesthe potential ambiguity of back systems(although they may be disambiguated syntactically. constituting a broad . Nor does the possession . such that when V is rotated around the array. This will get front . that is.44This typo logical implication . family of relative systems Not all languageshave terms glossing left/right . so that the Tamil sentence " The cat is on the left side of the tree" would on the relevant interpretation) mean ( " " The cat is on V ' s right of the tree.) . for rotation " " will get left and right wrong . they are used as binary relations " (as in to my specifying the location of Fwithin a domain projected from a part of G " ' " " ' " " " " left . The cat is to the left of the tree has left on ' the sameside as V s left . " at the chair s left ) . Hausa Hill 1982 a that G has G so onto ) ( . left . apart from showing the derivative and secondary nature of relative systems . as if we wrote the coordinates of Von a sheetof acetate . as left/right . at the animal s front . at the houses front .' Frames of Referenceand Molyneux s Question 143 tion or reflection). primary ( ) " front " before which the cat sits. back. so are left and right . front / back. etc.43 and (2) whether there is a ternary relation with viewpoint V distinct from G. Now . thus just as front and back are reversed glossed . mapping so that a the coordinates than rotate rather . But the point to establish here is that a large variation of systemsis definable. In Tamil . But it may not be the correct solution because sequence 42 other interpretations are possible . and placed it on G. Some languages that lack have encoded the odd isolated any such systematicrelative systemmay nevertheless " " relative notion . front / " ' in " to the left of the chair " vs. The test for a relative systemis ( I ) whether it can be usedwith what is culturally construed as a ground object without intrinsic parts. To get the English system right . we might supposethat the coordinates on V should be reflectedover the transverseplane. of such a system of oppositions guarantee the possessionof a relative system when intrinsic even in more or less terms a use such they way ( purely Many languages are primarily used with deictic centers ). not rotated. In English. as in F is in my line of sight toward G. or arguably reflection across the frontal transverseplane. and indeed more plausible. But English is also not so simple. flipped it over in front of V. and right at least in the correct polar around the secondaryorigin . (And this may suggestthat . in front of you. and many other languages translate " sentenceglossing " The cat is in front of the tree will mean what we would mean in " " English by The cat is behind the tree. the description changes(seebelow) . Thus " the cat is in front of the tree" in English entails that the cat F is between V and G on V appear to have been rotated in the coordinates the because the tree . the rotation is complete. That somerelative systemsclearly use secondarycoordinates mapped from V to G that thesemappings are by origin a meansof extending the intrinsic frame suggests of referenceto caseswhere it would not otherwise apply . languagesthat do indeed have a relative system of this kind also tend to havean intrinsic systemsharing at least someof the same terms.

and sides .so The primary coordinate systemalways has its origin on V. and front and bac on the vertical or y -axis (back has (the base ~ of ) F higher than G and/ or occluded by G ) .51Typo logical variations of such systemsinclude degreeto . but be defined. and G.4S ) Through projection of coordinates from the viewpoint V. and seebelow). and (arguably) reflection.systemsthat utilize relative systemscan be thought of as derivedintrinsic ones relations to extend and conceptual supplement intrinsic ones.46 For some languages . for example . this is undoubtedly the correct analysis . One particular reason to so extend intrinsic systemsis their extreme limitations as regards logical inference of spatial relations from linguistic descriptions. right. as if trees had inherent fronts. by rectangular coordinates on the two-dimensional visual field (the retinal projection) so that left and right are defined on the horizontal or x -axis. chapter 7.144 StephenC. but relative ones do (Levelt 1984 . Levinson the intrinsic systemis rather fundamental in human linguistic spatial description. backs. Coordinate systemsbuilt primarily on visual criteria may not be polar . where F and G are unrestricted as to type.47Thus many relative .front . back. the facets are thus named and regions projected with the samelimitations that hold for intrinsic regions. translation. and left may be assignedby clockwise normally polar. F. Terms glossedfront and back normally do involve secondary coordinates (but compare the analysis in terms of vectors by O' Keefe. this volume) . we assign pseudointrinsic facets to G. Intrinsic descriptions support neither transitive nor converseinferences . except that V and G must be distinct. chapter 3. Secondarycoordinates may be mapped from primary origin on V to secondary origin on G under the following transformations: rotation . for example rotation fromfront . Such coordinate systemsare . there may be a secondarycoordinate systemwith origin on G. although they sometimesdo (as when they have reversed application from the English usage ). Terms that may be glossed left and right may involve no secondary coordinates . with arguments V. this volume.48 More exactly A relative relator R express es a ternary spatial relation.

Speakersof such languagescan then describean array analyst to compassbearings " of . our " north . with appropriate socializa tion . or place in retinal field) are definitional of the terms. many languages make extensive . prevailing " 53 wind directions. reflection) for secondary coordinates. river drainages . one refers to the fixed direction provided by gravity (or the visual horizon under canonical orientation ) . to Nepal . the same idea of fixed directions can be applied to the horizontal . only systemswith conceptual simplicity and systems ground . fact these . for example . The conceptual ingredients for such systems are simple: the relevant linguistic expressionsare binary relators. and thus point to any named location from any other ( Lewis extraordinary accuracy 1976 . degreeof use of secondary coordinates. or culture types.) . to Australia . It perhapsneeds emphasizingthat this keeping track of fixed directions is. corresponding one way or another to directions or arcs that can be related by the .g. a spoon in front of a cup.g. People who speak such languagescan be shown to do so. which always have their origin on the In are the . " and in effect unrelated to . gaze ). " " AbsoluteFrame of Reference Among the many usesof the notion absolute frame of reference . many systems clearly abstractions and refinements from environmental gradients (mountain slopes . but we may presume that a heightened senseof inertial navigation is regularly cross 52 such are Indeed checked with many environmental clues.. They do so by fixing arbitrary fixed bearings . No simple ecological determinism will explain the occurrenceof such systems ." " east. type of mapping function (rotation . Levinson 1992b ) . across neighboring ethnic groups in similar . relative systems environments." ' south. celestial azimuths. races . Lessobviously of psychologicalrelevance . to New Guinea. These cardinal directions" may therefore occur with fixed bearings skewedat various degreesfrom . wide open desertsand closedjungle terrain) . In fact. not a feat restricted to certain ethnicities. Such a systemrequires that personsmaintain their orientation with respectto the fixed bearings at all times.' Frames of Reference and Molyneux s Question 145 which a systematicpolar systemof coordinates is available.for example . and differing degreesto which visual criteria (like occlusion. somealmost exclusive " " the horizontal . useof such an absolute frame of referenceon . for example . which can be found alternating with .. cardinal directions." and " west. with figure and ground as arguments and a system of coordinates anchored to fixed bearings . body axis vs.) of ' " cup without any referenceto the viewer/speakers location. How they do so is simply not known at the present time. differing anchoring systemsfor the coordinates (e. etc. they can dead reckon current location in unfamiliar territory with . as spoon to north / south/ east / (etc. . translation . ?) as shown by its widespreadoccurrence(in perhaps a third of all human languages from Meso-America. environments. and which occur in environmentsof contrastive kinds (e.

and similarly for many Austronesian systems . Second . found by clockwise / designated 54 rotation from north. but it is not clear which direction . for example . they are the only systemsthat fully support transitive inferences elegance acrossspatial descriptions. For example. ). Absolute systemsmay also show ambiguities of various kinds. the relevant expressions " " " " (e. our interpretation of " north " as top of a map) . the coordinate system is thought of . or to the abstractedfixed bearings . Third . . one axis is determined places and monsoons is a fixed abstracted axis but the other is determined . local stream). and the problem of unfeatured objects. in that they offer two labeled half lines " north " " south" but label both ends of the . is itself the anchor. uphill may imply further away in my field of vision. guaranteeingthat the two axes do not remain orthogonal when arrays are described in widely different . logical but psychological. where these do not coincide. regardlessof their location with respect to G. they require a cognitive overhead . local hill . (roughly. by by the location of the central mountain and thus varies continuously when one circumnavigates the island. . a north-south axis is primary . .. Other systemsmay have a primary and a secondaryaxis. some such systemsmay even have relative interpretations " " (e. the geometry of the designatedangles is variable.56 " Some systemslike Tzeltal are " degenerate . like a quasi-proper name.. and the complexities involved in mapping secondarycoordinates. Thus.g. north or south. orthogonal with the same terms. and relative ones do so only if viewpoint V is held constant (Levelt 1984 ) . Relative objectsallow the naming of facets systemsare dogged by the psychological difficulties involved in learning left/right distinctions.g. Even more confusing. Thus on Bali . each half axis having its own clear anchor or fixed central bearing. Levinson . Even where systematiccardinal systemsexist. together background will that for P which direction P is from ego' s reckoning specify any arbitrary point current locus (so that ego may refer to the location of P) .146 StephenC. some systemsmay employ true abstracted cardinal directions on one axis. One crucial question with respect to absolute systemsis how. Intrinsic descriptionsdo not do so. west . cf. It may be a polar system . often developedfrom intrinsic systemsthey display ambiguities acrossframes of reference " " (like English in front of ) .. but landmark designationson the other. namely the constant calculation of cardinal directions with a systemof dead . places of particular sociocultural importance may come to be designatedby a cardinal direction term. if we have four half lines based on orthogonal . conceptually. the differing degreesto which the asymmetriesof " " . so that .g. where the systemis abstractedout of landscapefeatures . The liabilities of absolute systemsare not .55Yet other systemsfavor no particular primary referencepoint . Intrinsic systemsare dogged by the multiplicity of object types. First . uphill or upstream ) may either refer to placesindicated by relevant local features(e. as in our north/south/ east west where north is the anchor and east south . on the other hand.

though we may think of north as a point on the horizon . " in front of the tree from Bill ' s point of view. to underline the necessity . More exactly An absolute relator R express es a binary relation betweenF and G. we must distinguish frame of referencequa coordinate systemfrom . and perhaps 135 on the other.3.it follows from Hockett ' s ( 1960 the flexible nature of linguistic reference ) design ' feature of displacement . let us first consider the logical . The geometry of the coordinate systemis linguistically/culturally variable. The " ' ball is in front of the tree (from ego s point of view) .2 " Logical Structure" of the Three Frames of Reference We have argued that . " north of the chair" .Framesof Referenceand Molyneux ' s Question 147 axes . Second . Conversely . Many Australian languageshave cardinal edge roots. Thus in the intrinsic frame one can say " in front of the chair" . Even in English. Just as relative relators can be understood to map designatedfacets onto ground " " objects (thus on the front of the tree assignsa named part to the tree)." Some of these stems can only be . northern edge analyzed as an interaction between the intrinsic facets of an object and absolute directions. so that in some systemsequal quadrants of 90 degreesmay be projected from G. we also usearcs of variable extent for informal description. Thus for the intrinsic frame. we can say. so absolute relators may also do so. which will then have unevendistribution around the horizon. as far as language is concerned . 4. while in others something more like 45 degreesmay hold for arcs on one axis. But to collect them together. The asserting X of the coordinate is origin system always centered on G. First . and of course in the relative frame. none of the three frames needhave a deictic center. " The ball is in front of me" (deictic center). deictic center qua origin of the coordinate system . The ball is north of me . we can say.3. and in the relative frame. that F can be in found a search domain at the fixed bearing R from G. F may be a part of G. the labels may describe quadrants (as in Guugu Yimithirr ). we need to show that we can in fact define the three frames of reference adequately without reference to the opposition deictic versus nondeictic center or origin . including ego or another deictic center. The literature also reports abstract systemsbasedon star-setting degrees points. We have already hinted at plenty of distinguishing characteristics for each of the three frames. for example . in the absolute frame. or they may have narrower arcs of application on one axis than the other (as appearsto be the casein Wik Mungan S7 ) . or Buhler s ( 1934 ) concept of transposeddeictic center. each of our three frames of referencemay occur with or without a deictic center (or egocentricorigin) . Still ." This is just what we should expect given . G may be any object whatsoever . for the absolute frame we can " " " say. the skeptical may doubt that this is either necessary or possible . then " affixes indicating . say.

3. Brown and experimentation (as illustrated in section 4. in spatial thinking . which hardly support any spatial inferencesat all without further assumptions(seeLevelt 1984and chapter 3.58Theseproperties have a special importance for the study of nonlinguistic conceptual coding of spatial arrays becausethey allow systematic . relative seemsto require intrinsic ). seealso Levinson 1992b Levinson 1993b . 1994 .1. for example . it seems . the direction viewpoint Von an observer ' s front or of the observer . Altogether then. In the intrinsic casewe can think of the named facet of the object as providing the anchor. these features are jointly certainly sufficient to establish the nature of the three frames of referenceindependently of referenceto the nature of the origin of the coordinate system . Linguistic expressionsmay be specializedto a frame of reference . and F2 is north ofF l ' then F2 is north of G ).3. 4. G is south ofF ) .148 StephenC. some use predominantly one only (absolute or intrinsic . We may conclude this discussionof the linguistic frames of referencewith the following observations : I . But absolute and intrinsic are distinguished in that absolute relators define asymmetric transitive relations (if F1 is north of G. with the anchor being constituted by. In this case . and . 2. while some useall three. just three frames of reference relative.3 Realigning Frames of Referenceacroa Disciplines and Modalities Weare now at last in a position to seehow our three linguistic frames of referencealign with . 3. in the relative casewe can think of the . From this. Certain important properties follow from the nature of the anchoring system in each case . some usetwo (intrinsic and relative. certain distinct properties under rotation emergeas illustrated in figure 4. so we cannot assumethat choice of frame of referencelies entirely outside language . thus fixing the coordinate system . Levinson properties. The samedoes not hold for intrinsic relators. But spatial relators may be ambiguous (or semantically general ) acrossframes.3. Danziger 1993 ). and often are. absolute and relative relators share logical features becauserelative relators support transitive and converseinferences provided that viewpoint V is held constant. The absolute and intrinsic relators sharethe property that they are binary relations whereasrelative relators are ternary. Pederson1993 . intrinsic . while in the absolute caseone or more of the labeled gaze fixed bearings establish es a conceptual " slope" across the environment.10. we may summarize the distinctive features of each frame of reference as in table 4. say. this volume) . or intrinsic and absolute ). whereconverses can be inferred (if Fis north of G. Languages use : absolute. as some have suggested . we may add further distinguishing factors. Although this is already sufficient to distinguish the three frames. Not all languagesuse all frames of reference .

NZ ~ yes yes " chair of north to ball " Absolute Z ~ A yes no " chair of left to ball " Relative 0 5 ! l o JJ yes fj " chair of front in ball " description ? description Intrinsic same same ground viewer Frames of Referenceand Molyneux ' s Question Rotation of: 149 .no no yes no description ? ? same array whole object Fiaure 4.10 Properties of the frames of referenceunder rotation .

But this mismatch origin system ego is in fact just as it should be. or fails to coincide. The motive.1). we may assigna definitionally . we need to fix the origin of the coordinate systemso that it coincides . particular from various kinds of nonlinguistic thinking to linguistic conceptualization. An immediate difficulty is that . we . but not . So to realign the linguistic distinctions with distinctions made across other modalities. though not necessarilyegocentric . but at the level of language more generally at the level of conception. we have a between and the various modalities where the . designed(as it were) so that one may expressother persons points of view. We may do so as follows. At the level of perception.59 Notice then that .4. origin and coordinate system come . These assignmentsshould be understood as special subcases of the usesof the linguistic frames of reference . and so on. Languageis a flexible instrument of communication ' . under the restriction concerning the nature of the origin : . Third . may concedethat the relative frame of reference is prototypically so. First . with ego in each frame of reference .150 StephenC. and perhaps most arbitrarily . If we make these restrictions. Second . we may note that the intrinsic systemis typically . openedup gulf language perceptual of the coordinate is so often fixed on some -center. Levinson Intrinsic Relation is Origin on Anchored by Transitive? Constant under rotation of whole array? viewer? ground? Yes Yes No Absolute Relative ternary viewpoint V A within V Yes if V constant No No Yes binary ground A within No binary ground " " slope Yes No Yes Yes the other distinctions in the literature arising from the consideration of other modalities (as listed in table 4. let us remember . they can vary freely and combine. take other perspectives . and perhaps presumably prepackagedas a whole. is to try to make senseof " acrossmodalities and in the very idea of " sameframe of reference . then we can align the linguistic frames of reference with the other distinctions from the literature as in table 4. non-egocentric non-egocentric origin to the absolute system . by establishingthat frames of referencein language should be consideredindependently of the origin of the coordinate systems .

so that we can talk of the leading car in a convoy as " the head of the line. while in either of the orientation -bound systemsit is inevitable. In certain respects . 2. The object-centerednature of the intrinsic systemhooks it up to Marr ' s ( 1982 ) 3-D model in the theory of vision.4 Table Classifications of Frames of Reference Aligning Intrinsic Origin ~ ego Object-centered Intrinsic perspective 3-D model Allocentric Orientation -free Allocentric Orientation -bound Absolute Origin ~ ego Environment-centered Relative Origin = ego Viewer-centered Deictic perspective 21-D sketch Egocentric Orientation -bound I . Intrinsic and absolute are grouped as allocentric frames of reference . as opposed to intrinsic . even the existing distinctions that have been proposed can be seenin many . while the intrinsic framework is Leibnizian. Thus the spatial frameworks sequence in the perceptual systems can indeed be correlated with the linguistic frames of reference . specifyingits orientation with respectto external coordinates. This correctly captures our theoretical intuitions . where the origin may happen to be ego but neednot be. and the nature of the linguistic expressionsinvolved that the intrinsic framework is a generalization from the analysis of objects suggests into their parts. absolute and relative frameworks are fundamentally similar on another dimension because they both imposea larger spatial framework on an assemblage .they are binary relations that are .Frames of Reference and Molyneux ' s Question 4. they viewpoint -independent are allocentric systemsthat yield an ego-invariant picture of the " world out there. thus in an intrinsic framework it is impossible to distinguish enantiomorphic pairs." On the other hand. A whole configuration can be seenas a singlecomplex object. as opposed to the egocentricrelative system . Absolute and relative are grouped as orientation -bound. absolute and intrinsic viewpoints are fundamentally similar. To summarize . the viewer-centerednature of the relative framework connectsit directly to the of 2-D representationsin the theory of vision. which is orientation -free. Indeed." On the other hand. I have sought to establish that there is nothing incoherent in the " acrossmodalities notion " sameframe of reference or inner representationsystems .6O Absolute and relative frameworks a Newtonian or Kantian spatial envelope presuppose .

would he recognizethe selfsameobjects under his new sphere 61 perceptual modality or not? The question whether our spatial perception and conception is modality -specificis as alive now as then. and so on)? output systems(touch. Which brings us to Molyneux s question. 's 4. It turns out that the three distinct frames of referenceare " untranslatable" from one to the other. we may ask whether the sameframes of referencecan in principle operate acrossall . whether perceptual or conceptual .. and intellectually profitable . Notice that this view conflicts in such a way that they have cross with the views of some that frames of referencein languageare imposedjust in the . must involve a frame of reference . for example .4 Molyneux Question In 1690William Molyneux wrote John Locke a letter posing the following celebrated question: If a blind man. movement. On the one hand. ( ) soberly. retinotopic imagesjust are. sight-restored individuals take a while to adjust . On the contrary. who knew by touch the difference between a cube and a . it seemsthat there is some innate supramodal system observable in monkeys and infants. gaze . . Klatsky and (e. Is there one central spatial model. once the special flexibility of the linguistic systemswith respectto origin is taken into account. in a viewer-centeredframe of reference But at least one major problem remains. willy nilly . Levinson detailed ways to correlate with the revised linguistic ones.731). language There have of course been attempts to answer Molyneux directly. and touch and vision are attuned to different properties (Howard 1987 is more attuned to weight and texture than shape . but it -modal thinking may even be dependent may be very restricted.g. neurophysiology suggestsdirect cross . immediately extrapolate from touch to vision (Meltzoff 1993 and the ). and from which instructions can be generated appropriate to the various . 730. Here I want to suggestanother way to think about this old question. Thus it should be possible . and sophisticatedcross 62 on language . Put simply. throwing further doubt on the idea of correlations and correspondences acrosssensoryand conceptual represen ' tationallevels . 94. monkeys reared with their own limbs masked from sight have trouble relating touch to vision when the mask is finally removed . Valvo 1971 (Gregory 1987 ). ). but seealso Stein 1992 wirings ( Berthoz 1991 ). to which all our input senses report. 81. so that some feel that the answer ' " Eilan 1993 237 . the tactile sense on the other hand human neonates Lederman 1993 .96. to formulate the distinct frames of reference -modal application .152 StephenC. I mapping from perception to languagevia the encoding process shall presumethat any and every spatial representation . but the results are conflicting . had his sight restored. More to the question is a " resounding ' yes .

for example . The same holds for an absolute description. and supposeyou code the scenein an intrinsic frame of reference " bottle in front of chair " . a kinaestheticsystemrelated to gesture language ' two related then becomes of s 1989 . but on a moment' s reflection. be determinate with respectto shape imagistic mode must. " What we should mean by " modality here is an important question. let us consider question 2. tactile warmth . Knowing that the bottle is at the front of the chair will .11 ) . as it . Now I ask you to happens " " : remember it . whether at least they can be translated into one another. self-evident fact: you cannot freely convert information from one framework to another. If so. for example . So without a " coding" or specification of the locus of the viewpoint V. be modality -specific perhaps . from this intrinsic description.dependingon your viewpoint . the bottle is also north of the chair (see figure 4. It is immediately obvious that .Framesof Referenceand Molyneux ' s Question 153 the modalities. All this would seemto rule out . you cannot later generatea relative descriptionif you were viewing the array so that you faced one side of the chair . and if not . discarding other information . Jackendoff 1991 ) . cannot be mode need not and . texture. Consider. there are specialized central representationalsystems for example. and so on (see . What hybrid monster would a a single supramodal spatial representation system such have to be to record disparate information ? All that representation system concernsus here is the compatibility of frames of referenceacrossmodalities. Do the different representationalsystemsnatively and necessarilyemploy certain frame~ nf reference ? 2. a propositional systemrelated to . you cannot generate a relative description from an intrinsic description. There is a striking . the hapticperhaps propositional kinesthetic modality will have available direct information about weight. and three-dimensional shapewe can only guessat from visual information ). and more generally " " to input/output systems . it seems 63 SO . In what follows I shall assumethat corresponding to (some of ) the different senses . while those in a . an imagistic systemrelated to vision. then the bottle would be to the left of or to the right of the chair . spatial representationsin an . Similarly . too . while the directional and inertial information (Klatsky and Lederman 1993 from the vestibular systemis of a different kind again. Suppose . . and the answerto it offers an indirect answerto question I . . in principle . that you view the array from a viewpoint such that the bottle is to the right of the chair. an array. with a bottle on the ground at the (intrinsic ) front side of a chair. . This is First . for example . can representationsin one frame of referencebe translated (converted) into ? another frame of reference Let us discount here the self-evident fact that certain kinds of information may . Our version Molyneux question : questions I . Levelt . translatability across frames of reference the easierquestion.

Levinson C '~ IOJ Z oS ~ c ~ RELATIVE bottle in front of chair INTRINSIC .154 ABSOLUTE k.Y L R ~ Lft ~ of ch bot to rig StephenC.-~ ---~ -""" .

More generally. ' The syllogism suggest . Indeed. in principle . For example . translations . visual) . you cannot get from an intrinsic description. then you can derive an intrinsic description. impossible This simple fact about translatability across frames of reference may have far. we can gestureabout what we have seen .the blind man upon seeing for the first time will not recognize by sight what he knew before by touch. a clearly false conclusion.g.for that . we can talk about. the only directions in which you can convert frames of referenceare. because the orientation of the ground object is irrelevant to the orientation -bound descriptions ." and likewise from the absolute description The chair is facing " north and the bottle to the north of the chair. Nonnally . By the samereasoning " " . representations What about conversionsbetween the two orientation -bound frameworks? Again .. you do not know what cardinal direction the bottle " lies in .. . though. for example .. Each senseutilizes its own frame(s) of reference(e. In short. in all other directions are in principle out . nor from " the bottle is north of the chair can you derive a viewpoint -relative " " description like to the left of the chair.e. Consider. that is. from the two orientation -bound frames(relative and absolute) to the orientation -free one (intrinsic ) . by touch alone. that the answer to Molyneux s question is no. a representationin one framework is not freely convertible into a representationin another). 2. the following syllogism: reaching consequences I . basedon the appreciation of form through three-dimensional grasping) . this remainsa translation only in principle . Framesof referenceare incommensurable(i. touch arguably uses primarily an object-centered frame. in fact a reductio ad absurdum. We can indeed fonD mental images of contour shapesexplored . haptic) cannot be freely translated into representationsin another (e. then..g. 3. Representationsfrom one modality (e.g. I take this to be a counterintuitive result. you can.64 For if the orientation of the ground object is fully spe cified. we will not be able to exchangeinfonnation across any internal representationsystemsthat are not basedon one and the sameframe of reference . From the relative description or coding " The it is clear that no conversion is possible " bottle is to the left of the chair .' Framesof Reference and Molyneux s Question 155 not tell you whether it is north or south or east or west of the chair. from the relative " The chair is the the bottle is to to and right of the chair facing my right description " in the same plane. in principle . while vision primarily uses a viewer-centered frame.an orientation -free representation to either of the orientation bound . you will need ancillary infonnation . arrive at the " ' intrinsic specification " The bottle is at the chair s front .

may be capable of adopting different frames of reference . proprietary frame of reference . and so on. Thus when we (presum ably) use the visual system to imagine rotations of objects. For example . he will not later be able to describeit . if a Tenejapan systemswithin a limited set of frames of reference man sees an array and remembersit only in terms of a viewer-centeredframework. On the one hand. To account for those facts. theseconspire to require that a speakercode spatial perceptionsat the time of experience ' s dominant in whatever output frameworks the speaker languageoffers. this seemsthe correct conclusion. either the frame of referencemust be the same -modal sharing of information or each acrossall sensorymodalities to allow the cross . for example . those involved in making spatial inferences . Thus more central. and (c) it may be desirable to describeany spatial experiencewhatsoever at some later point . gesture and language and so on. in principle. Thus the facts that (a) frameworks are not freely convertible. be used acrossdifferent kinds of internal representationsystems . Becausepremise I seems self-evidently true. levels of representation seemcapable of adopting more than one frame of reference . then. Representationalsystems of different kinds. (b) languagesmay offer restricted frameworks as output . more conceptual. we project from 3-D models (intrinsic ) to 2! -D (relative) ones . But. those involved in nonverbal memory. . we must then reject premise 2. Levinson or draw. on the other hand. as well as those in involved speaking .2. Here. those involved in tasks requiring systems visual memory. it will be necessaryto assume . at a higher level. showing that both are available. what we have felt with our fingers. that individual subjectsdo indeed actually utilize the sameframe of referenceacross modalities. it will not be sufficient to establish that the sameframe of referencecan. can adopt the same frame of referencewhen utilizing different representational subjects . But to account for the facts described in section 4. the assumption that each sensory modality or representational systemoperatesexclusively in its own primary . peripheral sensory . visual processing frame of reference seems to deliver 3-D analysesof objects as well as 2-D ones . In short.156 StephenC. This would explain how it is that Tenejapans . But now we have an explanation for this apparent fact: the untranslatability across frames of referencerequires individuals to stabilize their representational . is the first part of the answer to our puzzle.his languagesimply fails to provide a systematic viewer-centered frame of description. . low-level vision systemsmay operate in proprietary frames of reference 2 D know of while otoliths are restricted to a gravitational may only retinotopic arrays.those involved in generating gesture . specializedto different sensorymodalities (like visual memory) or output systems(like gesture and language ). or indeed Dutch . modality must allow more than one frame of reference Intuitively .

The frame of reference dominant in a given language " infiltrates " other modalities. Molyneux . The untranslatability across frames of reference greatly increasesthe puzzle. ' Actually . Therefore.the standardization of frames of referenceacrossmodalities in line with the began . 2. It is in this light that the findings with which we . the others must follow suit.now seemnot only less local language surprising. an affirmative answer to Molyneux s question is evidently requiredotherwise we could not talk about what we see . which suggests The secondkind of responseis an a priori argument: I . across the various " input " and " output " devices ? I have tried to sketch answersto thesepuzzles . other modalities have the capacity to adopt. Moreover.5 Conclusions 157 This chapter began with some quite unexpectedfindings: languagescan differ in the set of frames of referencethey employ for spatial description. Dts Ackaowledgme Thischapter is based on results of joint research Brownon Tzeltal . who have the research collaboratively developed programoutlined here(seealso Senft 1994 . and so on. . 2. . presumably to ensurethat speakerscan talk about what they see . or psychological faculties of quite different kinds? If it does make sense . but actually inevitable. Wilkins . or be able to " annotate" experiences with the necessary ancillary information .Frames of Reference and Molyneux ' s Question 4. all modalities must have different frames of referenceavailable. Frames of referencecannot freely " translate" into one another. which suggests a yes answer to Mr . This raisesa number of fundamental puzzles : What sensedoes it make to " across talk of " same frame of reference modalities. or adapt to . if the modality most adaptive to external influences . language . namely. in particularwith Penelope . other frames of reference a yes answer to Mr .there seemsthus to be a cross reference . feel. To do this. why should it be so? What light does the phenomenon throw on how spatial information is shared across the senses . Therefore. the options in a particular languageseemto dictate the useof frames of referencein nonlinguistic -modal tendency to fix on a dominant frame of tasks. but also with many colleagues in the CognitiveAnthropologyResearch Group. What is deeply mysterious is how -modal transfer is achieved this cross . The answersconvergein two kinds of responses to Molyneux ' s question " do the senses talk to one another?" The first kind of responseis an empirical argument: 1. adopts one frame of reference 3. Molyneux .

ideas . so that we can consider how they apply acrossdifferent perceptual or conceptual systems 8. and 1O . Pederson1994 different researchprograms challenged premature who have Institute . . But seesection 4. Seealso Rock ( ) " is voiced 7. building on an . The design of this experiment is by Eric Pedersonand Bernadette Schmitt. . but I think motivated. 25) . not through peripheral representationsystemsspecializedto modular representationsystems .4.) . I shall usethe tenn modality in a slightly special . . The contributions . a neighboring community to Tenejapa. for example debt to Levelt' s pioneering work on the the and Bowerman. But this is terminol . earlier design describedin Levinson 1992b 5. In addition . most likely in -modal another inner representationsystemappropriate to another sensoryfaculty . have been woven into the text. oglcal. 4. Assuming that these sensory input systemsare modules in the Fodorean sense . Lynn Nadel. way. respectively typology and logic of spatial relations will be particularly evident) . I receivedvery helpful comments on the manuscript from Sotaro Kita . through Psycholinguistics conclusions and emboldened others (see . John Lucy. theseobjectswould be said to provide the frame of reference ( Brewerand Pears1993 . One kind of disagreement . Levelt. theseare body-part tenns with no spatial uses . They are not however to be confused with in our perceptual and motor experiences " the system of coordinateswhich abstractly represent them (emphasis ) . . for our purposeswe wish preciselyto abstract out the properties of frames of reference . 1990 Gestalt notions. etc. Thus cross -specific . we are then interestedin how the output of one module. who subsequentlydemonstratedthe existenceof the samephenomenon group by Haviland ( 1993 in Zinacantan. and David Wilkins . a crucial part of ' ' what we need to know is what those objects are. used . is related to the output of some other module. Rock ( 1992 . and criticisms of other participants at the conferenceat which this paper was given . Although there are phrasesdesignatingleft -hand and right -hand. I am also indebted to colleaguesin the wider . 3. but still sense . es process 2. not all of which I have beenable to adequatelyrespond to. Hill 1994 ) . " When placesare individuated by their spatial relation to certain objects.158 1993 . chapters2. The design of this experiment was much improved by BernadetteSchmitt. The phenomenonof fixed bearingsin gesturewas first noticed for an Australian Aboriginal ). When psychologists -modal" effects in mind transfer of infonnation acrosssensory have talk of " cross . they " " modalities (vision. and Bernadette SuzanneGaskins. touch. in this volume Bierwisch. 3. As the tenn frame of reference is commonly " . 471) : Spatial frameworks are incorporated by Paillard ( 1991 . Danziger 1994 . in someparticular inner representation system . and Dan Slobin have beenimportant intellectual influences Schmitt and Laszlo Nagy have contributed to experimental design and analysis. not nearly exclusivelyfor objects in contiguity ' a relative one basedon the speakers viewpoint (seeLevinson 1994 ). Mary Peterson Notes 1. 6. effectscan be assumedto occur through communication betweencentral. while body-part tenns for face and back are used for spatial description and then on the basisof an intrinsic assignment . my thanks to them and the organizers of the conference Finally . which built directly on the ) is here commenting on Asch and Witkin 1948 .

Some notion of absolute spacewas already presupposedby Descartes coordinate systems . . Much behavioral experimentation on rats in mazeshas led to classifications of behavior . presumably similar to Marr ' s . who argue that allocentric behavior can always be mimicked through egocentric computations: " Perhapslanguage . This opposition is . respectively .472) has a broader notion of frames of reference than most brain scientists (and closer to psychological ideas ). a Euclidean spaceinclusive of both body and ( 1982 ) object-centeredsystem object. 14. egocentric frames of reference example ). Brewer and Pears( 1993 . presuming upright posture for action. but none of it all. Specifying a frame of reference would have to do with specifying how co-ordinates are to be assignedto eventsin the world on the basis of their spatial relations to certain objects. O' Keefe and Nadel' s 1978 classification. for example . SeeO' Keefe and Nadel 1978 . after Talmy 1983 identical to themeversusre/atum. as Einstein ( 1954 . Theseobjects provide the frame of reference . Students of animal behavior. 26) consider the role of coordinate systems . 16. relative and absolute. usually egocentricversusallocentric (or geographic response types to frames of reference Pick 1988 147 156 . have noted that maps consisting of relative angles and distancesbetweenlandmarks have quite different computational properties to maps with fixed : in the former . 13. and Markus 1990 ). Miller ' and Johnson. xiv ) pointed out . Despite that . for parallel to the notions of frame of reference . Chen. Jammer 1954 . . I shall use the opposition figure versusground for the object to be located versusthe object with respect to which it is to be located. . each time landmarks are added to the map. . most rat studies fail to distinguish between these two kinds of allocentricity. though. is in terms of body position responses(cf. he proposes that there are four such frames subserving visually guided action. all organized around the geocentric vertical: ( I ) a body frame. provides the only conclusive " macroscopicevidencefor genuine allocentricity .16. (3) a world frame. 473) .see . O Keefe and Nadel 1978 ). and various other terminologies. the bearings databaseincreasesexponentially (see . Work on infant behavior similarly relates behavioral .Laird 1976 . 10.'s Frames of Reference andMolyneux Question 159 9. There are many good sketch es of parts of this intellectual terrain (see . Mc Naughton. (2) an object frame. and place responses (involving allocentric mental maps) . Thesedistinctions are seldom properly made in the literature on mental maps in humans. He even provides a rough neural " wiring diagram" (p . ) 15. feeding the object and world frames. but not the latter . Paillard ( 1991 . trajector versuslandmark. referent versusre/atum. 29). Seealso Brewer and Pears( 1993 . and (4) a retinal frame. for example . 471. II . " " 17. 14. ' s introduction of 12. This association was in part due to the British empiricists like Berkeley whose solipsism made egocentric relative spacethe basis for all our spatial ideas . but what they have to say only increases our puzzlement : " Two eventsare representedas being in the samespatial position if and only if they are assignedthe sameco-ordinates." This fails to recognizethat two distinct systemsof coordinates over the sameobjectscan describethe sameplace. cue responses (a kind of allocentric responseto an environmental gradient).

" 22..160 StePhenC. thus giving our apparent ability to rotate mental mentally rotate it (Tarr and Pinker 1989 ' ) some evolutionary raison d etre. The conceptual systemis abstract over different perceptualclues . ). " " . and generations of philosophers since have agreed (see Van Cleve and Frederick " " an external orientation . .egocentric takes place seemshighly task. and Tudor 1989 21. 20. where principal axesare identified. there is no convincing explanation of the English deictic use of front . cannot be distinguished in an object-centered(or intrinsic counterparts . say. . which is spatially represented " into one or another . " above and to the left where one perceptual clue for . or (4) purely object-centered . The equation is Tversky' s. SeeAcredolo ( 1988 ).155 the vertical (namely gravity) is missing (Friederici and Levelt 1990 . outside the " " context of maps) may also relate to a more abstract absolute spatial framework where both . (3) rotation of the perspectivepoint around the object. I am grateful to Eve Danziger for putting me in touch with this work . SeeVan Cleve and Frederick .Radvansky and Irwin s view would seem to be subtly different from Levelt s ). where the processhas hardly begun by sixteenmonths. right say. . ( 1989 26. This leap from a perspectiveimage. or worse. viewer and landmarks are embeddedin a larger frame of reference 27. as shown by the fact that " astronauts can happily talk about. who gives sixteenmonths as an end point . . AsKant 1768made clear. in facing us. Brown 1994 to Tzeltal Levinson and and for the relevance 1991 . Yet others suggestthat images (Shepard and Metzler 1971 object recognition is achieved via a set of 2! D images from different orientations (Bulthoff 1991 ) suggestwe have none of thesepowers. for example . substantial freedom in putting the perceivedstructure. For example . is possible (Marr argued) only by assumingthat objects can be analyzedinto geometrical volumes of a specifickind (generalized cones ). the cube comparisonstest can be solved by ( 1) rotation using viewer-centered coordinates. but when we say. compansons : 24. while some( Rock.g. we do not (as. 379 ) display deep confusion about frames of reference in information of handedness that one can have orientation free representations they suggest " an orientation -free frame of referenceby utilizing the notion " clockwise. seebelow in text.dependent . . a silhouette. The cat is to the left of the tree. 154 ) ) . (2) rotation around an object-centeredaxis imaged with viewer-centeredcoordinates . seealso Pick ( 1993 a route-finding task. 23. objects differing in handedness (enantiomorphs or incongruent " in Kant ' s terminology). " : we " The cat in front of the tree " as if the tree was an interlocutor " back " " left " " . but only in an external coordinate system ) frame of reference . Others have suggestedthat what we store is a 2! -D image coupled with the ability to ). Levelt ( 1989 concludes that the spatial representation itself does not determine the linguistic description: " There is . propositional format . But asKant ( 1768 ) showed . Wheeler. Carlson. hence3-D models must be of this kind . For example . Levinson 18. for ). " " 28. the notion clockwise presupposes ' ' 25. SeeDanziger 1994for possible connections to linguistic distinctions. 19. . her survey perspectivein somecases(e. Thus Cohen and Kubovy ( 1993 . actually. 1991 ). The age at which this switch to the non.

However . the requisitecoordinatesystems gap is that the factshavealwaysbeenunderdescribed . Thus Englishin as in " the moneyin the piggy bank " " of the ground object intrinsic . 36. ." Or wherethe and spatialdescriptions earthdropsawayin onedirection of mountainranges . 34. 37. The importance of the distinction betweenbinary and ternary spatial relators was pointed . out evenin the mostrecent works beingproperlyspelled or relatorshavea complexrelation to framesof 29. The so -calledtopologicalprepositions in termsof coordinate . not . rather than being distinguished semantically see for . . ). Thus properanalysis of the " topological features . Eric Pederson . like the TorresStraits . as long as it meets the necessary axial and shapeconditions. or axial properties . andso forth." But noticethat wecould havesaid" at my feet " " " " here feet or foot clearlymeans differentin eachcase the notion of . Carlson. Somepsychologistshave beentempted to presume of the . although there are reasonsfor thinking that landmark systemsand fixed-bearing systemsare distinct conceptual types. systems " relators no or coordinate information and many" topological . for example express angular and often intrinsic at or near . We know one way in which this tripartite typology may be incomplete: somelanguagesuse conventionalized landmark systems that in practice grade into absolute systems . Exceptin someplaces . . the Australian language Guugu Yimithirr has (derived) lexemesmeaning " north side of " " south side of " and so on which combine both intrinsic and absolute frames . The readermay feelthat the notion of " front" is differentfor chairsand persons (and so " is somehow moreabstract than" in front of course it is). . wherethe tradewindsroar throughwestward " and " windward can be in termsof " leeward . andin particularthat " in front of me " or " at the foot of the chair" of the chair. Thus the " face" of a stone may be the bottom surfacehidden in the soil. because " " ambiguity of English spatial expressionssuch as in front . For example . of landmarkobjects notionsinvolvespartitioning their features between noncoordinate spatialinformation and the framesof reference mentioned belowin the features of informationdistributedbetween " is an intrinsic notion basedon text. example 35. importedinto the horizontalplane 31. The reason . an intrinsicpart of the relatumobject 32. and David Wilkins ) over the details of this scheme necessarily agreewith this particular version. This point is important . notethat frames of reference areheredefined reference . I am indebted to many discussions with colleagues (especially Balthasar Bickel. English on as in " the cup on the table" would seemto combine absolute (vertical) information with topological information (contact) and intrinsic information (supporting planar surface ). Less exotically. out by Herrmann 1990 33.Radvansky and Irwin 1993 ( . gravitycanbenaturally . . that frames of referenceare imposed on language by a spatial interpretation . bottom (vertical ) information (undersurface ) andabsolute 30. ason theedges . although they would not . First. of referencein a single word. underas in the dust under the rug compounds properties .' Frames of Reference and Molyneux s Question ' s left therefore for this explanatory Tamil) meanthecat is to the tree to our right. but shares something . othersdo involve the vertical absolutedimension " .

91). Thus Miller and Johnson.and Hausa-style interpretations offront . Miller and Johnson. 44.388) . suggestingthat there are two distinct subsystems maintaining left and right always the same " " " " involved. 401). for . 47. so that it is odd to say of a cat near me that it is " in front ofa distant tree. Kita has pointed out to me) . there is no contradiction in " the cat is to the front and to the left of the tree. becauseF and G need not be in the same plane at all (as in " the tree to the left of the rising moon" ). If an object has both an intrinsic top and bottom .Laird ( 1976 ) suggestthat the notion of intrinsic region may be linked to perceptualcontiguity within 10degrees of visual arc (p . 43. Levinson 1994 ). 's 45. As shown by the intrinsic system priority in acquisition (Johnston and Slobin 1978 ) . one of which is armature-based and the other based on the location of individual facets . First . This does not seem . It may be that left and right are centeredon V. and Levinson ( 1994 ) on Tzeltal. but such anthropomorphism may be ethnocentric. On the other hand. so that it lacks an intrinsic front . Levinson 1992bon an Australian language example ). But some languagesencode relative concepts based directly on visual occlusion or the absence of it . 39. for example . Evidence for that analysis comes from various quarters. and an intrinsic front and back. while someNilotic cultures make the assumption that a tree has a front . 46. I owe the germ of this idea to Eric Pederson . 41. " the mountain to the left of the tree" ). to the left of the ball does not ascribe a left facet to the ball. My own guessis that English is semantically general over thesealternative interpretations. while English " front " and " back" can be centeredon G. thinking of English speakers treat objects as six-sided. We tend to think of human prototypes as inevitably the sourceof such prototype parts. we think of a tree as unfeatured on the horizontal dimension. . For a nice contrast betweentwo apparently similar Meso-American systems .. Second . 40. the of " intrinsic left /right " is perhaps an indication that such systemsare not exclusively possession object-centered(becauseleft and right cannot ultimately be distinguished without an external frame of reference ).g. whilefront and back are indeed rotated and have their origin on G. once again." An alternative analysis of English would have the coordinates fixed firmly on V." Incidentally. Levinson 38. and give " F is in front of the tree" an interpretation along the lines " F is between V and G " (" behind" glossing " G is between V and F " ) . away from the way it leans . English left and right are not clearly centeredon G becausesomething can be to the left of G but not in the same plane at all (e. the right analysis for English left/right. while languageslike Japanese . thesedo not have intrinsic counterparts (as S. see MacLaury ( 1989 ) on Zapotec. in Mayan languagesplant parts figure in human body-part descriptions (seeLaughlin 1975 . some allow both the English. for example . and " " intuitively . the remaining two sides are intrinsically left and right . 387.162 StephenC. some languageshardly utilize an intrinsic frame of referenceat all (see ." Above all .Laird ( 1976 : " Peopletend to . 42. but that the conceptualcounterpart to this perceptual notion of region combines perceptual information with functional information about the region drawn from social or physical interaction (pp . Note that .

54. Our current polar systemis due no doubt to the introduction of the compassin medieval " " . 52. But seeBaker ( 1989 ). Although such a systemmay be basedon a solar compass solstitial variation makes it necessaryto abstract an equinoctial bisection of the seasonal movement of the sun along the horizon . sequence ' 57. 55.. The rotation and translation casesclearly involve secondarypolar coordinates on G. Before. for helping me to systematizethese 58. maps typically had east at the top . equinox. and thefront / back terms involving rotated secondarycoordinates on G. Environmental clues will not explain how some people can exercisesuch heighteneddead reckoning abilities outside familiar territory . For the cross cultural implications and a working out of the place of absolute systems in all this. and so on . see . Thus the environmental sourcesof such not generally explain how they are used . I am grateful to David Wilkins . Reflection will havefront toward V. by inertial navigation with constant checks with visual information and other sensory information (e.' Framesof Referenceand Molyneux s Question 163 in relative descriptions hold only on the presumption 48. wind source fluctuations . Translation will have back toward V. Warlpiri may be a casein point . right (as in English. and clockwise from back: left . on one analysis ) . hencethe expression orient oneself . or can be thought of (as seems for English) as the superimposition of two systems . showing times. but clockwise from front : left . observations ). Note that none of these environmental gradients can provide the cognitive basis of . Fmay be a part of G.g. left (as in Tamil ) . Rotation will havefront toward V. right (as in Hausa) . 49. and clockwise (looking down on G ) fromfront : right . Danziger 1994 . 56. other languageslike Tamil useit in more far -reaching ways. it is therefore less confusing to fix the system by referenceto a mentally constituted orthogonal . as in the bark on the left (side) of the tree. Although transitivity and converseness that V is constant. who believesin faint human magnetoreception. or how the systemsmay explain their origins but do " " cardinal directions are psychologically fixed. Once the community has fixed a direction . drainage. Guugu Yimithirr would be a casein points becausethere are no elicitable associationsof or priority betweencardinal directions. . SeeVan Cleve and Frederick 1991 for discussion of this Kantian point . back. sensingwind direction ) . that our use of polar coordinates is older than the compass . I presumethat such people have been socialized to constantly compute direction as a background task. The reflection casescan be reanalyzedas defined by horizontal correct and vertical coordinates on the retinal projection . it remains in that direction abstracted systems in local landfall . 53. front . 59. Table 4.4 owesmuch to the work of Eve Danziger (seeespeciallyDanziger 1994 60. back. 51. " " 50. See Peter Sutton s ( 1992 ) description of the Wik Mungan system (another Aboriginal languageof Cape York ) . and other colleagues . Conversely . the left/right terms involving only primary coordinates on V. or even regardlessof removal of the subject from the local environment.

166 and U . ix . 28. M . cognition . M . 174 1991 . 1. The possibility of getting from a relative representation to an intrinsic one may help to ' explain the apparent inconsistency between our findings here and Levelt s (chapter 3. ). (Eds. thus suggestingthat frames of referencemight residein the mapping from spatial representation to language rather than in the spatial representation itself. and Witkin . Ettlinger 1987 . First discussedin Locke. and culture: Where to orient oneself in Belhare . 38. December . 194 199 63. Oxford : Oxford Science . A . and Danziger. Cross-linguistic perspectiveson topological spatial relations. in principle possible between frames of reference . The issuemay be lessclear than it at first seems . Brain and space Bickel. Some semantic universals of German adjectivals. ( 1994 ) . the data are compatible with an analysis whereby the spatial representation is itself in a relative frame of referenceand the mapping is optionally to an intrinsic or relative description. This would seemto explain all the data that we currently have in hand. . M . 157 Hinsdale. NJ: Erlbaum. : Brain bases and development . whereas a mapping from intrinsic spatial representation to linguistic relative representation would be in principle impossible. H . Foundationsof Language . ( 1988 . Max Planck Institute for Psycholinguistics . In J. and Pederson . Annual Report of the Max Planck Institute for . 407. 4). ( 1989 . Berthoz. . . .36. S. 8). Human navigation and magneto reception Press . ( 1992 ) . Krit Acredolo ... seeTye 1991 . 5. E. subjectswho made ellipses always presupposedan underlying uniform . McCarthy . B. ( 1991 framesfor the perceptionand control of movement . 1993 Psycholinguistics Baker. as Levelt acknowledges . this ' volume) . ( 1948 ) . Asch. 121 ( 1992 . H . In J. Bierwisch. E. 64. . Perception of the upright with displaced visual fields and with body tilted . Nijmegen. Manchester : University of Manchester ) . Eds. But . Spatial cognition chevsky Bellugi ( . See : " languageservesas a cross . General . Reprinted in Journal of Experimental Psychology .418. Bowerman. M . . -modal bridge" . References -Davis. Infant mobility and spatial development .164 StephenC. E.477. Cognitive Anthropology ResearchGroup. as here described . ( 1994 Baayen ) . Spatial operations in deixis.. even when their spatial descriptions varied between relative spatial frame of reference and intrinsic . 13 question was brought back into philosophical discussionby Gareth Evans ( 1985 ). Nijmegen. 3. Stiles ) . and Brewer 1993explicitly addressit . The mapping from relative to intrinsic is one of the two mappings. Working paper no. Molyneux ' s : Ch. Reference . Journal of Experimental Psychology . Talk given at the American Anthropological Association. In Levelt s task. Dennett 62. Essay on Human Understanding(book 2. 81. ( 1967 ) . for example . 455. Painard ) .111 (Ed. L . Studiesin spaceorientation 2. Levinson 61. San Francisco.9.) . A .). and many of the papers in Eilan .

Brown . R. CognitiveAnthropology Research . and Irwin .). Cognitive Anthropology Research Group . 1982 Bulthoff . Uphill and downhill in Tzeltal. and Levinson. 65. ( 1973 . ( 1991 ) . D. Max PlanckInstitutefor Psycholinguistics in Mopan . ( 1993 ) .. and Brewer . Max PlanckInstitute for Psycholinguistics .30. 46 . Working paper no. E. McCarthy . S. J. Max Planck Institute for Psycholinguistics . A . 28. Landy and J. Brewer ( Eds. P.).Radvansky. . 223 guage Carlson. 25. 24. 236 Eilan. H . Cognitive .).Frames of Reference and Molyneux ' s Question 165 Brewer. J. Working paper no. and B. H . place and action. In N .). 9. H . . Jammer . mental representation . Max Planck Institute for Psycholinguistics " " " " Brown. L . McCarthy . Nijmegen Group. C. In N . Workingpaperno. N. 173 . ( 1993 ) . . H . . ( 1993 . Spatial conceptualization in Tzeltal.174 to themind . ( 1991 ) . Spatial representation : Problemsin philosophy and psychology . Spatial representation : Problems in philosophy and psychology . In TE . A. The role of physical objects in spatial thinking . Cognitive Maya. E. Spatialrepresentation : Problems in philosophy andpsychology . Frames of referencein vision and lan- Clark . Oxford : Blackwell. ( 1993a ) . ( 1987 Ettlinger ).. and Levinson. . ( 1934 ) . G. Journal of Linguistic Anthropology. . Cognitive Psychology . Cognition and spacekit version 1. Oxford: Oxford UniversityPress . Nijmegen Dennett .. In MS . S. (Ed. and fiat slopes ) . R. Oxford: Blackwell . Movshon (Eds. . Moore (Ed. McCarthy : Problems in philosophy . Eilan . D . Consciousness explained 's Eilan. Computational modelsof visualprocessing . and the child. A . Shape from X : Psychophysicsand computation . time. ( 1954 ).). R. Buhler. Boston : Little.64.255 . ( 1993b ) . Jarvella and W. ( 1993 .. Anthropology ResearchGroup . Brown. Oxford : Blackwell. 351.30. Brewer(Eds McCarthy . MA: HarvardUniversityPress space -modelsensory . Nijmegen.). Explorations in Mayan cognition . Cambridge. Space and the acquisition of language . MA : MIT Press . P. and Pears . Reprinted in R. ( 1993 Danziger ). Frames of reference Brewer (Eds. . ( 1994 ). 25. In R. 6. Campbell.).74. K . .0. B. ( 1993 ).382. Oxford: Blackwell . Eilan. B. Eilan .95. As freshmeatlovessalt: The logic of possessive Danziger relationships . 30 AnthropologyResearch Group. ) Molyneux questionand the idea of an externalworld. and B. and Kubovy . ( 1991 ). Spatialrepresentation andpsychology . Introductionto M. Cognitive ) . A . R. Einstein : Thehistoryof theories . Cambridge . semantics . New York : Academic Press development Cohen. 305. . In N. .. . 46. Gregory(Ed.330. N. Concepts of space of in physics . TheOxfordcompanion . D .) . New York : Wiley.. Mental rotation . Klein ( Eds. Cross integration . 3( 1). The deictic field of languageand deictic words. C.. M . Nijmegen. Speech . Brown . and B. P.244 : Where is above ?Cognition . ( 1993 ) . .

732. Nijmegen. D . In J. ( 1993 . The development of locative expressionsin English. J. Cognition. PsychologicalReview . Serbo-Croatian . Amsterdam: Benjamins. and Turkish . Collectedpapers StephenC.45. T . Hockett . 78. Paper presented at Pacific Conference on Contrastive Linguistics and LanguageUniversals. In N . 1991 space Klatsky . : Problemsin philosophyand psychology representation . . M . Oxford : Oxford University Press Jackendoff. G.172 and individual differencesin spatial ability . 47(3). Cambridge. ( 1986 ) . Zeitschrift fUr Liter . Dordrecht: Kluwer . 4/ . und links: Das 6H -Modell . A . Italian . J. 6. University of Hawaii . Spatial by the human haptic system . Hemnann . Van Cleve and RE . Howard .) The philosophy of right and left : Incongruent counterpartsand the nature of .. Hill . S. R. 208 p . R. rechts. Toward a theory of deixis. B. Image and mind. left/right : A contrastive study of Hausa and English. Perceptionand Psychophysics . A . Spatial configurations and evidential propositions. Scientific American ). and . Brewer ( Eds.34. R. 92(2). C. ( 1985 : Accounts of mental rotation ) . Translated as On the first ground of the distinction of regions in spacein J. . Cambridge: Cambridge . ( 1971 ) . J. C. E. 191 . 253. Frederick (Eds. front / back. and Carpenter. 529. Weissenborn and W . M . Spatial reference in weightlessness factors and mental representations . 9 45. I . University Press Hill ..). January.96. Kosslyn.42. ( 1994 ). Working paper no. P. MA : Harvard University Press . Languageand spatial cognition : An interdisciplinary study of the prepositions in English. D . M . In Studies in natural languageprocessing . ( 1991 ). Levinson Fillmore . ( 1985 . Vor . Spatial coordination of the senses . and Lederman. McCarthy and B. Oxford : Clarendon Press . P. ( 1982 ) .205. Cognitive coordinate systems . Max Planck Institute for Psycholinguistics . and Levelt. Jammer. Klein (Eds. ( 1993 ) .. 89. 27. Journal ) Anchoring iconicity in Guugu Yimithirr pointing gestures of Linguistic Anthropology.166 Evans. 727. L .545. Honolulu . and Slobin. Eilan. C. M . ). 137 Kant . Gregory ( Ed. 25. ( 1990 ). Cambridge. ) .266. ( 1978 ) . Parts and boundaries. ( 1960 . Oxford : Oxford University Press Haviland . Johnston. J. 3. In R. 3( 1). Gregory. ( 1990 : Perceptual ) . aturwissenschaft Herskovits. R.). companionto the mind.. R. S. W . The origin of speech . Up/ down. Von Dernersten Grunde des Unterschiedesder Gegendenim Raume. 11. ( 1987 . Cognitive Anthropology Research Group . L . Spatial and nonspatial avenuesto object recognition . Oxford companionto the mind. ( 1980 ) .140 und Linguistik . ( 1987 ) . 117 . 203. . Just. ( 1954 : The history of theoriesof spacein physics .). Conceptsof space MA : Harvard University Press . ( 1768 ) . hinter . Journal of Child Language . Friederici. L . Here and there: Cross linguistic studies on deixis and demonstration . The Oxford ) . F. Oxford : Blackwell.

( . Smithsonian . ( 1993 ). . ) amongthe Tenejapans . ). R. Max PlanckInstitutefor Psycholinguistics -part terminology : Tzeltalbody Levinson . G. ( 1989 ). : Cambridge . and J. Vision . M. . J. ( 1969 ). ( 1976 . 217 16 265 Brain Sciences and . 119 Linguistics . 1990 . I and2. CognitiveAnthropologyResearch descriptionin Guugu Yimithirr. . ( ) great Laughlin . 4 791 856 32 .). Towards . R.. Someperceptual . S. D. N.38 .. G. 1994 Brown S. ( 1989 ). ( 1976 ). cognition . Zapotecbody part locatives MacLaury . Eds and Brewer Eilan. W. Neuroscience . andChinese Psychological presented Paper speakers . DC: R. Pragmatics and conception of spatialdescription Levinson . : Cambridge . 219 -Laird. imitation . What and where in spatiallanguage . ). A. P. W. Cambridge andperception . : Anthropology . . Immanuel Kant P. ( ). . and on.41. and Markus . ( ). " Deadreckoning . . B. B. .' Frames of Referenceand Molyneux s Question 167 " " " " and spatial Landau . ( 1994 ). K. Primerfor the field investigation 2 I 5 47. A.235 . C. psychology . of spatial : The cognitiveconsequences and cognition S. W. C. ( 1993 ).358 .. Oxford: 43 . . M. of space in the neurophilosophy : An essay O' Keefe . ( 1993 ). Brewer (Eds . . London : Longmans a semantic Leech of English description . HarvardUniversityPress -horse . ( 1983 ). R. S. N. Blackwell 64. Behavioral . van limitationson talking about space Levelt . and Jackendoff . S.). Kant and the sea and : in Problems . and Levinson . Spatialinformationand cohesion McCullough American of the the Annual Convention at . Chen . J. MA: MIT Press . The . J. Koenderink(Eds . and Johnson ). E. andB. 13 . J. In N. D. L. extensions and metaphoric : Prototypes . ( 1993 ). Language ( 1992b . Journalof Cognitive andcomputational : A neurophysiological of direction the sense hypothesis . MA: Miller. NewYork: Freeman Marr. Pragmatics . 55(2). van der Grind. Press : VNU Science Utrecht . Journalof American International . Routefinding by desertaborigines 21. Vision of English in the gesticulation . andlinguisticdescription . ( 1984 ). Oxford: Blackwell andpsychology in philosophy Problems . Cambridge UniversityPress . Society " . 3. 29. In N. Cambridge Levinson UniversityPress . 323 . . . Limits in perception Doorn. and the mind . Levinson . Semantics Lyons . Vols. 3(2). Ethos asappliedphilosophy . C. shape . Spatialrepresentation ( . Eilan. R. Lewis . J. Working paperno. landmarklearning McNaught . ( 1992a ). C. McCarthy of thepreverbal . A. 191 ' -modalperception : Cross Meltzoff . 1975 dictionary of SanLorenzo . Journalof Navigation in Australia .154 . Nijmegen Group.. Speaking . Language . B. Cambridge to articulation : Fromintention Levelt . McCarthy philosophy . Special volumeof Linguistics and objectdescription . 22( I ).202 . Spatialrepresentation infant. ( 1982 ). In A. E. J. ( 1977 ). Molyneuxs babies : . Washington Zinacantan Tzotzil . C.

J. 33." Journalof orientation . 121 Experimental (4). research . Acredolo(Eds . and B. The representation of egocentric space in the posterior parietal cortex. . July. 691. I. NewYork: Plenum . MA: MIT Press ). Spatialrepresentation : Problems in philosophy andpsychology .703 171 . Wheeler . J. language ). Kritchevsky . ( 1983 ). In N. Spatialreference toy matching games study and in . E. 404 Psychology Rock . and Inhelder .). KeganPaul -Davis Pick. Language use . ( 1995 . ( 1989 ).. Amsterdam : Benjamins . . Organizationof spatial knowledge . Perception 21 1 59. . L. Stiles ).). Rock (Ed. R. Berlin : Springer . 2.). 185 viewpoints Psychology . Comment . ( 1990 ). Jr. . H. . L.). 31 Pinker andcognition . 243 268 . Thelegacyof Soloman Asch . Talk givenat the 1stAustralianLinguistic . ( 1988 Sorabji ). Paillard . J. . ( ).. 98 99. B. G. Frank and I.406 : General . J. Canwe imaginehow objectslook from other .311 . Behavior a/ and Brain Sciences . P. ( 1994 ). Sydney . objects . NJ: Erlbaum Pick in children . ( 1991 ). . space . Brewer(Eds McCarthy . ( 1989 ).. Campari(Eds . . In H. In J. and Metzler . I. Jr. ( 1993 ). Thegrammar of space Takano of rotated forms : A theory of information types . Hinsdale 156 . Cognitive . U. Mentalrotationof three . How language Talmy space . Brainandspace Pederson in two Tamil linguisticsystems and manipulable . Stein.282 orientation : Theory .on : Theories in antiquity and their sequel . 21. J. 294 theory Pederson ascontext asmeans : Spatialcognitionand habitual . London : Routledge and . ( 1992 ) . Spatial . ( 1978 ) . In I. 15(4). NJ: Erlbaum . 701 ..700. The frameof reference . and Tudor. Oxford: Blackwell .). . Levinson O' Keefe . F. R. Pick and L. Spatialinformation A. H. language Linguistics ' . Cardinaldirections Institute . Cambridge . E. London : . aspects M. L. .. ( 1992 ). Matter Duckworth . 6( 1). andU.62. Perceptual .. . Language linguistics Melanesia -dimensional . S. L . D. and Nadel. Y. ( 1994 ). Press . In . Geographic space . ( 1988 of spatialcognitivedevelopment . .210 ? Cognitive .). Psychology structures . ( 1993 )..A case Senft in Kilivila: The Tinker .168 StephenC. 225 . Oxford : Clarendon Press . Learnability Rock . Thechilds conception of space . . andapplication . S. N. 145 cognition . Science . Bellugi (Eds . ( 1971 Shepard ). Sutton in Wik Mungan . Eilan.. Cognitive . andmot . Spatial : Brainbases anddevelopment . L. ( 1956 Piaget ).42. Oxford: Oxford Science . Hillsdale . Rock 1992 on Aschand Witkin' s " Studies in space . (Ed. ( 1989 . I. The hippocampusas a cognitive map. 25. R. Svorou .

andPinkerS. Theimagery Tye. From part to person: Natural tendencies for cognates . Cognitive 21 233 282 . (Eds. . (in press Taylor. ( 1971 : The problems and behavior ) . ( 1991 . ( 1991 ) . York . B. ( 1993 of semanticchangeand the search ) . Cognitivemapsin rats and men . 109 Tversky ). . Psychology andMotivation . B. Nijmegen. Max Planck Institute for Psycholinguistics . Cambridge .. Thephilosophyof right and left : Incongruent . 23. Cognitive Anthropology ResearchGroup . ( 1991 ) . Journalof Memory . Working paper no. Psychology in spatialdescriptions . ( 1989 ). MA: MIT Press ). A . J. Perspective & Language . RE . . A. Tolman Review .. Spatialmentalmodels of Learning 145 . . counterpartsand the nature of space Vandeloise : A casestudyfrom French. 55(4). E. and Frederick. : Representation debate andmind . C. ( 1948 . M. Mentalrotationandorientationdependence recognition . . 189 ). Psychological . H. Spatial prepositions . ( 1991 Valvo . 27.. Chicago University of . 35. D . M. of visual rehabilitation New patterns Van Cleve.) .208 . Sight restoration after long-tenD blindness .Frames of Referenceand Molyneux ' s Question 169 in shape Tarr. C. Chicago Press Wilkins . andTversky ). Dordrecht : Kluwer .

. : the linguistic In this chapter. and perceivedby eye rather than by ear.Karen Emmorey Expressedby hands and face rather than by voice. seeWinston 1995 how spatial contrasts permeate the linguistic structure of sign languageAl though the discussion is limited to ASL .1 Multifunctionality Languages In this section. object orientation . Four insight into the relation between linguistic and spatial representations of a will be examined: how functions as part linguistic system space major topics . the impact of a visually basedlinguistic systemon performance with nonlinguistic tasks. aspectsof the neurolinguistics of sign language of Space in Signed 5. I do not discussthe use of Language(ASL ) . Meier 1991 ) . but the discussionshould illustrate spaceto create discourseframes. . Sign languagesare rich and complex linguistic systemsthat manifest the universal properties found in all human languages( Lillo -Martin 1991 ). Because use of physical space they directly use spaceto linguistically expressspatial locations. and finally . signed languageshave evolved in a completely different biological medium . I describe several linguistic functions of space in American Sign . passeddown from one generation of deaf people to the next ( Klima and Bellugi 1979 Wilbur 1987 ) . Used primarily by deaf people throughout the world . they from spoken languages have arisen as autonomous languagesnot derived from spoken language and are . sign languagescan provide important . and point of view. the relative efficiency of ) at various grammatical levels (American Sign Language signed and spoken languages for overt spatial description tasks. Deaf children with deaf parents acquire sign language in much the sameway that hearing children acquire spoken language(Newport and Meier 1985 . other signed languagesare likely to share most of the spatial properties discussedhere. I will explore a unique aspect of sign languages . The list is not exhaustive (for example ).

I have found no phonological contrasts in ASL that involve left -right in signing space . the location of a sign is articulatory and does not carry any specific meaning. For example and D Ry1 differ only in where they are articulated on the body. but linguists have recently broadened the term phonology to mean the " patterning of the formational units of the expressionsystemof a natural " . At the purely phonological level.172 Karen Emmorey Figure 5. the ASL signsSUMMER . Location is one of the formational units language (Coulter and Anderson 1993 of sign language phonology. whether a sign is made with the right or left hand is not distinctive (left -handers and righthanders .2 Sign languagesdiffer with respect to the part of its phonological representation phonotactic constraints they place on possible sign locations or combinations of locations. see Corina and Sandier ( 1993 ).1 Example of a phonological contrast in ASL .1 PhonologicalContrasts Spatial distinctions function at the sublexical level in signed languagesto indicate phonological contrasts. and morphology of ASL (seebelow) . For example. 5). DRY ' - ~ OJ " ~ / Ia UGLY SUMMER .1. That is. Where a sign is articulated is stored in the lexicon as .1. Such left -right distinctions appear to be reservedfor the referential and topographic functions of spacewithin the discoursestructure. as shown in figure 5. ) . claimed to be somewhat analogous to consonants in spokenlanguage(seeSandier 1989 . For all signedlanguages . For a recent and comprehensivereview of the nature of phonological structure in sign language . UGLY . 5. syntax. Sign phonology does not involve sound patternings or vocally based features . there are no phonological minimal pairs that are distinguished solely on the basis of whether the signs are articulated on the right or left side of signing space . Furthermore.what is distinctive is a contrast betweena dominant produce the samesigns and nondominant hand) . in ASL no one-handed signs are articulated by contacting the contralateral side of the face ( Battison 1978 ) . These signs differ only in the location of their articulation .

the visual system is well suited for simultaneously perceiving a large amount of information .The Confluenceof Spaceand Languagein Signed Languages 173 ~ ~ ~ GIVE base form . or reduplication are relatively rare. . the major articulators for sign (the hands) move relatively such that the duration of an isolated sign is about 1. complex forms are most often created by nesting a sign stem within dynamic movement contours and planes in space .= . the duration of an average spoken word is more like 500 milliseconds.2 Morphological Inflection In many spoken languages . Thus both sign and speechhave exploited the advantagesof their respectivemodalities. Sign languages preference for simultaneously producing affixes and stemsmay have its origin in the visual-manual modality . jaw) can move quite rapidly .2 illustrates the base form GIVE along with severalinflected forms. the articulators for speech(the tongue. producing easily perceiveddistinctions on the order of every 50.. iterative. Generally. That is. In ASL and other signed languages prefix . for example . In contrast. or continual. then there is strong pressurefor signedlanguagesto expressmore distinctions simultaneously . The articulatory pressuresseem to work in concert with the differing capacities of the visual and auditory systems for expressing simultaneous versus sequential information . For spoken simultaneous affixation es such as . by contrast. If language processingin real time has equal timing constraints for spoken and signedlanguages . Figure 5. morphologically complex words are formed by adding es or suffix es to a word stem. For example . in the Semitic languages ).g. prefer nonconcatenative process ' prefixation and suffixation are rare. whether the action was habitual .000 milliseconds slowly . infixation . thesedistinctions are marked by different movement patterns overlaid onto a sign stem. ASL has many verbal inflections that convey temporal information about the action denoted by the verb. lips.200 milliseconds . languages process templatic morphology (e. Signed languages es such as reduplication. This type of morphological encoding contrasts with the primarily linear affixation found in spoken languages .:::::-~ ~ GIVE GIVE continuative habitual GIVE reciprocal 5.1. and . whereasthe auditory systemseems particularly adept at perceiving fast temporal distinctions.

5. Recently. other researchershave argued that these spatial loci are agreementmorphemes or clitics that are attached to pronouns and verbs (e. That is. the nominal DOG is associated with a spatial locus on the signer' s left and CAT is associatedwith a locus on the signer' s right .g. the signer deictically points to a locus in the same way he would point to a physically present person.1. 1994 . In figure 5. As evidence for his position.3. Liddell ( 1993 1995 ) arguesthat just as there is an unlimited number of spatial positions in which a .174 Karen Emmorey 8Thedog bites the cat. nominals can be associatedwith locations . ) . Further description of coreferenceand anaphora in ASL can be found in Lillo -Martin ( 1991 ) and Padden( 1988 ). Janis 1995 . 1995 ) has argued that spatial loci are not morphemic.3 Example of the sentential use of spacein ASL .3 Coreferenceand Anapllora Another hypothesizeduniversal use of spacewithin sign languagesis for referential functions. In contrast. there has been some controversy within sign linguistics concerning whether spaceitself performs a syntactic function in ASL . Another device for establishing the nominal-locus association is to articulate the nominal sign(s) at a particular location or by eye gazetoward that location. Liddell ( 1993 . dog) are first associatedwith spatial loci through indexation. The direction of the movementof the verb (BITE ) indicates the grammatical role of subject and object. as shown in figure 5. Padden 1990 . This associationcan be establishedby " indexing" or pointing in signing space to a location in spaceafter producing a lexical sign.3. In ASL and other sign languages .3 is being useddeictically rather than anaphorically. Nominals (cat. BITE belongs to a subset of ASL verbs termed agreeing verbs3 whose movement and/ or orientation signal grammatical role. ASL pronouns also make use of established associationsbetween nominals and spatial loci. A pronominal sign directed toward a specificlocus refers back to the nominal associatedwith that locus.. The verb BITE moves between these " " locations identifying the subject and object of the sentence [ Thedog] bites [the cat] .8 Figure 5. He proposesthat spacein sentences like those illustrated in figure 5.

g.4 According to this view. extremely intimate mixing of linguistic structure representation . The association of a nominal with a particular location in spaceneedsto be part of the linguistic representation at some level in order to expresscoreferencerelations between . .1. the identity of each object is provided by a lexical sign (e. which can be embedded in locative and motion predicates . the location of the objects.g. The syntactic role of subject or object is assigned . TABLE . and many of the details have not beenworked out. and Bellugi 1995 ) . but either by word order or by the orientation or the temporal end points of the verb itself. there also appears to be an unlimited number of potential locations within signing space(both vertically and horizontally ) toward which a verb or pronominal form can be directed (seealso Lillo .. For example . not by the spatial loci . If this association is not part of the linguistic a proform and its antecedent then there must be an . The spaceitself. ASL has a set of classifier fonDS for conveying specific locative infonnation . signing space is most often interpreted as corresponding to a . T -V . If this is the case and therefore cannot be agreementmorphemesor clitics. space is used nonmorphemically and deictically is much as deictic ) . CHAIR ). is not part of a syntactic representation . their orientation . The use of spaceto directly represent physical location in real (or imagined) space in contrast to spoken languages . spread tight lips with eye gaze to the locus). for these predicates . Liddell has argued. Furthennore .in Signed of Space and Language TheConfluence Languages 175 physically present referent could be located. the phraseDOG INDEX . in which spatial spatial relations stands marked infonnation must be recovered from an acoustic signal that does not map onto .4 Locative Expressio The spatial positions associatedwith referentscan also convey locative infonnation about the referent. INDEX simply establish a referencerelation between DOG and a spatial locus that happens to be on the ' signer s left. In locative expressionsin the infonnation content in a one-to-one correspondence ASL . The flat B handshape is .Martin and Klima 1990 . Corina . and nonlinguistic representationsof space 5. produced simultaneously with the INDEX sign. then location specificationsare not listable or categorizable ) . and their spatial relation vis-a-vis one another are indicated by where the appropriate accompanyingclassifier sign is articulated in the space in front of the signer. To ensurea locative reading.3 could " be interpreted as " the dog is there on my left .. shown in figure 5. but such an interpretation is not es required by the grammar. evenif space example itself does not perform a syntactic function . signersmay add a specificfacial expression (e. it does perform both a referential and a locative function within the language(seeEmmorey. the particular location at which a verb begins or ends servesto identify the referent of the subject and object roles. This hypothesisis quite ( gesture usedwhen accompanyingspeech For radical. Under the nonlocative reading. rather.

surface tables or sheetsof paper. The C handshape is the classifier handshape for bulky boxlike objects like televisionsor microwaves . ASL verbs of location incorporate detailed information about the shape of objects. chapter 8.. Landau and Jackendoff ( 1993 ) have recently argued that languages universally encodevery little information about object shapein their locative closed -classvocabulary (e. both Tzeltal and ASL express locative information through verbal predicates that form an open-class category.. and a chair to rny right . Figure 5.g. and seatedpeople. Flat B handshape : ~ C handshape : ~ Bent V handshape :~ These handshapesoccur in verbs that expressthe spatial relation of one object to another and the rnanner and direction of rnotion (for rnoving objects/people) .4. ASL usesthe actual visual layout displayed by the array of classifiersignsto expressthe spatial relations of the objects. this volume) . An " English translation of the ASL description would be I enter the roorn. Levinson 1992a ). fiat -topped. First .Karen Emmorey 5. prepositions) cornpared to the arnount of spatial detail they encode in object narnes(seealso Landau. As one can surmisefrorn our discussionand frorn figure 5. unlike prepositions (although the rnorphernesthat rnake up these verbal predicates belong to a closed class ) . It is unclear whether these languagesare counterexarnplesto Landau and Jackendoff' s clairns for two reasons .176 Room of classifie const layout Description layout using spatlallze .4 illustrates an ASL description of the roorn that is sketched at the far left. ASL appears to have a rich representation of shape in its locative expressions ." Where English uses separatewords to expresssuch spatial relations. Like the locational predicates in Tzeltal ( Brown 1991 . there is a table to rny left . -prorninent objects like the classifier handshapefor rectangular. The distinction rnay hinge on whether theseforms are con- . The bent V is the classifier handshape " " for squat. legged objects like chairs. a TV on the far side. srnall anirnals.4 Figure Example of an ASL spatial description using classifierconstructions.

me houseis near the bike..e. The bike is near the house b.5 Figure Finalclassifier of either (2a) or (2b) . in ) . it still appearsto hold. the figure tends to be smaller and more movable than the ground ' ): (from Talmy 1983 object. the sentence . ( 1) a.5 (the signer s left hand shows the classifier form for . ) has proposed several universal features that are associated with Talmy ( 1983 the figure object (i . Thus. Second . the classifier in the first phrase is held in ) . as illustrated in figure 5. the located object) and with the referenceobject or ground . In English. the figure occurs first . as shown in (2a) and (2b) (the subscripts indicate locations in space . For example . the classifier handshape representing the ( produced figure can be located with respectto the classifierhandshaperepresentingthe ground ' object.in Signed andLanguage TheConfluence of Space Languages 177 5. and the ground is specified by the object of the as the figure preposition. For ASL the degreeof shape detail is less in classifierforms than in object names . the flat B handshapeclassifier is used for both TABLE and for PAPER example the count nouns encodemore detailed shapeinformation about theseobjects than the classifier form . although the contrast is much less striking in ASL than in English. This asymmetry can be seenin the following sentences . In these examples line indicated the extended ) during the articulation of the second phrase space( by . configurati 011 -classelementsor not (seealso Talmy 1988 sideredgrammaticizedclosed . When a large unmovable entity such as a house is expressed and This same between is odd. except that the syntactic order of the figure and ground is reversedcompared to English. In this with one hand ) way. figure ground semantically asymmetry objects occurs in ASL .

?BIKE VEHICLE CLAS SI FIE Ra -CLASSI FIERneara HOUSEOBJECT to describe a seriesof fifty-six pictures . cup in midair and then draw the table beneathit. prepositions among pick ground geo . a car behind depicting simplerelations a tree almostinvariablyexpressed thegroundfirst. For example . metries At first glance in ASL. her right hand showsthe classifierfonn for BIKE). Nonetheless that fact that signers conceive . in As shown ground (3a ) and 3b and in illustrated 5. one would draw the table first and then the rather than draw the generally cup. and then the figure is through drawing locatedwithin that ground . This figure respect ground object ordering may be an effect of the -spatialmodality of sign language visual . Talmy argues specifications across between and all out different . theclassifier example ) for theground(thehouse ) doesnot appear to be more geo metric to specifications ally complexthan the figure(the bike) with respect for shape(indicatedby classifierhand or for . Thus . HOUSEOBJECT BIKE VEHICLE-CLAS SI FIE Rnear a b. HOUSEOBJECT BIKE POINTnear a b. whendrawinga pictureof a cup on a table .178 Karen Emmorey HOUSE . . ( ) figure . The final classifier is the same for either(2a ) or (2b).what differsis phrasalorder.. along . languages Thefigureis oftenconceived of asa simplepoint. expression figure but expression of the groundcannot : -CLASSI FIERa (3) a. preferto initially express like English ) also arguesthat prepositions (for languages Talmy ( 1983 ) ascribe to and .g. the grammarof ASL reflects languages of thefigureasa point with respect to a morecomplex .6 of the can be to reduced a point.7 More cross linguistic work will help -spatial modality conditionsall signedlanguages detenninewhetherthe visual to the groundand thenthefigurein locativeconstructions . it appearsthat there is no such asymmetry in (2a construction . ?HOUSEPOINTa BIKE VEHICLE-CLASSI FIERneara . I askedeight native signers6 Recently between two objects (e. For example that the English . the ground tends to be producedfirst. a dogundera chair. configuration -CLASSI FIERa (2) a. The signers with to the . He evidence that all particular geometries figure ground objects presents ' characterize the figures geometrymuch more simply than the ground . to presenta scene visually . For . and thenlocatedthe ). The ) shape spatialgeometry locativeexpression in (2a that differentially ) doesnot appearto havea linguisticelement encodes in thewaythat prepositions do in spoken figureandgroundgeometries . whereas thegroundobjectcanhave morecomplexgeometric .

8.5 Frames ASL can express . In the the most likely for both dimensions figure 5. ) describing picture five of eight signers preferredtheir own viewpointand producedthe classifierfor for bowl articulatedaway from the chest banananear the chestwith the classifier . relations itself to encode that canusespatialgeometry for languages spatial of Reference 5. ). the origin of the coordinatesystemis the viewpoint of the signer All in 5. . ( 8 a relative frame of reference frames .6 Figure -ground complexityappears to hold even about figure Thus Talmy' s generalization . Final classifier construction for (3b ) . For this case . relative . articulatedon the rig~t. the shown were asked to describe ASL figure . chapter . To indicatethat the graphicway. Within spatialpropertiesof thesereference . eight picture signers example their one with the banana on their left but oneindicatedthat the bowl wason right ( without usingsigningspacein a toposignerprovideda descriptionof the scene ON SI -DE instead ). producingthe neutralphrase form for bowl on the left the classifier bananawason their right.7. or absoluteframe of spatialrelationsusingan intrinsic discussion of the linguisticand 4 this volume for reference see Levinson . signers produced was simultaneously form for banana a classifier and then side of signingspace . In who is signing of the person from the perspective scenes aremostoftendescribed . 5.The Confluenceof Spaceand Languagein Signed Languages 179 Final classifier consh"uction for (3a) .1. 's from the addressee viewpoint9turn out to be more likely in the Descriptions ' is still than in the left-right dimension front-backdimension (the signers perspective in shown .

Figure 5. Signer 's viewpoint (5/ 8 signers) . Karen Emmorey ---- a.---~ ~~ ~ : : :A :~ = ~~ -=~ 5.180 . .8 b .7 Figure Illustration of one of the pictures that signerswere asked to describe . Addressee 's viewpoint (3 / 8 signers ) .

9 and 5.8a. Note that the lack of an overt marker of point of view. in which the banana is behind the bowl . This spatial configuration of classifiersignsmaps directly onto the view presentedin figure 5. Only one signer placed the car on his right and the tree on his left. along with the corresponding pictures of a car in their different locations with respectto a tree.The Confluenceof Spaceand Languagein Signed Languages ISI behind the classifierfor banana.8 (rememberthat you as the reader are facing both the signer and the picture) . There were no switches in viewpoint within either the left -right or front -back dimension. the pictured female signer adopts her own perspective(describing the picture as she seesit ) . the ground object (the tree) is first and generally held in spacewhile the lexical sign for car is articulated expressed and the vehicle classifier is placed with respectto the classifier for tree. and the classifier form for vehicles encodes : the front of the car is representedroughly by the tips of the index intrinsic features .8. Note that the orientation of the car is consistentwith the point of view lo adopted. while the male signer adopts 's the addressee viewpoint . (seeLevelt 1984 Bananas and bowls do not have intrinsic front / back features . Again the majority of signersexpressed own view of the picture. Signers were also consistent within the intrinsic frame of .9 to the addressee viewpoint for figure 5. As you can see .9 and 5. all signersreversedtheir descriptions according to the viewpoint they had selectedpreviously.8b.10.10. In figures 5. cars do have these intrinsic properties.10 illustrate ASL constructions and middle fingers. which are extended using the vehicle classifier. the potential ambiguity . three signers 's describedthe picture from the addressee viewpoint .10 representthe final classifier construction in the description . as shown in figure 5. However. There were no overt linguistic cuesthat indicated figures doing when viewing which point of view the signer was adopting. The majority of signers described figure 5. As noted above.9 by placing the vehicle classifier to their left in signing . In contrast. signerswere very consistent in what point of view they adopted. although one signer 's switchedfrom her own viewpoint in describing figure 5. Figures 5. and the consistency within an adopted point of view also occur in English and other spoken languages ).the vehicle classifier is always oriented toward the tree. when the signerswere shown the reverseof figure 5. Again all space signerswere very consistentin which point of view they adopted. In contrast. The illustrations in figures 5. signersorient the vehicle classifier to indicate the direction the car is facing. producing the classifierfor bowl near the chest and the classifier for banana in line with the bowl but further out in . as shown in figure 5. and thus signers could not use an intrinsic frame of referenceto describethese pictures. For example . Also as noted. lexical signs identifying the referents of the classifier signs are given first. This configuration would be the spatial arrangement signing space seenby an addressee standing opposite the signer (as you the reader are these ) .9 and 5.

10 Fiaares t ~ ~ ~ ~ . signers 7 / 2 ( viewpoint s Addressee . ) rs signers ) in ( viewpoint Addresseels 182 's a.9 aDd 5.) . Signer ' 5. Karen Emmorey . b .

the expressionsshown in figure 5. . north . That is.. west. handshape N handshape :~ E handshape : ~ S handshape : f' ) W handshape : SlY ( Thesesignsare articulated in this manner.10illustrate such simultaneous expression of reference frames. further researchwill detennine how addressee and signer are established altered and discourse . Figures 5. frame of referenceambiguities can abound in ASL . . palm in . This situation contrasts sharply with how speakersgesture in cultures which employ absolute systems of reference such as . Thus the intrinsic frame of referenceis not dependentupon the relative frame. solo ASL signers (such as those in this study) are less explicit about spatial perspectivethan signers with conversation partners. Finally . . regardlessof where the person is standing. The car could actually be in any left-right or front -back relation with respectto the signer. The answer appears to be that a relative framework is not necessarilyentailed in locative expressionsin ASL . And just as in English ( Levelt 1982a .9 and 5. toward the left /right or away from /facing the signer). in ASL these two frames of referencecan be expressedsimultaneously. up. regardlessof true west or north . ASL signerscan usean absolute referenceframe by referring to the cardinal points east.g. That is. hand moves handshape SOUTH : S hand moves down. hand movestoward right .TheConfluence in Signed of Space andLanguage Languages 183 reference . almost always changing the orientation of the vehicle classifier appropriately 11 (e. 1984 ). The signs for thesedirections are articulated as follows: WEST: W handshape . NORTH : N handshape . EAST: E .9a and 5. linguistic expression within an intrinsic frame occurs via the intrinsic properties of certain classifierforms.9b could be interpreted as the rough equivalent of " the tree is in front of the car" 's without referenceto the signer' s (or addressee ) viewpoint . palm out . One question of interest is whether signerscan escape the relative point of view that is imposed " automatically " by the fact that signers(and addressees ) view their own articulators in spaceand thesearticulators expresslocative relations using this space . viewpoints disambiguatedduring Preliminary " " evidence suggeststhat .what is critical to the intrinsic expressionis that the vehicleclassifieris oriented toward (facing) the tree. like English speakers(Schober 1993 ). and south. The linguistic and nonlinguistic factors that influence choice of viewpoint within a relative referenceframe have not been determined . hand moves toward left12. and a relative frame can be imposed simultaneouslyon signing space if a viewpoint is adopted by the signer. although it is likely that severaldifferent linguistic and nonlinguistic factors are involved. that is.

palm in toward left . namely. Spoken languages have several different . the movement of thesesigns can be changed to label directions within a " map" created in signing space . centered and outward from the signer. As in English. In thesecultures. words. a spatial frame of referencecan be associatedwith a particular character . palm to left PATH NORTH " n" hand right hand shapetraces traces path samepath. the following directions were elicited from two signers describing the layout of a town shown on a map (from Taylor and Tversky 1992 ): STRAIGHT right hand traces a path outward from the signer " You drive " straight eastward. and the sign NORTH13 then traced a path similar to the one in (4). Tversky. Within narrative. such that north was centered outward from the signer. (5) UNDERSTAND MOUNTAIN R-D (4) YOU DRIVE EAST " e" handshapetracesthe samepath. Both signersand speakersuse linguistic devicesto explicit than in spoken languages indicate whether utterancesshould be understood as expressingthe point of view of " " the signer /speakeror of another person. The linguistic mechanismsused to expresspoint of view in signed languagesappear to be more . this volume) . that is. directional gesturesare articulated toward cardinal points and vary dependingupon where the speakeris oriented.184 Karen Emmorey certain Aboriginal cultures in Australia (see Levinson 1992b and chapter 4. and Coon 1992 . it is the direction words themselvesthat pick out an absolute framework within which the discoursemust be interpreted. 5. It appears that ASL direction signs are either fixed with respectto the body in their citation form or they are usedrelative to the spacemapped out in front of the signer. and Tversky. The signer who uttered (5) then shifted the map.6 Narrative Perspective In a narrative. near signer " Understand that Mountain Road " goesnorth in this direction.1. point of view can mean either a visual perspectiveor the nonspatial perspectiveof a character. The frame of referenceis relative. that character' s thoughts. and the origin of the chapter 12 coordinate system is the viewpoint of that character in the story . (seediscussionsof viewpoint in Franklin . Although the direction of the citation forms of ASL cardinal signs is fixed. For example . this volume) . or feelings .

and literary " " as well. and to the left of the boy was a tall tree.. he or shecould signal a referential shift . 14 ) have noted that ~ . That is.g. point of view (in either sense ) can be marked overtly (and often " " continuously) by a referential shift.g. an inherent limitation of spoken languages suggest linearize deictic and path information . indicating that the following sentences ) should be understood from the perspective of the boy. Poulin and Miller 1995 ). Supalla 1991 . (here ). styles(e. syntactic structure (active vs. Referential shift is expressed by a slight shift in and or in head . Engberg . Supposea signer were telling a story in which a boy and a girl were facing each other. A ." That spoken languagesexpressdeixis and path through separateelements discourse (either through two verbs or through a satellite expressionand a verb) reflects .. or facial expression (for -Pedersen1993 discussionsof this complex phenomenon . Padden 1986 . Signed languagesoffer important insight into how different frames of referenceare specifiedlinguistically .The Confluence of Spaceand Languagein Signed Languages 185 devicesfor expressingeither type of perspective : pronominal deixis (e. as is easily done in signed languages languagesbecausewords are articulated in the space surrounding the signer. I would further hypothesize that this simultaneous expression of deictic and other locative information within the verbs of signed languagesmay lead to habitual expressionof spatial viewpoint within discourse . and locatives. but also that character s spatial viewpoint through signsmarked for location and/ or deixis. signedlanguages usespacein severaldifferent linguistic domains. Deixis is easily expressedin signed . demonstratives . body position / changes eye gaze position. rather than expressthis information simultaneously . use of J vs. Gee and Goodhart 1988 ) . as well as with respect to points that may be established in to indicate the locations and viewpoints of protagonists set up in the signing space . the signer would produce the sign LOOK -AT upward and to the left. he or shewould produce the sign LOOK -AT and direct it upward and to the right . The following is an exampleof a referential shift that would require overt marking of a spatial viewpoint . including contrast co reference . see Loew 1983 . Signedlanguagesusethesemechanisms but in addition . they . The visual-gestural modality of phonological to influence the nature of grammatical encoding by comsigned languagesappears es (see pelling signed languages to prefer nonconcatenative morphological process also Emmorey 1995 . ' directional deixis p . Signers often ' ' expressnot only a character s attitudinal perspective . but also with respectto sender and receiver . there you). If the signerthen wanted to shift to the perspectiveof the girl . spoken languagemust . such that " toward " and " away from " can be encoded simply by the direction of motion with respectto the signer or a referential locus in space . . If the signer wanted to indicate that the boy looked up at the tree. passive ). Slobin and Hoiting ( 1994 . free indirect discourse ) . To do this. In sum. Lillo Martin 1995 . in that plays a key role in signedlanguages a path verb moves not only with respectto source and goal.

2. John ( 1992 ). In addition . Although spoken languagesalso have mechanisms to express deictic and locative relations.1 Solving Spatial Puzzleswith SpatializedLanguage To investigatethesequestions .11 : Subjects describe how to placeblockson a puzzle grid. I explore some possible ramifications of the spatial encoding of locative and spatial contrasts for producing spatial descriptions and . shifts in referenceare often accompanied by shifts in visual perspectivethat must be overtly marked on deictic and locative verbs. 5. and (4) how differences in linguistic encoding betweenEnglish and ASL affect the nature of spatial commands and directions. relations are directly encodedin space 5.186 Karen Emmorey unique aspectof the visual-gegturalmodality may be that intrinsic and relative reference frames can be simultaneously adopted.11 onto a sizes ) problem. what is unique about signed languagesis that such . (2) to what extent signers use lexicalized expressspatial locatives in spatial directions. and instructing experimenter14 fit all blocks must . Specifically and directions commands . L > ABCDEFGH I Figure5. ten hearing English speakersand ten deaf ASL native a task in which they had to solve three spatial puzzlesby were compared using signers where to place blocks of different colors. The data from English speakerswere collected by Mark St.2 Some Ramifications of the Direct Representation of Space In the studies reported below. shapes an . . To solve the see 5. figure puzzle grid ( within the puzzle outline. (3) whether the use of sign language provides an advantage for certain spatial tasks. I investigate ( I ) how ASL signersuse spaceto solving spatial problems. and a similar but not identical protocol was used with ASL signers [ ? [ ? 1 2 3 4 P ~ . Solvinga spatialpuzzle .

E2. by tracing a distinctive part of the board in spaceor by holding the nondominant hand in space . For example . p < . This prediction was confirmed: 67% of the English speakers commands referred to the puzzle grid.65. the subject and experimenter sat side by side. signerscould simultaneously describe the orientation of a piece (through the orientation of a ' classifier handshape ) and that pieces relation to another block through two -handed .for example . These categories were able to account for all of the orientation of a single piece the commands given by the twenty subjects . D2 . point to the puzzle or to the pieces but they were also not permit ted to point to the piecesor puzzle.12awas not statistically significant. ASL signers could use their hands. the numerical difference seenin figure 5. (8) Green pieceon EI . (2) commands expressinga relation between two pieces . The only difference was that in ASL . representinga often an . For both signers and speakers . Languagedid not appear to influence how subjectslabeled the puzzle pieceswithin this task. . Of course . Instead of referring to grid coordinates.we examined different types of English and ASL instructions. There were significant differences .encoded in either To explore how speakers and signers use spatial language spaceor sound.l2b ) .01) . (7) Place the red block in 3G H 2G. C2. ASL signers used spacein various ways to indicate the positions on the puzzle board. There were no significant differences placed for a given command (seefigure 5. and (3) . However. We hypothesizedthat ASL signersmay be able to usesigning spaceas a rough Cartesian coordinate system . and therefore would rely less on the coordinates labeled on the ' puzzle board. This differencein grid referencewas statistically reliable (F ( I . We had hypothesized that signers might make more referencesto shape becauseshape is often encoded in classifier handshapes(seediscussionabove). Puzzle commands could be : ( I ) commands referring to a position on exhaustively divided into three categories the puzzle board. whereasonly 28% of the commandsgiven by ASL signersreferred to the puzzle coordinates.in Signed of Space and Language TheConfluence Languages 187 English speakerswere instructed to si~ on their hands and were not pennitted to .12a in how either ASL or English was used to label a particular block. The following are sample commands containing references to the puzzle grid given by English speakers : (6) Take the blue L pieceand put it on HI H2 G2. 18 ) = 9. and D3. two command types could be expressedsimultaneously. ) part of the puzzle board ( edge We also compared how signers and speakersidentified the puzzle pieces to be ) . in the types of commands used by ASL signers and English speakers(see figure 5. however. such that each had the samevisual perspectiveon the puzzle board.

Examples of English relation commands are given in (9). 70 Other 8peakers 81gners P08lt1on - D88f Engl18h 8On Cortin .) '0 'E . p < . signing the letter coordinates moving crosswiseand the number coordinates moving downward. Subjectsdid not differ significantly in the percentageof commandsthat referred to the relation of one piece to another. they often specified these coordinates within a vertical spatial plane. . ~ P8 Shape Color 0 10 40 20 30 50 .8) appearsto account for this differencein command type. Type of command reference 5.15. as well as the constructions illustrated in figures 5. as if the puzzle board were set upright .05) .188 80 60 Karen Emmorey S JO 8. E E 0 (. e l signers Deaf . English ' reliance on commands speakers involving coordinate specifications (see examples 6. ( 10 ) Put it to the left of the green piece ( II ) Switch the red and the blue blocks. 18 ) = 4. Thus the true horizontal " " plane of the board laying on the tabletop was reoriented into a vertical plane within signing space .9. but clearly useof a vertical plane doesnot necessar layouts yet ily indicate a true vertical relation betweenobjects. It is interesting to note that even when ASL signers referred to grid coordinates.5. The linguistic and constraints on using a vertical versushorizontal plane to representspatial pragmatic are to be determined . speakers Orientation English m Relation on board Position puzzle 0 60 10 50 20 40 30 .47.12 Figure classifier constructions (see figure 5.( II ): (9) Put the other blue L next to the green one. 5. English speakersproduced significantly more commands referring to a position on the puzzle board compared to ASL signers (F ( I . 'a C .10 ). and 5. Type of puzzle piece identification b.

p < . ( 13 ) Rotate it 90 degrees ( 14 ) Flip it back the other way. ASL signers often overtly specified orientation . For example . ASL signers produced significantly more commands that referred to the orientation of a puzzle piece (F ( I . Figure 5. Figure 5.The Confluence of Spaceand Languagein Signed Languages 189 ASL signersalso produced thesetypes of commands . conveyed . Finally . note signe ~s perspective .13 . Examples from English of commands referring to orientation are given in ( 12 ). given other non-orientation -specific commands .15 illustrates the simultaneous production of a command indicating the [pictured ] :G 5. Orient the . green block in this wayo See green block in figure 5. the nondominant prepositional phrases example hand can representone block . Figure 5.05) .14 illustrates a command in which orientation change is specified by a change in the orientation of the classifier handshapeitself. ( 12 ) Turn the red one counterclockwise .11 .13 GREEN CL Figure CL:G-orientation . For English speakers . For betweenpieces .15 ).24. figure 5. rather than the relation .13 illustrates an ASL command that indicates a change in orientation by tracing a block ' s ultimate orientation in signing space (the vertical plane was often used to trace shapeand orientation ) . In piece contrast.( 14) : . but generally space . 18 ) = 5. and the dominant hand either points to a spatial locus to the left or right (somewhat like the construction illustrated in figure 5.6a) or the dominant hand representsanother block and is positioned with respect to the nondominant hand (seefigure 5. a change in orientation was often inferred from where the had to fit on the board.

NEAR . The lexicalized locatives that were produced by signers in this study included IN . ON . ASL also has a set of lexicalized locative signs that are used much lessfrequently than classifier constructions in spatial descriptions. AGAINST . Figures5.14 and 5.15 .Qthwiseat the top of another block [the green block ] . and these were almost always produced in conjunction with commands involving classifier constructions.Move the red L so it is oriented len . Only about 20% of ASL commandsinvolved lexical locatives.long end outward [pictured] 5. and BETWEEN . The grammatical structure of theseforms is not well understood. Signersalso used orientation of an L-shapedpiece and its relation to another piece the sign ROTA TE quite often and indicated the direction of rotation by movement of the wrist (clockwise vs. 15 RED L CL:B Figure CL:L -orientation .190 Karen Emmorey [pictured ] Figure5.and their ) or verbs (seeShepard they adpositions (seeMcIntire 1980 . counterclockwise ).14 BLUE L CL:L-orientation -Move the blue L so it is orientedwith the .are -Kegi I985) ?.

UNDER. For example woulddirectpieces to beplacedIN CORNER . two linesmeetto form . the reference ( ) category spatialentity in area " areasarisingfrom a dividing surface (p. 153 ). and OUT ) . but neverIN : ).11 .The Confluence of Spaceand Language in Signed Languages Figure5. IN wasalsousedwhena block (most " " often the smallblue square ) wasplacedin a hole createdby other blockson the board or whena part of a block wasinsertedinto the part of the puzzlegrid that stuckout (seefigure 5. 16in this case Herskovitz1986 a typeof container(see . PUT RED L IN G2 H2 ON 3E 4F 3F 3G The useof the prepositionin for describing grid positionson the puzzleboard falls " ' s 1986 " " under Herskovitz . ASL signers ( 15 ) PUT RED LON G2 H2 1213 15 ) PUT BLUE ] ( 16 [CL:G. signers (falling under Herskovitzs spatial entity in a container ). The linguistic data from our study provided some of IN and ON (these signsare shownin figure interesting insightinto the semantics 5. applyingonly whenthereis a clear appears . Illustrationby Frank Allen Paulin Newell( 1983 ASL lexicalized locativesigns ). containment relation . This objectmust be one of several structuredoesnot appearto be availablefor the ASL sign IN . namely . IN semantics has not beenwell studiedeither (seeMcIntire 1980for somediscussion of IN .shape tracedin shape verticalplane ( 17 ) . ( sample example above usedthe lexicallocativeON in this context . particularsemantic -like as container did use IN when of the puzzlecould be construed aspects Signers " ' " . in and on interchange usedthe prepositions ably to specifygrid Englishspeakers " " " " in H2 H2 commands 6 and7 for G2 or on G2 see coordinates . the reference ) . In both cases objectforms a type of container . . The useof the ASL lexicallocativeIN into which a block could be placed to be more restrictedthan Englishin. 149 ).16 ).16 .

subjectstended to place the most constraining piece first (the green block shown in figure 5. For example .] in upper to lower L classifier(right hand) is horizontal left oriented and positioned with respectto B classifier plane (left hand) as in figure 5.shape ] THAT -ONE ROTATE [CL :L . However. Finally . the iconic properties of ON lead one to that its use might expect depends upon a support relation . ASL signers and English speakersdid not differ significantly in the time they took to solve the puzzles ..reference shapetraced obj. The support requirementsdescribedby Herskovitz for on in English do not appear to apply to the lexical locative glossedas ON in ASL ." English speakers never produced commands relating one block to another using " only the preposition on. Signers used both vertical and . the ability to linguistically represent objects and their orientations in spacedid not provide signerswith an advantageon this complex spatial task. with the nondominant hand representingthe support object. This difference in semantic structure highlights the difficulties of transcribing one -Kegl ] 985) .Lorientation] new orientation traced in horizontal plane " Move the red one so that it is oriented " lengthwisenext to the green. Given the nature of the puzzle. for further discussionof languagevariation and topological concepts ). the red piece ON the green in ): figure 5. In addition . are not compatible with this hypothesis .] I ) .192 Karen Emmorey One might conjecture that the iconicity of the sign IN rendersits semanticstransparent . In summary.g. The data from our experiment.one hand representsa container.] 5 ON GREEN " Rotate that red L shaped block clockwise so that it is oriented lengthwise at the top of the green. this volume. chapter ] 0. ( ] 9) RED [CL :G .11 ON GREEN ( ] 8) RED MOVE [CL :G .orientation] clockwise [CL :B. ASL signersusedON when placing one block next to and contacting another block (e. English speakersand ASL signersdiffered in the nature of the spatial commands that they used for positioning objects. and the other locates an object within it . For example . iconicity can be misleading. however. and both groups appeared to use similar strategiesin solving the puzzle. English of another (seealso discussionin Shepard languageusing glosses on is not equivalent in semanticsor syntax to ASL ON (seeBowerman. subjectsnever said put the " red block on the green one. Signers and speakersdid not differ in the number of moves required to solve the puzzlesnor in the number of commands within a move.

Changesin horizontal planes of spaceitself as a rough Cartesian coordinate system in the were orientation position of spatial changes directly through expressed object . . . I way . . . In the following study. The heart of this different useof spatial languageappearsto lie in the properties of the aural vocal and visual manual . . I . In contrast in orientation and and . ~ A . J .17 ). . the . ( ~ ~ Describer . I ! f . set Experimental . . and interruptions . . . . For example by their own orientation in space such direct representation within the linguistic .2. feedback ) interchanges " describer (the person giving the instructions) could not see the manipulator . English classifiers signing space by tracing shape speakerswere lesslikely to overtly expresschangesin orientation and relied heavily on direct referenceto labels for coordinate positions. Figure5. . the hands can directly expressorientation linguistic modalities. in ASL . In order to elicit very specific instructions " and to eliminate (or vastly reduce . but the manipulator could seethe describer through a one-way mirror (seefigure 5. I . . and the semantic structure of the ASL locatives cannot be extracted from the iconic properties of the forms. Finally . we further explore the effect modality .17 -up for room descriptions .of Space TheConfluence Languagein Signed Languages 193 . . one Manipulator ~ ft ~ 8 Q ~ n g . mirror . ASL and English differ in the signal is not available to English speakers semanticsthey assign to lexicalized locatives for the topological concepts in and on. may exert on the nature of spatial languagefor both spoken and signedlanguage 5.2 Room Description Study Eight ASL signersand eight English speakerswere asked to describe the layout of " " objects in a room to another person ( the manipulator ) who had to place the objects 17 (pieces of furniture ) in a dollhouse.

Mean time for ASL was 2 min ) description . at least to some degree . p < . and their type Room Room type Haphazard arms 80 Speakers Signers English . the speedof the signers descriptions is quite striking because . on .05. . 5. However. However. The position of the hands in spacerepresentsthe position of the objects with respect to each other.18a .194 100 Karen Emmorey 0 . In one way. b. Emmorey and Corina 1990 Bellugi 1979 ) . . 48 sec to describe the same average ' rooms. as we have seenthus far in our discussion of spatial language in ASL . c 0 " 'S . Clothier . ASL signs take twice as long as English words to articulate (Klima and average . their position. the two hands can represent two objects simultaneously through classifier handshapes . 4 ~ 60 40 20 Haphazard Normal 8.18 Figure The manipulator could not ask questions but could request that the describer pause or produce a summary. Subjectsdescribed six rooms with canonical placementsof furniture (" normal rooms" ) and six rooms in which the furniture had been strewn about haphazardly without regard to function (" haphazard rooms" ) . Accuracy of manipulators . and the orientation of the hands can also simultaneously representthe objects' orientation . For example .o c C ) 'I a C I ~ 5 0 5 0 5 0 . . b). The simultaneousexpressionof two objects. E .00. 14 )= 5. ~ Deaf c E .. languages Signerswere significantly faster than speakersin describing the rooms (F ( I . there are several modality -specific factors that would lead to efficient spatial descriptions and lessenthe need for discourse linearization ( Levelt 1982a . .4 signers ' sec an of 2 min . English speakers required . _ 3 1 3 2 2 1 . and McCullough ) . seefigure 5. certain results emerged from the study that illuminate some ramifications of the direct representation of space for signed . The linguistic data and analysis arising from this study are discussed elsewhere (Emmorey. Doll house room description .

as measured by the percent of furniture placed correctly by the manipulators in each group (seefigure 5. . representedby height of the bars in figure 5. We here images memory ( ) hypothesized that theseimagery abilities are integral to the production and comprehensionof ASL and that their constant use may lead to an enhancementof imagery skills within a nonlinguistic domain. -accuracy trade-off. ) Bellugi ( useof visual mental imagery. On the other hand. Emmorey.TheConfluence of Space andLanguage in Signed Languages 195 orientation standsin contrast to the linear strings of prepositions and adjunct phrases that must be combined to expressthe sameinformation in English.18b . from the spatial medium of signed languages . If thesehearing native signershave visual-spatial skills similar to those found for deaf signers . we examinedthe ability of deaf and hearing . any impact on nonlinguistic spatial processing and 1993 examined the relation betweenprocessingASL and the Kosslyn.3 Interplaybetween andSpatialCognition Spatialized Language We now turn to the relation between general nonlinguistic spatial cognition and processinga visual-spatial linguistic signal.18b ) . ) To summarize . compared to the auditory transmission of spoken languages S. These differences arise. shown by the line in each bar in figure 5. we also tested a group of hearing subjectswho were born to being deaf parents. . Does knowing a signed language have ? In a recent investigation. at least in part. if these cognition arise from the useof a visual-spatial language have visual skills similar to those found in signers . In the previous study. regardlessof whether a lenient scoring measurewas used(object misplacedby more than 3 cm or misoriented . Specifically . There was no significant differencein percent correct. this second study suggeststhat the spatialization of American Sign Languageallows for relatively rapid and efficient expressionof spatial relations and locations. and to maintain in this last skill will not be discussed . we saw that ASL signersand English speakers focused on different aspectsof objects within a spatial arrangement . In order to distinguish the effectsof using ASL from the effectsof deaf from birth . this would suggestthat differencesin spatial .18b by 45 degrees ) or a strict scoring measurewas used(object misplacedby I cm or misoriented by 15 degrees . to generate mental images subjects to mentally rotate images . as reflected by instructions for the of blocks within a coordinate differing placement plane. this would spatial hearing subjects that differences in suggest spatial cognition may be due to auditory deprivation from birth . These subjectslearned ASL as their first languageand have continued to useASL in their daily lives. The difference in description time was not due to a speed Signers and speakersproduced equally accurate descriptions.

In addition .e. in order to understand and would must perceivethe reverseof what they themselves processsign. this transformation processmay occur frequently. describing the position of objects and people ) must therefore observes be understood as the reverseof what the addressee during discourse actually (assuming a face to face interaction) . As discussedearlier. and Bellugi ( 1993 ) originally suggestedthat ASL signers . particularly because they were faster even may be faster in detecting mirror reversals when no rotation was required (i.19 ).g. they signer easily . . compared to when mental rotation is not required on any of the trials. Given theselinguistic processingrequirements must be usedto produce the signs . The spatial loci usedby the signer to depict a scene(e. during sign comprehension the between signer and addressee must mentally reversethe spatial arrays created often i. a spatial locus established on the right of example by the signer such that . we at than would be better that mentally rotating hearing subjects signers hypothesized .. preliminary results from Emmorey and Bettger indicate that when native ASL signersand hearing nonsignersare asked to make mirror -samejudgments in a . hearing subjects ' s articulations into the reversal that transform a do not ASL . the person signing (and thus on the left of the addressee ) is understood as on the ). for .10a 's ' scenes are most often describedfrom the signer s perspectiveand not the addressee . both deaf and hearing signershad faster of rotation . see 5. at zero degrees ) . The crucial difference for ASL is that right with regard to the speaker thesedirections are encodedspatially by the signer.196 Karen Emmore~ We hypothesized that mental rotation may playa crucial role in sign language processingbecauseof the changesin spatial perspectivethat can occur during referential shifts in narrative (seeabove) and the shifts in visual perspectivethat occur . Emmorey Kosslyn. The problem is not unlike that facing understandersof spoken languageswho have to keep in mind the directions " " left " and " . Subjects were shown two forms createdby juxtaposing cubesto form angular shapes were asked to decide whether the two shapes were the same or mirror images . and this indicates that signersdo not actually rotate images faster than nonsigning subjects .19 of orientation ) ( figure regardless Our results support the hypothesis that use of ASL can enhancemental rotation skills (seethe top illustration in figure 5. aspect learning great difficulty produce Anecdotally. To test this hypothesis in which a task similar to the one devised by Shepard and Metzler ( 1971 ) subjects . the addressee of with this have . Furthermore. recent researchby 18 es may be involved when Ilan and Miller ( 1994 ) indicates that different process within a mental rotation experiment made at zero are mirror samejudgments degrees . However.e.9a and 5. Note that the slopes reaction times compared to nonsignersat all degrees for the angle of rotation did not differ betweensigning and nonsigning groups.. the addressee ) perceiver ( . Because right in the scenebeing describedby the signer (seefigures 5. we used imaged objects and making mirror imagejudgments.

i e i 8 . ~ m ! ! ii ~ ~ o < t & a . -0r The Confluence of Spaceand Languagein Signed Languages . c 0 I I ~ ' i 6 0 ~ .18A81 ) I UO . m I I ~ I 0 '. s ~ i ~ I r fD c 6c ~ ~ t t 6J . ! ~ mm c c I I i . .

It would be incorrectto signthe verbat the heightof the signer is present . addressee. a short .ASK . The faster response times exhibited by signers on the mental rotation task may reflect faster times to initiate mental rotation or faster times to generate a mental image ( as suggested by the next experiment ) . Finally . these groups do not differ in accuracy or reaction time . that is . signers may imagine referents as physically present . it appears to be due to experience with a visual language whose production and interpretation may involve mental rotation ( see also Talbot and Haude 1993) . image generation may be an important process underlying aspects of referential shift .20) : To directthe verbASK towardan imagined referent mustconceive of the location . Illustration fromLiddell Agreement ( 1990 ).198 Karen Emmorey comparison task that does not involve mental rotation . the ability to create an image ( i .imagined tall referent Figure5. the signer ' s head . . For example andaddressee wereto imagine that of theimaginary referent . the finding that hearing native signers performed like deaf signers indicates that enhancement on this mental rotation task is not a consequence of auditory deprivation .term visual memory representation ) on the basis of information stored in long .20b ]). if thesigner Wilt Chamberlain wasstanding beside themreadyto givethemadviceon playingbasketball . 1985) . Since the signermustconceptualize agreement a. In ASL . standing the locationof body partsof verbreflects this.imagined tall referent b. if the [5.20a ]).20 verbsand referents imagined as present.. * addressee.e. ' s head thesignASK wouldbedirected upwardtowardtheimaged heightof Wilt Chamberlain 's chin (figure (figure[5.ASK . the heightand directionof the as layingdown on a chair. etc referentis imagined . Naturally workswhena referent . Another visual imagery skill we investigated was the ability to generate mental images.. Rather .term memory ( see Kosslyn et al . Liddell gives the following example involving the verb ASK which is lexically specified to be directed at chin height (seefigure 5. This is exactlythe way agreement . and these visualized referents are relevant to the expression of verb agreement morphology . Liddell ( 1990) argues that under referential shift .

both groups generateequally good images nonsigners subjects appeared to image letters in the same way: both groups of subjects required more time and made more errors for probes located on late-imaged segments . Results experience ability mentally generate images from a perceptual baselinetask indicated that this enhancementwas due to adifference in image generation ability . present Our results indicated that both deaf and hearing signersformed imagesof complex letters significantly faster than nonsigners(seefigure 5. thereis a sense imagined sucha body in order to properlydirect agreement verbs .19 that ) . this difference in responsetime basedon probe location is not found when image generation is not involved. This result indicatesthat neither group of subjectsgeneratedimagesof lettersas completewholes . on the first stroke of the letter F ). ability experience .19. This was not enough time for the subjectsto complete forming the letter presented thus image.. Kosslyn et ale 1988 ). the finding that hearing that their enhancedimagegeneration signersperformed similarly to deaf signerssuggests is due to with ASL rather than to auditory deprivation . and both groups imaged segmentsin the sameorder. and we might . subjectshave faster reaction times. then the speed of forming these images would be important . that is. when both the probe X and letter (shaded gray) are physically . Furthermore. Crucially. Roth and Kosslyn 1988 serially from parts (e..g. . Subjectstend to generate letter images segment by segment in the same order that the letter is drawn.The Confluenceof Spaceand Languagein Signed Languages 199 the referent to be present in whichan invisiblebody is present . Therefore. A lowercaseletter precededeachgrid . and theseeffectswere of comparable magnitude in the two groups. suggests although signers create complex images faster than . The signing and nonsigning subjectswere equally accurate which that . compared to when the probe is located under a late-imaged segment . The .. Kosslyn and colleagueshave used this task to show that visual mental images are constructed . The crucial aspectof the experiment was that the probe mark appearedin the grid only 500 ms after the lowercasecue letter was . (Liddell signermustconceptualize 1990 . The imagegeneration expectsignersto developenhancedabilities to generateimages task we used is illustrated at the bottom of figure 5. deaf and hearing . This finding suggests with ASL can affect the to visual . 184 ) If deaf subjectsare in fact generatingvisual imagesprior to or during sign production . Subjects first memorized uppercaseblock letters and then were shown a seriesof grids (or setsof brackets ) that contained an X mark.g. rather than to differencesin scanning or inspection . when the probe X is covered by a segmentthat is generatedearly (e. responsetimes reflect in part the time to generatethe image. and subjectswere asked to decide as quickly as possible whether the corresponding uppercaseblock letter would cover the X if it were in the grid . Again.signersand nonsignersdid not differ in their ability to evaluate probe marks when the shapewas physically present .

conceptual structure. Klima . this volume) and spatial representationsis a topic for future research andSpoken for Signed 5. Note that these experiments have focused on ASL processing . In contrast.. Poizner. and Bellugi 1987 the visual-spatial modality of signedlanguages . Bellugi. Poizner. Poizner. For example them to match a model (the W AIS blocks test). whether there is a more direct relation in sign between linguistic representations(e. but does not produce sign language aphasias given testsof sign languagecomprehensionand production (e. doff . colored blocks and asked to assemble have signers great difficulty and are unable to capture the right hemispheredamaged . signers terized by a lack of morphological and syntactic markings and often accompaniedby halting . effortful signing. whereasthe right hemisphere es linguistic functions is dominant for visual-spatial functions. When many visual-spatial abilities. For example " " charac with left hemispheredamage may produce agrammatic signing. when given a set of nonlinguistic tests of visual-spatial functions. and Bellugi 1987 damage perform normally .200 Karen Emmorey esa relation betweenvisual-spatial imagery within linguistic This researchestablish and nonlinguistic domains. from the Salk Sign ). Despite (Poizner. Ursula Bellugi. Given that ASL express brain contrasts what is the . In general .g. An agrammatic signer will produce single-sign utterances that lack the grammatically required inflectional movements and use of space (see discussion above) . organization for sign by manipulating spatial ? Is sign languagecontrol led by the right hemispherealong with many other language visual-spatial functions or does the left hemispheresubservesign languageas it does of the ? Or is sign languagerepresentedequally in both hemispheres spoken language that the Edward Klima have shown and brain? Howard Poizner. brain honors the distinction betweenlanguageand nonlanguagevisual spatial functions . sign languageexhibits properties for which each of the cerebral hemispheres of hearing people showsdifferent predominant functioning . but these same signers show marked impairment on . Klima . Image generation and mental rotation appear to be es that must obviously be deeply embeddedin using ASL . and Klima 1989 ) . whereasthe right hemisphereis specialized . see Jacken language . adult aphasiassimilar to classicaphasiasobservedin speakingpatients.g.. linguistic processingoccurs primalily within the left hemisphereof deaf signers . chapter I . and for nonlinguistic visual spatial processing in these signers Klima have shown that damage to the left hemisphereof the brain leads to sign .4 NeuralCorrelates Languages Finally . right -hemispheredamage produces impairments of . and theseare not process involved in both visual imagery and ASL perception. signers with right -hemisphere Aphasia Exam. the left hemisphere has beenshown to subservelinguistic functions. Bellugi.

Nevertheless . 19a young hearing signer (age 39).. Corina. Her attempts to place one set of objects in relation to others were particularly impaired. speaking subjectswith right -hemispheredamage 1987 also that some Poizner.in Signed of Space and Language TheConfluence Languages 201 overall configuration of the block design. Although her English description had no evident spatial distortions . bilingual in ASL and English. Emmorey. from this case that there be more Hickok . The comparison test betweenEnglish and ASL spatial commands (seebelow and figure 5. Corina ( 1989 ) developeda specificset of tasks to investigateD . and Bellugi 1995 . but they are incorrect when compared to the actual layout of objects. She underwent surgical evacuation of a right parietal-occipital hematoma and an arteriovenous malformation . Her ASL description showeda marked disorganization of the elementsin the room. Evidencesupporting this hypothesiscomesfrom a map to space bilingual hearing patient with right -hemisphere damage studied by David Corina and colleagues(Corina et al. sheexhibited a striking dissociation betweenher ability to comprehendand produce spatial descriptionsin English compared to ASL . Examination of a magnetic resonanceimaging (MRI ) scan done six months after the surgery revealeda predominantly mesial superior occipital-parietal lesion. unlike spoken language . 1990 . Emmorey. who was exposed to ASL early in childhood.N . to set up . . One of thesetasks required DiN . while the inferior parietal lobule was spared .. not to sound) . linguistic description for signedcompared to spoken languages The caseinvolves female patientD . Their descriptions are not ungrammatical. The superior parietal lobule was involved. and sheincorrectly specifiedthe orientation and location of items of furniture (seealso Emmorey. for example when describing the layout of furniture in their room or apartment. Klima . and she showed no linguistic deficits Sign Diagnostic Aphasia Exam was excellent for English. she was impaired in her ability to describe the spatial layout of her room using ASL .e. although someof the deepwhite matter coming from this structure may also be involved. ASL requires that the cognitive representationof spatial relations be recoveredfrom and instantiated within a spatialized linguistic encoding (i. Similar impairments on this task are found . and Bellugi 1995 ).Nis comprehension of locative relations in English and ASL . Corina . One hypothesis for this dysfunction following right -hemispheredamageis that .21) was conducted by Corina approximately one year after DiNis surgeryD . was not aphasic for either English or ASL . and Corina 1993 The data may ) suggest information encoded within a involvement when right hemisphere processingspatial . cognitive spatial relations . with hearing.N . and Bellugi ( ) reported signing patients with right hemispheredamageshow a selectiveimpairment in their ability to use spaceto expressspatial relations in ASL . Her performance on the Salk .

pen. 1 : ) CL instruction paper on ( ASL PENCIL to response B : ~ CL incorrect PAPER DiNis 8 . but better than other right -hemisphere age damaged subjects who were given the English test (69% correct) . This result is particularly striking . made many more spatial errors. given the iconicity of the ASL descriptions (seefigure 5. The instructions were .21) . Her score was worse than her normal -matchedbilingual control ( 100 % correct). : paper the response instruction on is instruction lish correct En pencil ~ to DiNis he English ~ . shown are not PENCIL and ) (the lexicalsignsPAPER spatialcommands real objects in accordance with spatial descriptions given in either English or in " " ASL .21. she performed relatively well. was given instructions in English to locate objects with respectto one another. when presentedwith similar information in ASL in which spatial relations are presented topo graphically in sign spaceD . but fails with the ASL instructions. cup. was later given 36 different spatial commands ( 18 in English and 18 in ASL ) which involved from two to four objects (e. DiN . This particular example was elicited through informal testing by Corina in which the same instructions were given in both English and ASL . When matched for number of spatial relations that were encodedin each language D .g.N .21 's differential ASL in comprehending English versus Illustrationof a RHO patient performance .N . DiN .. An exampleof a simple English instruction would be The pen is on the paper. correctly interprets the English command. However. scoring only 39% correct.83% correct.202 : instruction ASL Karen Emmorey Figure5. book ) . The English and ASL instructions along with DiNis responsesare illustrated in figure 5.

Nis comprehensionimpairment is not linguistic per within the right hemisphereD se . and for linguistic structure. All current evidence indicates that signed languagesare constrained language and space . for spatial cognition in general . chapter 10.Laird . Thus far . (3) spatial attention in sign language perception and nonlinguistic visual-spatial perception (do signers show differencesin spatial attention ?). DiN . spatial relations must be recovered from a visual-spatial signal in which much more information is encoded about the relative position and orientation of objects. this volume. Future researchmight include investigations of the neural substrateof sign language the following : ( I ) the semanticand grammatical structure of locative constructions in different sign languages(how do sign languagesvary in the way they utilize physical ?). but stemsfrom the fact that linguistic information about spatial relations must be recoveredfrom a representationthat itself is spatialized. and ( 5) the neural substrateand psychological mechanisms . doesnot have difficulty understanding ASL spatial contrasts that do not encodeinformation about location or orientation . (4) how signersbuild that could be attributed to experiencewith sign language spatial mental models (doessigning spaceoperate like a diagram? SeeJohnson. the requirement of reading off spatial relations directly from the orientation and position of es classifier signs in spacemay make additional demandson spatial cognitive process . What is different and unusual compared to spoken languages(see Supalla 1982 about signed languagesis their visual spatial form . for languageprocessing .The Confluence of Spaceand Languagein Signed Languages 203 We hypothesize that the dissociation betweenD . this volume). DiN . chapter II .the fact that spaceand movement can be used to linguistically representspaceand movement in the world . 1995for a complete discussionof this dissociation and for additional -processingexperimentswith normal ASL signers evidencefrom language ). this volume. there is no by the same principles that shape spoken languages evidence that signed languagesgrammaticize different aspectsof the spatial world ). Furthermore. and Bowerman. Thus the caseof DiN . This chapter has explored the ramifications of this spatialized encoding for the nature of . also bears on our earlier discussionconcerning . compared to English. That is. referential versustopographic functions of spacein ASL . exhibits a dissociation -level between the use of signing space as a linguistic device for marking sentence referential distinctions and the useof signing spaceas a topographic mapping device (seeEmmorey et al. signed languagesoffer a unique window into the relation between . for discussion of spatial cognition and spoken language acquisition). (2) when and how signing spaceto representtopological and other spatial concepts the children acquire locative vocabulary (what is developmental relation between spatial cognition and sign languageacquisition? SeeMandler .Nis comprehension of English of the highly specificspatial realization of and ASL spatial commands arisesbecause ASL classifier constructions. chapter 9. In conclusion.

here . Descriptions signis articulated may be givenunderneath shape areprovidedin quotes . who weremy primary language .204 Karen Emmorey that underlie the mapping betweena linguistic signal (both signed and spoken ) and . Whethersubjectis associated with the beginning or end of the verb depends " " 1988 of verb(cf. Classifierformsare abbreviated in italics(CL:Gof theclassifier anda description of themeaning followedby thehandshape . A subscripted of the unmarked root form of a sign word followinga . Some pronouns maynot be specified signssuchaspersonal Lillo-Martin and Klima 1990 .. I am particularlygratefulto the GallaudetUniversitystudents these studies . who has made similar 7. .g. . systematic changein meaning connected areusedwhenmorethan oneEnglish GI VEbabltu by hyphens . that the sentence considered . Brentari1988 ). verbsare indicating 3.. upontheclass that a sentence is mark (1) indicates 5. Mark Williamshelped in this chapter . The glossrepresents I . I alsothank BonitaEwanand providedvaluable consultants and who werethe sign language SteveMcCullough . OthertenDSthat havebeenusedfor these ) and inflecting (Liddell 1995 1988 (padden ). the gloss of how a classifier ). Subscripts verbsare markedwith a subscript to indicatethe . Hoiting 1994 . ' s movement 4. Followingtraditionallinguistictypography . pronouns . models for the figures create of many the figures who participatedin Finally. Englishtranslations in the lexiconfor location(see 2. LOOK-AT) . Liddell 1994 ). backwards verbs . I thank David Corina Hickok and Ed Klima for many 00146 R37 . Acknowledgments This work wassupported . nativesigners to me by Dan globin. and agreeing spatialloci.)' Multiword glosses are usedto indicate word is requiredto translatea singlesign(e. nouns CL. unmodulated meaning in form associated with a that the signis madewith someregularchange signglossindicates . ROI DC by National Institutesof Health grantsROI DC 00201 and HD 13249 . INDEX . Merrill Garrett and Mary Peterson about the issues presented insightful discussions comments on an earlierdraft of this chapter . The exampleof drawing was suggested about scene (Slobin and arguments settingand the effectof modality on signedlanguages ).g. a question is unacceptable . Wordsin capital lettersrepresent for ASL signs the Englishglosses . BIT~ ). loci at which they are signed(e. and thus indicatesgrammatical morphologyin ASL (e. Greg . a star(* ) indicates marginal weredeafindividuals who wereexposed to ASL from birth. Theseare only someof the areasin which the study an amodal spatial representation of sign languagecould enhanceour understanding of the relation betweenlanguage and space .g. Notes . In this study .. 6. Padden .

Except for the sign LEfT . This study was conducted with Marci Clothier and StephenMcCullough . I thank Mary Petersonfor bringing this work to my attention. 16. palm out (try it ) . the experimenter was either a native speakerof English (for the English subjects ) or a deaf ASL signer (for the deaf subjects ). This study was conducted in collaboration with Shannon Casey . Also . For example which also move acrossthe body. . When the signs NORTH and SOUTH are used to label paths within a spatial map. The addressee is assumedto be facing the signer." For both left.9 and 5. there is somechangein articulation for left -handers . 18. it might be useful for you the reader to imagine that you and the signer viewed the display from the same vantage point . with opposite directions of motion (left to right vs. 14. the orientation of the palm is reversed : outward for WEST and inward for EAST. 17. say. Sign linguists often use " frame of reference . WEST is perhaps the only sign that is specified as moving toward the signer' s left rather than toward the " nondominant side. and now the signer is facing you (the addressee ) to describeit . as might be implied by figures 5. they often retain someof their upward and downward movement. a classifierconstruction in which the shapeof the blue puzzle piece is traced in the vertical plane (seefigure 5. 13. pointing the vehicleclassifierto the right with the right hand. the sign WEST moves toward the left . IS.13 for an example ). signs. In understanding this discussionof point of view in ASL . right to left . but usethe pseudonyminitials A . 19. CORNER is a frozen classifier construction produced with nominal movement (Supalla and Newport 1978 ) . This is not an orientation command but a shapedescription. The direction of movement is fixed with respectto the signer s left and right . It should be noted that occasionally a signer may ignore the orientation features of the vehicle classifier . However. when in actual fact the car is facing away from the tree. signersdid not always alternate which hand produced the classifier for TREE . pointing the vehicle classifier toward the tree classifier . namely. 9. For EAST and WEST. unlike other . say. The sign can be articulated at various positions in spaceto indicate where the comer is located (e. referring to anaphoric nonspatial sense referencein a discourse(seeespeciallyEngberg Pedersen1993 ).and left -handers would articulate the signs illustrated in figure 5. Signersdescribedthesepictures to a video camera rather than to an actual addressee .S. This may occur when it is difficult to produce the correct orientation . II . 12. This changein palm orientation also occurs when a right -handed signer articulates EAST or WEST with the left hand (switches in hand dominance are phonologically and discoursegoverned ). top left or bottom right) .The Confluenceof Spaceand Languagein Signed Languages 205 " in a 8. and the sign EAST moves toward the right -handers ' right .g.. There were only six examples(out of thirty -five) in which a signer ignored the orientation of the car becauseit was awkward to articulate. 10. respectively ) . right . Poizner and Kegl ( 1992 ) also discussthis patient.1.and . perhaps due to phonological constraints.10.

D. ). U. K . Paperpresented spoken language at the Academy of Aphasia . 20(5).206 Refereaces Karen Emmorey Battison. October -spatial . In Papers re-opened . 507 Gee andthehumanbiological . and the brain. NJ: Erlbaum space . 49. D. andCorina . H . 139 181 . B. Tversky .). . 165 Phonology Coulter and . Working paperno. . Reilly (Eds .. ( 1989 ). 6. syntacticfunctionsof space Tucson . In K. Differentialprocessing referentialfunctionsof space . ( 1990 ). InternationalStudies on SignLanguage Research and of theuseof space language Communication of the Deaf . La Jolla. Space Engberg SignLanguage in a visual . . D. Emrnorey and J. vol. -Pedersen in Danish : Thesemantics andmorphosyntax . andNonnan . modality. D. and Coon . Hickok.. Kosslyn . M. 1O . . Trendsin . AZ . . Strong(Ed. Nijmegen thropologyResearch Group. 16 Theory . ( 1988 ). Corina . Baltimore Corina in signlanguage structure . ( 1993 Emmorey ).. On the natureof phonological . andMotor Skills.27. R.518 models .. Salk ). W. M. CA: Academic phonology in signlanguage : Effects of phonetic . K. E. and Sandier . Lexica/ borrowing in American Sign Language Press . U . vol. R. ( 1993 Emmorey ). W. . Poizner. language imagery hearing signersCognition .1252 structure andmorphology . In M. Neurosciences . Paperpresented at the Academyof AphasiaMeeting . D. R. Coulter(Ed. Visual imageryand visual : Enhanced abilities in deaf and ASL ... and Bellugi . . V. ( 1988 ). U. Silver Spring. Dissociationbetween topographicand in ASL. Hillsdale . D. 2. U. P... 10 . New York: Cambridge learning . L. . 19 Franklin. American SignLanguage capacity for language . 24 .388 . andAnderson . O Grady . 7J. CognitiveAn. Chicago Parasession on Agreement in Grammatical : Chicago . ( 1992 ). Spatial in relations signed versus : Cluesto right parietalfunctions . D. andGoodhart . 46 .. ( 1993 ). MemoryandCognition . Languagegestureand . J. Kritchevsky .62.. and Bellugi .. no. 1227 of topographic and . Lexicalrecognition Emmorey .) Phonetics : Currentissues in ASLphonology . G. Switchingpoints of view in spatialmental . S. ( 1995 Emrnorey ).207 . K. ( 1993 )....74 . Unpublished . and Klima . Language anddeafness . ( 1991 ). LinguisticSociety Brown in Tzeltal . Max PlanckInstitutefor Psycholinguistics Corina relations testbatteryfor ASL. Topographic manuscript Institutefor BiologicalStudies . Language . F. ( 1978 . Bellugi . Perceptual . ' -Batch Corina . ES . G.. Introductionto G. and Corina . . Hamburg : Signum . 380 Brentari verbsin ASL: Agreement . K. UniversityPress BeUugi.. Backwards from the . Press . SanDiego . R. 43. S. N.. Spatialconceptualization . ( 1993 ). . ( 1990 ). CA. ( 1989 . MD : Linstok ) ..

beliefs . andtokenspace ). W. Liddell. andWallach . 12 . 251 Levelt in describing .. S. Utrecht . K . ProvostD . Language . Durham structure . Speech Klein (Eds .). Washington . . 199 (Eds Levelt .. ). saarinen . Nijmegen Levinson andcognition : Thecognitive of spatialdescription . gesture .. R. ( 1995 : Grammatical in ASL. . . generationCognitive Psychology " " " " Landau and spatial . . 195 . ( 1993 ). Behavioral andBrainSciences . 13 . surrogate . Language studyof theprepositions in English .268 . Brennan ). What and where in spatiallanguage . Working paperno. Four functions . In S.243 mentalimagery . S. S. S. ( 1994 . Cambridge . 16 . Cave . A violation of pure insertion reaction time. shape . B. In K.. Paper Workshop at thePhonology andMorphologyof SignLanguage signed languages presented . J. NewYork: Wiley. gesture . ( 1982b ). J. Some perceptual . . andspace . Amsterdam . A cross linguisticperspective . A. 19 . limitationson talkingaboutspace . 323 Science Press . on ASL verbagreement . . in . and linguisticdescription ). Journalof Experimental : Human andPerformance . Hillsdale andJ. Jarvella andW. Peters and E.The Confluenceof Spaceand Languagein Signed Languages 207 Herskovits : An interdisciplinary andspatialcognition . ( 1985 ).536 520 . J. and Bellugi . 195 ability: A computational analysis esin image . S. W. W. CognitiveAnthropologyResearch Group. A. and Miller.). Reilly (Eds . U. Vision and object descriptions . L. W. Real consequences . Cognitive styles . van de Grind. Max PlanckInstitutefor Psycholinguistics . Cave ..198 Lucas(Ed. . . S. Emmorey Janis andJ. and Jackendoff . Cambridge : Cambridge . S. ( 1979 . 20 319 343 .). . UK : ISLA. Tokensandsurrogates (Eds onsignlanguage . ( 1986 ). S. . Sequential process .). place . CognitiveAnthropologyResearch Group. 217 238 cognition Levelt in the useof spatialdirectionterms . B. . Bergman . ( 1993 ). ( 1995 ). and Von Gierke . S. K. UniversityPress : Mental rotation and choice nan. NJ: Erlbaum Reilly(Eds Klima. .224 . Signlanguage research : Theoretical issues . Individualdifferences Kosslyn . Thesigns of language Press . M.. Ahlgren . ( 1992b ). Language . 18 .358 A. In K. Psychology Perception . Hillsdale . ( 1988 Kosslyn ).42. and J. DC: Gallaudet Press . B. Linearization spatialnetworks . -part tenninology Levinson : Tzeltalbody . Brunn . In R. S. In C. ( 1992a . and M. R. Working paperno. Conceptual spoken . W. 176 . ( 1984 ). . MA: HarvardUniversity ). Limits in perception : VNU . Perspectives Liddell. College Liddell andlinguisticissues in spatialmapping : Comparing and . R. August . Cognition . andspace .220 . W.. 20(3). andaction . ( 1994 ).).). In I.. E. Language consequences in GuuguYimithirr. NJ: Erlbaum Emmorey . J. vanDoom. ( 1990 : Reexamining ofa locus thestructure of space in ASL. In A. Koenderink(Eds . Process es . Max PlanckInstitutefor Psycholinguistics .).. Nijmegen Liddell. . andquestions . ( 1982a ). Dordrecht : Reidel .

J .170 . . . . In C . Construction . of the third dimension in mental imagery . . andspace out Emmoreyand Lillo . ( 1986 National ) . On J . : National Association ) . DC . ) .132 . MA . of the sign : Linearity and nonlinearity in American perspective taking in conversation .256 What and . . . acquisition . : Theoretical issues . ( 1987 ). Los Meier Newell of the . 4 . ( 1988 in Linguistics . ( 1989 Language . the motor behavior : Perspectives . NJ : Erlbaum . Universal : Kluwer ) . University San Diego Padden Lucas University Poimer from Poimer Cambridge Poulin Language Hillsdale Roth . . ) ( 1983 acquisition sign by deaf children . E .132 . ( 1985 linguistic ) . Dordrecht ) . representation : Foris . Chicago : University Loew . Neural . ( 1994 and ) . in ASL . Sign Language : A developmental perspective . Language . . NJ . . . ( 1990 P .70 . basis of . 344 ) . Padden and Teaching ( Eds Research . American Sign Language . Emmorey . gesture of . Fischer and theory 191 . diss Outstanding of Dissertations California . I . . ) . hands reveal about the brain . Siple of Sign . Cognition . ( 1990 ( Ed . NJ . S . M . . C . J . Basic communication Spring : National Newport Siobin Hillsdale Padden Fourth MD .208 Lillo . MD . Language The data . I . ) . Roles . The ( Eds Klima grammar . New PhiD . . 219 U .57 . and American H . R . ( 1988 ) .. D . ( 1992 Language ) . . . : Erlbaum in differences issues pronouns language syntactic . ( 1995 J . Spatial Sign 117 . 79 . C . Phonological . Aphasiology Bellugi . Reilly . Language and point . 1 24 . diss .361 . In : Gallaudet C . 118 Washington Kegl Sign . narrative Reilly ( Eds discourse . In K . C . Locatives and ( Eds . Silver . Klima . ( 1991 ) . Sign Press . view and in space Quebec . The language relation research between space and grammar in ASL verb morphology . University ) . E . 1983 in ASL . University of California . diss ) . gesture ) . H .MartinD . R . vol . and . M . In S . 60 Association ) . C . of morphology York and : Garland syntax . C . S . I . . PhiD . Garland . Theoretical Press . W .210 . Vol Sign . 20 . . ser . Pointing in American and American sign language Karen Emmorey : Setting the nullargumentparameters . Proceedings . ( Ed Deaf .938 .MartinD . 155 : ASL in sign Language . In 881 D . / nteraction the Deaf Padden . Silver . E . . Dordrecht Lillo . In : MIT Miller K . The Meier Cross . The study of acquisition language of American . R . ( 1980 Angeles . ) . and : Erlbaum Kosslyn Cognitive Sandier Sign Schober Psychology . language . W .) . . .MartinD . ) . Hillsdale .) . ( 1991 . and point predicate . 6 ( 3 ) . Language . M . . . 44 of Spring the . PhiD McIntire . American Scientist . ( 1993 . . ( 1983 . research Chicago in reference of Minnesota in American . 47 . ) . and Press . and ( Ed . Verbs Symposium and on of role Sign - shifting Language . of view .

PhiD. ( 1995 ).19 Linguistics Society Linguistic in the serviceof a task. Mental rotation of three Shepard 703 . D.114 Reilly (Eds . Acredolo(Eds structures L. University Language in a chair?Thederivationof nounsand . ( ). ( English ) Manually Supalla research in signlanguage . Science the AnnualConference Fourteenth . and Newport Supalla . Boston : Little. Spatial . ( 1978 ). ( 1985 word formation . frames . D. objects . Press . NJ: Erlbaum . Topics .207 . CA: Berkeley Society . Language .. Journalof Memory Language . 261 descriptions . Brown andJ. 1983 .). K . Reference AnnualMeetingof the Berkeley . Rudzka to cognition (Ed. R. In Proceedings St.). F. Perceptual -dimensional objects spatial visualizationability: Mental rotation of three . Ph. Understanding verbsin AmericanSign Language throughsign language . research in . John of the . (Eds .132 language skill and Talbot. Hinsdale . . of Institute . Spatialmentalmodelsderivedfrom surveyand route Taylor. Emmorey discourse in comparative Winston . SipleandS. Science . Chicago 85 of verbsof motionandlocationin American andacquisition Sign . NewYork: Academic research . -Kegi. diss . Massachusetts . T. 165 cognitive linguistics . Fischer . E. ( space ). H. F. Locative relationsin AmericanSign Language Shepard . The relationof grammar Talmy .. Pick and L. ( 1987 SignLanguage ). . S. Learninglanguage . and . Amsterdam . H.signed : The coded 1991 . ( 1992 ).. How language Talmy Press . 87. Berkeley . Siple(Ed. ( 1988 ). 77 1387 1391 3 Motor Skills . ( 1971 . American . Society Cognitive of development in. T. . N. dimensions andapplied : Linguistic Wilbur. andapplication . Technology . L. In P. and Tversky . 2. In P.109 Press . Proceedings of the Nineteenth Typological considerations . J..). gesture . language modalityquestion . : Universityof Chicago . How manyseats . 91. : Benjamins . In K . Theoretical issues . Structure Supalla . In B.. and Haude sign language . E.The Conftuenceof Spaceand Languagein Signed Languages 209 -dimensional . Spatialmapping . J. diss and discourse . vol. Erlbaum Hillsdale NJ: . syntax : in spoken and signed to movement Slobin languages . SanDiego . of California .D. R. 31. and Metzler ). 1. .). ( 1982 ).). M. ( 1994 ). In H. NewYork: Plenum : Theory orientation . 171 701 . . R. The relationshipbetween and . ( 1993 ). ). B. andspace . and Hoiting. ( 1992 ).Ostyn .292 .

is a particular manifestation of the overlapping systems model of cognitive organization .1 Introduction This chapter proposesa unified account of the extensivecognitive representation of .especially forms of motion . our framework will also cover visual instances in which one : the perceived" apparent .g.1 OveraUFramework Our unified account of the cognitive representationof nonveridical phenomena . for example from the plateau to the valley. a picture frame) as having been tilted from a vertical-horizontal orientation .Chapter 6 Fictive Motion in Language and " Ception " Leonard Talmy 6. cognitive systems . and Thescenery steeple rushedpast us as we drovealong. We will mainly consider similarities between two such cognitive representations : languageand visual perception. conceptual. The vacuumcleaneris downaroundbehindthe clotheshamper . the perceived " induced motion " of a rod when only a surrounding frame is moved. and the " " of a vertical stroke possible perception of a plus figure as involving the sequence followed by a horizontal stroke. The cliff wall faces toward/away from the island. the es like indentation perception of a curved line as a straight line that has undergoneprocess and protrusion . to give an immediate senseof the matter.1. Thus. the framework posited here will cover linguistic instances that depict motion with no physical occurrence : Thisfence goes . 6. or other . the possibleperception of an obliquely oriented rectangle(e. for example perceivesmotion with no physical occurrence " motion in successive flashesalong a row of lightbulbs. just " " exemplified. In a similar way. as on a marquee . I looked out past the .. This model seespartial similarities and differences across distinct cognitive systemsin the way they structure perceptual..both as they are expressed nonveridical phenomena linguistically and as they are perceivedvisually.

we usethe lesscommon term veridicalrather than. Another example of a dimension is state of change . in saying that the two discrepant representations differ in their assessed degreeof veridicality . where one of the representationsis assessed We presumethat the two representationsare the products of two different cognitive itself is produced by a third cognitive . reality Of the two discrepant representationsof the sameobject." In the general fictivity pattern.or vice versa. And the termfictive has beenadopted for its referenceto the imaginal capacity of cognition. the individual need not have any active experienceof conflict or clash betweenthe two maintained representations . and that the veridicality assessment subsystems . a term like true. this cognitive pattern of veridically unequal discrepant representationsof the sameobject will here be called the pattern of " general fictivity . we will characterizethe to be more veridical as " factive" and the representation assessed representation assessed to be less veridical as " fictive.disagreewith respectto somesingle dimension. One form of this last veridical representationincludes fictive change -time is the more specific dimension when applied to a physical complex in space dimension state of motion. they would be inconsistent or contradictory .to signal that the ascription is an assessment . representing though not exclusively opposite poles of the dimension. as being more veridical than the other." Adapted from its use in linguistics. Further . the two cognitive representations consist of different contents that could not both concordantly hold for their represented object at the sametime. One example of such a dimension is state of occurrence observed . Here. On the other hand.that is. the term of greater veridicality. factive of someentity in the more veridical representation (the presence presence ) is coupled with fictive absence(the absenceof that entity from the lessveridical representation ) or vice versa. Here. this pattern : a discrepancy within the cognition of a single individual . say. the more veridical representation of an object could include factive stasis . subsystemwhosegeneral function it is to generatesuch assessments In the notion of discrepancy we intend here.212 Leonard Talmy The particular manifestation of overlap we address involves a major cognitive . with no appeal to some notion of absolute or external produced by a cognitive system . but might rather experiencethem as alternative perspectives . the two discrepant representations frequently. factive is hereagain intended to indicate a cognitive assessment but not to suggest(as perhaps the word factual would ) that a representation is in somesense objectively real. the more veridical representation could include . As a whole. while the less . not to suggest(as perhaps the word fictitious would ) that a representationis somehowobjectively unreal. Specifically discrepancy is between two different cognitive representationsof the same entity. Several different dimensions of this sort can be . as judged by the individual ' s cognitive systemsfor general knowledge or reasoning . Here.

1. Of thesetypes. fictive absence ness . representation is assessed the literal representation is fictive. As will be discussed .ve to the lesspalpable representation " " the factive representation . and the other representationis the . Moreover. fictive change . we focus on at a particular lower level of palpability . frequently in conjunction with their factive opposites . Here the literal literal referenceof the linguistic forms that make up the sentence as less veridical than the representation based on belief. belief is factive. and the other is a particular . The general fictivity pattern can be found in a perhaps parallel fashion in both . this fact reflectsa cognitive bias toward dynamism. Here. while the representation based on Accordingly . less palpable percept the individual has of the same sceneconcurrently. perception generally and of domains with what is usually associateddifferentially perception separate . Parallel to the linguistic case as the less veridical of the two representations . Thus. we here mainly treat the linguistic pattern in generally with factive stationariness which the literal meaning of a sentenceascribes motion to a referent one would otherwise believeto be stationary. Further these . fictive stasis . In language where one of the discrepant representationsis the belief held by the speakeror hearer about the real nature of the referent of a sentence .or vice versa. fully palpable representationis generally of stationariness To accommodate this account of visual representations that differ with respect to their palpability . too . the present chapter focuseson fictive motion . We will say that an individual sees termfict . the pattern is extensively exhibited in the case language and vision. while the less veridical representation has motion .9." and" Ception Motionin Language Fictive 213 stationariness . one main form of the generalfictivity pattern is the casewhere one of the discrepant representationsis the concrete or fully palpable percept an individual has of a scene on viewing it . In vision. usually in combination . and fictive motion . It will be seenthat such fictive motion occurs preponderantly with factive stationariness more than doesfictive stationarinesscoupled with factive motion . and the visual more to the be the term factive may . general fictivity can accommodate " " any fictive X. fictive stationari. All of these palpabilityrelatedparamet " are characterizedbelow in section 6. we can expect to find cases of fictive presence . representation palpable applied " " . while the palpable . to be discussed visual less where the fictive motion . one may identify a number of additional cognitive parameters " that largely tend to correlate with the palpability parameter. representation is of motion . Here the less palpable percept is assessed . we posit the presencein cognition of a gradient parameter of palpability . In fact. Given our focus on the pattern in which fictive motion is coupled . but only senses the fictive representation(when it occurs later) . parameters than that domain a to extend larger cognitive continuously through appear of in the combination that fact covers alone one with associated . to a large extent.

the present study of fictive motion is basedin language . there is generally the potential for any linguistic in space -domain example to have an analogue in a visual format . we advancethe idea of a single continuous cognitive domain. fully palpable of the appearance corresponding visual display. one aim for the presentchapter is to serveas a guide and as a call for such experimental research . its target domain is the nonphysical one of mood states motion in space . whereasmetaphor theory is cast in concepts and terms more suitable for languagealone. its conceptsand terms can apply as readily to visual representationsas to linguistic ones . The other parallel would then hold between the less veridical literal referenceof the sentenceand a less palpable associatedimage perceivedon viewing the display. although its source domain is -time.214 Leonard Talmy conception. correspondence expect find two component parallels. which we call " ception. and coverage paths. These categories include emanation. but extendsout from there to considerationsof visual perception. If we view this correspondence starting from the languageend. pattern paths. In this way. a linguistic example -time of general fictivity whose representationspertain to physical entities in space in can. For example space . in a cross of this sort we could to . One parallel would hold between the two factive representations . General fictivity can serveas the superordinateframework because . as discussedlater. the other between the two fictive representations . perhaps the type of fictive . frame-relative motion . Experimental methods are needed to determine whether the parallel between the two fictive representations holds. among other reasons . one parallel would hold betweenthe linguistic believedto be veridical and the concrete representationof a sentence . linguistic metaphor as a whole fits as a category within the framework of generalfictivity . be mappedonto a visual exampleof generalfictivity . to accommodatethe full range of each such parameter . Accordingly. 6. This last category. In fact. a metaphor like Her mood wentfrom good to bad would be excluded .1. In particular . the linguistic referential difference betweencredenceand literality is then translated in the visual domain into a difference in palpability . The restriction of the present study to the representation of physical forms in -time excludestreatment of nonspatial metaphor. In sucha mapping. advent paths (including site manifestation and site arrival ). However . accesspaths. effect . Using the perspectiveand methods of cognitive linguistics. Accordingly ." In the presentchapter we largely restrict our study of general fictivity in language to the casewhere both of the two discrepant representationsare of a physical complex -time.2 FictiveMotion in Language Fictive motion in language es a numberof relativelydistinct categories encompass (first set forth in Talmy 1990 ) .

The " mountain range" example in ( I ) avoids this problem. the mountain range. the one that is assessed thesetwo representations . the one directly depict the mountain range as moving. b. some speakerswill report a strong semantic evocation of motion . for the same instance of constructional fictive motion . which . Of . there appears an additional range of differences in what is conceptualized as moving . The factive representation assessed and experiencedas more veridical. This factive representation is the only representation present in sentence ( la ). These latter two sentences manifest the general fictivity pattern." " Fictive Motion in Languageand Ception 215 motion most familiar in the previous linguistic literature . and subjective motion in Matsumoto.what can be tenDed experienced fictive motion. however. But ( 1b) and ( lc ) representthe static linear entity. Thus. extension in Jackendoff ( 1983 " " ( 1987 ). in a way that evokesa sense or aconceptualization of something in motion respectively . consistsof our belief that the mountain range is stationary. is that every speakerexperiences a senseof motion for somefictive motion constructions. But a purer demonstration of this type of fictive motion would exclude referenceto an entity that supports the actual motion of other objects (as a road guides vehicles ) or that itself may be associated with a history of actual motion (like a TV cord) . Our current tenD coverage paths is used as part of the more comprehensivetaxonomy of fictive motion presentedhere. the fictive representation and experienced as lessveridical consistsof the literal referenceof the words. abstract motion in Langacker ). That mountain range goesfrom Mexico to Canada. Illustrating coveragepaths can serveas an orientation to fictive motion in general Modesto This category is most often demonstratedby fonDSlike This road goes from to Fresnoor The cord runsfrom the TV to the wall. That mountain range goesfrom Canada to Mexico. though. Here ( 1a) directly express es the more veridical static spatial relationships in a stative fonD of expression without . was called " virtual motion " " " " " in Talmy ( 1983 ). over the degreeto which such expressions evoke an actual senseor concep tualization of motion . evoking fictive motion . while other speakerswill report that there is none at all. That mountain range lies betweenCanada and Mexico. They each involve two discrepant representationsof the same object. Most observerscan agree that languagessystematically and extensively refer to stationary circumstanceswith fonDS and constructions whose basic referenceis to motion . Speakersexhibit differences . c. What does appear common.that is. ( 1) a. from north to south and from south to north . which accordingly does not manifest the generalfictivity pattern . We can tenD this constructionalfictive motion. the mountain range. . Where an experienceof motion does occur. This conceptualization can vary .

of which the main ones are shown in (2) .216 Leonard Talmy acrossindividuals and types of fictive motion .1. for example . the emanation category are listed . The fictively moving entity is itself factive/ fictive. which of emanation the category . The strength and character of experiencedfictive motion . in the mental imagery of the speakeror hearer. .3 Propertiesoftbe EmanationType as a Whole Amid the range of fictive motion categories . if observer -based -based -neutral/ observer 3. The severaldistinct categoriesof fictive motion indicated above differ from each . this chapter selectsfor closestexamination . scans the observer is factive/ fictive and moves . by some unnamed object that entity. appearsto have been largely unrecognized in The other indicated categories of fictive motion will be more briefly discussed section 6. the fictive motion may . / 4. Factive motion of someelementsneednot / must be presentfor the fictive effect. the intangible on somedistal terminates and its emanation continues by impinging path along entity exhibited by 2 that are of fictive features of the values . for example moves with respect to the named entity. and its fictive motion does not depend on any factive motion by some tangible entity nor on any localized observer. or by a senseof abstract directednesssuggestingmotion relative to the named entity . Out of the range of fictive motion categories . by the imagistic or conceptual equivalent of their focus of attention moving relative to the named entity . 2. as well as its clarity and homogeneity . eventhe sameindividual may deal with . are a phenomenologicalconcomitant of the presentstudy that will needmore investigation. What is conceivedas fictively moving is an entity/ the observation of an entity . . a car or hiker relative to the mountain range. The ( ) general object particular the in . In most subtypes of something intangible emerging from a source . of motion moving relative to the named entity by some abstracted conceptual essence . Included in the the sameexample of fictive motion differently on different occasions manifested be by the named conceptualizationsof this range. intangible entity is what (3) Specifically moves fictively and is itself fictive. (2) Principalfeatures distinguishingcategoriesof fictive motion in language I .1 6.8. Each category of fictive other with respect to a certain set of conceptual features motion exhibits a different combination of values for these features .and. The fictive effect is observer . by the mountain range in ( I ). emanationis basically the fictive motion .

in section 6. What is conceivedas fictively moving is an entity . but examples from other languages can be readily cited. the entity might be conceivedor perceivedas some intangible abstraction moving along a stationary line or shaft. perceptual analoguesto the emanation types that have beendiscussed 6. or a point -type front . The fictive effect is observer 4. The fictively moving entity is itself fictive.of an orientation path is of a continuous linear intangible entity emerging from the front of some object and moving steadily away from it . Then." Fictive Motion in Languageand " Ception 217 (3) Thefeature values for emanationpaths in language I .that is already in place and joined at one end to the front of the object. First . we will specifically suggest . The demonstrations of at least constructional fictive motion will rely on linguistic forms with basically real-motion referentssuch as verbs like throw and prepositions like into and toward.2 Orientation Paths The first type of emanation we consider is that of orientation paths.6.the only characterization used below. consisting of an approximately planar surface on a volumetric object. including whether the front is a face-type or a point -type. In the exposition. We presentfour of theseemanation typesin sections6. and whether the fictive motion of the intangible line is axial or lateral. The category of emanation comprises a number of relatively distinct types.itself equally intangible. or on the object s motion along a path. radiation paths. the " front " of an object is itself a linguistic conceptualiza tion or perceptual ascription based on either a particular kind of asymmetry in the ' ' object s physical configuration. consisting of an endpoint of a linearly shapedobject. -neutral. " such a front can be either a planar or " face -type front . shadow paths. where the 2 leading side would generally constitute the front . In addition to fictive motion along the axis of such a line. Alternatively . 3.5: orientation paths. The illustrations throughout will be from English only in the present version of this chapter. though.7. and sensory paths. wherever some form of linguistic conceptualization is posited. .and possibly a corresponding visual perception. In the main casesrelevant here. though. This entity may be conceivedor perceived as a moving intangible line or shaft. 2 . The linguistic conceptualization. Presentednext are five subtypes of orientation paths that variously differ with respectto severalfactors.2. Factive motion of someelementsneednot be present for the fictive effect. in somecasesthe line can also be conceptualized or perceivedas moving laterally . we will raise the possibility of a corresponding perceptual configuration . In this characterization.

1983 Object (in Talrny s [ 1987b ] terms) along a path indicated by directional . The linguistic constructions. In English. the object has a particular prospect. or vista relative to sorneother object in the surroundings. We would argue that what is thus being crossedis the posited intangible line conceivedto emergefrom the front of an object.218 Leonard Talmy we note the occurrenceof constructions that are sensitiveto the fictive presence of an line with the front of an before we to its fictive intangible aligned object. b. suchconstructionsgenerallyemploy verbslike/ aceor look out. rnore veridical representation consisting of our belief that all the referent entities in the sceneare static and involve no rnotion.2. treat this line as Figure rnoving relative to the other object as Ground or Reference ' . The orientation that an object with a face-type front has relative to its surroundings can . lessveridical representationin pattern. but only poorly when walking behind.3 This usagepattern seems to suggestthere is linear to walk across in front of an something present directly object. Shecrossed?behind/ * besideme/ the TV .1 ProspectPaths The first type of orientation path that we exarninecan be termed a prospectpath. the vertical side of a cliff acts as its face prospect upon its surroundings is characterizedin terms of a fictive course of rnotion ernergingfrorn its face and rnoving along the path specifiedby the preposition relative to a valley as ReferenceObject. this exarnple rnanifests the general fictivity of its words depicts a fictive. adpositions -type front . be conceptualizedlinguistically .and perhapsperceived " " " " With its front face.2. but not elsewhere with respectto that object. Shecrossedin front of me/ the TV . The literal sense which sornething rnovesfrorn the cliff wall along a path that is oriented with respect to the valley. Consider the sentences in (4) : (4) a. proceed motion . 6. that will next be seento exhibit fictive motion in a further set of construction types. This prospect is characterizedas if sorneintangible line or shaft ernergesfrorn the front and rnovescontinuously away frorn the rnain object relative to the other object. Again .in terms of fictive rnotion. The orientation of such a linear object is here conceptualized . The cliff ' s In the exarnplein (5). 6. But this representation is discrepant with the factive.2 Alignment Paths The alignment path type of orientation involves a stationary straight linear object with a point -type front . in effect." " exposure . The sentences here show that the verb cross can felicitously be used when walking transverselyin front of an object with a front . (5) The cliff wall facestoward /away frorn / into /past the valley. and not at all when walking to one side.

the addition of a path preposition in this construction has the effect of forcing a fictive alignment path interpretation that requires a straight-line ' contouring of the snake s body.2. The hypothesis that fictive orientation paths emerge ' from an object s front and move away from the object correctly accountsfor the fact that the sentence with " toward " refers to the head end of the snake as the end closer to the light . pointing only its head at or away from a light . that coercesthe verb' s semanticproperties . in the examplesin (7). This function of directing a person s attention can be the intended end result of a situation.3 DemormtrativePaths The demonstrativetype of orientation path also involves a linear object with a point . lIThe arrow on the signpost pointed toward /away from / into /past the town. such as an arrow or an extendedindex finger. physical gaze along path specifiedby preposition (7) a. or motion the the . But here the fictively moving line type front from which an intangible line emerges ' s attention functions to direct or guide someone along its path . lie. and the light is the distal object. Or it can be a precursor event that is instantiated or followed by another event. But in the normal understanding of (6).4 object. with a path preposition. As it happens the English constructions that evoke this arrangement are not free to representjust where the linear object is aligned with any orientation .and perhaps perceived along the axis of the object. Of note. a linear object with a front end. and continuing straight . 6.in terms of something intangible moving linguistically . A sentence with lie alone would permit an interpretation of the snakeas coiled and. . say.Fictive Motion in Languageand " Ception" 219 .the front being the end either closer to or further from the distal in (6) illustrate this type. while the sentencewith " away from " indicates that the head end is the further end. emerging from its front end. The particular orientation of the linear object can either be an independent factor that simply occasions ' s attention or can be an instance of directing someone . seems to emit an intangible line from its front end. intentionally set to servethe ' purpose of attentional guidance. This line ' s attention movesin the direction of the object' s orientation so as to direct someone . along a prepositionally determined path relative to somedistal object. or moving bodily along the fictive path . gaze Thus. the sentences (6) The snake is lying toward /away from the light . . toward or awayfrom . such as the person' s directing his or her . bIpointed / directed him toward/past/away from the lobby . the snakesbodyforms an approximately straight line that is aligned with the light . Here the snake is the linear object with its head as the point -type front . That is. this construction combines a verb of stationariness . but are limited to the two cases the distal object.

rather than below under sensory paths along with " looking .2. The reason is that the act of looking is normally treated differently in English from " " the act of photographic shooting. This fictive motion esa path along which the Agent further intends that a particular subsequent establish motion will travel. camera One might ask why the camera example is included here under the targeting type " of orientation path. with a so-conceivedphotographic probe emergingfrom the ' s front . (8) I pointed/ aimed (my gun/ camera Here the caseof a bullet shot from the aimed gun exemplifiesreal motion following the preset fictive path . where the distance between the two objects progressively decreases . Consider the sentences ) into /past/away from the living room. the object with the vision-equippedfront . 6. that is. But what within the situation depicted by the example sentencecould be .swivels . A path preposition ' like toward normally refers to a Figure object s executinga path in the direction of the ReferenceObject. an Agent intentionally sets the orientation of a front -bearing object so that the fictive line that is conceptualizedor perceivedas emergingfrom this ' front follows a desired path relative to the object s surroundings. thus causing the lateral motion of the line of sight that emergesfrom that front . and we do not conceive of the act of looking as involving first the gaze establishmentof a targeting path and then a viewing along that path . targeting.whether my head with its eyesor the camera with its lens. something like this sequence ~ single or double fictive path. In contrast. Although of intentions and actions. The presentdiscussiondealsonly with lateral motion of the line of sight.4 Targeting Paths In a targeting path. the camera provides an instanceof fictive motion following the fictive path. We normally do not speak of " aiming or pointing " our . The path preposition specifiesthe particular path that the line of sight follows. seemsto underlie our conceptsof aiming. This subsequentmotion either is real or is itself fictive. with shifts in its orientation .5 Line of Sight Line of sight is a concept that underlies a number of linguistic patterns. or in (8) in this regard.5 on sensorypaths. sighting. with comparatively complex. It is an intangible line emerging from the visual apparatus canonically located on the front of an animate or mechanicalentity .2. Axial fictive motion along the line of sight will be treated in section6.220 Leonard Talm }' 6. Additional evidencefor treating the shifting line of sight as an orientation path is that the sentences exhibiting this phenomenoncan use not verbs like look but also nonsensoryverbs like turn~ just sensory In the examplesin (9). Consider how fictive motion is at work in the caseof a sentencelike I slowly turned / looked toward the door. and perhaps also a component of perceptual structure.

and the irradiated object. This linguistic . then. the radiating event can be characterizedas involving three entities: the radiator . As I turn my head in the motion of the line of sight that emergesfrom my eyes this line of sight does indeed direction . Apparently what the preposition toward in this sentencerefers to is the . in the case of . This additional particularization is the only type treated here." Fictive Motion in Languageand " Ception 221 ? The only object that is physically moving is my turning exhibiting thesecharacteristics head. directly . though. In this type. We can note that English allows each linguistic form in a successionof path indications to specify a different type of fictive motion . hence detect . and then sight initially my swivelsdownward so as to align with the axis of a well./away from the window. one can often indeed detect the presenceof the radiation .III slowly turned my cameratoward the door . not moving closer to it . thus entering the well. light light . indicates that once my line of sight is oriented at a downward angle. the likely interpretation is that type discussedin this section. in ( 10 ). Thus. A radiation path differs from an orientation path in that the latter consistsof the motion of a wholly imperceptible line. the three process motion of the radiation along a path. What one cannot see the one can radiation .for example ./ around the room. (9) I slowly turned/ looked. The linguistic conceptualization of a radiation path is of radiation emanating continuously from an energy source and moving steadily away from it . of the specifying form . ( 10 ) I quickly looked down into the well. This radiation can additionally be understood to comprise a linear shaft and to subsequently impinge on a second object. 6.and. Under its specification " " I am horizontal of is line ( looking straight ahead ). indicates a lateral motion . In a radiation path. to the tapestry. And this radiating event then involves es: the (generation and) emanation of radiation from the radiator . appropriate clockwise or counterclockwise follow a path in the direction of the door and shorten its distance from it . the satellite down. past the pillar . and the impingement of the radiation upon the irradiated object. this radiation motion of is what remains imperceptible any The sentencesin ( 11 ) reflect the preceding characterization of radiation for the particular caseof light in the way they are linguistically constructed.1 from the painting. the radiation itself. the first pathof a line of sight.3 Radiation Paths The second type of emanation we consider is that of radiation paths. the preposition into. then the fictive motion of my vision proceedsaway from me axially along the line of sight. The secondspatial form . yet that object stays in the samelocation relative to the door .

The sun draws this energy toward itself when there is a straight clear path between itself and the matter. the task is to suggesta number of viable alternatives to the normal conceptualization. At this point . ( II ) a. and we will speculateon those factors in section 6. However. And it is certainly absent from extant linguistic constructions. Although physicists may tell us that photons in fact move from the sun to the irradiated object. it must be other factors that lead to a conceptualization in terms of motion away from the sun. Imagine the following state of affairs. An account of this sort is in principle as viable as the usual account. All matter contains or generatesenergy. but that it moves in the reversedirection from that in the prevailing conceptualization. and of prepositional object. becauseany phenomenon that could be explained in terms of imperceptible motion from A to B must also be amenableto an explanation in terms of a complementary imperceptible motion from B to A . any correspondencebetweenthe scientific charseeany such occurrence acterization and the conceptualization of the phenomenonmust be merely coincidental . the so-conceivedmotion of radiation from the radiator to the irradiated must be fictive motion . Now . but at least its locus of origination ). In fact. In both sentences . however. and that the light movesinto the cave or impinges on its back wall to illuminate that spot. that the light emanatesfrom the sun and moves steadily as a beam along a straight path through space . for all its equality of applicability . The sun (or a comparable entity) attracts this energy. in the end.our normal conceptual reverse apparatus.222 Leonard Talmy construction mainly involves the choices of subject. Thus ). In other words. Therefore. The sun glows when energyarrives at it . of path-specifying preposition. the generalunderstanding is that the visible light is a radiation . we certainly cannot actually . the fact is that this -direction scenario is absent from .6. Matter glows when its energy leavesit . tion possible One alternative conceptualization is that there is a radiation path. as compelling as this characterization of light radiation may be felt to be. / onto the back wall of the cave. then. purely a conceptualization. that the sun is the source of the light (perhaps its generator. and we suspect that any counterpart English lacks any sentencelike that in ( 12 formulation is universally absent from the languagesof the world . ( 12 . Thesealternativesshow that the unique appearanceof this conceptualization cannot be explained by virtue of its being the only conceptualiza .even resistedby. it is necessarilyso. ) * The light is shining from my hand onto the sun. it is. Becausedirect sight does not bring a report of ' light s motion . The sun is shining into the cave b. The light is shining (from the sun) into the cave / onto the back wall of the cave.

(Such warmth is of course also the casefor the sun. in addition . itself a constellation of factors." Fictive Motion in Languageand " Ception 223 The conceptualization that an object like the sun.the pole blocks the transit of energy in the no energyis drawn out of the portion of the ground behind reversedirection. The sun as Source conceptualization here has the pole as blocking the light ground that would otherwise proceedfrom the sun onto the ground directly behind the pole. Here one can seethat the surfacesof oneself facing the fire are brighter than the other surfacesand. -direction conceptualization works here as well. one can feel that they are warmer as well. a vertical pole and its shadow on the different circumstances . any -direction alternative attempted to invert . Thus we can check the viability of a conceptualization in which light originates at a point between the two salient objects and fictively moves out in opposite directions to impinge on each of those two objects. The reverse ' additions in this exampleare that when the fire attracts energyfrom the parts of one s body facing it . for example. the fireas-Sourceof both light and heat is not the only possibleconceptualization. do glow . Perhapsthe radiation does not exhibit fictive motion at all . However. Further . And the one saliently associatedwith fire. Another assumptionin the normal conceptualization we can try to challengeis that the radiation movesat all. whereasthe potions of ground adjacent to it . The sun attracts But the reverse energy from the side of the pole facing it . Because the pole. Or consider a fire. But we can also test out the factor which holds that a radiation path originates at one of the salient physical objects and terminates at the other. The same -direction conceptualization used for the sun holds as well for the fire. ( 13 ) tries to capture this conceptualization. counterintuitive * ( 13 ) The light shone out onto the sun and my hand from a point betweenus. or a flashlight produces light that radiates from it to another object is so intuitively compelling that it can be -direction conceptualization in of value to demonstrate the viability of the reverse . hence saved for the present example ' further factor here is that the attraction that the fire exerts on an object such as one s body is stronger the closer it is. but cannot do so from the portion of the ground directly behind the pole becausethere is no straight clear path betweenthat portion of the ground and the sun. the departure of that energy causesnot only a glow but also the sensationof warmth. The reverse one of which can be challenged the directionality of the fictive motion in the prevailing conceptualization. it fails to glow. a fire. Consider. -direction conceptualization is not the only feasible alternative to the The reverse prevailing conceptualization of a radiation path. this sentence es seemswholly does not work linguistically and the conceptualization it express . from which energy is being directly drawn. this effect is stronger the closer one is to the fire. Once again. but more ) .

eventhis argument of " sha . Sentences like thosein ( 15 to that surface aconceptual ) showthat Englishsuggests . thepredicate . For thereis no theory of particle physicsthat positsthe existence paths " . The treethrewits shadow the wall. anda path . The pillar cast /against /projected could conceivably be made We can note that with radiationpaths . the Experienceremits a Probe that movesfrom the Experiencerto the Experiencedand detectsit upon encounter with it . downinto/ across ( 15 ) a. Thusthese sentences setup the izationof this sort throughits linguisticconstruction it is asthe shadow asthe Figure nominalthat refersto the shadow . suchasinto. a shadow onto b. somethingintangible Experienced Experiencer moving in a straight path betweenthe two entities in one direction or the other. in particular . including visualpaths. because a weakargument . the argument from the sun to my hand that the directionof the fictive motion proceeds . is thus treated either Experienced as a probing system that emanatesfrom or is projected forth by a viewer so as to . herefunctioning . . project . or against preposition the valley . . But sentences like ( 14) show that this conceptualization. the surface cast or asa motionverblike throw asGoal. across . But however tenable that is the directionthat photonsactuallytravel in the case of shadow not be used could like this may be. too . has neither linguistic nor intuitive viability . By the other branch of the conceptualExperiencer ization . onto .5 SelB) ry Paths One category of emanation paths well representedin language is that of sensory paths. of its shadow dowons that movefrom an objectto the silhouette 6. By one branch of this conceptualization. This is the -as-Sourcetype of sensorypath .224 Leonard Talmy but rather rests in space as a stationary beam . This type of fictive motion involves the conceptualiza and of the the and tion of two entities. Sight. the experiencedemits a Stimulus that moves from the Experienced to the Experiencer and sensorily stimulates that entity on encounter with it . ( 14 6. ) * The light hung betweenthe sun and my hand. The linguistic The third type of emanationwe consideris that of shadow paths is that the of a shadow also a and path perception perhaps conceptualization hasfictivelymovedfrom that object shadow of someobjectvisibleon somesurface . the objectwhose is locatedasthe Groundobject on which the shadow Source . This is the -as-Source type of sensorypath .4 Shadow Paths .

where some forms with favor the Experienceras Source the Experiencedas Sourceoffer difficulty to somespeakers . The old wallpaper shows through the paint even to a casual passer Despite theseforms of alternative directionality . The bidirectional conceptualizability of sensorypaths can also be seenin alternatives of lexicalization. thereby stimulating a visual experience so as lexicalized verb a We can first illustrate this phenomenon using nonagentive . I can hear/ smell him all the way from where I m standing. namely. thereby promoting the interpretation of the Experienced . But show is thereby promoting the interpretation of the Experiencer as Source lexicalized to take the Experienced as subject and can take the Experiencer as the object of the preposition to. see is lexicalized to take the Experiencer as subject and the Experiencedas direct object. Further . Here the conceptualization appears to be that the Agent subject volitionally projects his line of sight as a Probe from himself as Sourcealong the path specifiedby the preposition relative to a ReferenceObject (the Experienced ." Fictive Motion in Languageand " Ception 225 detect some object at a distance . I can hear/ smell him all the way from where he s standing. ' ( 18 ) a. This is the casefor English. among the nonagentive vision verbs in English. -as-source sentence like ( 16b Somespeakershave difficulty with with an experiencer ). such as those for audition or olfaction shown in ( 18 ). ' ( 17 ) a. Even a casual passer ( 19 -by. agentiveverbs of vision in showis minimal relative to that of a verb like see English are exclusively lexicalized for the Experiencer as subject and can take directional . The enemycan seeus from where they re positioned. ' b. . distal object and arrives at an individual . We can be seenby the enemy from where we re standing. We can be seenby the enemy from where they re positioned. which takes the Experiencer as subject and allows a range of directional prepositions. as shown but this difficulty generally disappearsfor the counterpart passivesentence . ' b. -by can seethe old wallpaper through the paint . And generally no problem arisesat all for nonvisual sensorypaths. Thus. We illustrate in ( 19 as Source ). In ( 16 to take the Experiencer as subject. ' b. this is the case phrasesonly with the Experienceras Source with the verb look. b. in ( 17b ). or elseas a visual quality that emanatesfrom some . fictive visual paths may generally . . see ) the two oppositely directed : two different are fictive motion of by path phrases represented paths ' ( 16 ) a. As shown in (20a). ) a. and the useof a verb like . ' rrhe enemycan seeus from where we re standing.

This is understood from such evidence as that in total darkness or in fully diffuse . ' looked into / toward/past/away from the valley. Thus. By the operation of the active-determinative . ' where am located outside the < valley> 6. say. But underlying such diversity. the following cognitive principle appearsto be the main one in operation: the object taken to be the more active or determinative of the two is conceptualized as the source of the emanation.1 The Principle that Detenninesthe Sourceof Emuadon For the emanation types in which a fictive path extendsbetweentwo objects. a pole and the shadow of the pole. as betweenthe sun and my hand. This greater brightness seemsto lead to the interpretation that the sun is the more active object. Another application of the active-determinative principle can be seenin shadow . the conceptualizations associatedwith the different types of emanation were treated as distinct. (20) a. and perhaps perceived .226 Leonard Talmy is not named in this type of construction) . we can seekto ascertaina cognitive principle that determineswhich of the two objectswill be conceptualizedas the source of the emanation.6. It is now time to consider the principles that ." We can proceed through severalrealizations of this principle that have functioned in the earlier examples . This will be called the " activedeterminative principle . rather than any of the alternative feasibleconceptualizationspresentedearlier. this chapter has laid out the first -level linguistic phenomena that manifest different types of fictive emanation. while the other object is understood as the goal. However. the pole is the more paths. * ' looked out of the valley (into my eyes ). the sun will be conceptualized radiation moving through spaceto impinge with the other object. b. On examination. 6. while the shadow is the more contingent or dependent entity .6 A UnifyingPrincipleandan Explanatory Factorfor Emanation Types So far . As between determinative entity . in particular . or the sun and the cave wall . the sun is perceivedas the brighter of the two objects. there is no (20b)-type construction with look in which the visual path can be represented as if moving to the Experienceras goal. more energeticor powerful . govern and the context that generalizesthesephenomena In the preceding part of the chapter. We presenthere a unifying principle phenomenathat can account for their existence and an explanatory factor. one may discern commonalities that unite the various types and may posit still deeper . as the sourceof the principle .

and this. a nonagentivelyacting Experiencer . ones with an Experiencer that acts as an intentional Agent as well as with an Experienced entity . while the referent of valley is understood as a nonagentive agentive Experiencer Experiencedentity. I looked into the valley. one can move the pole light . positioned at the active-determinative locus. Here it seemsthe very property of exercisedagency leads to the interpretation that the Agent is more active than the Experiencedentity . rather than any alternative interpretation such as that the shadow moves from the indicated surface to the physical object. because " " . the agentive agency Experienceris conceptualizedas the Sourceof the sensorypath. For example the one as . . But this . the pole is still there but no shadow is present and the shadow will move along with it . from whom a detectional probe is taken to emanate . The task remaining. it will be this object. that is. for example where the active and determinative entity in the situation is the agent who fixes the ' orientation of the front -bearing object. is to ascertainthe additional cognitive principle is saved criteria that ascribe greater activity to one set of phenomenaor to a competing set. such as a cameraor the Agent s own arm with extended index finger. then. though. The fact that nonagentivesensorypaths can be conceptualizedas moving in either of two opposite directions might at first seemto challengethe principle that the more active or determinative entity is treated as the source of fictive emanation. the shadow-bearing object is thus conceptualized as generating the shadow. . which is either inanimate or currently no~ manifesting relevant . in fact. it is by the operation of the principle that this interpretation of the direction of the fictive motion prevails. that will be conceptualizedas the sourceof the fictive emanation. That is. By the operation of the active determinative principle . Further . By the operation of the active-determinative principle . by one set of interpreted criteria . targeting paths and agentive demonstrative paths. With our principle applying correctly again. only available interpretation for the sentence The active-determinative principle also holds for those types of orientation paths that are agentive . Thus the active-determinative . A further realization of the active-determinative principle can be seenin the caseof agentive sensorypaths. whosefictive motion . whereasthere is no comparable operation performable on the shadow. each be need not be the case that is more active than the other. is interpreted as more active than the entity probed." " Fictive Motion in Languageand Ception 227 . which then moves fictively from that object to an indicated surface. It may be that either object can. But under an alternative set of criteria . In the visual example presented thus proceedsfrom the Experiencer to the Experienced the referent of " I " is understood as an earlier. is the . by different criteria . the active-determinative principle requires that the Experiencer be conceptualizedas the Source of the fictive sensorymotion . the Experiencedentity taken to emit a stimulus is interpreted as being more active than the entity stimulated by it .

the nonagentiveorientation path types: prospect paths.6. and nonagentive demonstrative paths.g. It is active becauseit involves the generation of intentions and of actions. The exerciseof agencycan be understood to have two components . the generation of an intention and the realization of that intention (cf.228 Leonard Talmy and that are in effect in the absenceof the principle ' s already known criteria (e. the directionality of the fictive motion may be set indirectly by the conceptual mapping of principle-determined cases onto the configuration. is one' s carrying out of the actions that bring about the new state of affairs.what can be called the agent-distal object patterns In this pattern an Agent. reach to it with a body part . The model-relevant characteristics . If it is innate. Here the fictive motion emanatesfrom only one of the two relevant entities.2 Poaible Basisof Fictive Emanation and its Types If we have correctly ascertainedthat the more active or determinative entity is conceptualized as the Source of fictive emanation. If it is learned in the course of development . and it is determinative becauseit remodelsconditions to accord with one' s desires . Finally . then each individual ' s experienceof agency leads by steps to the conceptualization of fictive emanation. then. In these cases . say. the next question to ask is why this should be the case . but this entity is not apparently the more active or determinative of the two. or cause(as by throwing) someintermediary object to move to it . Such exerciseof agency is experiencedas both active and determinative. We speculate here that the active-determinative principle is a consequenceof a foundational cognitive system every sentient individual has and ' . namely. that of agency . then somethinglike the samestepsmay have beentraversed by genetically determined neural configurations as theseevolved. there is a remainder of emanation types to which the active-determinative principle does not obviously apply in any direct way. We remain agnostic on whether the connection is learned or innate. intending to affect the distal object must either move to it with her whole body. The realization component.. In this way. Talmy 1976 . greater agencyor energeticness ). Specifically . we can suggestsomething of the stepsand their consequentinterrelationships. the individual s exerciseof agencyfunctions experiences as the model for the Source of emanation. the characteristicsof agencymay provide the model for the active-determinative principle . as described in the next section. Either way. An intention can be understood as one' s desirefor the existenceof somenew state of affairs where one has the capability to act in a way that will bring about that state of affairs. forthcoming) . alignment paths. 6. however. The particular form of agency that can best serve as such a model is that of an ' " Agent s affecting a distal physical object.

the Experiencerhas a front . we can see how the agent-distal object pattern can serve as the model for the two main agentive fonns of emanation. such a mapping might either be the result of learning during an individual s . In both cases . like the Agent s motion continuing along this line until it Experiencer as Source es until it encounters the distal reachesthe object. and agentive visual sensory paths. Possibly some role is played by the fact that the more acute tapered end of the ann . the extending ann typically exhibits actual motion away from the body along a line that connectswith the target object. To consider the fonner casefirst . the specific agent-distal object pattern of extending the ann to reach for someobject may directly act as the model for agentive demonstrative paths. the intangible line of sight movesin a straight ' line betweenthe front of the Experiencerand the distal object. namely. the more active or detenninative entity is the Source from which fictive motion emanatesthrough spaceuntil reaching the less active or detenninative entity. leads during the extension and is furthest along the line to the object . agentive demonstrative paths and agentive sensorypaths. Specifically ' the Agent. and whose tapered end is the end closest to the distal object and the end conceptualizedas the Source from which the demonstrative line . the active-detenninative principle is based In particular . one associatedwith nonagentive type. the act of intention . the visual sensory path progress ' s motion in the the of . in which an Agent executesfactive motion toward distal object. such as an Agent extending his ann and pointing with his finger. emanates Similarly. Hence one can posit that the pattern of agencyaffecting a distal object is the model on which . But these are object progress also the characteristicsof the active-detenninative principle : namely. the fingers. and the ensuingactivity that finally affects the distal es through spacefrom that initial locus to the object." " Fictive Motion in Languageand Ception 229 of this fonn of agencyare that the detennining event. when fully extended . Such an agentive demonstrative path might then when the ann is fully extended for the in turn serveas the model . we can see parallels between the agent-distal object pattern. ' Again. like the Agent. like this line s moving away from the initial locus of the Agent. for example a figure like an arrow . ' where. in which an Experiencerprojects a fictive line of sight from himself to the distal . like object. the visual sensorypath movesaway from the ' . the distal object. whose linear axis also coincides with the line between the arrow and the distal object. or might have beenevolutionarily incorporated into the perceptual and development . Thus the physical world appears to Agent object perception be mapped onto the conceptualization of an intangible entity moving along a line. the ann s linear axis coincides with its path of motion . takes place at the initial locus of the Agent. the Experiencer is active and detenninative. like the Agent s moving along a straight line betweenhis front and the distal object.

the sun. an organism s production of factive motion can becomethe basis for the conceptualization of fictive motion .that of an ' Experiencers front emitting a line of sight that proceedsforward into contact with a distal object. and henceto justify this conceptual superimposition of an Agent onto it . A reinforcement for this mapping is that the Experiencer is determinative as the Agent and the solid object is determinative over the shadow dependenton it . onto light visible in the air (especially . for example . in the prospect type of orientation path. and with the distal object mapped onto the vista toward which the prospect line progress . which is thus conceptualizedas moving from the brighter mapped to the duller object object.onto situations amenableto a division into comparably related components . Once again.where the Experiencedemits a Stimulus. Thus. and the distal object is mapped onto the less bright . is associated .on We may account for the reversecase the grounds that it . the visual path moving from this Experiencerto the ground location of the shadow is mapped onto the conceptualization of the fictive path that the shadow itself traversesfrom the solid body onto the ground. In turn . What is required is simply the conclusion that the conceptualization of an object emitting a Stimulus can be taken as active enough to be treated as a kind of modest agencyin its own right . this modeling may occur by the conceptual mapping or superimposition of a schematizedimage. the schemafor the agentivevisual path may get mapped onto the radiation situation . with her visual path mapped onto the conceptualized schematic component of a prospect line moving away from the wall . Either way.the brightest component with the most energeticcomponent of the radiation scene in the caseof light . say. An association of this sort can explain why much folk iconography depicts the sun or moon as having a face that looks outward. . can serveas a receptiveframe onto which to superimposethe model of an Agent emitting a visual path . this agentive visual type of fictive emanation may serve as the model for severalnonagentive emanation types. the Experiencercomponent may be superimposedonto . as through an aperture in a wall ). a light beam. too . In particular . The direction from Experiencerto Experiencedis clear becausethat is the sameas for agentive viewing. The only emanation types not yet discussedin terms of mapping are the nonagentive sensory paths that can proceed in either direction. The direction of motion conceptualizedfor the visual path is also object in the scene onto the radiation . . As for shadow paths.6 es In a similar way. as the active determinative Agent.230 Leonard Talmy ' conceptual apparatus of the brain. say. The visual path is mapped onto the radiation itself. a cliff . where the Experiencer . the model may be the situation in which the agentive Experiencer herself stands and views her own shadow from where she is located. with her face corresponding to the cliff wall . say.

as well as to folk iconography. and some fundamental structural properties that it has in common with all systems . Moving the array in the direction of one of one or another of their common vertices of the common verticesblasesthe perception of the pointing to be in the direction of that vertex.1 " OverlappingSystems Cognitive Organization ' ' Converging lines of evidencein the author s and others researchpoint to the following picture of human cognitive organization. A brief description of our model of cognitive organization. referred to in the introduction . of conceptual structure. reasoning other cognitive systems .between language and each of these : visual perception.7 Relation Systems In this section we present a number of apparent similarities in structure or content between the emanation category of fictive motion in language and counterparts of ." " Fictive Motion in Languageand Ception 231 in OtherCognitive to Counterparts in Language of Emanation 6. triangles.in particular . " Model of 6. This research has considered similarities and dissimilarities of structure compass . Human cognition comprehends a certain number of relatively distinguishable cognitive systems of fairly extensive . systems 6. We call this view the overlapping by the strict modularity notion (cf. We mainly consider emanation in cognitive systemsother than that of language similarities that language has to perception and to cultural conceptual structure. The general finding is that each cognitive system has some structural properties that may be uniquely its own. One perceptual phenomenon related to orientation paths has beendemonstratedby Palmer ( 1980 ) and Palmer and co oriented of in certain who found that Bucher ( 1981 equilateral arrays consisting ). kinesthetic perception. and cultural structure. One might needexperiments . will first provide the context for this comparison. for example .7. .2 Fictive Emanationand Perception The visual arrays that might yield perceptualparallels to the emanation type of fictive motion have been relatively less investigated by psychological methods than in the caseof other categoriesof fictive motion (seebelow) . We assumethat each such cognitive systemis more integrated the cognitive systems and interpenetrated with connectionsfrom other cognitive systemsthan is envisaged " ) . some further structural properties that it shareswith only one or ~ few other cognitive . Fodor 1983 " model of 7 cognitive organization. which may be regardedas a concretesymbolic expression of perceptual structure. subjectsperceiveall the triangles at once pointing by turns in the direction . memory.7. although theseexperimentsdid not test for the perception of an intangible " " line emerging from the vertex currently experiencedas the pointing front of each triangle or of the array of triangles. attention . planning.

on viewing these arrays. The robust and extensiverepresentationof fictive emanation in languagecalls for psychologicalresearchto test for parallels to this category of fictive motion in perception . on viewing a scenethat contains procedureswill need to probe whether subjects an object and its shadow. .232 Leonard Talmy ' that test for any difference in a subject s perception of a further figure depending on whether or not a fictive line was perceivedto emergefrom the array of triangles and pass through that figure. . This work involved the more forward locations. But this effect is presumably due to the factively forward progression of the figure. Freyd 1987 ) does not demonstrate perception of orientation paths. The subjects sequential presentation of a figure in successively the last did exhibit a bias toward perceiving presentedfigure further ahead than its actual location. Similarly . hencewith the concurrent perception of two discrepant representations . One would need to determine whether subjects . but concurrently ' sensethe fictive motion of something intangible emanating from the objects fronts at a faintly palpable level of perception. concurrently with having moved from that object to the surfaceon which it appears a factive and palpable perception of everything within the scene as stationary. demonstrative paths. the question remains whether the appropriate experimental arrangements will show particular perceptionsfor this category that accord with the general fictivity pattern. one Finally . the factive and more palpable perception of the beam as static. Similarly . That is. in a direction away from the brighter object. But confirmation of a perceptual analogue to emanation . paths must await such research " ' " We can also note that Freyd s work on representationalmomentum (e. one of them more palpable and veridical than the other. to check for a perceptual analogue of visual sensorypaths in language ' can use either a scenethat depicts someonelooking or a subject s own processof looking at entities to ascertain whether subjects simply perceive a static array of entities or superimposeon that array a lesspalpable perception of motion along the probing line of sight.. to probe for visual counterparts of linguistic radiation paths. and targeting paths. that is concurrent with .g. designed to test the perception of fictive orientation paths in their several distinct types.prospect paths. for example visual arrays that include various front -bearing objects. see the factive stationariness of the depicted objects at the fully palpable level of perception. research will need to test for anything like a fictive and less palpable perception of motion along a light beam. experimentsof this type would need to test subjectson the presentation of a single picture containing a forward -facing figure with an intrinsic front . less palpable senseof the shadow as . experimental . alignment paths. have some fictive. perhapssuperimposedon. To check for the perceptual counterpart of linguistic orientation paths. . to test for a visual parallel to linguistic shadow paths. Consider.

movies and comic Agent can project forth from his extendedfingertips. seemsalso to parallel a type of iconographic depiction. as in Jane lookeddaggersat John. For example . daggersgoing from the experiencers eyesto the body of the experienced The linguistic conceptualization of fictive demonstrative paths emerging from the point -type front of a linear object.7. But this is exactly the character of ' " " Supermans Xray vision as depicted in comic books. This is the depiction of magical power beamsthat an . eyes conceptual emits a sensoryProbe appearsto hold sway in the cartoon imagery. Such toward and into Superman eyeswhere they might be perceptually registered an Experiencedto Experiencer path direction might have been expected from our understanding of Xray equipment.cleaner sculptures to explicitly ' . An example to be cited can sometimesbe modeled by fully palpable representations below is the use of stick figure drawings or of pipe. it is the author s observation. iconographic . cartoon depictions actually show a line of ' . as from a pointing finger.3 Fictive Emanationand Folk Iconography Fictive representationsthat are normally only sensedat a lower level of palpability . which is normally only sensed various other aspectsof fictive emanation normally only sensedhave been made explicit in the concrete depictions of folk iconography. but also by metaphoric expressions . reflecting a current feeling of hate for John. Thus the expression" to look closed ' " daggersat . Note that Superman otherwise obscured object and permits it to be seen Xray vision is not depicted as stimuli that emanate from the obscured object and proceed 's . Supermansendsforth from his eyesa beam of Xrays that penetratesopaque materials to make contact with an 's .though a careful study would be neededto confirm this.that in the processof drawing the sun. Further . representsthe notion that Jane s mien. this iconographic procedure reflects the linguistic conceptualization of fictive radiation paths as emanating and moving off from the brightest object. not inward toward it . can be elaborated as the projection of weapons from her eyes to John. Comparable examplesbasedon the linguistic conceptualization of an Agent emitting a visual Probe are representednot only by grammatical constructions and other -classforms. represent its radiation body with lines drawn radially outward from the circle. image objects schematicstructure. If so. fictive sensory paths of the agentive visual type are linguistically conceptualized as intangible lines that Agents project forward from their eyes through space into contact with distal objects. This plate might have onto a photographic plate on which the image is registered 's in which the Agent but the model been analogized to Superman . ' Finally . For example books often have two battling sorcerersraise their extendedhands and direct destructive beamsat each other. after completing and adults for the of the sun both children a circle .Fictive Motion in Languageand " Ception" 233 6. where the radiation moves from the equipment . In the sameway. indeed. schematically .

for example . such as invisibility or the ability to passthrough walls or other solid objects. The exceptional phenomena found to occur in ghost physics may be the same as certain cognitive phenomenathat already exist in other cognitive systems . we can suggestthat the pattern of standard and exceptional properties is structured and cognitively principled . and (2) it passesthrough walls. turns around and then movesin the direction of the town. building. 6.234 Leonard Talmy representationsof the sun and moon often depict a face on the object. regardless of the structure around me. which on other grounds might have seemed just as suited for conceptualization as specialproperties.6. An exampleof this is temporally backward causality.and the properties exhibited by ghosts or spirits in the belief systemsof many traditional cultures. through gesticulations . The linguistic expression of fictive demonstrative paths and its gestural counterpart may well afford the relevant properties. On the contrary. That is to say. itself has the following crucial properties: ( I ) it is invisible. with only a few exceptions that function as " attention attractors. I will not . However. if I . To consider gesturefirst . that is. much as it may be mapped onto other fictive motion types. the findings reported in this chapter may supply the missing account. cultural belief systemsseemuniversally to lack a concept that a ghost can at one point in time bring about somestate of affairs at a prior point of time. but other kinds of potential exceptions . Boyer has no explanation for the selection of particular exceptions that occur in ghost physics and may even find them arbitrary . As noted in section 6. Boyer holds that ghost and spirit phenomena obey all the usual causal physics expectations for physical or social entities. through open doorway.in particular.2. In fact.4 Relation of Fictive Emanationto Ghost Physicsand Other Anthropological Phenomena We can discern a striking similarity between fictive motion . instead appear never to occur. as if to represent the object as containing or comprising an Agent that is emitting the radiation of light . the demonstrative path. The anthropologist Pascal Boyer ( 1994 ) seesthese properties as a culturally pervasiveand coherent conceptual system which he calls " ghost . and then are tapped for service in cultural spirit ascriptions.the very same properties ascribedto spirits and ghosts. " . .7. a representationof this sort can be attributed to the mapping of the schemaof an agentive visual sensorypath onto the radiation situation . orientation paths. indicate a at leads the out the exit of the that path begins my finger. I will simply extend my arm with pointed finger in the direction of the town ." Certain of these exceptions have widespread occurrence acrossmany cultures. effectively conceptualizedas an intangible line emergingfrom the finger. am inside a windowlessbuilding and am asked to point toward the next town .

the use of any of the directional prepositions suggeststhe conceptualization of an intangible line emerging from the front end of the arrow . visual perception. by Keil ( 1989 ) and Carey ( 1985 ) for other categoriesof cognitive phenomena . in the set of sentences expressionof fictive demonstrative paths. an agent who bearsmalevolent feelings toward another person is able to transmit the harmful properties of thesefeelingsalong the line of his gazeat the other person. Once again. as propagating in one or more directions away from that entity. fields of life force. power. . This is the sameschema as for a fictive visual path : the Agent as Sourceprojecting forth something intangible along his line of sight to encounter with a distal object. this imaginal line is invisible and would be understood to passthrough any material objects presenton its path . or magical influence emanating from entities. as being (generated conceptualized(and perceived and) emitted by someentity. Rather. One may look to such broadly encounteredcultural conceptsas those of mana. For example arrow points to/ toward/past/away from the town. that of agentive " visual paths. systemsof language .just like the fictive emanations of linguistic construals. In addition to such demonstrative paths. That is. following a straight coursecoaxial with the arrow ' s shaft. found in the conceptual systems of many cultures. these forms of imagined energy. and moving along the path representedby the preposition. we can observefurther relations between cultural conceptualizations and another type of fictive emanation. but also of cultural cognition. It thus seemsthat the general fictivity pattern generatesthe imaginal schemasof fictive motion in the cognitive systemsnot only of languageand of visual perception. the structure of such conceptions as power.are ?) as being invisible and intangible. specifically in its conceptualizations of spirit and . in the cognitive culture system harmful influence and . Relations between fictive motion and cultural conceptualizations extend still further . general systems fictivity of this sort is thus one area of overlap across at least the three cognitive . In terms of the " overlapping systems . for example . In a frequent conception of the evil eye. Nor does it exhibit its own mode of construal or constitute its own domain of conceptual constructs of the sort posited. and cultural cognition . and in some forms as then contacting a second distal entity that they may affect.Fictive Motion in Languageand " Ception" 235 Thesesameproperties hold for the conceptualization that accompaniesthe linguistic this . The structural parallel between such anthropological concepts of emanation and the emanation type of fictive motion we have here described for language is evident and speaksto a deepercognitive connection. it is probably the same as or a parallel instance of conceptual organization already extant in other cognitive " framework outlined above . Consider the notion of the " evil eye . ghost phenomena magical energyappearsnot to be arbitrary .

The spots thus enclosed within the envelopeor positioned along the line can then be cognizedas constituting a unitary Gestalt linear pattern.8 Further Categories of Fictive Motion Leonard Talmy As indicated earlier. if the sentencewere to be were to be treated interpreted literally . if the literal referenceof the sentence as factive one would have to believethat the spots of paint physically slid forward along the floor .8 The purpose of this section is to enlarge both the linguistic -perception parallelism.1 Pattern Paths The pattern paths category of fictive motion in languageinvolves the fictive conceptu. For the fictive effect to occur. [cf.8.236 6. but that motion is vertically downward in falling to the floor . or appearance /disappearance change fictive motion . Such is the forward fictive motion of the configuration . it is the pattern in which the physical entities are arranged that exhibits the fictive motion . the literal alization of some configuration as moving through space senseof a sentencedepicts the motion of some arrangement of physical substance along a particular path. with that follow . For this fictive effect. In the illustrations scopeand the scopeof potential language are provided. Rather. Consider the example in (21) . By contrast.that is. shown within brackets. for each. while we factively believethat this substanceis either stationary or movesin someway other than along the depicted path . we suggestsomeparallels in visual perception that have already been or might be examined. (21) Pattern paths As I painted the ceiling. the fictive motion sentences factive motion counterpart sentences . We here briefly sketch five further categories . the physical entities must factively exhibit some form of motion . (a line of ) paint spots slowly progressedacrossthe floor . In this type.] Here eachdrop of paint does factively move. (a line of ) ants slowly progressedacrossthe floor . one must in effect conceptualize an envelope located around the set of paint spots or a line located through them. rather. languageexhibits a number of categoriesof fictive motion beyond the emanation type treated thus far. 6. The fictive motion . The appearanceof a new paint spot on the floor in front of one end of the linear pattern can then be conceptualizedas if that end of the envelope or line extended forward so as now to include the new spot. but thesein themselvesdo not constitute the . qualitative . . As I painted the ceiling. as a foil for comparison. is horizontally along the floor and involves the linear pattern of paint spots already located on the floor at any given time.

is that the emanation type does not involve the factive motion of any elements within the referent scene . so that no additional principle need come into play .the fictive motion of a seeminglysingle light progressing along the row of bulbs. one may needto subdivide apparent language motion into different types. as factive.2 Frame -Relative Motion With respect to a global frame of reference .Fictive Motion in Languageand " Ception" 237 In one respect . subjects perceive a single dot moving back and forth between the two parameters locations. This condition is illustrated for English in (20a) and is diagrammedin figure 6. Within this frame. the perceptual representation most palpable to subjectsis in fact that of motion . But to establish the parallel correctly. a subject would have an experiencethat fits the generalfictivity pattern . But the pattern paths type does require the factive motion or the change of some components of the referent situation for the fictive effect to occur.to fix the source and direction of the fictive motion .the active-determinative principle. the pattern paths type of fictive motion is quite similar to the emanation type. Accordingly. an entity that is itself fictive. indeed. involve different perceptual mechanisms Most researchon apparent motion has employed a format like that of dots in two locations appearing and disappearing in quick alternation. the stationary state of the bulbs.and assess of veridicality .moves fictively through space . a language can factively refer to an 's observeras moving relative to the observer stationary surroundings. In both these categories of fictive motion .la .an imaginal construct. and thus would not correspond to the . there may exist a slower type of apparent motion that can be . 6. The subject will perceiveat a higher level of palpability . this determines the direction of the fictive motion . One difference . In this fast form of apparent motion . linguistic case On the other hand. though. it must depend on a principle . Such types are perhaps largely basedon the speedof the . as well as the periodic flashing of a bulb at different locations. Here.8. within certain . process viewed and. it may be sUr Dlised . But a languagecan alternatively refer to this situation by adopting a local frame around the observer as center. that is. But the subject would concurrently it as being at a lower level perceiveat a lower level of palpability . the observer can be represented as stationary and her surroundings as moving relative to her from her perspective . One example might consist of a perceivedand that would parallel the linguistic case a row of bulbs in which one after another bulb is briefly turned subject viewing light on at consciously perceivable intervals. one may speculate . The perceptual phenomenagenerally termed apparentmotion in psychology would seemto include the visual counterpart of the pattern paths type of fictive motion in . This condition is . Here.

Globalframe : Fictive motion absent I rode along in the car and looked at the scenerywe were passingthrough . we term the fictive effect here observer based frame relative motion. B. part factive and part fictive. Stressingthe depiction of motion . * Leonard Talmy illustrated in (20b) and diagrammed in figure 6. In a complementary fashion. Local frame : Fictive motion present I sat in the car and watched the sceneryrush past me. one in which the factively stationary surroundings are fictively depicted as moving . I sat in the movie set car and watched the backdrop sceneryrush past me. For instance situation within a single sentence . accordingly. which suggeststhe adoption of a perspective point midway betweenthe observerand her surroundings. and accordingly. But one condition no language seems spatial conditions to a fictive representation able to representis the adoption of a conceptualization that is part global and part local.] . this condition also contains a form of fictive stationariness . Thus English is constrained against sentenceslike (220 ).238 [!] 0 Figure 6. for the factively moving observeris now fictively depicted as stationary. [cf. shifts from a directly factive representation of the . (22C) shifts from the global frame to the local frame and.1 Frame-relative motion : global and local. (22) Frame-relative motion : With factively moving observer A . Shift in midreference from global to localframe . Further . a languagecan permit shifts between a global and a local framing of a . andfrom factive to fictive motion I was walking through the woods and this branch that was sticking out hit me.] C. This condition is thus a form of fictive motion .1b. I was walking through the woods and this falling pineconehit me. [cf.

If such a local frame were possible sentences that the observer as and the observedas acceptable fictively depict moving . part-fictive motion * We and the sceneryrushed past each other. (23 Ba) with a attempting this depiction for example uniform local framing and (23 Bb) with a shift from global to local framing . but a sentencemay also " ratchet down" its representationof the situation to the basic condition in which the observer is stationary. We can suggest an account for the difference between moving and stationary observers in their acceptanceof fictive local framing . its water rushed past me. a stageat which it will agentively bring about optic flow itself. * As I sat in the stream.lc . The unacceptable fictive local framing that they attempt is diagrammed unacceptable in figure 6.] In the precedingexamples . We can suggesta developmentalaccount. b. But sentences stationary .g. one could find stationary observer as center.ADg \ lage and Ception 239 D . In a complementary fashion. if the observer is already stationary. We and the [ logging truck rushed past each other. a sentenceis free to representthe situation as such. the next question is why it should be that stationariness is basic for an observer. An infant experiences optic flow from forward motion while being held by a parent long before the at stage which it locomotes. it has months of experienceof optic flow without an experienceof motion . the observerwas factively in motion while the observed (e. The stream flows past my house . This framing in which. cf. However. (23) Frame-relative motion : With factively stationary observer A . then a sentencemay only representthe situation as such. part -localframe withpart -factive. Local frame : Blocked attempt at fictive motion a. This earlier experiencemay be processedin terms of the . If this explanation holds. Globalframe : Fictive motion absent a. if an observeris factively moving. Ab ) . * My house advancesalongsidethe stream. the scenery ) was factively stationary. a sentencecan also expressa global .." " Fictive Motion in I . B. factively. Lacking: Part -global. As I sat in the stream. b. that is. this complementary situation differs from the earlier situation in that it cannot undergo a local reframing around the . That is. However. Accordingly. and is not free to " ratchet up" its representationof the situation into a nonbasic state. before the infant has had a chanceto integrate its experienceof moving into its perception of optic flow . The main idea is that stationarinessis basic for an observer .are .properties expressedexplicitly in the global framing . the observeris stationary while the observedmoves situation is illustrated in (23 Aa . already in the basic state. I rushed through its water.

we watchedthe heavens spin around us. this is readily found .240 Leonard Talmy surrounding world as moving relative to the self fixed at center. Indeed. Closer to our linguistic caseis the " motion aftereffect. we should not expectto find a linguistic effect that treats observerstationarinessas basic relative ' s arc-sized to an observer . somesubjectsfictively perceivethis frame as stationary while the rod moves in a complementary manner. However. prototypically . this genre -basedin our sensebecausethe observer is not one of of experimentsis not observer the objects potentially involved in motion . representationslike thosejust seen One possible corroboration of this account can be cited. A sentence sort would be used only for special effect. typically turning motion . hencefictive. it should behave like forward translational motion and pennit a linguistic refraIning. although not optic flow just discussed extended rotation . for one language pennits only factive representationsof such turning by an observer . I looked over all the room s decorations . Indeed.assessed as lessveridical. This experiencemay be the more foundational one and persist to show up in subtle effects of linguistic . spedby front of me. Psychological experiments have afforded severalprobable perceptual parallels to frame-relative motion in language . The a perception subject factively knows that he is stationary. As I quickly turned my head . English. namely. On the other hand. Rather than the forward type of . as in English sentenceslike As our space shuttle turned. Becausethe infant can thus integrate the experienceof motor control in with experienceof transverseoptic flow at a foundational level. not in the everyday colloquial way the forward motion case is treated. for example ' . ratchet down to a fictive state for the observer as in As I quickly ." present where a subject has been spun around and then stopped. It does not . . Infants at the outset do have one fonn of agentive control over their position relative to their surroundings. and a fictive representation of herself as stationary with the surroundings as moving toward and past her.of the surroundings as turning about him in the complementary direction. The question is whether such a subject will concurrently perceivea factive representation of herself as moving through stationary surroundings . much as when riding in a car. One parallel is the " induced motion " of the " rod " and frame genreof experiments . Perhaps the experimental situation closest to our linguistic type would in fact be a subject' s moving forward through surroundings. typically stationary 's decorations turned my head the room in of the latter . Here. as still further corroboration . or I rode on the carouseland watchedthe world go round. this action brings about a transverse type. becauseextended spinning is not part of the infant ' s early experience . but concurrently experiences . turning the eyesor head through an arc. while a rectangular shapethat surrounds a linear shapeis factively moved.

a.] For a closer look at one site arrival example . and site manifestation .located in the sites they occupy. / distributed all over / strewn/spread [cf. involving the fictive motion of the object to its site. which is not fictive the object at its site. But the concurrent factive representation of this scene is around the oasis contained in our belief that the treeshave always beenstationary. Site manifestation d. / shows up near /appears /reappears [cf.Fictive Motion in Languageand " Ception 6. or manifestation at the site it occupies whereasits depicted motion or materialization is fictive and. The two main subtypes of advent paths are site arrival . Comparably. the site manifestation examplein (24d) literally represents the location of the rock formation at the sites it occupiesas the result of an event of materialization or manifestation. With active verbform . (24a) uses the basically motion specifying verb to cluster for a literal but fictive representation of the palm trees as having moved from some more dispersedlocations to their extant neighboring locations . This rock formation occurs/ recurs /reappears / shows up near /appears volcanoes.8. Termite mounds are scattered / distributed all over the / strewn/spread plain .] B. often wholly implausible.3 . namely This category is illustrated in (22) .] b. / tilted away from the wall. The palm treesclusteredtogether around the oasis the ice cream clustered around The children cf: together [ quickly truck . Pentland . Site arrival I . With passiveverbform c. This fictive representation is concurrent with our believed factive representation of the rock formation as having stably occupied its sitesfor a very long time. The beam leans / tilts away from the wall. Advent Pat I L " 241 ' An adventpath is a depiction of a stationary object s location in terms of its arrival . fictive manifestation of the motion but fictive change . The stationary state of the object is factive.] [cf: The loose beam gradually leaned 2. (24) Advent paths A . Gopher traps were scattered the plain by a trapper. Ball lightning occurs/ recurs volcanoes. in fact. We can cite two psychologistswho have made separateproposals for an analysis of visual forms that parallels the linguistic site arrival type of fictive motion .

but rather as the result of some processof deformation applied to an unseenbasic form . depending on the . ' some body part of a person. certain forms are regularly perceivednot as original patterns in their own right . of the static will be that more as palpable. whether this is plausible or implausible. The cloud is 1. has the remaining portions . or the focus of a person s attention . squashing es correspondsto the psychologically salient tance. a smooth closed surface is describedas the deformation . for example . What is factive here is the representation of the object as stationary. without any entity traversing the depicted path.8. particular sentence (25) Accesspaths a. Comparably.242 Leonard Talmy ( 1986 ) describesthe perception of an articulated object in terms of a processin which a basicportion of the object. [cf.] . Though it is not specified . To consider this last example in terms of our general fictivity pattern. 6. andresis . [cf.] c.000 feet up from the ground. as can be seenin the examplesof (25) . The bakery is acrossthe street from the bank. and ends with that wedge exiting or being removed from the circle. indentation . The balloon rose 1. say. what is fictive is the representation of some entity traversing the depicted path. In a similar way.000 feet up from the ground .4 AccessPadis ' An access path is a depiction of a stationary object s location in tenns of a path that some other entity might follow to the point of encounter with the object. that starts with a circle. The vacuum cleaner is down around behind the clothes hamper. a subject looking at such aPac -Man shape may concurrently experiencetwo discrepant perceptual . proceedsto the demarcation will consist of an imagined sequence of a wedge shape within the circle. held to be the more veridical and representations PacMan configuration per se. its central mass moved into attachment with it . The ball rolled acrossthe street from the bank. for example simple surface of a sphere . The factive representation . I extendedmy ann down around behind the clothes hamper. of a bent pipe or a dented door. [cf.] b. An example is the -shapedpieceremoved perception of aPac-Man -shapedfigure as a circle with a wedge from it . Leyton ( 1992 ) describesour perception of an arbitrary curved surface as a deformed version of a . as describedin the tradition of Gestalt psychology. felt as being perceivedas lesspalpable. He shows that this set of process causal descriptions that people give of shapes . one that has undergone protrusion . An example is the perception of a clay human figure as a torso to which the limbs and head have beenaffixed. the fictively moving entity can often be imagined as being a person. perceived and less veridical The fictive representation .

depending on the particular sentence examplesof (26) . the fictively moving over the configuration of the object." Fictive Motion in Languageand Ception " 243 In greater detail. Subjectscan a kind of experimental design that might test for the phenomenon be shown a pattern containing some point to be focusedon. This path could be followed physically by a person walking. and then recrossto find the bakery. advanceto the bank. A subject might factively and at a high level of palpability perceive a static representation of this . as when I use (25a) to direct you to the bakery when we are inside the bank. a depicted access path can also be physically implausible or that in That quasar is 10 million lightlike impossible. What is fictive is the representation . Further . a B to be focusedon. with the B simply located on the left.8. or the object often can . (25a) characterizesthe location of the bakery in terms of a fictive path that beginsat the bank. Though it is not specified of attention the focus an observer as be . that is further correlated with a second fictive change . 6. redness ) (increasing . and in (26c) it is the lateral motion of a line (a north -south line advancing eastward ). The depicted path can be reasonable for physical execution. Perhaps example A at the top point and.it is unlikely that you will first cross the street. radially outward over a ( ) two dimensional plane. or location of a spatially ' extendedobject in terms of a path over the object s extent. But concurrently figure much as just described . as when I use (25a) to direct you to the bakery when we are on its side of the street. where the whole can be as involving paths perceived factively as a static geometric Gestalt and/ or fictively " " be a would an plus figure with the letter leading to the focal point .5 CoveragePaths A coveragepath is a depiction of the form . What is factive here is the of any entity traversing the representationof the object as stationary and the absence some of entity moving along or depicted path . or perceptually by someone . as in The bakery and the bank are oppositeeachother on the street. Such is the casefor referents . orientation . at the left hand point . Apart from the useof fictive access ' s location can in a factive characterized also be representation directly generally object . say. either slants directly toward the B. the subject might fictively and at a lower level of palpability perceivethe B as located at the endpoint of a path that starts at the A and. an paths such as these yearspast the North Star. it is in 26b is linear Note that in (26a) the fictive path . But the samedepicted path may also be an improbable one. . proceedsacrossthe street. or solely conceptually by someone shifting her shifting the focus of his gaze attention over her mental map of the vicinity . Does the fictivity pattern involving access paths occur perceptually? We can suggest . imagined being entity in the be seen as can itself. and terminatesat the bakery. or moves first down and then left along the lines making up the " " plus.

to the other end of the fencein the valley. in which an observer or our focus of attention or some of the fenceitself advancing . we have a factive representation of the fence as a stationary object with linear extent and with a particular contour . Babcock and Freyd 1988 ). The fencegoes /zigzags / descends from the plateau to the valley. as I view a particular figure before me. The oil spreadout in all directions from where it spilled. moves from one end of the fence atop the plateau. a. If further distinctions have been adduced . is my identification designation. 6. perhapsa subject viewing a plus configuration along its delineations. / descended [cf. [cf. concurrently. On the one hand.] Consider the fictivity pattern for (26a) . . perhaps image along its own axis. while implicitly sensingsomething intangible sweepingfirst downward along the vertical bar of the plus and then rightward along the horizontal bar (cf. One motivation for challenging the traditional categorization is conception that psychologistsdo not agree on where to draw a boundary through observable psychological phenomenasuch that the phenomenaon one side of the boundary will be considered" perceptual. or the contrasting of the whole category of perception with that of /cognition. and location in geographic space . We can ask as before whether the generalfictivity pattern involving coveragepaths hasa perceptualanalogue . together with representationsthat appear in language . perceivedfictively at a lower level of palpability in terms of pathways " " . The field spreadsout in all directions from the granary. orientation . Concurrently. we have the fictive representationevoked by the literal sense of the sentence .] c. they have been the separatedesignation of part of perception as sensation . For example . Much psychological discussion has implicitly or explicitly treated what it has termedperceptionas a singlecategory of cognitive phenomena . though. (2) The weather front advancedtoward the east. The soil reddenstoward the east. I went/zigzagged b.244 Leonard Talmy (26) Coveragepaths from the plateau to the valley. we suggesta general framework that can accommodate the visual representationsinvolved in general fictivity . The phenomenonmight be found in a visual configuration perceived factively at a higher level of palpability as a static geometric form and.9 " Ception" : Generalizingover Perceptionand Conception In this section. along its length. ( I ) The soil gradually reddenedat this spot due to oxidation." while those on the other side will be excluded from that . For example " " will seeit explicitly as just such a plus shape . [cf.

It seemsthat theseparameters tend to covary or correlate with each other from their high to their low ends. advisable to establish a theoretical framework that does not Accordingly. perhaps later. that characteristic. it seems a cognitive discrete imply categoriesand clearly located boundaries. mental imagery. To this end. that is. conception. throwing the baby out with the bathwater. The main advantageof the ception framework in conjoining the domains of perception and conception is not that it eliminates the difficulty of categorizing certain . too . We here propose thirteen parameters of cognitive functioning that appear to extend through the whole domain of ception and to pertain to general fictivity . seemamenableto alignment with the . but also on whether there is a principled basison which one can even adduce such a boundary. but these . Further . what about my thought of potential danger that occurs on viewing the object? Moreover. Most of these parameters seem to have an at least approximately gradient character. in that it by fiat discards a potentially useful distinction simply becauseit is troublesome. Though helpful. rather.conception distinction in a graduated form . such parametersare largely gradient in character and so can reintroduce the basis of the discrete perception. While perhaps best limited to the phenomena of current processing ception would include the processingof sensory stimulation . is precisely that it allows for the positing or recognition of distinctional parametersthat extend through the whole of the new domain. Someof the parameters distancealong the gradients of the respectiveparameters seemmore to have discrete regions or categorial distinctions along their lengths than to involve continuous gradience .with their highest perhapsranging from a fully smooth to a merely rough gradience value at the most clearly perceptual end of the ception domain and with their lowest value at the most clearly conceptual end of the domain. The strength of the ception framework . and that recognizes domain encompassingtraditional notions of both perception and conception. An individual currently manifesting such " " 9 processingwith respectto someentity could be said to ceive that entity ." Fictive Motionin Language and" Ception 245 of it as a knife to be understood as part of my perceptual processingof the visual ? And if stimuli . we here adopt the notion of " ception" to cover all the cognitive phenomena . cognitive processing such identification is consideredpart of perception. parameters whose unity might not be readily spotted across agerryman dered category boundary . Such a framework would then further allow for the positing of certain cognitive parametersthat extend continuously through the larger domain (as describedbelow) . psychologists not only disagreeon where to locate a distinctional boundary. any particular cognitive representationwill tend to merit placementat a comparable . taken by itself problematic cognitive phenomena could also be seen as . understood by the conjunction of perception and . or instead part of some other. and ongoingly experiencedthought and affect. consciousand unconscious .

There is no assumption that levels along theseparameterscorrespond to . More generally. an entity is experiencedas being abstract. To serveas referencepoints. or murky . and . that of the palpability -relatedparameters This entire proposal of palpability related parameters is heuristic and programmatic . One of the thirteen parameters other parameters . Thus perhapssomeof the parameterspresentedbelow should be merged or split . and definite. and. Given vision involved with be the most to centrally appears that the other twelve parameterslargely correlate with this one. unmanifest. the one that we term palpability . conversely . In this section.246 Leonard Talmy . four levels can be designated . These levels of palpability are discussedthe next four sections and illustrated with examplesthat cluster near them. At the low end. we present the . manifest. At the low experiencedas being concrete end. indefinite. The parameter of clarity is a gradient at the high end of which an entity is experienced as being clear. implicit . intangible. One issue is whether the set of proposed parameters is exhaustive with to . 3. 6. indistinct . tangible. the semiabstract . explicit .9. At the low end. the proposed parameters are all wholly appropriate to those phenomena Another issueis the partitioning of general visual fictivity that results in the particular cognitive parametersnamed. The parameter of palpability is a gradient at the high end of which an entity is . though. these thirteen proposed palpability -related parameters thirteen parametersare treated strictly with respectto their phenomenologicalcharacteristics . we would first need to show that our proposed parameters are in synchrony.aligned from high end to low end. and palpable. we term the whole set . distinct . we would need to show that the listed parameters are sufficiently . and impalpable. Conversely justify their being classedtogether as components of a common phenomenon . whether respect palpability and generalfictivity (presumably not ). an entity is experiencedas being faint or dull . the semiconcrete along this gradient: the (fully ) concrete the (fully ) abstract. from the fully palpability with which some entity is experiencedin consciousness concrete to the fully abstract. The parameter of strength is a gradient in the upper region of which an entity is 1O experiencedas being intense or vivid . . It will require adjustmentsand experimental confirmation with regard to several issues .sufficiently to .1 Palpability and RelatedParameten The parameter of palpability is a gradient parameter that pertains to the degreeof . instead of independent from each other to justify their being identified separately treated as aspectsof a single complex parameter. 2. As they are discussedhere. other cognitive phenomenasuch as earlier or later stagesof processing 1. -related general fictivity . an entity is experiencedas being vague.

" " Fictive Motion in Languageand Ception 247 4. The parameter of identifiability is the degreeto which one has the experienceof recognizing the categorial or individual identity of an entity . the structural end reducesor boils down and regularizes this for In to its . that one can assign it to a familiar category or equateit with a familiar unique individual . At the low end of the gradient.the assessments with the high end of other parameters pertain to the substantive makeup of an entity . At the high end of the gradient. one may experiencethe entity as having a location but as being unable to detennine it . product cognitive an entity as 6. for its 8. one s experienceis that one recognizesthe ceivedentity . when such a fonD is considered overall in its entirety. At the low end of the gradient. At the high end of the ' gradient. and in the taste modality . ostension amounts to an entity s overt sound qualities. Either way. its flavors. and that this location occupiesonly a delimited portion of the whole spatial field. can be detennined. On the other hand. and as having its experiencedas being real. Progressingdown the gradient. At midrange levels of the gradient. texturing including ' modality .a body . and that it thus has a known identity . one s experience is that the entity doeshave a location . 7. In the movements of and coloration its fonD auditory . the parameter of ostensioncomprisesthe degree to which an entity is experiencedas having such overt substantiveattributes.which correlates . the " " . The ostensionof an entity is our tenD for the overt substantive attributes that an entity has relative to any particular sensory modality .specifically there.thus. A fonD can be a simplex entity composedof parts abstractedor idealized lineaments or a complex entity containing smaller entities. the structure end can . While the content end deals with the bulk fonD of an " " entity. As a gradient. The parameter of objectivity is a gradient at the high end of which an entity is . At the structure end of the parameter which correlates with the low end of other parameters the assessments pertain to the schematic " " delineations of an entity . ostensionof an entity includesits appearance and motion . as external to oneself . In the visual modality . Such own intrinsic characteristics experiencedas being out entity ' ' . the componentsof this experience diminish until they are all absent at the low end." that is. the content end can provide the comprehensive ' summary or Gestalt of the fonn s character. . The parameter of /oca/izabi/ity is the degreeto which one experiences having a specific location relative to oneself and to comparable surrounding entities ' within somespatial referenceframe. one can have the experience that the concept of location does not even apply to the ceived entity . as having autonomous physical existence " further is an . pattern . the entity is experiencedas being subjective ' s own mental !! . and is in fact known. of one a construct activity . 5. if not also one s . At the content end of this gradient. more specifically . The content /structure parameter pertains to whether an entity is assessed content as against its structure. to one s mind .

qualitative or approximative . At the low end.remains an issue . At the low end.248 Leonard Talmy reveal the global framework . but it can also occur in their absence . respectto each of the thirteen parameters . Our characterization of each level of palpability below will generally indicate its standing with . the entity is regularly inaccessibleto . but attention. one experiences uncertainty about the entity . At a . or network of connectionsthat binds the components of the form together and integrates them into a unity . consciousness 11. the assessments pertain to the content of an entity and are (amenableto being) geometrically Euclidean. The parameter of certainty is a gradient at the high end of which one has the experienceof certainty about the occurrenceand attributes of an entity . and are (limited to being) geometrically topological or topology-like . But for the sake of simplicity . metrically quantitative . the entity is currently not in consciousness could readily become so. At the low end.or . But the manner in which the various modalities behave . 9. At the high end of this parameter. The parameter of stimulus dependence current on line sensorystimulationI in order to occur. form . one experiences doubt about it . At the high end of the parameter or in the foreground of attention. 12. The terms for all the above parameters were intentionally selectedso as to be neutral to sensemodality . At the lowest end. is the degreeto which a particular kind of 13.in possibly different ways. We with respect to the parameters briefly addressthis issuelater. The parameterof actionability is a gradient at the high end of which one feelsable to direct oneself agentively with respect to an entity . and absolute. In the midrange of the gradient. preciseas to magnitude. the first three levels of palpability presentednext are discussedonly for the visual modality . movements end of the parameter. Along the gradient parameterof accessibilityto consciousness . experienceof an entity requires At the high end. one feelscapable only of receptiveexperience of the entity . more actively. 10. sensory require. the experience stimuli .for example . an entity is accessible to consciousness everywherebut at the lowest end. the entity is in the periphery of consciousness or in the background of or attention . stimuli must be present for the experienceto occur. to inspect or manipulate the entity . together with the degreeof its precision and absoluteness acterization. schematic . and relational or relativistic. the experience can be evoked in conjunction with the impingement of does not . and so on. At the low . to stimulation for its occurrence relation or has no . pattern. the entity is in the center of consciousness lower level. The type of geometry parameter involves the geometric characterization imputed of one' s charto an entity. Still lower. the assessments pertain to the structure of an entity.

and is regardedas an instance of substantivecontent. change position relative to it . For example . An apparently further .at the corresponding point of whatever scene one now looks at. A first example " at each intersection of a semiconcreteentity is the grayish region one " sees (except the one in direct focus) of a Hermann grid .2 ConcreteLevel of Palpability At the concrete level of palpability . It is dependent cognizing accordingly experiencedas being out " ' there. The viewer can experiencethe entity with full consciousness and attention . This grid consistsof evenly spacedvertical and horizontal white strips against a black background and is itself seenat the fully concrete level of palpability . in the visual case . with the ostensivecharacteristicsof precise form . 6.for example .9. palpability related parameters Examplesof entities experiencedat the concretelevel of palpability include most of the manifest contents of our everydayvisual world . As one shifts one' s focus from one intersection to another. The entity is usually recognizable for its particular identity . a representationceived at the concrete level of palpability is generally experiencedas factive and veridical. after a bright light has beenflashedon one spot of the retina. Another example of a semiconcreteentity is an afterimage. has a sense of certainty about the existenceand the attributes of the entity. a spot appearsat the old locus and disappearsfrom the new one.hence not as entity is experiencedas having real.9. and movement. In short one the at the . one must be actually at the . where this precision largely involves a Euclidean-type geometry and is amenableto metric quantification . The . coloration . before outlining their generalcharacteristics . physical. one ceives a medium grayish spot. or perhaps manipulate it to expose further attributes to inspection. looking entity experiences entity high end of all thirteen .Fictive Motion in : and " Ception" 249 6. Comparably. after staring at a colored figure. that is. It can function as the background foil against which a discrepant representationat a lower level of palpability is compared. and feelsvolitionally able to direct his or her gazeover the entity. With respectto general fictivity . suchas an apple. Outside of abnormal psychologicalstates(such as the experiencing of vivid hallucinations). as clear and vivid . and with a precise location relative to oneselfand to its surroundings. or a street scene . an entity that one looks at is experiencedas fully manifest and palpable. texture.3 Semiconcrete Level of Palpability We can perhaps best begin this section by illustrating entities ceived at the semiconcrete level of palpability . one ceivesa pale image of the figure in the complementarycolor when looking at a white field. not as a construct in one s mind. autonomous existence ' s own " on one of it . this concreteexperience of an entity requirescurrently on-line sensory stimulation .an " artificial scotoma" .

it moves about freely relative to the concretely ceived external . or ghostlike. But . being able to ' manipulate them only by moving one s eyes about. But an afterimage. artificial scotomas . it is one of fictive presence .a shifting pattern of light that spansthe visual field. as lessclear. but the entity is otherwise afterimage cases entity is partially preserved largely not amenableto categorization as to identity . Rather. is assessed spots. The type of discrepancypresentbetweentwo such concurrent representations of a single scene is generally not that of fictive motion against factive stationariness. a representationceivedat the semiconcrete and more fictive as a scene is on viewing generally experienced relatively palpability -level representationthat is usually being ceived at the lessveridical than the concrete sametime. . and phosphenesbecausethese entities . say. level of With respectto generalfictivity . are themselves or relation to that grid . pressureon the eyeball. It has the quality of seemingsomewhat indefinite in its ostensivecharacteristics . though perhaps not to the fullest degree becauseof their appearanceand as one shifts one' s focus. afterimages semiconcrete . even with one' s eyesclosed. is experiencedas lesstangible and explicit . as against of Hermann for that is the fictive factive absence . the other cited semiconcreteentities can be ceived for a while without further stimulation . and one can only exercisea still lower degreeof actionability over them. . Of the semiconcreteexamplescited " above. and phosphenes . in fact. The " out there" status is still lower or more disappearance dubious for afterimages . the grayish spots of the Hermann grid may be largely experiencedas out " there. of an artificial scotoma. although each is fixed with respectto one' s visual field. by comparison with the fully concrete level. grid example representation as being .250 LeonardTalmy semiconcreteentity is the phogpheneeffect. while the factive representation of the scenebeing viewed is taken more veridically as lacking any such entities. ' Generally. translucent.which results from . to experiencedoubt about any such corporeality . or of phosphenes present only in a relatively fictive manner. as mainly treated so far. Although one has " the experienceof directly " seeing the entity . phospheneimage ranks lower on the localizabilityparameter because . . artificial scotoma. . perhaps hazy. one may be fully consciousof and direct one s central attention to such entities as Hermann grid spots. of an afterimage. an entity ceived at the semiconcretelevel of palpability . scotomas ' but one experiencesless than the fullest certainty about one s ception of them. its lessconcrete properties may largely lead one to experiencethe entity as having no real physical existenceor . and as less intense or vivid . In general . The Hermann grid spots are fully localizable move along with one' s eyemovements ceived only in with respectto the concretely ceived grid and. The ception of Hermann grid spots requiresconcurrent on-line sensorystimulation in the form of viewing the grid . The identifiability of a semiconcrete environment in pace with one' s eye movements in some . at least. once initiated . .

as well as vagueor indefinite and relatively faint . In viewing a scene " " " its " is that one does not see such an entity explicitly .9. representation that may seemso minimally discrepant with each other they are rather experienced as complementary or additive. it would be with respectto some approximate or vague category.4. Because the semiabstractlevel of palpability is perhaps the least familiar level. General fictivity works in approximately the same way for all three types: object structure. we presenta number of types and illustrations of it .) history and future has its own fictivity pattern.Fictive Motionin Languageand " Ception" 251 6. and is difficult to act on. which will be described separately Much of visually sensedstructure is similar to the structure represented by linguistic closed -class forms. but not as having autonomous physical existence . A sensed or attention . and with no quality of direct visibility . Its type of geometry is regularly topology-like and approximative.9. Unlike most forms of general of such structurality from the concrete representation the of concrete content and that of sensed structurality fictivity . One experiencesan entity of this sort as out " there.4 involving structural . This fictive presence contrasts with the factive absence . and this parallelism will be discussed later in section 6. (The type in section 6. . characterizing the pattern of general fictivity that holds for three of thesetypes. The representation of structurality one sensesin an object or an array is generally experiencedas more fictive and lessveridical than the factive representationof the concreteentities whose structurality it is. Insofar as such a sensedentity is accorded an identity . but they lack by representations this degreeof explicitnessin their original condition of ception. The representation of structurality is a case of fictive presence rather than of fictive motion . we refer to them here collectively as structurality . Often a sensedentity of the present sort is understood as a structural or relational characteristic of the concrete entities viewed. In fact. It has little or ' no ostension . seems less entity is of relatively low saliencein consciousness certain. such as line drawings or wire sculptures .9. we will adopt sensingas a technical term to refer to the implicit presence ception of an entity at the semiabstractlevel of palpability . referenceframes.4 SemiabstractLevel of Palpability An entity at the semiabstractlevel of palpability is experiencedas present inassociation with other entities that are seenat the fully concrete level. but rather senses . one s experience . and force dynamics . In order to characterize the general fictivity pattern for these three types " " together. while engagingin on-line " 12 viewing of something concrete. Such sensedstructures or relationshipscan often be captured for experiencingat the fully concretelevel schematic . perhaps localizable as a genuinely present characteristic of the concrete entities viewed.11. but it itself is intangible and nonmanifest.

object structure sensedas being of the -neutral. But in addition . one seesconcretely certain particulars of ostensionsuch as outline .9. A sensedstructural schemaof this sort can be made concretely visible. coloration and shading. an object of this sort is sensed portion and a hollow interior .as appropriate to the semiabstractlevel of palpability This pattern of ception shows .that the type of geometry (parameter 9) here sensedin the structure of an object is topological or topology-like. and closure neutral discontinuity For a more complex example . and shading. may categorizeas the inside schema wherein the first object is inside the second -object case like a number of also exhibits this As in the single . The person may sensein this object complex a structural schema " " . one seesat the fully concrete level of palpability that person' s outline and form .in . delineation. But one does not concretely seesuch a schemawhen . like a well and a trench. color . topology object array . like a beachball and a in shape . in completeness in and of bowl . A structural schemaof this sort is generally sensedin the object in a form that is abstracted away from each " " of a number of other spatial factors. More precisely terms of an idealized schematization as consisting of a plane curved in a way that definesa volume of spaceby forming a boundary around it . However. Thus not only can the first object and the second object themselves characteristics each vary in magnitude and shape . the delineations of the garments . of closure. as when a stick figure drawing or a pipe cleanersculpture is shaped to correspond to such a posture.252 Leonard Talmy 6. but in addition the first object can exhibit any orientation relative to the secondobject and can be located throughout any portion .1 Sensingof Object Structure One main type of sensed entity is the structure we senseto be present in a single object or over an array of objects due to its -object case . for example .what she above. a person may ceive one object as located at a point or points of the interior spaceof another object that she sensesas having the envelope / interior structure described . like a thimble and a volcano. like a bell jar and a birdcage. when in a squatting or leaning posture. as well as being -interior type is magnitude-neutral and shape envelope neutral. In particular . This envelope / interior structuring can thus be sensed equally acrossobjects that differ in magnitude. when one views a arrangement in space certain kind of object such as a vase or a dumpster.4.one only senses 1982 ) that representa human figure in terms of an arrangementof axesof elongation provide one theoretization of this sensedlevel of ception. one may sensein the object a structural pattern comprising an outer . degree continuity /discontinuity. A comparable sensingof structure can occur for an array of objects. textures. and so on. one does not seebut rather senses the person' s bodily structure in its current configuration . punch . To illustrate first for the single . on viewing a person. For example . The Marrian abstractions (Marr its presence looking at the person. texture.

The visually sensed" across would then also exhibit the topological property of being shape -neutral. but for the most part. one can sense a grid of compass directions amid this scenery . while still being sensed the " inside" schema .4. one might include some lines to representthe rectilinear frame of the hall . This " " visually sensed across schemawould then exhibit the topological property of being " schemain magnitude-neutral.9. when one views the interior of a restaurant. though it can be so . Perhapsone can seesome of the hall ' s framing delineations concretely. For a more intricate example . one may sensean abstract structure in this path . but represented rather only sensed as present. Further casesperhaps also belong in this object structure type of sensing . in seeingthe sceneryabout oneselfat the concrete level. The path itself would not be a caseof fictive motion . Thus. This may apply to the part of an object being occluded by another object in front of it . 6. for example . the patterned arrangement in spaceseemsto be sensed . one senses a hierarchically embeddedstructure in spacethat includes the schematicdelineations of the dining hall as the largest containing frame and the spatial pattern of tables and people situated within this frame. for the path is factive. or to the back or underside of an object not visible from a viewer' s current perspective . For example . Comparably. Thus parts of objects not concretely seenbut known or assumedto be presentin particular locations may be sensedas present at those locations. spots or circles for the tables and someshort bent lines for the people that mark their relative positions within the frame and to each other. 6. and it is " this schemathat is fictive. But in addition . However.4. together with some.3 Sensingof Reference Frames Perhapsrelated to the sensingof object/array structure is the sensing of a reference frame as present amid an array of objects.9. One may even have a choice of . some ceiling-wall edges . if one were to represent this sensedstructure of the scenein a schematic drawing. But the path is sensedas instantiating a particular idealized path schema . and in the path ofa deer running from one side of the field " schema to the other along a zigzag slanting course . one may equally sensean " across the path of a deer running in a straight perpendicular line from one boundary of a field to the opposite boundary. Thus one may senseas equal instantiations of an " across ' schemaboth the path of an ant crawling from one side of one s palm to the opposite side and the path of a deer running from one side of a field to the opposite side.2 Se18ingof Path Structure When one views an object moving with respect to other objects. one concretely sees the path it executes as having Euclidean specifics such as exact shapeand size. this is an abstraction for the most part not concretely seenas such." FictiveMotion in Language and " Ception 253 or amount of the secondobject' s interior space as manifesting .

4.4 Se18ingof Structural History and Future Another possible type of sensed phenomenon also pertains to the structure of an object or of an array of objects. state. One may even sensetoys that are lying over the floor not simply as . The examples of visual counterparts already given in that section were of a figurine perceivedas a torso with head and es like limbs affixed to it . That person can sensewithin this frame. one may not ceivethe frame as a static diamond shape . on viewing a boat leaving an island and sailing an increasing distance from it . in which the bike is to the left of the church. Levinson ( 1996 ) have performed experiments on ) and Pederson( 1993 . but may rather senseit as a square manifesting the result of having beentilted away from a more basic vertical-horizontal orientation .254 Talm Leonard }' ) . result as the rather but own in their subsistent self right an on . In effect. One sensesa set of clay shards not as an arrangement of separate distinctively shapedthree-dimensional objects but as the remains of a flowerpot that had beenbroken. or pseudohistory of activity that led to the present structure senses . Another example is the sensingof a dent in a fender not as a sui generiscurvature but as the result of a deformation. with findings of strong linguistic-cultural biasing for the particular exactly this issue . one a probable. For example that is hanging on the wall at an oblique angle. 6. this structure is sensednot as statically present but rather as having shifted into its particular configuration from someother configuration . it can be proposed that one regularly not as static configurations certain complex forms within everydayscenes senses of deviation from someprior . A sensedhistory of this sort is the visual counterpart of the fictive site arrival paths described for language in section 6. Alternatively . Or shesensethe presence of a viewer-basedframe radiating out from herself. consider a person example based an earth scene manifest bicycle at its rear.3. Or she can sensethe presenceof an object-basedframe. frame sensedas present reference of type One may also sensethe presenceof one or another alternative referenceframe for the case of a moving object executing a path . one can senseits path as a radius extending out from the island as center within the concentric circles of a radial reference frame. Here. however. equal-sided picture frame viewing generally more basic. in which the bike is west of the church. Thus. In addition to such relatively schematicentities. default. For alternative referenceframes to senseas present (as described in Talmy 1983 with a the toward at a church who is right facing looking .9. one can sensethe island as the origin point of a rectilinear ' reference frame and the boat s path as an abscissal line moving away from an ordinate. wedgeremoved. of an irregular contour perceived as the result of process with a a circle as Man aPac and of and indentation figure perceived protuberation .8. in which the bike is behind the church.

but not palpably see . and that the route-sensingprocess . of changesaway from potential probable future succession its current configuration .4. one may senseits potential return to the true ( probably as part of imagining one' s manipulations to right it ) . standing the opposite corner. In terms of generalfictivity . It applies to a viewer. rather. Here we do not refer simply to unconsciouscognitive computations that . a thrown ball sailing in a curve through the air. volitionally proceed to executethrough some region of space at one corner of a restaurant crowded with tables who wants to get to say. For example . The main difference is that there the viewed entity was itself stationary. Before starting out . Such a sensedfuture might involve the return of the entity to a basic state that it had left. It may be that the initially projected route is inadequate to the task. with respectto the picture frame example . One fonn of path projection is based on motion already being exhibited by a Figure entity. Accordingly . It is superimposed on the factively and veridically seenstatic representationof the entity . A viewer observing the concretely occurrent path of the object can generallysense . whereashere it is in motion .Fictive Motion in Languageand " C . whereas here the sensedpath segmentsare projections mostly basedon one' s naive physicsapplied to the viewed motion . One may also project backward to sensethe path that the ball is object will traverse to have traversed before it was in view. 6.5 Sel8ing of Projected Paths Another type of sensedception can be tenned projected paths.9. Thus. Another fonn of projected path pertains to the route that an agentive viewer will . there the sensedchangesbefore and after the static configuration were largely associationsbased on one' s experienceof frequent occurrence . such a viewer will often senseat the semiabstract level of palpability an approximate route curving through the midst of the tables that he could follow to reach his destination. the difference betweenthe factive and the fictive modes of ceiving the frame is the difference betweenseeinga static diamond and sensinga squarewith a past and a future. not only a history of its current configuration Viewing an entity may lead one to sense but also to sense a or . the sensingof an entity ' s structural history or future is a less veridical representation of fictive motion in a sensory modality .eption 255 comprising some specific spatial static pattern but rather as manifesting the result of having beenscatteredinto that configuration from a home location within a box. we refer to the conscious experiencea viewer often has of a compelling senseof the specific route that the . The viewer might sense the shape of this path virtually as if it were taken by an aerial photograph . Path projection of this sort is thus likely akin to the in the preceding wholly sensingof structural history and future discussed section.the path that it will subsequentlyfollow . say. on viewing the previous picture frame hanging at an angle. for example . enable the viewer to move to the spot at which she could catch the ball .

these . leaning at a 450 angle against the outer wall of a rickety wooden shed. resistanceto such opposition . Included forces such as in such sensed force dynamicsare the interactions of opposing 's ' s intrinsic another rest or motion an object . This sensedforce structure might structure implicit throughout theseovert elements include a force (manifestedby the shed fully but tenuously resisting ) that is now success an unrelenting outside force impinging on it (manifestedby the slab). disappearance .) To illustrate . whereasthe . but also would sensea force dynamic . object resting on a horizontal surface a scenein which a large concrete slab is consider For a less schematicexample .e. Rubin ( 1986 ) report that subjectsperceive ) and Engel and Rubin ( 1986 ) forces at the cusps when viewing a dot that moves along (in our terms. general fictivity . 6.4. block (i.9. only the physical surroundings are seenconcretely. A person viewing this scenewould probably not only seeat the concrete level the slab and the shedin their particular geometric relationship. just .256 Leonard Talmy is regularly updated and reprojectedas the viewer movesalong his path . Complementarily. appearance presence an analysisof the semanticcomponent of languagethat pertains to force dynamics. the force is perceived as being dissipated. When the bounce is progressively heightened . when the original configuration is rotated 90 degrees ) . to a stuck attaching of an perception object viewing force" when the samesmall block is similarly positioned on the top face of the larger . analogously to what is sensed " an of such is no But there wall. or absenceof blockage. sense . But throughout such a process . perceived only .6 Se18ingof Force Dynamics Also at the semiabstractlevel of palpability is the sensingof force interrelationships among otherwise concretely seenobjects.7 SeI Wingof Visual Analoguesto Fictive Motion in Language Finally .. This form of projected path is akin to the fictive access is sensed paths follow to path 6. then a path like a bouncing ball.8. when the perception is that a force has been added at the cusps the ball ' s bounce is reduced . Jepson and Richards ( 1993 ) also note that when a block is drawn with one face coplanar to and in the middle of the vertical face of a larger block .4. then the percept is as if the smaller " in the block is " attached or glued to the larger block . the fictive motion types presentedbefore this section on ception can now be recalled for . and the this tendency . expected . Further of palpability that is.4. (SeeTalmy 1988bfor . they are sensed . Most of the visual patterns suggestedas their relevanceto the present discussion abstract level motion fictive of the types seemto fit at the semi linguistic counterparts in of terms . and that is capable of incrementally eroding and giving way at any moment. describedin section 6. the overcoming of resistance . object opposition to tendency toward . In this latter case an in be would as is attachment not contact viewing .9.

5 Abstract Level of Palpability The casescited thus far for the first three levels of palpability have all dependedon . to as fictive motion . alignment paths. such entities are thus largely impalpable. one can range from certainty to ' to from a and capacity manipulate them in one s mind to an puzzlementover them. radiation paths." Fictive Motion in Languageand Ception " 257 visual analogueshave involved the sensingof fictive motion. pattern paths.9. they do not involve the " " sensingof fictive presence(as was the casefor the representationsof structurality ) . As a summary. and. accesspaths. At this level. Thus. Finally . They do seemto exhibit a range acrossthe remaining . this is a complete list of the fictive types proposed. frame-relative motion . . .2 6. one experiences conceptual or affective entities that do not require on line sensory stimulation for their occurrence and may have little direct relation to any such stimulation. experience knowledge representation relationships among concepts . Largely clustering near the lower endsof the remaining palpabilityrelatedparamete . Such phenomenamight include the following : the awareness 's the within one . they can exhibit either content or structure. abstract. artificial concurrent on-line sensorystimulation (with the exception that afterimages scotomas ) . can exhibit a range. and phosphenesrequire stimulation shortly beforehand a level still further down the palpability gradient. demonstrative paths. experienceof being only a passivereceptor to them. But we can adduce . though perhaps tending toward the approximative and qualitative type. vague in to localization not amenable and characteristics in ostensive faint . the experience entities normally ceived as a result of on-line stimulation . They might include not only the imagined counterparts of . sensory paths. and . they can range from full salienceto elusiveness palpability -related parameters or virtual inaccessibility to consciousness . Such abstract entities may be ceived as components in the course of general ongoing thought and feeling. this. With the addition of the casesof structural history/ future and projected paths characterized just above. the (fully ) abstract level. we may senseat the semiabstract 6. experienced subjective spaceor identification as to category They " " in oneselfrather than out there. Thus. insofar as they manifest a type of geometry. and targeting paths). and coverage paths.for example -line while on sense otherwise would one the structure of in imagination only . have a visual representationsensed 6.8.but also phenomenathat cannot normally or viewing an object or array in space ever be directly ascribedas intrinsic attributes to entities ceivedas the result of on-line of sensorystimulation . shadow paths. in this chapter.5 and just seen . advent paths. we can list here the fictive types from sections6. too . all of which participate in this phenomenon level of palpability the fictive motion of the visual counterparts of orientation paths (including prospect paths. lacking perhaps hence as often are .

while closed . at least insofar as it is linguistically associated First . Despite such upscaling lent . We can next identify a number of conceptual categories expressedby linguistic closed -class forms that are seemingly never directly produced by on-line sensory stimulation .concepts representmore structural . purely ideational level in the absenceof -classlinguistic form current sensory stimulation by hearing or thinking of a closed on For that refers to the sameschematicstructure. and of courselessconcretewhen imagined in thought . these to consciousness cognitive manipulability (actionability ) and access are phenomenathat . Thus the conceptual category of tense . we next cite a number . it is easiestto give further examplesof ceptually abstract by linguistic representation the meanings of certain linguistic forms.258 Leonard Talmy of implications betweensetsof concepts .13 end of the palpability gradient. A secondlinguistically represented category can be termed reality status. and the formation of inferences of change occurring over the long term. For any event being referred to . one might sensethe presenceof a structural across " " .and hence -classmeaningsso as to further convey the character of the fully abstract of closed . but the degreeof concreteness they do have tends to lend a measureof explicitnessto the conceptual and affectivephenomena associatedwith them. " " a wide range of affective states . . However. assessments . Many cognitive entities at the abstract level of palpability are the semantic referents of linguistic forms and thus can also be evoked in awarenessby hearing or thinking of those forms. These forms themselvesare fully concrete when heard. and propositional attitudes (such as wish and intention ) . assessments social influence (such as permissionsand requirements . when experienced directly without associationwith such linguistic forms. may be at the fully abstract level of palpability .can also be ceived at the fully abstract. But one can also ceivethe same across schemaat the abstract schemain that scene level of palpability by hearing or thinking of the word across either alone or in a sentencelike The log lay acrossthe road. a schematicstructure one might otherwise senseat the semiabstractlevel of palpability through on-line sensorystimulation . more abstract. and future . expectationsand pressures ). This category is well representedin the event relative to the presenttime of speaking languagesof the world but has seemingly scant homology in the forms of ception higher on the palpability scale that are evoked by current sensory stimulation .a type largely included under the traditional linguistic term mood. And with such greater explicitness may come greater . . pertains to the time of occurrence of a referent . experiencesof of veridicality .as by looking at an object or scene . viewing a scenein which example " " a log is straddling a road. with such specific member concepts as past. present. Becauseopen-class phenomena by citing -classforms tend to forms tend to representmore contentful concepts .

in Wintu. Finally. . andwouldalsoincludethesimplenegative e. accepted " " inferred " shared throughgenerally from . Smithin hers . Again. In particular . a linguisticcategorythat can be termedparticularity to whether an entity in reference pertains is to beunderstood asunique( Thatbirdjust flew in). or counterfactual .g. and "judged as probable reported . conditional .." Another linguisticcategoryof the cognitive 's ' s inference statetypecan be termedtheaddressee status . Similarly . purpose . the Englishdefiniteand indefinitearticles theanda or an. . In a numberof knowledge languages (e. asin Johns relationships 's book motherand John . too. or asa particularoneout of a setof comparable entities(A birdjustjiew in). particularizes 's thespeaker of theeventthat sheis referringto.14 . counterpart To continuethe exemplification . wish . it apparentlytendsto . may not directly evokethe relationalconcepts of kinship and possession the linguistic formsdo. . these . suchas kinship and relationships ' . Thus the ceptionof genericness in human cognitionmayoccuronly at theabstractlevelof palpability . a further setof categories at the abstractlevelof -classformspertainto the palpabilitythat can beevokedby closed cognitivestateof some sentient entity. or viewingJohnin thedoorwayanda book on the table . and regret For somefinal examples . " inferredfrom " " entertained temporalregularity aspossible . the linguisticallyrepresented of modality category . heard . potential . seem categories at the higherlevels of unrepresented 's . desire . aspects of situations that arecurrentlyseen . with suchmember notionsasthose expressed by English can must and should . manylinguistic -classforms specifya variety of abstract closed . ( g English not). Onecommonlinguisticform representing specifying this category is that of determiners . smelled .for example exclude the generic case . because of havingbeen " . Furthergrammatically areintentionand represented cognitivestates volition.Fictive Motion in Languageand " Ception" 259 this category would includesuchindications as that the eventis actual . suchas viewingJohn in his house andMrs. Thusa conceptual palpability that can be termedspeaker category status knowledge " . on-line ception . This is the speaker knowledge ' as to the addressee s ability to identify somereferentthe speaker is currently . feathers ) But the on-line ceptionof an entity at the concrete or semiabstract level this rangeof options may not accommodate . knowledge evidence accompanying . looking at a particularbird doesnot tendto evokethe ceptionof all birds generically . represented by linguisticforms called" evidentials the statusof . or sensed . The Englishendings express possession esboth of these . "" member notionsas : " known from personal as factual experience asfactual . haslittle concrete . or generically as an exemplarstandingin for all comparable entities(A bird has . .for example that mark definiteness . whereit is expressed inflections on theverb thiscategory by hassuch ). and soon at the concretelevel or sensed at the semiabstract level are seemingly not ceivedas havingany reality statusother than the actual . Again.

In fact. one can imagine the sight and sound of it starting to bark . whether related to vision or to other sensory modalities. as in the case of many . mental imagery. on seeinga dog. it can be evoked in association with an entity ceived at the concrete level during on-line stimulation by that entity .10 FurtherTypesandProperties D of Ceptio Leonard Talmy The full structure of the entire system of ception certainly remains to becharacterized. (2) actions one might undertake in relation to the entity. Or . As examplesof associatedaction (2). asjust discussed . Along the gradient parameter of stimulus dependence . entity . there be a of to what we have for may gradient palpability parallel posited ception that to motor control . one might experienceor executethe gyrations of body English as if to effect ' a correction in the ball s path .10. that is. Such associatedphenomenacould include the following type: ( I ) mental imagery. We 6. (4) particular conceptsor aspectsof one s knowledge one associates and (5) inferencesregarding the entity . we can here illustrate the remaining four Having already discussed of thesetypes of associativeception. For example .2 AssociativeFOrlDS of CeptioD What can be tenDedassociativeforms of ception pertain to ceptual phenomena evoked in associationwith an entity during on-line sensorystimulation by it . (3) affective statesexperiencedwith respectto the ' with the entity. imagistic ception seemsto fall in the midrange. on viewing a bowling ball inexorably heading for the side " " gutter.1 Imagistic Fonns of Ception What can be termed imagistic forms of ception include mental imagery. conceptual categoriesof language 6. But imagistic ception can also occur without on-line stimulation . as during one' s private imaginings. That is. one might experiencea motoric impulse to manipulate the frame so as to right it . from the least to the most applies Proceeding palpable. whether aspects of it are unrelated to sensory attributes.260 6. in the midrange would be one' s . on viewing a tilted picture frame. with respectto such kinesthetic effects . It needs to be determined whether imagistic ception can also occur at the low end of the stimulus dependence parameter. but some brief notes here will sketch in a few lineaments of that structure.10. at the low ' end would be one s experienceof intending to move. but not ascribed to that entity as intrinsic attributes of it . as well as the sight and kinesthesiaof one' s walking over and petting it .

4 Diaociatio18 amongthe Palpability-Related Parameters While the thirteen palpability -related parameterslisted in section 6. Smith is John' s mother from the visual apparencyof their agesand of their resemblance . where a particular phenomenon is placed along the in trinsicality gradient varies according to the type of phenomenon . the phenomenadescribedin section " at the 6. or ' inferring that a book on a table belongs to John from the surroundings and John s manner of behaving toward it . they would be ceived as actually present and perhaps inherent attributes. This intrinsicalityparameter . But at the lower end of the intrinsicality gradient. some .1 generally tend to correlate with one another for the types of ception that had beenconsidered . that is. that is. the sensedphenomenawould be experienced as intrinsic to the entity being ceivedat the concretelevel.that the ceiver is " detecting" in the concretely seenentity .9. of roadkill .4 as " sensed semiabstractlevel and the associativephenomenareported here may belong together as a single group ceived at the semiabstract level of palpability .Fictive Motion in Languageand " Ception" 261 experienceof all -but-overt motion . they would be experiencedas incidental phenomenathe ceiver brings to the entity . much like the person s height. is actually just the objectivity gradient (parameter 5) when applied to phenomenaconnected with an entity rather than to the entity itself. and at the high end would be one s experienceof one s overt movements Associated affect (3) has such straightforward examplesas experiencingpleasure . or of a mugger. English. what might be termed intrinsica/ity .3 Parameterof Intriaicality Associative forms of ception like thosejust adducedmay be largely judged to cluster near the semiabstractlevel of palpability .10.g. e. and the occasion . 6. or . one may experiencethis beauty as an intrinsic ' attribute of the person seen . the culture.9. disgust. And examplesof associatedinference(5) might be gathering that Mrs . including checked movement and covert body ' ' . or fear at the sight of something.10.. of a child playing. if one ceivesbeauty in conjunction with seeinga particular person. the associative phenomenapresentedhere would be experiencedas merely associatedwith the concretely ceivedentity . alternatively. For a classicalexample . the individual . At the high end of this gradient. or thinking of one' s childhood home on smelling fresh bread. To be sure. as a personal interpretive responseby the beholder. however. But the sensedtype and the associativetype within this group would still differ from each other with respect to another gradient parameter.such as structure and patterns of force impingement. In fact. 6. Associated knowledge or concepts(4) could include exampleslike thinking of danger on seeinga knife .

which can exhibit all levels along the palpability gradient except perhaps the most abstract. But the conscious manipulability of the linguistic forms expressingtheseconceptual phenomena ranks them near the high end of the actionability gradient (parameter 12 ) . Comparison of the sensorymodalities with respectto ception requires much further investigation. as engagedin the ingestion of food . already -classlinguistic forms are generally noted.10. seemsto rank low with respect to the localizability gradient (parameter 6) . The observation of further dissociationsof this sort of the parametersadducedand ultimately justify their can argue for the independence . At the same time.5 ModaHty Differencesalong the Palpability Gradient In the discussion on ception. level and the semiabstract however. the cognitive phenomenaexpressed by closed at the most abstract level of the palpability gradient (parameter 1) . murky on the clarity example gradient (parameter 2). at least for humans. Or the low on most of states rank some affective parameters for may quite again. visionand language .262 Leonard Talmy . may by its nature seldom or never rank very high along the palpability . For example dissociations can be observed . 6. and ostensiongradient (parameters I . For example of one' s current body posture and movements . and 4). 4).11 Content / Structure Parallelisms between Vision and Language The analysis to this point permits the observation of two further between . The modality of smell. . having relatively definite form and movement. But we can briefly note that each sensory modality may have its own pattern of manifestation along the various palpability -related ' . with respect to the imagistic forms of a can have visual mental fairly high degreeof ostension(parameter imagery ception. we have mostly dealt with phenomena related to the visual modality . 2. including one s sense parameters adduced. and nonostensiveon the ostension gradient (parameter 4) while ranking quite high on the strength gradient (parameter 3) becausethey are experiencedas intenseand vivid . it may rank somewherebetweenthe semiconcrete level along the palpability gradient (parameter I ) and at a comparably midrange level along the clarity gradient (parameter 2) . the kinesthetic modality . For another case of dissociation. intangible on the palpability gradient (parameter 1). perhaps hovering somewhere between the semiconcrete and the semiabstract level. may range more over the content region than over the structure region of the content/ structure gradient (parameter 8) . identification as distinct phenomena 6. And the modalities of taste and smell. for instance . clarity .

1 ComplementaryFunctionsof the Content and Structure Subsystems and Language First . while the structure subsystemis foremost at the semiabstract level of . so as not to bump into the tables' corners. have a content subsystemand a . the referents of open-class forms largely manifest the content palpability . . to revisit an earlier example dining area of a restaurant will likely plot an approximate. without an overarching schematic map for guidance. . through the tables. a person operating without the overall topol ogy-like subsystemwould be reduced to inching along. . Within on . subsystem . The two subsystemsthus perform complementary functions and are both necessaryfor optimal navigation. as will be demonstrated next. The two subsystemsserve largely distinct and manifesting the structure subsystem complementary functions. To do this. the content subsystemis relevant for fine-grained local calibrations path through space while the structure . each of which specifiesa rich complex of conceptual content. the person will attend to the Euclidean particulars of the tables. . using the guidelines of the precision subsystemto follow the sides of the tables and the curves of the chairs. For a caseinvolving motor planning and control . a person lacking the precision subsystemmight set forth on an approximate journey but encounter repeated bumps and blockages for not being able to gauge accurately and negotiate the local particulars. both cognitive systems . as in executing a particular . as well as other forms of motor activity .11. . These are the verb rustle. vision and language line vision for structure subsystem . in the viewing of an object example or array of objects. On the other hand." and " Ception FictiveMotion in Language 263 in Vision 6. absoluteness and a substantive as . . consider the sentence forms in any single sentence This sentencecontains just three open-class forms. we can at work in language We can next illustrate the two subsystems -class observethe distinct functions servedby the open-classforms and by the closed A rustler lassoedthe steers . a person wanting to cross the approximation . subsystemcan project an overall rough-and-ready first . qualitative coursecurving semiabstractlevel of structure in a spatial array . In language -class forms are generally limited to while the referents of closed . Euclidean geometry as against topology. first for vision and then for . holistically against relativity precision against approximation summary as against a unifying frameworkS We can first illustrate the properties and operations of the two subsystemsin vision. If such were possible . the content subsystem is foremost at the concrete level of palpability . Thus. which specifiesnotions of . A number of properties from both the content/ structure gradient (parameter language 8) and the type-of -geometry gradient (parameter 9) align differentially with the . Included are properties pertaining to distinctive functioning of thesetwo subsystems as bulk as against lineaments . Thus. using the concretelevel of specificbulk content. using the sensed But in the processof crossing.

specifying ' unitary instantiation . which specifiesa illegality . specifying the speakers assumption of ready . we are still in a Western cowboy landscape . These include the function a specify relatively spare concepts serving structuring event of the current suffix ed specifying occurrencebefore the time . the article the. while keeping the other alternately changing one type of form in the above sentence -classforms. the two cognitive subsystems and complementary functions as they are in operation. First . for human consumption notions of a particular animal type. the grammatical category of noun ) indicating ). as in a sentencelike Will constant. we are now transposed original sentence to an office building . specifying multiple instantiation .264 Leonard Talmy . both in ceiving and motorically negotiating a visual sceneand in cognizing of content and of structure the referenceof a sentence . the identifiability for the addressee suffixer . but with the content-specifying forms altered.4 on ception at the semiabstractlevel of palpability proposed that we can sensethe spatial and force-related structure of an object or an array of that any structure of this sort is sensedas objects when viewing it . specifying the opposite of this. the precedingsection has shown that the type of geometry. and that of direct object. in the referential and discoursecontext of a contribute the majority of the content. and castration. specifying the performer of an action. all the structural delineations of the depicted scene and of the speechevent have been altered. then. but becausethe content-specifying openclassforms . sentence . the verb lasso that is in knotted a and swung around. recall that section 6. and livestock. the open-classfonDSof the sentence -classforms determine the majority of the structure. indicating an Agent. the institution of breeding . whereasthe closed Thus. and the grammatical relation of subject.9. and particular configuration rope looped ' steer. the change only the open class forms. and the article a. the suffixs speech " " .2 ComparableCharacter of the Structure Subsystem The structural subsystemsin vision and in language exhibit great similarity . In sum. The distinct functions servedby thesetwo types of fonDScan be put into relief by . indicating a Patient.11. But we can now are the same . which specifies the noun and in a certain s head an animal circled over way. With respectto language . the sentencecontains a number of closed . It was suggested consisting of an idealized abstractedschemawith a topology-like or other qualitative . performing equally necessary interact with each other. Thus we can changeonly the closed the lassoersrustle a steer? Here. indicating an object and that of verb (for lassoed (for rustler and steers a process . -class fonDS that On the other hand. Here. theft . in Vision and in Language 6. as in A machinestampedthe envelopes structural relationships of the sceneand of the speechevent are the same as in the . and the zero suffix (on rustler). property ownership. cast.

9. The structure subsystems of vision and languageexhibit a further parallel.1 that one can visually sensethe presenceof an " inside" type of structural schema on viewing a two -object complex in which one object is sensed as located at a point or points of the interior space defined by the other object. consists of idealized abstracted schemaswith topology -like properties. This schemacan be topologically or qualitatively abstracted away from particulars of the objects' size . Talmy 1975 . too .9. discontinuity . With respectto the structure of an array of objects. the extensivebody of linguistic work on spatial schemas(e. the structural schemas -class forms. then we can supposethat both visual sensingand linguistic closed -classrepresentation are connected with . 1983and Herskovits 1986 . In fact.whether sensed all thesecasesof abstracted or conceptually imposed schemas visually or specified by linguistic closed -class forms can be understood as a form of . Accordingly. relative to the products of other neural systemsfor processingthe concrete ostensionsof ceived entities.here.4 that the structural schemas one semiabstractlysenses to be presentin an object or array are assessed as being fictive. They constitute not fictive motion but fictive presence presenceof structure. relative orientation . those pertaining to expressedby linguistic closed . the character of the structuring yielded by visual sensingand that yielded by the linguistic closed -class system appear to be highly similar. If we can heuristically hypothesize that some particular neural systemis responsiblefor) ' rocessingschematicstructure in general . In particular. That is.here. we consider severalparallel vision. 1994 ." Ifwe can now extend the hypothesisof a neural systemresponsiblefor processingschematicstructure . and relative location. it . we can add that the products of its processinghave ascribedto them the character of being fictive. Herskovits has made it a cornerstone of her work to treat the spatial schemasshe describes as " virtual " which are to be distinguished structures" (previously called " geometric conceptualizations ). . among much else ) constitutes a major contribution to fictivity theory . Proceedingnow to demonstrationsof similarity . it was proposed languagecases in section 6. " " '' from the canonic representations of objects as they are. the fictive fictivity .4." Fictive Motionin Language and" Ception 265 -classforms is dedicated to specifying the structure of the whole or systemof closed some part of a conceptual complex in reference . state of closure.g. relative to the factive status of the way one concretely sees the object or array . shape .. Now . the spatial schema specified by the English preposition in -classform can thus be usedwith equal exhibits all thesesameproperties. or " tap into . We can now point out that when such linguistically specified structure pertains to space or force. Recall the observation in section 6. This closed . Now ." that single neural system for this common characteristic of their mode of functioning . relative to the factive character of spaceand force are also fictive representations the objects and arrays that a languageuser understandsthem to pertain to. specifically.

4. on viewing certain scenes . in a well. Further in a trench. It has a component for representing in an form the kinds of schematic structures explicit generally sensedonly implicitly at the semi abstract level of palpability . Thus English awayfrom indicates motion from a representedby closed -type axis within a point on an ordinate-type boundary progressingalong an abscissa rectilinear grid . as in The bus drove across the country. But out from indicates motion from a central point along a radius within a radial grid of concentric circles. to be present For a final case that ." pictorial representation .266 Leonard Talmy .9. . topology-like . in a volcano in a . the topology -like properties of the structure Comparably. Here. appropriatenessto refer to someobject as located in a thimble. this schema is shape along neutral. it can belllar . whether this sceneis viewed visually or referred to linguistically . In a related way. apunchbowl.3 suggested . 6. both referenceframes are clearly fictive cognitive impositions upon the scene . as to a jagged path. in a beachball in in a or . the ) a schemainvolving motion preposition through specifies(in one sector of its usage a But a line located within medium. And . We here call this the component for " schematic . But thesetwo alternate schemas -classforms. But this type of visually sensedstructure also has linguistic closed -classparallels.2 addressed in the of a viewed sensed path moving object. the structure subsystems vision and languageproduce a fictive representation .4. as well as to paths of thousands of miles. This is evident from the fact that it can be applied both to paths of a few centimeters . visually linguistically imputed to a again topological are relative to the Euclidean fictive path representations particulars seenor believed . Thus the English preposition across which specifiesa schemaprototypically involving motion from one parallel line to another along a perpendicular line betweenthem. relative to the concreta of the object array . section 6.9. as in Theant crawledacrossmy palm. as in I zig-zaggedthrough the woods through the woods in the schemas thus sensed or . one may sense the presence of either a rectilinear or a radial referenceframe as the background can also be against which an object executesa path . birdcage of both be said that in abstracting or imposing their schema .exhibits the topological property of being magnitude-neutral. section 6. These alternative conceptual schematiza in like: The boat and out tions can be seen sentences drifted further further away/ from the isle. . thus through can be applied equally as well to a looped path.3 Stnlctural Explicitnessin Vision and Language The cognitive system pertaining to vision in humans has another feature that may have a partial counterpart in language .11. . as in I circled . or Thesloth crawled 10feet away/outfrom the tree trunk along a branch.

however. explicitnessis the closed The linguistic linkage of overt morphemes to the structural schemasthey represent lends some concretenessto those cognitive entities. a full -blown pictorial depiction manifests the content . permit increasedactionability upon them. and the like. and planes .are of this schematickind . The form of such morphemes . in depicting an object or sceneviewed. The component of languagethat may partially correspond to this representational -classsystemitself. . textures. Marr 1982 at ). and later as a circle atop a vertical line from which two lines extend laterally right and left at a midpoint and two more lines slope downward from the bottom point . this languagecomponent differs they represent . more so than to the cognitive systemsfor concretely ceiving the full ostensionof objects. and this way. In fact. much cognitive processinghas to occur between the responsesof the retinas " and these hand motions. both of single objects and of object arrays (cf. Thesemorphemesconstitute tangible counterparts to the abstract forms. The very first pictorial depictions children drawings. It then appearsthat the component of the visual systeminvolved in producing external depictions taps specifically into this sameabstractional . Thus. lines. shadings . in the developmentally earliest systemfor schematicstructure in general ' phaseof operation. For example .a child might draw a human figure at an early phaseas a circle with four lines radiating from it . What impinges on one' s retinas are the particularities of ostension : the bulk . a child representsnot so much its -level characteristics as the structure that he or she can sensein it at the concrete semiabstractlevel of palpability . Accordingly . they constitute a major part of what is sensed the semiabstract level of palpability . wire sculptures " " produce.their stick figure drawings. otherwise located at the fully abstract level of palpability . and perhapsafford greater consciousaccess to them. such structural abstractions are in any casenecessary for the ception of visual form . colorings. but rather one-dimensional lines forming a structural schematic delineation. It must be emphasizedthat such schematizationsare not what impinges on one' s retinas. a child s iconographic capacity would appear to be linked mainly to this structuring mechanism . bulk into delineations. a mechanism already in place for other functionsstructuring system where this mechanismmay be the sameas the earlier heuristically hypothesizedneural . as in both static and filmic cartoons and caricatures . Yet what edges ' emergesfrom the child s hand movementsare not such particulars of ostension . This processingin a principled fashion reduces . as characterizedin the precedingsection. and so on of an entity looked at." Fictive Motionin Language and" Ception 267 In iconographic representation . or boils " down. As proposed in this chapter. line . does not reflect the form of the schemas in . But the structure subsystem can be made explicit through the component subsystem of schematicpictorial representation by schematicdepictions involving the use of points.

respectively . 1988a ) and here. the degreeand types of variation such structural featurescan exhibit acrossdifferent cultural groups. first . systems . The representationof an entity within supply the two discrepant representations the target domain is understood as factive and more veridical. it remains to chart their differences in vision and languagediffer in various respectsas to what they treat subsystems as structural . the fact that vision and language to consciouscontrol .11.can generally render the structural representationsof the structure subsystemexplicit suggests that theserepresentationswere not in access ibly in the first implicit place. 6. the where system would seemto comprise only a part of the structure subsystembecausethe former pertains to the structural representationof an extendedobject array . It may be expectedthat the structure . . They posit explicit and implicit forms of visual perception of objects' apparently the concepts in the literature closest to this chapter s concepts of the concreteand semiabstractlevelsof palpability . Separatecognitive systemsfor representingobjects and spaceshave been posited ' by Nadel and O Keefe ( 1978 ). in particular as expoundedby Lakoffand Johnson ( 1980 ). . rather than being wholly inaccessibleto .4) can in fact be experiencedin awareness least at a vague or faint degreeof clarity . Although this section has pointed to content-structure parallelisms betweenvision and language .9.both largely amenable consciousness .whereasthe latter respectto which the location of a figure object is characterized also includes the structural representationof any single object.4 SomeCompariso . second . 6. these systemsfit well. And . that entities such as structural representations sensedat the semiabstract level of palpability at (like those treated in section 6. and by Landau " what" and the " where" and Jackendoff ( 1993 who characterized them as the ). The representation from the sourcedomain that is mapped onto the entity in the target domain. their degreeand type of geometric abstraction.the field with . is understood as fictive and lessveridical. We would claim instead. But they claim that their implicit form of perception is inaccessibleto consciousness . accords readily with general fictivity . However. into the content and structure " " subsystems posited in Talmy ( 1978a .268 Leonard Talmy crucially from the pictorial schematicrepresentations . by Ungerleider and Mishkin ( 1982 ). on the other hand. which do correspond in structure to what they represent . The source domain and the target domain of a metaphor . To be sure. with Other Approach es The present analysis raises a challenge to the conclusions of Cooper and Schacter " " " " ( 1992 ) .12 Relation of Metaphor to Fictivity Metaphor theory. and the times and sequences in which thesestructural featuresappear in the developingchild .

factively. the subject will . " Love is a " " " " " journey . and may experience a . Argument is war. One observation arising from the fictivity perspective . But at present . (27) Fictive: X is Y Factive: X is not Y Thus. is that any of the Lakoff and Johnson' s ( 1980 ) three term formulas. This can be seen in sentenceslike The ordeal still " lies aheadof us. they might characterize the perception of the C figure as involving the mapping of a sourcedomain of continuity onto a target domain of discontinuity. love is a . while in some fictive expressions . while the dynamic " " " . factive temporality is here expressedliterally in terms of fictive spatiality . both psychologistsand linguists balk at the notion of closure as a metaphor. if its terms were to be extended to cover vision.its fictive representation . she will sensea complete circle at the semiabstract level of palpability . She will experiencethe former representation as more veridical and the latter one as less so. then. and have somewherein their cognition an understanding of the discrepancybetweenthesetwo representations . An extension of this sort should indeed be assayed . As for the framework of linguistic metaphor. the outline of a might generalframework for addressingsuch phenomenaacrosscognitive systemsis here in place. In terms of general fictivity . Meanwhile. Concurrently likely seea C at the concrete level of palpability . Consider. Seeingis touching . a subject viewing a round and narrow-gapped C-like figure. One reason for choosing to adopt fictivity theory over metaphor theory as an umbrella aegis is that it is constructed to encompasscognitive systemsin general rather than just to apply to language . love is not a journey . and Christmas is coming . perhaps not noted before. This .for example . linguistic expressions onto time as a target domain. The characteristic that renders an expressionmetaphoric what metajourney very phoricity dependson is that speakersor hearershave somewherewithin their cognition a belief about the target domain contrary to their cognitive representation of what is being stated about it . ." Fictive Motionin Language and" Ception 269 For example often involve spaceas a sourcedomain mapped . where the static spatial relation of frontality " is " " mapped onto the temporal relation of subsequence .is actually a cover term for a pair of complementary formulas. as representedin (27) . one of them factive and the other fictive. for example . so that the subject experiencesa visual metaphor of continuity ." In terms spatial relation of approach is mapped onto temporal succession of general fictivity . is the way that degreeof discrepancy between the two representations the framework of general fictivity would characterize the Gestalt phenomenon of closure.its factive representation for the same figure.

The factive fictively as motion even more than factively as stationariness representation of a stationary referent directly as stationary is what Talmy ( 1988a ) calls the " synoptic perspectivalmode" . can be understood as a form of cognitive " staticism. it can be observed that .270 6. This sort of mapping. which explicitly refers to my . progressiveforward motion ." Given this framework .notably . linguistic exhibits a strong bias toward conceptualdynamism as against staticism. the physical material is factively moving. stationary phenomenaconsideredby themselves can in somecases be represented . this volume) calls the survey form of representation ." . We can now compare the relative occurrence of fictive motion and fictive stationariness in language and. In the examples given. in language . . fictive motion in languagecan be interpreted as the mapping of motion as a source domain onto stationarinessas a target domain. its fictive representation in terms of motion exemplifies Talmy ' s " sequential perspectival mode. when the scenery is .8. Correspondingly. Visual counterparts of fictive stationarinesscan be found in viewing such phenomena as a waterfall or the static pattern of ripples at a particular location along a flowing stream. expression The cognitive bias toward dynamism in languageshowsup not only in the fact that . Here one ceivesa relatively constant configuration while all the physical material that constitutes the configuration constantly changes . the observer is fictively treated as fictively treated as moving toward the observer In . is the reverse ping of stationarinessas a sourcedomain onto motion as a target domain. linguistic expressions manifesting fictive motion far outnumber onesmanifesting fictive stationariness . fictive motion occurs preponderantly more than fictive stationariness . fictive stationarinesshas already been seen in frame-relative motion . in a related way. This is illustrated in (28a) .1. In stationary phenomenaare fictively representedas motion more than the reverse addition . then. perhaps also. phenomenaother than motion . A mapping of this sort can be seenas a form of cognitive " dynamism." Fictive stationariness : the map. stationariness can have fictive status in both languageand vision. In terms of metaphor theory. in vision. addition certain formulations treat motion as if it were static. There the physical substance was for the most part factively stationary. I can say My path wasa circle with the tree at its center which confines the fact of motion to the noun path and presentsthe remainder of the event as a static configuration . stationary linguistic For example. while the fictive pattern that it formed moved. That is. This situation is the reverseof the pattern paths of section 6.13 Cognitive BiastowardDynamism Leonard Talmy As we have noted above. In other words. while the fictive pattern that it forms is stationary. instead of saying J went around the tree. that is. in turn . it is what Linde and Labov " " " " ( 1975 ) call a map and what Tversky (chapter 12.

Thus an individual who suddenly ceives all the components of a conceptual domain as concurrently copresentin a static pattern of interrelationships is said to have an " aha" experience ." Fictive Motion in Languageand" Ception 271 and. AnnetteHerskovits . and for muchvaluable . For . Notes 1. if an equilateral trianglemoves along . configuration motionasa basis for ascription of " front" status . Kean Kaufmann . other examples virtually a static preclude . while an individual who ceivesa successionof one consequentevent after another through time as a simultaneous static pattern of relationships is sometimesthought to have had a visionary experience . on viewing a picture hanging on a wall at an angle. AckD Dts Owledgme I am gratefulto Lynn Cooper Palmer . one usually progress through a proof step by step rather than seeingthe full complement of logical relationships all at once. a person may example more readily ceivethe picture as a squarethat has beentilted out of true and calls for righting . which unfortunately included in thepresent version of this chapterfor lack of space . Bucher andPalmer canprevailover ( 1985 ) haveshownthat. The wells get deeperthe further down the road they are. And my thanksto KarenEmmorey discussion for corroborative Mary Peterson data on fictive motion in AmericanSignLanguage could not be . Stephen . permit ting only a representation in terms of fictive representation motion for colloquial usage . comparably. representation (28) a. what both Linde and Labov and Tversky call the " tour " form of . This chapteris plannedas the first installmenton a more extensive treatmentof all the fictivecategories . There are some housesin the valley. There is a houseevery now and then through the valley. While this example allows both modes of representation . ' rrhe wells depths form a gradient that correlates with their locations on the road. as illustrated in (28b) . Thus . b. whereashe may require a special effort to ceive the picture statically as a diamond. Comparably. In a similar way. whenin conflict . 2. In fact. as seenin (29). in vision. in the cognitive systemof reasoning es . b. ' (29) a. cognitive dynamism is so much more the normal mode that the cognizing of staticism is often regardedas a specialand valued achievement . factively static phenomenain cognitive systemsother than language may also be more readily cognizedin fictively dynamic terms than in static terms.

Whether the ' triangle s vertex leadsalong the line of motion or trails . the triangle is simply seenas moving backward. 1988a . / looks out toward the butte. Such distinction . for example its head pointing past the light . this construction cannot refer to nonaligned fictive .. * The snakeis lying past the light cannot refer to a snake lying straight with paths. corresponding to our coveragepaths. Comparisons of language ( 1978 structure to the structure of the reasoning systemappear in Talmy ( 1988a ). 5. corresponding paths points / The houseis over the hill ). the line is still seenas the front ..g. 6. 9.. though perhaps without much recognition of conferring that advantage . and this chapter) and in Jackendoff ( 1987 ) . Note that the notion of crossing behind a front -bearing object may be partially acceptable . And the most in 1995a this and to the attentional 1995 and . Talmy ( extensiveidentification and analysis to date of the foundational structural properties common " to all the cognitive systemsappears in the " Parameters section of Talmy . as well as the parametersnext posited to extend through it . then that line is seen as defining the front -back. . the with reference to a putative cognitive subsystemunderlying the analysis is presentedprimarily structure of narrative. actually extendsfully along the front -back axis of the object. Already in common usage are other terms that are neutral to any perception. The cliff wall faces b. Colllparisons of language structure to the structure in visual perception appear in Talmy . ). is why this construction " cannot also be used for aligned arrangementswith path geometriesother than " toward or " " * away from . The term and perhaps the basic concept of ception derive from a short unpublished paper " " by StephenPalmer and Eleanor Rosch titled Ception : Per. " and end location town to our demonstrative to toward the . The bedroom window faces / looks out /opens out toward the butte/ onto the patio . 4. 3.A . corresponding to our access paths.conception . in (i ). 1983 . The " e. Where the vertex trails . Still needing explanation. This mapping may be reinforced by the fact that the prospect path ascribedto an inanimate with an actual viewer located configuration . such as a cliff wall or a window . Thus. but the analysis is intended to be quite general across the range of . as in Thesnakeis lying int% ut of the mouth of the caveto refer to a snakelying straight with its head pointing into or out of a cave mouth.).272 Leonard Talmy one of its axes of symmetry. to the structure of the cognitive culture system . belong to the present approach. extension (e. In this work . The road goes " " arrow from New York to L . To note the correspondences . in ). in Talmy ( 1988b . Probably poorer as models are such other forms of agency as an Agent' s affecting some cognitive state that she herself has or somephysical object that she is already in contact with . though more salient in front . Due to the constraint noted above. one readily imagines a viewer standing at the cliff edge or in the bedroom looking out along the samepath as is associatedwith the cliff wall or the window. 7. cognitive systems " 8. Jackendoff ( 1983 ) has abstracted a concept of pure di " " " " " rectedness with four particularizations : actual motion . orientation (e. however . But the structuring of the ception concept found here. is often associated at that configuration and directing her or his visual path along the samepath as the prospect line. ( g.and Con.g. (i ) a. to the structure of kinesthetic perception. Talmy ( ) system chapter). possibly due to a conceptualization like this: the posited intangible line.

. note further the following .Fictive Motion in Languageand " Ception" 273 tenns include representation . It should be recalled that the palpability gradient has here been introduced fictivity in language to mainly help characterizegeneralfictivity in vision. . Once stimuli from the entity impinge on the body' s . but the original two representationsdo not. cognize ." not becauseit fits a category of a theoretical ontology according to whose tenets the " " entity is out there. -related representationsinto the visual modality does tend The mapping of two such language to involve a palpability contrast. conjunctions.g. 10. not from any other usesthis word may have beenput to in the psychological colloquial usage literature. verbs. allowing increasingly greater degrees that would tend to correlate with the high ends of the other parameters should be located within its upper region. All thesetenns have their particular applications and will be used in this chapter. The adoption of the verb to senseas a tenn for this purpose is derived from its everyday . Though linguistic referencecan be located along it . Insofar as it is concluded in our scientific ontology that an entity is in fact located external to one' s body. abstract fonns like grammatical categories(e. Perhaps alone out of the thirteen. including the portion that leads to sensory receptors consciousexperiencingof the entity.g. but the new tenn ception is specifically intended to emphasizethe continuity acrossthe larger domain and the existenceof largely gradient parametersthat span it . the neural processing of the stimuli . Linguistic categories like the preceding have been presented only to help illustrate the abstract end of the palpability parameter. II . As treated extensivelyin Talmy ( 1988a ). " nounhood" and " verbhood" per se ). We lack any direct consciousexperiencethat our processingof the entity is itself internal. As discussed . In physiological tenns. and adjectives large and easily augmented Closed-classfonns are categoriesof fonns that are relatively small and difficult to augment. never again leavesthe body. like the others. and sometimescognition. open-classfonns are categoriesoffonns that are . On the contrary. free fonns like prepositions. the parameter of strength has an open-ended upper of intensity. consisting primarily of the roots of nouns. grammatical relations (e. this parameter is not suitable for characterizing generalfictivity for language . we experience the entity as external. experience . Included among them are bound fonns like inflectional and derivational affixes . is intended as a phenomenologicalparameter . Despite this fact.. subject and direct object). the processingis system ' specifically organized to generate the experienceof the entity s situatednessat a particular external location. and complexes like grammatical constructions and syntactic structures. An entity is assignedto the high end of this gradient becauseit is experienced as being " out there. general fictivity in languageinvolves the discrepancybetweenthe representation ' s literal reference of one' s belief about a referent situation and the representationof a sentence . Thus the point along this parameter region. we apparently lack brain-internal senseorgans or other neural mechanismsthat register the interior location of the processingand that transmit that infonnation to the neural consciousness . 14. and detenniners. and word order patterns. The parameterof objectivity . 13. 12. . not becausethat parameter is relevant to general .

Language . . andspatialcognition manuscript language . MA. Talmy ( 1978a . and Rubin . A.. Detecting of Engel on Motion: Representation andAnalysis IEEE the Workshop . J. 16 (2). andcognitive . Hirschfeldand S.. Mappingthe mind : Domain . M. ( 1983 : An essay onfaculty psychology . 38. Dissociations Cooper .274 Leonard Talmy 15. L. 217 238 cognition . A. ( 1987 ). . ? Technical . Perception of dynamic infonnation in static handwritten ~ J. B. 111 and : Natural ontologies on cultural representations . Charleston . and Schacter .. Current Directions Science of visualobjects . and Richards . 227 ambiguous triangles . ) . On beyondzebra . MA : MIT Press . kinds . and PalmerS. MA: MIT Press . . and Jackendoff .: ). References Babcock. Stanford : StanfordUniversityPress . and Freyd.401 . Cognitiveconstraints Boyer . New York: Cambridge . A. ( 1986 ). ( 1986 studyof the prepositions ). J. F. and the interface between Herskovits . Cambridge : Cambridge Press University " and " " " Across .). Chicago Lakoff. S. ( 1993 ). Gelman(Eds religiousideas . ( 1993 ). Whatis a Percept report RBCV Jepson . Behavioral andBrainSciences . Unpublished . G. Perception andPsychophysics . R. : The relation of linguisticand visual information Jackendoff . in childhood . ( 1985 ). ( 1985 ).146 in Psychological .. What and where in spatiallanguage . A. N. ( 1989 . M. Foundations of cognitive grammar Langacker .236 . . Modularityof mind MIT Press . Cambridge . . ( 1983 ). A. 1(5). 26 . 19(3). . . 101 . ( 1994 ) along : Lexicalorganization . Conceptual Carey change between structuraland episodic .. R. In L. 89 Cognition -TR-93 -43 . Semantics . In Proceedings visualmotion boundaries . Concepts development : Universityof Chicago welive by. Toronto: Universityof Toronto Department of ComputerScience . 7. MA: MIT Press ). Cambridge Keil. ( 1994 ). D. and Johnson . These papers also present an expanded form of the linguistic demonstration synopsizedin the text below. ( 1980 ). ( 1992 ).. W. Cognitive momentum .114 . L. American fonDs Journalof Psychology . Explorationsof representational Psychology Freyd . . ComputerSociety SC . A. 141 representations . Metaphors . R. 1988a ) first observed the homology between vision and language as to a content/ structure distinction . . Cambridge Fodor.9 May.130 . . E. in cognition andculture UniversityPress specificity Bucher . P. in English . . ( 1987 ). A.. Effectsof motion on the perceived pointing of . ( 1988 . M. J. S. . Press " " " " and spatial Landau . A. R. M . Cambridge Jackendoff andcognition . ( 1987 ). 369 andspatialcognition : An interdisciplinary Herskovits .

165 205 Amsterdam : . Relativity in spatial conception and description. Press . ( 1993 ). Categories . Artificial . D . C. PhiD. 7.).. Oxford : Clarendon ." Fictive Motion in Languageand Ceptinn 275 Levinson. and Labov . The relation of grammarto cognition Talmy TIN LAP 2 Issues in Natural : University Proceedings ( Theoretical ). L. . The hippocampusas a cognitive map. S. Acredolo(Eds .. diss . .905 . N. Cambridge: Cambridge University Press . research . Urbana of Language Processing of Illinois. . In Language . thought Language Marr . Perceptual organization . A. P. Nadel. ( 1985 Talmy ). J. ( 1981 ). E. Syntax and . Jr. Cambridge : Cambridge .100 ). ( 1988a ). In Universals sentences . Press Palmer in configurations of . . Vol. Semantics Talmy . ( 1992 Linde . ( 1982 ) . andBucher . causality. Forcedynamics Talmy . ( 1986 ). Rethinking linguistic relativity . L. H. S. 57 UniversityPress -Ostyn(Ed. .). In . U. Journalof Experimental : HumanPerception andPerformance . Geographic space A. ( 1988b . M. Pick. L. Vol. S. E. ). Semantic Talmy types .). Levinson (Eds. ( 1978b ). (in prep. J. Frank and I. structures . Spatial networks as a site for the study of language and 51 924 . 225 Spatialorientation : Semantic structurein lexicalforms .). Figureandgroundin complex Talmy of human language . Vol. Topics to cognition . ( 1980 ). M.238 . 49. Symmetry. and O' Keefe. . T. ( 1986 ). ( 1978a ). 625 . Cognitive Linguistics. C. . ( 1983 ). Greenberg . CA: StanfordUniversityPress (Ed. 12 . . Gumperz and . Thegrammar of causative : A synopsis . Shopen and the typologyand syntacticdescription (Ed. In J. MA : MIT Press Leyton . ( 1976 ). Matsumoto . Spatialinformation : Springer . 4.149 lexicon . What makes triangles point: Localandglobaleffects . San Francisco: Freeman.282 : Theory . ( 1975 ). .). Lexicalization patterns . 939. NewYork: Plenum .). 3. Rudzka in . E. . Configural pointingof ambiguous . Subjective motion and English and Japaneseverbs. 88.116 (Ed. J. Technology and syntax of motion. L . Therelationof grammar Talmy . Intelligence . ( 1975 ) . . Stanford . Cognitive . NewYork: Academic constructions Press . andapplication . L. How language Talmy space . cognitive linguistics Benjamins in language andcognition . L. In D. 12 . S. Grammatical categories .. Waltz (Ed. . L. 181 causative . Cambridge. Vision: A computationalinvestigationinto the human representationand processing of visual information. 4. Shibatani . In Syntaxand semantics . J. Kimball (Ed. Berlin theory Pentland andthe representation of naturalform. M . In H. In J. . Cognitive Science . 285 ambiguous triangles Psychology Palmer in perceived effects .).).331 28 293 . New York: Academic semantics Press . L. ( 1978 ).). Y .114 triangles Psychology Pederson and manipulable in two Tamil linguisticsystems . Rubin of visual motion.649 . In B. M. . ( 1996 ) . vol. mind. Campari(Eds . 6. andLP .) . W .. Syntax . 43. Massachusetts Institute of . L.. L.

81. Monist .116 ) . L . J. Plenary addressat Talmy. L . J. Cambridge. MA : MIT . L .). ( 1995 . 1990 . ( 1982 Goodale. and Talmy . Hillsdale. G .. L . Mansfield (Eds.). H . Fictive motion and change in language and cognition . Conferenceof the International PragmaticsAssociation. In D . and Mishkin . 78( 1). W. 421. Barcelona. Duchan. : Theirform and meaning . Grammatical constructions ) . Analysis of visual behavior . July.). A . and R. NJ: perspective Erlbaum. Deixis in narrative: A cognitive science . In G. . Ingle. ( 1990 . Press . ) .460. Oxford : Oxford University Press (Eds. ( 1995a . ( 1995b L . Bruder. M . In M .276 Leonard Talmy ) . Thompson ) . Shibatani and S. The windowing of attention in language Talmy . The cognitive culture system Talmy. L . . Narrative structure in a cognitive framework. M . Hewitt (Eds. Two cortical visual systems Ungerleider.

if a satiated animal notices food in a location during its initial an environment of . it can on a subsequentoccasionuse that information exploration . we had to allow for the . Neuropsychological studieshad suggested " " cerebral hemisphereis specializedfor visuospatial processing . Second demonstrated in the been has that of function lateralization repeatedly impressive that while much of the right human brain. This theory can account for a substantial part of the experimental literature on the infrahuman hippocampus. Vector Grammar. what Tolman ( 1948 ) called a cognitive map. Milner showed that this memory function respectedthe generallateralization . is involved in the construction of an allocentric spatial representation " " of the environment.1 Introduction In this chapter I wish to return to a subject that Lynn Nadel and I first addressedin our book The Hippocampusas a Cognitive Map ( 1978 ) nearly two decadesago. interest are labeled map to direct as and used theselocations can subsequentlybe retrieved into the map goals behavior. however. placesof use for future stored locations label and their in the and . In order to extend the theory to account for the human data. we neededto extend it in two ways. Evidence from animal experiments proves strong evidencethat the hippocampus. information . The gist of the argument presentedthere was as follows. First . and the Cognitive Map Theory JohnO' Keefe 7.7 Chapter - The Spatial Prepositionsin English. During the initial exploration of an environment and subsequently . a cortical area in the mammalian forebrain . or episodic . For example . the left side has been given over to languagefunction . need a to satisfy hunger Upon finding itself in the sameenvironment it can retrieve the location of the food and useit to direct its behavior toward that location. Following her dramatic demonstration with Scoville of a memory function for structures in the mesial temporal lobe (Scoville and Milner 1957 ). we had to incorporate a temporal senseinto the basic map to account for the ability of humans to process and store spatiotemporal. Constructed and modified during exploration (a cognitive behavior). this map provides the animal with a representationcenteredon the environment and locatesit within that environment .

this semanticmap would incorporate linear temporal would serveas the deepestlevel of the linguistic information and in consequence . Before returning to the semantic map idea ." " The went from Mary to the library . therefore we sought to create a more general framework by following the work of Gruber . (Smith and Milner 1981 that this lateralization of function might be due primarily to Nadel and I suggested differencesbetweenthe inputs to the hippocampal map on the two sidesof the human brain and not necessarilyto any fundamental differencesin principles of operation. identificational sense possessional senseof the prototype . not according to their physicalloca . directions . They proposed that the parallels in surface structure reflected parallels in underlying meaning. languageis clearly not reducible to the set of spatial sentences . It would operate in the same way as both right and left . Nadel and I interpreted this to mean it might be possible to envisagenonphysical spacesthat located items. which would come primarily from the language centers of the neocortex and would consist of the names of objects and features and not of their " " sensoryattributes. In addition . Frisk and Milner 1990 ). In contrast. system However. I intend to explore the adequacyof one of thesein particular (O ' Keefe 1990 ) as the basisfor a semantic map . The book " " The rock went from smooth to pitted . We suggestedone tion . 1976 ). and there are now severalcomputational models available.1) . we have learned a great deal about the working of the infrahuman cognitive map at the physiological level.278 John O' Keefe of function: patients with damage restricted to the right mesial temporal lobe were amnesicfor visuospatial material. Places are patches of an environment that can vary in size and shape . Evidence gathered since has strengthened this conclusion . Spatial relationships are specified in terms of three variables : places. In the years that have intervened since the first publication of the idea. In the cognitive map theory . but according to their location in these other dimensions suchdimension might be that of influencebut did not develop this notion any further . whereasthose with left -sideddamagewere amnesic for linguistic material. it will be helpful if I elaborate some of the details of the basic theory as developed for physical space. the left human hippocampus would receive infrahuman hippocampuses a new set of inputs. who noted the similarity in structure between ( 1965 ) and Jackendoff ( 1976 " " went from New York to Los Angeles sentences such as " The message . " librarian went from laughing to crying . 1989 . providing comprehensionand narrative memory. In this chapter I would like to develop further this idea of the semanticmap. narrative the basis for . and a circumstantial sense . The right human hippocampus would receive inputs about objects and stimuli derived from the sensoryanalyzersof the neocortex and attributable to inputs from the external physical world . in this case the substitution of for the positional sense . and distances (figure 7. entities are located by their spatial relationships to each other .

In animalssuch as the rat. thesecan be identified in one of severalways: either as the local gradient of a universal signal such as gravity. they can also be located by their direction and distance from other places . movements n of features in that depending on the size of the environment and the distributi < ? environment. they are computedin real time on the basisof actualmovements in highermammals autonomous . They are located in terms of the spatial relations among the invariant features of the environment.The Spatial Prepositions 279 ELEMENTS FORA MAP -AB B ""z (~ ::::: =:: . ~~ ~ :::::::) MAP = PLACES ABC L AC I Aci L CB I CBI DIRECTIONS LAB DISTANCES I ABI Figure7. or as having a specifiedangle to a . As with places . whereas from actual they may become ." Directions are specifiedas a set of parallel vectors. The place codeis carried by the pattern of firing of the place cells in the cortical region called the " hippo campus . or olfactory currents. Distances and di~ tions can be represented by v~ tors drawn from oneplace to another . as the vector originating at a place or object and passing through another place or object (or passing through two places ). geomagnetism .1 and the distances and directions Cognitivemapsconsistof a set of placerepresentations between them .

gating the metric signals arising from such sources by the headby. 1990 .g. A path may be marked by a continuous feature such as an odor trail or a road but neednot be. nonspatial. for example direction signals so that only movements when the animal is heading in the same direction are integrated. and Ranck 1990 ).. postulatesthat the deepsemanticstructure of languageis intrinsically spatial and that other. Furthermore. translations of position in an environment are spe whose tail at the of movement and whose cified as translation vectors begins origin head ends at the destination. something about the structure of the systemby analyzing the way it representsspace in revived within case . This is the contention of a group of linguists called " location ists" or meanings " localists" Anderson 1971 Bennett 1975 . The direction codeis carried by the pattern of firing of the head direction cells in the postsubiculum (see . In either case . through updating the current direction on the basisof angular head movements ) . then it might be possible to learn . I have explored Euclidean polar models (O' Keefe 1988 . It is still not clear whether the spatial coordinate framework is a rectilinear or a polar one and whether the metric is Euclidean or otherwise. Distances between objects or placesare given by a metric. ) ( ( consistsin a verb and its a recent review) suggestthat the basis for spatial sentences . which can be marked by the negativeof that vector. Paths can be identified by their end placesor by a distinct name. For every direction there is an opposite direction . perhaps by means of metaphorical extension of their core spatial . for example Taube. places along the path can be identified and associatedwith the path . propositions are in some way parasitical on theseprototypical formulas. Within this spatial framework. Conversely . 1991 ). another cortical region that neighbors on the hip pocampal region and is anatomically connected to it . If the cognitive map theory is on the right track in its contention that the left human hippocampus is basically a spatial mapping systemthat has beenmodified to store linguistic as opposedto physical information . The basic unit of this metric might be derived from : either there is a reafferencesignal from the motor systemwhich one of two sources estimates the distance that a given behavior should translate the animal or use is . . grammar linguistics recently linguistically long theory. A path is an ordered sequence of placesand the translation vectors betweenthem. Theselocalist theories seeCook 1989for . and an example of an interoceptive input would be a vestibular signal. Muller . Vector addition and subtraction allow journeys with one or more subgoalsto be representedand integrated.280 John O' Keefe previously identified direction (e. the geodesicdistancebetweentwo objects or placesneedsto be computed . In recent papers . on a journey with more than one destination the optimal or minimal path can be calculated. made of environmental or interoceptive inputs which result from such movements An exampleof an environmental input would be a changein retinal location of visual stimuli . A tradition .

however. many of the spatial relations respectively described in spatial sentencesare conveyed by the prepositions. In most cases more objects or places or two between . an increasein vocabulary would eventually obviate the device need for the externalized map entirely. in most cases a point along the vector.The Spatial Prepositions 281 . Over time. these spatial relationships and their modifications can be representedby vectors. identifying the associatedcases initiator of the action. calculated from the spatial relationships or vector of the head the and ) which specify the origin and termination (or the tail . In some cases(for example. By contrast. The spacecoded by the locative prepositions is a mixed necessary polar rectilinear one. seebelow). This view suggests person to another (O Keefe and Nadel 1978 the basic to elaborate hominids cognitive map by began point in their evolution substituting sounds for the elementsin the map or for someof the sensoryaspectsof . These maps were probably primarily transmitted as drawings in the theseelements sand or dirt with different icons standing for different environmental objects. In this chapter I will set out the basic framework of vector grammar and show how it accounts for many of the spatial meaningsof the spatial prepositions. Different grunts would enrich the detailed information in the map and might serve the additional purpose of acting as an encrypting . temporal) senseas in English have a spatial (or in one or two instances . the thing acted on. but the neural substrate would retain the structure of the original mapping function . and locative. seeabove) that the prepositions . it needs to be signal such as the force of gravity . . object. As Landau and Jackendoff ( 1993 ) have pointed out in their recent article. Much of the work of the locative prepositions involves the identification of these two variables. In an uninflected languagesuch as English. My thesisis that the primary role of the prepositions is to provide the spatial relationships among a set of places and objects and to specify movements and transformations in these relationships over time. . with vertical prepositions. Typical casesmight be agent. there are only a limited If this be the case then it is possible that a description of the . The location of an entity within this notation is given by a vector that consistsof a direction and a distance from a known location. In this chapter I will assume(following the location ists. Nadel and I speculatedthat the origin of language might have been the need to transmit information about the spatial layout of an area from one ' that at some . these number of representationsset up by the spatial prepositions might provide the basis for a more general linguistics. One of the roles of the preposition for is to supply the metric information . distancesare lesswell specified the metric is an interval one. and the place or places of the action. In this way one group of a family could forage a patchy environment and report back the locations of foods to the rest of the family . the direction is given by an environmental . 4O1n ) .

The syntacticrules specify generating . which differs in stateor locationcanbe . In . over. If time can be coded prepositions a fourth dimension is it to other . Although thesehave antonyms (above . For example the difference between the active ." They differ from each other in interesting ways that will allow us to explore the properties of the spacethey depict.2 Physical Spatial Meaningsof the Vertical Prepositions In this section I shall analyzethe spatial meaningsof four related prepositions: be/ow. I will higherdimensional conclude with a discussion of the metaphorical usesof the verticalstativeprepositions to represent the nonphysical relationsof statusandcontrol. Thus motor programmer . the orderin whichtheelements the differentpartsof the vectorsystem areto be translated into surface elements asa functionof thewaythat theyareread . and the passivevoice in the surfacesentence dependson the direction of travel along the underlying vector (head to tailor vice versa ) relating an agent and its actions. The four prepositions have in common that they denote 1 spatial relationships betweenentities in one linear dimension. although to showhow the temporalprepositions codefor a fourth dimension . just asthereis an associated system that translates informationfrom the spatialmap into instructions to the motor sothat the animalcanapproach desirable planningsystems places containing objects or avoid oneswith undesirable . and beneath(or underneath ) . 7. thesyntaxof sucha system general structureof the deepsemanticnarrativeto the temporalstructureof the surface informationtransmission .2. be/ow. down. by possible incorporate nonspatial relationships by axes aswell? As a preliminaryexplorationof this question . and on top of>. I will concentrate on the locativeprepositions and in particular thosedealingwith the verticaldimension otherswill alsobecovered . under . Consequently in anydetailthe role of syntaxin this kind of grammar . I will not address will consist of a setof rulesfor relatingthespatial .John O' Keefe their basicmeaningand that the other meanings are derivedby metaphor . which I shall call the " Z -dimension. be/ow relatestwo entities (A and B) in terms of their relative location along the Z -direction. so also there is a production systemfor objects sentences from the map narrative . among otherthings of the narrativeareto be readand how . 7. and how changes only slightlyfrom the threespatialones codedby the translationalmeanings of the same . I will thenextend theanalysis .1 Below Let us begin with what I believeto be the most basic of the four prepositions. On my reading. up. in this chapteris to setout the premise that a vectornotation My primaryconcern cancapturemanyof thebasicmeanings of the spatialprepositions in English . Consider the simple deictic sentence . I shall refer to theselatter only when they contribute something extra to the discussion .

which of the two . a way of locating A and B along that direction . the question of whether A is below or above B can be assessed by comparing their relative magnitudesalong the Z -axis. and means for assessing whether A or B has a larger component along that direction. and abovethe observerif Az > obs.3 shows this situation ). In this notation a direction is designatedby a set of parallel vectors of unlimited magnitude and unspecifiedmetric. the speakers (or listener s) place is B. The most convenient notation for accomplishing theseis vector algebra. there must be a second suppressedterm.The Spatial Prepositions 283 ( 1) John is below. This vector can be specifiedby a magnitude R and an angle. the previous sentences depend more on the topic of the discussion entities has already been located. In order to make the assertion in ( 1). In the deictic example of sentence I . is the angle that A makes with the Z -direction vector at observer(obs) . . where A is the magnitude of A and . or to assess its validity . which is easier to locate perceptually.Bz < 0. The location of each entity is specifiedby a vector drawn from an observer to the entity .2) . and the relationship betweenthem is as follows: the magnitude of the component of A ' s place in the . Again . Note that the relationship betweenA and B is perfectly symmetrical and that neither A nor B can be considereda referenceentity in the deep structural representationof the relationship. we needa notation for specifying the Z -direction . A above B . if Az > Bz and Bz > Dz. The be/ow relationship is a transitive one. . . . Choice of one or other as the referent in the surface sentencemay . The component of the vector A along the Z -axis can be computed by calculating the inner product of A and Z given by the formula : Az = A cos. The sameconsiderationsallow the observerto decide whether A is below B when neither is located at the observer (figure 7. Az > Dz . and other . which I shall argue is the place occupied by the speakeror the listener. Becausebe/ow is a bipolar preposition. A is be/ow the observer if Az < obs.Bz > 0. . with the direction vector through the point observer (figure 7.Z -direction is ' greater than the magnitude of B s place. If Az . John or his ' ' place is A . A below B. If Az . By simple transitivity of considerations arithmetical relations on the Z -dimension.

. The location of an entity A can be representedby a vector drawn from the observerto that entity . Figure 7.. In the example . A BELOW Observer . Observer " r '"". C. The relativelengths of the projection onto this axis determine which itemsare belowwhich . . z B.~ tk ~ Q . measuredwith respect to a direction Z . The projection of the vector onto the Z direction is shown as Az . Band C have identicalprojections andaretherefore both equallybelowA. The vector is characterizedby a distance R and an angle .D Below A Fiaure7.284 John O' Keefe 0 . and D hasa projectiononto the Z -axis .3 Eachitem A.A ..2 Vector location and the below relation. B.C.

chapter 3. There are. chapter 4) . In general then nonnal . 2 The new ( ) planet appearedbelow the moon. which are probably labeled by referenceto their canonical orientation relative to gravity (see Levelt. the conversantsmust ensurethat they are similarly aligned to each other relative to the entities or that there is a conventional orientation relative to the gravity signal that enables the Z -direction to be labeled conventionally . (4) Hitting below the belt. seemto require the use of an allocentric framework for most purposes . Further . that is. Does this imply that the spatial relationship denoted example by be/ow can be computed only within an allocentric framework? Can we say anything about the constraints on frameworks that can be used ? In general . This is most obvious with the specializedcase of the parts of the human body. Sentences 2. this volume) .3. The case of the bottle and similar manufactured objects that refer to body parts (back of a chair . for the reasons pointed out by other contributors to the present volume ( Levelt. leg of a stool) would seem to follow the same rule. as on a page or a video display unit . and object-centeredvectors are fixed to the entity or entities related.5 are examples . I assumedthat independently of the locations of the observers the distancesfrom each observerto the entities was known or could be computed.2 and 7. by movementparallax. In the fonner case . Where it is used to describean allocentric relationship. (5) The label below the neck of the bottle. Egocentric vectors are fixed to the body surface of the observer . Even the ability to see things from another' s point of view would appear to involve computations basedon an underlying knowledge of the two ' observers locations in allocentric space .The Spatial Prepositions 285 In figures 7. other. however. in the case of the allocentric framework. the use of be/ow relies on the availability of a direction vector shared between the speaker and listener. for .D in an allocentric framework . I assumedthat they existed in an environmental framework independent of the location of the observersand that their relationship within the framework could be assessed . I choseto represententities A . the framework cannot be a simple polar coordinate system . (3) Below this line on the page. Levinson. this is provided by the universal gravity signal. or (b) where the entities are constrained to lie on the XZ plane. A secondconclusion can be drawn about the underlying framework on the basisof our discussionof be/ow. chapter 3. The egocentric use occurs under circumstances(a) where the entities are very far away from the observerand therefore do not changerelative locations with observer location . more limited uses of be/ow that employ egocentric and object-centereddirectional vectors. but must have at least one . conversation would .

The question is whether the comparative notion is an intrinsic part of of be/owness . are equal to eachother. with at least an ordinal scale provide information about the differences between the scale values. For the purposesof the scaling. and b the origin . and only transforms classes are allowed. For example . the below relationship equivalent to the relabeling of the classes . a > 0 In this linear transform . they differ in the number of properties of the real number systemthey . they assertthat somedifferences Transformations that preservethe differencesbetweenvalues as well as their ordering .2) . Can we say anything more about the metric at this stage . This follows from the simple observation that in a polar coordinate system the below relation cannot be specified by one variable alone. We have. a changesthe gain of the metric. It would appear that the be/ow directional scalecomesclose to fulfilling the requirements for . The relationship betweennumbersis transitive because m > nand n > pimplies m > p . Nominal scalesare simple classification scalesin which the labels the thing measured . ordinal . implies possible say in addition .286 John O' Keefe rectilinear axis.n = p . the elementswithin stand for the namesof classes each classare consideredequivalent and different from all the elementsin the other . The usual categoriesof scalesare the nominal. m . One way of testing this is to ask whether it is possible to apply the an interval scale comparative operator more to the preposition and thus to derive equivalent intervals . No other relationship among the elementsis implied . Specifically are permissible . how are distances . This is most easily characterizedby the types of transformations that can be respect applied to the assignedvalues without transforming the relationship of the scale to . As we shall see . It follows therefore that the most parsimonious theory would specify the Z -direction by a single dimension in all . but requires two variables: a distance and an angle (seefigure 7. constant and added to another constant without consequence Z2 = aZ . but no significanceis attached to the interval betweenthe numbers. interval . that scales . Roughly. and all mathematical transformations that maintain . evidencefor a single dimension along which entities can be located. ratio . this does not necessarilyimply that the other two (non-Z ) usages dimensionsare also rectilinear. then. the valuesof one scalecan be multiplied by a positive to relationships. + b. in of the differ ? Scales2 this dimension type metric employed specifiedalong the to or measurements this describesthe relationship of the observations systemof real numbers. Becauseit is the monotonic ordering of the numerical assignmentsare permissible are dealing A we below C B below A and C B below that to . Clearly. and absolute. and if so. Ordinal scales consist of a series of numbers such that satisfies a nominal scale observations equal to each other are assignedthe samenumber and an observation larger than another is assigneda larger number. Interval scales are ordinal .q. In particular .

I would argue that becauseit is always legitimate to ask for the relationships set out in (8). the final category we shall consider. it is intuitively obvious that changesin scale do not affect the relationship either. These suggestthat it falls short of a ratio scale . a > O In absolutescales . (7a) . the scaleis an interval one. the metric of the be/ow relationship is at least ordinal . In addition . It can. Indeed. 7. Here the only permissibletransform is the gain of the scale .Z -direction. This line or surface is the object of the preposition down. Z2 = aZl . (9) a. the directional component of down is relative to another entity. (nominal) (7) A is more below C than B.2 Do H '" (and Up) The locative meaning of down is related to that of be/ow in that it specifiesthe direction of the entity as lying in the . A is more brighter compared to B than to C.The Spatial Prepositions 287 the meaning of below or merely an extension of it .2. Ratio scalesare interval scalesthat do not have an arbitrary origin . A is three feet below the surface. which in this case is . b. it requires a line or surface that is not orthogonal to the Z -direction and on which the entity is located. no transfers are allowed and the underlying assumption is that the real number system uniquely maps onto the observations Z2= Zl . But is it ratio? Here the fact that the be/ow relationship can be assessed from any arbitrary observation point and can use any origin suggests that it does not rely on a fixed origin but is invariant under arbitrary translations. chapter on material ordered on lessthan an interval scale (6) A and B are below C. As with be/ow. . (interval) Compare theseto (6a) A and B are brighter than C. (8a) . however. Furthermore . be elevated into a ratio or even an absolute scaleby the provision of explicit metric information . however. and probably interval. A is twice as far below B as Cis . it may not be possible to compute the vector calculations suggestedin this . As we have seenalready. (ordinal ) (8) A is as far below Bas C is below D . A is more brighter than B by the sameamount as C is more brighter than D.

the preposition of down only takeasdirectobjects in the nonhorizontal entitiesthat haveor canbe treatedashavingmonotonicslopes down the more to our . . preposition comparative plane Applying of the the Z on to is sense its that did with below component operate . ( 12 ) *The boat wasdownthe ocean : a planeor line that I shallcall the " reference entities Thustherearetwo reference " " " entity. the hill has a projection onto the Z the dimension . as we . ( 10 ) The house ( 11 ) Justdownthe treefrom Samwasa largetiger. one dimension in which system down between difference the above direction below with the As to this. and John hasa larger . regardless . canbea one Intuitively functionof Z over the decreasing entity shouldbe a linearor at leastmonotonically 's of the person on the other side of the hill . it is or tail of a directionvector explicitly source supplied . indicatingthat the non-Z -dimensions that either or surface line a from Z the to extract suggests sloping component ability or that non Z Z and into two thesecan be decomposed ) ( orthogonalcomponents our of the basis then on It seems . relationship ) down thehill than Jill. This can be tested of the preposition of the reference planeand the sense steepness more Who is 7. primitive . that we are dealingwith at least a two dimensional analysisof down more dimensions one or other the and is vertical . onto the Z axiscanbecomputed . of signand thereareno obviousasymmetries and its antonymupis merelya change axis would the Z scale of If A is downfrom B. Someone relevantrange .Z-coordinate . their projections coordinate . In generalthe preposition governedby the preposition is not information . To put it anotherway. the Jill thanthe other. / orthogonal .Z than Jill. we find. . Similarly the reference and located the hill between the of theslope entitydisruptstheuse entity down can .4 in three of the the ) down (farther people figure by asking question or Jim? the hill from Jill? John of downis that neitherJohnnor Jim is moredown from of the meaning My sense . . surface dimensional a two line or dimensional . it ) asin ( 12 entity is not horizontal(perpendicular this reference . Themeasurement of a true 0 or the absence of evidence is clear and there one interval to be an appear . As long as the plane and a placeor objectthat I shallcall the reference to the Z axis reference extended ). Thereis no interactionbetween . ( 13 ) Johnis more(farther John and Jill are both locatedon the hill . However areirrelevant . is not downthehill from you.288 O' Keefe John from identifiesthe from. a localminimumon relative. is the deicticlocationhere that the referent assumed is down thehill (from here ). thenB is upfrom A. If this .

JohnandJim areequallyfar downthe hill from Jill. then an overlap in any location in the . be garneredfrom an analysisof the third of our prepositions.4 Downmeasures in the Z-direction the relationship .5 shows this relationship for three pointlike objects. however. B is not under A . The projection of the entity onto the X -direction is determined in the same way as that onto the Z -direction by calculating the inner product of the vector drawn to the entity from . however. then it must have a more negative value in the Z -dimension. If B is under A . however. 7. it placesrestrictions on the location of theseentities in one or two directions orthogonal to the Z -direction. under. In addition . In addition . Figure 7.2. despite differentlateraldisplacements .component of the meaning. if the entities are extendedin the XY plane. and therefore the scale is not a ratio one. it must have one or more locations in common in at least one orthogonal dimension (let us call them X and Y for the moment without prejudice to the question of the best representationof relationships in this plane) . the under relationship can be assessed by the samealgorithm . For example . origin (this is relative to the referencepoint identified byfrom ). The scale of the other two dimensions is not clear from the two prepositions below and down becausethe use of the comparative operator more in conjunction with theseonly operateson the Z .3 UlUler Under is similar to down and be/ow in that it also codes for the spatial relationship betweentwo entities in the Z -direction. When one or more of the entities is extended in one or more of the non-Z directions.The Spatial Prepositions Jill 289 Farther Down the Hill Figure7. Evidence about theseother dimensionscan. The relation an observer depicted is conveyedby the sentences ( 14 ) C is under A but not under B .

( 15 ) Under the canopy of the heavens ( 16 ) Under the widening sky. which I shall designateunder1 the more frequent usageof under. . and down and be/ow. .5 Under representsa spatial relationship in the XY -plane as well as the Z -direction.componentof the vector to the entity . Another interesting differencebetweenunder . . C is not underB because Cx lengths differ. on the one hand. the effect of the comparator is not fixed but dependson the relative dimensionsof the two entities. B under A and C under B does not mean extended entities that are that C is underA . Note that unlike be/ow. In The comparator cannot be applied to theseusages . Compare the following : two sentences . Let us leave aside for the moment the small number of usagesthat seemto mean that there is no intervening entity betweenthe two relata: . John O' Keefe Bx Direction 0 B II Bz C - Cx Ax C under A . ariseswhen we examine the locus of operation of the comparator more. . on the other.Z -length. the comparator is more often found to operate on the orthogonal X -dimensionthan on the primary Z -dimension.but not under8 Figare 7. When applied to under . Recall that when applied to be/ow and down.290 X Direction . under is not transitive when applied to XY -plane suffices in the XY -plane. more acted to increase the length of the Z . C is under A the Bx and because it has the sameX -length and a greater .

( 18 implies greater depth ). and the substitution can usually be made transparently. that is. ( 18 ) The box was farther under the table than expected . implies a greater . In contrast. ( 17 Ignoring the metonymic usesof table and water. it is clear that the first usage ). Further . and there is no information contained in the preposition about the lateral variable. . have locations projections in common. under2is responsiveto thesedimensions . this magnitude will be a length along some vector (e. It follows that any change in the lateral location of the lower one will not affect the amount of overlap. where both relata have a limited extension in the XY -plane. 291 IE ~ A B more under than A Figure7.. Y in figure 7. the magnitude of some aspect of the projection of B onto the table is greater than that of A .6: ( 19 ) Stick A was under the table. a or Z dimension . In general .The Spatial Prepositions . ( 17 ) The wreck was farther under the water than expected . Consider sentence( 19 ) and related figure 7.6) measuredfrom the edgeof the table to the farthest edge .6 StickB isfarther (more thetablethanstickA because thereis a greater ) under lengthof overlap with theprojectiononto the XY-plane . In the first usage . We can use this fact to explore the properties of the second and third dimensions of spatial language and the relations between these and the Z dimension. under acts as a synonym for be/ow. These usagesmay be confined to situations in which the upper entity is very long relative to the lower one and completely overlaps with it . which I shall designateunder2 length in the X -dimension. while the second . but stick B was even farther under it .g. I read sentence( 19 ) to mean that both sticks A and B and the table (top) have onto the XY -plane and theseprojections overlap.

D more under2than C in figure 7. The sentencealso confirms that both measurements scaleand that the samemetric applies to each. This slight asymmetry appears not to relate so much to size as to relative mobility . This last sentencealso suggeststhat the meaning of under2 in the XY -plane is a distance and not an area. Applying the comparative test to the preposition under revealsthat the metric is the sameas that for the . that is. any differencesin the projections of the objects in the Z -direction are irrelevant. Evidence for this can be gained by imagining the sameor different objects of different projection sizesand exploring the meaning of (23) A farther under than B.Z -direction .292 John O'Keefe of the object projection .7 suggests dimensions either of which could be taken into consideration.. Furthermore. My claim that A more under2refers to the absolute length of A might appear to be contradicted by sentences such as (24) Mary got more under the umbrella than Jane and thus got lesswet. in which caseeach distance is measuredfrom the edgeof the table intersected are on an interval by the chair. as theseobjects are positioned in different ways undera constant-sizetable (seefigure 7. over does not show complete symmetry with under2 In somesubtle sense . the distance under2is taken from the longer length." and does not constitute a refutation of the presentproposal. Consider (25) and (26) : . This conclusion is strengthenedby the fact that it makes sense to say (22) Chair A was as far under the table as it was below it . Thus (20) Box A was farther below the shelf than box B and farther under it . It is interesting to note that .7) . a greater proportion ) under " " " " the umbrella. the table is lessover the chair than the chair is under2the table. (21) Chair A was as far under the table as chair B. the orientation of the objects need not necessarilybe the samebecausethe relevant length is taken from the intersection of the object with the edgeof the table or from the nearestedge( C is asfar underas B) . This clearly implies that Mary got more of herself (i. In this usage . Note that this sentence can be used even when the chairs are at right anglesto each other. that when an object has two Finally . (for down ) and above(for be/ow). Figure 7. it is clear that more modifies Mary rather than " under.7 shows that the judgment of which objects are more under (or more under2 ) does not depend on the relative proportion of the length that intersectswith the referenceobject (B more under than A ). however.e. an interval scale . unlike the antonyms up .

.. First . I I I I I I I I I I I I I I I I I I I I I I I I I I.- A B Figure7.. Whentwo objects differ in morethan onedimension . it has a more restricted sensein the XY -plane...4 Belleatll (or Underlleatll ) Beneath (or underneath ) has a meaning that is close to that of under but differs in two ways. -I I I I I I I I ' I ' . 0 r -' I I I ' I I I ... Whereas under means an overlap between the projections of the reference entity and the target entity. is determined dimension of eachandnot by thetotal area(D > C).. may places map primarily by theinvariant features of anenvironment andonlysecondarily andtransiently by which them . .. Thereason forthis ... farther under by thelargest red carwas under thestreet ..~. (26 ) The lamp Sentence in most contexts .2... objects occupy 7.. objects or theorientationof theobject(C > A). butless (26 ) isnotincorrect likely . beneathmeans that the target entity is wholly contained within the limits of the reference entity projection . and in part as a of this restriction. (25 ) The lamp street was over theredcar . It follows that the projection of the lower entity in the XY -plane must be smaller than the upper.- . at least in part be that the in thecognitive arespecified .The Spatial Prepositions D 293 .7 The relationship moreunderis determined the two by the total lengthof the overlapbetween in the XY-planeand not by the proportionof the total objectwhichis under(B > A). Furthermore. the application of the comparator more (or farther ) consequence to beneathoperateson the Z -direction and not on the XY -plane..

It follows that . Near is not simply derived Euclidean the line determined the by geodesic along of the proportional distances a sense from nearer but contains in addition among the items in question. provide the necessary and sufficient condition for useof the comparativesnearerandfarther . S 7. Let us begin by examining the meaning of near when points are being related. (28) This road goeson for three miles. Note that the directions of the points from each other are not confined to the samedimension but are free to vary across all three dimensions . they are not. with only two points.or three-dimensional entities shows it is the surface points that are important and not any other aspect of their . in figure 7. (29) The housewas near (far from ) the lake. and C.8a. and that the distance is measured metric. For gives the length of a path. Consideration of the near/far relationship of two. Thus. Changing the relations of other items in the set can alter whether two items are near to or far from each other.Z of to have a slightly more restricted meaning in the sense seems direction. but in figure 7. B. In somecases . A . to be calibrated relative to distances The distancemeasureincorporated in near seems betweenthe items with the smallest and largest Euclidean distance separation in the set. Theseitems act as anchor points that control the meaning of the terms for all the others. Beneaththen meansthat the target element is contained within the volume of space defined by its XY -projection through a large (or infinite ) distance in the . dependent neither is near (or far from ) the other. (30) A is not near B but it is nearer to B than Cis . near and far from give relative distances that are . Three points. one or more of the contextual referentshave contextually dependent been omitted. and I will pursue that line here. 8) observedthat the meaning of near was context. Band E are near each other. Underneath limiting the projection in the Z direction.294 John O' Keefe (27) The red tray was farther beneath the top of the stack than the blue one. More underneathsounds lessacceptable than more beneathand might indicate that underneathis a three-dimensional volume of spacerestricted to the immediate proximity of the . O ' Keefe and Nadel ( 1978 .8b.3 Distance PreJ ")Sitioll near ( to) and its antDistances are ~iven by the preposition for and the adverbials onymfar (from ) as in (28) and (29) .Z or undersurface of the reference element.

TheSpatialPrepositions (a) B F E D 295 AB C B Figure 7. Finally .8c and ask which is nearer to A ..g. shape (e. because . centroid) or mass (center of gravity) . shapeB or shapeC. the presenceof barriers seemsnot to influence our judgment of near or far . we will seethat B is. If we inspect figure 7. by virtue of point x . In (c). : Reprise These considerations of the meanings of the verticalprepositions the following suggest conclusions : . 7. (31) is permissible (31) The houseis nearby.8 Nearnessis context-dependent . B is nearer A than C is by virtue of point x . but it will take a long time to get there sincewe have to go the long way around.4 VerticalPrepositio . In (a) A is not near B but nearer than CE is near B in (a) but not in (b) .

and distances . directions. the distance between objects is calculated from the nearest surface of each entity and not from some alternative derived location such as the geometric centroid or center of mass. the horizontal direction vectors are local and need to be coordinated relative .S HorizontalPrepositions The original cognitive map theory suggestedthat . which is constant regardless of location . irrespective of the animal' s location in that environment. or . to each other.296 John O' Keefe 1. Interestingly .1) . specify . In the vertical dimension . Foremost among thesewas their relation to other placesas determined by vectors that coded for distance and direction (figure 7. Static locative prepositions relate two entities. whereasthe latter is a universal signal. The directions are given relative to the places in terms of directions and distances . direction can be given by the universal gravity signal . 7. The resultant direction this computation is independent of the observer vector functions in the sameway in the horizontal plane as the gravitational signal in the vertical direction. The primary differenceis that .9) of the two vectors from the observer to each of the entities. These cells are selectivefor facing in specificdirections relative to the environmental frame. The direction vector originating in one placeor entity and running through a secondcan be computed by vector subtraction (seefigure 7. and ' s location. in the horizontal plane. nothing comparable to this signal is available and the direction vectors must be computed from the relative 3 positions of environmental cues. but there are also circumstances in which the two non vertical dimensions may be expressed in polar ( or other ) coordinates . Prepositions identify relationships between places . The nonvertical dimension ( if present) may be rectilinear . The metric of vertical and nonvertical axes is identical because it is possible to compare distances along orthogonal axes. 5. The scale is an interval scale with a relative origin detennined by one of the reference entities of the directional prepositions (usually the vector source or tail ) . in common with their vertical cousins. and static distance prepositions also relate three entities becausethis is the minimum required to give substanceto the comparative judgment that they imply . 4. This is achievedby mapping them onto the global directional system Locative horizontal prepositions. In the horizontal plane . 3. 2. The space mapped by the prepositions is at least two -dimensional and rectilinear in the vertical direction . places could be located in severalways. static directional combinations of these prepositions relate three entities becausethere is always an (often implied) origin of the directional vector. In that the direction component of this a recent paper (O ' Keefe 1990 ) I have suggested vector is carried by the head direction cells of the postsubiculum.

5. . A C ) terminating on different parts of the referenceobject or place. . . it acts in a manner analogous to be/ow in the vertical dimension.The Spatial Prepositions ~ . As shown on the left side of figure 7. No restriction is placed on the location of the entity in the vertical direction .. According to this definition ..10. .1 Beyond Let us begin with an analysis of the spatial meaning of the preposition beyond . '" " c " W " " . '~ ~ irection Vector AB ~ i f ij" :~ : . '. 7. Furthermore.. " . Observer '. as can be seenfrom sentence (32) : (32) Janecamped beyond and above the woods. this specifiesa three-dimensional region located with--to a specificrelationship to the referencedirection and a pair by the set of vectors--to of referencevectors (AB . Vector AB 297 Figure7. the effect of the comparator more is to act on the length of the vector in the horizontal plane: (33) The tower was farther beyond the mound than the castle. and distancesare given relative to the length of a standard vector drawn betweenthe two referenceentities along the referencedirection. The region beyondthe mound is specifiedby the set of vectors originating at A whose projection onto the direction vector (inner product) has a greater length than --to the larger of the two reference vectors coincident with the direction vector (AC ).9 The directionvectorthroughtwo objectsA and B can be computed by taking the difference between thevectors A and B. direction vector. " ." " '.

(34) me tree was behindthe trench. Beyond is the setof all places with locationvectorsgreaterthan AC and anglewith of beyond and includes only the places thoseplaces the directionvectorsmallerthan AD.10 . behind by their relation to the Beyond -. with thedirectionvectormustexceed -used behither means that the The opposite of beyondis the seldom .298 John O' Keefe A Beyondthe Mound Behind the Mound A Besidethe Mound Figure7. An object behinda referenceentity on location than doesbeyond . is located by the set of vectors with a ---+ larger magnitude than the referencevector (A C ) but with an angle less than vector AD (figure 7. center) . Behind AD). and the test for this is an overlap in the projections of the two in the XZ -plane. and this ---+ simply location vector has a length lessthan the referencevector AB .10 detennined can be represented as places . In the sameway that farther under can refer to the amount of overlap in the XY -plane between two . (35) me cottage was behindthe lake.. 7. Beside represents havinga projectiononto In additionthe angle AB less than AC. an entity can be partially .. The application of the comparator test shows further similarities. entitiesand a setof reference vectors directionvectordrawnthroughtwo reference (AB. than and the reference directionof magnitude greater that of AD. AC is a restricted subset with a lengthgreater than AC. As with under behindthe referenceentity.5. and beside . This need for overlap accounts for the awkwardnessin using behindwith referents that are not extendedin the vertical dimension.2 Be1lill4 Behindfunctions in a manner analogous to underin that it placesgreater restrictions .

right) . of before . this is easily overridden by the motion of the vehicle: (39) The car careeredbackward down the hill . it would be legitimate to omit the first clauseif the previous (37). At . However. about . across. the source of the direction vector is the implicit deictic here. for example . or more usually in front of 7. 7.to . the front to the back of a car. More usually. as. from .5.TheSpatial Prepositions 299 entities separated in the vertical dimension. (36) The red toy was pushed farther behind the box than the blue ball.-:iC ' but whose angle with the referencedirection is greater than that of referencevector AD (figure 7. The source of the direction vector can be specified explicitly as the object of the preposition from . around . Jameswas hidden behind the boulder. Familiar objects have " natural " behindsestablishedby a vector drawn from one differential part to another.6 Omnidirectional Prepositio . along . to . The opposite of behindis before . opposite . In a pool game it might be the cue ball: (38) The last red was behind the eight ball.3 Bes . About relaxes the precision . so farther behind can refer to greater overlap in the XZ -plane of entities separatedalong a horizontal referencedirection.5.de Besideidentifies a region at the end of the set of vectors whose projections onto the referencedirection fall between the referencevectors All and . 7. against . In sentence . via . (37) From where Jane stood. among (amid ).one substitution operator that locates the entity in the same place as the reference entity . and beside with a slight preference for the latter . being inferable from the previous context. scattering pedestriansin front of it and leaving a trail of destruction behind it .4 By By is the generalized horizontal preposition and includes the meanings behind . between. for example narrative had establishedthat Jane had been looking for James .10. and through locate entities in terms of their relationships to other entities irrespective of their direction in a coordinate reference framework and therefore can be used in any of the three directions . At is the general one . beyond. the source is implicit . More often.

if the circle it lies on several contiguous places along . (40) The shop was around the comer.5). Betweenlocates the entity on the geodesicconnecting the two referenceentities. The computation is the sameas that for deriving a direction vector from the subtraction of two entity vectors (seeabove discussionin section 7. About is . Therefore the area covered by about is relative to the distribution of the other distancesin the set under consideration in the sameway that the meaning of near dependson the distribution of the entities within the set. except that the order in which theseare taken is ignored.. more compact. in almost all instancesthe radius of the circle is left undefined. but not when about is used. Among increasesthe number of referenceentities to greater than the two of between reference The interesting issuehere. Among roughly meansthat the target entity is within some imaginary boundary formed by the lines connecting the outermost items of the set. But clearly the membership of the referenceset itself is not immediately obvious. there is the weak presumption that they all lie on the samecircle when around is used . . and this would appear to apply to about as well. there is little to choosebetweenthe use of about and around when single entities are located. except that it Because be small relative to the averageinterentity distancesof the other membersof the set. An equivalent definition of betweenis that the sum of the distances from each of the referenceentities to the target entity is not greater than the distance betweenthe two referenceentities. Consider a cluster of trees with an individual outlier pine tree somedistance from the main group . as with many of theseprepositions that usemultiple entities. . In the cognitive map theory the size of the equivalent to at plus contiguous places place fields is a function of the overall environment. (42) He was not among the trees This suggests that the application of the preposition amongdependson a prior clustering to determine the numbers of the referenceset. the set of vectors of a constant R originating at an entity) with the referenceentity at its center. When multiple entities are located.e. (41) Those who could not fit around the table sat scatteredabout the room. Around has at least two distinct meanings . is how the referenceset is defined. on that circle. Amid operation that is necessary is a stronger version of amongthat conveysthe senseof a location near to the center .300 John O' Keefe of the localization and introduces a small uncertainty into the substitution. both related to the underlying figure of a circle (i. it lies at one place on the circle perhapsat the end of an arc of the circle. the angle made by the vectors joining the target to each of the referencesshould be 180 . but stood betweenthe thicket and the lone pine. Alternatively . If it is extended The first meaningis that the locatedentity is somewhere .

from and to mark places at the . along. theory (O Keefe 1990 Across . and oppositeare like down in that they situate an entity in terms of its relationship to a referenceentity and a one. Because my emphasisin this chapter is on the prepositional . and the present . not attached to it but is supported independently in the vertical dimension. One possibility is that the centroid or geometrical center of the cluster is computed. there is the weak presumption that the (including 0) of intersections distance from the target entity to the last intersection is roughly the sameas from the referenceentity to the first intersection. grammatical and lexical features and the temporal prepositions. The centroid is a central concept in one computational version of the cognitive map ' ). _ ion . aspect . The location of these time events is also based on vectors and thesecan be oriented in either direction from a reference point . this constraint being due to the fact that changesin time can take place in only one direction (from past to future) . . however. time is representedas a set of vectors along a fourth dimension at right angles to the three spatial ones. Times future to the referencepoint have vectors of positive length. 1976 ) / 1985for detailed discussions In the presentsystem .7 Temporal Prepositio The incorporation of time into the mapping systemis accomplishedthrough various . both are roughly the same distance from the referenceline or plane. Each event is representedas a vector that is oriented with its tail to the left and its head to the right . The primary grammatical featuresare tense . which can be the present moment of the utterance or someother time. (43) Oxford Street goesfrom Tottenham Court Road to Marble Arch via Bond 't Street but doesn passthrough Hyde Park. Two dimensional features are usually more extended in one direction than the other. In the present scheme . . and amid denotes a location not too far from this. I will mention tense and aspect only in passing (see Comrie 1975 system .The Spatial Prepositions 301 of the referenceentities.or two-dimensional feature. These different times are representedby the tensesof the verb. times past have vectors of negativelength. Across specifiesthat the vector from the referenceentity to the target intersects the reference line or plane an odd number of times. Along specifiesan even number . a vector of 0 length. In addition . that is.andtheFourthDime 7. It is. Against specifiesthat the entity is in contact with the surfaceof the referenceentity at at least one point . Oppositerestricts the number of intersectionsto one and the intersection angle to 90 . and via and beginning and end of a path that consists of a set of connected places through specify someof the placesalong the way.

of eventsonto the time axis is shown at the bottom The projection of thesesequences of the figure. this leavesthe nonphysical vertical prepositions free to representspecialized relationships betweenentities. third . they can be related to the present time of the speaker / listener or . the state AB CD and . they can be placedin isolation independentlyof any other representation as might occur at the beginning of a story . in the latter casesthe two referencesare confined to orthogonal to the spatial axes that axis and are therefore collinear. In my the fourth dimension of the entities and eventsof the other three dimensions brief summary I will classify them according to whether they use one or more reference points. In instances theselatter . to some other previously identified time. My discussionof the meaningsof the temporal prepositions will be basedon the abstract eventsportrayed in figure 7. The upper event shows a state of affairs in which an entity occupies a vertical location before time A . es(Mary ' s move) and The eventsthemselves are states(dropping down) or process are representedas vectors that must move from left to right (no time reversal ) . These points can be located in three . specify the location. The lower event shows a processof movement 44 and 45 as examples of the over a period of time.11. the beginning and end of the dropping down. respectively process . Second . The punctate events A and B. we will expectthat the senses are also unidirectional . (44) Mary moved from an apartment on the top floor to one on the floor beneath ' (45) Sarah. that is. The general rule seems only spatial prepositions that can operate in the single. Becausethe temporal dimension appearsto be confined to a single axis . The temporal prepositions. the location vector is drawn with the tail at the referencepoint and the head at the located time. Mary s roommate. The . order. from right to left (with a negativemagnitude) if the event occurred prior to the reference point . nonvertical dimension of the line can be borrowed in this way (but seethe specialcasesaround and about) . dropped down to tidy up the new apartment for an hour during the move. For example . then. then jumps to a new location and remains there for a short period AB . Let us use the sentences . and direction within . Because of the temporal prepositions lying parallel to the fourth axis. ways. to be that but not all spatial prepositions can be so employed. First . are marked as points on the time axis. As we shall see . after which it returns to the previous location.John O' Keefe " The choice of the present time as a 0 referencepoint is traditionally called absolute " " " tense while that of a nonpresent referencepoint . and from left to right (with a positive magnitude) if it occurred later than the referencepoint . for further discussion ) . most of the temporal prepositions are similar to (diachronically borrowed from?) their homophonic spatial counterparts. relative tense (seeComrie 1985 the vectors representingtime are all unidimensional.

c """ D ~ -----_ . but other0 reference points adopted three eventsof the top sequence (the dropping down and the presuppositionsof being in and returning to the upstairs apartment. C [) ~ AS TIME .----------. Here I am assuming that all events have some projection in the time domain. respectively . for example .TA . I suggestthat the meaning of the temporal prepositions is as follows.) . .M.Tc.. The representation assumes that the representation events occurred in the past could have been . and is represented vectorAB on the time axis . droppeddown is represented by a physicalmovement endsat time B. --+ The processof moving representedby vector CD has a similar representation on the time line. An event such as " Sarah Temporalprepositions " on the Z -axis that begins at time A. The event CD has a time vector which begins at Tc (noon) and ends at To (2 P. and + BT .M. The usual representationof a processsuch as CD is (46) The move took placefrom noon to 2 P. the difference betweena state and a processresiding in changesin the nontime dimensions . C A B D . Referring to figure 7.11. A process suchas " Mary by " has a similar moved on the time axis . 303 8 -------. but that this can be ignored. . when the length of the event vector is short in comparison to the length of the location vectors..N A ----- ----------- lJ -------~ . Present Figure7.The Spatial Prepositions . . The tail of the secondand head of the third are left indeterminate. --+ T (CD) = To .11 as relationships in a fourth dimension . where D and C are the respectivelocation vectors. are representedon the T-axis by vectors --+ --+ --+ AB . .

. In neither caseis the origin or tail of the vector specified . The length of the vector is given by the preposition for . before Mary finished moving. tomorrow Mary will have beenmoving sincenoon..M. An event that occurs after one time and before another occurs during the interval. It and ( Friday) ( first of April ) therefore be an use to these from the hours of may idiosyncratic distinguish pointlike ' the day (at 5 o clock) on the one hand and the extendedmonths of the year (in May) . (50) Mary has (had) beenmoving sincenoon.-+ setsthe length of vector CD. T. in locate an entity by referenceto a single place on the fourth axis. (51) ?By 2 P. :::::. by the end of the move sets T. During specifiesboth the head and tail of the temporal vector. the temporal useof on seems general on and to dates on the is not usedin any general sense . Sincehas the additional restriction that the temporal reference point acting as the source of the location vectors for the event in question must be later than the event. (49) Sarah visited the new apartment during the move sets Tc < T. whereassince specifiesthe time at which it began. that is. T. This is to account for the acceptability of (50) but not (51) . In suggeststhat there is an point as the maximum of a set of possible places extent of time that is considered as the referent and that contains the entity . (48) Sarah dropped down after Mary beganmoving. a vertical dimension that would place the entity at a location above or alongside of the time point . . before setsit to the first place to the left of that place. however. < To. This is given as the object of from . By fixes the location of the reference . > Tc. In to be restricted to the days of the week . On is somewhatmore difficult . Whereasby and to set the head of the temporal vector at the referenceplace. The simple temporativesat . by.. Until specifiesthe time at which a state or processended.304 John O' Keefe (47) The move lastedfor two hours . the location vectors must have negativemagnitudes . Sinceand until are two temporal prepositions that do not have spatial homologues . To. At operates in the same way as it does in the spatial domain by substituting the place of the referent for the entity . :::::.. Other simple temporal prepositions give the location of the event or duration of the condition by referenceto a time marker that fixes the beginning or end of the time vector. it would seemto introduce the notion of a secondtemporal dimension. About and around also suggesta second dimension. TB < To.

The referent in beyonddenotesthe value that the end of the event before the second . and the head or terminus of the vector . The roof had an icicle on it . The change is encoded in the attributes of the object. The representation of the second type of change . the important distinction betweenpast and beyondin the location of the entity in the orthogonal axis of the spatial domain does not apply. where there is no change in the . As Nadel and I noted (O' Keefe and Nadel 1978 to the event of the sentence ). this time a transformation vector. and and the of the event later than the first time two times locates the start until. also involves a vector. the circumstantial change . In the figure.The Spatial Prepositions 305 As with the spatial prepositions. The icicle was in the garden (after time I) . The second of these relates to the circumstantial mode of Gruber ) and Jackendoff ( 1976 ) . the is read as an and its location between object relationship (53) a. It consistsof a four -dimensional The representationof this is shown in figure 7. Changesin state are representedby a vector drawn from an object at time 1 to itself in the same location at time 1 + I . some of the temporal prepositions require two referencepoints for their meaning. Changesin ( 1976 location of an object are representedby a vector whose tail originates at the object in a place at a particular time and ends at the same object in a different place at a subsequenttime. Becausethe time axis is basically a unidimensional head of an event vector exceeds one. it is possible to incorporate the notion of changes into the semantic map. The middle of the figure shows the translation vector that representsthe event of the sentence . the origin or tail of the vector is the object of the locative preposition from . . These include between .12 structure with time as the fourth dimension. These take two forms: changes in location and changesin state. since . The icicle was on the roof (before time I ) . I have shown two spatial dimensionsand one temporal dimension. and the two prepositions . The left sideof the representationshowsthe unstated presupposition that the icicle was on the roof for some unstated time prior . Between . (53) c. past. appear to be interchangeablein most expressions 7. Both changesare representedby vectors. In both types of change . and the right hand the postsupposition that the icicle continues in the garden for someduration after the event. b. beyond .8 Translation and Transfonnation Vectors Once one has a temporal framework . is the location identified by the locative preposition 10 (52) The icicle fell from the roof to the garden.

1 will explore the metaphorical meaningsof be/ow and beneathas used within the restricted domain of social status. I shall explore the metaphorical usesof the vertical stative prepositions. In the course of this discussion I shall (including social influence ask some of the same questions about these metaphorical uses as I did for their : what are the properties of the spacesrepresented . or changedfrom solid to liquid ) . The vertical movement . Thus each object has associatedwith it a list of attributes. labeled icicle . 7.2 will deal with under . thetwo places vectordrawnbetween between thetwo places at 1is represented by a translation location of the object. The four. In addition to the time axis translationvector . Objects are formed from the collection of inputs that occupy the samelocation in the ' map and that translocate as a bundle (see O Keefe 1994 for a discussion of this Kantian notion of the relationship between objects and spatial frameworks) . Section 7. and so on? Section 7.9.) and in the placelabeled garden at all timesafter 1 (1+ ).a vector representsthe changein one of theseattributes at a time t. but a change in one of the attributes assignedto the object. I hope to show that they apply to two restricted domains: influence ) and social status.12 Figure .9. Figure 7. (54) The icicle melted ( = changedfrom hard to soft at time t . what type of scaleis physical uses used . In a circumstantial change .306 TRANSLATION VECTOR John O' Keefe 4 t + ~ in locationof an objectin the semantic by a mapat a particulartime 1is represented Change .13 shows the map representationof sentence54. is shownon the placelabeled roof at all timesprior to object " " 1 (1. one spatial axis (Z ) is shown " " " " dimensional .9 Metaphorical Usesof Vertical Prepositions In the following sections . whose 7.

. When looking at be/ow and beneathwithin the domain of social status.... .-J ( solid } . and that status vertical scale would appear to be transitive: if A has a higher status than Band B than C. In general . (55) Shewas acting below (beneath (56) Shewas acting under his orders... Shewas acting under her station. but appears to be restricted to the domain of influence or control . 7..- I CICLE long cold liquid) ' ..---- .. The basic mapping systemappearsto be a kinematic one which does not representforce relations. the representation of ideas such as causation .an entity which is vertical to another and in contact with it might exert a gravitational force on it or that an entity inside another might be confined by it .1 Below.- ( 4 ROOF t+ . force and influence in the semanticmap presentsa problem.... This might explain why the prepositions that convey theserelationships... the first thing to notice is that people are ranked or ordered in tenDs of their social status on a .... Beneath ....The Spatial Prepositions 307 VECTOR TRANSFORMATION --------- ------------ICICLE long cold ....13 vector in stateof an objectin the semantic by a transformation maparerepresented Changes in the old propertybeforet and whoseheadendsin the new property whosetail originates after t... The closestone comesin the physical domain is the implicit notions that . and Dow" Contrast the following legitimate and illegitimate metaphorical uses of be/ow and under: ) her station.. Figure7... are used to representinfluence in the metaphorical domain. One person has a higher or lower status than another.. ( 58) Shewas acting below his orders.9..-------.- . semantics is more complex . it follows ... (57) . such as under and in.....

.14 ) seemsto be one in which each status token is confined to a vertical line in the status dimension. but now it is quite in character. The use of be/ow and beneathin this senseis restricted to reflexive status. relationship .15 ). It to be of influence or control . I am ignoring here the possibility that status might be context-specific becauseI do not think this is reflected in the semanticsof the prepositions. one can have a disparity between the value assignedto an individual act and the longer-term status. so that (61) Until recently that remark would have beenbeneathyou.2 Under in the meta the most interesting useof the verticalprepositions Underhasperhaps . but these are free to vary in the other dimensions such that John can move so as to be beneath himself but not beneath Sally.e. A sequence of such actions. although it is not much used (62) Sally was getting above her station. In The domain seems confined to the domain phorical as a CognitiveMap ( 1978 that one of the Hippo ). and thus one could not say ' (64) John acted in a way beneath Sally (Sally s station) . Finally . Nadel and I suggested campus . HereI will pursue theideathat this domains wouldbethat of influence metaphorical " " is represented by an additional vertical dimension (figure7. This gives rise to sentences such as (59) John acted in a manner beneathhim. (60) That remark was below you. The stative preposition down seems to have almost no usein the nonphysical sense The closestone comesare colloquial forms of verbal ranking such as (65) Put him down. the speaker ) . note that there is no vantagepoint (egocentric point ) from which thesejudgments are made or which would change them (i . Now within the vertical scale of status.9. but not * (63) That remark was above you.308 John O' Keefe that A has a higher status than C. ' s status is not relevant . however. His status is below hers. but at the sametime can be compared in the vertical dimension with " " Sally. The antonym of be/ow/ beneathin this context is above .. 7. will result in a status change . Thus the best model (seefigure 7.

TIME ~ o " .15 Fipre Influence of one entity. 309 Sally Status John Tom INFLUENCE :# ' ~ " " ' : ~ ~ ~ ~ ~ :~ :! II . ' 0 " v - - - - - - - ~ ~ ~ : : : : : : : :: : ~ - - - - - . usually an agent. over another entity or an event is representedby a superior location of the first on the vertical influence axis.The Spatial Prepositions CI . ( ~ 7.

To increase or decreasethis influence requires a movement or expansion of one or the other entity in the horizontal plane. outside of . . As the last examplessuggest . (76) John was more under control than Sam. under John s influence b. and ' (69) a. The first meaning of undercannot take a comparative form . under Sally' s control with (67) . (75) He was out from under the control of his boss. ' (78) She slowly extricated Sam from Harry s influence. * (71) He was above. (77) John was more under the influence of Mary than Sam. and has no antonym. (73) Shewas free from stress (74) The car was out of control . the referent in this meaning of underhas an extent in the vertical dimension. But surprisingly the antonym of this under is not over in many examples . * (70) More under the aegisof the King is not transitive. free from the aegisof the King . . In contrast. (72) More under her influence every day. and to be more under a cloud than X has the same senseof a greater overlap in the projection onto (one or more) horizontal dimension as in the physical meaning.310 John O' Keefe There are two homophones (under! and under ). but varies with the direct object. and this may require force in that direction. the secondmeaning follows all the rules for the secondphysical under . which follow different rules and which are derived from the two meaningsin the physical domain: (66) under a widening sky (67) under the table Compare (68) Under the aegisof with (66).

On the face of it . (79) Jane increasedher influence over Harry until shehad complete control .The Spatial Prepositions 311 There are two types of relationships that conform to this pattern. . they . in terms of the scaling of the metaphorical vertical prepositions. * b. Thus one can say: appearto havethe sameinterval scaleas their physicalcounterparts (85) Jane is as far below Mary in status as John is above ' (86) John is lessunder Sam s control than Jim is and it will be easierto extricate John. would be the occurrenceof an event underneath the control of an agent' s influence. it does not seemobvious how they could be reduced to a single dimension becauseone wishes to preserve the possibility of the following types of relationship. ' * (87) John is more under Sam s control than Sam acted below himself. * The King s aegiswas over John. ' (83) a. Notice that the underrelationship is not transitive. what would the Z -axis be? Perhapsthe higher the status. (82) Jane holds sway over John. I wish to remark briefly on the fact that there appear to be two nonphysical vertical dimensionsthat are orthogonal to eachother and to the physical vertical one. control and influence. If this were the case . Note that . on this reading. Finally . The King held his aegisover John. ' (80) Jane s influence over John (81) Jane lords it over John. The antonym of under2is over. Now we come to the most difficult part of the theory: the relationship betweencontrol and causation. and thesevary in the amount of freedom left to the referent object. John can be under Jane' s influence and Jane can be under Joe' s. we cannot compare the Z -axis and the non-Z -axis directly . unlike the three dimensions of physical space . to act below his station in order to maintain control over (84) Jack felt it necessary Jane. Perhaps here oile should consider the possibility that overlapping representations symbolize a control or influence relationship while nonoverlapping ones stand for a status one in the same 2-D space . Causation. the more possibility for control? Finally . but John is not necessarilyunder Joe' s.

The sentence (92) Mary did not causeX is ambiguous. This can be represented in their locationsat the time and showingmomentarysynchronous changes space : of the event. 7. with two possible underlying structures: one in which Mary has influence but the event did not happen.16showsthis ' as a momentary increasein Mary s influence to symbolize an active role in the event. while figure 7. influence is representedby an under relationship between the influencer and the influenced. Figure 7. which differ . If the influence of an agent over another agent or .17 shows a continuing influence but no change to symbolize a passive role in the event. Consider the : closely related sentences (90) Mary made (caused ) the icicle fall from the roof to the garden. (91) Mary let (did not prevent) the icicle fall from the roof to the garden. and the distance between them on the vertical dimension. then it might object can be representedby the location of the first above the second be possibleto representthe influenceof an agent over an event such as that portrayed in (90) an (91) by an action or movement along the influencedimension. theseare five-dimensional sentences According to the present analysis in the control exerted by the agent over the event. (89) John causedthe book to go to the library . This type of representationcan also it can show how capture someof the more subtle featuresof causalinfluence. the complex influencerelationship also allows for the following sentences . For example . and the other in which the event did happen but the causalinfluencewas not exertedby Mary . The lateral overhang between the two representsthe amount of control exerted . On the simplest reading. As we saw in the previous section.10 Causal Relatioll Sin the SemanticMap O'Keefe John Our analysisof the metaphorical useof be/ow and underhas led to the suggestionthat the causal influence of one item in the map over another might be representedby relationships in the fifth dimension. but that Mary s was the by placing Mary at a higher level than John in influence superior one. because influence can selectivelyact on parts of the event as well as on the whole.312 (88) The book went to the library . the sentence (93) Mary made John throw down the icicle ' meansthat both Mary and John had agentiverolesin the event . . the amount of influence exerted. causation is representedas a pulsatile increasein influencecoincident with the physical spatial event.

(96) Mary made John drop the icicle. I have said very little about the way that surface sentences and paragraphs ' could be generatedfrom the static semantic map. /~"~ v ~ ] INFLUENCE . (95) Mary allowed John to drop the icicle.11 Syntactic Structuresin Vector Grammar Thus far . It also permits one to representrelative degrees of influenceover an event in a manner analogous to that over agentsor objects. (94) Mary allowed John to throw down the icicle. ( ~ ~ = - - - - - Figure 7. as in (97) Mary had more influence over the course of eventsthan John. Nadel and I (O Keefe and Nadel TIME ~ o ~ & -------.16 Causal influence is representedby a pulsatile changein the vertical inftuen ~ dimension at the sametime t as the physical event. or the idea that an event of continuing duration can have variable amounts of control at different times. (98) Mary took over control of the event from John on Monday .The Spatial Prepositions 313 THEEVENT MARYCAUSED - - - - - - - . 7.

Figure7. In particular . Economy of expressionis analogous to the optimal solution to the traveling salesmanproblem. The corresponding systemin the semanticmap would comprise the syntactic rules of the grammar. we postulated a system that extracts the distance and direction from the current location to the desired destination... For example . : .- OBJECf ~ ~~ ~ ~ ~ ~ ( ::: . TIME ~ o ~ & v MARY INFLUENCE ...17 in the verticalinfluence dimension Permissive influence is represented of change by theabsence . a large number of narrative strings can be generateddepending on the point of entry and the subsequentroute through the map.314 EVENT THE ALLOWED MARY John O' Keefe OBJECf -. Recall that the cognitive map systemin animals includes a mechanism for reading information from the map as well as for writing information into the map. of theinfluencer during theevent 1978 ) likened this operation to the way in which an infinite number of routes between two placescould be read off a map. This information can be sent to the motor programming circuits of the brain to generate spatial behaviors. The syntactic rules operate on both the categoriesof the deep structures and the direction and order in which they are read.. In an important sense there are no transformation rules for reordering the elementsof sentencesbecause theseare read directly from the deep structure.~ ~ " ~ ~l .... Given a particular semantic map. reading the relationship between an influencer and the object or event influenced determines whether the active or passivevoice will be used.

28. Studiesin lexical relations. Amsterdam: North Holland . Casegrammar theory. J. mental computation P.. L . and Jackendoff. I have deliberately chosenthe tenD entities to refer to the relationships because . R. .. of any language References : Towardsa localistic theory. University Press : An essay in stratificational BennettD . ( 1988 ) . 349. Neural connections . : . MA : MIT Press ' O Keefe. Harnish (Eds. Cook . J. 16. As far as I am aware. Cambridge.150 " " " " Landau. What and where in spatial language and spatial 217 265. Washington. Lexical structuresin syntax and semantics . ( 1976 ) . J. . Toward an explanatory semantic representation . I have relied heavily on the classicdiscussionby Torgerson ( 1958 ). Nadel. Behavioraland Brain Sciences ' O Keefe. I am assumingthe geomagneticsense for spatial coding. Tense . DC : Georgetown University Press Frisk . J. ( 1990 ) . London Longmans semantics . .TheSpatialPrepositions Acknowledgments 315 contributions Cartwright for herextensive I would like to thank Miss Maureen helpand substantive . . ( 1985 ) . P. The grammar of case . Spatial and temporal uses of English prepositions . 7. B. The role of the left hippocampal region in the acquisition and retention of story content.). A . . In what follows . ( 1965 ). In L . R. Cambridge: Cambridge University Press Comrie. Amsterdam: Elsevier. and Milner . Storm. M . 225. ( 1989 ) .359. W . In O. C. ( 1971 ) . Gruber . J. Gruber. Progressin Brain Research . Culicover. ( 1993 ) . V . ( 1976 ) . and so on. B. and R. Computations the hippocampus might perform . 83. but wish to include places . to limit my discussionto objects. cognition . . there is no evidencefor it in the prepositional system . . vol. A .). Massachusetts . B. Cambridge: Cambridge Anderson. 89. Institute of Technology. Cooper. is absentor so weak in humans that it is not available 3. ( 1975 ). PhiD. Jackendoff. features 2. M . ( 1976 ) Aspect .300. The version an earlier on comments made . diss. Cambridge: Cambridge University Press . A computational theory of the hippocampal cognitive map.Mathisen (Eds. OUersen and J. B. experimental to this chapter Neil Burgess that forms the basisfor the cognitivemap model was supportedby the Medical research Councilof Britain. Understandingthe brain through the hippocampus 287. Neuropsychologia .284. Research Notes I do not wish I . Comrie. Linguistic Inquiry .. ( 1990 ) .

B. Oxford : Oxford University Press (Ed. ( 1957 ). 19. J. . J. Description and quantitativeanalysis .435 Neuroscience .208 Tolman . M .793. ( 1981 ) . 83.21. and Milner . 273. and Psychiatry. J. Cognitive . B. R. J. Theory Torgerson ofsca . Paillard ) . Brain and space O' Keefe. L . Cognitive maps. Journal of Neurology. ( 1994 ) . 189 mapsin ratsandmen Psychological andmethods / ing. M . Proceedings . J. and Milner . ( 1958 ). O' Keefe. NewYork: Wiley. Smith. In J. W. S.316 John O' Keefe O' Keefe... and Milner . ( 1991 . L . The hippocampusas a cognitive map. U. time and causality. . 781. C. 10 . 55 . ( 1948 ). Neurosurgery lesions . Press Scoville.. E. Right hippocampal impairment in the recall of spatial Taube .. and Nadel. 27. Oxford : Clarendon . . 20. L . Journalof . Muller.. and Ranck . The role of the right hippo campusin the recall of spatial location. B. Head direction cells recordedfrom the postsubiculum in freelymovingrats.. I . ( 1978 ) . location : Encoding deficitor rapid forgetting ? Neuropsychologia . 11. Neuropsychologia . 71 Smith.81. ( 1989 ) . . B. .. 420 .45. ( 1990 ).295.). Review . Loss of recent memory after bilateral hippocampal . of the British Academy 35. W. B. The hippocampal cognitive map and navigational strategies .

b. learning the language of even the plainest spatial preposition. in Talmy s 1983 terms) and one that servesas the reference 's " " object (Talmy ground object) .we ? In calling each object by its name. This is one kind of representation . . It has some fruit in it . Such descriptions could be captured within a volumetric framework such as that described by modem componential theoriesin which object parts and their spatial relationships are represented (e." " cup.say. If we were to describe this scene . in describing the spatial relationships between or among objects. Linguistically encoded spatial relationships most often representrelationships betweentwo objects." we seemto recruit a different representation .requires that the child come to represent objects in terms of geometrical descriptions that are quite abstract and quite distinct from each other.is critical . we seemto recruit representationsof a quite diffe . Crucially. Lowe 1985 . There is a cup in front of the bowl and a vasenext to it . However.a " container" . " The bowl has some fruit in it . one in which the surface of the object is relevant. suggesting that we are recruiting relatively detailed descriptions of the objects' shapes . d. Binford 1971 . There is a bowl. Biederman 1987 ) . When we say. " The bowl has flowers painted on it .rent sort.1." we recruit a relatively global representation of the object' s shape . When we say. Marr 1982 . in or on. The bowl has flowers painted on it .Chapter 8 Representations of Objects in Languages and Geometric Multiple Language Learners Barbara Landau Central to our understanding of how young children learn to talk about spaceis the question of how they representobjects. Consider the still life arrangement in figure 8. we might say any of the following : ( I ) a. the one that is being located " " ' (the figure object." bowl ." " vase distinguish among three containers that have rather different shapes(and functions). c. in which its status as a volume.g. but no further details are. What are the geometric representationsunderlying thesedifferent spatial descriptions " ..

318 Barbara Landau Figure 8. as a surface in the case of the term on. and as a set of axes in the . representingan object as a volume in the case geometricaldescriptions for example of the term in.one in which the principal axesof the bowl are relevant.1 Each object in this scenecan be representedas a number of different geometric types. if the objects cannot be represented properly . There is a cup infront of the bowl . chapter 3. we recruit yet . These few examples show that learning the meanings of spatial terms requires learning the mapping betweenspatial terms and their corresponding regions where " the relevant regions are defined with referenceto geometrically idealized or sche " matized representationsof objects (Talmy 1983 ) . a different representation " " The region in front of the bowl spreadsout from one of its half axes(and whether these axes are object-centered or environment-centered depends on a variety of factors. learners must possessthese object representations case of in front of and behind the correct before learning mapping. When we say. Therefore a crucial part of learning the mappings is properly representingobjects in terms of their distinct relevant . this volume) . . the terms cannot be learned. In fact. " " but nothing elseis. seeLevelt.

for example . but not the differencebetweena square and a triangle. the first two years of life are devoted to constructing a system of knowledge that can support the general permanenceof objects in the face of continually changing perceptual and motor interactions betweenthe infant and objects in the world (Piaget and Inhelder 1948 . However. Later. or simply blobs ) . That we can talk easily about bowls. sometimeduring later childhood. the developmentof spatial of stagesin which children would knowledge would proceed through a sequence . language The idea that very young children might possess such rich and flexible representations of objects is at odds with traditional theories of spatial development . cups.. According posit changes spatial knowledge to Piaget' s theory. by the fact that twoand three-year-olds draw a variety of geometric figures as simple open versusclosed no specific metric properties. and the kinds of spatial relationships into which they enter suggests that we possess a cognitive systemthat allows for flexible " schematiz " of ing objects (cf.: Geometric Representationsof Objects ultlplf 319 The brief analysisjust given suggeststhat there is a variety of object representations . which substantial in over the first six years of life. the child s knowledge of spaceis still incomplete. Although even infants might be capable of discriminating betweenobjects having different metric properties (e. possessing develop. relationships such as . a triangle). developed objects. such impoverished representations were evidenced . projective properties would figures. such as the straight line. they must also be representedat a . Talmy 1983 ) . Piaget proposed that the child possessing a topological representation of spacewould only be capable of representingthe difference betweena line and a closed loop.g. Once such knowledge has " " of the child is said to possess true representations . the early acquisition of spatial terms among children suggeststhat thesemultiple representationsof objects may exist early in life and may be used to guide the learning of spatial . Piaget hypothesizedthat from around age two .highly general properties such as first representonly top logical properties of space connectedness and openness versusclosedness .simply as a set of axes " " is quite coarse (as volumes.the languageof objects and . Piaget 1954 ) . a square vs. metric properties such as angles and distances would come to be represented even later. and vases . ' Extending Piaget s view to the realm of spatial relationships. or a relationship specifiedby location along such a line. Similarly. Objects must underlying spatial language places be representedat a fairly detailed level of shape .representations ' that go beyond perception. surfaces . but could not support the representationof a distinction between contact with a vertical versusa horizontal surface . For Piaget.and they must be representedat a level that skeletal level. Central to the present discussion . a topological representation could support understanding of a contact or attachment relationship between two objects.

of degree shape from children learning other evidence of young children learning English. which eliminate all details of shapeexcept the relative length and orientations of the three principal " " axes . by Talmy 1983 spatial terms appear to strip away details of shape(as suggested -based representations of objects are also accessibleto young fine-grained. shape children." The studies also indicate that . Next I present evidenceshowing that young children learning English show strong blasesto ignore fine-grained shapewhen learning novel spatial terms or when interpreting known English spatial terms. The evidenceindicates that both coarse and axial representations ' of objects can be elicited by engaging children s knowledge of known and novel spatial terms (in English. would be unable to representobjects in terms of their axesin order to learn such basic . emergingduring late childhood. but that they show equally strong blases . In this chapter I review evidenceshowing that such nontoplogical representations . horizontal surface ). Further . I focus on three different of different parts support acquisition " : ( I ) " coarse kinds of representations . the cat " " " " or mat in the sentence The cat is on the mat ). Pa~ icular emphasis differ from those relevant to similar spatial terms in other languages will be placed on comparing English to other languages whose locational terms appear to incorporate much more shape information than those in English. which preserve a considerable . and would be unable to learn the distinction spatial terms as in front of or behind betweenGerman auf and an (attachment to vertical vs. The axial representations in particular illustrate that young children naturally representobjects in terms of skeletal ' descriptions in which the object s principal axes are the major components of . Theserepresentationstend to emergewhen children are engagedin learning . For example cally would be incapable of using precise object shape for naming bowls or cups. spatial prepositions) . and (3) fine-grained representations is primarily based on studies I will describe The evidence detail. are indeed accessibleto young children learning the language of space it appears that young children possess multiple representationsof objects that can of the spatial lexicon. and how theseobject descriptions .320 Barbara Landau that encoded by the terms in front of or behind would require at least projective . although languagesis consistent. representationsof space While topological properties might seemcongenial to the analysis of spatialloca tional terms (Talmy 1983 ). bloblike representations of objects. a variety of evidencesuggeststhat a topological representation ' of objects and relationships is too weak to characterize young children s . (2) axial representations . although the representationsunderlying its " shape " " ). object names In the following sections .. I first outline how objects are representedwhen they are " " encodedby noun phrase argumentsof spatial prepositions in English (e.g. the child who was limited to representingobjects topologi knowledge. which " " eliminate all details of shape information .

Other languagescontain as few as one generallocational marker (e. a kind of static spatial relationship: example stand representsthe vertical posture of an object.1. Most of theseprepositions are two-place predicates . ta in Tzeltal. languages 8. becausespatial prepositions in English encode location only . they provide a well-defined domain within which to intensively examine the kinds of spatial relationships that languages encode . verbs such as stand and recline ) and to locational terms in other 2 . some languages by spatial collapse several English distinctions into broader categories(e. among . Spanish en covers English in and on).g. one can compare these meanings to those encoded by other spatial terms in English (e.places expression by prepositional phrasesheaded by spatial prepositions. German auf and an cover English on but distinguish betweenvertical . the vase is the referenceobject. This empirical evidence will raise a number of questions which I will outline . as long as they are somewhere . where placesare canonically encodedthis way. Levinson 1992 ). With that knowledge. the relevant region is actually any portion of the surface of the vase : The sentencewill be true regardlessof where in particular the flowers are located. spatial locations.g. etc. for example . and the spatial preposition " on" maps a region of spaceonto the referenceobject. In a simple sentencesuch " " " " " as " The flowers are on the vase . while others split a single English distinction into several finer categories(e.Multiple Geometric Representationsof Objects 321 to attend to fine-grained shape when learning novel object names . For . and there is variability in the preciserelationships that are encoded in other languages terms : Considering prepositions only . the flowers play the role of figure.g. adjectives such as long and wide. although there are some with a greater number of arguments . including issues of possible structures and mechanisms underlying this gross difference in object ...are encoded canonically ? In English. representation 8.! contiguous with the surfaceof the vase Note that spatial prepositions do not exhaust the possibilities for talking about spatial location . . However..g. there exist verbs that describeposture.1. nouns such as top and bottom. even in English. Although the upper surface of an object may be the preferred reading for on in English..1 English Spatial Prepositio18 The spatial prepositions in English form a relatively small closed class numbering somewhere aboveeighty (not consideringcompounds suchas right next to) . A sample list is given in table 8.1 Waysof Representing in Places Objects How are objects representedwhen they serveas figure or referenceobject in a locational . amidst. crouch and kneel other postures . recline representshorizontal posture .

. for example Johnson. respectively in and how to be universals however there this . However .322 Barbara Landau Temporal only during Intransitivies here there upward downward inward since outward afterwards ) upstairs downstairs sideways until ago south east west left right backwards ) away apart together north and horizontal attachment. Phrasessuch as in the bowl or in the house are easily understood becausebowls and housesare easily construed as volumes. As one example . Miller and .Laird 1976 . the preposition in requires a referenceobject that can be construed as having an interior : If one object is in another. the abstract nature of thesegeometric descriptions can be seenthrough other ' cases . Korean ahn and sok cover English in but " " " " " " ). in which the preposition will coerceone s reading of the referenceobject. distinguish between loose and deep or tight containment. Jackendoff 1983 ) . respectively . Talmy 1983 . Theseuniversalscan be revealedby considering the geometric that are encoded restrictions imposed by a spatial term on its arguments (see . the latter must have some volume or area within " " " " which the object can be located. figure appear variability Despite referenceobjects are geometrically schematizedand in the kinds of spatial relationships . For . Herskovits 1986 .

A ball bounced along the road. (2) a. however. Given coercion by the verb. object principal orthogonal 8. b. with the figure often representedas a relatively blob . A snake lay along the road. whereassentence it is difficult to construe a ball (2c) is marginal because as a " linear" object. a liquid ). dirt in the mat" ). that sentence(2d) is completely natural . although the term in seems to expressstraightforward " containment" (with the referenceobject " somesort of container" ). Terms such as in.any object of any shape . c. we can interpret a sentencesuch as " John drank marbles." where marbles are taken as a continuous stream (cf. he proposed an asymmetry in the geometric descriptions of figure and referenceobject. below . Thus. respectively relationships appear require figure object that can be construed as a " linear" object. b) both are easily understood. There do exist.3 Thus sentences (2a. or " customersin a line" a virtual line .g. d. For example .. size . the prepositions listed in table 8. Specifically . Note . the verb eat requires an argument construable as an edible (hopefully. " This processof " schematizing objects has been describedby Talmy ( 1983 ) in his seminal work on the geometry of figure and reference object where he suggested strong universal constraints on the geometric properties relevant to the figure and referenceobject. and the referenceobject represented more richly . ?A ball lay along the road. in a phrase such as in the dot or in the mat the dot or mat will be example " construedas a 2-D area or evena 3-D volume (e. the comparable imposed by verb to drink requires an argument construable as a continuous quantity (centrally. food). while throughout express es . Such ( ) semanticallymotivated restrictions appear to restrictions verbs on their arguments . One further distinction mentioned by Talmy is the figure object' s distribution in space : through is used for nondistributed objects. and many others do not impose any specialgeometrical requirementson the figure object. we can useit equally well for " coffee in a cup" (where the reference object is a physical container).2 Geometryof the Figure Object Taking examples from English. in this ' case . on. the ball s path (as it bounces ) becomesthe figure. " birds in a tree" (a virtual volume). or type can play the role without violating the meanings of the majority of prepositions. a few restrictions for certain terms. however.Geometric of Objects Multiple Representations 323 " " " " . Terms such as acrossand along representrelationshipsof intersectionand parallelism . * " John drank a marble" ) . and so forth . and these to a and reference .1. above . Trees stood along the road. often in terms of the shapeless ' s three axes .1 show very few constraints on the figure object.

the opennessof the ground ). . and what role they play in . the orientation of the ground is distinguished (German aufvs . shiny. there is only one basic spatial preposition in Tzeltal (ta. a surface as a volume (in. " " " limp . lumpish object to move/ be located.. and for ) between for two still other terms. For other terms.g. English makes similar distinctions in certain verbs (e. As another example " figure object distinctions in locational verbs. and direction toward or away (Korean has two separateterms for English through hin others. from the speaker( ).g. including roots meaning small. the referenceobject tends to be representedfairly coarsely. At this point . slimy. icky material to move/ be located" (Talmy 1985 ) .ta of flexible bulging bags objects whose opening is larger than their height. to rain. linear object suspendedby one end to move/ be located.in which a greater amount of geometric information is incorporated These examples into the figure object. " " " spherical object to move/ be located. pachal. to spit). say.1. Thus waxal. chepel es a considerable number of . although other languages have locational verbsthat do impose shape restrictions on the figure object. the number of referenceobjects is distinguished ( referenceobjects. It remains to be determined than that shown . For example .are challenging becausethey raise the question of whether there are universal blasesin the kinds of information typically incorporated . Tzeltallocational predicates . aspect ratio . there do not appear to be any other requirements on the geometry of the figure object for spatial prepositions in English. amongor amid for more than two ) . an).3 Geometryof the Reference Object Like the figure. it is representedas a shapeless point or blob (e. Nor do I know of any in the spatial prepositions of other languages . according to Talmy . inside or as the referenceobject is represented (on). terms such as near or at do not require that any specificgeometricinformation be preserved ) . Atsugewi possess (Brown 1993 ) . a general relational marker). among . In other languages . it should be noted that into the figure object in locational expressions the degree of shape information exhibited in .324 Barbara Landau distribution of the object in the ground (compare " There were raisins throughout " the pudding" to " ' !fhere were raisins through the pudding ) .ta of . and runny . is however . exactly how fine-grained theseshapedescriptors are. the overall systemof spatial language 8. Levinson 1992 ) .ta predicatesused for locating objects (seeBrown 1993 is predicated of objects whose opening is smaller than their height. German her vs. or the ratio of height to width ) can appear as part of different spatial . but information about an object' s axial structure (specifically. although this particular pattern of conflation is not dominant in English. Aside from these few distinctions.. For certain terms. by English prepositions greater .

The rough shape parameters required for such . At least one languagepossess setsof terms to refer to the object-centeredversusenvironment-centeredapplication of theseterms. people do appear to have blasesto interpret theseterms with regard to different referencesystemsunder different conditions ( Levelt. and then decideson markednessusing detailed shapeinformation a clear (e. which specifiesan object s principal dependent on the object axial system dimensions . " at the head of the table. These axes are also engaged by certain spatial nouns and adjectives in English: top/ bottom. however.g." the foot of the bed. in front of/ behind) . but also for different ends of the axes(top/ bottom. English also has such expressions(e." and so on ( Levinson 1992 ) . but the Tzeltal systemis richer in its range of locational terms. Carlson. a number of spatial prepositions require that the reference . " Thesespatial terms appear to be insensitiveto referencesystem . front / back. such as " at the " " head of . and side expressregions defined by referenceto the axes /..Multiple Geometric Representationsof Objects 325 Most critically . front / back. the Tzeltal body-part systemis massively ' . and another to refer to (environmentally determined) regions adjacent to the object (Levinson 1992 ). . then that would be marked " head. . Levinson suggests that the assignmentof body-part terms dependson a object axes strict object-centeredalgorithmic assignmentthat analyzesthe object into its principal and secondaryaxes . bottom) . shape assignmentprovide a challengeto the generalization that ground objects are stripped of detailed shape elements . Each of theseterms. long. For example . right/left. and the two sets of horizontal axes (right/left or beside .g. a novel object might possess " head of " and " foot of " axis for which would be relevant if but one end principal .Radvansky and Irwin 1993 es different ) . The vertical axis object be construed in terms of its three principal axes (above / below )." " at the noseof .. For example spatial terms of other languages . of the axis has a distinct protrusion . however." " the arm of the chair " ). even though there is still quite a broad range of shape variation sufficient for assigning" nose of " to an object part . The axial representationsas a whole appear to be the richest geometric representations required by English spatial prepositions. and Tai The spatial nouns are marked not only for different axes . thin. the endsof which are often labeledwith locational terms. at the butt of . they also playa major role in the . dependson very much the samekind of analysis into principal . The Tzeltal body-part systemutilizes one set of terms to refer to object parts. and wide expresssize differencesalong different axes . The " star is above the flagpole can be used to describe a location with respect to an object-centered framework (the region near the top of thejlagpole. for top vs." or perhaps " nose " consistent with its . However. chapter 3. regardlessof its orientation ) or an environment-centeredframework (the region adjacent to the gravitational top). with the viewpoint varying application of the latter being quite difficult to learn) . this volume. For example .

as axial. However. the referenceobject. The basic vocabulary for object names in English includes proper nouns (e. These geometric descriptions appear quite different from those which might be engagedduring object naming. incorporating distinctions such as volume. Piaget. To the extent that these terms are linked with schemesfor object recognition.2. languages such as Tzeltal appear to include more shape information . English appearsto incorporate the least information in figure objects. this system has also been objects in English and in other languages posited to be developmentally complex. grouping ' together objects by the relative proportions of the object s principal dimensions(e.. as fine grained? The following empirical evidence representobjects ascoarse . evenTzeltal incorporates relatively little shapeinformation . when compared with the much richer information available to identify objects. disregarding almost all shapespecification of the figure object. a tree) . and most critically . 8.2. principal axes (of either the figure object. As for the referenceobject. Mother ) and count nouns (a dog.1. pachal vs. a number of investigators have ) representations development nonlinguistic ( proposed that the spatial prepositions recruiting axial representationsmay be relatively difficult to learn (seeJohnston 1985for review). Inhelder. they would seemto require geometric representationsthat preservemuch more fine-grained spatial . number. ) suggested there appears to be an asymmetry between the figure and reference object. it engagesan axial representation of the referenceobject in order to also recruit the axial representation describethe relevant region.326 Barbara Landau The axial system thus appears to be critical to the representation of reference .. but not much more. Basedon this proposal for the axial of . At the other end of the dimension. with children coming to representprojective geometric properties such as straight lines only during middle childhood (Piaget and Inhelder 1948 . Other languages . information than the ones so far described How do young children appear to represent objects (both figure and reference ) when learning spatial terms. If we consider the degreeof geometric specificationto be a dimension. Interestingly. waxal ) . or both ) . and how do they representthe sameobjects when learning ? Can they ? Are young children flexible in their representations object names . I return to this issuein section 8. for example . surface . English again incorporates very little shapeinformation . evidence for eachof thesetypes of representationin young learners positive provides .4 Summary The geometriesof both figure and referenceobject are relatively coarse . Fred. with the figure incorporating relatively less geometric specification than the reference object. apparently.g. and Szeminska1960 ) . As Talmy ( 1983 .g. at the most.

in Tzeltal spatial predicates 8. attend only to the object s principal axis) . then one might wonder whether English-speaking children could readily learn to incorporate the somewhat more detailed (axial) information captured.2 for objects) . These studies have shown that children can ignore shape information altogether and that they can treat objects in terms of their axial representations .2. These studies have shown that relatively fine -grained shape information can be used to assign objects to named categories . As the object was placed.1 CoarseRepresentatiol B : Scbematizingthe Figure Object . independent of its location . for example . The entire display then was set aside . One is whether young children learning a novel English spatial preposition ' will tend to ignore shape entirely (or perhaps . If the answer to this question is positive.1. One of the objects was identical to the standard. we have conducted a variety of studies examining children ' s treatment of objects when learning novel spatial prepositions and when comprehending familiar prepositions .2. the figure object is generally treated quite coarsely shapeless point or blob .Multiple Geometric Representationsof Objects 327 8. Recall also that other focusing object principal languagesmay incorporate somewhatmore detailed shapeinformation into the figure object. . 's on the axis. subjects heard. 8. In addition .2 EmpiricalEvidence for DifferentKindsof ObjectRepresentation among Young Learners In order to determine whether young learners possess the different kinds of object representation underlying figure and reference object . Two questionsarise. or (for terms such as along and across ) as a linear object. Spatial predicates in Tzeltal include terms that incorporate information about the figure' s aspect ratio (height-to-width proportions). and subjects saw each of three different objects being placed in each of five different locations on and around a secondbox.1 Ignoring the Shapeof the Figure Object Landau and Stecker ( 1990 ) posed the first question by modeling a novel spatial preposition for young English-speaking children and then asking to what new figure objects and locations children would generalize this term. Three-year-olds and adults were shown a novel object " " (the standard ) being placed on the top of a box in the front right -hand comer (the " standard" location: see " figure 8. for example. and the other two were different from it in shapeonly (see figure 8. and curvature. flexibility . See " this? This is acorp my box. using the novel term acorp in a syntactic and morphological context compatible with interpretation as a novel preposition. we have conducted a separate line of investigation to determine how children treat objects when they are learning a novel name for the object .2). Each time subjectsviewedan object being placed on the second .either as a Recall that in English.

also rejecting all locations that were off the box.might itself predisposesubjectsto ignore object . Figure 8. in the standard location front hand corner ization to any object ) right only (top One might wonder whether the context of the experiment. children then generalizedto all locations on the top of the box. they were asked either " Is this acorp the box?" or " Is this a " ' corp? Subjectshearing the novel preposition ignored the object s shapeand generalizedon the basis of its location. Would they generalizeonly to the standard in its standard location? Or would they generalizethe term in a way consistent with the general pattern of English spatial prepositions. and generalizingto a range of locations? In this condition . Is this acorp your box? The question was how children would generalizethe meaning of the novel term. although they were somewhatmore conservativethan the children. acceptingall three objectsequally (summedover locations) . they were asked . They heard either " Seethis? This is " " " acorp my box (novel preposition) or Seethis? This is a corp (novel count noun) . they did attend " to the object' s location. rejecting all locations off the box. Adults showed a similar pattern. Then they were shown the three different objects each being placed one at a time on and around the box in different locations.2 ) . Someof them confined their general 4 . Children and adults were shown a Objects and layout used by Landau and Stecker ( 1990 novel object being placed on the top of a box. generalizedon the basisof its shape u " " box. However. We found evidenceagainst this interpretation in a secondexperimentalcondishape .in which objects are being placed in various locations. as shown.ignoring the particular shapeof the standard object. Having beentold that the object was " acorp the box (when placed on the top front right -hand corner of the box).328 Barbara Landau . both children and adults ignored the shape of the standard. Subjectshearing the novel count noun ignored the object' s location and . Each time.

This pattern of findings shows that young children are capable of representingthe -grained level. they observed a test object being placed in one of the locations. and the test objects included a replica of the standard. and attended to the object itself.for example . and a 2" x 2" x ] " block. However. many of the three-year-olds tested behavedjust as they had in the first experiment . with one critical . we have modeled novel spatial prepositions using figure objects experiments that possess a very clear principal axis. regardlessof its location . but each time . The standard object was now a 7-inch straight rod ." the standard ' object was placed perpendicular to the box s main axis. some Tzeltal roughly linear figure object as does across terms appear to require even more shapeinformation . ignoring object shapeand generalizingsolely on the basisof location.2 Incorporating Axiallnformatio D into the Figure Object One experimentwas . The results of this experiment again showed that subjectstended to ignore shape and generalize primarily on the basis of the object' s demonstrated location. . But they do figure object at a very coarse not show that children are incapable of incorporating any elements of the figure ' object s shape when learning a new spatial term. In this condition . This time. while subjects hearing a novel preposition (" acorp the box" ) ignored shape and attended to location .of Objects Multiple Geometric Representations 329 tion . And . Thus one might ask. along requires a require attention to the figure object s principal axis. except that different objects and locations were exactly like the one just described used( Landau and Stecker] 990.3 for standard object and standard location ) . In both . In fact. seefigure 8. Test locations included this samelocation as well as one slightly to the left of it . and one diagonal to it . completely ignoring shape . subjectsnow generalizedonly to the standard object. As subjectsheard. How readily will young children incorporate shapeinformation into the figure object? We have approached this question through two sets of . we told subjects " " Seethis? This is a a the same as for the novel . a wavy rod of the same extent as the standard.2. Even in English. exception. one parallel to the box' s principal axis. 8. as the standard was being placed on the box. subjects hearing a novel count noun (" a corp" ) ignored location . rejecting both of the objects that were not identical to the standard. as mentioned above. Subjectsthen were shown the same test objects placed in the same test locations as in the first condition . certain terms ' . That is. they were asked " With this " Is this a ? context as a mental to a count corp syntactic serving pointer noun reading. phonological sequence corp using preposition (acorp). The question is whether such modeling might more strongly elicit at least an axial representationof the figure object. we followed the sameproceduresas above. " Seethis? This is acorp my box.1. but placing the new word in a syntactic and morphological context appropriate to a count noun interpretation .

" When asked . Thesesubjects the pattern shown by subjectsin a secondcondition of this experirnent were shown the sarneobjects and locations. That is. accepting objects that were s~ ciently long to intersect the box (when placed perpendicular to its rnain axis) . they showedsorneattention to an abstract cornponent of shape . although sornedid attend to the axis. F~ 8. In doing so. the basis of its location and its detailedshape and on . This contrasts rnarkedly with .2. but heard the novel term in the count " " noun context. Thesesubjectsalso tended they disregardedthe details of their very different shapes to generalizeto the two locations in which the test object was at perpendicular intersection with the box. younger subjects (three-year-olds) still tended to cornpletely ignore detailed . Older children (five-year-olds) and adults shape . the horizontal and diagonal locations were considerably less favored (seenote 4) . thenovelcountnoungeneralized on the basis of the object hearing shape sornethree-year-olds and rnost five-year-olds and adults acceptedboth the standard and the wavy object while rejecting the block.3 Objectsand layout usedby Landauand Stecker( 1990 ) in a second study using the same ' in figure8. Is this a corp? . whereas . " See this? This is a corp . that is. All subjects tended to attend to one skeletal cornponent of shape in this preposition condition also attended to location. Subjects methodasdescribed hearingthe novelpreposition ignoredthe objects axis . a rnore salient principal Thus. principal Subjects generalized ' s exact .the principal axis. these subjectstreated the two objects as sirnilar with respectto their principal axis.330 Barbara Landau ~ ~ . when we rnodeledwith a standard object possessing axis.

4 shows displays appropriate to the two terms waxal. lechel example " " is used for wide flat objects lying flat (Brown 1993 ). Figure 8. when learning a novel preposition. oblong-shaped bottle on the " ' top right hand corner of a box. might talk about locating objects.4) . regardlessof location. flat objects (seeright column. we told subjects . but rather. In a relatively new approach to this issue . children geometry learning Tzeltal must learn the range of application of thesetwo terms. ignoring its location. The locative ta is a relational marker and the predicates waxal and lechel each is used when locating a particular geometric figure type. Given that theseterms are found in a natural language . as well as quite a number of others that encodedifferent geometric distinctions. we have been modeling novel spatial terms using figure objects whose shape properties are representedin Tzeltal spatial . we conducted an experiment quite similar to the studiesof novel spatial prepositions describedabove (Landau and Hseih in progress ). All subjects then saw a series of eight objects being placed in various locations on or around the box.t disk in the samelocation on " ' " a box. See this? I m putting this lechel my box (bottom left . Thus the dissociation betweenshapeand location that we had found in the first set of experimentswas replicateo with entirely new objects and locations.4) .4) . The object on its box was then moved aside . to schematizeit in terms of its principal axis. For a secondgroup of three-year-olds and adults. Waxal is used for vertically oriented objects. For one group of three-year-olds and adults. See this? I m putting this waxal my " box (seetop left . Our question. " figure 8. fla. This illustrates once more that children' s responses were not forced by salience(or lack thereof ) of either object shapeor location.Multiple Geometric Representationsof Objects 331 subjectsnow generalizedthe novel count noun to objects of exactly the same shape as the standard. Half of the objects were tall . As each test object was placed in its location . we said. was not whether the terms are learnable. figure 8. at best. and a second . identical box was placed in front of the subject. In order to answer this question. oblong-shaped objects.ta and predicates lechel ta. As we placed a tall . for " " . However. As we placed a wide. therefore. how difficult it would be for English speakersto infer such meaningsfrom a relevant modeling situation. figure 8. and half of them were wide. Tzeltal. we modeled the meaning of lechel . Both children and adults were capableof generalizing on the basis of the object shown. a tall oblong-shapedcontainer or solid object canonically standing . What . the conflation of specific with location must be All learnable. each of which describes the location of an object. we modeled the meaning of waxal. they tend to ignore the figure object' s shape . subjectswere asked. or . We then modeled two different locational situations. We introduced the experiment by telling subjects that we were interested in how . and people speakinga different language that we would use some words that Tzeltal speakersmight use .

332 Barbara Landau .

Locations on the top of the box were acceptedmore frequently than those off the box. there was an interaction between the modeling condition subjectsobservedand the test objects to which they generalized . " This is waxal the box" ) generalized more often to other vertical test objects. then we should expectthem to generalizeto a compound of shape and position . subjectsmight ignore the object' s overall shape .) Alternatively . Of the twenty adults tested . previous studies The overall pattern of resultswas consistentwith previous findings. if " lechel. Only one subject Figure 8. and were asked whether each was waxal/ lechel the box. Subjectsshown the vertical object (upper left ) were told ." they should generalize to all vertical objects in the relevant location . Generalization to novel positions was consistent with previous results. (And this region might be the top surface of the box. and nine more generalized to all objects located on the top surface of the box. Subjects shown the horizontal object (lower left ) were told . Is this waxal (lechel ) the box? If subjectsattended to the overall shape(verticality or horizontality ) of the figure object as well as its location . this effect was small. to all in located the relevant as had in done the generalizing objects region. nine generalizedto all objects located in the standard position . Adults entirely ignored the vertical/ horizontal aspect of the objectswhereasthree-year-olds tended to generalizeon the basisof the object' s principal axis." then to all horizontal objects in that location. If they had heard " waxal. " " they were asked again. Subjects who saw the vertical standard (and heard. However. with children showing an overall tendency to say yes to novel object/position combinations more frequently than adults. sometimesin combination with its location. and there was no reliable interaction reflecting differential effectsof the standard in both object shapeand position . Examination of the individual responsepatterns shows that few subjectsactually generalizedon the compoundbasisof object shapeand position . Subjectstended to generalize the novel term to new locations and to new objects. All subjects then were shown four vertical and four horizontal objects (right column) being placed on or around the box. " I ' m putting this waxal my box" (using the Tzeltal spatial predicate for tall oblong objects " sitting canonically" ) . Subjects were shown either a vertical object or a horizontal object being placed on the top of a box.Multiple Geometric Representationsof Objects 333 " about now? Am I putting this waxal (lechel ) the box? After the object was placed. subjects . and adults tended to be more conservative than children. while subjects who saw the horizontal standard (" This is lechel the box" ) generalizedmore often to other horizontal test objects. as it had been in the previous studies. " I ' m putting this lechelmy box" (the Tzeltal predicate for flat objects lying on a surface ) . . as shown in the left column. Most crucial to the design of the experiment.4 Objects and layout used by Landau and Hseih ( in progress ) . saying yes to the standard position and no to the position off the box more frequently than children .

. four accepted the standard object (vertical or horizontal ) in the standard position . : SchematizingReference 8. the linguistic contrast between such parametersmight readily serveto partition the geometric spaceso as to respect ' Tzeltal contrasts the verticality or horizontality of the object s axis. and so forth . Why should children have been more likely to conflate axial information and location in this study when they had shown strong blasesin the other studies to ignore axial information ? At this point . Removing from consideration the children who said yes to all queries left seventeenchildren. the general pattern is intriguing . while only 2 of 20 adults had consideredthe object s axis at all relevant to " " the novel spatial term. they might lead children to partition the geometric object descriptions in a different way from those invited by the partitioning of the object space in English. seven children accepted either vertical or horizontal objects (but not both ). three children accepted all objects on the top of the ' box. but it is possiblethat the contrast between vertical and horizontal objects in this experiment led to relatively strong weighting of this object property . If so. flexible objects (pachal).2. This overall pattern is quite different from that shown by the threeyear-olds. Thus. and three acceptedeither vertical or horizontal objects that were on the top surfaceof the ' box. English-speakingadults appear to be firmer in their conviction that object shapesimply should not be conflated with position for novel spatial terms.1.that is. and this subject said yes to only the standard object in its standard position . 14 of 17 children (who did not show a yes bias) did so. while three children did so. because include vertical objects (waxal ). For example . while the contrast betweentwo long and one short object (all of which were horizontal ) in the previous study could have diminished attention to the axis. we do not know . While these results are only suggestive . Of these . Not a single adult had actually generalizedon the basis of the compound axis-plusposition . he did not generalizebeyond the modeled context. That even a small number of young English-speaking children are willing to conflate vertical/ horizontal axis together with location suggeststhat the learning processis not over by age three. flat objects (lechel). languagestend to incorporate a greater . In this ' tended to conflate the direction of the object s axis but not adults children . study with position .2 Axial Representadorm Object That young English speakingchildren resist incorporating axial information into the figure object raises the question of whether they show similar limitations for the referenceobject.334 Barbara Landau respondedin terms of both shapeand position . this would suggestthat the parametersof the contrast set used in such studies might lead to different conjecturesabout which object dimensions are important . and fourteen respondedon the basis of the standard object s axis. Of the latter . In real languagelearning. As describedin section 8.

The empirical results from acquisition studies have indeed suggestedthat these . reported in Jackendoff. Tversky. this comparesto terms such as in or on. establishingthe relevant regions for such terms requiresconsiderablestructure . this volume. According to ). here it would seemcritical to acknowledgethe role of spatial representationinconstructing . and object itself (seeNarissiman 1993 Levinson 1993for some rules of application) . Moreover. which appear much earlier and do . for example .g. chapters 3. Johnston and Slobin 1978 ). Inhelder. for discussion of the com' ) . and establish regions centeredon thesevirtual axes In fact. the use of reference be difficult (compared to mastering changing systemsmight quite in or on which do not such see Levelt and . It is not obvious terms appear later than other terms not requiring axial representation . Identifying sucha region and mapping it to its respectiveterm might orthogonal axes seemsimple. form meaning transparency (e. object. and right/left representregions surrounding . with the particular region defined in terms of the object' s three principal . and in particular . 12. The observer can derive the three axes . They could be due to more data-driven causessuch as morphological complexity.that is. respectively . The axial representations that must be recognizing it (Marr 1982 extendedoutward from the object are not directly given in the stimulus either. the representation of axes does not emergeuntil Piaget (Piaget and Inhelder 1948 well into middle childhood.. and Szeminska1960 ) . Both of theselimitations would ' impose serious restrictions on the child s ability to learn terms requiring representation of the object axis. extend them outward from the . chapter I . and 13.Multiple Geometric Representationsof Objects 335 degreeof geometric detail in the referencethan the figure object. terms such as in front of/ behind . A prominent view of this difference not appear to undergo much developmentalchange in acquisition time is that object axes are difficult to represent . Consistent with this view is Piaget s argument plexities of referencesystem usage that representation of the straight line is not achieved until middle childhood . representationsand rules to ensure that on the part of the observer the correct axes are found and that they are extended in a linear fashion from the . . . theseextendedaxes Theserepresented axesmight be difficult for the learner to construct. although many theories of visual object representation suggestthat ' ' . a number of studies have shown that terms such as in front of and behind are not completely mastereduntil around age four or evenlater. in addition . hence to recovering an object s axes is critical to reconstructing an object s shape . the difference between in back of and behind) or even . The object axesare not given directly in the stimulus. terms that require extension of the axis outward into space(see . In English. however. above / below an object. that this is the result of a representational problem in the child. this volume. Piaget . using engage systems Logan and Sadier. Leyton 1992 ) . and that sensitivity to viewpoint differencesis not complete until this time (Piaget and Inhelder 1948 .

theseenrors tend to cluster around cardinal points. comparison across . in and on are ranked among the 20 most frequent words.neverthelessappear . in progress a clear principal axis. it possessed no principal axis. Perhaps . balls.who have not completely mastered in front of/ behind to possessrepresentationscongenial to the mature understanding of these terms. In English. while behindis ranked 450th (Francis and Kucera 1982 ). axial representations seeKucjaz and Maratsos 1975 . In a separateset of experiments . from a geometric analysisof their shape are not clearly accessible In the experiment. whether this axial representation permitted extension of the object' s axes to the larger region surrounding the reference -based ) properties of the . a second reference of its proportions and symmetry. and five-year-olds and adults one of three different referenceobjects placed flat and directly in front of them on a table . studies have shown that very young children two more to the separate point . we asked whether certain structural (shape object. etc. nonlinguistic task in which they were place objects " " to place dolls and toy animals on a table such that they could either talk to each other or follow each other in a parade. Even the youngest children tended to orient the toys properly . and a third reference object was round and therefore possessed " " " " . then there still may be conditions that can be accessed under which this accessis impeded. Levine and Carey ( 1982 ) gave two -year-olds a linguistic task in which they were to " in front of " another and a .. Tanz 1980 ) .. One referenceobject was Ushaped . for example . seefigure 8. tiallanguage These observations motivated us to investigate in detail the nature of young chil dren' s representationsunderlying the single spatial relationship encoded in English " " by in front of . This again suggests endpoints of the objects principal axes be accessed for learning spawhich can of axial representations objects may possess . Tanz in front of one children make errors showed that when placing object ) young ( 1980 " " or behind another. A number of studies have found that young children are especially poor at assigning fronts or backs to objects that themselveshave no intrinsic orientation (e. but was marked with eyes and a tail (simple object was identical to the second ) . we showed two -. that is. trees . Subjects were tested on one of these reference objects. we asked whether young children possessan axial representation of objects that could support learning of theseaxis-based spatial terms. for objects whose principal axes . First . suggestingthat they recognizedthe fronts and backs of the objects and knew how to align them with each other. Second reference object might more readily invite an axial interpretation .5) . and because (Landau. and critically . and might therefore induce better performance than the round object. three-. the ' that very young children .336 Barbara Landau input frequency. We had two principal questions. Theselatter properties might induce assignment piecesof fabric glued to the surface of a principal axis. If children do possess for spatial term learning.g.

Each of the figure objects were placed one at a time in each of the four cardinal locations plus a fifth . the object was placed in (up to 16 ) additional locations in the region fanning out from the side of the object facing the subjects Figure 8. Following ten such trials . 21) . Locations were probed in a particular order. . Each time the small object was placed in a location.and five-year-olds and adults performed in a yes / no task in which they were shown a range of small novel objects (the figure) placed in a variety of locations around the referenceobject and were asked to judge whether the figure was " in front of " the referenceobject.6) . 19 regions surrounding the axis (locations 12. 18 . Objects were presentedin the horizontal plane in front of subjects .6. 8. . the ' region extending directly from the object s principal axis (locations 10. Variation eyes and tail possessed in these properties affected the nature of young children' s and adults' judgments of the " " region in front of each. subjectswere assigneda placement task in which they were given a seriesof four small objects with no distinguishing features . 7. the round object possessed no such axis. " " . The U-shapedobject possessed a clear principal axis. separated into regions that ' correspond to (A ) the broad rectangular region following from the object s principal axis and ( B) the broad triangular region surrounding this. 9). directly on top and in the center of the referenceobject (seefigure 8. among youngest Three. Referenceobjects varied in how clear an axis they exhibited.6 shows all 21 locations. and the round object with " " " " cuesto indicate the probable location of the principal axis. .5 Three referenceobjects usedin study of the structure of regions. the . the entire yes Following / no procedure. referenceobjects would determine whether cues to the location of the principal axis " " (in the case of the Ushaped or eyes objects ) might induce better performance the children. Is this (figure) in front of this (referenceobject)? (indicating subjectswere asked each object at the relevant moment) . as indicated in figure 8. in order to ensure obtaining responses for critical areas such as the region closest to the object (locations 6. . 13 ) and the regions farther away from the referenceobject (locations 14 15 16 17 20 . Subjectswere asked to place objects " in front of " each referenceobject and to judge what locations were acceptableinstancesof being " in front of " each.Multiple Geometric Representationsof Objects 33 Figure 8. II ).

The locations " " shadedwithin the block representthe proposed canonical region for the term in front of and adults. The locations shadedwithin olds and olds five three were widely acceptedby . year year " " the triangular area surrounding that block (the external region) tended to be lesspreferred " " by children and adults. in numerical order. of acceptanceby adults (seefigure 8.6 Layout of locations probed in regions task.8 for comparison of canonical and external regions .338 Barbara Landau Figure 8. each time twice. followed by one query each on locations 6 20. Subjectsfirst were querled on locations 1 5. which elicited a high proportion ). except for the eyes referenceobject.

Athough there is some diffusenessin the responsesof the two-year-olds. that is. The secondmost common pattern occursat both agestwo and three. and leastprominent among adults judging locations around the " eyes object. The relatively high acceptanceof locations 2 and 4 among three-year-olds judging the round object suggeststhat lack of a clear object axis invited subjects to entertain more than one axis as the relevant one for determining when one object was . this is shown in the figure as subjects lined up to each other. Note that both children and adults do vary somewhat in their preferred location for this initial placement . and finds children locating objects at the end of the side / side axis. This pattern of accepting both 1 and 3 (in " " " " English. Even someadults consideredthe far end of the object to be their first choice for locating the figure in front of the referenceobject (a pattern that is the preferred one in Hausa. This pattern was most prominent among three-year-olds judging locations around the round referenceobject " reference . At age two change appears most children place the object at one of the cardinal locations.Multiple Geometric Representationsof Objects 339 and were asked to place each " in front of " the reference object. Individual dots representdifferent subjects placementof the first object they were given. the point at which its front -back or side-side axis would have projected into the space surrounding the object. one group for each referenceobject. Trailing behind 1 and 3 as the subjects choices were locations 2 and 4 the locations falling at the ends of the side/ side axis. For easeof representation .7 for each of the reference ' objects.. this disappearsby the age of three. Adding eyesappearedto drive subjects of all ages to locate the figure object directly along the half axis extending outward from the eyes . the canonical locations for in front and behind ) occurred almost always with the Ushaped and the round reference objects. This variability occurred only for the Ushaped and round referenceobjects. The round object with eyes " elicited " yes responses only to 1.e. This is consistentwith previous findings suggesting that young children more often correctly place objects in front of or behind objects with clear fronts and backs (Kuczaj and Maratsos 1975 ). The results for the placementtask are shown in figure 8. especially favoring both endsof the object' s front / back axis. but the major developmental to occur between the ages of two and three. seeHill 1975 ) .the locations falling at the two ends of the front / back axis. When subjectswere asked to judge whether a small object was in front of the reference object. The adjacent pattern becomesstronger over age . were also assignedthis placementtask. in front of ) ' second the eyes . the location directly adjacent to (i. they tended to say yes to locations 1 and 3. however. The most frequent responseacrossall ageswas to place the object in line with the referenceobject at one of its cardinal points. The results of the yes / no method tell a similar story about the cardinal locations. Three separate groups of two -year-olds.

. . 6 I I . Adults I 8 0 81 8 I I .. 0 : RF6TTFiAL ~ T TASK PLAC () . . I .: .1 1 . . .340 Barbara Landau 218 [il ] . . . . 315 . . .[ijJ a . I 51s I . .

object eyes objects Although the regions appear to be similar for the Ushaped object and the round object at all ages . each locationarisingfrom extension of axes on the object .7 ' first " Can in response to thequestion Subjects placements you put X in front of the(reference " at all ages showed a preference for oneof the four cardinallocations )? Subjects object . The " external" region representsthe triangular region extending outward from the front edgeof the referenceobject and surrounding the canonical region. Adults accept a greater proportion of the locations as of " in front of " than do children. Figure 8. but is due to expansionaround theprincipal axis rather than by a random increasein the acceptanceof locations or by expansion outward from the front edgeof the object. Overdevelopment " " . subjects at all ages and for all referenceobjects accept the canonical regions more frequently than the external regions.Multiple Geometric Representationsof Objects 341 in front of another. Finally ." Second . it becomeslarger by expanding outward from the object' s extended(virtual ) axis. Three noteworthy trends appear in figure 8. the strong axis-based responsesfor the Ushaped and " " eyes referenceobjects suggestthat young children are quite capableof representing an object' s axis. This suggeststhat even the youngest subjects represent the region directly adjacent to the referenceobject as the preferred region for the spatial relationship encoded by " in front of .8 showsthe proportions of " yes to the two different regions. . First .8. ' The third significant developmentconcernsthe subjects treatment of the reference with relative to the other . the region for in front of does not becomelarger by seepingoutward from ' the object s edge ." Three-year-olds show the samepattern for eyesas they do for the other two eyes reference . That is. The causeof this differencecannot legitimate cases be ascertainedfrom the figure. their relatively poor performancewith the round object suggests that objects lacking a clear principal axis may be lesseffective in allowing young subjects to show their knowledge. responses " " The canonical region representsthose locations falling within the rectangular region extendingdirectly outward from the front edgeof the referenceobject (seefigure 8. the analysis of the regions surrounding the cardinal points showed that subjectsof all agesgeneralizedtheir interpretation of " in front of " to regions with a well-defined geometry that was basedon extension outward of the object' s principal " axis. rather. the pattern differs for " . mentallyimposed .6) . younger children tend to accept positions directly along the axis while adults accept the entire canonical block. although their preferencefor the canonical over the external regions objects is slightly more pronounced for eyes . That is. Five-year-olds show an overall dampening Figure8. there appears to be growth in the size of the canonical region over age.

locations these of for details preferred 8. : .8 " " " )? to the question" Is this is front of (the reference object Proportionsof yes responses " " " " reference each object surrounding arewithin the canonical and external regions Locations canonicalloca the at all . 1 2 402 1 46 1 - Canonical External s U Adult s Figure8.6 ages see ) Subjects ( figure " " . There . 5'8 F 0 . exceptfor adultsjudging the eyes reference 5 . they object tions to the externalones .342 Regions Rarhara Landau [1] 3'5 0 . 1 .

and II in figure8.Multiple Geometric Representationsof Objects 343 " of " yes . Theseregions seemremarkably similar to those describedby and Tversky by Logan and Sadier (chapters 12 and 13 . this volume) among adults in and attention tasks see also participating imagery ( Hayward and Tarr 1994 ) . asif theyobjectwasmobile fanningoutwards object ' -year -olds and couldexplorethe environment . There may be a of wherein to themselves not only an object' s region reactivity perceiversrepresent region of influence when static. This basic pattern 8. The geometric structure of the referenceobject itself has some effect during the early years . both for the canonical and external region. the development of regions appears to begin with the axis. Whether or not such regions are also part of young children' s representationsis an intriguing question.8 (continued Figure ) all locations from the reference accepted . By the age of three. A number of subjectswere quite explicit on this issue . The idea that the geometry of a region in front of an animate object might be different from that in front of an inanimate one is reminiscentof Michotte ' s ( 1963 ) notion . remarking that the object could " look over this " way (indicating a location in the external region) . the children' s regions appear to be more narrowly defined with respectto the object' s axis. the foregoing studies strongly suggestthat young children can and do representthe principal axesof referenceobjects by the age of two. I 0. . theseaxial representationscan be extendedoutward from the object and can serveas organizing referenceframes for setting up regions relevant to basic spatial terms such as in front of or behind .6). and broaden with development . Thus. but by and large. contrary to the pattern ' predicted by Piaget s theory. This appears to be due to their ' that the relevant is affected assumption region by the object s status as an animate. young children appear to be capableof setting up object axesevenin cases where the perceptual clues to the location of the axesare weak. where a number of subject insist that the only acceptablelocations are those falling along the extensionof the axis. these effects seem to be imposed on a basic pattern in which any referenceobject can be representedin terms of its axes and surrounding regions. Five in acceptance of the canonical depression " to from their reluctance to say" yes those regionstems anylocations except alongtheextended axisitself(locationsI . To summarize . adults show an overall increasein the size of responses " reference the external region for the " eyes object. but also the region from which it potentially could influence another. Although the geometric and conceptual nature of the referenceobject may modulate the geometric details of the relevant region. This conservatismcausesan overall " reduction in " yes . To the extent that they differ from those of adults. Finally . Inspection of individual responses responsesby the five-year-olds suggeststhat the critical change is in the canonical region.

naming has beendemonstratedin a variety of other kinds of studies In many of these studies . the scenewas describedusing a novel term in a syntactic and morphological context suitable to a preposition. size weighting of shape . property that is engagedwhen young children are learning object names A variety of evidenceover the past twenty years has hinted at this pronounced ' . there is strong evidencethat children are not limited to these descriptions ..1. for example . and texture in generalizing the novel count noun to new . In this context.3 Fine-Grained Representatio ' focus of this the chapter so far has been the young child s ability to sche Although bloblike or coarse axis based either skeletal matize objects in terms of ) geometric ) ( ( descriptions. for example " or a dog " kitty -cat. including shape A number of developmental studies have shown that children find it easy to learn -basedcategories(Bomstein 1985 namesfor shape ). subjects are shown a series of " " additional novel objects and asked for each . the novel term on the basis of the object s shape scenewas described using a novel term in a syntactic context suitable to a count noun.344 Barbara Landau appears to exist quite early in life and is mapped onto the corresponding spatial terms betweenthe agesof two and three. Attention to object shapeduring object as the modeled one. in another condition . both children and adults generalizedthe novel word only to objectsof the sameshape of its location. as. However. . In one condition . and hear it labeled. ). In this context. subjectsare shown the novel object. Recall that in some of the experiments described in section 8. In another context. Is this a dax? In another version of this study. Then. and five-year-olds) and " adults are shown a novel object and they hear it labeled. neither children nor adults generalizedthe ' . Rosch et al. organized in terms of certain key properties.2. as if the object itself (and not its location ) was being named.are . Which one is a dax? The results from these two methods tend to convergeand suggestthat object shapeis a privileged perceptual . ( 1976 ) argued that our basic level categories setsof objects named by count nouns in responseto the question " What is that?" . with the novel object still in view. young children (two -. three. and Jones( 1988 was . . However. . regardless . but then they are " " shown pairs of objects and are asked . rather objects. who compared children s reported by naming .2. Clark ( 1973 role of object shape ) reported that children s early overgeneralizations " " tended to be based on shape . when a child calls the moon ball . children and adults were shown an object being placed in a location on a box. these but can also representobjects in terms of rather detailed shape in the namesof children are to when tend learning engaged emerge representations than their locations. Seethis? This " is a dax. the role of object shapein the developmentof object A systematicattempt to asscss ' Landau Smith. : SchematizingObject Kinds 8.

For example . " Is this a dax?" ) Both children and adults tended to weight shapemore strongly than either sizeor texture. Landau. Smith.Multiple Geometric Representationsof Objects 345 objects. they were asked whether the test object was also a member of the named category (e. and quite strong among five-year-olds and adults (see . In the basic experiment. c Shape Changes NO :? r!lJ l != Sponge Figure 8. size .9) . or texture. Each time. they tend to generalize the object' s name to others of similar shape of sizeor texture.g. . Wire Chicken Cloth Standard " (2 wooden) U U " ' " " " Changes Texture ~ - 1 1 . ! 1 ) 1988 . for example . a developmentalpattern emerged : The " shape " bias appears to be weak among two-year-olds. when told a novel object was " a dax. t [ .. even if they were much larger than the original or a quite different surface texture (seefigure 8. regardless and Jones( 1988 ). Smith. children were shown a novel object and heard it labeled. moderate among three-year-olds.9 When children and adults are shown a novel object and hear it labeled. Bias Jones and . After Landau. 2 The ( Landau nges YES : ~ . then they were shown objects varying from the standard in either shape ." subjects then generalized the word dax to other objects having the sameshapeas the original object. Shape Smith . possessed In this study and severalfollow -ups.

For example . but tended to generalize on the basis of surface texture when instructed with a novel adjective (" This is a daxy one" . e. its shape . When an object is being named. an input bias may act in ( g. the shape bias appears most robustly in the object-naming context. The particular pattern of context dependencesuggeststhat the bias is closely linked to the representation of objects and. see also Smith. the more willing he or she is to accept objects of different shape . The bias appears to begin around the time when children have fifty words in their vocabulary. adults reject even a small change in shape from the but acceptan object of the sameshapethat is ten times as large.in this case children bring to the language learning task certain a priori categories . The younger original the child . Carey. often to substance guided children s attention away from shape -basedrepresentations What are the geometric properties of theseshape ? One striking fact is their level of detail. suggestingthat the bias may becomesharper as children learn more about which properties are the best basis for generalization ( Joneset al.and five-year-olds using count nouns that guided attention to shapeand massnouns (" This is somedax" ) that ' . ' concert with the children s own representational blases . compared to those representationsrecruited forlocating -based representation appears objects. and Jones 1988 ) . As another example have learned a bit more syntax can be guided by syntactic and morphological information to attend to properties other than object shape . object and substance whose existencemight constrain the range and type of inferences . For example. Sola. other objects sharing the sameshape . and Spelke young children show different preferences ( 1991 ) found that two -year-olds showed a strong shape preferencewhen shown a rigid object. This suggested . . Recentstudies indicate that the growth in the shape bias is correlated with children' s productive vocabulary. while in other contexts. Smith.6 . in turn . although by the age of about two . Subrahmanyam ( 1993 ) showed similar effectsamong three. Jones . . the preference for Although the shape bias emergesquite early in development sameshapeis highly context-dependentamong both children and adults. That is. Same hencesuch a generalization would in generalbe safe. and Jones ( 1992 ) showedthat somethree-year-olds and most five-year-olds tended to generalize on the basis of object shape when instructed with a count noun (" This is a dax" ). Smith. By the age of two . and Spelkethat young (seealso Subrahmanyam 1993 ) . object naming. the children may sharpen their conjectures as they learn that words for objects may safely be generalized to -shapeobjects are often of the samekind . Carey. Landau.346 Landau Barbara and Jones 1992 ) . and Landau 1992 ). but a very weak shapepreferencewhen shown a massof gooey substance to Sola. the basic level terms common in maternal input). children who they can project from a single exemplar. children show a reliable tendency to generalizeon the basis of sameshapein the naming context (Landau. 1992 ). because many words do indeed refer to objects sharing the sameshape .

children and adults who were shown a straight rod and heard it named tended to reject a roughly linear object of the samelinear extent. which weakenedthe object s .ultlple Geometric Representationsof Objects 347 to preservea good deal of detailed infonnation about its part structure and arrangement (while such elementsseemto have only limited relevancefor locating objects) .1. but whosepart structure and suggested rigidity was altered. one might expectthat . ( 1992 ) sought to detennine whether children would make different predictions about the range of acceptableshape transfonnations relevant to a named object. in the studies described above. and three successivelylarger size changes . this object did not qualify as a member of the same named category. In the studies described in section 8. . the parts capture specific arrangment of parts will be subject to some variability or range.10 ) . except that it possessed curved edges . In a first experiment. Although little is known about the range of object-internal motions that must be captured by theories of shape . both children and adults extended the novel label to objects of different sizesfrom the standard. Several young children would respect both in their generalization of object names recent studies suggestthey do. The range and degreeof detail necessaryto include objects into the same named category is as yet unknown. As in most of the studies on object shape and object naming. especially as it interacts with imputed malleability . and it is the arrangement of these that would seem to what we call an object' s " shape . and then were asked which of a set of shapeand sizetransfonnations were also membersof the namedcategory. Three-year-olds and adults were shown a novel line-drawn object. For example ." Further . They were testedon three successively .2. subjectsin one experiment saw an object identical to the standard of the ' first experiment. adults who were shown aU -shaped object tended to reject objects with evena small defonnation in overall shapedcaused by bending one of the object parts slightly outward. heard it named. as many objects possess parts that regularly undergo motion . subjects were shown standard objects comparable to thosefrom the first st:udy . However. In one seriesof studies . shape In subsequentexperiments . ) Because of the importance of both part articulation and part motion in the characterization of object shapeand in theoriesof object recognition. but did not extend the label to objects of different . Marr example and Vaina 1982 . Biedennan 1987 ). Leyton 1992 . subjectswere shown a rigid -looking object with sharply delineated part boundaries more extremeshapechanges (figure 8. many modem theories of object recognition propose a strong role for object parts as components of object shape(Hoffman and Richards 1984 . Landau et al. there do exist models for characterizing limited classes of motions (see for . apparently becauseits overall contour was wavy (as compared to the straight contours of the modeled object) . For example . depending on its part structure.

subjects tended to accept weakened . eitherby curvingedges wereadded . 1992 ). was tended to even small . When angularobjectswere shownand named affectedby subtledetailsof object shape . Wheneyes accepted quitesubstantial (middletwo panels changes that theyobjectcaneasilybedeformed . subjects shape ). aspart structure (top"panel ). changes .10 ' Children and adults judgmentsof which objectsbelongin the samenamedcategoryare . asif theynow assume (Landauet al. However reject shape changes subjects moreshape or adding" wrinkle.348 Barbara Landau Standard No No Yes Yes No Yes Yes Figure8.

In particular. However. And in a fourth " experiment. progressiveweakening of the part boundaries (and correlated destruction of cues to rigidity ) led them to generalizeto shape changesas well. Although subjectshad generalizedsolely to sizechangesin the first experiment. In one of these experiments . subjectsbecamemore and more willing to accept a larger range of destroyed " -shape . changes they now assumed that the object was a tubelike. In another experiment. further suggestingnonrigidity . These studies begin to suggestthat the spatial system underlying object naming . subjectsgeneralizedto all shape shapechanges as if . This suggests that a bias for " same shapechanges objects must engageobject representationsthat admit of flexibility in the face of varying rigidity and changing part structure (seealso Becker and Ward 1991 ). we have beeninvestigating children' s inferences about the kinds of shape changesthat might obtain under mechanical transformations . The results of the four experiments showed massiveeffects of part structure and suggested rigidity . One group of subjectswas shown each of the three standard objects. In a different seriesof experiments . (Objects undergoing rotation now had hinges at their joints ) . Thus. subjects also saw one part of eachobject undergoing a small motion .10 ) . Children and adults who saw the standard object with no motion generalized to few shape . we showed children a novel object that was of distinct composed parts arranged in a particular configuration (seefigure 8. This is a dax ). their spatial relationships. eachof which was consistent with a more extensiverange of object part motion . and then was shown a set of new objects whoseconfiguration would obtain if the standard object' s parts were capableof motion . a of saw an object identical to that of the second separategroup subjects .Geometric Representationsof Objects Multiple 349 part boundaries and suggestedmalleability (figure 8. except that it possessed massively textured edges . it must incorporate a incorporates information about object shape relatively detailed (possibly hierarchical) representationof object shape . as rigidity and strong part boundaries were successively capable .. in which the ' object s parts. Another group of subjects was shown objects of the same configuration. which were possible motion -based shapechangesof the standard. and their ranges of relative motion are . When eyeswere added to the objects.11 ). a cue to animacy) would affect subjects judgments -changedobjects could still be membersof the named of which shape category. nonrigid object of internal motion .g. those who saw the standard undergoing a small amount of part changes motion generalizedquite freely to the novel configurations. All subjects were then shown the same set of test objects. heard it labeled " " (e. This last type of object was meant to test whether certain powerful ' cues to object kind (in this case . but this time. Theserepresentationscould provide a powerful systemthat would allow the present . Both the curved and curved/ wrinkled objects led subjects to accept a moderate number of . different subjectssaw a curved and " wrinkled " object with " eyes placed at one end.

11 Children and adults are sensitiveto the range of shapetransfonnations likely when an object has pennanently fixed versus moving parts. Subjects who viewed an object with no motion (left) and heard it named then tended to reject objects with even small changesin configuration as an instance of the named category. The detailed representations of object shapethat seemto be engagedwhen young children are learning object .3 Some Uenge Croatinguistie that Proves Rule different kinds that young children possess The empirical evidencereviewedsuggests are engaged children of object representations .350 Barbara Landau No Motion Motion Figure 8. each of which is selectively active when in tasks involving different parts of the vocabulary. subjectswho viewed an object undergoing that would be the product of more extensive a small motion then accepted shape . Theserepresentations seemquite different from those engagedwhen young children are learning ' that describean object s location. In contrast. terms or using : TzeltalmaybeException Cha Notesanda Possible 8. young child to link up an object name with its shapeand to generalizeto classesof transformed shapesconsistentwith certain principles of object constancy. changes motions.

this volume).g. covering both .have begun to that children suggests very young form spatial linguistic categoriesconsistentwith those found in their native language . as does Italian . Considering the variation in how locations are over languages . Continuing with the examples provided by Bowerman (chapter 10 language . Suchcross linguistic differencespoint to a strong role for early learning. In other cases . language . one should be suspiciousof conclusionsbasedonly on one expressed . children learning Korean . A number of languagescollapse the distinction between locational and directional terms that is made in English by in versus into . Choi and Bowerman 1991 ) have found that children learning Korean are likely to " " " " distinctions between respect tight fit and loose fit contact and containment relations ignored by children learning English spatial prepositions (though of course English-speaking children must respectsuch distinctions when they learn adjectives such as tight and looseor verbs such as to fit ) . Moreover evidence on the structure and acquisition of other languages . For example . as in The plane flew over the house other ).g. German distinguishes two types " " of contact (on) relationships by the orientation of the reference object (auf for gravitational contact. and English all differ somewhat in the range of object types that are included in basic spatial notions expressedin English " " " " by the action of putting in compared to putting on. Dutch distinguishes between various types of attachment all covered by the English preposition on. but obencan be locational only) . usually horizontal . For example .. the two languagessplit (e. " and that the geometriesof both figure and referenceobject are considerably " sparser (in terms of shapedetail) than the representationsof thesesameobjects as members . English over can be locational . Choi and Bowerman (Bowerman 1991 and chapter 10.. and an for nongravitational contact. for example . but they do not invalidate the search for the universals that underlie the expression of spatial . or line. such as attachment to a wall ) . blob .well before the age of two years. Dutch . as in The plane was over " " " whereas the house or directional . But one might object that all of the experimental evidencereviewed thus far has concernedchildren speaking English. that the reference object tends to be representedas those or as a set of orthogonal axes . it is English that collapseslocational and directional " meanings (e.7 Yet none of thesedifferencesseemto provide major counterexamplesto the claim that the figure object tends to be representedas a point . Other languagesmake yet further distinctions that are not found in English. as noted earlier. v ( vo before certain consonant clusters ). Russian has a single term.Geometric Representationsof Objects Multiple 351 names do not appear to be engagedwhen children are learning words for object locations. this volume. German " ber can be either directional or locational. Korean distinguishes between " " " " degreeof fit and among actions involving putting on different types of clothing .

where the Queen sits) . by a strictly . there are predicates its closed " and others that describe" horizontal " that apparently describe bulging bags . whereas " the term for " mouth would be used to locate at an object part with an edge or " orifice. To take an example ' the table is a region at the end of the table s principal axis. Tzeltal has therefore been offered as a counterexample things. the Tzeltal to particular body-part systemusesspecificelementsof shapeto assignparticular terms " would be used " nose for the term Levinson to locations. So much is actually quite similar to " " " " " " " " English. For example . the term for head is usedto describethe tops of objects and the term " " for bottom is used for the bottoms of objects. but rather. which usesanimal to the names and human body part terms to assign spatial regions of objects. The evidencecomes primarily from the Tzeltal body-part system . The shapedistinctions do not appear to be part of the spatial meaning of the term.8 For example . Could this be a universal. Levinson claims that much more so than in English. To put it another way. including locational terms. nose or tail poisedon someobject (though they would if they were usedas nominals) . " " For example . but rather. what used to region of the object it maps onto ) is separatefrom the geometric algorithms " head" of the from English. Rather. which end is usually decided by a variety of criteria (e. This language has often been described by investigators as particularly " visual" in that it appears to encode a large range of shape distinctions in -class items. as first suggested ? 1983 ) ( Some recent evidence from Tzeltal might appear to provide counterexamplesto such a universal. Just becausethe term " " headis used to name the region does not mean that each and every head of a table . We often refer to the head or foot of the table. However. and the term for " tail would be used for long thin protrusions.. front . the body-part terms do not appear to refer to the distinctive shapesof . The meaning of the term (i. and that this therefore erodes the shapedistinction between objects-as-named and objects-as-located? I believenot. when usedas locatives. assign the term to the object. lying ( Brown 1993 to the notion that very little shapeinformation is encoded in the figure and ground for purely locational terms (Levinson 1992 ). Does this mean that fine-grained shapeinformation is part of the spatial meaning of the locational term.. and side. regions that are encoded in English by terms top. say.g. or the arm or leg of a chair. are distinctions used ' to identify particular regions relevant for the term s meaning. According to Levinson ( 1992 The caseof Tzeltal seems ).352 Barbara Landau by Talmy of categories . the terms refer to spatial regions whose locations are defined with respectto some salient geometric property . back. according to locate something at an object part with a particularly sharp protrusion . bottom. named by nouns. . sitting " ) . Tzeltal assigns analogous most body-part terms not by a metaphorical extension .e. must be similar to a real head in shape .

suchproperties are in generallikely to Leyton . English presumably languages a striking counterexampleto the general claim that the specificsof shapeare absent from the figure or referenceobject. protrusion .4 Structure : Some . and objects in particular named categorieshappen to share greater overlap in shape than they do with objects in different categories( Rosch et al. the axis is found by using properties such as elongation. Such examplesmight be found . but they do possess shapethemselves that demands encoding in terms of three principal axes . 1992 remain robust over a variety of viewing conditions (e. As a (perhapsnecessary ) bonus.properties that are likely to be universally important in such assignments(seeJackendoff. and symmetry. As it stands from Tzeltal doesnot suggestthat spatial terms map onto specificcomponent shapes . More Questio What causesthe differencesin geometric representationbetweenfigure and reference object on the one hand.g. this volume. possibly suggestinga restricted set of sarily lead to modification of the hypothesis . explanation suggestthemselves One possibility is that this difference reflectsnothing particularly interesting about either language or spatial cognition. blurring . andMechanism Possibilities . and this would neces . may provide a particularly compelling exampleof how vast apparent cross linguistic differencesmay ultimately rest on deep similarities in how languagemaps onto spatial representation . but rather. in some languages .. Function . spatial . and objects as named on the other hand? Severalkinds of . one should not to find terms that to expect any spatial correspond spatial relationships holding between specific volumetric components or specific arrangements of components . ) . Rather.Geometric : Representationsof Objects Multiple 353 geometric algorithm that analyzes the object into its major components and their relative orientations. Perhaps object shape doesnot matter to location because locations do not demand suchinformation . is a direct consequence of how the world is. Objects in the world actually do come in an astounding variety of shapes . 1976 ). 8. This is as true of Tzeltal as it is of region (usually dependent on the object s axes and of all natural . rather than providing . it suggests that using spatial terms requires being able to locate the relevant ' ). What kinds of counterexampleswould disconfirm the hypothesis that both figure and referenceobject do not contain any particular shapeinformation necessaryfor describing the region? As suggested by Landau and Jackendoff ( 1993 ). chapter I . Thus the location of the region " head of " an object is defined with respectto the object' s principal axis. flatness . the evidence shapeproperties that is relevant to spatial terms. rapid exposure ). . Thus Tzeltal. Locations in the world do not possess a three-dimensionalstructure . thereby supporting the assignment of axes and directions to objects during a wide 9 variety of spatial tasks that humans normally perform . however. of course.

it is not a foregone conclusion that all organisms will encode objects and spatial relationships in just this way. achieve we must world and on the tasks How .354 Barbara Landau . Mishkin to such features(seeSchneider 1960 . but these systemsare not direct reflections of some objective system " the world out there. one specializedfor the task of object identification ( what ) and " " the other for object localization ( where ) . and Ungerleider and two streamsof Converging evidencefrom human psychophysicalstudies suggests " " is specialized . One parallel concerns the separation between object and place in language . far enough not to . A variety of evidencesuggests " " and in humans. Thesecortical areascontain neurons with quite different receptive field properties. and their nonlinguistic underlying the language of objects and places counterparts? Is the structure of spatial language driven at all by the structure of spatial cognition? There appear to be severalintriguing parallels betweenspatial languageand spatial cognition that suggestpossible relationships. Why not encodeobjects in terms of ? Why not encodelocation in terms of generalproximity relative size . then. on the one hand." More of plausible is the possibility that our representational description in have evolved responseto constraints on both the physical systems . on the other neurological and cognitive studies of the what and where systems fuller discussion 1993 for Jackendoff Landau and 1982 see Mishkin and . Those in the inferior temporal lobe have a large receptive field falling within the fovea and are driven by complex sets of features . without regard to the to oneself ? Given that there are different possibilities for how we represent objects three axes and places . this possibility seemswrongheaded that locations come specifiedmetrically true and also in . experiments on monkeys have shown selectivedeficits in the two tasks. objects come many shapes in three dimensions . whereasdamage to the posterior parietal cortex disrupts various localization tasks (but not object recognition) . The bifurcation a similar reflect that system parvo may processing . but (basically) ignore it when locating those same ? The structure of the world surely imposes some constraints on our representational objects . those in the posterior fovea and are insensitive include the does not field that a lobe have receptive parietal 1982for review) . do we repre ~ent the world? The study of spatial language can tell us how we representthe world linguistically . but does this have any bearing on how we representthe world nonlinguistically? Are there any communalities betweenthe representations . the question is. Damage to the inferior temporal cortex appearsto disrupt object identification (but not object localization). (Ungerleider the existenceof two systemsin monkeys of this parallel) . and that found through " " " " . For example. Although it is certainly true that As stated. rather than shape .things closeenough to reach. What gives rise to the particular way in which we do ? Why do we attend these aspectsof space for the purposes or language represent to shape when naming objects.

the fact that object example as one pressurein the design of spatial language " " . certain named locations must be supported by quite specific and detailed perceptual representations (e. Recently. depth. for example The low-passfilter preservedonly coarse information about the scene . whereasthe magno systemis insensitiveto color but is special ized for properties relevant to localization. Warach. whereas shape and color (but not location) are representedin the what" system " object location (but not shape or color ) is representedin the where system is reminiscent of the distinctions uncovered by linguistic analysis and documented through experimentation among young children. Lincoln Center )..motion . Objects must be assembledfrom parts (and this requires assignment of relative location). tion . the city) and a high passfilter of the other " " . this parallel betweenspatial languageand spatial cognition will undoubtedly undergo " " " " . For . (say. of one scene (say. but seeVan Essen ) . revision as we learn more about the coordination of the what and where systems For example. ( on the one hand. While intruiging . for example the conditions . ? Landau and evidence relevant to the structure of spatial language is this Why these of different the that 1993 systemsmight serve Jackendoff ( properties ) suggested . Similarly . In different . Anderson and Felleman 1992for evidencethat and Hube11989 . and vice versa ( Fara et al. hybrids were createdfrom a low passfilter highway scene the scene highway) . the lack of locational information in object namesmay be due to the lack of such information in the systemsunderlying object recognition. is that there are different for the tasks of object identification and object location . A recent study by Schynsand Oliva ( 1994 illustrates how this might occur. and spatial (locational) language Coslett 1994 ). functional consequences differences functional and that these give rise to differencesin the kinds of properties ) most readily processedin the two tasks.g. not inconsistent with the first . and Farah 1985 color and between naming functional for a object evidencehas appeared separation Saffran Breedin other the on . It is possiblethat the relative lack of shape information in locational terms across languagesis due to the lack of shape information in the cognitive and neurological systemsunderlying object location. Subjects were shown a target scenefollowed by a mask and a rapidly presentedimage that was a hybrid of two of different kinds of . (Shiffrar 1994 A secondintriguing parallel. " " " " Dodger Stadium.Multiple Geometric Representationsof Objects 355 " " for color and shape . Levine. Human clinical the systemsare coordinated at relatively early stagesof processing evidenceindicates that object recognition functions can be spared without localiza) . and perception of certain kinds of motion (a " " where systemproblem) may be constrained by the specificsof object identification ). a variety of evidencepoints to the necessityof coordinating information at levels likely to precedelinguistic encoding. and . and location (Living stone . a combination of a city sceneand a scenes(each a possible target). 1988 .

the reversehybrid would contain the overall geometry of the with the fine details of city buildings. while the other operateslater by extracting the finer infonnation . Schyns (personal communication) -grained infonnation is indeed processedmore rapidly than comments that if coarse " fine grained. sequential operation would allow the perceiver to extract infonnation about general geometric composition first . this is most likely becauseobject shape is an excellent predictor of other properties held in common by members of many object " kinds" . this would be most beneficialwhen the scene and the perceiver had to categorize it quickly . Thus one city highway hybrid might contain the overall geometry of the city with the fine details of the highway vehicles . followed by focused attention to the details of an identified scene was unknown . It is a fact that the human visual systemcan distinguish among an enonnous variety of object shapes . The high-pass filter preservedfine-grained infonnation . While both might be used to identify scenes . One scheme operates earlier by extracting only coarse infonnation about scenegeometry. they identify pass infonnation . Consider object shape and object name. Any learning device that beginswith some broad set of distinctions is likely to converge on a solution more quickly than an unconstrained device. it is highly likely that universal predispositions in object representation interact with learning quite early in life. Indeed. subjectstended to identify the scene representedwith low -pass filter (coarse ) infonnation . The resultsshowedthat at the fastestpresentation times (30 ms).as long as the set of universalsis correct. then the where" system might be incapable of doing anything but selectingcoarseinfonnation about objects and their generalgeometric relationships. while children learning English must learn to ignore theseattributes. Does it make senseto attribute the design of spatial language to such language causes ? And certain facts about spatial languagemust surely be learned(or ignored)" or " flatness " children learning Tzeltal must learn to attend to an object' s " bulginess when describing its location . Schyns and Oliva ( 1994 ) interpreted this pattern as evidencefor two different processing schemesthat operate in sequence .356 Barbara Landau ' s overall " " it preservedthe scene geometry but eliminated all fine-grained boundary and edge infonnation such as would be required for identifying particular buildings or vehicles . These two parallels betweenlinguistic and nonlinguistic systemsplace the burden of explanation on the design of systems that presumably evolved independent of . The question was whether subjects highway would identify the hybrids on the basis of coarse or fine-grained infonnation . at slower times 150 ms tended to the scenerepresentedby highpresentation ( ). What are we to make of this? Note that none of thesepossibilities is inconsistent with the others. and how this would vary with exposuretime. It is also a fact that sameness in object shapeis strongly correlated with sameness in name.

learning all languagesshould learn terms shape objects -shapeobjects. there . it would be reasonableto assumethat the possibilities outlined above are all mutually reinforcing . Becauseobject names often do apply to (though clearly not all . In this way. Remaining clear that objects can be represented More puzzlesthan answersremain. corresponding roughly to loose. repercussions system might shape .and tight -fit versionsof the English term in. Although it seems in terms of very different geometric descriptions (for different purposes ). From the perspectiveof learning. basedon structure or function . they with these same kinds that are correlated for ) ( object to could learn that shape is important object naming. that differentially select may exist different systems information relevant to naming objects and to locating them.S Concluding Comments . it remains unclear just what the status of thesedescriptions is. appearing as differencesin the coding of objects in linguistic representations of " what" and " where. the differential representation -based information in these systems may propagate up to the of shape highest level. usually named by detailed shape . children should learn to discount the particular shape of an object when learning those terms. and places (more precisely .Multiple Geometric Representationsof Objects 357 ) . That is. For example . whereascoarse language " " in then we see direct of the where is a function . see Bloom 1994 in children similar that are . what is the status of these descriptions with respect to dividing up spatial " " ? If detailed shapeis really a function of the what system . becauselocational terms such as spatial prepositions tend to apply acrossobjects that vary enormously in shape . while the same figure and ground do possess objects obviously can be and are representedin detail when they are recognizedor named as object kinds. and the kinds of distinctions that appear relevant acrossall languagesare completely . that the distinctions betweenfigure and ground geometry. Similarly. Objects (usually named by count nouns) preserve different portions of spatial language -functions. place ." Puzzles 8. with respectto at least . but from the requirementsof a communication systemthat must rapidly . four different issues First . It is possible . unrelated to the facts about structure and processingof objects compared to places external It is also possiblethat the facts about spatial languagederive not from causes to language . given that some languagesdo incorporate somewhat more object information than English in their stock of basic spatial terms. But if this is true. we are still left with a puzzle of why convey complex meanings comparatively little fine-grained detail. of course. A role for learning would seemto be crucial. the child learning Korean will have to learn the difference between ahn and sok.

In other languages . long. for example systems coarse descriptions in one system but not the other? A recent study by Breedin. relatively detailed object ) and positional verbs. this volume) . Should we expect the different shape descriptions to cleave neatly along lines of " " " " form class . some of which are selectedas special by ? languages Third . axial) to be found ? Or does that system give rise in any principled sensewithin the visual system " " to a variety of different descriptions. One of their patients sustaineddamage on this issue . and axial representation are encoded in spatial adjectives(e. what do we do with the persistentappearanceof the same coarse shape " . or along lines of some other distinction such as what and where ? " " And if so. The deficit to the infero temporal lobeand possessed was specific to object naming. I have presentedevidencesuggestingthat learning and development . 1993 shapecan be encodedin verbs (Japanese . nor on . supporting the functional axial coarse and the detailed between / descriptions outlined in this chapter. what is the descriptions as they articulate with ? In this chapter. this patient showed no impairment on spatial prepositions. But ? Even can we really connect the object/place representationsto different form classes within English.that show " " " " " " " . precise shape is encoded in certain verbs (posture verbs such as to kneel and to crouch . at least with respectto language a severeobject naming deficit. what is the status of object descriptions relative to representation in the " " brain? Do the different object descriptions enjoy different status in the what and " where" ? Can we find evidencefor the existenceof axial and . and perhaps manner verbs such as to undulate and to spin) . for example ). and the object representations language learning for cornerstone as a critical serve in life flexible accessto them early learning.g. flat ? spatial predicates A secondpuzzleconcernsthe status of thesedifferent object descriptions relative to . So far .the patient could recognize objects. seeSinha et al.358 Barbara Landau spatial prepositions) preserveonly coarse or axial descriptions. thin .. wide. which require labeling the ends of the object axes based terms are functionally separatefrom object names . Saffran. thin . may different with coordinated Discovering precisely how these representationsbecome parts of vocabulary and how they becomemodified by learning remains a challenge . Is the three-part division (detailed. and Coslett ( 1994 . Despite the naming deficit. separation these status of Fourth and finally . The existence . chapter 2. and up in classifers descriptors. long. so good. Breedin and Saffran in preparation) may shed somelight . verbs. for future research .round . Thus the axisobject-part terms. Allan 1977 coarse or axial shape can be encoded in classifiers (see . probably prior to multiple representations of objects exist early in development these different of . coarse visual representations . see Bierwisch.

This is clearly not the case -class However. This focus does not entail that . SeeHerskovits ( 1986 of many other contextual constraints. while the open class languageis likely to representthe fine semanticstructure of a language . 4. Thesepossibilities are currently being tested that shapematters for object names . There are severalpossibleexplanations for the sharpeningin the shapebias with vocabulary growth . grammaticizedportion of the ). but with eachsurrounding the single location most frequently acceptedby the child. If the flowers are real (rather than painted). Jennifer BloomandManishSingh . even for English. such as the ability to detect sharpening of the shapebias are a consequence which words are count nouns (hence object names ) . which we can build richer theories of the kinds of spatial meaningsencodedin languages " " 3. The tenD acrossis described by Talmy ( 1983 ) as requiring a ribbonal figure and ground that people judging the acceptability of a 1991 showed Williams . 6. . This suggeststhat the figure must have a clear principal axis ellipses intersecting"rectangles " (making it a linear figure) in order to best satisfy the requirementsfor this tenD. seeLandau and Gleitman 1985 's the child of reflection is a bias of the the is that learning genuine shape sharpening possibility . Probe trials were conducted using the same span of locations. I thank Paul March of Dimesand by National Institutesfor Health grant ROI HD-28675 of thechapter versions on previous for thoughtfulcomments . sharpening representations tional resources(seeLandau 1994for discussion ) or an enhancementdue to input that reinforces . One possibility (described in the text) is that children begin with a representational in tenDSof shape bias in which objects are represented . A separateseriesof questions probed subjects generalization patterns for known tenDS such as across . Vim for helppreparing Nolan andJessie figures Notes I . A secondpossibility is that both vocabulary growth and the the importance of shape of a third factor . following Talmy ( 1983 . This procedure was modified for the few children who said yes only to locations other than the one directly in front of them. " " 5. This chapter will focus on spatial prepositions in English. and another bias in which object names bejust to connect up the two pairs of would of The function kinds.Multiple Geometric Representationsof Objects Acknowledgment 359 -FY93 -0723from the Award 12 Sciences This work wassupported by Socialand Behavioral . this assumption Should of wider a verbs meanings range ) may represent (including spatial prove wrong. the patterns were not the sameas those found in the learning study (seeLandau and Stecker 1990for details) . Syntactic growth (with which the child could detennine which words are count nouns) has long been thought to be a possiblecauseof ) . spatial infonnation is coded only in thesetenDs. then pragmatic constraints would force the ) fordiscussion interpretation that they are on the upper surfaceof the bowl. I assumethat the closed " " . A third the so-called vocabulary explosion (for discussion . 2. An ( ) experiment by object circles across found instance of as an rectanglesmuch lessacceptablethan intersecting display . the analysis of English spatial prepositions can still provide a framework within . are linked to object learning in noise with expandedcomputaa decrease either reflect could the . It is worth noting that neither children nor adults were simply translating known prepositions ' .

In L. B. Recognition by-components . S.. and Saffran ). ( 1991 ). Cognitive semantic dementia . ChildLinguage . Cognitive development acquisition language . M. haptics. 46(3).Barbara Landau 7. most of the shape 8. 41. R. : A theoryof humanimageunderstanding Biederman . Paper at IEEE Systems Binford. H. 3. J. Frames . : Cognitive vs Bowerman . and Coslett ).. Hammond ). ( 1988 . P. L. Color-nameversus shape .. Miami. A. ( 1991 ). 9. Levine . I thank Manish Singh for illuminating discussionof this issue References . Paper Brown . Levinson determinants J. ( 1977 ). andBowerman .. ( ). 297 -namelearningin youngchildren . S. 12 . Cognitive . ( 1993 ). . Cambridge . ( 1971 . 20. ( 1985 ). Review . I. 65. -semantics in the acquisition of : The role of syntax Bloom names . ( 1994 mappings ). Science . most notably. . Reversal . Rethinking linguistic Gumperz linguistic . M. Manuscript loss : A case . The characterization of Tzeltal as especially" visual seems distinctions it carries can also be representedby other spatial systems . M. T. Childrens useof shape . ( 1987 ). 92. and Calvanio . . Visualperception by computer presented andCybernetics Conference .147 ..660 . O. 223 : in English andKorean motionevents to express Choi. C.122 -specific . 6. . 617 Neuropsychology . . Moore Ed. Lexicalacquisition . K.. . .16 Development objects posturalchange . . 83 patterns ' ' in his first Clark. In TE .110 of language ( ). K . I thank Paul Bloom for reminding me of this fact. D. Possible . M. A. ). In and S. University April Language presented Stanford . Learning . and . MA: Cambridge UniversityPress relativity in the faceof semantic Breedin . Visual and spatialmental .462 : Dissociable of representation . Cognitive : Identity versus animate . at the 25th Annual Child . P. ( 1991 ). . The role of shapein the acquisitionof Tzeltal (Mayan Stanford Research Forum. H. ( 1994 . Cognition lexicalization Theinfluence of language . Special volume .. and Ward. 94 . ( 1993 ) locatives ). (in preparation . . M. D.). E. B. Press NewYork: Academic Farah .332 . ( 1973 ). 387 393 ' . What s in a word? On the child s acquisitionof semantics . V. 439 Psychology imagery systems . . . Gleitmanand B. Sentence processing . 53 .. S. and Irwin. " unmotivated. R. Whereis above ? Cognition . Lingua .244 . 115 Psychological . EM . Language Allan . CA. Saffran . -Radvansky : in visionand language of reference Carlson . Landau(Eds nominals . E. Classifers (2). 285 ' in extending novellabelsto Becker . 11 . TempleUniversity study effectwith of theconcreteness Breedin . I thank Misha Becker for helping collect data on thesedistinctions.. The origins of childrens spatial semantic categories Eds .311 . Journalof Bomstein .

6. B. L . 35 .. B. 1990 ) . ( 1993 ) . St . Cognition. In Proceedings ) . JonesS . and Jackendoff. Theoreticalissues . Journal of Memory and Language . L . ( . ( 1992 ) .1004 . 202. 31 . A . Boston. ( 1983 . W. and object kind . and Kucera. L . The development of locative expressionsin English. ( 1994 . L . Mo . Boston: Houghton Mifftin . Journal of Child Language Serbo . B . M .). R. The of in lexical . 807 . 5 : Syntactic 312 . and Gershkoff-Stowe. . ) . University Press " " " " Landau. M . and object naming . ( 1992 kind ) .oral and Brain Sciences . Variation in the use of front and back in bilingual speakers of the First Annual Meeting of the Berkeley Linguistics Society. 18. 529. Langauge Development Conference Kuczaj. . J. R. and JonesS . in B . L . E. ) . Johnston. I . Medin ( Ed. bias in children s Landau . N . Objects and places .. B. W . R. 299 321 . B. and Richards. linguistic study of languageacquisition 961. Languageand experience . ( 1984 ) . D .210. M . Landau. . . globin (Ed. in M . Syntactic context and the shape 825 . H . 2. 16. On the acquisition of front . ( 1982 : Lexicon and ) . ( 1985 : The evidencefrom children learning English... and side. back. 31. October. and Gleitman . Smith. Cognitive prerequisites In D . D . Frequencyanalysis of English usage grammar. R.. Vol . 217 238. ( 1994 ).96. Paper presented at the Psychonomics Society . Vol . Louis . Hoffman . Hill .. 253. D . Neurology . New York : Academic Press Landau. Parts of recognition. 46. Languageand spatial cognition: An interdisciplinary study of theprepositions in English. What and where in spatial languageand spatial .. and geometric representations early lexical learning Cognitive Development . cognition . and Stecker . ( 1985 . Cognition.Multiple Geometric Representationsof Objects 361 Francis. W . ( 1985 ) disorders . Cambridge. The cross . . Hayward . 65. Spatial languageand spatial representation Herskovits. and Maratsos. MA : MIT Press Johnston. B. ( ) importance shape early learning Cognitive Development . S. NJ: Erlbaum. what Warach " and . 287 Levine . 1010 1018 .. Object shape . Landau . ( 1975 . malleability . Hillsdale. and globin . Leyton . object .304. Lynch.). . C. Rigidity . J " . ( 1978 ). 255 265. Smith . Smith . object name. . Paper presented at the Boston University . and Moore . -Croatian . In D . Psychologyof learning and motivation. J. On the origins of the shape bias in young children' s novel word extensions . and Tarr . ) . ( 1992 ) . ( 1986 ) . adults B ' and lexical learning . 1988 .. 3 . Landau. and Turkish .. C. Jackendoff. Two visual systems to bilateral in mental imagery cerebral : Dissociation where - imagery due posterior lesions . Cambridge. Semanticsand cognition. Berkeley: University of California . ( 1975 ) . Behav Landau. Child Development . and Farah " . . MA : Harvard ) . . ' Landau . of D " ..545. and JonesS . Cambridge: Cambridge University Press .

Up front . movement.. ( 1992 ) . 5.. B. PsychologicalScience .. ( 1963 / causality. causality. Thorseng. ( 1994 Schyns . Current Directions in PsychologicalScience . New York : Freeman.. Symmetry.902. Mervis.. C. English. width . E. 240. ( 1954 ' . / reality in the child. and perception. 1981 -Braem. ( 1960 / geometry ) . London : Methuen. ( 1993 ) . . Glasgow. M . 9. 895. Max Planck Institute for Psycho linguistics. Michotte . adjectives . 96. ( 1989 . MA : MIT Press Leyton. Vision. Languageandperception . Dordrecht : Kluwer . Perceptualorganizationand visual recognition. Science . . ). . 3. andCarey .. 12 . P.. London.. ( 1982 ) . 382. and Plunkett . A . 8. . Count nouns. ' . N . E. . Journal Levine . J. Basic objects in . Unpublished manuscript. Carey.362 Barbara lAndau : Theacquisition of a concept and a word . When what meets where... Inhelder. ( 1982 ) . Segregation of form . 740. E. shape . L . D . New York : Basic Books. Vision . physiology.. B. 273. 501 524. and Landau. Representationand recognition of the movement of shapes o Proceedings / the Royal Society. Boston University . ( 1994 ) . Japanese the International Conferenceon the Psychologyof Languageand Communication . ( 1993 ) . ( 1976 Rosch.. Cognitive Anthropology ResearchGroup . Reprint. Onto logical categoriesguide young children s inductions Sola. A . and Johnson-Laird . MA : Harvard Miller . L . ( 1992 . S. color . D . 179 211.439. The child s conceptiono Piaget. 2/ 4. ( 1985 ) .. Cognition. and Spelke of word meaning: Object terms and substanceterms. and Inhelder. C. 195 scaledependentscenerecognition. and perceptualproperties Smith.. 645 Language of Child Levinson. mind. From blobs to boundary edges . ( 1992 . M . DevelopmentalPsychology . W . S. S. / 63.749.. and object description Working paper no. The lexical semanticsof length. The child s conceptiono Piaget 1967 . and height. Cambridge. J. M . ( 1969 ) . S. Marr . The constructiono Piaget. ) . and Oliva . A . 28. Paper presentedat and Danish Evidence from : and language acquisition . ( 1976 ) . Shiffrar . and Hubel . Reprint. M . J. Hayashi. Anatomy . Press University " " " " " " Narissiman. G . . New York : Norton . Schneider . C. 38. ( 1948 ) . Gray . P. ) . Science LoweD . Johnson. Comparative spatial semantics . ( 1991 ) . B.. ' .. Cambridge. S.657 . and linguistic description: Tzeltal body-part tenninology ) . and Szeminska . Jones ) . (1982 ). .286. G.200. Marr . and depth: Living stone. / space . D . New York : Norton . K . B.100 Sinha. Theperception0 . and VainaL . P. Cognitive Psychology natural categories . Two visual systems : Evidencefor time and spatial ) .. and Boyes ) . in children' s novel word interpretations. D .. Nijmegen.

Acredolo(Eds . orientationTheory . ( 1993 ). L. application . G. P. and Mishkin. Mansfield . Anderson . Science . Analysis ( . Press Plenum York: . Two corticalvisualsystems .. In T. Informationprocessing Van Essen . Los . MA: Eds .. ( 1992 ). Press . J. University .. ( 1991 ). Cambridge University in theacquisition Tanz of deicticterms . description syntactic typology ( ). Cambridge lexicon UniversityPress . . Cambridge R. categories Ed. W. . C.149 . A. and . nouns Angeles andmass . In D. ( 1980 ). C. L. 255 . Spatial Pick and L. ( 1983 . Childrens and adults understanding .Geometric Representationsof Objects Multiple 363 of count contextin the learning esandsyntactic process . D. 419 : An integrated visualsystem perspective systems ' ' . L. Universityof California . ). MIT Press in the primate . Diss . M. D. PhiD.423 .). Columbia Williams . M. Studies .. 57 : Cambridge . How language Talmy. Ingle . structures space ). ( ) Talmy and the 3 Grammatical . and Goodale of ). K. ( 1982 Ungerleider . J. Language . Vol. Perceptual Subrahrnanyam . In H. Honorsthesis of across . andFelleman . New 225 282 and research : . Shopen structurein lexicalforms : Semantic . Lexicalization 1985 patterns . 549 586 behavior visual . : Cambridge .

Rather." One further introductory comment.Chapter PreverbalRepresentation and Language Jean M . I will suggestthat language is structured in spatially relevant ways becausethe meaning systemof the preverbal languagelearner is spatially structured. Pace Piaget.1 Sensorimotor Schemas Are Not Concepts The more I delve into cognition in the first year of life the more it becomesapparent that many of the most basic foundations on which adult conceptsrest are laid down during this period. " " spatial information has been redescribedinto meaning packages . It is that many of the most basicmeaningsthat languageexpresses both semantic and syntactic are basedon spatial representations . What I hope to contribute are a few suggestions so structured. Such a point of view will hardly be newsto cognitive linguists such as Ronald Langacker or Leonard as to why languageshould be Talmy . the subtitle of this chapter should read: " How SpaceStructures Language . I will argue that someof the categorical packagesretain somespatial characteristics or packaging characteristicsoften ascribed to languageitself are actually due to the prepackaging that is accomplished during the preverbal period. the preverbal system forms the foundation on which language rests . So with apologiesto Leonard Talmy for twisting his words. Mandler Although my interests lie in the character of the preverbal conceptual system rather than of language itself. and thesemeaning . I shall argue that preverbal conceptual representationis largely spatial in nature and that the relationship betweenspace and languageis therefore far -reaching and pervasive . and it constrains what is learnable. 9. Babies do not wait until the onset of languageto start thinking . spatial meaningsplace constraints on spatial terms. To say that the preverbal meaning system is spatially structured is not to say that it is the same as spatial perception. It is not just that spatial terms tell us something about spatial meanings or that . the first year of life is far from being an exclusively . the problem of packaging meanings into workable units is thus a prelinguistic one.

this dilemma was handled in different ways by people studying . if for no other reason than the need to account for the . On the . developmental psychologiststo explicate ably enough. Fundamental concepts months (perhaps to seven of animals and vehicles are learned by around six resting on an even earlier conceptual distinction between animate and inanimate things). Mandler . and utensils follow soon after. Instead. These es :ning process conceptual domains in turn are used to control inferential reaso has In addition the . but have avoided specifying what that systemis like. The researchthat Laraine McDonough and I have been conducting indicates that the foundations of the major conceptual domains are being laid down during this period (Mandler and McDonough 1993 ) . we pay lip service to the idea that to onto a conceptual mapping learn languagerequires a preexisting conceptual system . McDonough in press . furniture . For a rich many people working in language acquisition conceptual developed this will come as no surprise. the higher cognitive functions that (among other things) sensorimotor stage will support languageacquisition are being formed in parallel with the sensorimotor learning that is going on. In its place we find a baby that has already ? It seemsto have disappeared life. All this is happening before children learn how to speak. in . Where is the familiar sensorimotor infant we Such findings should give us pause who has not yet achievedconceptual representation the creature are usedto hearing about. like. On the in it various transforms instead thought precedes ways . He did devote a good deal of effort to describing how sensorimotor schema schemasmight be transformed into conceptual (symbolic) representation . are themselves As best as I can tell . The neglect seemsto be due in part to a conflict within Piagetian theory . to learn languagerequires representationalsystem base . babies are not supposedto have a conceptual . .366 Jean M . which then language other hand. episodic memory system ) press (Mandler and McDonough es have begun (Mandler and become operational and long-term recall process ) . but he said little about how the new type of representationdiffered from the old. Piaget tended to characterize early verbalizations as just another kind of sensorimotor . As a result. Workers in language languageacquisition and those studying cognitive development for learning language the various notions to necessary acquisition attempted specify it to the left reason and then. and the domains of plants. According to Piaget ( 1951 ). But the complexity of the concepts that newly verbal children expressin language in literature the current researchdoes make evident a tension that has been lurking for many years. yet according to linguists. . Sensorimotor schemas little is said about what the concepts but and concepts are mapped into language . becauselanguage begins before the sensorimotor period ends . but one hand. Piaget ( 1967 ) said that conceptual thought is not created by language . The result are said to be transformed into concepts is a gap in his theory .

than onto the sensorimotor schemasthemselves . sensorimotor schemasare not the right sort of representation . For example . agent. Thus in many accounts themselves were the sameas the sensorimotor schemas the sensorimotor achievementswere assumedto be the baseonto which languageis . For example (with the exception of the nativist position that grammatical categories are innately given) there seemsto be widespread agreement that the underlying concepts needed to learn grammatical " " " " " " " " categoriesare notions such as actionality . A sensorimotor schemaprovides something like meaning in that it enables recognition of previously seenobjects to take place. and " Maratsos 1983 . the meaning might be a representationof one object containing another. But where the " ) developmental psychologists were to possession ( take over. which Piaget considered to be the sine qua non for both the development of the higher cognitive functions and languageacquisition. Spelke et al. sensorimotor schemas are neither conceptsnor symbols. 1992 . location . objecthood. It is this conceptual simplification onto which propositional languageis mapped. rather . In it a of the greatly simplifies .and Language Preverbal Representation 367 the representational status of these notions. it seemsto have been assumedby default that the relevant conceptual categories . and thus for the world to seem familiar . For the most part . consider putting a spoon into a bowl. there was largely a blank. but . But a sensorimotor schema does not " means to its parts for purposesof denotation or to enablethe baby allow independentaccess to think independently of the activation of the schemaitself (Karmiloff -Smith 1986 ). fonning summary this case . until the recent work on objects and agencybegan to appear (Baillargeon 1993 . There are other ways in which sensorimotor knowledge also appears to be the structure perception . How ? Some kind of are such schemasto be mapped into a discrete propositional system interface betweenperception (or action) and languageis needed . a conditioned stimulus predicts or " that some other event will follow . such as object permanence acquisition to stage6 sensorimotor accomplishments thesewere not very successful(seeNelson and Lucariello 1985for discussion ). Sensorimotor schemas wrong sort of basefor learning language and control action. It also allows each component of a familiar event to signal the next component to come. something that will allow an analog-digital transformation . Typical examples of this approachwerethe various attemptsto relatelanguage mapped . This kind of reaction is indexical. Piaget provided some of the reasonswhy procedural forms of for learning language representation such as sensorimotor schemascannot in themselvesserve a semiotic function . These schemasconsist of a large number of parameters that monitor continuously varying movementsand rapidly shifting perceptual views. but the conceptual system event that constitutes its meaning. In short. This requires an intricate sequenceof movements . BecausePiagetian theory was silent about conceptual representation at the end of the sensorimotor period. Leslie 1984 ).

Elmas. Quinn to more abstract or theory-laden concept formation takes place and Elmas ( 1986 ). the traditional doctrine of the British empiricists. and in so doing theseperceptual categories becomeconceptual in nature." before children develop abstract conceptsabout the world they categorizeobjectson the basisof their physical appearance according to the laws of perceptual similarity . as I . this approach does not explain how the transition from perceptually basedcategorization . and one could imagine how it might becomeassociatedwith the perceptual category of dog.2 Differences between Perceptual and Conceptual Categories Jean M . and cats from both dogs and lions. quickly distinguish from zebras . More importantly . But it is difficult to understand how properties that are less clearly perceptual are represented " " " " . Nevertheless . but Quinn and Elmas (in press ) believe assumedo many others. Mandler In addition to Piaget' s view that at the end of infancy conceptsare constructed out of sensorimotor schemas . espousedin modern times namely. categories of animals after a very few exposuresto pictures of contrasting classes For example both three olds and six month olds learn to horses . This view is given support by the recent findings of Elmas and Quinn and their colleagues(Elmas and Quinn 1994 . and Rosenkrantz 1993 ) showing that as young as three months. such as animate or interacts with me. that these perceptual categories form the kernel around which the first conceptswill develop. and only much later generalizefrom theseconceptsto form a superordinateconcept of animal. have pointed out that no one taking the traditional empiricist view hasever satisfactorily explained how abstract or superordinate conceptsare derived from the perceptualconceptsof infants. A property such as barking might be a perceptual category in its own right .g. which Keil ( 1991 ) has called the by philosophers such as Quine ( 1977 doctrine of " original sim. In this view. but it would seemto be a processalong the lines of the doctrine of original sim. Indeed. or how theorybasedassociationsbegin to supplant perceptually basedones(seealso Fodor 1981 ). there are both theoretical and empirical difficulties with this view . there is an even older view of the onset of concept formation . Quinn. in my opinion . . Theoretically. among others (e.. babiesfirst form conceptssuch as dog and cat on the basis of the similarity of the exemplarsto each other. dogs from cats. babies form perceptual . This associativedoctrine of the creation of conceptsis exemplified in current theory by the view that the first conceptsto be formed are at the basic level ( Mervis and Rosch 1981 ) . Once theseperceptual categoriesare formed. It is agreedthat these are purely perceptual accomplishments . various types of information becomeassociatedwith them. it does not specify in what form the that have never been resolved information to be associatedwith the perceptualcategoriesis itself couched. Keill99 I ). The details of this process have never been worked out . In this view.368 9. ) .

this difficulty might be finessed . in particular . Second . The quintessential example of this dilemma is shown by infants in our experimentsdistinguishing betweenlittle models of birds and airplanes. Nelson 1985 ) . For example . Bauer. basic McDonough and I have suggested that the infants at these different ages are actually engaged in different kinds of .. whereasin our work we have measured times to manually explore objects.2 Furthermore. and plant were late acquisitions. such as -level classes of land vehicles dogs and rabbits (and also basic . several reasonsto make the distinction are already known. Rather. vehicle. whose exemplarsdo not look alike. whose exemplars do look alike.to six-month -old infants are apparently so much more advancedthan seven . The details of the development on theseconceptual domains is not my main concern here. perhaps languageacquisition itself contributes to superordinateconcept formation (e. on our tasks infants differentiate animals and vehiclesfrom sevenmonths onward. Because most conceptual aspectsof thesetwo developmentshave not yet been . differentiation among various basic-level classesof mammals. such as cars and trucks) is still not well establishedat eighteen months (Mandler . while failing to categorizethe basic -level classes within them. the traditional . while at the same time not distinguishing between dogs and fish or 3 dogs and rabbits. whoseshapesare quite different ( Mandler and McDonough 1993 ). However. and McDonough 1991 ). This researchshows that on some tasks infants distinguish global categoriesbefore they distinguish the basic-level categoriesnestedwithin the animal class . But even by elevenmonths.to eleven month -olds. First .Preverbal Representationand Language 369 As long as it was assumedthat superordinate concepts such as animal. all of which have outstretchedwings and therefore very similar overall shapes . Apparently . why the younger infants make fine discriminations among -level classesthat the older infants do not.g. it would be difficult to explain how infants could manageon any kind of task to categorizetwo superordinate domains. and other global concepts such as plant are in place at least eleven months by (we have not yet tested younger children on this concept) . a purely perceptual account of categorization cannot explain why three. I want to emphasizethat the developmentof perceptual categories (which are sensorimotor accomplishments ) does not look like the development of ones. ! For example . if there were only perceptually based categories in infancy. the studies of categorization in young infants have measuredtimes to look at pictures. However. even though superficially there seemto be similar task demands in the processing various experimentsthat have beenconducted. they do not differentiate dogs and rabbits or dogs and fish. specifying the differencesbetween them is still problematic. Nevertheless investigated . research in our laboratory has shown that infants have formed conceptsof animal and vehicle as early as sevenmonths of age (Mandler and McDonough 1993 ). The experimentsfor both age ranges have used a habituation -dishabituation paradigm.

but do not find thesedifferencesimportant enough to treat them differentially . but having constructed fewer concepts about the world . the primary meaning to accrue to a basic-level category such as dog is that it is an animal. . Babiesseethe differencesin the appearanceof dogs and rabbits. but for most purposes treats them as the samekind of thing . or are man s best friend. they show intenseinterest and concentration and subject loss is virtually nil . there is often high subject ). it is secondary(not only for infants. dogs. as having meaning. that is. This different processconsistsof treating objects as kinds of things. but that when they are older and are given a task that engagestheir interest. the meaningsthat make up this concept need to be derived from information that babies can learn from observation alone.370 Jean M . namely. Thus perception of biological versusnonbiological motion of is one early sourceof knowledge that could be usedto divide the world into classes in and inanimate . that move animate ways things . when infants are given objects to explore. although the kinds they have conceptualized are fundamental cuts that give meaning to the perceptualcategoriesthey are also making. not just as things of differing appearance This early conceptual processing is crude in comparison to the fine perceptual discriminations that infants make. This situation is essentially the same as when an older child or adult seesthe differencesin the appearanceof poodles and collies. they have just begun to handle objects and are still unskilled -month -olds have held any kind at doing so. So what kind of information is at their disposal first that seems ) has shown likely to be relevant is biological motion . They appear not yet to havedivided the world into very many different kinds. our findings suggestthat very young infants begin to perceptually categorizethe world in the absenceof meaning. Bertenthal ( 1993 that three-month -olds already differentiate biological from nonbiological motion . exactly what doesthis initial concept of animal consist of and question then becomes how is it learned ? Unless one wants to posit that the concept of animal consists of a set of innate ideas . animals. I am suggestingthat the babiesin our experimentscan seethat dogs look different from fish or rabbits. That is. Although this issueneedsfurther study. but adults as well) that dogs are four ' 4 leggedor bark . The . Mandler looking -time habituation -dishabituation experiments do not engage infants very . a different processis brought to bear. for example deeply (Mandler and McDonough 1993 loss in these experiments even when the infants are given something to suck on to keep them awake and happy. On the other hand. It is also unlikely that most seven ? The of real animal in their hands. . It seems insofar as the parametersof people' s motion are concerned likely that they do the same for other animals as well becausethe parameters governing animate motion are quite general. namely. for most purposestreat them as the samekind of thing . babies are not yet independently locomoting. By sevenmonths.

There are. As young as three months. if the differenceis not just to remain a sensorimotor distinction but to represent a meaning. interactive. by four months. babiesdifferentiate causedmotion from self-motion (Leslie 1984 ). whereas those things that move mechanically and get pushednever interact from a distance. causing -to-be-moved. 9. There are undoubtedly other properties babies learn about before they begin to handle objects themselves . dolls do not. By six months. Indeed. contacting a surface move. They expect an object that losescontact with a surface to fall . babieshave learned something about containment. Another characteristic to be noticed is that the things that move in the biological way and start on their own also interact with other objects from a distance . containing these are all observable and or kinetic .3 How Meanings Are Created -toSelf-starting. but theseare someof the main onesthat have beenstudied to date. By sevenmonths. Similarly. One of the ways to do this is to notice that the things that move in the biological way start up on their own. major properties pick up acuity is still not well developed to Responsivity thesecharacteristics of motion can explain why babies as young as two months of age respond differentially to people and to dolls (Legerstee 1992 ). that other objects cannot pass through them. unless it is supported by a hand (Baillargeon in press ) . and that objects still exist when they move out of sight (Baillargeon 1993 ) . so that various insubstantial objects such as a horizontal finger touching a large box are expected to be sufficient to provide support. many other properties of objects that babiesobserveas well.Preverbal Representationand Language 371 Once these categoriesof motion are formed they must be characterized in some way. By four months. This is one of the reasonswhy I have proposed spatial / properties it that is spatial properties (including motion ) babies analyze and abstract from perceptual . a processof perceptual analysis begins to take place . mechanically moving. infants have begun to learn about the properties of object support . biologically moving. caused . of course . babies know that objects are solid. Notice that each of these . This is an attentive processthat occurs when an object is being . these are some of the properties is available even to very young babies that babies can when their . whereasthe things that move in the mechanical way start only when another object contacts them. Peopleinteract with them. Slightly older infants expect that any contact implies support. it can explain why. babieshave learned enough about contact and support to predict that something seento overlap its supporting surface by only about 15% of its basewill fall . I have suggestedthat as infants are learning to displays to form meanings parse the world into objects. 1992 (Mandler 1988 ) . they know that containers must have bottoms if they are to hold things (Kolstad 1991 ) .

Even if we agree that the earliest meanings . I claim that thesespatial abstractions are sufficient in themselves to meanings the initial of such as animate inanimate represent meanings concepts thing. The mechanismof perceptual analysis I have describedmakes it unnecessary to posit inmate ideas or concepts . it is a constructivist account. or move around them) for meaning to begin to be created . there doesnot seemto be any good meaningsthemselves reason to translate them into propositional form . Once language is learned. move them around. and container. After all . perceptual analysisalone can build up meaningsand can do so continuously throughout infancy (and for that matter. on the contrary. it may take no more than an intelligent eye and a mechanism to transform what the eye observes . Thus babies have a mechanism that enablesthem to abstract spatial regularities and to usetheseabstractions to form the . Thus babiescan create a beginning concept of animal even though it is crude genes compared to the biological theory they will eventually espouse(Carey 1985 ) . It is theseinva