Series Foreword

The JeanNicod Lectures are delivered annually in Paris by a leading philosopher of mind or philosophically oriented cognitive scientist. The 1993 inaugural lectures marked the centenary of the birth of the French philosopher and logi Jean Nicod (1893 1931 . The lectures are sponsored by ) ~ the Centre National de la RechercheScientifique (CNRS) as part of its effort to develop the interdisciplinary field of cognitive sciencein France. The series hosts the texts of the lectures or the monographs they inspire .

Bouveresse , Presidentof the JeanNicod Committee Jacues Andre Holley, Secretary the Cognitive Science of , Program

Fran ois Recanati Secretaryof the JeanNicod Committee , ~ and Editor of the Series
Nicod Committee Jean Mario Borillo Jean Pierre Changeux Claude Debru Jean- Gabriel Ganascia Michellmbert

PierreJacob Mehler Jacques Philippe de Roullhan Dan Sperber


I am grateful to CNRS, The Centre National de la Recherche Scientifique of France, for inviting me to give these lectures in Paris in the spring of 1994 at the second annual Conference Jean Nicod de Philosophie Cognitive . Without the stimulus that this ' conference provided - actually , the threat of embarrassing myself before a French audience these ideas would probably never have been assembled in such a .timely way . They might not have been assembled at all. My weeks in Paris were made particularly memorable both socially and philosophically - by a group of philosophers and scientists , faculty and students alike (most of them , I think , at CREA , Centre de Recherche en Epistemologie Appliquee ), who were not only gracious hosts,. but challenging critics after each lecture . It was a humbling experience to find myself arguing technical issues in the philosophy of mind with French philosophers - in the middle of Paris no less- in English . Imagine what they would have done to me in French. I particularly want to thank Pierre Jacob for everything he did - and there were many things - to make my visit so pleasant.


Thanks to the Stanford Humanities Center I was able to devote almost the full year to preparing these lectures. I was a fellow there during the year 1993 1994 and I am deeply to the support I received from the director, Wanda grateful Corn , the associatedirector, Charles Junkerman, and the fellows and staff. No one everknocked on my door. They obviously know how to run a researchinstitute . I really started working on these lectures (without realizing it ) in 1991 when I was preparing talks to give in Valencia, Spain. The Valencia lectures were early stabs at these topics - so early, in fact, that I did not want to publish the results. I wish to thank the philosophers in Spain, and especially Josep Corbi , for their hospitality and helpful discussion . Students and friends read some ---and in a few casesallof these lectures and gave me useful comments and criticism . . I want to thank Sven Bemecker, and Tll11Schroeder here at Stanford, Murat Aydede now at Chicago, and Eric Schwitzgebel at Berkeley for their help . They made more of a difference than might be apparent to them . David Rosenthal read most of the lectures and provided useful comments - especially on chapter 4. Berent Enc, an old friend from WISCOnsingave me months of grief by not letting , me ignore problems I didn ' t know how to solve He made me start thinking about chapter 5 two years before I got around to writing it . Georges Rey not only read the manuscript - several times as far as I can tell but gave me a truly astonishing number of useful suggestions and criti cisms. Every author should be so fortunate . I am deeply grateful to him . I tried , as best I could , to respond to his crit icisms in final revisions, but some of our differences went far too deep. To keep the book a reasonable approximation to the lectures I gave in Paris , I had to ignore many of

I' m sure.ound. No matter. how really hard these problems ~ are. We are now both wor ried about them.Acknowledgments ' Georges s most perceptive and challenging questions. Sorry about that. I thought it was supposed to go the other way ar. come back to them. I would like to thank Giiven Giizeldere . He arrived at Stanford a few years ago and promptly got me worrying about his problems. Finally . patiently and . fu great insight . I thank him for making me realize. We will . .

the first -person. at least. between what we experience (reality ) and how we experience it (appearance ) . aspect of our sensory and affective life . Representational Naturalism (as I shall call the view defined by the Representational Thesis) helps one understand . it is the only approach to the topic of consciousnessthat has much to say about the baffling problems of phenomenal experience. is where my money is. why conscious experiences have that peculiar diaphanous quality the quality of always being . though . plus or minus a bit ) (1) All mentalfacts are representation alfacts. present when but never where one looks to find them. remove all the mysteries. The thesis. in naturalistic terms. to justify putting one s money on the nose of this philosophical horse. for example . as far as I can tell . is that. It does not .Prologue The purpose of these lectures is to promote a naturalistic theory of the mind . It provides a satisfying account of the qualitative . ' It removes enough of them. The reason facts arefacts aboutinformational I am interested in this thesis is that . and (2) All representational functions. in two parts. These lectures are an attempt to convince others that the smart money belongs there. That. .distinguishing . to be sure. In providing this .something I call the Representational Thesis.

something about what camerasdo. Since the manipulation by and use of representations is the primary job of the mind . perforce. help to be told about I -stops. a deeper understanding of the mind . Nonetheless. What you need to know is something more basic . and the like . but it is not enough. I have given my own account of the propositional attitudes .of the biological machinery by means of which the mind does its job . judgment .some would say completely implausible .in Oretske ( 1988) . derive from conceiving of the mind as the representational face of the brain . If you don ' t know what a camera is.that dimension of our conscious life that helps to .for sensory affairs . The Representational Thesis is plausible enough for the propositional attitudes . something about pictures . thought .a biologically plausible answer.belief . . it establishes a framework within which subjectivity can be studied objectively . The topic is qualia . focal lengths.particularly belief and desire . Without that you don ' t know what a camera is no matter how much you know about the machinery means of which cameras do their job . shutter speeds and ASA numbers .Prologue xiv account. in these lectures I concentrate on perceptual experience. it is of no . And it provides an answer. The thesis is less plausible . It demystifies introspection : the mind ' s knowledge of itself no longer requires an internal " " eye observing the clockwork of the mind . a deeper understanding of the nature of representation and its naturalistic basis questions about the function or purpose of consciousness Thesebenefits .no matter how detailed and precise . and more. Information about the wetware may be useful . A working premise behind the Representational Thesis is that a better understanding of the mind is not to be obtained by knowledge . for the phenomenal or qualitative aspects of our mental life .

. There are. 1. however.a general feeling of depression . as I apply them . hunger . I focus here because frankly . I gave in Paris in ~ ay 1994 In these lectures I concentrated on the positive aspects of the Representational Thesis. belief . These lectures are already too long . . and the possibility of knowing what it was like to be a bat or a dogfish . . . forexample about which I do not know what to say. . still other experiences .Prologue define what -it -is-like-to-be us. inverted where progress . This is an omission that I think could be removed without fundamental alterations to the explanatory machinery developed in chapter 1. What little I do say about it occurs in chapter 4. this is where progress is most difficult . That is the purpose " of the " plus or minus a bit in the statement of the RepresentationalThesis. etc:). troublesome aspectsto a representational approach to mental phenomena. . The first four chapters of this book are the Nicod lectures . I do not . then.awareness of one s own esbodily states and process even though this is the source of some of our most obtrusive experiences (pain . for ' example.and applied . Unlike the conceptual side of oUr mental life . however. This. Many perhapsmost philosophers are convinced that this is an area where representational ideas are leastapplicable. sensations experiences and feelings cannot . consciousness points of view . and judgment can be understood as internal representations of external affairs. There are. to sense experience. examine proprioception . experi - . problems that I deliberately ignored in these lectures This is especially so when representational ideas are given a naturalistic spinseepart (2) of the RepresentationalThesis . on what it could tell us (if it were true ) about self-knowledge . There is much that is relevant to my topic that I do not discuss. qualia. Even if thought . thirst .if there is any will be most significant .

publish these lectures without addressing these doubts . powerful intuitions . quality that defies representational (not to mention naturalistic ) treatment . . There are.4. This concluding chapter represents my answer to those who will think .ing chapters 1. that the virtues of representational naturalism are all well and good. but . . therefore. . well . . after read. I think it can. a what -it -is-like .that support this skepticism .sometimes even arguments .Prologue XVI ences have a phenomenal . I could not . it just can' t be true. furthermore . I do so in chapter 5. .

Experience is a special kind of representation. taste and feel of things . 128) thinks that without it we would not . phenomenal experience . My experience of an object is the totality of ways that object appears to me. is to say what kind of special representation this is. and not all representations are mental . constituted by the properties things are represented as having . the quality of experience how things seem to us at the sensory level. Since not all facts about representations are representational facts. we think of all mental facts as representational facts. then. what ? A zombie? If . have a concept of consciousnessat all . .1 The Representational Character of Sense Experience Sense experience is the primary locus of consciousness. Remove it completely and one becomes. and the way an object appears to me is the way my sensesrepresent it . . The first order of business. Whether or not this is true . I begin by making some relevant distinctions among kinds of representations . I conclude the chapter with a short preview of the explanatory benefits of a representational approach to per- . is .a nonconceptual for In of representation .dominates our mental lives. sound . Marcel (1988 p .the look . in accordance with the Representational Thesis.

. F.) correspond to different car speeds (24 mph . and the fact that pointing at 37 means 37 mph are representational facts about the instrument and this state of the instrument . 37 mph . it can fail to do what it is supposed to do.). how fast the car is moving (F . its different ) states (pointer positions " 24." " 37. and so on. what it is supposed to do. etc. is to indicate." etc. f2.1 The way performs its function (when it performs it ) is by ' occupying different states sl ' s2 . enough . for reserved later chapters 1 The. if and only if 5 has the function of indicating (providing information about) the F of a certain domain of objects. provide information (to the driver ) about. represents a property . " 24" the information it is going 24 mph . of F. sn corresponding to the different determinate values f1. like any fallible system. is misrepresentation. If a registration of " 37" on a properly installed instrument fails to carry information about the speed of the car. Its job. doing its job . . often . 5. its function . . The result . . ceptual experienceIn 5 I describethe way it naturalizes and in 6 the way it provides a satisfying intentionality are account of why. though experiences in the head. Given the function of this instrument . . each of its states is supposed to carry a different piece of information about the speed of the car: a " " registration of 37 is supposed to carry the information that the car is going 37 mph . one ' t find them there. Other doesn explanatory benefits are . or carries the same information that a registration of " 24" carries then it is not . The fact that the speedometer has a " " speed indicating function . This is what the instrument was designed to do .1 Chapter . A speedometer ) (5) represents the speed (F of a car. and. fn.Nature of Representation The fundamental idea is that a system. When it is doing its job .

but it is not a representational fact. not (merely ) facts about mental representations . Nor does one know what (or whether ) a system represents by being told that it supplies information about speed. and mode of operation . The instrument couldn ' t do its job without this cable. indeed . It is a fact about a representational system. a fact about what information it is supposed to carry. Neuroscientists may know a great many facts about the brain . That does not make knowledge of such facts knowledge of the mind . For the same reason. On the representational account of mind I intend to give .facts about their color. according to the Representational Thesis. is a fact about a representationa device. is not a representational fact about this instrument . is. One does not understand more about the representational life of a system by being told it contains mercury or has an orange pointer . but the fact that there is such a cable does not imply that the device hasa job .The Representational Character of SenseExperience The fact that the speedometer is connected to the axle by a cable that transmits information about speed. of mental facts.let alone has the job of providing inf ~rmation . a metal whose volume serves to indicate temperature. material constitution . There are facts about representations . These facts may even turn out to be facts about mental representations . knowledge of representational facts. A representational fact about S is a fact about what S is designed to do.that do not tell one anything about what information they are supposed to supply or. shape.facts about experiences and thoughts . whether they are supposed to supply information at all. the difference between representational facts and (merel ) facts about representations is the difference between the mind and the brain . Knowledge of the mind . but it is not a representational fact. It is not whether it . on the other hand . the fact that a particular thermometer is filled with mercury.

misreP. We may be misled by the column of smoke. and damaged instruments ). Not all events that carry information have the function of ' ) carrymg it .. incidentally . even when it fails to provide the information it is its job to provide . but the smoke does not misrepresent wind speed the way an anemometer can. surely.1 Chapter supplies infonnation about speed that is important . The smoke cannot misrepresent wind speed. Representation is here being understood to combine teleological with infonnation -theoretic ideas. Since an object can retain a function even when it fails to perform it (think of the heart. This . it is whether it has the function of doing so. If the concept of . the idea of something having an information -carrying function . It captures the normative element inherent in the idea of representation. the power to say that something is so when it is not so.e.resentation It must include the power to get things wrong . The power of the color television to misrepresent color lies in its color-depicting function . is doing in the present theory . but that. to illuminate the nature of . representation is to do a useful job in cognitive science if it is to be used .continue to represent (i . the angle acolumn of smoke makes with the horizon carries information about wind speed. misrepresent) something as going 34 mph . but there is no representation without functions . This is what the teleology. thought and experience it must be rich enough to allow for . There is information without functions . To use Matthen s (1988 example. for instance.even when things go wrong . We may take it to indicate something it doesn' t (thus ourselves misrepresenting wind speed). is why black and white television does not misrepresent the color of the sky while color television sometimes does. in particular . a device can retain its indicator function . the kidneys . is not the function of the smoke. The images on the two screenscan .

Before applying these representational ideas to mental affairs . Anne Treisman (1992 p . Sensory organs and mechanisms are commonly described in terms of what they are " for. as signals or events which have been assigned the " burden " function ? of ( ) carrying meaning (information ?) . the utricle and sacculefor indication of linear acceleration. operative in many scientific approach es to mental representation. yet." The semicircular canals of the middle ear are said to be for the detection of angular acceleration. I think . This . is the difference between not representing color and misrepresenting color. Gallistel (1990 describesthe brain as representing aspectsof ) the environment when there is a functioning isomorphism between the environment and the b ~ain process that ' " " adapts the animal s behavior to it . are thus evaluable in terms of how well they do their job . The terminology is not always the same. but because that is their job . one misrepresents the other does not . This way of describing perceptual mechanisms and process is a representational way of thinking about es them.mental representation . in more precise terms . ' s 1982 well -known David Marr ( ) theory of vision assigns to ) early vision the function or task (Shapiro 1993 of depicting the spatial and chromatic properties of distal scenes. we need to know . The sensesyield representations of the world . Something like this conception of representation is.The Representational Character of SenseExperience be identical . 227 for example. ). but the intuitions and theoretical purposes are similar . and the states they produce by way of performing their function . The senses . not just because they (when working right ) deliver information about the world . though . speaks of representations . exactly what kind of representation a . and the retina for encoding information about light for transmission to the brain .

furthermore . sensors. We arrange things so that certain liquids . the job of the various senses to provide information about the world . . representing that world is it their job to source of these functions . Armed with these distinctions we can give a preliminary taxonomy of mental representation and begin the argument that conscious experience is a speciesof natural representation. on the . I begin my answers by explaining a number of pivotal distinctions : (1) natural vs. detectors and gauges are functions they get from us .thus . by design.providing functions of measuring instruments . and how in representational terms .do our experiences of objects differ from our thoughts about them ? What . conventional representations (2) representational states vs.their makers and users. Each gives rise to a different kind of function and. representational systems and (3) represented properties vs.what it is. represented objects. the difference between natural and conventional representations.the systems in which they occur conscious what is being represented? Good questions. supposed to do. 2 Natural and Conventional Representations The function of a system (or state) is what it is designed to do . indeed. The information .1 Chapter is supposed to be. what information provide . . thus.hence. . We design them. We give them a job to do. by their placement in transparent tubes adjacent to a calibrated scale provide us with iri formation about temper. One important difference (for our purposes ) is the difference between naturally acquired and conventionally assigned functions .what is the present account . If it is. There are different sources of design. a different form of representation. is it about certain representations that of makes.

4 As a result . they derive from their evolutionary history . flag poles and paper clips do not represent what thermometers represent. When a thing ' s informational functions are derived from the intentions and purposes of its designers. Their volume is also reliably correlated with temperature .The Representational Character of SenseExperience ature . I call the resulting representations conventional . That is not what they are designed to do. 380) " ' puts it thus: one of Darwin s important discoveries is that we can think of design without a designer. ). 287 321 and Searle 1992 p . (Dennett 1987 especially pp . biological functions .3 I do not argue for this . but for the present I follow the lead of Wright (1983. designed by anyone to do it . Godfrey . is not universally accepted . That . thus . though . in the relevant sense be designed to do a certain .and. Flag poles and (metal) paper clips. assume. )." The senses I . Millikan ( 1984). They do not represent anything . I assume that there are naturally acquired functions and. natural representations . deny it ). they have quite a different job . I assume it . perceptual (including proprioceptive ) systems produce representations of those conditions (external or internal as the casemay be) they have the function of informing . thus. Kitcher ( 1993). 1987). is not their function . This view . and users in this way. We haven' t given them that kind of job . Though they carry the same information . Neander (1991a 1991b Papineau (1993 Bennett (1976 and . I return to this point in chapter 5. Representationsthat are not conventional are natural. We call the resulting artifacts thermometers . builders . have the function of doing it .without being job . Philip Kitcher (1993 p .Smith ( 1994) . ) many others in supposing that bodily organs and mechanisms can. . As the names suggest. I know . also carry information about temperature . for example. the volumes of which are proportional to temperature. have information -providing functions . 52. .

5 represent We have. This makes the thesis a form of philosophical naturalism . The representations they produce by way of carrying out their informational functions have a content... that does not depend on the existence of our purposes and intentions .Otapter1 about. something they represent . more precisely. the following preliminary classification..unlike those in . speedometers and television sets .. .. All mental representations ) (not to mention nonmental natural representation s6 are . .. the internal states (experiences feelings) the . then. Conventional Indicator functions are conventional Instruments Language Natural Indicatorfunctions arenatural Sensory systems Concepts Hereafter.. That is why the perceptual representations in biological systems .or .. that they do not get from us.. ~" " ' " . This is why the senses. Representational systems Systemshaving indicator flmctions ... . I win mean the thesis that all mental states are natural representations . say. sensesproduce by way of performing their function . 3 Representational Systems and Representational States The above classification is crude.. " " " " " " " "" " ' "" " . or mean..have original intentionality ... something they say or mean..make the systems in which they occur conscious the objects they of . " " . when speaking of the Representational Thesis... laptop computers.

has a close tie with behavior. Thoughts and beliefs are classified with . Mice who hear pianos being played do not believe pianos are being played .that the toast . Both are forms of natural representation . but not all such representations are conceptual representations . by a visual experience . judgment .a belief. though . is burning . For our purposes . too feeble to believe this even though their hearing is good enough to hear it . but the visual experience represents the piano playing in much different ways than does the belief . One can seeor hear a piano being played without believing a piano is being played .The Representational Character of SenseExperience classified together.7 Experiencesof piano playing do not require the concept of a piano (at least not in the same way as a belief or judgment requires it ). it normally brings with it an ability to say what one is aware of. These are different kinds of mental representation . They require no understanding of what a piano is or what it sounds like . thus . Believing is something else. in part . Until these experiences occur one has not seen or heard the piano . experiences This is correct. It requires the concept of a piano . A conceptual awarenessof facts . as far as it goes. This is not so with sensory awareness One can see or smell (and. Their understanding is . Even mice can seeand hear pianos being played . hearing by an auditory experience. it is I1t imp Orta to distinguish seeing and hearing from knowing and believing . For those who have language. or knowledge that the toast is burning . All representations are representations of (purported ) fact. be perceptually aware of) burning . and one can believe a piano is being played without seeing or hearing it being played . I can see Paul playing the piano and believe he is playing the piano . Seeing a piano being played is constituted . I assume. some understanding of what a piano is.

therefore . The cook and the mouse differ in what they think about what they smell.8 The difference between experiences of k (as F) and thoughts about k (that it is F). They differ in how they regard the smell. They both have some kind of sensory representation of this event. They both experience it . to ) describe someone as seeing ( being conscious of) the difference between A and B is to imply that the person sees (is conscious) that they differ . colors . A mouse in the kitchen (the one who heard the piano being played ) can smell. I take a moment . or feel. They may also differ in how the burning toast smells to them.g. however . burning toast) and events (e. numbers.differences. burning but not smell that it generally clear enough when we describe ourselves as conscious of concrete objects (e. ambiguity appears. of.. piano playing ). believe) that toast is burning . answers. and thus have sensory awareness of burning toast but it will not (like the cook) be aware (i . It will smell the toast . " What is that strange smell ?" might be the remark of someone who smells the toast burning but is ignorant of what toast is and what it means to burn . sizes. The mouse will have little or no conceptual awarenessof this event. When we use an abstract noun or phrase to describe what we see hear.1 Chapter toast while having little or no understanding of what toast is or what it means to burn . But they both smell it . we start describing ourselves as being aware of abstract objects . Thus .what we are aware or conscious .between sensory and conceptual representations of k. to remark on this ambiguity since it would otherwise muddy the discussion to follow .what is being described is normally a conceptual awareness of some (unspecified ) fact.g. problems.. not smell it as burning . The abstract noun phrase stands in for some factive clause (Dretske 1993 . To describe someone as being .e.

as aproblem .and . ' between experiences of k s blueness and beliefs or judg - . Cats are not color blind just because they ignore differences in color. If this is hard to imagine. etc. think of the perceiver as an animal. One has to . Until one conceptualizes it this way. A ' child or an animal might be visually aware of the shirt s color (their visual experience of the shirt being.without being conceptually aware . know what color the shirt was . at some conceptual level . as they say . and this may not be present . according to a representational theory. that anything is (or looks) blue.thus representing the color in some conceptual way (as blue .) It would be odd to describe a person as seeing (thus being aware) of the color of his shirt if the person did not .. thus.without sorting (or having any disposition to sort) the shirt with other blue objects. we must be careful not to describe an experience of a ' shirt ' s color as an awarenessor consciousnessof the shirt s color. These differences are important in thinking about ways of representing properties and. ' Likewise .to ' One can experience blue (the shirt s color ). I will try to keep these matters straight by distinguishing between sensory and conceptual representations of facts. suffused with blueness) without their knowing or thinking that the shirt is blue . Since the topic is our experience of objects not our beliefs about . the properties we are conscious of. in this sense be aware of blue . For this form of words implies9 a conceptual representation of the shirt ' s color. one is not (as we say) aware of the problem . the color of the sky.The Representational Character of SenseExperience aware of the color of his (blue) shirt is to imply that he or she is aware that the shirt is blue . to be aware of the problem it isn t enough to see (experience) the thing that is the problem (e. conceptualize what one sees the clogged drain . them.g . a clogged drain ). One has to see (the fact) that it is clogged.

and .1 Chapter ments that k (or something ) is (or looks ) blue . token states have two different sources for their indicator functions . is the basis of this ordinary . What . Call these systemicindicator functions (= functionss) and the representations they give rise to systemic representations (= representationss. and of burning toast without being ~onscious that anything is blue . its representational status . a thermometer ) is supposed to provide information about temperature. shape. say 32 . representations Nonetheless . Phenomenal awareness is a mode of awareness that does not require although it may. then. shape. familiar . If a system ) (e.from the system of which it is a state. and B is the state (e. State > B therefore representssthe temperature as being 32 . and texture ? In representational terms. and I will sometimes use it . what is the difference between an olfactory experience of toast burning and a belief or judgment that toast is burning ? Experiences (sensations ) of burning toast and beliefs . acquire its indicator function .g. function (functionS of indicating a temperature of 32 . hence. awareness One can be phenomenally conscious of a shirt ' s color. of a piano being played . and texture. that a piano is being played . (2) A token state may. I hope .distinction between an experience of color.. then B has the systemic . and a belief or judgment about color . (1) A state may derive its indicator function . The word " " phenomenal is often used to describe this sensory mode of awareness. on the other hand . in fact be accompanied by . not from the system of which it is a state.and .g .. mercury at such-and-such level ) that is supposed to carry the information that the temperature is. (thoughts ) that toast is burning are representations and all are particular (token ) states or events . or that something is burning .conceptual . in representational terms. but from .

systemic functions ) acquiredindicator functions (= function sa)' The indicator function sa (hence representationsa) of token states may . If I use the instrument in a car with normal up tires. You put the number 1601 there. it meansthat the temperature is 38 . I put the number 501 at position / / B of the pointer . indicating an axle rotation rate of N rpm . This mechanism was designed to be used in cars equipped with different sized tires. If we print the number " 381 at this / . or a different . Suppose we have a simple speedometer mechanism that represents vehicle speed by registering the rotation of the axle. this state of the system . it might acquire (or be given ) a special . the mercury s rising to this point signifies danger. indicator function . be different from their systemic functions (what they represents . and what it representsa need ' not be what it representss . point on the scale then each and every time the mercury reaches this point . my car is going 50 mph . When our axles are turning at a rate corresponding to a pointer position of B. If you use larger tires.if any. It nonetheless via . the manufacturer left the job of calibrating the face of the instrument to the user. If we print the word I/DANGERI / at this (and higher ) ' points .The Representational Character of SenseExperience the type of state of which it is a token. I calibrate the dial in one way . No matter what B systemically represents (what it is. An example should make this difference clear. supposed to indicate ). Since (at a given speed) the axle rotates more slowly with large tires than with small tires. )' As we shall see a token state might represent in both systemic and acquired ways. Pointer position B. Call such functions (functions that state types are assigned or acquire independently of their . your car is going 60 mph .that of . 1 / / you calibrate it differently . So we calibrate differently by assigning different numbers to this pointer position . has the same systemic function in both cars . by design of the system. .

varies from car to car. 60 mph in your car. what it representsa(about speed). Their experience of this sound may well be the same. sees an axle rotation of N rpm as a speed of 50 mph . It rep resentsa 50 mph in my car. B representss (i . this particular function of indicating state of the system has the systemic an axle rotating at N rpm . . " sees the same thing . of an axle rotation of " as haVing the same experiences ( . In my car state B representsa50 mph . Yet.1 Chapter different calibration . They hear the same sound . 246) describestwo of Pavlov s dogs. it representsa something different in my car than it does in your car." ' Walker (1983 p . Once calibrated " " . although this state representss the same thing in both systems. is an experience of) the same thing in all cars: an axle rotation of N .an axle rotation of N have the same but sees it as 60 mphOurspeedometers " but the " rise to different experience gives experience "beliefs . " calibrated. Their response is also the same: both salivate. what it representss in both your car and my car. Now imagine the two dogs hearing aclarinet playing middle C. is conditioned to salivate when middle Cisplayed on any instrument . has a different acquired function in our two cars. That is what this state means. their responses .. Nonetheless. To suggest ( by way of tendentious description ) the way this distinction is to be applied . Your speedometer differently . however. The other is conditioned to salivate to a clarinet playing any note..e. ' " " N) but as having different belief s (about speed). in your car it representsa60 mph . my speedometer as it were. Given the information ' (about axle rotation ) this system has the function of supplying . One . we can describe individual " " systems in which this perceptual mechanism is installed " viz . what it was designed (at the factory ) to do . State B has the same indicator functio ~ in all cars.

The way a belief represents the world . We can. not the things look . for this reason. the kind of visual experience k produces in me. properties are systemic Thought (conceptual states in general). are states whose representational . the dogs hear it differently . change our calibration . change what we see. experiences have their representational content fixed by the biological functions of the sensory systems of which they are states. We can change what we see something as what we . through learning . The way their experience representss the sound may well be the same. sound .l1 How an experience representss the world is fixed by the functions of the system of which it is a state. on the other hand . modular in Fodor ' s (1983) sense. I can change what I believe when I see k. As my description of these examples is intended to suggest . but I can' t much change the way k looks (phenomenally ) to me. take it to be . and feel at the most basic (phenomenal) level. The quality of a sensory state .even if we cannot. upon seeing it . but the way their experiencerepresentsait is different .is thus determined phylo genetically.12Through learning . is onto genetically determined . anyway ) hard -wired . As a result of different learning. . on the other hand . Since we inherit our sensory (as we . the other as the sound of a clarinet . properties are acquired As a result . not in the same way . since they are (at a fairly early age.The Representational Character of SenseExperience are mediated by different acquired representations . Experiences are. we cannot (not easily anyway ) change the representationals character of experience . might put it ) hears it as middle C. experiences are to be identified with states whose representational . This is why a representations of k as red (a sensation of redness) is different from a representationa of k as red (a belief that k is red) even though both are representations of k as red.

As their construction suggests. Another manufacturer produces a more sophisticated .1 Chapter The above example is a bit too simple-minded to reveal the full power of the representedsrepresenteda distinction to illuminate the sensation. The pointer positions of these instruments .cognition (experience belief) difference . these two sources of information are combined to yield a reliable measure of car speed no matter what size tires are used. Since the height of the axle above the road surface provides a useful measure of tire size. Pointer positions are determined by two sourcesof information : not just by the rate at which the axle is rotating . not about how fast the axle is rotating . in fact .thus have a speed-indicating systemic function and are therefore calibrated at the factory. The early phase in this informational process uses information about axle-rotation . but about how fast the car is going . but by the height of the axle above the road surface. no matter what size tires they use. but ( by integrating it with information about elevation ) it sacrifices this information in order to produce a final indication of speed. Information about axle rotations is thereby lost . "ucture of reveal its potential for capturing some of the sb actual sensory systems (I am thinking mainly of constancy mechanisms). indicate how fast the axle is rotating . these more sophisticated instruments are designed to provide information . speedometer It is designed to be used on all cars.driven by two sources of information . Perhaps therefore. a minor embellishment will better . the final representation (pointer position ) does not . es Although the information handling process in these fancy instruments use information about axle rotation to generate a representation of speed. This is an instrument .

In this case since we want the car to be in 3d . This.( having the func tio ~ of indicating a speed of 48 mph ) . at both 48 mph and 50 mph . however. if ) the instrument was to be used as a speedometer one would .The Representational Character of SenseExperience whose individual states (pointer positions) representsspeed. not as a . The state . what they have acquired (or been assigned13the job of indicating . having the functio ~ of indicating a speed of 50 mph ." But if speed is what this instrument " experiences" (= what its :various stateshave the systemic function of indicating " " ) what does it believe ? This depends on what the states of this instrument representa. If the various statesof the device acquire thesefunctionsa' then it will end up " experiencing" 50 mph and 48 mph ( these are the speeds these two states . might be the functions these states acquire if we used the device. in fact. assign functions to its various states in the way that reflected their systemic functions . unlike the crude one described above. not axle rotation . of course." Then B (having the functions of indicating a speed of 50 mph ) gets the same function a as the state ." " medium . we could partition the dial into several sub-divisions and label them " slow. would be given this function aby printing the number " 50" . Normally .at this position of the pointer } 4 We are. speedometer but as a mechanism to control an automatic transmission. " experiences" speed (not axle rotation ) even though the information -delivery process depends on information about axle rotation to generate this " experience . we give these two states gear the same function a. we pleaSe If we aren' t particularly interested in exactly how fast the car is going .the functiona' namely . of indicating a mediumspeed. free to assign any function sa . This instrument . B." and " fast.

At the conceptual level I operate with .. a few dozen categories for the colors I experience. then. I experience the differences (in the same way the speedometer described above registers a difference between 50 mph and 48 mph ). for making judgments about (= representationsa of) the differences I experience . so to speak. direction . At the sensory level I can discriminat hundreds of different colors. but . The information needed for such recalibration is already there in my experience. yes. sees both these . Learning calibrates me no more finely than I need to satisfy needs and carry out purposes. I am sensitive to (i . frequency . but it . looks like this : . It "experiencesa difference between 50 " " mph and 48 mph . that I could not " recalibrate " in a more discriminating way if the need arose. The taxonomy. objects affecting my senses I nonetheless have a limited for categorizing these sensory differences conceptual repertoire . speeds as medium (or 3d gear) speeds Conceptually. I treat these differences as different instances of the same kind . can discriminate ) all manner of differences in the light . at the conceptual level . of course. at best. The last example is meant to be suggestive of the way innate sensory systems might have the function of delivering fuformat ion about more or less continuous quantities surface reflectances . pressure. this " abstracts" from the difference between 48 system sensory and 50 mph . temperature .e. sound . speed . and chemistry of .Chapter 1 )' represents but representinga them both in the same way (as a 3d gear speed). At the level of experience. etc.) ( while the cognitive system might " chunk " this information into parcels that are more useful to the individual system' s needs and purposes. This is not to say.

."... not all representationss (not even all natural ) are mental . in our (second) speedometer system . They are the states whose functions it is to supply information to a cognitive system for calibration and use in the control and regulation of behavior example .The Representational Character of SenseExperience Representations indicator functions States having . process (the ones carrying information about axle rotation ....".. as noted earlier . representationss Experiences are those natural representationss that service the construction of representation sa' representationss that can be calibrated ( by learning ) to more effectively servIce an ' ' organism s needs and desires.exercising and reasoning " 16This is why . " Conceptual Sensory with acquired Stateswith systemic States indicator functions indicator functions Thoughts Experiences Sensations Judgments Beliefs Feelings A few clarificatory remarks are in order: (a) Experiences are representationss' but ...ISEvans (1982 chapter 7...... In order to qualify as conscious experiences ....... idea when he describes internal states whose content depends on their phylo genetically ancient connections with the motor system..... ... """' "" "" Conventional Conventional functions Natural Natural functions """ "" " . .let alone experiences . 4) expresses a similar . the early stages in the information . Evans requires that these content-bearing states serve " as input to what he calls a concept.....

comparatively modular in . therefore. so to speak. after learning .Chapter 1 for example) are not appropriate analogues of experience.Fodor ' s (1983 sense : ) hard -wired . of course. and so on. natural representationssthat are not mental. The gauge. We begin by hearing (experiencing) sounds and end by hearing (recognizing ) words . but doesn t . after the kind of calibration occurs that is involved in language learning ." Pointer positions representings pressure now (also) representa altitude . es Though these early process have a functions. 8 experiencesacquire an added representational dimension } Sensory adaptation may be an instance of such recalibration . informationally incapsulated. Their function is to supply information to representationssthat are. There is . We can use a pressure gauge (functions = delivering information about pressure) as an altimeter by recalibrating its face in " feet above sea level. We still hear sounds. This same transformation occurs as we learn to identify and recognize objectsand conditions we experience. The information vanish es (absorbed into information about speed ) before it reaches the relevant control structures that determine behavior (in this case . then .pointer positions ). this information is not available for calibration . This is not to say that the states (experiences) they produce are immune to changes in their representational status. make it available to calibrational " (cognitive ) processes.that of providing information about axle rotation (information the system needs to generate a representation of speed ). The instrument usesthis information ' .17 (b) Sensory systems have phylogenetic functions and are. no way this information can be given " conceptual" form . experiences pressure as altitude .and therefore represents axle rotation . These are . Just as we can recalibrate a speedometer when we . but . as it were .

but about timbre and ( binaurally ) the directional properties of sound. there seemsto be . (c) The states by means of which a representational system performs its informational functions have a structure that enables them to acquire functions without explicitly acquiring " " them. This . there is no corresponding represen tationala change: we still seeblue and yellow . . all the remaining indicator states (positions of " " the clock hands ) receive an implied indicator function . through some distinguishable evolutionary process All the indicator states may receive an implied indicator function by one state explicitly acquiring it . red and green. In the caseof vision . touch educates vision and objectsbegin to " look " to be where they can be grasped. there is a gradual yellowing of the macula (spot on the retina ) that changes the signals sent from the retina to the brain about the wavelength of light (example taken from Clark 1993 p . After wearing spectacles that make " " " " everything look 300 to the left . (d ) Audition provides information . red and green.The Representational Character of SenseExperience change the size of our tires (the state B acquiring the function of indicating a speed of 60 mph . as it did with normal tires. is one way of describing the experiential change that accompanies adaptation . Despite this constant change in the information repre170 ) sentationals states carry. As we age. asblue and yellow . for instance). a speed of 50 mph ) so also can sensory systems recalibrate when there is a change in the information an experience delivers . ' When there is this kind of structure among a system s indicator states (as there is in our experience of color and sound. at least . not only about pitch and intensity . Once we assign 12 to a certain position of the clock hands. . there is no reason to suppose that eachrepresentationals state has explicitly obtained its indicator function . not .

147 . Stein and Meredith 1993 Dretske 1981 pp . It would be more appropriate . We are sometimes talking about the properties objects are representeds as having (the phenomenal appearances). both describe (i. . 56. representa) how things seem to them "in the same way . are complex questions that . If I can use the speedometer analogy again. and movement (Fischback 1992 p . Zeki 1992 p . Justification for thinking of experience and thought in this way will come later. though . . 145. about the way things seem. often enough. form . things seem same to a vehicle speed of so-and-so many miles per hour. 71). If we are the talking about vehicle speed. to describe a given sense modality . . about what it is like to have an experience . the subjective quality of an experience the phenomena . straddle the sensation-cognition fence. then. but as having multiple indicator " functions . I am quite aware that questions about qualia. And there is no reason experiences " associated with a given sense modality (smell or vision ) might not . ) (e) Finally.g ...1 Chapter . exploit information from other modalities (see e. Once calibrated. trying to indicate how I intend to use the distinction between systemic and acquired representations. location. the first speedometer has axle-rotation qualia .e.19 . The second has " " speed qualia . . I am. different systems for infonnation about color. with these brief remarks. appearances are the way experience representss things to be. I return to these tangled issues in chapters 3 something and 5. at other times about the ones they are representedaas having (what the phenomenal appearancesprompt us to believe. not (as I have been doing ) as having the function of indicating F. or what we . as). on the basis of the experience are inclined to take . . separatepathways .

It does not represent the relationship (i. not the left rear tire . It isn t the function of this instrument to say what it is . Hang the same thennometer on you the wall and it tells you the temperature of the room. often enough . The instrument says how hot this is.infonnation about these external connections. living room . between the reference and the senseof a representation ) . not at the gauge. The same is true of all representational devices. Gauges do not supply they do not have the function of . whose pressure it registers.whose temperature it registers. supplying Representations have a sense (the properties they have the function of indicating ) and . but at the external connections between gauge and world that make it the right front tire . a reference (an object whose properties they represent). Two representations with the same sensecan have different referents. Put one in your coffee and it tells how hot your coffee is.. Put it in your mouth and it indicates something about you ' whether you have a fever. If you want to know which tire (if the gauge represents which topic it is commenting on. but the sense does not determine the reference. any) you have to look . or you . . where this is whatever medium the thermometer is in .The Representational Character of SenseExperience 4 Represented Properties and Represented Objects Ordinary thennometers represent the temperature of whatever medium they are in .coffee. Pressure gauges do not tell you which (if any) tire it is that they represent the pressure of.e. its being in this ) that makes it this rather than that it represents the temperature of. is the same distinction Nelson Goodman (1976 was getting at in his contrast between a picture of a black horse ("black horse" here specifying the object the picture is . The difference between represented object and represented property .

Some pictures of black horses do not represent the black horse as a black a so-and -so. Although my present black -horse experience is an experience of a black horse (I am.1 Chapter a picture of) and a black-horse picture ("black -horse" here specifying what the picture depicts the object as). 1987. Or imagine taking a picture of a black horse disguised to look like a brown camel. This picture of a . which . dreams. Although token experiencesmay be individuated ) by being experiences of a particular object. say. black horse is a brown -camel picture . they are de re modes of representation (Burge 1977. imagination ). I can nonethelesshave exactly the same type of experience a .g . but a certain external causal or contextual relation I will designate as C. black-horse experience without its being an experience of a . Recanati 1993 .. with extensional content. Experiencesare like pictures in this regard. Nothing guarantees that black -horse pictures are pictures of black horses. a spot in the distance . The object a representation is a representation of is not determined by the properties that object is represented as having . The picture depicts the black horse as. black horse or. The extensional content is what object it is that is being represented. 14) make the same distinction by contrasting explicit . It could be anything . nothing the representation says. seeing a black horse). that is. Bach 1986.20 What determines the reference for a de re mode of representation (the object it is a representation of) is not how it is represented . k. indeed. The explicit content is the way this object is being represensted. Imagine taking a picture of a black horse at a great distance. Lloyd 1989 p .21 There is nothing in the content of the representation . They are not black-horse pictures . an experience of any object at all . tokens of the same experience type can occur with a different k or with no object at all (hallucinations. Others (e.

then. to my car that it can do what other speedometers cannot do . They represent k as being blue . ). representing different objects . not your car. say something (whether truly or falsely) about my car.viz . at different times. Other speedometers cannot even say something false about my car. then we can say that .representational fact about S. ) necessarily As I am using terms. the case may be) the speed of my car. the fact that 5 represents k is not. and k stands in C to 5. but they do so without representing it to be k that they represent to be blue. in part . changeswhat 5 represents (what objectit represents but not what 5 says (represents about it . then 5 will indicate the F of k. The speedometer in my car is connected to the axle of my car. If k is the object whose F is represented at time t . C makes it the case that k is the object 5 represents as blue. the fact that it is k (rather than some other object or no object at all ) that stands in relation C to the representation is not what the representation represents. It therefore represents (or misrepresents as . when a representational system 5 is functioning properly . cannot) represent context. 5 represents k. at time t .at least not a pure . Representations do not (indeed . changing the object that stands in relation C to 5. Changing the way a system is deployed . 5ince the veridicality of an experience depends on its reference . C is the relation such that. De re modes of representation have their reference determined contextually.on what object (if any ) it is an experience of . by context (C). It is a . At other times 5 may be deployed differently thus.The Representational Character of SenseExperience makes it about this object rather than that object or no object at all .the veridicality of experience is determined .. Nonetheless. C ~ ereby helps to determine whether the representation (that k is blue ) is veridical or not. not your car. It is in virtue of this special relation . C. by the relation I am here calling C.

for some F. what 22 object (or whether there is an object) one is experiencing. When we describe 5 (whose function it is to indicate F) as representing k. thiS implies that. But it also implies something that is not a representational fact. then . " " in the Clearly . part not .an important fact for many practical purposes . 5 represents (possibly misrepresen ) the F of k. therefore.1 Chapter fact about a representation . say) that is F. but it does not have the job of indicating does not therefore represent . at least not by introspection .hence. (according to the Representational Thesis) representational. Hybrid representational facts (that 5 represents k) are mixtures of representational facts and facts about representations. That 5 represents k.viz . This will be important when we discuss in chapter 2) introspective knowledge . implies a representational fact. 50 facts about the object of representation are hybrid facts . 5 represents the F of k.which objects . that k stands in relation C to 5. but that this (whatever it is that stands in C to the instrument ) is going 25 mph . This is why one cannot know .facts. The fact that there is something in the world that is going 25 mph does not mean that a " " speedometer registering 25 mph says something true about the world . H representational system 5 says anything at .. It may turn out that 5 is misrepresenting something to be F when there is something (in the next room. For what is ( known by introspection are mental . For what the speedometer " says" is not that there is something in the world going 25 mph . my use of the word something description of 5 as representing something as being blue is not an existential quantifier .but it is not a fact that has to do exclusively with what the system has the function of indicating . for some F.part representati . 5 has the function of indicating the F of those objects which stand in C to it .or even whether there is an object.that.that stands in C to it .

. It should be understood . An experience can misrepresent by (1) saying something false... owing to a failure of presupposition . represents something as being (say) blue---it is that this is blue where this is whatever object (if there is one) to which S stands in the Crelation. this merely shows something about the nature of sense experience that we knew all along : viz . ""' " -. something stands in for a failed indexical . " " however. that when there is no object.The Representational Character of SenseExperience all when it represents color .that this is blue when there is no this that is not blue.and . misrepresents) this as being blue when there is no this.. by saying that this is blue when it (the object of representation ).... . I shall continue to describeS as representing something to be blue . and (2) by saying what is neither true " " nor false . When there is no object of representation. but there is no object that S represents to be blue / Veridical The objectS represents to be blue is blue '" Misrepresentation The objectS represents to be blue is not blue .. then. nothing that S represents to be blue . what S represents to be so (the content of the represen~ tion ) is neither true nor false.e. We thus have two different kinds of misrepresentation: S representscolor S represents something to be (say) blue / . If there is no object of representation.. thus . Object 01representation There is aI) objectS represents to be blue No object of representation There is no objectS represents to be blue Misrepresentation S represents something to be blue .. If we take this to mean that . S represents (i .is not blue .. that misrepresentation takes two forms.

mental events. In each case we find that a representational account of the mind provides a satisfying explanation of intentionality . The Power to MisrepresentChisholm (1957 describes this ) as the first mark of intentionality . but it turns out . Indeed. exactly. Unlike symbols. of part of the natural order. it is a power that natural representations do not get from us. has thoughts about him and desires for him . . hears him . diagrams. one can seewhere intentionality comes from and why it is there. as FOdor (1987 suggestsit must . and whether or not he was right about its being a feature of all . and instruments . is the power or capacity of one state of affairs to refer to or be about anotherS seesTom. as manifestations of overall biological and developmental design. moreover.and. most philosophers take intentional characteristics (variously understood ) to be distinctive of a great many mental phenomena . those about senseexperience . Brentano meant by inten intentionality tionality . By conceiving of mental facts . Beliefs and experiences have the power to " say" or " mean" that k is F when k is not F. to be really something eIse. we can understand where they get this power . What follows is a brief catalog of those aspects of intentionality that have figured most prominently in the recent literature . Whatever. ) 1. 2. Aboutness This. If we understand representations as states having indicator functions (systemic or acquired). They are states that exhibit the first mark of intentionality in an original (nonderived ) form .1 Chapter 5 Intentionality Brentano (1874 conjectured that a mark of the mental was ) . In each case intentionality is real e~ough . These are thingsS cannot do unlessS occupies states that . gauges. and only. in particular . they have the power to say this (see above) even when there is no k.

I could be experiencing these properties (during a dream or hallucination ) when there was no golf ball . But if Tom stands in the right causal relationship to these experiences . . after all . Echoing Goodman. about the manifold s having that pressure (a condition or state). My experience is of or about these properties as much as it is of the golf ball . no object that had the properties I experience. In representing the pressure in an intake manifold . as what they refer to.experiencein order to be an experience of Tom. S' s experiencesof Tom may not depict who (or what ) it is they are experiences of. movement . ' are experiences of him . state of affairs . and think about their properties . and dimpled texture . the manifold (the one standing in relation C to the gauge) would be the object the gauge was conscious of. as what they are thoughts about experiencesof and desiresfor . Even the simplest measuring instruments exhibit these dimensions of aboutness albeit in a derived way . Tom. If pressure gauges were conscious . having a pressure of 14 psi would be the property the gauge perceived it to have . It is not only about the manifold (an object ). I see a golf ball experience and . We not only experience and think about objects we also .The Representational Character of SenseExperience have Tom as their object.. in seeing it . and texture. I experience its color. S s experience they need not be a Tom. see Tom on a foggy night and mistake him for someone else. the experience is still an experience of whiteness. In seeing . movement . roundness. therefore. and its having a pressure of 14 psi would be the condition . it is about the pressure in it (a ' property ) and. if their intentionality was original rather than conventional . or fact the gauge was aware of. shape. Whether or not there is an object. a pressure gauge " " says something about the manifold . . S can.

as stationary not moving . This. yes. If 5 did not have this function . The system to which k stands in this relation must be a representational system. The behavior of the balance " says" or represents something about the objects on it ." A drunk ' s behavior is not abouthis blood chemistry becausewe can use his behavior to tell something about just the chemistry of his blood . 5 would not be about k. taste it . doesn' t make the teeter-totter ' s state aboutthe children in any relevant senseof " about. but in desiring an apple I desire to eat it . What makes 5 aboutthe object k is not simply the fact that k stands in relation C to 5. 5tanding in relation C to 5 is necessary . if there wasn ' t something 5 was supposed indicate about k. this should not be taken to mean that this dimension of intentionality is being equated with a causal or informational relation (whatever the relation C comes down to ) between two objects. It " says" that one is heavier than the red not blue. as round not square . This is not to say that we could not learn something (exactly what we learn from the balance) about the relative weight of the children from the behavior of the teeter-totter. A playground teeter-totter ' s position stands in the same relation to the children playing on it that a beam balance stands to the objects whose weights it has the function to compare. and reference. one having thefunction of indicating the F of those objectswhich stand in C to it . but not sufficient . AspectualShape thinking about a ball I think about it in In one way rather than another. is fixed by contextual relation C. then even to if k stood in relation C to 5. 3. I can desire an apple. to make 5 about k. in turn . throw it . or simply have it .Chapter 1 Though aboutnessis here being understood in terms of the reference of a representational state. hold it . These are the aspects . These are the aspects under which I think of the ball . though . look at it . The teeter-totter does not.

Our mental states not only have a reference an aboutness an object (or purported object) that . It can.rough and cold. they represent that object in one way rather than another. Experiences are of objects yes. therefore. Doing X doesn t automatically give you the function of doing X. The objects I touch feel a certain way to me . see smell. Even when remains. When an object is represented. This follows immediately from the fact that the experienced object is simply that object. round . and having . Aspects come finely individuated and this .red. One cannot . if there is one. there is no object. too. An instrument can have a pressure-indicating function without having a temperature -indicating function even when it cannot deliver information about pressure without delivering information about temperature. one). an object without experiencing it under some aspect. or taste an object without experiencing " it as having determinate properties . the other. If we suppose that. whose determinable properties the sensory system represents. . the aspect The same is true of the objects we experience. A pressure gauge is not a thermometer just becauseit delivers information about temperature (though we could certainly ' use it as. nonetheless represent one without representing . and stationary. and therefore make it . there is always an aspect under which it is represented. in accordancewith gas laws. is part of what is meant in speaking of aspectual shape. an instrument cannot help delivering information about temperature (that the gas is 38 C) in delivering information about pressure (that it is the 14 psi ). forms their topic . The objects I see looKa certain way to me .The Representational Character of SenseExperience " " ) (the language of aspects is taken from Searle 1992 under which I desire the apple. . the pressure of the gas in a (constant volume ) tank cannot be 14 psi without the gas having a temperature of (say) 38 C. but one cannot experience .

yet. Miller (1984 explains Husserl ' s technical ) term " noema" in this way : Since act hasdirectedness an of regardless whetheror not it hasan object. a contextual. however . therefore (according to the Representational Thesis) the aspectual character of experience and thoUght. It is supposed to be part of what it is like to have that experience.( ) . whether or not they are about something in the sense described above.Chapter 1 the function of providing temperature information is what is needed to represent temperature . both be black -horse . something' other than the object must be that which " accounts the act s directedness for . experiences and.directed ness. Two experiences can be subjectively indistinguishable . Directedness The objective reference of an experience what object the experience is an experience of. one be an experience of a black horse. Experiencesare supposed to exhibit this kind of intentionality . experience and its reference .23It is the explanation for why experiencing a certain pressure is different from experiencing a certain temperature even when these aspects cannot be separated in the object one experiences . Following Fo . H I understand it (and I' m not sure I do ) it is the quality of experienceby means of which experience purports to refer to some object or fact. Husserl is the acts noema16 ' . the other arelational . to be a quality of experience that is supposed to be intrinsic to the experience. 4. Directedness on the other hand . It is the specificity of indicator functions to this property rather than that property (even when the properties are nomically correlated ) that explains the aspectual profile of representations and. a connection between the noema of an :llesdal ( 1969 ) . This " something according to . property of the experience.whether or not they have an objective referent . There is . is supposed .

it sounds as though there is supposed to be some subjective quality of experience that is such that if I am (in fact) seeing a black horse. for it to have a noema (p . Neither are experiences I' m not sure. When I am experiencing an object . . . if any. . Representationsare not like that. Miller (1984 p . not within " the subjective " accessible character of the experience itself . For an experience to be directed is. once again . then . 31). If this means. simply . nothing in my experience of it determines which object I' m experiencing anymore than "there is something about a gauge' s representation of a tire ' s pressure that determines which tire it is registering the pressure of . 69) says that the part of the noema (or noe . directedness comes out the same as aboutness as already defined. It does not represent an additional aspect of inten tionality . If this is what directedness is supposed to be. 17 takes the noema to be that which determines ) which object (if there is one) is the object of the experience. it seemsclear that experiencesdo not have it . it determines which black horse I am seeing.The Representational Character of SenseExperience Miller (p . that what object. 758). that I understand noema and the directedness they are supposed to confer on experiences. is to say that they have a subjectively accessiblequality that is such that if they have an objective reference (some object they are o . as it seems to mean. then .f) which object it is. matic sinn: the reference determining aspect of the noema) which determines the purported object of the act (in abstraction from properties ) is indexical (also Christensen 1993 p . To say that experiences have this quality of directedness. though . this quality determines . an experience is directed at is determined . however . but by the contextual or relational properties of the experience " " (what I called relation C in 3). " If we understand directedness in this way.

y. It does not make much sense to suppose that the Qbject being represented as here is really being represented .z]. not with objectreference (if any) of the experience but with certain special . that this purported object is represented as having properties . It is the object that is there . as there I can. the representation of it as being here determines which object (if there is one) is being represented that way . Something is represented as being here and something is represented as being there. it likens experience to something like a story. determines which object it is that is being represented as being there. This is especially true in vision where experience represents objects as being in particular places . In thinking about a story we can be thinking about : (1) the words that tell the story or (2) what those words mean.z ] .y. But if I represent an object as being here (where it is understood that only one object can be represented as being here). It is the one being represented as here. If this is what directedness comes down to. of course. then directednesshas to do. the story they tell .y. If there is an object there. Even simple instruments can have this kind of directed" ness (in a derived way ) . the properties being represented apply uniquely to only one object. 6 Mind and Brain : The Whereabouts of Experience Insofar as a representational account of sense experience depicts experience as representational. Call the first the .Chapter 1 Sometimes. at a given time .at [x. represent more than one object as red so that an object' s being represented as red will not determine which object (if any) I am representing. places that (given the way we represent objects) can only be occupied by one object.z] . Radar is " directed at whatever object it represents as being at coordinates [x. the representation of it as being there. at [x.

The Representational Character of SenseExperience story-vehicle and the second the story-content. that there is something red over here and something triangular over there. Aside . how could they possibly explain my trip to the closet? The itch I feel has to be in me to explain why I scratch. if we sometimes do what we do becauseof what we think and experience then beliefs and experiences . for example . from their role in explaining behavior (some philosophers are sceptical of this ).but . I ceaseto experience . are the experience vehicles.that say or mean .e. Thoughts and experienceshave to be in the head (or at least somewhere in the body ) if our having them is (sometimes) to explain why we act the way we do. though .e. story vehicles) are in books. Stories (i . for instance. the states that tell (express) a story about the world . have to be in the systems whose behavior they explain . there is this kind of ambiguity . the ambiguity between representational vehicle and representational content . the physical states that have a representational -content. If my thought (that my clothes are in the closet) and desire (to get dressed) were not in me at least states or conditions of me. experienceshave to be in us in order to explain why . thoughts and experiences are in heads. of course. The maiden was rescued by a knight in shining armor . though this story is in a book .. a knight . or a daring rescue in (i. but what happens in the story (content ) does not happen in a book . no one expects to find a maiden .that episode was certainly in the story. between the covers of) a book. If mental states are to have an explanatory role. I go to the closet because I want to get dressed and think that is where my clothes are. Whenever one talks about representations. Our talk about thoughts and experiencesexhibits this kind of ambiguity . In the same way that stories are in books. It need not happen anywhere .. What is in the head.

Just as what makes a story the story it is is what it is a story about (i. Stories about blue dogs . make the experience the kind of experience it is. . And this something has to be in me or a state of me.f to be well camouflaged. We do not find the content of experience the properties that . are the same sort of thing we find when we look in books: representational vehicles that in no way resemblewhat these representations represent. what we find by looking in the brain of a person experiencing blue dogs is neither blue nor doglike .e.the blue dog vehicles . an experience . .1 Chapter a room full of people when I close my eyes. ) looked into the brain of his patient . the facts that make what is in the head mental. What we find . in particular .. We find the experience vehicles (see . the story -content). so what makes a mental state . he did not observe the colors and sounds the patient reported experiencing . cheesy brain matter. 118 for the same point ) . 1 there are sensationsand feelings in there. are electrical and chemical activity in gray. It is all black and white . but something ceasesto exist when I close my eyes. a patient who was (as a result of electrical stimulation by Penfield ) experiencing a variety of sensations. What we find . in other words . the facts that convert elec- . Lycan 1990 p . Why else would its existence be so utterly dependent on the position of my eyelids? But though the thoughts and experiences have to be in us.are neither blue nor doglike . According to the Representational Thesis. Just look in the book.and. 111 Tye 1991. The room full of people does not vanish . we all know what we will find when we look inside the skuU. We can have stories about blue dogs (blue -dog stories ) and blue -dog experiences . Just so. they seem .the experience it is is what it is an experience of. When Wilder Penfield (1957 1959 . This is exactly what one would expect on a representational theory of experience. p . instead .

What sometimes helps to obscure the difference between representation. This enables users to tell . What makes a certain pattern of electrical activity in the cortex into a blue. has the function of indicating . and. revolutions per minute . One could as well hope to figure out what . represents Instruments come with their representational " faces" calibrated in degrees Fahrenheit . at what is in. are facts that are not identifiable by looking .content . at the representations. function of indicating . whether there is meaning.there at all. if anything . One merely looks at the representation (e. One could as well hope to figure out what . but the experience -content isn' t there. just like stories. ancient scratchings on a cave wall meant by peering at them with a magnifying glass. The experience vehicle is there. it is the content that makes the experience (the story ) the experience (the story ) it is.The Representational Character of SenseExperience trical and chemical activity in the cortex into blue. a voltage difference between two electrical contacts in a computer means (represents. indeed . not only what the instrument is designed to measure.g . but how it represents that quantity .. and so on . the head. ohms. if anything . signifies ) by taking ever more exacting measurements of the voltage . What something represents what it . miles per hour .dog experiences . That isn ' t going to tell you what anything means or .dog experience is a fact about what this activity represents what it has the . a " " pointer pointing at the numeral 900 on the face of an .vehicle and representation-content is a confusion (fostered by conventional measuring instruments ) about how one goes about determmmg what a representation . pounds per square inch. According to the Representational Thesis. that is why looking at the brain of a person having an experience isn ' t going to tell you what that person is experiencing. exclusively . are not facts one can discover by loo ~ g in the cortex.

In this sense the mind isn t in the head . on the one hand. and. by looking in (or at) :ee the instrument . represents . An animal s brain doesn t come neatly labeled in this way .. on the other hand. . You couldn t do it with an instrument either without the help of labels.Chapter 1 ' instrument labeled TACHOMETER and calibrated in rpm s) " what it " to discover what the instrument is saying." S what its experience is like . but ' their content is not. what story it is telling about the engine. see what the instrument is seeing " " . This obscures the distinction between the representation itself . as it were. tells you what the representation-content is . Looking at " " the representation -vehicle . With everything neatly labeled in this " way. the story being told . one can.e. what is being represented (= content).that the engine is turning at 900 rpm . something (the vehicle ) that is in the instrument . story contents) are in books. The representations are there. something ' ' that is not in the instrument . by looking at (or in ) it . That is why you cannot tell . what the animal whose brain it is is ' experiencing. something (the vehicle) that tells a story about the engine. any more than stories (i . the pointer pointing at 900.

knowledge the mind has of itself . How do I know it does? These are not questions about how I know that today is . Natsoulas (1983 defines ) one form of consciousness. Introspection is the process by means of which we come by such knowledge . There is nothing from which I infer that A looks longer than a privileged ability to be noninferentially aware of (all or some of) one' s current mental occurrences We seem to have .2 Introspection I think today is Wednesday. How do I know I think this? From this angle A looks longer than B. this ability . In telling you what I believe I do not have to figure this out (as you might have to) from what I say or do.of what I think and how things look to me. They are questions about what John Locke called reflection the notice that the mind takes of its own operations and states. It just does. That we have introspective knowledge is obvious . Philosophers may disagree about the mode of access we have to our own thoughts and feelings and whether the . I will call the mind ' s direct knowledge of itself introspective knowledge . They are questions about how I know something about myself. Wednesday or that A is longer than B. That is somehow given .reflective consciousness.

The objects and events one perceives to learn these facts. but by an awareness of A and B.2 A representational theory of the mind is an externalist theory of the mind and. facts about one s mental life thus (on a representational theory ) representational facts. an awarenessof external (physical ) objects. not by an awareness of the experience that represents A as longer than B. what IS going on in the mind ? . but most are willing to concede that there is a dramatic difference between the way you know how I feel and the way I know it . by their indicator " function ) then how is it possible to know . by looking inward . to be sure. the objects the experience is an experience of. about internal representations . On a representational theory of the mind . What one comes to know by ' introspection are. as it struck Wittgenstein (1974 412) as the . are seldom internal and never mental. There is no Imers monitoring appeal to---no longer a need forinternalS Ca the clockwork of the mind . not by the intrinsic character of the events occurring inside.1 One becomes aware of representational facts by an awarenessof physical objects. If ) mental facts are constituted . One learns that A looks longer than B. These facts are facts.g .2 Otapter authority we enjoy is that of infallibility or something less. faces the same problems that all externalist theories confront (Boghossian 1989 . therefore. but by the relationsthese internal events bear to external affairs (e. " queere. It gives an account of introspective knowledge that is compellingly simple and plausible . if you will ." by introspection .one.stthing there could be a representational theory dispels the mystery. introspection becomesan instance of displaced perception knowledge of internal (mental ) facts via . If the idea of the mind turning its attention onto itself " strikes . The problem is to explain how we come by such knowledge and what gives us this first person authority .. however.

that I have gained five pounds by looking at the bathroom scale there is a conceptual representation of me as having . there would be no point in doing so. One looks at a gauge on the dashboard in order to see how much gas remains in the gas tank . objects the pilot actually sees. During instrument landing all the important facts are outside. In such casesone comes to know that k is F. perceptual fact in another. The fact I learn is a fact about by introspection are not there . k. with the aid of mirrors .3 Perceptual displacement. This problem " evaporates once one understands that introspection is not a process in which one looks inward . but some other object.sees that k is F. The mental facts one seeks to . occurs when there is conceptual.are inside. that one does not see One listens for the sound of a timer to find out . Even if we could look inside . The perceptual fact is displaced from the perceptual object.seeing that k is F by seeing not .by seeing and hearing .perceptual object in one place.In b' ospection That would be like trying to determine . not k itself . but the instruments and gauges . a tank located beneath the car. but no corresponding sensory representation of k4 In seeing .a fact about an object.othat I weigh 170pounds . whether one was a father or a citizen of the United " States . when a cake in the oven is done and reads the newspapers or watches television to seewhat is happening on the other side of the world . 1 DisplacedPerception Many of the facts we come to know through ordinary sense perception are facts about objectswe do not perceive.the . This pattern . h. .is familiar . but h. I seehow much I weigh by looking at the bathroom scale on which I stand. The object I seeis the bathroom scale.

Perceptual displacement enlarges the number of facts one perceives without a corresponding enlargement in the number ' of objects one perceives.).Chapter 2 gained five pounds . Those who perceive h without an appropriate connecting belief (as I will call it ) . Perceptual displacement requires belief (knowledge ? jus 6 that an tified ~elief? ) appropriate connection exists between the properties of the perceived object (h) and the properties of the target object (k).. orientation size. one' s knowledge of what the objects one can already seesignify . Those who do not believe this will not ' see that k is F when they see that h is G. ) believe that the positien of its pointer indicates the weight of the person standing on it . If you don t know ' t therefore that the object you are standing on is a scale don ( .g. ) objects one can see .e.the object the perceptual belief is a belief about. not me. One must believe that h would not be G unless k were (probably ) F. etc. on a representational theory of perception .will perceive the same objectsbut not the same facts. but a sensory representation of the scale. The properties sensuously represented (i . They help (becausethey increase' the number and kind of . ) representeds are properties of the scale (its color. you will not see how much you weigh by looking at it . representeda ) having 5 gained 5 pounds . The property .a belief about the relationship between h and k. one does not need microscopes and telescopes..e. about the objects one cannot see This is what connecting -confirmed theories) provide .. not by seeing more objects but by expanding . beliefs (e. but they aren t necessary One can also seemore facts. conceptually represented (i . well Before turning to the demonstration that .is a property of me. To expand one s observational powers . not properties of me. not the scale. introspective knowledge (at least of certain mental states) is a form of displaced percep- . shape.

but not as a representation. A picture of a picture that does not represent the picture as a picture7 is not a metarepresentation of the picture . The copy would . metarepresentational Pylyshyn (1978 p . They are representations of them as representations. If we think of a photograph as a (pictorial ) representation (of whomever or whatever it is a picture ). representations of representations.Introspection tion . in this se~ e. mental facts. Since mental facts (according to the thesis I am promoting ) are representational facts. I thus produce a metarepresentation . A description of the picture as weighing 2 grams is not a that clearly identified the shape and arrangement of the letters . a word about the facts that introspective knowledge is knowledge of. a photocopy of this page . as of Clyde where " " Clyde is understood to be a description of what the picture is a picture of . It is. Introspective knowledge is knowledge of the mind . I represent the object as a representation . Using this terminology in a different way. . and as having certain colored areas on its surface. as rectangular.i. In describing it as a picture of Clyde . introspective knowledge is a (conceptual ) representation of a representation. Uoyd (1989 p .e. Metarepresentations are not merely . (see and Ferner (1991 p .would be a metarepresentation according to the way I understand Uoyd . we can represent this photograph in a variety of different ways .its representational content ) . Thus. We can represent it as a piece of paper. 35 .. 169 calls a metarepresentation any representation of a representation ). simply .of the fact that something (else) is a representation or has a certain representational content. as a piece of paper weighing 2 graIns. We can also represent it as a picture of Clyde (or. It represents a representation. 593) . in which the representing properties (of the representation ) are explicitly represented.

a representation of something (a thought . of. 2 Knowing Another's Mind ' If introspection is direct knowledge of one s own mind . The photocopy would not . One comes to know (the fact) that one is experiencing blue by or not they are meaningful . introspective knowl - . The photocopy would not represent the marks on the ' page as representing anything . an experience) as a thought or an experience or (more specifically ) a thought about this or an experience of that. but some displaced object. say) is conceptually represented as an experience of blue via a sensory representation not of the experience but of some .Chapter 2 explicitly represent the properties of the symbols their . and we restrict attention to sensory affairs.e.i. If E is an experience (sensory representation ) of blue . As we shall see this displaced object . introspective knowledge is a speciesof displaced perception. then. be a metarepresentation as I am using this word . therefore . other object. That isn t the function of the picture . is . Introspective knowledge . then introspective knowledge of this experience is a conceptual representation8 of it as an experience of blue (or of color ). a metarepresentation . Its function is to copy the symbols on the page wheth . If . being a form of representation. however . the blue object one sees Introspective knowledge of E requires no other sensory representation of objects than those already being represented by Ethe experience one comesto know about. It will copy gibberish as faithfully as meaningful prose. is (typically ) the object the experience of blue is an experience . it is an instance in which an experience (of blue .by means of which individual shapes and arrangement they represent what they represent.. not the experience of blue .

5ince we are conceiving of how things look . and (2) for some F. Let me start with an even simpler question. indeed. As we have already seen (chapter 1. pressure. weight ) of k. and this pointer presently occupies position P. " Before trying to answer questions about how one can know one' s own mind . speed. or so I the way objects (if there are any) are systemically represented knowledge of sense experience is knowledge of how things are being representedsby the person in whom they are representeds .Introspection edge is direct knowledge of facts about phenomenal appearances. whether it representsan object at all. feel . (1). what it currently represents the F of k as. To say that 5 represents k is to say both that (1) k stands in the C relation to 5. will also appear child ishly obvious. 5ince the topic (given the thesis I am defending ) is the epistemology of representation. temperature. It is important to notice that the question is not how one might come to know what objects an instrument represents or. If an ordinary measuring instrument represents the F (speed. 4). then my answer. how would one go about finding out what it currently " says" about the F of k. and smell as modes of sensory representation . That there is an object whose pressure. For if it does. that 5 represents k is a hybrid representational fact. sound . or curvature 5 represents is not itself a representational fact about 5. we can gain some leverage by asking how one might figure out how instruments represent the objects they represent. however. If the instrument performs its indicating chores by means of a pointer . . 5 representsthe F of whatever objects satisfy (1). it will help if we begin by asking how one might know the corresponding facts about another mind . how does one find out what position P says or means about F? I hope this question sounds childishly simple .

or that there is an object. To know the " mind " of an instrument is to know (2) but not (1). How . The face of the gauge is. Even if it is not connected to anything .be concrete suppose the question is a question about a .they represent as so and so." The gauge is. a mental. therefore. but it is a method that depends on our trusting someone to tell us what pointer positions are supposed to indicate . If an F representational system knows there is an object it is representing to be F. how they represent what (if anything ) they represent. then. and representational systems do not have the function of indicating what object.that they represent in this way . As what ?9 It might seem that the answer to this question is simple : just look . it will not ion. side. what determinate value of F they represent it as having. O misrepresenting) something to be 14 psi } That . particular pressure gauge.fact. know it by'introspect Since we are interested in introspective we can set this kind of knowledge to one knowledge . the facts that define what was " " going on in the representational mind of an instrument ? To . if they could know their own representational states we would expect them to know . representing whatever it is connected to as having a pressure of 14 psi . therefore. after all .not . but not what object. and the pointer is pointing at the number " 14. after all . . How is it representing the pressure of the tank to which it is connected.or that there is an object.e. is " 14 " means . S represents what it has the function of indicating . would we find out how an instrument represents ? the objects (if any) it represents How would we find out the representational facts. it certainly is representing (i. what the pointer pointing at psi This is a way of finding out what the gauge is saying..2 Chapter is not a representational fact. If instruments could introspect. marked off in pounds / square inch (psi ).

to make it more like the question we actually face in asking about the representational states of living systems. So. There is nothing there (or the numbers are no longer legible ) to tell a curious onlooker what the pointer positions mean. but this is not a courtesy that nature extends to us in the caseof natural systems of representation.Introspection Manufacturers do this ( by printing numbers . in fact. that it is. going to tell you what pointer position P means. the various pointer positions representspressure. a pressure gauge. this would not tell the gauge what it needs to know to know how it is representing k. of course. it would do it no good to " " peer inwardly at its own representation of the pressure. but " " you cannot find out what it says about pressure by looking at it . The . let us suppose there are no explanatory or interpretive symbols on the gauge. Wormation about how a representation of k represents k is never available in .on the face). pointer is presently occupying position . therefore .l1and that . Even if it could somehow " feel" where its pointer is pointing . The pointer occupying position P is the representation . We have to find this out for ourselves. what is going on in its representational " mind . we will not know how the instrument is " " experiencing the world .P Our question is: what does P mean. on.nothing to tell curious neurobiologists what (if anything ) the electrical and chemical activity in there means. or around the representation of k.12If this gauge " wanted " to find out how it was representing the pressure in k. in order to impart some realism to our investigation . . I will assume that the gauge has a function . what does it represent the pressure as? Until we know this . There are no convenient numbers in the brain ." It should be obvious that the way to answer this question is not to look more carefully at pointer position P That isn ' t .

e. does one go about determffiing what P means? If we knew the instrument was working properly (as it was designed to work ) and knew it was connected in the right way (i . This. find out what pressure it has the function of indicating . External ' observers are often better positioned to observe a system s internal representations than is the system itself. yes. According to the Representational Thesis. If the pressure is 14 psi. One would simply determine . are inside . How . though .Chapter 2 The idea that introspective knowledge is achieved by " " " some Lockean " inner sense a looking ( spicere ) inward . by independent means. then. psi .. then. That will tell us what Prepresentss the pressure to be. To find out what pressure it has the function of indicating . a misunderstand of this fundamental point . Neither would it tell the instrument . does not tell us what we want to know . " intra " at internal ) ( representations. It cannot. will be what P means . what information P has the functio ~ of providing . That would be what P means. would have to know if it was to know how it is representing things . what the pressure in k is when the gauge is in state P. but perceiving them will not tell you what you need to know to reach knowledge of the mind . find out what the pressure is when the gauge is in state P and functioning the way it is supposed to function . then. we will know what the system itself . is. To find out what P means about pressure. That. then the state it occupies must have the function of indicating 14 . what we have ' to find out about the gauge s current pointer reading is what information P is supposed to carry about pressure. then there would be a straightforward way of telling what P means. by C) to k. if the gauge is doing what it has the function of doing . if it could introspect. The representations . Once we know what P says about pressure. We can see ' the instrument s pointer positions . I suspect.

Even when no one tells us what P means (or we can' t interpret the symbols). therefore . the kind of information it was " designed " (by natural selection ) to provide . there is no special problem . So assign meanings to the states by determining what the actual values of F are when the system is working right . and with the kind of objects for which it was designed? In the . We can. The states it occupies will mean whatever the actual value of F is. This is. we can (knowing something about such devices and the commercial intentions with which they are made ) make reasonable conjectures about how the instrument is supposed to be used. then things will appear to it the way things really are. of course. as it were). Such devices come with instruction manuals and carefully calibrated scales that tell the user what the pointer positions mean. then. in effect. Perhaps the indicator function of a system. and whether it is performing well . when it speaks truly .Introspection The reasoning goes like this: If you have a system whose function it is to indicate F. we have some information . how it is to beconnected . There is . make reasonable inferences about what its various pointer readings mean by seeing what the pressure is when the instrument is (we conjecture) working properly . of much depending on the animal and the senseorgan in question. In the case of natural systems (the sense organs in animals ).a process in which one determines what states me"anby comparing what the system says.about when these systems are functioning in the way they were (we conjecture ) designed (by natural selection ) to function . caseof artifacts.13 But how does one tell that a system is functioning properly and is being used in the kind of circumstances. is not . with the facts about which it speaks. if it is functioning properly " " (perceiving veridically . a calibrational process. no guarantee that we will get it right .

and pituitary glands are functioning properly . So what an external observer needs to know to determine how a system (whether natural or artifactual ) is representing an object is what its reaction to that object means and . in occupying state P misrepresents the pressure in k. to be sure. But if there is no special problem about telling whether hearts. If it turns out that the pressure in k is actually 14 psi. But the same can be said about any bodily organ to which we attribute a function . We cannot go back in time to observe the selectional es . Even if we could . process that.2 Chapter what we think it is. The evolutionary history may not have been what we think it was. P is the system s way of reacting to a pressure of 10 psi. practical problems about figuring out what anything is for. then the system is representing k as having a pressure of 10 psi . in proper working order. and given that the system is in state P. kidneys .the one(s). Given that the system has the function of indicating what value F is. then the system. Putting the . when it is functioning properly . what would the value of F be if the system was working right ? If a system' s reaction to k is ' P. therefore. indicator properties ) were the ones cau Sally responsible for its selection. what the reaction of a system means is what value of F the reaction is a reaction to when the instrument is functioning the way it was designed to function . we would still have the problem of determining which of the many kinds of information a system supplied was the kind or kinds it was selectedto provide . there should be no special problem in finding out whether visual or auditory systems are functioning properly . and connecting it to a tank gauge . it has the function ' of supplying . That is a question about which of a system s traits (in this case. There are. We might be wrong about what its function is. and. in the caseof the senses gave these systems their indicator functionss.

puts us right back into the situation of an external observer. is not equipped to tell that it is in proper working order. but the ' gauge doesn t have to be in proper working order for P to mean 10 psi . on independent grounds . it might be thought . about two instrumerits . surely. This . And even if it was. We are talking . of which is used to calibrate the other. by an elaborate calibrational process . indirectly and inferential - . is the way the gauge would have to find it out . objects it represents how would the device itself go about this out? By means of occupying a certain state P . to the pressure.Introspection at 10 psi .and. How might the gauge come to know what Prepresents the pressure in k as? It might seem as though the gauge has to do what external observers have to do. finding our gauge represents the pressure in k as 14 psi .one system knowing . but how one part of a composite representational syste~ knows how the other part represents the world . how might it gain accessto pressures that are known .natural or artificial . But is this possible? The gauge. This being so. this . responsesto pressure? If the system has access on independent grounds . 3 Knowing One ' s Own Mind If this is the way external observers go about determffiing how a device . Since this is the way we find out what P means. then we are no longer talking about a simple pressure representing the . It means that even when things go wrong . we are no longer talking about a representational system knowing how it represents the world . minimally . for the purpose of calibrating its own . though . is the way we tell that P means 10 psi . It has to go through the elaborate process described in 2.

occupy the state whose representational content is under investigation . This reasoning overlooks a crucial asymmetry. In representing k as 14 psi 5 occupies a state . say 14 psi .whatever information is carried by state another representsthe world . it does carry information about S . representational facts about itself. (vii . I have to look at both the gauge and the world . whereas 5 does.. try to determine what state P means in system 5. therefore. an external observer.about the way 5 is representing k. That is the difference between the representational system itself and external observers (you and me) trying to find out how it is representing the world .at the gauge to tell that it is working right . For in representing k as.has absolutely reliable information about what the pressure in k would be if 5 was working right . and the world (i . I do not .. If this is what it takes for a representational system to know how it represents the world . in a direct and unmediated way. 5.that I do not. 14 psi ) if P were functioning properly . Information (about what Prepresents k as) that an external observer can only obtain by an elaborate calibrational procedure is information that 5 itself obtains by .e. ' has information .e. When I try to find out what the gauge is representing k as.2 Chapter ly . P) that is perfectly correlated with ( hence carri ~s information about) what k would be (i . Thus. The system itself necessarily occupies a state that carries information about what the world would be like if the system was functioning properly . You and I do not. then it seems to be impossible for a representational system to know .and necessarily . These two sourcesof infor - . although P may not carry information about k. merely representing k. That is why we must go through the elaborate process described above. 5 automatically .. When I . k) to find out what the world is like when the gauge is in state P.

On a representational theory of the mind .Introspection mation tell me what the informational function of the state P is.iS precisely what is needed to make self-knowledge an instance of displaced perception . though . Otherwise simple instruments (which provide information to us ) would end up knowing things . but something else.representational systems have. But something more than information is needed to know. Animals do not. most adults do not . This is . The final section is devoted to this particular aspectof introspective knowledge . or some of one s internal states as representing the pressure to be 14 psi . the power to represent oneself.e. But the instrument itself . What more is needed is fact. And . Before we take up this matter. in occupying P need look at the world . not many . This does not mean that every representational systems has self knowledge .a process whereby a system gets information about itself (sufficient for knowing facts about itself ) by perceiving . k) to get this information . This is especially important because it bears on a paradoxical aspect (mentioned . this is the source of first -person authority . for most of the representational states they occupy. Gauges do not. This result . doing its job .. All we have so far shown is that every representational system. in representing the world . at whatever it is already " looking " at only (i. what the world is like when the gauge (in state P) is . P (the state the instrument " " gets itself into by looking at k) is a state that representsthe world and carries information about itself . the power of metarepresentation ' . something that not every.a result of thinking about perception in representational terms . not itself . a word should be said about exactly what information is available to a system that has representational powers . has all the information it needs to know how it is representing the world . Wants do not .

the fact that first person authority seems jeopardized by conceiving of the facts (i . Let me explain . thus. isn ' t in knowing what you believe and experience. mental facts) that are known introspectively as facts constituted by relations (including . Nonetheless. The present chapter is my way of saying why I agree with them.historical relations ) existing between what is in the head and what is not . but on the attitude (the relation ) one has to that content. That would be paradoxical .e. . . If one understands that introspection is a process in which information about internal affairs is gained in the act of representing (perceiving ) externalobjects. an externalist theory does create genuine obstaclesto self-knowledge . though I agree with them that an extemalist theory of mental content is no bar to direct and authoritative self-knowledge of what that content is (of what it is one thinks and experiences).2 Chapter earlier ) of extemalist theories of the mind . The extrinsicness of the mind is only a problem if one thinks one comes to know these extrinsic facts by looking inward . The problem centers not on the content.does not threaten first -person authority .in particular . what is knowable by introspective means.but something should here be said ~out exactly what information (about its own representational .. I shall come back to these issues in chapter 5 where I consider a variety of objectio ~ to externalist theories of the mind . states I take to be available to a representational system ) and. at the experiences these facts are facts about. however. ) ) Tyler Burge (1988 and John Hell (1988 1992 argue that externalism about the mind .at least externalism of the 14 right sort. representatio theories . The problem is in knowing that you believe and experience it . The problem .on the present account of things . the fact that what one learns (introspectively ) about oneself are extrinsically constituted facts is no longer a problem .

and Carol a Dodge. but ask someone else whether he owns it . Both Burge and He'll argue that higher -order belief about (lower -order ) thought (and . So you really cannot go wrong in identifying the content of the lower -order belief. But. But just as knowing that it is a Buick Clyde owns is only part of . just show me the car. absolute authorities . If people own the cars they drive . if push comes to shove. from what you believe. I am not very good about facts pertaining to ownership .. not that they own it . I think .in fact.Introspection I am (let us suppose) very expert at identifying cars. so to speak.knowing that he owns a Buick (you also have to . in fact. has. Ask someone else.about what we think and experience . I' ll tell you whether . If this is how it works . then it has to be true that it is P you (second order) believe yourself to believe since the secondorder belief attributes to the first -order belief whatever content it . but not very good (in fact. If you want to know what kind of car Clyde owns. eXperience belief to ourselves (as we do in higher -order belief that we believe P) the higher -order belief (= metarepresentation ) attributes to the lower -order representation exactly the same content (i. Tom a BMW. The legal niceties are beyond me. In self-ascribing a ) ably. Likewise . we are all very good. When someone buys a car. even if externalism about content is true . you really can' t Pe wrong in believing that it is P you believe since what you believe you believe is borrowed . I' ll tell you . if truth be known . I am the world ' s leading authority . presum has a reflexive character. Is it the bank or Clyde that owns the car? I don ' t know .e. P) that the lower -order representation carries. just when does ownership transfer? I don ' t know . then. but .it is a Buick that Clyde owns . I know Clyde owns a Buick. If it really is P that you (first -order ) believe . I know what car they own . very bad) about the attitudinal aspect of these mental states.

if the lower -order state has content. Representational systems have privileged information . All they argue is that. One also has to know one believes it . knowing it is P one believes is only part of knowing that one believes P. we get information (I have argued) about the way we experience (represent the objects we experience (as blue . Or no attitude at all. that it stands in those relations to the property F that constitute representation. not red or ) green).2 Chapter know he owns it ). in the case of higher -order belief . They have not argued that you cannot be IS wrong about believing P. 1 did not argue for this because 1 think it is false . information about the antecedent. But 1 have not argued that in . in both cases there is no authoritative . You cannot represent . identification of the content as P is authoritative . In the case of self-knowledge . it is (you can take my word for it ) certainly a Buick that he owns . Burge and He'll do not argue that . All they argue is that if you believe P. there is any information about whether the lower -order state (that has P as its content) is one of belief or some other attitude . necessarily occupying a state that carries the information that it is F (not G or H ) that you are representing something as. Nor have I . But . that one stands in those relations to P that constitute believing . say. something as F without . 1 have argued that in representing something as F a system has information about how it is representing things as F (rather than G or H ) . representing something as F a system necessarily carries infonnation that it is representing F. If Clyde . you cannot be wrong (in present-tensed self-ascriptions) about its being P you believe . but 1 have not argued that we get information that we stand in that relation to what it is we experience that justifies calling the resulting state an experience of blue. If it is an experience it is certainly an experience of blue. owns it .

but that we experience and ( think it (that there is a mind it exists in ).let alone privileged information .knowledge of what is going on outside our mind . I can feel it . If I want to make absolutely sure.see below ) about how they are representing things . The situation with respect to introspective knowledge knowledge of what is going on inside our own mind . we do not know this by introspection . not that there is an external world . what exists " in " the mind ) exactly parallel to perceptual knowledge . If we know there is an external world . I cannot see or feel that my visual and tactile experiences are veridical . But I cannot.. What the Representational Thesis provides . seeand feel that I am not hallucinating .Introspection (which is not to say they know. It tells us how we know it is F we experience. If we know not only what we experience and think i.that that is what they are doing . an apricot . And no more. is what it should provide . the way things are being represented. The same is true of minds in so far as they are representational systems. or a pencil ) in front of me becauseI can Seethat it is. it is not by 16 seeing or smelling . I know it is a tomato (not a banana. then. not the job of telling us that they are doing their job right . What we know by introspection is no'! that we have a mind . The sensesare charged with the job of telling us what is in the external world .i . but they have no information . Knowledge of internal affairs is no different . It does not tell us how (or whether ) we know . smell. of course.. but what is " in " the mind .) is what there is in the external world . that there is an external world . not about themselves. It reveals the source of firstperson authority about the contents of the mind . What we see (hear. about what it is we think and experience. Representational systems have the function of supplying information about the world . etc.e.

natural in the caseof animals and small children ). They do not know things about their own representa . facts about its own . representational states that it always . If I don t know what a posbnan is. the fact that the dog only barks when the post . The argument . rather.e.. a mental content). The conclusion is that a system could know (in some privileged way ). What is also needed are the conceptual resources for representing the fact to be known and an 11 appropriate connecting belief" if the knowledge is reached ' by means of displaced perception .Chapter 2 we experience F. has not been that representational systems know how they are representing the world (most of them do not ). They do not have the power to represent tiqnal efforts themselves (or anything else) as representing things in acertain way. how (or whether ) we know that we stand in those relations to F that make F part of the mind (i. that they necessarily have information about how they are representing the world that is more direct than that of external observers who can observe both the external affairs the system represents and the system' s internal representations. 4 Knowledgewithout Experience Gauges do not introspect . not that it does know (in any way whatsoever ). Or small children . therefore . It shouldn ' t. but . But information sufficient unto knowing facts can be available without those facts being known .and necessarily has information about its own representational states sufficient unto knowing what their content is. That is the business of episte17 mology. Nor do animals . They have representational powers (conventional and derived in the caseof instruments . but they do not have metarepresentational powers .

Perner (1991 has nicely summarized the evidence bearing ) on a child ' s ability to understand representational facts and . Displaced perception will not occur. appears to develop between the agesof three and four years . Wellman 1990 Gopnik 1993 . IS They have the information . For the same reason. therefore (on a representational theory ). beliefs needed to use the dog ' s bark as a means of perceiving ' (displaced) the postman s arrival . They lack the right concepts. knows (as I do ) that the postman has arrived . . Such an understanding . That. ) does not prevent one from representing things . Not having a concept of representation . basically. A small child (or the cat. the power of metarepresentation. not enough to have direct knowledge of mental facts. They can hear the dog barking as well as I can. but not the understanding . is the same reason they do not have introspective knowledge .Introspection man arrives will not enable me to hear ( by the dog ' s bark ) that the postman has arrived . hence . also see Flavell 1988. it cannot conceptually represent (hence cannot believe. Until a child understands representation. in hearing the dog bark .it cannot occur. 189. for that matter ) has access to the same information about the postman ' s arrival as I do. 82. but it certainly prevents one from believing (hence. and the dog ' s bark carries the same information (about the postman ) to them that it carries to me. knowing ) that one is doing it . even if they had those concepts they might lack the appropriate connecting . it is not enough for introspection .until I learn something about postmen (enough to enable me to believe that a postman has arrived ) and acquire the connecting belief that the barking dog is a reliable sign (= carries the information ) that the postman has arrived . and. mental facts. that one have accessto information about those facts. But neither the child nor the cat. (Perner 1991 pp .

What they lack is the power to give conceptual embodiment to what they are getting information " about.makes it into a form of inferential .1 can know how I am representing some object without " " truly representing the intermediate fact that provides me with this information about myself . possibly misrepresenting) something as F.. There are two ways of knowing that the postman has arrived : by seeing or hearing him arrive and by seeing or hearing something else(the dog ) that " tells" you he has arrived .that one represents the world thus and so . -as a speciesof displaced perception. that is. The first point is that although I cannot hear that the mailman has arrived by hearing the dog bark unless the dog is. and they needn t look inward to get it . then the representational account fails to give selfknowledg the immediacy we know it to have.including itself . barking (and I hear it ). It may seem as though this account of introspective knowledge. They ' already have all the information they need. knowledge . What prevents small children and animals from introspecting is not the lack of a mysterious power that adults have to look inward at their own representations . This point is certainly relevant. then there is an intermediate " step " in the reasoning that makes knowledge of the target fact indirect .is representing (and. therefore. If introspective knowledge of oneself.Chapter 2 cannot know ) that anything . If . in fact. My representatio~ . I truly represent the the intermediate fact that " tells" me the postman is here .unless. but there are two important differences between introspective knowledge and other forms of displaced perception that neutralize the objection. and thus indirect .has the same structure as knowing that the postman has arrived by hearing the dog bark . on a representational theory . introspective knowledge is more like the latter than the former .

about me. about how I am representing things Displaced knowledge is still possible . it is a strange case of inference: the premises do not have to be true to establish the conclusion. my representation of the dog will not carry information about the posbnan . The premises need not be true . say) does not have to be veridical for this representation to carry information about the target fact. In the case of introspection . It may turn out that the dog changesits habits . in this way. the perception is displaced . I can no longer come to know (though I may still be led to believe) that the postman has arrived by hearing the dog bark. Should this happen. Secondly the connecting belief in the caseof introspective knowledge is not defeasible. you cannot go wrong .that from this " fact you are representing k as blue . but the necessaryinformation is there whetheror not the intermediate representationsare veridical . As long as the " inference is from what you " see k to be (whether this is veridical or not ) the conclusion must be true: blue must be the way you are representing k. If you " see k as blue and infer " . it is a very unusual form of inference.Introspection of k (as blue . if this is inferential knowledge . The posbnan cannot fail to be there when this dog barks.the " fact" that k is blue . In the case of self knowledge . If this is inferential knowledge . fallible . my sensory r ~presentation of the intermediate fact need not be . But the connecting belief in the caseof introspective knowledge " is not . .19 Once again. The connecting belief is no longer true. If my representation of the dog is not veridical (if I hear it as barking when it is not ).it starts barking at other people. veridical for it to carry information about me. yes. As a result . there will be no displaced knowledge . about how I am representing k. I do not have to truly represent the color of k in order to get information about myself from my sensory representation of k. though .

one to know what experiencesare like by " scanning" them. This.2 Chapter and the inference cannot fail . like external objects. for all the world . this internal scanner. the experienceslook . If one is aware of experiences in the way one is aware of external objects. this account of introspective knowledge explains a fact that is otherwise " " puzzling on inner sense theories of introspection . e~actly how the wine " . If there is an inner sense some quasi-perceptual faculty that enables . if it does. As long as one' s conclusion (about how one is representing things ) is based on premises " " concerning how things are perceived to be. if you can. If one is asked to say what one' s current visual experiences are like . one seems able to know this without having any identifiable experiences other than the visual ones one is able to describe. Introspection has no phenome- .one finds oneself ' tastes attending. . There seemsto be no other relevant place to direct one' s attention.20 A representational approach to the mind gives asatisfying explanation of this fact. It explairis why introspection has no phenomenology or. I submit . At least one does not have experiencesother than the wine. unlike the other senses has acom .experienceone is asked to describe. not to one s experience of the wine . is the source of the directness and " " of immediacy introspective knowledge . why it always has the samephenomenology as the experience one is introspecting . for example. It suggests that there is not really another sensein operation at all. If one is asked to introspect one' s current gustatory " experience . It does not " present" pletely transparent phenomenology experiences of external objects in any guise other than the way the experiencespresent external objects. but to the wine itself (or perhaps the tongue or palette).Tell us . Besides explaining first -person authority . one cannot go " " wrong . This is very suspicious .

by introspection . but the knowledge is obtained without any experiencesbeyond the ones one comes to know about. Given that I understand . in my experience of blue . I have. To know I am experiencing bluely . I don ' t need anotherexperience. to know that that is the kind of color experience I am having. come to know about experience . . I need only the experience of blue. One can. !. all I need to know what kind of experience I am having .Introspection nology becausethe knowledge one gets by it is (itself) experience -less. There is no need for an experience of the experience.heconcept of experience and its qualities . All one needs is a belief about it .

what it is like to be that creature. you can get the same information I have about the character of my experiences . This is a result of thinking about the mind in naturalistic terms. If the subjective life of another being. in principle knowable by others.qualia with the properties objects are repre sentedsas having . this must be because we fail to understand what we are talking about when we talk about its subjective states. 2) . but there are no person-privileged facts. seemsinaccessible . Though each of us has direct information about our own experience (see chapter . For materialists . Subjectivity becomes part of the objective order . facts that only one person can know about. circumstances of time and place that make one person authoritative . There may be circumstances of time and place that enable one person to know something others do not . and its feeling some way is a material state of S. If S feels some way . how can it be impossible for us to know how S feels? The Representational Thesis identifies the qualities of experience.3 Qualia For a materialist there are no facts that are accessibleto only one person. . there is no privileged access If you know where to look . this is as it should be. The properties that S representssthings to have is.

Shewill not seeit as or seethat it is. and sheepdogs look . Does it look like a poodle to her? Susanwill not say or think it is (or looks like ) a poodle . He has never seen a dog . The first one he sees . Arthur is a young toad of normal toad. nonetheless . makes no difference to how the dog looks to her. ' it is (or looks like ) a poodle . She doesn' t know what dogs are. Poodles look like blurry spots to half -blind people---the same almost everything else this size looks. a poodle . The first one she sees is a French poodle . This is not the way he will conceptualize what he sees Will it . way look like a poodle to her? .eyesight and toadintelligence . Poodles do not look like poodles to half -blind people despite the fact that they look the way poodles normally look to half -blind people.different from the way bulldogs . has never . seen a dog . . Though the dog looks to Arthur the way poodles look to normal toads. You and I can describe the way the dog looks (" like a poodle " ) and Susan cannot. Arthur is different .Chapter 3 1 French Poodles and French Wine Susan a child of normal eyesight and a poodle to him ? . So poodles look to toads the way poodles look to half -blind people: . Toads have the visual acuity of half -blind people. nonetheless look . He will not see it as. surely. or see that it is. it seemsclear that the poodle looks different to Susan than it does to Arthur . but this . The dog looks to Susan the way poodles look to you and me . This is not the . she will conceptualize what she sees Will it . happens to be a poodle---the same one Susan sees Does it look like a poodle ? Like Susan Arthur will not say or think . Given what we know about the eyesight of human beirigs and toads. a poodle . and Arthur is a normal toad . the dog does not look like a poodle to Arthur . terriers .

They. I do not want to quarrel about words . Good fakes are supposed to cause the same kind of experiencesas the originals . produce .e. seealso Millar 1991 p . 32). the manifest properties of poodles . too .Qualia In saying that the dog looks like a poodle to Susan I am ' not saying that Susan s visual experience of the poodle represents it as a poodle and. No . To keep things straight I will call these uses the doxas right tic (= belief) senseof " look " (" appear. it would not " seem" to that person that there was a pumpkin there." " seem " etc. not conceptual .). but if that person did not have the concept of a pumpkin . representations of the poodle . 42) describes as . 40. . and so on. 30]. ' Susan s experience of the dog representsthe dog as a poodle in the sense that it represents the dog as having what McGinn (1982 p . did not know what a pumpkin was. Millar (1991 p . ubscript it accordingly Oackson [ 1977 p . therefore misrepresentsit in some if it is not a real poodle . conceptually representing) a good fake to be a poodle . for instance. says that a person . Some will feel it wrong to describe anything as looking like a poodle to someone who does not have the concept of a poodle . poodle qualia in Susan. disguised terriers. Susan is not misperceiving (sensuously misrepresenting ) it when she visually represents a fake poodle in the same way she visually represents a real poodle . These objects also look like poodles . poodle robots. pictures of poodles. Although she would be making a mistake in believing (i. .) and will . etc. we are here speaking of way sensory . There " " " " may be a sense of look and other appear words (especially when used with a factive clause) in which Millar is . those properties that make poodles look so much different from other dogs (not to mention bicycles.. A variety of nonpoodles also have the manifest properties of poodles: trompel 'oeil. could have pumpkin experiencesof exactly the same type I have when I see a pumpkin . calls it the ~ .

There is. terriers. in the absenceof countervailing consideration ' . To say that a dog looksd like a poodle epistemic to 5 is to say that. this is what 5 would take the dog to be. If you are color blind . another senseof these words . and it looks to me like a poodle . though . Describing the dog as lookingd like a poodle to 5 implies that 5 has the concept POODLE. (seeMillar 1991 p . The discrimination condition is here left deliberately vague to reflect the context-sensitivity and circumstantial relativity of thesejudgments (k can look like a poodle to you in circumstancesC but not in C'). then it must look to 5usan like a poodle whether or not she understands what a poodle is . I will call this the phenomenal senSe " look " (loo ). what 5 s perception of the dog would (normally ) prompt 5 to believe .is necessary to prevent a dog looking like a poodle to a toad (or a half blind person) to whom all dogs look the same. whether or not she has the concept of a poodle . You do not satisfy the discrimination condition . Even if there is some rare breed of dog that you cannot distinguish from a French poo- . 19). a sensein which if the dog looks the same to 5usan as it looks to me. That isn' t what her perception of the dog causesher to believe. and classifies or identifies (or would do so in the absenceof countervailing considerations) what she sees in this way . To say that the dog looks .3 Chapter use of " look " ).phenomenally of ~ .call it the discriminatory clause. In the doxastic sense the dog 5usan . if you cannot discriminate one color from another. and (2) the dog looks different to 5 from other dogs (bulldogs .like a poodle to 5 (= looksp) is to say two things : (1) that the dog looks to 5 the way poodles normally look to 5. Following a long tradition . an object does not loo red to ~ you simply becauseit looks the same as red objects normally look to you .)} The second clause. understands what a poodle is. seesdoes not lookd like a poodle to her. etc.

This is why a dog can loo like a French ~ poodle to one even though one cannot discriminate it from a very good fake. None of this implies that the object looks to 51 the ' way it looks to 52.2Hence. we can say that the dog like a poodle to 5usan but not to Arthur (it looksp like a poodle to neither .3 ~ It should be noted that simply becausean object F looksp to two people does not mean it the same to them. in part . The fact that a dog like a poodle to 5usan and (say) to another looksp poodle does not mean it looks the same to them.Qualia dIe. then toads cannot see differences that nonrial human beings see . on the toad ' s powers of discrimination . Fake poodles are not .not normally . of course). one does not have to be able to discriminate French poodles from very good fakes in order for a dog to loo to you like a French poodle . The proof of this lies in the inability of toads to perceptually discriminate the shapes of middle -sized objects. If . despite the discriminatory clause . It looksp may turn out 51 and 52are equally good at discriminating Fs from other objects and that a given object. With these terminological stipulations . . anyway . as I have been assuming (on the basis of neurophysiological and behavioral data4 ) middle -sized animals that are moving (toads don ' t seem able to see the stationary ones) look the same to toads. More of this in a moment. a dog might nonetheless " loo like a French poodle to ~ " you as long as the implied set of contrasts did not include the rare breed. maybe it doesn t. Maybe it does. Just what poodles look like to a toad depends. looks to 51 the way Fs usually look to 51 and looks to 52 the way Fs usually look to 52. the contrast-class. A toad ' s visual looksd system was not designed to discriminate the forms of middle -sized objects in the way the visual system of humans was. The possibility of inverted qualia is not foreclosed by these definitions .

" " Maybe not.the gustatory equivalent of blurry spots. both taste like an ordinary red " jug " wine .01.3 Chapter differences in form and detail that make poodles look different from bulldogs to us. So poodles . I infer . must loo to toads the way poodles ~ look to us when we see them in poor light and without our glasses. if numbers designed beyond the decimal point are not (as they say) significant digits for it .99 and' anthat (a bulldog ) shape. not if I cannot discriminate it from aChiantis If Burgundies taste the same to me as Chianti .00. No . presently tasting tastesplike a red Burgundy to me? It does after all . Arthur ' s visual system " does not have the requisite " resolving power .00 and 7.01.pretty much the way bulldogs and terriers look . . This. If I stand (gustatorily ) to wines the way toads stand (visually ) to dogs. taste the way red Burgundy normally tastes to me. We are also in a better position to describe certain (otherwis ~) puzzling situations . A toad ' s visual system does not represents poodles the way our visual system representssthem.00 to an insb' umentthat has no internal state with the function of indicating a value of between 6. Burgundy and Chianti are to me what poodles and bulldogs are to toads . It depends on whether I -can distinguish jug wine from Coca. Suppose all red wines taste the same to me. basically. Nothing can " seem" like 7. Application of the same criteria for telling how things " look " or " seem" to a system is evident in the way we think . is the reason nothing can " seem" like a poodle to Arthur . chancesare neither Burgundies nor Chiantis taste like Burgundy to me. Does this mean that the red Burgundy I am . then the insb' ument cannot represent anything as 7. and talk about insb' uments If a measuring insb' umentis not to discriminate between 7. In all likelihood . Nothing in Arthur has the function of indicating that yonder object has this (a poodle ) shape rather .

Any two items that present different hues to the unadapted eye will present different hues to the adapted eye. If the wine example seems far .about the data in making inferences about appearancesp ' of someone s experience . so the effects will soon disappear . . matic appearance of most things . That this is falla cious . sipping an expensive Burgundy .at least suspicious . The adaptation difference between the two eyes vanishes quickly . But the apparent hue of everything shifts. We must . be careful in using discriminatory . maybe they all like exquisite Burgundy and not . tastep as I was assuming . Matching items will continue to match as well . as it were . While they last. however. experiences 7. This gustatory anomaly may be hard (impossible ?) to detect . these adaptation effects are similar in several ways to the purported differences between people . All wines taste to me the way fine Burgundies taste to a connoisseur . stare at some bright colour for 30 seconds and . like ordinary red table wine .fetched . That is exactly the way it tastes to me . Every time you . All wines taste to you the way cheap Chianti tastes to a connoisseUr . ap ~ earsp to them in that mode . p . 167) gives a more compelling example : H you close one eye. Clark ( 1993. then you and I might be the same in our discriminatory powers (we both embarrass ourselves at wine tasting parties ) even though all wines taste different to us . however . then blink your eyes successively you will note shifts in the chr0. Such inferences move from quality premises about what animals can and cannot discriminate in a given sensory mode to conclusions about how something .6 She.can be seen by considering the wine example .00 shapes and Arthur does not .Qualia ' Susan s visual system is a more sensitive insb " ument. Discriminations made with the adapted eye will match those made with the unadapted eye. If this were so. say that the wine tastes like that (gesturing toward a cheap Chianti ). Even though all red wines taste the same to me . I agree .

different in the way things taste to us .who think the quality of experience must .3 Chapter Clark concludes by saying that the possibility of interpersonal differences of this sort seems fatal to the project of defining sensory qualities in terms of what can be discriminated . be defined in behavioral or functional terms. of course.different in the way we sensuously represent the objects we perceive. The Representational Thesis is a naturalistic theory that avoids this problem . behaviorists and functionalists8. yet. (2) It provides an account of sense experience which makes the qualitative aspects of experience objectively determinable . is the inverted spectrum " " 7 problem .sometimes (from a practical standpoint ) impossible to discover what the functio ~ of a certain state is. In identifying qualia with experienced properties. This. physically definable . that we might be indistinguishable in our . ) It seems then. Shoemaker (1975 makes the same point . . It is. a representational approach to experience makes qualia as objectively determinable as are the biological functions of bodily organs. It may be hard. but there is nothing essentially private or exclusively first person about functions . however .g . it concedes is not functionally definable. experienced properties with properties ' representeds and the latter with those properties the senseshave the natural function of providing information about .. discriminatory behavior and.9The qualitative character of perceptual . a representational approach to the mind does two things : (1) It respectsthe widely shared (even by functional ists1 ) intuition that qualitative aspects of experience are subjective or private : they do not necessarily express themselves in the behavior (or behavioral dispositions ) of the system in which they exist. By identifying qualia with the properties that the experience representssthings as having . experience. The problem is a problem for thosee . somehow .

What I hope to do ) in the remainder of this chapter is to show that such pessimism can be avoided. In accordancewith the Representational Thesis.ask what would be the caseif the device was working the way it is supposed to " work .Qualia Eduardo Bisiach (1992 a neuropsychologists. If you want to " know how things " seem to a ringing doorbell . we look to the purposes and intentions of the designers and builders . " no way in which a natural science of consciousnesscould have anything to do with qualia " (p . In the case of conventional representations . There is. 115 . questions about what properties these stateshave the functions of indicating . how the bell is currently representing the front door. the way it is supposed to work . A representational account of experience not only makes room for qualia. 2 Qualia as Represented Properties It is hard to find a description of qualia with which two (let alone all ) philosophers would agree . " " finding an objective way to study qualia . 2. As we saw in chapter 2. This ' ' means that questions about another person s (or animal s) qualia are questions about the representationals states of the person (or animal ). but it seems safe enough to begin by saying that the qualia in sensemodality M (for 5) are the ways objectsphenomenally appear or seem to 5 in M . I continue to identify qualia with phenomenal properties those properties that (according to the thesis) an object is sensuously represented (representeds ) as having . the way an experience representss an object is the way that object would be if the representational system were working right . despairs of ). it provides an objective way of studying them. the way it was designed to work . Using .. he says. That will be how things " seem to the doorbell .

The state of a representational system means whatever it has the function of indicating . though Susan could not .. but that is not what the bell means. This is what the wine example shows. even under optimal conditions . believe) it to be a bill collector.I emphasize. hallucinating impaired poodles when she looks at dogs. A person hearing the bell might represent (ie . We can imagine Susan dreaming of poodles } 2 We can imagine her hallucinating . not a visiting relative . in point of fact. It does not mean that a bill collector is at the door even if . not what it says or represents. We can imagine someone with virtually no ability to discriminate wines experiencing an exquisite . The reason it cannot " seem" to a doorbell that a bill collector is at the door is the reason it cannot seem to Arthur as though there is a poodle nearby. That is why doorbells do not " lie " about who is at the door. too .ll That is why we say that the bell means that someone is at the door-':'-not a bill collector. that is who is there. This is not . Arthur ' s visual system cannot represent something to be a poodle for the same reason a doorbell system cannot represent someone at the door as a bill collector. Yet. Given the operation of this system . but someone .Chapter 3 this test. That is what the ring means whether or not there is someone at the door. We can even imagine Susan with vision so she cannot distinguish poodles from bulldogs .because Arthur cannot distinguish poodles from bulldogs . we can still imagine her having poodle qualia . cannot distinguish between bill collectors and visiting relatives. poodles.thus having poodle qualia when she seespoodles. when the bell rings. it must " seem" to the system that there is someone at the door . not. For we can imagine Susan' s eyesight so bad that she. the bell . They never say who is there. thus impaired . cannot distinguish one from the other. distinguish poodles from bulldogs .

Arthur occupies no such state.e. When it registers 78. He can tell a " " 77. K s speed palate is much less discriminating than J. Its registration of 78 means not 77 (and below ) and not 79 (and above) } 4 There is no state of K (as there is of J) that has ( by design ) the job of . but in what information their visual 13 systems have the function of providing . that (as the digits after the decimal point suggest) means not 77. whether or not they enable her to distinguish poodles from bulldogs . J has speed quale that K never experiences .01 speeds Since no state has this function . Susan occupies perceptual states that . We can imagine their discriminatory powers being the same and . Another person always experiences the taste of cheap Chianti . designed to provide rough infonnation about speeds.00. the functions) of distinguishing middle -sized objects (like poodles and bulldogs ) from one another.any wine . the key difference between Susanand Arthur is not the difference in their discriminatory powers .. J and K. nonetheless have the function (i . That is why he can no more experience poodles as poodles than standard issue doorbells can represent bill collectors as bill collectors.00. All these speeds feel the " " " . whether or not they can any longer perfonn their function . No .Qualia Burgundy taste when he or she drinks wine . their having quite different experiences The key difference is not in what infonnation their visual system provides " (this might . Instrument K is a less expensive device. J is a precision device manufactured to measure speed in hundredths of miles per hour." same to K. yet. be the same). J is a speed connoisseur. imagine two instruments .98 speed.92 speed from a 77. J has a discriminating speed palate .99 and 78.01 (and above) . no state of K represents the " " ' speed as being 78.99 (and below )" and not " 78. indicating that the speed is between 77. .

00 the same way it representsa speed of 78.05 .e. J and K are responding to the same objective condition just as both Susan and Arthur are responding to the same poodle .as a blob . and was designed to be sensitive ' to .00 to 78. representing speed differently ) while being indistinguishable in discriminatory behavior and capacity.50. these differences in speed.00 mph and 78 mph respectively. Suppose speedometer J. When the speed rises from 77 to 78. loses sensitivity . Arthur represents poodles the same way he represents bulldogs . Speeds that . these differences in shape. were represented as) different to J now seem the same.23 J' s needle (just like K ' s needle ) doesn ' t budge (it may help to think of these as digital devices ) .05" if the ' ) speed changed from 78. via damage or wear. and (we are assuming) was designed to be sensitive to . these differences. " seemed" (i . In representing . before damage...00 to " 78.e. Small changes in shape make a difference to Susan.05.just as K represents a speed of 78. the speed as 78. J (like K ) sluggishly " " " " responds with a shift from 77. Just as J s mechanism is sensitive to. ' " " Just as J s registration would change (from 78. Neither of them would register. This is not true of K and Arthur . We can imagine our two instruments " experiencing different " qualia (i .as 78.84 and 78. It becomes as insensitive to speed as K. As the speed varies between 77.00 to 78. the two instrwnents provide the same information about speed. nothing in them was designed to register. After damage to J' s information delivery systems. The two instruments thus become functionally . But Susan represents this object in a more discriminating way . Susans visual system is sensitive to.00 at approximately 77. Susan s " needle" would move if the repreSentational objective shape ' changed from that of a poodle to that of a bulldog .3 Chapter J is the instrumental analog of Susan K of Arthur .

Despite this equivalence. She has sensory states that mean this even if they no longer carry this information . speeds There is no state of K that has this representational content .00" on J means something different from a registration " 78" on K. J. have an internal state that was designed to indicate a speed of 78. a registration of 78.00 mph . who ). but J can still about this speed. therefore. He cannot " dream " about 78. Damage changes what information a state carries (what it correlates with ) but not what it means (what it has the function of correlating with ).00" has (and retains) the function of " " indicating . The difference between J and K is that J does.00 on J means something different " 78" on K even if the two from a registration of responses carry the same information . The fact that of J no longer delivers the information it is its function to provide does not mean that it loses the function of providing it . while K does not . I agree. The fact that when things are not working properly the internal states of the two devices indicate the same thing does not imply that they meanthe same thing . K cannot. with impaired vision . That is what a reading of " 78. This is why we can imagine Susan . therefore.00 mph . but not Arthur . not even what they can do. As long as it retains this function " " . with Shoemaker (1991.00 mph qualia and still doesn ' t . a registration of " 78. .00 mph to J. whatever ) things still " seem" like 78. agrees with Ned Block and Jerry Fodor (1972 that qualia . therefore . After damage (injury .Qualia indistinguishable .00 mph . J can no longer (veridically ) perceive a speed " hallucinate " or " dream" of 78. The representational difference between J and K lies. K cannot. can represent the speed to be 78. seeing every dog as a poodle. K never experienced 78. an internal state that .00. has this function . 508). not in what they do. but in what their various stateshave the function of doing . p . old age.

in physical terms. temperature . functionally equivalent ) and. it does not mean that they are not physically definable . theory of the mind that does not abandon the subjeCtivepoint of view. 3 Pointsof View Thomas Nagel (1974 said that " . and it seemsinevitable that an objective. Both 51 and 52 represent 0 ." A representational account of subjective phenomena constitutes an objective. If we compare two representational systems that represent the same set of detenninables (speed. qualia . occupy different representational states. physical theory will abandon that point of view . . yet . we have a naturalistic theory of representation and . of the conditions in which systems have information . But this does not mean that qualia are not capturable by a representational account of the present sort . It does not abandon points of view . it explains them. etc.3 Chapter are not functionally definable. Experiences can thus be different " even though this difference can no longer " express itself in discriminatory performance.). but 51 represents it as blue while 52 represents it as cyan (a shade of blue ). As long as we have a naturalistic theory of indicator functions . For two representational devices can be equivalent in their discriminatory powers and capacities (hence.carrying functions. . hence. Or both representsome- . naturalistic . weight . They are physically definable as long as there is a description . every subjective phenomeno ) is essentially connected with a single point of view . their points of view are fixed by the objects they represent and the detenninateforms of these detenninable properties they represent these objects as having . Though this means that qualia are not functionally definable.

The difference in their point of view is determined .Qualia thing to be cyan. thus.or different parts of the same object. one that is determined by the representational facts. Experiences do not differ subjectively just because they are experiences of different objects and. Biro points out that there is nothing mental about fixed points of view . They represent different objects in the world . constitute different points of view. about systems. different parts of the world .them sensitive to. points the same. One cannot tell .nothing it is . just by looking at the radar screen that they have different . This difference in point of view is not . The screens after all .. In either case they have a different point of view .a point of view that can be exchangedby changing ' places with each other. What gives the two radar systems a different perspective on the world is the different way they are connected to the world (a relation I earlier dubbed " C" ).thus affording any two systems representing different objects different views (i . If we suppose that any objects necessarily occupy different places in the world .thus coming to have the same (fixed ) point of view as the LaGuardia radar. of view . but which part of the sky they represent .to be . The radar at O ' Hare (Chicago ) and LaGuardia (New York ) represents the location and movement of planes in their vicinity . not by how they represent their part of the sky. and thus able to represent.e. in the case of natural systems not a mental ) fact (: . The O Hare radar can be moved to LaGuardia. The instruments have different points of view. views of diffe ~ t places) of the world .then the difference in " view " created by the fact that different objects are being represented is a difference in what Biro (1991 1992 calls afixed point . cyan. however. a connection that makes. ) of view . arepresentation al ( hen e. but it turns out that they represent different objects. might look exactly .

What makes a difference in what it is like to have an experience.00 mph while" " another represents it as 78 mph . We could exchange fixed points of view without so much as a ripple in our conscious life . but of differences in the way the points are viewed . In Section 2 we saw how an instrument could " " provide a close up view of the same magnitude .3 Chapter " like " in ' ( Nagel s sense to occupy or change fixed points of ) view . The O ' Hare and the la Guardia screens although having . embody differing " perceptions " of this objectthe same kind of difference that we experience when we look at small print with the naked eye and through a magnifying glass. Such differences in points of view help to define what it is . even if these instruments are veridically " perceiving " (representing) the same object. but our experiencesare the same. different points of view . Differences in experiential state are not the result of differences in point of view . therefore. so to speak. One speedometer represents the speed as 78. One has better vision " or it is. To a less sensitive instrument these " " speeds look the same.05 looks different from a speed of 78. If we see different objects . -representational systems are representing Even when F the same object they can represent it . They can. we occupy different (fixed ) points of view . at different degrees of magnification or over different ranges.but the way these IS objects are represented from that point of view . in the quality of experience. Different points of view can be experientially the same. therefore . The first has a close up view that changeswith small perturbations in the speed. Hence.not . is not the objects the experience is an experience of . could be in exactly the same representational state. but they are twins . they see it differently .00. standing closer. It has a better view of the object. so to " " " speak. To " " a sensitive instrument a speed of 78. our (fixed ) point of view .

fish . a scienceof the mind . within a representational idiom. then there is reason to think there are facts you cannot know about the experience of others.certainly any other animals . How can we know what a neighbor ' s experience is like . It also suggeststhat the hope for an objective a will o' -the-wisp .at least that part .experience things you do not . of the mind given over to subjective experience . The argument easily generalizes . then.Qualia liketo havethat point of view but theyareeasilycaptured . Knowing what bats .that are inaccessible " to objective determination . This.unless we can somehow get into their heads and experiencewhat they are experiencing? The Representational Thesis reveals what is wrong with this reasoning . makes it seem as though there are aspects of subjectivity what it is like to experience F. 4 What Is It Like to Experience Elecbic Fields ? There is an argument due to Frank Jackson(1986 that if you ) do not experience color (hue ) for yourself . in turn . blue . This version of The Problem of Other Minds strikes many people as being persuasive. How can s~meone without a senseof smell know what lilacs or burning toast smells like ? There are. If you have to experience F to know what it is like to experience F. and other colors . then a certain piece of factual knowledge remains inaccessibleto you . if there is reason to think (as there is) that other people . is reserved for those who experience them. know without having had certain experiences But if this is . You do not know what it is like to experience red . it seems facts you cannot . knowledge of what those experiences are like . so. then some knowledge . and neighbors .not to mention the experience of aliens and animals .

to illustrate the " point . But she has never experienced an electric field .it does not require the conceptual lI impossibility of getting inside ll the head of another being . all Mary wants to know is how things seem to dogfish when they. What . The host does not have to be 18 C. . What sort of qualia are produced by .and attaches itself. our parasite attaching itself to hosts when . impossible . from a practical standpoint . in principle . In both casesit is a question of determining how a system is representing the world . She understands all about magnetic and electric fields . a sensory system that detects deformations in electric fields? Since dogfish are fairly complicated animals. it has evolved an acute thermal sense It is capable of the temperature within fractions of a degree of registering bodies with which it comes in contact} 8 The parasite senses a potential host as being 18 C. and so on.sometimes .the ideal temperature . Mary is an expert on electromagnetic phenomena . Imagine. Though this is difficult . would it like to be a dogfish and experience electric fields? Mary is not asking to be a dogfish . 's I borrow Jackson protagonist . All that is necessary to bring about attachment to the host is that the .3 Chapter experience is. She has heard that sharks and dogfish have an electromagnetic sensethat enablesthem to detect prey by 5el1 Singdistortions ' often created by prey ) in the earth s electric field . then. the host is at a certain temperature : 18 C. Since the temperature of the host is critical to the survival of the parasite . it will help if we begin with something simpler . and only when . no different from knowing how " 11 things seem to a measuring instrument . via their electromagnetic sense detect prey. Mary . she ( wonders . In wanting to know what it is like to be a dogfish .a mono -representational 17 6 parasite} I return to dogfish in a moment. Maxwell ' s equations. quantum electrodynamics.

but at what the parasite is looking at the host. look at th ~ host. if things are working right .Qualia parasite senseit as. to know what it is like to be one. how it represen the objects it perceives. I am merely drawing out the consequencesof facts that almost everyone accepts . what the host is . you know how the host ' " feels" to the parasite.18 C. if a tomato looks red and round ' to 5. then redness and roundness are the qualia of S s visual . one looks. I am not denying it . Deaf people can know what sound waves are without knowing what it is like to hearsound waves. 18 C. The question is: what is it like to be this parasite when it sensesa receptive host? If you know what it is to be 18 C. you do not have to be a ts . I do not wish to deny this. in the remainder of this section. If you know enough to know what it is to be at a temperature of 18 C. For. All you have to parasite know is what temperature is. and especially in the next section. though . To know what it is like for this parasite. So if you want to know how things seem to the parasite. represent it to be. You know what the parasite s experience " senses the host. I will . For the moment .19Surely knowing what temperature is will not tell one what it is like (if it is like anything ) for a parasite (or even another human being ) to feel a temperature of this kind . If " is like as it knowing what it is like to be such a parasite is knowing how things seem to it . The first fact is that qualia are supposed to be the way things seemor appear in the sensemodality in question. you know all there is to know about the quality of the parasite' s experience. try to say more fully why I am not ' denying it . not in the " " parasite. This will surely sound preposterous to some.facts that are quite independent of the representational point of view being defended in these lectures. I beg the reader s indulgence .is how things seem to the parasite . So~for example.

properties the properties the object being perceived has are exactly when the perception is veridical . represents . . and it bulges . They know . the property the object (the host) has when the perception . then one of the two facts that led to it .3 Chapter experience of the tomato. and knowing the geometry of the field . If that result is absurd. what it is like to be that parasite. Mary . There is then there. knows what quale the parasite' s experiencehas. that has to be th ~ quale of the parasite' s experience of the host whether or not the parasite is perceiving the host veridically .is to blame. then it is representing patterns in the " " electric field . representing. an irregularity We can describe the configuration of the field . Back to dogfish . Just as . in the field whose geometry is describable. . Mary draws . (whether or not the host actually has this property ). If the dogfish electromagnetic sense is functioning normally . rocks. So anyone who knows what 18 Cisknows what this property is. The field is normal here. In drawing. What she draws (describes . things ever arethe way they seem it follows that qualia . and plants deform them. . the that define what it is like to have that experience . then (second fact) if . is veridical is the property of being 18 C.not the RepresentationalThesis . who knows all about electric fields and how fish . I just did . She knows everything there is to know about electric fields in the way you know everything ' there is to know (there isn t much ) about moving at a 's speed of 4 mph . it is pinched in " " between here and there. with respect to this single quale. Mary knows all about electric fields and irregularities in electric fields. Since in the case of our parasite. describing . If this is so. or knows ) about the electric field is what the fish sensesabout the electric field in which it finds itself. describes represents and knows what it is like to be a dogfish (veridically ) sensing that kind of field . could draw an exact picture of the field .

To know what the quality of an experience is is not neces sarily to be able to recognize something as having that quality when you experience it yourself . In knowing what an experience of a field of this shape is like . If I know that you hear a change of key.knowing what it is like to have four sides. Your experience has that change of-key-quality .20Mary. other color experiences Likewise . color blue since the color blue is the color one experiences It ' s blue is the quality that distinguish es one experiencesfrom . it is like to have an experience of this sort. T is the quale of this experience.Qualia you . and ) ) others have pointed out . There is nothing more to know about how an electric field of that shape seemsto the dogfi ~h. she knows all there is to know about the quality of experiencesof this type . On almost any theory of experience. I do not myself have the skill or ability to recognize a change of key when I hear it myself . If Mary knows what a field of type T is. Nonetheless. with respect to this single dimension of the experience what . there is no more to experiencing an electric field of type T than there is to being an electric field of type T since T is exactly what makes this electric field experience different from other experiences of the electric field . knows what an experience of a field of this shape is like . I know enough musical theory to know what a change of key is. in knowing what the shape of the field is .certainly on a representational theory there is no more to the quality of 's one experiences in experiencing blue than there is to the . and I know what this quality is. Mary knows . I know exactly what your experience is like in this single respect. despite knowing this fact about your experience. there is an ability sense of knowing -what -it -is-like that I need not possessto know what it is . She knows this ' quale of the dogfish s experience.know what it is like to look (phenomenally ) four-sided. As Lewis (1983 . Nemirow (1980 .

The fact that it is an electric field that has this shape is not what is being represented . dogfish have to know about the configuration of the electric field .of-key-quality . she has to know what shape they represent it to be. that something is that. They do not have to know that it is an electric field that has this configuration . it is like to hear a change of key. If we suppose that the temperature of a gas is the ' energy of the gas s molecules. then in representing the temperature of a gas to be 78 a thermometer represents the molecules to have an energy of 78 .Chapter 3 like (in the fact sense to hear a change of key. This does not mean I could tell . And the same is true of Mary . its shape. Knowing the quality of a dogfish ' s experience does not require Mary to be able to identify . Thermometers represent the energy of the molecular constituents as being 78 (rather than 77 or 79 ). via any sensemodality . electric sense It is important to understand that the dogfish is not being said to represent the electric field as an electric field . In order to eat. its geometry . in the fact sense I know what . by hearing . That is an ability that is not required to know what quality your experience has when it has the change. The theoretical constitution of whatever it is that has that shape is another matter. the thermometer does not represent the gas as having molecular constituents much less as having molecular constituents with an energy of 78 . not as being the energy of molecular constituents. If I know ) what a change of key is. then.much less an electric field of such-and-such form . what the dogfish identifies by its . In order to know how the fish represent the field . The same is true of Mary . It is like that where the " that " refers to what I know to be a change of key. But in representing the molecules to have an energy of 78 . What the fish represents about the electric field is its configuration . She does not have " .

In charting the discriminatory powers of the dogfish willi respect to electric fields (assuming the electrical senseof these fish has the function of informing them of the shape of this field ). but only I can express what we both know with the words : " My keys are on the table. Why should they ? So. Not being aware of any electric field . Once this is known . Suppose a fish experience an electric field of pattern P. You and I both know my keys are on the table. Mary is determining the character of the fish ' s perceptual experience how their experience represents . Mary is thereby learning exactly what it is like (for dogfish ) to sense electric fields. May knows exactly what the fish is experiencing . Mary knows nothing that can be expressed in quite this form . the fish experiences in the way the fish experiences it . surely . Mary cannot. Nonetheless Mary cannot herself describe what she knows . If these fish could think .and there is nothing preventing Mary from knowing all of it there is nothing more to know about the dogfish ' s awareness of electric fields. does not show that the dogfish experiences ' something that Mary doesn t know about. inrunning discrimination tests on dogfish . the electrical environment . She does not because dogfish do not representselectric fields as electric fields. For Mary there is no object she is aware of (a this) that is an electric field of pattern P.Qualia to know what representing something as an electric field might be like . But this . fields ." . this fish could think that this (referring to an experienced electric field ) has pattern P. Knowing what she does about electric .21 There is one thing dogfish can do that Mary cannot do. Mary is determining just which defonnations of the electric field the " " dogfish is sensitive to and which it is blind to. exactly what the fish is referring to with its demonstrative " this " and what shape the fish is experiencing it to have.

What I am experiencing when I experience blue is something I can describe by saying. " This is the color blue. not be able to express what you know in the way I do. This assumption conceals an important problem .. The problem has proved particularly vexing in the caseof color -called secondary property ) perception . And the fact that ' Mary doesn t stand to the elechic field in the relation that allows her to refer "to it in the way the dogfish experiencesit (as a this) does not show there is something about the dogfish ' s experience that Mary does not know. There is nothing you can refer to as this that is the color quality I happen to be experiencing.yet. The fact that we cannot refer to what we know to be on the table in the same way does not mean that there is something different we know . becauseyou are differently situated.thus know exactly what I know . And the fact that Mary cannot experience electric fields does not show she cannot know exactly what dogfish experiencesare like . what objective property do experiences of color have the functio ~ of indicating ? Is it wavelength of the incident (on .3 Chapter You have to use different words to refer to what you know to be on the table (what it is you know I know ) . This does not show you do not know what the quality of my experience is. but what objective property is this ? If color qualia are identified with the properties that color ' experiencesrepresents and experiencesrepresentsthe objective properties they have the functions of indicating .e . ) of? Red yes.g . What (and other so property is my experience of red an experience (representation .22 I have been assuming throughout this discussion that there was no particular problem about the determinable . being represented only a question about which determinate form of that determinable the system represented at a given time ." You might know I am experiencing blue . " Your " keys are on the table.

all those that give rise to the same color called metameric matching . light of different wavelengths will arouse the same color experience. there are no objective conditions our experiences of color inform us about. we air left with the question of .. Depending on conditions . If . in one class all the diverse objects we designate as red is the fact . then we cannot ? . Even under ideal conditions . that they all give rise to the same sort of color experience the experience of redness But if this is so. properties that vision has the function of informing .Qualia the retina ) light ? Is it a more global property of the entire optic array (see Land 1977)? Is it . perhaps . not objective .e a system (like vision ) should have been selected to why indicate the presence of this highly disjunctive (and altogether ' motley ) collection of distal conditions? Isn t the truth of the matter that the only thing that holds together. If this is so. some surface ? property (reflectance ) of the objects that reflect light to the eye? The difficulty in finding a plausible objective property that the visual system is responding to in the case of color has led many philosophers (Hardin [ 1986 being a recent ] one) to subjectivism about color.the fact that a wide (infinite . in fact) set of objective circumstancescan give rise to the same color experience . There is no particular spectral reflectance a surface must have to be seen as green . Subjectivists do not think that colors are objective properties . a host of different surface qualities can give rise to the same color experience. This general phenomenon. then. about. Depending on illumination . surface spectral reflectance then either color vision in humans is doing ).g . a strikingly bad job or colors are complex disjunctions of objective conditions . and color is identified with some objective property of objects (e. the function of color vision is to indicate color.thus . Therefore the color experience does not " tell " us ~ e wavelength .

This is a complex issue and I cannot hope to do justice to it in the space of a few pages.23But.that used information about the rate at which the axle rotates to provide information about speed. Since it is cheaper to manufacture . not axle rotation . the representational states of this device (pointer positions ) thereby have the systemic function of indicating speed. we have a speed-analog of metameric matching . in fact) number of different speeds that can .a speedometer . there is no need to go to additional expense . Such a device is cheaper to manufacture than an instrument that ( by taking into account elevation of the axle above the road surface) automatically compensates for tire size.Chapter 3 define the quale red in terms of some objective property an (thus experience has the functions of indicating representings . 3) I described an instrument . On any rational selection procedure. There are a variety (infinite . Earlier (chapter 1. The designers selected the cheaper mechanism for installation in cars becauseit did the job well enough well enough to justify its selection over a device that was more versatile and reliable but substantially more expensive. once again. and every car in which the device is to be installed nonnally comes equipped with standard size tires. an analogy with a conventional system may help to indicate the proper response to this problem . Once we have a vehicle equipped with such a device. Given that it was selected for this job . For we are specifying the objective condition > ) (the highly disjunctive set of surface spectral reflectances the experience has the functions of indicating by means of the subjective qualia (experiencesof redness these objective ) conditions produce . even though the system uses information about axle rotation to do its job . this mechanism (rather than the more sophisticated one) would be chosen for the job .

rpm indicator the function of indicating vehicle-mph . etc.what it therefore has the function of doing . We say that it is a speedometer We the same device into an AXLE -RPM METER by could make giving it that job . In the case of instruments . in vehicle speed. All it means is that to determine what a system s functio ~ is one has to sometimes determine the conditions in which it was selected to do it . So the insh" umentis given " " a job to do . we increase the number of ways it can go wrong without malfunctioning .Qualia give rise to a representation of 78 mph : a speed of 78 mph with normal tires . If we gave it that job . We are interested. But we didn t do that .indicate speed. The fact that does not mean it does not have that functions of indicating ' speed. a speed of 62 mph with small tires .indicate car speed. We give the .indicated speed in a certain specified set of conditions (normal size it cannot do that job in other conditions tires). By giving our axle. primarily .depends on the (possibly ) special conditions existing at the time at which selection occurred. If people start using tires of different sizes. It was selectedto do a certain job becauseit did that job . we (designers and users ) ' determine which of the quantities on which the insh" uments behavior depends are the ones it represents. insh" umentits function . then this device will no longer be able to do what it is its function to do . It is hostage to these circumstances in the sense that unless these circumstances are right . the system will not be able to do its job even though it performs as well as it can perform . We increase " the number of what Matthen (1988 called normal misper ) .what it therefore represents . not (except as an indicator of speed) axle rotation . a speed of 93 mph with large tires. For what it was selected to do . tire size would be irrelevant to performance of its function .that makes it hostage to circumstances. irrelevant to the ' accuracy of its representations.

and you cannot always tell what a system is designed to do . in addition .to more " distal " conditions . consequenceif there were physically possible but not actual (nonplant ) metameric matches to green plants that the system " could not discriminate . states are. By extending outward . we make it harder and harder to tell . For what its states have the function of indicating depends less and less on what information the system delivers when functioning properly and more and more on the . exactly what it is supposed to do.the number and variety of different objective conditions that can bring about the same representational state. ) . tires) as going 62 mph . performance of that function in special circumstances A number of authors (Hatfield 1991 1992 Shepard 1992 . of color vision is to enhance the discriminability of. It is using designed . thus . We. . We enlarge the number and variety of metameric matches . it would be of no green plants from soil and rocks. Hatfield (1992 observes that if the function ) .what the system has the function of indicating .what it . . increase the number of ways a system can go " " wrong while performing optimally . represents-:-by looking at what it actually does in present conditions even if you know that what it is doing is what it is supposed to be doing . on whether it is doing it in the kind of conditions that resemble the conditions in which it was selected to do that. it is. When the speedometer described above misrepresents the speed of a car (with under -size . say then " . Judgments of function depend on what a system is designed to do. made this same point about the relevance of actual historical have conditions in evaluating the function of a system and. by examining the insb" ument what its representational . thus (on the present construal of representation) its representation efforts. therefore .what it thereby has the function of doing . in a sense doing exactly what is to do .Chapter 3 " ceptions. It depends.

but it doesn' t show that it doesn' t represent that infinite variety of conditions in a single way . The same is true of color vision .Qualia information about axle rotation to tell . This is true . but that is because we have been considering. in exactly the same way. not with the instrument . when working properly .do I really understand what it is like to be a bat? Things are not this simple . how fast the car is going . but with the ecology in which it is ' being used. 362) view : that color is whatever property it is the function p of color vision to detect. It only shows that which property it is may no longer be obvious from the variety of conditions that cause us to experienceit .not with the kind of reliability for which it was selected It would be wrong to infer . to an infinite variety of different speeds.thereby knowing exactly where the moth seems to the bat to be . The trouble lies. thinking to its logical conclusion. we reach Hilbert ' s (1992 . S What Is It Like to Be an Experiencer of Elecbic Fields? But does this really mean that we can objectively determine what it is like to be a dogfish ? Or a bat? Can things really be this simple ? If a bat ' s sonar system represents the moth as . being there and I understand just where (and what ) there is . not what it is like to be a dogfish . The fact that so many different conditions cause us to experience red does not show that what we experiencewhen we experience red is not an objective property . If we pursue this line of . of course. that this device does not have the function of indicating a simple objective property (speed) simply because. it responds. These aren t the conditions in which the instrument was selected to perform } 4 The algorithm it uses to compute speed no longer applies . in accordancewith a built -in algorithm .

Until we know everything they do by way of representing their surroundings . know what it is like to be a dogfish or a bat. in seeing objects we . If did not understand color and that. They do not understand what it is like to seeobjects. properties that S senses In knowing exactly what configura tional qualities the dogfish experiencesin sensing the deformation of an electric field .3 Chapter or a bat. there is an aspect of our experience of objects that they do not understand . size.the qualia we experience (and they knew we experienced) in seeing objects. To know what it is like to be S we have to know (with regard to sensory qualia ) all the . but what it would be like to do one of the things that dogfish and bats do . If he knows what movement is. in other words . Does the fish not only sensethe spatial layout of the field . An experience of movement whether it be visual . they if that is. shape. Mary may still be ignorant about what else there is to the fish ' s phenomenal world . what other properties of its surroundings or the electric field (besidesits geometry) the fish is experiencing. but variations in its intensity ? Is it also sensitive to induced magnetic fields? Phase relations? It would be a mistake for aliens to conclude that they knew what it was like to see objects because they knew everything there was to know about location . We won t . But that is surely not all that dogfish and bats experience. a difference between knowing what it is like to senseF (which S does) and knowing what it is like to be an S (which sensesF). . therefore. and orientation . There is.sense the geometry of electric fields or the spatial location of moths. we will not know everything there is to know about how ' they experience the world . indeed. also represent their reflectance properties ( what color is). A blind person may know what it is like to visually experience movement . that is enough.

or kinaesthetic. determinable properties . 264) makes a . they are. with respect to this single property . we do not necessarily know what it is like to be a bat. Even when the sensesoverlap in their representational efforts . similar point . One also experiencesthe object' s shape. in knowing only how his sonar represents his surroundings (the film might give us that). and if these experiencesare all of the same property .much more . color.Qualia they do in the case of spatial properties .that in seeing what the bat senses (via his sonar apparatus ) as he swoops about the cave we do not " feel the additional " sympathetic sensations appropriate to such movements. direction of movement . I also have to know what it is like to sensean adrenalin rush. 35) represent -different ranges of .has its qualitative character defined by what it is an experience (representation) of .they ( McGinn 1991 p . In describing a hypothetical film (visual representation) of a bat ' s sonar perceptions . the same kind of 25 experience. I do not know what it is like to be you watching an angry lion charge just because I know what angry lions look like . In knowing what it is like to visually experience movement. Akins (1993 p . . This is why seeing and feeling movement are much different even though the same thing (movement ) is represented in both modalities . subjectively . a blind person does not necessarily know what it is like to visually experience an object that is moving . That would require knowing how else the bat was representing his surroundings and his own internal state. though .involved in seeing an object move than experiencing the object' s movement . size. ~or there is more . and a host of other properties . So.

but are not conscious of having them . after all . Yet . This point is simple enough . This chapter is an attempt to clarify some of these issues and to give .a difference in what the person knows about the experience he or she is having . then . Are there . E. a fuller picture of the role experience plays in our mental lives . Are there . It is . thereby . like cancers . Others have it .4 Consciousness Some people have cancer and they are conscious of having it . but are not conscious of having it . others we are not . . Some of them we are conscious of having . ' Moore s ( 1903) meticulous ac~ount of this difference . It is adifference in the experiencer . two forms of cancer : conscious and unconscious cancer ? Some people are conscious of having experiences . then . It may even appear obvious . tWo sorts of experience : conscious and unconscious experiences ? Experiences are . in this respect . despite G . and the disastrous consequences of ignoring it ) failure to distinguish the object of awareness from the act of awareness contaminates contemporary accounts of consciousness . a straightforward distinction between what we are aware of and our awareness of it . Others have them . But the difference is not a difference in the experience .

we are conscious ofthings . unlike the creatures in which they occur.Chapter 4 1 Conscious Beings and Conscious States Stones are not conscious but we are. . are not conscious of anything or that anything is so although their occurrence in a creature may make that creature conscious of something or that something is so. I call ) all these creature consciousness. fears.). II II II Accordingly . by ignoring these nuances.1IHe also describes the different liaisons they have to noticing . White (1964 describes interesting ) differences between the ordinary use of IIawarell and IIconscious.of objects (the bug in my soup). or es processin that being . States(process . nothing essential is lost. condition . events (the commotion in the hall ). Following Rosenthal (1990. I think . Being conscious of a thing (or fact ) is being aware of it . but to some state. and facts (that he is following me). . conscious beings) are said to be either conscious or unconscious . conscious awareness and consciously awareII are redundancies. The word is applied to beings who can lose and regain consciousnessand be conscious of things and that things are so. attending . and experiencesas being conscious or unconscious we attribute or deny consciousnes .3 A . When we describe desires. For purposes of this discussion I will regard IIconsciousll and IIawareII as synonyms when used as transitive verbs.the sensein which certain mental states process . events. etc.2And so are many animals . Creature consciousnessis to be distinguished from what Rosenthal caIls stateconsciousness. and realizing . and attitudes (in or of es . not to a being. properties (the color of his tie). We are not only conscious (full stop). Though my use of these expressionsas synonymous for present purposes blurs some of these ordinary distinctions .

to suppose that creature consciousnessof something (whether object.French horns never become aware that they are aware of anything . seeing a person is one way of being conscious of the person. I have. then one' s experience of it must itself be conscious . Mice who hear. not later . but if you see smell or hear it .and thus become aware of. relevant sense conscious of it . that seeing. in the . or smelling him . " The words " in the relevant sense are important . . One can also become conscious of the person by feeling . hearing ." Wrong again! Hearing a French horn is being conscious of a French horn .without . event. Given this usage. that is. .absorbed in one s work ) not be paying much attention at all ." Wrong ! Not being aware that you were aware of a French horn does not mean you were not aware of a French horn . If later asked whether you heard a French horn .much less French horns.4You may not pay particular attention to what you see . being aware that one is aware of an F (see the earlier discussion of this point in chapter 1. or hear. you might well reply " No . seeing. it may seem natural to suppose that if one is conscious of some object.not necessarily that it is a French horn . merely been describing common usage. 3) . One can be aware of (hear) the sound of a French horn without being aware that that is what it is. So much for terminological preliminaries .Consciousness I assume. One ' might think it is a trombone or (deeply .perceptual fonns . therefore. tasting. hearing . furthermore . hearing . smelling . and smelling are species Thus. but when you are hearing the French horn . you might (thinking it is a trombone) reply " No . proper - . smell. For one can be aware of an F (seeor smell an F) without being aware that it is an F. you are. I hope. and feeling are specific fonns . of consciousness Consciousnessis the genus. It may seem natural . Even if you are asked this question .

in some sense conscious of one neces .that characterize . We are often aware that we occupy such states . Conscious mental states. in particular . not to be accepted at all. ' s auditory experience of the piano playing must be a conscious experience. French horns ) the states themselves can be described as conscious.experiences . there are dangers in making it . For the same reason.that we are experiencing this and thinking that . be conscious of them. shapes movements .as an experience one is. If . we must . " " Though it is natural to make this inference. the experience. to be sure. a person is still conscious of whateyer properties . then her belief (fear) that someoneis following her is a consciousbelief (fear).as many people do . or fact ) requires some state in the creature to be conscious . sounds. . . but also of one' s experience of the tree and piano .are . The experience of these properties is conscious .thus being aware of the piano playing . But we must be careful not to conclude from this that because the states are conscious. It is not to be accepted lightly . no physical object is actually perceived . in fact.Chapter 4 100 ty. states in (or of) us without which we would not be conscious of trees and pianos . If one hears a piano being played.but that no more makes the states (of which we are aware) conscious than it makes a cancer (of which we become aware ) conscious . It is. This conclusion is most peculiar . if a woman is aware (or simply fears) that someone is following her. trees. There are. That doesn' t follow .colors . Since these experiences make us conscious of things (pianos . in the case of dreams and hallucinations . For if one thinks of a conscious experience . sarily accepts the conclusion that one cannot see a tree or hear a piano without being aware not only of the tree and the piano . We call these states experiences . The dangers arise from a possible misconceptio of what it is for a state to be conscious.

4 for this kind of misrepresentation). I mean to be speaking of the properties. not stateswe are conscious of They are states that make us conscious. and feel. is conscious when we are. then it is also conscious when we are not. can be . these experiencesareor . ( that when we perceive something we are in a conscious state which we could be in even if we were not perceiving that thing . make conscious by being conscious of them. The person in whom this representation occurs is. They are states that enable us to see hear.Consciousness 101 statesthat we are consciouswith . as I say denying this. aware of these proper . . When I speak of experiences . . In such cases the internal representation says or .as conscious as veridical experiences As Millar 1991 p . not states that we . I am not . " means" that something is pink and rat shaped. So if the experience experience can be the same as when . In dreams and hallucinations . " Nothing seems less problematic than . one occupies representational states that are not " tethered'i ( by contextual relation C) to an external objectthe kind of object these stateshave the function of informing " " about. to be sure. experiencesmake one conscious of some object The experiences one has during dreams and hallucinations do not make one aware of any objectS yet. A point of clarification before we move on to defend this general picture of consciousness. not states that we see .any object the we are." When we are not perceiving anything . not the objects (if any) the experience makes the creature aware of. making a creature aware of something. or feel. hear. but nothing stands in the relation C) is pink and rat (to which one shaped. I am not saying that all . The sensory representation representssthis as having these properties when nothing does (see chapter 1. 2) puts it .

Without this representational state. Neither would " see" pink rats.) are not qualities of a mental event.7 Pain is not a mental event that is made conscious by one' s consciousnes of it . The other does not. not (therefore) in the qualitative character of their experience. Though I have not explicitly addressed the topic in these lectures. strained . the cluster of properties one hallucinates or dreams that some object has. itches. a pink rat . etc. Just as a visual experience of a tree is an awareness of a nonconscious object (the tree) pain is an awareness of a nonconscious bodily condition (an injured . not in what properties they experience not in the properties they . They are.Chapter 4 102 ties. thirst . say that the hallucinator or dreamer is aware of certain images as long as we understand that these " images" are not objects (let alone mental objects) that have the properties one hallucinate -5or dreams that something has. They are exactly the properties (pinkness. In the case of hallucination . hunger. nausea. though there is no consciousnessof any object that has these properties . tickles. hunger. fear.) as those who see pink rats are aware of. etc. I would say the same about pains. they are properties of the . One person seesan object. rat -shapes . What makes a creature conscious of theseproperties is the same thing that makes a person who sees pink rats conscious of them : an internal state that represents something to be pink and rat -shaped. if we like . instead. The difference between them lies . are aware of. The qualities we are aware of when we experience pain (thirst .6 We can. that has the properties the experience represents something to have. and a variety of other feelings and emotions that are commonly thought of as conscious. neither the dreamer or the perceiver would experience the properties they experience. or diseased part ). but in what objects they experience.

When things are working right . the type of experience one has is determined . hunger . but by the properties one s experience represents something to have . thirst . Though we can be. but the bodily states that these representations (pains ) represent . or nausea. perhaps. What we are conscious of when we feel pain ( hunger.have it to others. . experiences are awarenessesof objects not objects of which we are aware. As we saw in chapter 1. as I say is a topic I have neither the time nor (I . My project in these lectures is more restricted in . If the properties represented are the . and itches stand to physical states of the body ) (they are experiencesof these physical states the way olfactory . admit ) the resources to effectively pursue . yes. visual . respect to these experiences then it will be time enough to turn to our experience of internal affairs and. But this. like visual .and most often are . and anxiety. and auditory experiences stand to physical states of the environment . It is an argument about senseexperience the kind of internal representation that (normally ) makes us aware of external objects. pains. So I leave the surface . scope. but not because we are conscious of them .Consciousness 103 physical state of the body an awareness of which is the thirst . In all cases the experiences are . conscious. tickles. even " " the sort of diffuse emotions and feelings ( diffuse because not necessarily directed at or upon anything ) mentioned above. pains. despair.) are not the internal representations of bodily states (the pains). not by the object experienced (there need be no ' object). but becausethey make us conscious of the relevant states of our bodies. which do not at least not on a representational component . If the argument can be made good with ' . etc.aware that we are in pain . Nor do I have anything (plausible ) to say about such emotions as joy . dread.

and perceptual consciousness If there is an inclination to doubt . if you will . whatever objects ( bearing C to the representation) these properties are properties of . particularly at night . theory of consciousness Conscious states are natural representation . They make one conscious of whatever properties the representation is a representation of and. That .Chapter 4 104 same. and as part of that mental activity . The coming -to is an alarming experience. 2 Higher -order Theories of State Consciousness After qriving for long periods of time . It is natural to describe what went on before one came to by saying that during that time one . but because they make this creature conscious of something .representationssin the caseof experiences rep. both the person who perceives a pink rat and the person who does not are in the same representational state hence. in the caseof thought . Call this a horizontal theory to distinguish it from the vertical theories (HO theories ) with which I now contrast it . Conscious creatures are resentationsa creatures in whom such states occur. the same perceptual state. this. There was mental activity. it is " " possible to come to and realize that for some time past one has been driving without being aware of what one has been doing . then consider the extraordinary sophistication of the activities . consciousnesswas present. That is to say there was minimal consciousness . in the two sensesof the word that we have so far isolated. lacked consciousness Yet it seems clear that. Both states are conscious (I have alleged) not becausethe creature in whom they occur is conscious of them (this mayor may not be so). if there is such. is the representational . there was perception .

and the curves in the road. . Ahigher good " order awarenessof an experience (something the truck driver lacks) is what makes the experience conscious. activity ) is conscious only if the person (or animal ) in whom it occurs is (creature) conscious of it . So the driver lacks an awarenessof certain items (his " own visual experiences that drivers in more alert statespre) have.8 A mental state (process. The driver not only saw the other cars. attitude . Arm strong concludes. despite possessing these two forms of consciousness.Order ( hereafter HO ) theory of state consciousness . What the exhausted driver lacks is not experiences of the road. By " perceptual consciousness he means that the driver was conscious of other cars. he saw them as cars (at least things to be avoided ). as curves in the road (how else explain why he turned ?) . Nonetheless. the curves. curves in the road. Arm strong is here propounding what has come to be called a Higher .that he is seeing the road . The driver is not aware of his own perceptual activities . If . There are two speciesof HO the- . as stop signs (why else did he stop ?). 59 ) Armstrong means In speaking of " minimal " consciouSness " that the driver was not in a coma or asleep. one is not intro sumably spectively aware of a mental state (like the visual experience the driver is having of the road). What the driver lacks is introspective awareness a perceptia . These objects were represented at both the sensory and (in some appropriat ~ way ) at the conceptual level ." success undertakenduring the period of " unconsciousness fully (p. the stop signs . the stop signs. but conscious experiencesof the road. . stop signs. and so on. then it (the experience is ) " " in one sense of the word unconscious .Consciousness 105 " . :n1ike awareness of the current states and activities of his own mind . Arm strong thinks there is a form of consciousness the driver lacks.

It isn ' t a higher -order awareness that one is experiencing the road that makes the road -experience conscious . One monitors the internal environment . the truck drivers experience of the road is nonconscious if the driver is not aware of having (i . What makes the driver ' s experience of the road conscious is that there is some higher -order sensory. 1990. Lycan (1987 1992 is another such theorist and he describes the theory as a " Lockean " inner sense theory of state . not necessarily conceptual .. ) Arm strong ( 1968.4 Chapter 106 " " ory corresponding to the two senses of aware we have already distinguished . as its object (what it is a t J1ought about ) the (lower -level ) experience. What makes the sensory representation (of the road) conscious is this higher -level conceptual representation of it . have a sensory representation (i . or somehow conceptually represents this experience (or itself as having this experience). -consciousness The idea behind this variant is that a mental state becomes conscious by one' s having a higher -order perception ( hence experience) .e.with a mental senseorgan the 's way one scans the external environment with one eyes. ears.e. ' s own " " experiencesand thoughts . ) thought of as a HOE theorist. There must be a higher -level thought (or thoughtlike state ) that takes . aware that he has) such an experience. of it . David Rosenthal ( 1986. representation of it . 1993b has been an articulate spokesman for such a view.. and nose. without . an experience ) of burning toast . but ahigher order awareness of the lower order experience. 1980) himself is perhaps lo better . Just as one can smell (be aware of) burning toast. The first maintains that what makes an experience (the sort of mental state we are here concerned with ) conscious is that the creature whose experience it is believes. According to HOT . HOT (for higher -order thought ) theories and HOE ( higher-order experience) theories. knows .

) think animals do not have conscious experiences. do not have the requisite higher -order . nonconceptual) relationship to the . Since HO theories trace the consciousness of a state to some higher -order representation of that state. the experiencescannot be conscious becausethe animals. having no concept of experience (or pain ). something in the nature of a thought or judgment about a mental state (like an experience ) to make that state conscious. according to HOE theory.without conceptually representing the road experience as a road experience.. judge . pains). for the experience to be conscious) anymore than one has to think . You do not have to think . hence.. on the other hand . This. smell) it . nothing in HOE theories that would preclude animals (not to mention young children ) from having conscious experiencesand thoughts . or believe that you are having a certain experience for you to be aware of it (and .e. HOT theories require. be aware of a road experiencewithout knowing or judging that it is an experience (let alone a road experience). one can. Though they may have experiences(e. This is why some HOT theorists (see e. Carruthers 1989 . es HOE theories from HOT theories.. then.ll If the truck driver s experience . judge . for purposes of assessing such theories. it is.g .e. important to understand how the higher -order representations are supposed to represent ' the lower -order state. since they require something more in the nature of a sensory (i. There is. experience do not deny conscious experiencesto organisms on the basis of their lack of appropriate conceptual simply capacity. thoughts that they have such experiences HOE theories. while HOE theories do not . HOE theorists do not .Consciousness 107 knowing what it is.g .. or believe that the toast is burning to be aware of (i . as far as I can see is the only thing that distinguish .

as an experience.4 Chapter 108 of the road is conscious only if the driver is somehow aware of it . brain . It will not yield a representation of an experience . therefore . the mind is there and that. would this make the road experience conscious? All one can become aware of by ~ anning (monitoring choose your favorite word ) internal affairs are activities of the nervous system. HOE theories appear to be making the mistake of inferring that because mental representations (representational vehicles) are in the head. would it be enough for the driver to see (with the help of neurosurgery and mirrors ). All that experience (in whatever modality ) of an experience gives one is a sensory representation of some part of . a material state of his .thus becoming (noninferentially ) conscious of his own experience? If the truck driver became absentmindedly conscious of his roadexperience in this way.that part or processin the brain. the facts that make these vehicles into thoughts and experiences You cannot represent a thought or experience . the brain . we can become aware that we think and experience and of what we think and experience by looking within . by seeing. as a thought or experience . They identify particular experienceswith token states of the brain or nervous system. is all that is in there. the relevant state of his brain . hearing . not the contents. To make this state of the brain conscious. in the same way he was conscious of the road. you cannot achieve metarepresentation. after all . therefore. how must this experience be represented in the dri ? ver ' s awareness Must the experience be represented as an ? As a representation? Would it suffice to represent experience it as a state of the brain ? HO theorists are typically materialists. That. by ( ) . smelling or tasting the thought or experience itself . The truck -driver ' s experience of the road is. or process in . All that is in the head are the representational vehicles.

HOT theories . is how most people . This. An internal scanner is as useful in mental affairs as would be a high -resolution camera in deciphering the meaning of coded text. All that is necessaryis that one be aware that one is having them. That is the same mistake as inferring that because words (meaningful symbols ) are in books.e. etc. the meanings are also there. according to a HOT theory of conscious states be .) meaningful symbols is not the way to represent their meaning (or the fact that they have meaning). some sensory. some thought or judgment that it is an experience (perhaps . sense them) for these experiencesto be conscious. HOT versions of state consciousnessavoid the mistake of HOE theories by not insisting on some quasi -perceptual .e. without sensing the state itself. why suppose.. not sensory awarenessof an experience but a conceptual awarenessof it . Why should his? . require. Since one can conceptually represent an experience as an experience without an awareness of the experience (see the discussion of displaced perception in chapter 2).riot by perceiving the . one need not . presumably. as HOE theories do. that higher order experiences of lower -order experiences makelower -order ' experiences conscious ? Why should the truck driver s awarenessof a processin his brain make that process a conscious thought . relation to the mental states that are to be made conscious by higher -order awareness. find out that they have cancer. and one can be aware that one is in a certain state without being aware of.. Why. seeing.becoming aware of thoughts and experiences. Experiencing ( a way of representing them as mental or as meaningful ? But if it isn' t . touching . then. should experiencing meaningful brain states. an experience of such-and -such kind ).Consciousness 109 experiencing what is inside . aware of one' s own experiences (i . My awareness of it12would not have this effect.

There are. How could they have higher -order thoughts that things appear to be F. to be sure. First . however.the one being conscious. If that is a consequenceof a HOT theory. It is hard to see therefore. The question is not whether a two -year-old knows what a six-year-old knows (about its own experiences. ) is. therefore . be conscious that they have experiences ' . ) of the mind . the fact that they only begin to understand the mind at about the same time they begin to understand the concept of representation (see Pemer 1991 pp . 82. it strikes me as very close to a reductio (it would be a reductio if we really . the other not. in my mind . but that isn t the question . but whether the experiencesof ) a two. decisive. of course. Given these developmental facts.where this is understood to mean that things need not actually beF for them to appear that way if they lack concepts for experience or thought ? If they are unable to hold higher .year-old and a six-year-old are. they .order beliefs about lower order thoughts and experiences. 189 . two objections to HOT theories that are. children only begin to gain a conception of thought and eXperience(as ways of representing the world that may or may not be accurate ) around their third year (Perner 1991 Flavell 1988 Wellman 1990 . report ?) that tells them they have cancer. children before the age of three years are unable to conceptually represent themselves as experiencing or believing . things could have a higher -order thought of the requisite kind .Chapter 4 110 cancer but by perceiving something else (Xrays ? a medical . at this early age. . no accident. are we to conclude . how. as developmental studies show. This possibility makes HOT theories plausible in a way that HOE theories are not. that none of their thoughts and experience are conscious? They may not . fundamentally different . as a result of this fact. On a representational theory .

that . the requisite higher -order thoughts need not be conceptually very ) sophisticated. deafness. therefore . for instance. their experiences are pretty much the same as ours? That is what we say about their diseases . It may be (and probably is) different . In Rosenthal (1991a. etc. but the fact that they have no concept of experience is surely not the reason !t is different .). know or believe that they have experience .Consciousness 111 knew . and hears things render its experience of them .instead of merely having str. of course. I see no reasons to think that because animals have no concept of experience . because the dog experience dog has no conceptual resources for thinking . When a dog scratches. their experience is somehow different from ours. and he suggeststhat. Why not about their experiences Why collapse the ? ' s awarenessof X and the X of which S distinction between S is aware in this place. contrary to what I have been arguing . injury . smell the same things (as HOT theory acknowledges to be the case why should the child ' s ignorance of the fact that it ) sees smells. ng intuitions .do not .that it is an itch . and .that o their experience was not fundamentally different ). therefore . of course. that it is irritating . barring other deficits (poor eyesight . he says that .year-olds know less about their experience of the world but . or whatever (on a HOT theory ) one has to think about an experience to make it conscious? Rosenthal is aware of these difficulties . unconscious ? What is the point of insisting that because . 's is totally different from ours. but nowhere else? The Sameshould be said about animals. if they see hear. If two year-olds are as perceptually conscious of external events and objects as their older playmates. they know less about their thoughts and experiences their and experiences are different ? Why not just say thoughts what I just said: that two. are we to believe that the itch is not conscious or that the .

next to the tree) without knowing what it is. (32. Experiences (and all other lower -order mental states that become conscious) must be " viewed as existing in a " relevant sensory field of their own . and a thought can " refer" to an experience without the person (who has the thought ) understanding what an experience is. is an " inner sense theory of what makes a state conscious. The way this nonconceptualized referenceoccurs. resources" Or. he . It is a This won thinly disguised conflation of higher -order thought with higher -order experience. I can see an object (off to my . as I like to put it there are conscious differ - . A second objection to HOT theories relates to the fact that. thoughts can similarly refer to sensory statesby way of their " position in the relevant sensory field .33) He goes on: " Something of this sort presumably explains how higher order thoughts can be about sensory states even though conscious differentiation of sensory detail quickly outstrips " our conceptual resources (33). though . as Rosenthal admits (see above) " conscious differentiation of sensory detail quickly outstrips our conceptual " .left . but if this is the model we are to use to understand the way thought can be about experience without representin the experience as an experience then our relation . " .4 Chapter 112 the higher -order thoughts need only " refer" to lower -order experience in order to make those experiences conscious. One is asked to imagine an inner eye scan" " ning experiences and becoming aware of them without needing to know what they are in the way I can visually scan a room full of people without knowing who any of them are (or even that they are people). suggests is the way we can pick out or refer to an object in our visual field even if we don ' t know what it is. ' t do. " This . . to our own experiences is being likened to our relation to the objects we see and hear. .

but a difference in what we see a difference. or people .as I think one must . This type of argument is more fully . or people and think there are eight .Consciousness 113 ences between experiences of which the experiencer is not conscious: differences in state consciousnessfor which there is no corresponding difference in creature consciousness If . that there are differences in state consciousness for which there is no corresponding difference in creature .mind . When we marvel in thosemomentsof heightened selfconsciousness .that a conscious experience of 8 objects (all of them fully in view . experiencethe richness we marvel at is actuallythe richness . conscious differences in experience cannot be traced to differences in higher order thoughts about these experiences . in other words . The next time we seeeight and think there are eight . This being so. ) For anyone willing to admit that one object can " enter " a conscious. in oUr conscious experience of the cards. are differences in conscious experiences that a person may not know about. the (if you will ) eight-ness is in my experience of the . the richness ) .may not be conscious of. this seemsfalse.simultaneously . at the gloriousrichness our conscious .but about which we can and do make mistakes. My conscious : experiencescannot differ unless my judgments differ . but is simply available (408 . trees. Dan Dennett (1991 denies that the multiplicity .the world outside in of . consciousness The numbers " 8" and " 7" are more or less . one grants . all of them seen ) is different from one of 7 objects then one grants that there . No difference in what we thInk . world if it is not in my thoughts about the world . therefore . " all its ravishingdetail. trees. It doesnot " enter our conscious minds. One grants. of . ) developed in Dretske (1993c 1994 . I selectedthem in order to represent the number of arbitrary things we all think we can see. . The ravishing detail of the . We see seven cards.

it -':'-perhaps even becomesaware that one was hearing it (all along ) without really being aware of hearing it . I remember seeing the moustache. are not always reflected in my judgments about the world (or my judgments . I realize I was having a certain kind of experience . up to the moment one became attentive. My experience of this ravishing detail does ceaseto exist when I close my eyes. can the consciousnessof experiences be identified with higher -order thoughts about them? There are. One is sometimes even able to count the total number of chimes even though some of them have already occurred (and are no longer audible ) after one starts counting . How . years. I remember seeing something that . though . Or I talk to an old friend whom I haven' t seen for many . It then became conscious. if I make them. He has grown a moustache (got new glasses etc. then. What better explanation for why it becomes conscious than that one becameconscious of having it . about my experience) . in thinking about our conversation . . I actually experience this ravishing detail.) but I do not notice the addition . at the time . The details I experience. Suddenly. I was unaware of seeing.Chapter 4 114 world does not cease to exist when I close my eyes. One may be so absorbed in one' s work that one doesn' t " hear" the chiming of the clock or the sound of children in a nearby playground . It is " in " my experience of the world . So the ravishing detail is not only " in " the world . as we like to say one becomes aware of . So my judgments do not always perhaps they never.track my conscious experiencesof the world . situations in which it seemsappropriate to speak of experiencesbecoming conscious in a way that seems to lend support to HOT intuitions . There is a temptation to describe this (along with Armstrongs truck -driver example ) as a case of an experience that was . unconscious. to be sure.or a great deal of it . Later. That is.

) that I did not realize were there.does not require you to know or be aware that you are hearing a clock chime that you are having auditory experiences of such-and-such kind . We even have experiences that we never become conscious of having . I sometimes become conscious of coins in my pocket (stains on my tie. that they are hearing anything at all (I assumehere that mice do not have the concept of hearing .experience if you will . If we do not carefully distinguish consciousnessof X from the X of which we are conscious.onto the objeCtswe know about. indeed. then creature unconsciousness .the knower . Mice can hear the clock chime and they are not aware that they are hearing a clock chime or. But this is no reason to suppose that the experiencesthemselves are different from the experiences we know about.can eas ily be mistaken for state unconsciousness. Are not these experiences unconscious at the time I have them? If so. we are not conscious of having . We do become conscious of having experiences that . I was unaware of having .thus having an auditory experience of these events . etc. They are as conscious of the chimes ) as we are. What happens after we " wake up " and realize .not know before. isn' t this fact to be explained by the fact that I was not aware of having these experiencesat the time I was having them? Such phenomena are certainly common enough. I know something I did . at the time I was having it .the fact that one is not conscious of X. It is me that is different . The coins (stains) are still the same. at the time they are occurring . that .Consciousness 115 a moustache. Hearing a clock chime . This is especially true of mental affairs where the possibilities for confusion are so much greater.13Certainly animals and small children do.that the X itself is unconscious . and I do not wish to deny their occurrence. We must not project differences in us .

mind you . Aware of the chimes . The person becomes conscious of certain experiences (in the sense of becoming conscious that she is having them ) that she was not conscious of before .Chapter 4 116 that we have been seeing or hearing something for a long time (~ ce never do " wake up " and realize this ) is a change in creature consciousness not state consciousness No state . but because these experiences make one aware of different things .not . Hearing the clock chime without realizing that it is chiming (or that one is hearing it ) involves conscious auditory experiencesbecause these experiences makes one (the person in whom they occur) aware of the chimes. that they are chimes or that one is hearing them. Conscious states are the ones that make you conscious of things . . . The experience of eight objects and the experience of seven objects are different conscious experiences .14If some mental states and processes are conscious .involve changes of state (acquiring a belief that one did not have before). These changes in creature consciousne . . ~long with Rugg (1992 p . objects in the one case seven in the other. of us changes from unconscious to conscious. but not necessarily any change in the state of which one becomesconscious. 3 The Function of Consciousness One of the benefits of thinking about consciousnessin the way proposed here is that it yields a plausible and natural answer to questions about the function or purpose of consciousn . not because one is aware that they differ . One needs no higher order thoughts to have conscious experience of things . others not .eight .changes in what one is aware of . one can ask. All one needs is experience of things .

5) conclude . If this is not the case then. aware of E. " topic .Consciousness 117 275). as Davies and Humphrey . somehow. then it is clear that E' s causal powers (as opposed to S' s causal powers ) are unaffected by the fact that it is conscious. HO theories of state-consciousness make questions about the f ~ ction of consciousness hard to answer. then consciousnessis epiphenomenal} 5 Mental states and processes would be no less effective in doing their job . Why should the causal powers of a thought or an experience be any different ? If the consciousness of mental states and processes comes down to -order experiences of them. or higher -order thoughts higher about them. too bad for consciousness : " Psychological theory need not be concerned with this . exactly. According to HO theories of consciousness . 4. of es? having conscious states and process Those that are conscious must differ in some relevant way from those that are not. Or. asking about the function of conscious states in mental affairs would be like asking what the function of conscious diseases. whether conscious ones are more effective than unconscious ones. we take that job to be . worse . . The causal powers of a rock (as opposed to my causal powerS ) are not changed or enhanced by my observing the rock or having thoughts about it . .was in medicine.w ~ tever. ( 1993b. If what makes E (some experience) conscious is the fact that S (the person in whom the experience occurs) is.those we knew about. the biological advantage.if they were unconscious. they make the answer obvious : it has no function . What is the point . pp . but by its making its possessor conscious of something else whatever it was the state represented . being conscious of it . But if a mental state is rendered conscious not by its possessor .then the value of a conscious state or process would reside in what it makes its possessor conscious of .

of experience is. how could they ) find food and mates . for some people. or whatever ). conceptually representing it in some behavior relevant way (as a hungry lion .16 This.Chapter 4 118 And here the answer to questions about the biological function of consciousness would appear to be obvious . surely. avoid predators . get around obstacles and. I expect to hear. smell. what is important (one might suppose) is not seeing a hungry lion (sensuously representing a hungry lion ) but knowing (seeing) that it is a hungry lion . It isn' t what you see that is important in the struggle for survival . environment (all species of consciousness. it is what you know about what you see. of avoiding predators and finding mates. according to the present theory . is too quick . things that have to be done in order to survive and multiply . headed this way. in the business . knowing (seeing) where it is and where it is headed. is why animals are conscious. If animals could not see hear.compete with one who is not and the outcome is clear. do the thousand .year-old child might do ) isn ' t much use to one who is on the lion ' s dinner menu.i .and you are left with a vegetable. Let art animal who is aware of its enemies. and taste the objects in their . Seeing hungry lions and conceptually representing them as tawny objects or a large shaggy cats (something a two. not sim- . may be) not what the evolutionary advantage of perception is. It is the representation of one animal by another as a receptive mate. After all . spin webs. in general. but what the evolutionary advantage of senseperception. when you take away conscious states .e.where they are and what they are doing . build nests. Take away perception as you do . See g or smelling poisonous mushrooms ~ is no help to one who cannot see or smell (thereby coming to know ) that they are poisonous. The one who is aware will win hands down . That. dangerous. The question is (or..

It is the well as certain behavior . a discriminating ear. and a sensitive nose is of no help in the competition if experience always (or often ) yields false beliefs about the objects perceived . experience it is no trick at all to see sexually willing (or. and so on? Why aren ' t we all . not philosophical . without color experience (Stoerig and ' ).to know that the willing are willing and the others are not. that it is moving toward us. the beliefs. Good eyesight. That is the skill .the lion . the sort of discriminations normally found only in those who can perceive colors . Cowey 1992 why doesn t everyone? What use is experience in cognition if the same job (the processing of information needed for the determination of appropriate action) can be achieved without it ? These are respectablequestions. unwilling ) members of the opposite sex.17If a person can receive the information needed to determine appropriate action ) 8 without experience. not the experiences that normally give rise to such knowledge . As we all know from long . its color .Consciousness 119 ply the senseperception of receptive mates.and it is a cognitive skill .that gives one a competitive edge in sexual affairs. size. answers. in each sense ) modality . They deserve answers scientific . the equivalent of blindsighters who (it seems get information about nearby objects (indicated by statistically significant performance toward those objects) needed to detennine appropriate action without experiencing (seeing) them . So why do we have experience? Why are we conscious of objects and their properties . the knowledge .that (for iristance ) there is a lion in front of us. as the case may be. The trick is to seewhich is which .at . that is important in the game of reproduction . that is important . location . why don ' t we? If people can discriminate wavelengths. But the answers .relevant facts about these objects . and movements .

she was indeed picking up " visual " information from somewhere . she no longer saw it as being there. As Humphrey described Helen in 1977 (Humphrey and I . But I do want to suggest that . she regained certain visual capacities . I am not suggesting that Helen did not eventually discover that she could after all use her eyes to obtain information about the environment . 89).and that her eyes had something to do with it . p . perfect.would appear to be clear. .g .e between objects that differed in size or brightness became almost . . r 3-D spatialvision and her ability to discriminate . H. Although Helen originally gave up even looking at things . Humphrey (1970 1972 1974 worked (for . fly. . she no longer knew how that information came to her: if there was a currant before her eyes she would find that she knew its position but . a carrot ) . lacking visual sensation .Chapter 4 120 least in a preliminary way . There is an enormous number of disabilities . after six years she remained unable to identify even those things most familiar to her (e. . . ) seven years!) with a single monkey.. She did not recover the ability to recognize shapes or colors . Sheimproved so greatly over the next few yearsthat eventually shecould movedeftly througha room full of obstacles pick up and tiny currantsfrom the floor. as her training progressed it began to dawn on her that . . whose capacity for normal vision was destroyed by surgical removal of her " entire visual cortex . She was a clever monkey and I have little doubt that. . (Humphrey 1992 p . 88) Nonetheless . Helen.would call the sensations of sight . even if she did come to realize that she could use her eyes to obtain visual information . She never regained what we . The information she obtained through her eyes was " pure perceptual knowledge " for which she was aware of no substantiating evidence in the form of visual sensation. .not the least of which (Marcel 1988a is an inability to initiate intentional ) action with respect to those parts of the world of which they are not conscious. Shecould evenreachout and catcha e passin.

89) . and this deficiency can be crippling " . The function of sense experience the reason animals are conscious of objects and their properties is to enable them to do all those things that those who do not have it cannot do. . or to inter -relate them in space and in time . Blindsight subjects even if they can guess their presenceor absencein aforced choice situation . " All these subjects lack the ability to think about or to image the objects that they can respond to in another mode.a reasonable empirical conjecture for the purpose of experience . but S will not be able to te~ what they are. we have a suggestion (at least) of what the function of experience is: to help in the identification and recognition of objects. they are still not very good (Humphrey 1992 p . 8). then . That is or may be . .for why animals (including humans ) are. human patients never recover their vision t ~ anything like the same degree that the monkey did . At least Helen couldn ' t. had no (visual ) experience of them. 8). cannot avoid bumping into lampposts . or at least a function . in perception. This is a great deal indeed. it remains clear that people afflicted with these syndromes are always " deeply disabled" ( Weiskrantz1991 p . was unable to seethem there. of sense experience. . Despite the attention generated by dissociation phenomena . If we assume (as it seemsclear from these studies we have a right to assume that there are many things people with experience ) can do that people without experience cannot do. This being so. there seemsto be no real empirical problem about the function . conscious of objects and their properties .Consciousness 121 If we follow Humphrey and suppose that Helen . Unlike Helen. Remove sensation from S and S might still be able to tell where things are. " " Though they do much better than they should be able to do . Furthermore . ( Weiskrantz1991 p . though still able to see where objects were (conceptually represent them as there).

There would only be aproblem about the function of conscious"states and process if es there was a problem about the advantages of seeing . then . for state consciousness The function of conscious states is to make creatures conscious. . Maybe something else besides experience would enable us to do the same ' things . All it would show is that there was more than one way to skin a cat. since conscious states are simply the states that make one conscious there is a biological purpose. some competitive advantage.Chapter 4 122 that is a perfectly good answer to questions about what the function of experience is. animals aware of what is happening around them . and a great many other animals . a biological function for creature consciousness If there is a . are conscious of things . .more than one way to get the job done. the same purpose. in having biological purpose. and hearing .of whatever they need to be conscious to survive and flourish . That is why we . . It would not show that the mechanism that did the ' job didn t have the function of doing it . The moral of this story is that in seeking a biological function for consciousnesswhat we should be looking for is . but this would not show that experience didn t have a function . smelling .

as ' Johnny Mercer s lyrics put it . do not explain anything . a naturalistic account of experience by showing how it provides satisfying . account of the qualitative aspect of our mental life . in particular . if possible. subjectivity. the possibility of qualia -inversion . self-knowledge . It is time to confront and. spike this rumor . of course. The first four chapters are an empty exercise. however. If this is so. It -Jim-inate the is time to e negative. experiencer what it is like to be the ejcperiencer cannot be . to ac-cent-u -ate the positive . I keep reading that the quality of experience . It cannot be a correct thought . the sensation. Many of the doubts about the adequacy of a representational theory spring from more general doubts about the adequacy of any externalist theory of sensation.4) were designed.cognition distinction . There is a rumour going around that a representational theory is known to be false if not of then certainly of experience. then .5 Externalism Supervenience and The Jean Nicod lectures (chapters 1.intentionality .how things seem to the . and so on. Theories that are known to be false. it does not explain anything . of a variety of otherwise baffling phenomena explanations . in these lectures was to promote anaturalist My strategy c ~ theory of the mind and .

Chapter 5


reduced to the way internal events are (or were) related to external affairs .! Not only are experiences, like thoughts , inside the head , what makes them the experiences they are - unlike what makes thoughts the thoughts they are - is also in there. I dealt with one such objection in chapter 2 The fact that . experiences are identified and individuated by their relational properties (what they have the systemic function of indica"ting = what they represents does not prevent the per)' s9n in whom they occur from knowing what kind of experience he or she is having . You are in a better position than I to know what is going on in your mind - what sensations you are experiencing - even though what is going on in your mind is not in your head. That, though , is only one of the objections to externalism and, therefore, to a representational account of the present type . It is the task of this final chapter to show that these . problems are manageable It is not enough to say- as I have sometimes been tempted to say- that all plausible theories of the mind have the same problems and that , therefore , these problems should not be my problems. Since I think the only plausible theories of the mind are externalist in nature, this may be true, but it is not much of a defense. Critics will be quick to point out that it is an admission that all plausible theories of the mind are false.
1 Seeming and Supervenience

The Representational Thesis is an extemalist theory of the mind . It identifies mental facts with representational facts, and though representations are in the head, the facts that make them representations - and , therefore, the facts that . make them mental - are outside the head. A state of the

Externalism and Supervenience


brain is an experienceonly if it representsthe world in acertain way, and a state representsthe world in this way, or so I have claimed, only if it has an appropriate information -carrying function . Since functions (whether systemic or acquired) have to do with the history of the states and systems having these functions , mental facts do not supervene on what is in the head. What is in heads A and B could be physically indistinguishable and yet, becausethese pieces of gray matter have had relevantly different histories, one is a representational system, the other is not ; one is the seat of , thought and experience the other is not ; one makes the person in whom it occurs aware of the world , the other does not . For most philosophers these are unpalatable consequences . This is putting it charitably : for many they constitute um a reductio ad absurd of externalist theories of sensation " . Even if one agreeswith Putnam (1975 that meanings ) ' s 1979 1982 examples ' t in the head" and finds ain , ) Tyler Burge ( in support of the social character of thought (a version of externalism) convincing - thus finds oneself accepting the idea that thought (or some thoughts2 have a content determined ) by external factors - few are willing to say the same . about sensations H two individuals are physically the same, then, surely, the fact that they live in different habitats, the . fact that they have had different individual histories, or the fact that they evolved in different ways, is irrelevant to their . current experiences H one has a headache so does the other. , H one is having an auditory experience as of a piano being played (whether veridical or not ), so is the other. Even if thought depends on the environment in which one exists (or existed), feelings do not. Theories that deny this are not just implausible , they are false. It is all very well for a theory to . have materialistically attractive consequences Other things

Chapter 5


being equal , these would be reasons for materialists to accept the theory. H the theory were true , it would provide ' explanations of these facts. Ah , but there s the rub . H it were true! The Representational Thesis cannot be true if it makes what one is experiencing here and now depend- not just causally (no one denies this ), but as a matter of logic - on what happened yesterday or in the remote past. Not only does the RepresentationalThesis make what one think S and feels externally determined , that one thinks and feels is likewise hostage to environmental and historical circumstanc . Steve Stich ' s (1983) Replacement Argument dramatizes this fact. Imagine replacing a thinking -feeling " " being- you , say- with a duplicate , a person that not only 3but lacks lacks your history, any history that would give its -providing systems the relevant information biological and learning -theoretic functions . Such a. being would get the same information you get (through its " eyes," " ears," and " nose" but these systems lacking the appropriate history, , ), would not have the biological function of providing information - at least not if biological functions are understood (as here understood ) as products of a certain selectional " " process (see chapter 1, 2) . The senses (if we can any longer call them that ) of your duplicate would not generate representations. They would , to be sure, supply the information needed to drive the motor programs in ways that mimicked your behavior , but there would be no internal representations of the objects about which information was delivered . There would , therefore, be no experiences of no , beliefs about no desiresfor , these objects. There would be no , qualia . Putnam ' s Twin Earth examples have convinced many that the mental does not supervene ---at least not locally on the physical . That is to say, it does not supervene on the

the other not . Fred and Twin Fred can be physically indistinguishable (in all intrinsic respects and.and7ftherefore. who think that physically indistinguishable beings must be in all the same mental states. or what have you ) that the mind .data . we come to know what our experiencesare like . Physically indistinguishable heads can harbor different thoughts . among other things .4 I know that not everyone is comfortable with this result. I will have nothing to say to those who reject all forms of externalism.or evolved.e. experiences are subjective surrogates (sense. I' m trying to win a battle. true. But I have to start someplace so I start here. a representational theory of experience . at least as it is understood in our commonsense folk psychology .and. not the war. I begin my defense of extemalism. thoughts (understood as what is " ' " thought . I suspect from a certain traditional picture of how . I am arguing the contrary : that if an externalist theory of thought can be true. thought content ) ain t in the head . I am arguing with those who .indifferent environments ) even when the brains in question are now indistinguishable .. Since these causal relations can differ (if Fred and Twin Fred grew up . just as importantly . The refusal to tolerate externalism about sensation derives. is determined . The content of thought . an externalist theory of experience cannot be true. Some flatly reject it . yet. percepts . one be thinking that the liquid ) he seesis taking this much externali Sm for granted .Externalism and Supervenience 127 physical constitution of the person having the thoughts . According to this familiar story.i . an extemalist theory of experience can also be . existed between ' events currently occurring in the brain and the person s physical sUrroundings. by the causal relations that exist. think that although an externalist theory of thought may be true.

then it is hard to seehow something could look red to Fred. If this is what is involved in something ' s looking red . green to Twin Fred if they are in the same physical state. is red and bulgy . becomes directly aware of this datum . two utterances (in different languages . say) could be the same (consist of the same sounds) and .. : Theexperience we mean to be describing when we describe ourselves as seeming to see a red ball is identified with either an internal image (a sensedatum ) or the mind ' s awarenessof this image. Maybe thoughts are like utterances. Twin Fred. H this datum . the sensedatum . the external object that caused it if there is one) is also probably red and bulgy . yet . on the other hand. they cannot.Chapter 5 128 becomes directly aware of in (indirectly ) perceiving the external world . Thinking that k is green is not in the same physical state. this internal least not if materialism is true. the only spectator in this theatre .and only indirectly (" inferentially " of ) the properties of the external object that caused it . It is. that is red. On this way of thinking about experience an object' s looking . not quite clear how thoughts must differ in their intrinsic properties to be different thoughts . For an object to lookred is for it to causea red datum to appear in the theater of the mind . The mind . express quite different propositions .its color and shape . in fact. things must seem the same to both of them.maybe a ( tomato. Since they have different things inside them. make quite . Thinking that k is red is not itself red. For if it looks red to Fred. H they are in the same physical state. This argument doesn ' t work with thoughts because thoughts are not conceived of as having the properties that objects (thought about) are thought to have. red is constituted by direct and infallible awareness of an internal object. then Fred must be directly aware of something internal that is red. must be aware of something green.

and that is very lame . to an act-object view of sense experience. indeed. it will be said.Externalism and Supervenience 129 different assertions. yet. are not like that. This is another argument is as lame . because " " different they are in languages. The resistance I submit . as we all know . a view of experience that regards different experiencesas experiences of different internal objects. Since the image or datum we experience when we experience different colors is subjective . a speculation about why some philosophers might resist an externalist account of experience without resisting an extemalist account of thought . If one conceived of sense experience in this way. But. But an experience of red and an experience of green. one person s modus pon ~ ' s modus tollens. They must differ intrinsically . It is only a speculation about a possible motivation . But this is not an argument . If that going to object to a representational account of experience becauseit is incompatible with a sensedata theory of perception . it would be understandable why one would regard an externalist theory of experience as ' incoherent. internal different color experiencescannot occur in indistinguishable beings. unlike a thought about red and green. may be traceable to an implicit commitment . therefore. they manage to bedifferent thoughts ( have different content). the (or a) moti - . Nobody at least none of the people I am arguing with . are constituted by an awarenessof an internal image or datum that is the color the experience is of. if one took a phenomenal appearance to be an internal object (or an awareness of an internal object) that had the properties things appeared to have. indeed . Why ? Becausean experience of red and an experience of is a sense datum theory of perception . Maybe thoughts are like that: the same " " physical utterances occur in two heads and.

Twin Fred thinks it (the same stuff ) is flamis Given my polemical purposes . Though physically indistinguishable . Why not ? different experiences But enough about possible motivations . It is a question of whether what Fred thinks he (Fred) is thinking is the same or different from what Twin Fred thinks he ( Twin Fred) is thinking . 3) should be enough to convince one that what Fred thinks he (Fred ) is thinking will not be the same as what Twin Fred thinks Twin Fred is thinking . if Fred thinks about it . physically indistinguishable fellows who . Fred ' s lower .will suffice to eliminate it . will he think he is thinking ? Will it be the same as what Twin Fred. It is not a question about how (or even whether ) they could find out that what they think they think is different . What about arguments? Let me begin with Fred and Twin Fred. and friends .5 Chapter 130 vation for the resistance my hope is that exposure . if he thinks about it .without differing internally . becausethey have had different histories . wives .. Fred thinks that yonder stuff is flim . If different sense .e.making . this is an admissible way to set the stage.not to mention thoughts and desires . in this sense seem (i . A moment ' s thought (not to mention the results from chapter 2. His higher -order thought . With the stage so set. will think he ( Twin Fred) is thinking ? ffisofar as they have introspective accessto their own thoughts . it explicit . will what it seemslike to Fred that he is thinking be the same or different from what it " seems like to Twin Fred that he is thinking ? This is not an epistemological question. ) order thought is that k is flim . what . seemd6 different to them . their mental states will . experiencesare not experiencesof different internal objects should we suppose that persons that do not differ why internally could not have different experiences? They can have different parents. currently have different thoughts .

If (as we assume) Fred does not have the concept FLAM . what they think they think is different . they said they said different ? Yes Though the sounds they make in telling you what they said are the same.order belief ' " " says he believes the stuff is flim while Twin Fred s higher " he Twin Fred thinks it is flam . Fred ' s higher . For exactly the same reasons Twin -Fred will think he ( Twin Fred) is thinking it is flam . he certainly cannot be thinking that he is thinking k is flam . one cannot think (or doubt ) one is thfuking something is flam .Externalism and Supervenience 131 must be that what he is thinking about k is that it is f1im since the concepts available to Fred to represent what he is thinking about k are exactly the same as those available to him for thinking about k. Imagine two people making the same noises: they both " " " " say that k is flum ( flum is the sound they both make " " when they say what k is). " order belief says ( ) So Fred and Twin Fred not only have different beliefs about the same stuff . Though not apparent from the activities in their brain (these. Not only does k lookd different to them . what they say in making these sounds is different . are the same). the second that he said it was flam . Ask the second what he said. The first reports that he said it was flim . He will say he said it was " flum " (this is the sound he will make when he tells you what he said k was). So what they think they are thinking is different . He will say he said it was " flum " (this is the sound he will make in . One person means flim by flum . they say and believe different things about their beliefs. If one cannot think anything is flam . we are assuming . thinking k is flim . telling you what he said k was) . Ask the first what he said. Is what . the other means flam . ' The same is true of Fred ' s and Twin Fred s beliefs about their beliefs. Nor can he be wondering whether he is thinking k is flim or thinking it is flam . So Fred will think he is .

the . None of this should be surprising . Before we ask whether there is. not the content of their respective beliefs . . in so far as one can speak of how their own mental process es appeard to them . it is important to note that what Fred and Twin Fred think about the quality will differ . in the doxasticsense .that his experience of k is exactly like his experience of flim . or could be. Such differences in their (first order ) mental states will " show up " in differences in what they believe and say about their (first order ) mental states In the . becauseit prompts them to believe something different . Fred will think . does not address the main question. How things unlike how seemp ( ) things seemd is independent of what one believes (or is disposed to believe) about the k one experiences Just becausek . then if they have beliefs about their beliefs.correctly as it turns out . We have merely been examiriing the way differences in thought content (at the first order ) ramify throughout higher levels of thought . their own thoughts seemddifferent to them. just does not mean k looks different to them phenomenallyand it .5 Chapter 132 (perception of k prompts them in them different beliefs about k) but . This . is with the phenomenal appearances . flam to Twin Fred.that we are now concerned. If Fred and his twin believe different things about k. looks flim to Fred. Fred seemsd(to Fred) to bebelieving that k is flim while Twin Fred seemsdto Twin Fred to be believing that k is flam .the quality of their respective experience. they will believe they believe different things . It will be true nonetheless . phenomenal quality of their experience how things see ~ to Fred and Twin Fred . a difference in the quality of the twins ' experience of k. sorts of caseswe are imagining ( Twin Earth situations) none of this will be apparent from either the outside (to us) or inside (to them). though .

~ This is not so. believing it is Betsy believing k looksp like Babs is the same as believing k like Betsy since Babs and Betsy loo ~ looksp the same. . How their phenomenal experie~cesseemdto them is different . I can have the one belief without are ' having the other. Twin Fred. on the other hand .that k flim to him . not about the loo of Babs and ~ the loo of Betsy (these may be the same). Fred' s belief that k looks flim is really no different from Twin Fred ' s belief that k flim and Twin Fred flam .correctly . of k. then what Fred and Twin Fred think looksp about the loo of k is the same. Hence.Externalism and Supervenience 133 things . seeing the looksp -like-flam belief. 100ksp 100ksp These clearly not the same.once again correctly . they look alike. . being twins .that k looksp looks the way flam things have always looked to him in nonnal conditions . fact. in fact. will think looksy that k flam to him . If Fred thinks that k looksp looksp thinks k flam . For the same reason. Although believing someoneis Babs is not the same as . Babs looks like Betsy Even if . He will think . So the twins will not only have different beliefs about their beliefs . but . Dut about the ~ -like -like -Babs-belief and the Betsy belief . Babs is Betsy And we are now making a claim . Babsand Betsy are not the same person. they will also have different beliefs about their experiences . So Fred will think . The belief that k looks like Babs is adifferent belief from the belief that k looks like Betsy even if . about the way k looksp As a result of his perception of k Fred is prompted to have a -like-flim belief about k while Twin Fred. It may be supposed that this does not follow . in . is prompted to have a looksp their phenomenal experience of k seeinsddifferent to them. So the twins beliefs about their experiences to them. same k. must be different . Just because flim things are not flam things does not mean that flim things do not look(phenomenally ) like flam things .

Without the concept of F. 5 cannot be made aware of this fact. What this means is that if k does not lookd F to 5. The access one has to the quality of one ' s experience (unlike the access one has to the qualities of the external objects the experience is an experience of) is only through the conceptsone has for having thoughts about experience. He cannot be made con- . to the F-aspect of the external objects this experience is an experience of) . Hence. become aware of the watery this concept.that one is having an experience of this kind .. then. If we even if it looksp . there is no way one can become aware of the phenomenal character of one' s experience of a puddle (as one can become aware of the qualities of the puddle itself) except through a belief that it has this quality .. that is. of course.except . seeing) a puddle one can be made aware of its watery qualities (i .g. even if they have the same phenomenal experience of k.e. 5 is "blind " to the F-aspect of his phenomenal experience (though not . For with no experience lookpof the puddle without 's of one experience awareness of the watery loo ~ of . One cannot. As ':Vehave already argued (see chapters 2 and 4) one does not experience one' s experiences (of puddles ) in the way one experiences puddles . it will not seemdthat way to them. . that the reason k does not lookd F is imagine that 5 completely lacks the concept of F so that nothing can lookd F to 5. And this requires the concept WATER. then although something can stillloo ~ F to 5. By experiencing (e. without (therefore) the puddle lookingd like water. F 5 will not be aware of this fact. the puddle can only be awareness that the puddle looksp like water . furthermore . it can loo ~ like water ) without possessingthe concept WATER. however . by becoming aware that something looksp to have this quality .Chapter 5 134 It prompts them to have different beliefs about their own experience of k. Hence.

and that . How could it be? There is as much difference between seeing eight fingers and seeing seven fingers as there is between seeing one finger and seeing two fingers .but he cannot be made aware of it . things do not as easily seemdto be what they see ~ to be. That. I say eight .7 To illustrate this important result. I mistakenly take there to be eight . But how many did there to me to be? What are appearp the phenomenal facts? Believing I saw eight fingers. looksp to S therefore .things can loo F to him . without concepts we are blind to our intuitions . I am wrong . let us first take a case where k F and thoughS has the concept F (is. The only difference is that. why would ~ I think that I saw.Externalism and Supervenience 135 scious of it . as the numbers get larger. is how many there appeareddto me to be. Without enough time enough to count. That is how many I think there were.eight fingers ? I must think the fingers produced in me an experience of the kind that eight fingers normally produces in me. As the numbers grow larger. the correct answer to how many fingers there " " appeare~ to be is seven. That is how many I think I saw. harder to notice . the difference is harder to appreciate. though . In this case. At the phenomenal level seeing (exactly) eight fingers is not the same as seeing (exactly) seven fingers . With ~ apologies to Kant . and } seeall seven. nothing 8 looksd F to S. Given that I saw only seven . able to believe that k is. F). there were . You hold up seven fingers. but enough time to seeall seven. or looks . therefore. If I didn ' t think there appeare to be eight fingers." I have no choice. When asked how many fingers there appeared to be. . therefore. His experience can havethis quality . harder to detect . Since then. I will (if I understand the question) describe the phenomenal appearances by using the number " eight . . one experience is different from the other.

If things see F to S. Knowing how to count .like a change of key to me . in this exchangeabout the music. become 9 I will then become aware this fact by counting the fingers .the concept needed to make k appeard to S to have the properties it to S to have. If truth be told . It did not sound to me like they changed key. or would normally believe. that my conscious experience of the fingers ( before I counted ) was not what I took it to be: it was a seven-finger experience . If I don t know who your sister is or what she looks like .of which I am not appearsp aware . of course. We are listening to a appearsp ' recording of Goible s 93d Symphony .ll So I answer " No " to your question . you turn just .e. to me ask whether it didn ' t sound that way to me..I can. a case in which Slacks the relevant concept .is like supposing that an attractive young woman looks like ' your sister to me.someone ignorant of what a change of key is and what a change of key sounds like .Chapter 5 136 there is a fact about my phenomenal experience . but how it soundsp to me. but uncertain about it . To suppose that anything sounds .. Consider. learn ) SEVEN. S is not aware that ~ ~ this is so. Hence. Thinking there has been a change of key.that there to me to be seven fingers . about .what I believe . how can anyone look like your sister to me? Clearly. about what I hear on the basis of hearing it . and I do not know what a change of key sounds like .to I am musically ignorant .not (as I took it to be at the time ) an eight -finger experience. next. But if we take your question to be a question not about how the music soundSdto me. I am describing how things soundd to me . nothing ever sounds like a change of key to me. I do not know what a change of key is.having the concept of conscious of (i . when the way things see ~ the way ~ things seemd' the experiencer is necessarily unaware of how things see .

I will have no awareness that I am aware of them. but for all ' ways of describing the way things loo ~ . I me under normal viewing conditions ? You are probably in a better position to tell me whether she looksp like your sister to me than I am. CHANGE OF KEY is not a concept I have available for describing my auditory experience. And this holds . to the fact that I have an experience of this kind . I cannot be made aware of the fact that anything looksp red to me . How am 1 not aware that it or looksp soundsp someone who has never met your sister . Not knowing what a change of key sounds like .supposed to know whether the woman we now see looksp like your sister to me . The fact that a person would not say. that my experience has this quali ~ tive character. things will still loo red and green to me. I will be aware of them in the way that I was aware (see above) of seven fingers (when I mistakenly took there to be eight ) and a change of key (when I do not understand the quality of sound to which this refers). not only for concepts like SISTER and CHANGE OF KEY. . not to the quality (I hear thus I experience . become aware that this is so. whether she is causing in me an experience which is like the experience your sister causes (or would cause) . I am not an expert on what the music soundsp like to me. even if the music soundsp to me exactly the way a change of key sounds to me) 2 I will not . If I don t know what it is to be red. but to the fact that I hear it . does not think . A conceptual deficit " deafens" me.the change of key ). I will still be awareof these colors ~ . All it shows is that the person is ~ this way .Externalism and Supervenience 137 my auditory experience. It will seemdto me as though I am not aware of them. Hence. then the answer is not so clear. perhaps would F does not show that k or even deny. that k soundsp looksl does not loo or soun~ F. Without the concepts.

We are sometimes aware of a difference between the way k and h loo different . The key question is whether 5 groups or classifies liquids according to taste.5 Chapter 138 This is not simply a matter of knowing . the right labels or words for experienced differences. consistently distinguishing Pepsi from Coke. then whatever quality or qualities it is that leads him to do so are the qualities that Coke and Pepsi seemd to 5 to have. He will have no thought of the form . know that this is the Pepsi taste and that the Coke taste. a matter of lacking certain discriminatory powers . but as this (call it a P-tasting) liquid and that (call it a C-tasting) liquid ." This is the situation I imagined myself to be in when listening to the music. or not knowing . If so. they tasted different . but he doesn t know which is which . but not know that this is the Pepsi and that the Coke taste. therefore. 5 can taste the difference between Coke and ' Pepsi. If not . we give 5 credit for rudimentary knowledge of taste types and . Nonetheless . if 5 responds (or can respond . He can tell the difference. He may not have heard of these soft dr ~ may not . instead. " Aha . Coke and Pepsi not only tastep different to 5. he won ' t tasted it . Even if I heard it . then even if 5 tastesp the difference. of course. this is the same as that. as a result of hearing sounds with this auditory qual - . and these two are different . and we realize two things appearp ( ~ this ) without knowing how to describe the of appearancep either k or h. he may not choose to respond) to one taste in a way different from the way he responds to the other. Having the words or concepts for Coke and Pepsi (or even for P-taste and C-taste) is irrelevant . I did not (not even mentally ) group or classify changes of key together. I do not . such discriminatory capacities with respect to taste types may be sufficient for crediting 5 with rudimentary knowledge of what the liquids like .not as Coke and tastep Pepsi. It is. for present purposes .

think . and not simply because I lacked words for " " describing a difference I heard. That is why it may have soundedp like a change of key to me without soundingd like a change of key to me. and that denials that it could be are based on a faulty picture of the relation between thought and experience.experiential qualities . until one acquires the conceptual . I do not need the concept RED to see red. say is a much more indirect process a process that requires the possession and . inaccessible . on display in the shop window of the mind . That was not my purpose . as it were . or judge them to be the same in some respect. believe . but that it could be. The strategy is not to show that experience is externally grounded . That. is what made me deaf to a change of key. a processin which one becomesaware of the qualities of experience in the way one becomes aware of the qualities of external objects that the experience is an experience of. But I do need this concept to become aware of the qualered. Awareness of phenomenal properties (that one is experiencing redness. to become aware that I am having an experience of this sort. Awareness of phenomenal properties . I am made aware of the color red by object merely opening my eyes.Externalism and Supervenience 139 ity .it was not intended to showthat they have different experiences of the puddle . qualia (understood as qualities that distinguish experiencesfrom one another) remain " hidden " . When illumination is normal and there is a red looksp) in front of me. use of the concepts needed to think that something is (or red. This being so. I have tried to do this by showing that qualia . that the puddle looksp different to them.that something looksp red. Where does this leave us with respect to Fred and Twin Fred? Nolie of this shows .are not . or a change of key) is not achieved by a processof direct inward inspection. the taste of strawberries.

an externalist about belief) is willing to accept this result. even if they are having the same phenomenal experience of k. If Fred and Twin Fred. though differing conceptu- . if they are conceptually different . If a conceptual externalist (that is. both of the Q ish sort). the quality of this experience that makes it the sameexperience must remain inaccessibleto them. despite this difference in the way their experiences seemdto them. If . Sheno longer has a reason to resist phenomenal externalism (= externalism about experience ) . If (as we are assuming) Fred and Twin Fred have different concepts for sorting and identifying the objects they perceive . let us say. cannot be made aware of. and this is something one cannot be made aware of without understanding what it means to be red. I can become aware of the color red without the concept RED Gust as I can be made aware of seven fingers without the concept of SEVEN). What Fred' s experience seernsd like to Fred will not be what Twin Fred' s experience seernsd like to Twin Fred. I think . these experiences are. capitulated . For if one accepts the conclusion of this line of reasoning . From a subjective standpoint . For to become aware of the quale red is to become aware that one is having an experience of a reddish sort . it will be as if their experience (of the puddle ) was not Q .Chapter 5 140 resourcesfor becoming aware of them. They cannot be made aware of it . then she has. They will be as unaware of the phenomenal quality Q as I was (see above) to a phenomenal change of key. They will be introspectively blind to that aspect of their experience that makes it the same experience. but I cannot be made aware of the qualered without this concept. then. nonetheless ~ the same (they are. then the qualities of experience which allegedly must be the same in physically identical persons are qualities that the persons in question. then the quale Q will be a quality that neither Fred nor Twin Fred can be made aware of.

suppressed inclinations to believe (Pitcher 1971 or micro -judgments ) (Dennett and Kinsbourne 1992. If experiencesare distinguished from thoughts . . After the nuts -:and bolts settle .such things as potential beliefs (Arm strong 1969).Twin Tercel an exact duplicate of . without anything lookingd F to you ). Lo! . what reason is there to insist that qualia must be the samein physically identical beings? Indeed. . why suppose it must be the same? The conclusion can be put this way . on the other hand. are having the same experience of the puddle . so that k can loo~ F to you without your believing . If . Either phenomenal experiences are identified with thoughtlike entities. .e. or (as in the present study ) phenomenal appearances(= the way things appearp) are distinguished from thoughts (from the way things appeard). then it turns out that one may be ' completely unaware of one s qualia (in the way I was completely unaware of the fact that the music like it is soundsp ..Externalism and Supervenience 141 ally .in which case the conclusion ) that sensationshave their properties externally grounded if thoughts do follows trivially . that anything is F (i . then phenomenal experiencesmust be externally grounded if beliefs are. or being disposed to believe . If it is not something they can be made aware of. the respectsin which their experiencesare the same (call it Q) is not a qual ~ of their experience that they can be made ty aware of. If one takes qualitative states as essentially changing key) knowable . as many philosophers do . one is willing to tolerate unknowable qualia . . what better explanation is there for why the experience of Fred and Twin Fred seem so different to them than that it is? 2 Replacement Arguments and Absent Qualia Lighming strikes an automobile junkyard .

a duplicate and not my own car miraculously "transported to the junkyard by the electrical disturbance ." Who (a skeptic asks) is to say that the gauge in Twin Tercel is supposed to register the amount of gasoline? hideed . indeed . Twin Tercel' s does not . More careful inspection reveals a minor difference . who is to say it is a " " " " " " gas gauge ? Do E and F really mean Empty and " Full " ? Even if they do. We ' know what it was designed to do and that it wouldn t be " " working if it failed to register amount of gasoline . tom upholstery. . We know the corresponding part in Tercel is a gas gauge. stamped into the engine block . stone chips in the windshield . a dented bumper . Even the identification numbers . Nobody put the symbols " E" and " F there. are the same. then . Why . is enough to convince me it is. But nobody designed Twin " Tercel. is unresponsive to the amount of gasoline in the tank.who knows?. Tercel' s gas gauge works . As far as the Klingons are concerned. What they don ' t agree about is whether to describe this as " not working . At least some people say it doesn' t work . We know what its symbols mean.Twin Tercel also resembles a 1932 Klingon Space ' Shuttle. Lightning can do such wonderful things . Everyone agrees that the pointer on the gauge in Twin Tercel . suppose they stand for anything ? Maybe . What we see as functioning parts of Twin Tercel s exhaust system. Twin Tercel has all the same identifying features: a small rusty scratch on the rear fender. Only a quick check.Chapter 5 142 my 1981 Toyota Tercel. why suppose they refer to amount of gasoline? Maybe the gauge is wired wrong : they stand for the amount of brake fluid in the master cylinder or the amount of solvent in the windshield washer reservoir. back home. unlike the one in Tercel . Others are not so sure. the Klingons seeas malfunctioning parts of its forward thrusters.

until we do so. by failing to do. amount of gas. of course. If gauges and instruments are objects that have a function . their " not working right . It would not be working right . and therefore functionally . It can' t be. If the gauge in my Tercel behaved that way. nothing that ." We could . give the parts of T' \ inTercela function . If I had made those marks . they are not a copy. there are no gauges and instruments in Twin Tercel nothing that can. Nonetheless. nothing can work wrong . therefore. nothing can work right in Twin Tercel and . It would be misrepresenting the amount of gas in the tank. There is nothing in Twin Tercel that can (as the corresponding parts in my Tercel can) misrepresent anything . an imitation . or areproduction of anything that was designed to do something. You cannot tell them apart . indistinguishable from Tercel. something they are supposed to do. and oil pressure. it isn ' t. therefore. after all .Externalism and Supervenience 143 the only thing that is working right in Twin Tercel is what we call the " fuel gauge. None of this is true of the part in Twin Tercel. given the origin of the marks ." Not only were the parts of Twin Tercelnot designed to do anything .the same function the corresponding parts of Tercel have . that Twin Tercel even hasa gauge. is physically . is like supposing that a word has been misspelled when the wind and waves carve out " anser" in the sand of a deserted beach. it would be broken. To suppose that Twin Tercel' s " " gas gauge is not working right . Twin Tercel. it might have been a misspelling .but . as the instruments in my Tercel can. represent speed. despite indistinguishability . If we forget about the gas gauge for the moment (the only point of difference between Tercel and Twin Tercel) Twin Tercel would coast through a Toyota Turing Test. There is. nothing the parts of Twin Tercel are supposed to be doing . but . would count as . all is dark in the .

it will be said. but Twin " " Tercel will not do so becauseit thinks the temperature is too high . Tercel that has the function of registering coolant temperature . Tercel will fan its radiator even when the temperature is normal . It does this whether or not it represents is getting hot. seat belts and to initiate appropriate responses Twin Tercel behaves in the same way . If Twin Fred materialized in a swamp the way Twin Tercel materialized in that junkyard . As a result of this arrangement . There are circumstancesin which Twin Tercel will also fan its radiator when the temperature is normal . . merely shows that experience is not representational. nothing that is supposed to detect unbuckled seat belts.e. Tercel has a thermostatically control led fan that cools the radiator when sensors tell it that the coolant temperature is getting too high . There is. " This is why Tercel can be " fooled and Twin Tercel cannot. It will do so when it " thinks " the temperature is too high . There is the nothing in Twin . Tercel also emits a loud " " screechingsound when it thinks the seat belts are not fastened " : Tercel comes equipped with " sensors whose function it is to detect high coolant temperature and unbuckled . as a result. This. Twin Tercel is a 13 Toyota zombie. fans itself when it thinks it is (i. If Fred fanned himself because he felt hot Twin Fred would be fanning himself for the same reason - . Tercel so to speak. For the same reason. There is nothing in Twin Tercel (as there is in ' Tercel) that means the temperature is too hot when it isn t. nothing in Twin Tercel that represents the conditions that cause it to behave the way it does.but it doesn t do these things for " reasons" Tercel does them . Twin Fred would be having exactly the same thoughts and experiences as its radiator when coolant temperature rises and emits a loud screeching sound when ' seatbelts are not fastened. itself as) getting hot.Chapter 5 144 representational mind of Twin Tercel ..

of the experiencer. But it doesimply an equivalence of experience. nothing to provide it with experiences in the way that Twin Tercel had no gauges to produce representations this does..and who knows ?~ maybe everything to do with the physical constitution of the experiencer Keep the physical constitution the . same ~ d the experienceswill also be the same.e. to have the experience supervenes on the eonstitution .. What it was designed to do. Tercel and Twin Tercel illustrate that . of representation makes it the casethat Swamp Fred theory (spontaneously materialized in a swamp the way Twin Tercel materialized in that junkyard ) has no sense organs. it is like to have an experience. The reaction is so widely shared that it deservesa name. All it shows is that a naturalistic theory of representation is not a satisfactory theory of experience. So says the internalist Intuition . The internalist Intuition gives expression to the conviction that experience (i. they must have the same experiences If a naturalistic . . the quality of experience what it is like . It was not intended to do this . I will call it the Internalist Intuition . So experience is not representational. Physical indistinguishability may not imply representational equivalence if representation is understood the way it is here being understood (i. It does not show that this judgment is false or even implausible . as historically grounded ). n ' t show (as it did with Twin Tercel) that everything is " dark " in the mind of Swamp Fred. instead. is to reveal how unreli able ordinary intuitions are about miraculous materializa - . The Twin Tercel parable does not directly challenge the Intemalist Intuition . This . The quality of experience what . has something . If Fred and Twin Fred are physically the same.e. will be a widely shared reaction to my story about Twin Tercel. I assume.Externalism and Supervenience 145 because he felt too hot .and ) for materialists this can only mean physical constitution .

therefore. This is perfectly reasonable .5 Chapter 146 tions and instantaneous replacements. influenced by a striking resemblance inappearance and placement14 of parts . We are. in normal circumstances . for instance. lifeless gas gauge) is not working right .we blithely ignore the fact that the resemblancein both appearanceand placement is ( by hypothesis) completely fortuitous and. If I find Twin Tercel in the middle of the desert. that it is. as a consequence this (pointing at its . a paperweight . we see to be quite irrelevant . It is . We seem driven by what I call the Paley Syndromel6 an irresistible tendency to use resemblance and placement as a basis for attributing purpose and design. To persist in giving them a purpose anyway ~ omething one does (if .automobiles . I will infer. There is . that none of them were made or placed there for any purpose .and I expect that reasonablepeople will join me in inferring . Yet. on deeper reflection . a perfectly respectable form of analogical reasoning . Yet. Our judgments about to what it makes sense say. and happens to be placed on papers (thus holding them down ). nothing irrational about the Paley Syndrome .one is told in advance that the parts were not designed.what it would be true to sayabout such bizarre casesare influenced by factors that . irrelevant to determining the function of parts .that it was designed (probably by the same people that designed my 1981Tercel) and that .as in the Twin Tercel story . thus.Is We ignore this and proceed to assign functions on the basis of resemblanceand placement anyway. of course . no one thinks that becausemy doorstop looks like your paperweight . for example . Just becausea bushing looks ' Ilke your wedding ring doesn t mean it becomes your wed ding ring if you slip it on your finger to keep track of it . when asked to render judgments about more complex objects . What is not reasonable is to make the same inference when .

these prereflective judgments . merely a plea for methodological irrational . Intuitions are generated by. Once again. Even if we had clear and distinct intuitions about Twin Tercel.into two Tercels both of which have the history . The Paley Syndrome is such a habitual part of our inferential practice that we use it to generate " " " intuitions about cases even when we know it doesn ' t apply . not a Toyota-free junkyard .amoeba fashion . exactly the patterns of thought that are not applicable in the bizarre circumStancescharacteristic of philosophical thought experiments . these intuitions . and. is in the midst of a philosophical thought experiment . That is where they are leastlikely to be reliable. they bubble out of. customary patterns of thought . begin to change. what do we say if the lightning strikes. gauges and instruments ) of the original . to say that Twin Tercel had a broken gas gauge. At least they become less " intuitive . The time to suspendintuitions . We ( can imagine all kinds of intermediate cases Even if we have . the time to not trust snap judgments . this is not an argument . There are other reasons to be cautious . but my Tercel when it is parked in a junkyard ? Which of the two resulting Toyotas is my Tercel and which the duplicate ? Or is neither a duplicate ? Maybe the original split . that it not only made perfectly ." This is typical of philo sophical thought experiments in which one is asked to make judgments about situations that differ profoundly from the familiar regularities of daily life . We are dealing with casesthat combine the most complex problems in epistemology and metaphysics.Externalism and Supervenience 147 only implicitly ) by calling them gauges and instruments . I assume that most readers will have supposed. therefore. After thinking about what it takes to be a gauge and about whether Twin Tercel' s parts have what it takes. . good sense but that it was true . on first hearing the Twin Tercel story.

5 Chapter


clear and distinct ideas about what it takes to be a gas , gauge, do we , in each of these cases have a clear idea about which of the resulting Toyotas has a gas gauge? It would be wise, then, to be cautious in our judgments about what must be the case when Swampman (a Donald Davidson doppelganger ) materializes in some swamp . We have a right to ask why the Intemalist Intuition should be reg~ ded as more reliable than our initial judgments about . Twin Tercel . Why must Swampman have all the same thoughts and experiences as Donald Davidson ? Indeed , ? why must Swampman have any thoughts and sensations Is this judgement based on some deep insight about the nature of thought and experience or is it merely a brute intuition , a metaphysical axiom , the sort of fact that is available as a premise, but never as a conclusion, in reasoning about the mind ? Brute intuitions are all well and good . We have to start somewhere. Not all premises can be conclusions. But such premises are suspect when applied to fantastic situations that violate normal expectations and background presuppositions . It would be silly to wonder whether a stranger found wandering around in my yard , a person that looked and acted exactly like my wife in the kitchen , was completely devoid of feelings and experiences(i .e., a zombie). That is the stuff of Hollywood science fiction , not everyday life . I fail to understand , though , why it would be any more silly than wondering whether an automobile parked in my driveway , one that looks and acts exactly like the one in my . garage, really had a carburetor, brakes, and a speedometer Both possibilities sound silly . Yet, in the second case the , silliness comes, not from the impossibility of a physically indistinguishable car lacking these parts , but from the improbability of the events (spontaneous materialization ) that would have this result . What reason is there to think that the silliness of the first isn' t the same?

Externalism and Supervenience


Aside , then, from its intuitive appeal- an appeal that we should mistrust - is there any reason to think the Internalist Intuition valid in the extraordinary circumstances described in Swampman (and similar " replacement" ) thought experiments ? I have already discussed and , I hope, disposed of two reasons for thinking that the Internalist Intuition is true . There is, ~ t , the argument from self knowledge : the argument that if the quality of one' s experience was determined by extrinsic factors (thus making it possible for doubles to have different experiences ), it would not be possible to know , by introspection , by looking inward , what it was like to have those experiences The answer I gave to this objection ? (in chapter 2) was that although one has privileged information about the character of one' s own experiences , one does not look inward to get it . Knowledge about one' s own experiencesis obtained not by experiencing one' s experience , but by simply having the experience one seeks knowledge of. The experience itself - normally an experience of external objects- carries all the information one needs to know what the experience is like . A second possible source for the intuition that experience cannot be extrinsically constituted was discussed in the previous section. H one conceives of experience in act- object terms - as the mind ' s awareness of internal objects or qualities - then it becomes impossible for a materialist to think of Fred and Twin Fred as having different experiences H different experiences . are awarenesses of different internal things , creatures in whom different experiencesoccur must , on this conception of experience differ . Fred and Twin Fred cannotbe , . different experiences The way to shed this prejudice having externalism is quite easy: shed (as almost everyone against has) this quaint conception of senseexperience.

Chapter 5


In the next section I will turn to a third possible source of the intuition that Fred and Twin Fred must be having the . same thoughts and experiences This is the idea that if the twins were having different thoughts or experiences then, , since much of what minded creatures do is caused by what , they think and experience one would expect this difference to show up in different behaviors. At least it should be possible for it to show up . But , since Fred and Twin Fred are physically indistinguishable , Fred and Twin Fred cannot behave differently . So any alleged difference in what they think or experience would have to be completely epiphenomenal : it could never reveal itself in anything they said or did . If nothing the twins can either say or do can evince the alleged mental difference, though , what is the point , what is the sense in saying that such differences exist? The logical , positivists liked to say that a difference that did not make a difference was not a difference at all. That may have been too strong, but , surely, a difference that cannotmake adifference , not even to the person in whom it occurs, is not a real difference. Or, if it is, it is not one a theory of the mind need worry about. Aside from these possible grounds for the Internalist Intuition , I can think of no other support for the Internalist Intuition that is not a variation on one of these three themes. I conclude, therefore, that the Internalist Intuition is a brute intuition , one that is not justified by any defensible claim about the nature of thought or experience. This does not show that externalism is true and that , as a consequence , can have different experiences physically indistin;guishable beings . It does not even show that internalism - especially about experience - is implausible . Not at all . Most philosophers take it as plausible enough not to need argument . , Arguments have to stop someplace and this seemsas good

Externalism and Supervenience


a resting point as any. I confess to myself feeling the pull of the Internalist Intuition . Indeed, I would not have thought to question it but for the fact that unless it is challenged, an even more (for me) obvious fact must be rejected- the idea, namely, that what goes on in the mind what we think , feel and experience - and , therefore , the qualities in terms of which we distinguish thoughts , feelings , and experiences from one mother , are nowhere to be found in the headwhere the thoughts , feelings, and experiencesare. The prob lemsabout conscious experience are so baffling that one can reasonably expect the right answers- if right answers are ever to be found - will require abandoning somedeeply held convictions about the mind - body relation . One will have to make a hard choice about what to give up . My choice is the Internalist Intuition .
3 Explanatory Relevance

One reason, not yet discussed, for supposing that externalism is objectionable is that it allegedly makes the mind . Causality, it is said (e.g., McGinn 1989 Fodor ; epiphenomenal 1991 is a local affair. The causal efficacy of an event resides ) in its intrinsic properties (and, of course, the circumstances in which this event occurs). Put the same kind of eventone with the same in Uinsic properties - in the same conditions and it will have the same effects. That is why physical duplicates Fred and Twin Fred, for example - behave in the same way . The internal events that produce their behavior are in Uinsically the same. Hence, they produce exactly the same bodily movements . The internal events and ' es process that cause Fred s arm to execute movement M are ' exactly the same as those that causeTwin Fred s arm to execute movement M . Since these causal processes are the

Chapter 5


same, whatever extrinsic (e.g., historical ) differences exist between the events that cause their arms to move must be causally irrelevant to why their arms move the way they do. ff the mental is extrinsic, if it does not supervene on the present physical constitution of Fred and Twin Fred, as it clearly doesn' t on a representational theory of experience and thought , then it will be irrelevant to why Fred and Twin Fred behave the way they do. The Representational Thesis makeS what you think and feel irrelevant - causally irrelevant - to what you do. Something like this reasoning is probably somewhere in the background when people invoke the Internalist Intuition . Clyde winces because it hurts . He goes to the fridge becausehe wants a beer. He grabs this bottle , not that one, becausehe thinks this is the beer and that the catsup. The only way to understand how the pain , the thought , and the desire can explain these behaviors is if the pain , the thought , and the desire are insidehelping to bring about the associated bodily movements. By locating the mental outside the body , by making what we think , desire, and feel depend on our history , externalism makes such ordinary explanations of behavior sound like magic. There are several possible confusions in this reasoning, the most important of which is the conflation of behavior with bodily movement. What thoughts , desires, and feelings explain is not why your arm moves (when you move it intentionally ), but why you move your arm. Once the difference between behavior and bodily movement is clear, it also becomes clear that externalist theories of the mind are not threatened by epiphenomenalism . Mental content can explain behavior without supervening on the neurophysiological events and processes that cause bodily movement . The mental is not robbed of its explanatory relevance by being extrinsic.

Externalism and Supervenience


Before enlarging on this point , though , a clarification . Historical events, things that have happened to a person, things that are, therefore, extrinsic to the current state of the person, can certainly make a causal difference to that person' s behavior. Jimmy stutters becausehe was dropped on his head when he was an infant . We can think of the fall as a remote cause It produced in Jimmy a pennanent neurological . condition , N , that now operates as a proximal cause of ' Jimmy s stutter . We can imagine a twin , someone in the same neurological condition , but who was not dropped on his head. A brick fell on Johnny ' s head. It produced the same neurological condition that Jimmy suffers from . Same proximal causes, different remote causes. The fact that ' ' Jimmy s fall has its effect on Jimmy s behavior through a proximal condition (neurological damage) that can be produced by other causes does not show that the fall is not ' among the causes of Jimmy s current behavior. It does not show that knowledge of the remote event isn ' t relevant to understanding why Jimmy stutters. All it shows is that different causes can have the same effect , that different remote) explanations can explain behaviors that have the ( same proximal cause. There is, then, no metaphysical problem in having historically remote conditions acting as causes of current behavior - no problem in conceiving of extrinsic facts - conditions that do not supervene on the current state of the systemexplaining the current behavior of the system. Remote causes are causeseven though their effects are brought about via more proximal events. There is, then, no particular problem about conceiving of different explanations for Fred ' s and Twin Fred' s behavior. If Jimmy ' s fall explains why he stutters even though a similar fall does not explain why someone with the same neurological condition stutters , there

now behavior. By making the causal/ explanatory relevance of the mind (temporally ) remote.e. proximal ) explanatory relevance -and then (i . The objection to the RepresentationalThesis.e. ' what (despite Horgan s unshakable conviction ) is not so clear. (1991 pp . and experiences This is not in dispute } 7What is in dispute . and experiences are to explain here-and. the mental states that explain a person s current behavior. Horgan is certainly right about one thing : beliefs. then. This is objectionable because as . desires. desires. and " for feelings . That cannot be the objection since there is nothing objectionable in that .e. desires ' and experiences. .Chapter 5 154 should be no special problem in accepting the idea that the ' explanation for Fred s behavior might be much different ' from the explanation for Twin Fred s behavior even though Fred and Twin Fred ( being physically indistinguishable ) have the same proximal causesof their behavior..the reasons which we act have an immediate " causal / explanatory relevance to our intended behavior. then it must exist in the head of the trip -taker at the time of the trip . the facts that underlie the content of belief and quality of . must exist at the time of the actions they help explain . If beliefs. If wanting a glass of beer is to explain a trip to the fridge . Feeling a pain is not going to explain why you wince unless you feel the pain at the time you wince.. 88. of experience and thought into a there . remote) causal relevance. desires . is whether the facts that give mental states their identity . the RepresentationalThesis denies this basic intuition . The point is rather that the Representational Thesis somehow makes the here and-now (i . they have to be here-and-now beliefs. one has (at least Horgan Horgan " " has) an unshakable conviction that beliefs . historical ) facts relevant to ' the explanation of a person s behavior.. should not be that it makes extrinsic (i .89) puts it .

Do these facts have to exist at the time and place of the behavior they explain ? I do not think so. but they do things . It Will be enough. to give a remote cause of the plant ' s Changeof color. and . Hawkmoths prefer white blossoms . this is why it does it . I hope. The flower changes color " in order to " exploit this seasonalalteration in its circumstances The plant sets more . and hummingbirds are more attracted to red blossoms. Why does the Scarlet Gilia do this . however. I have developed this point at great length elsewhere ! 8 What I give here is a brief sketch and a few examples . This is plant behavior. events in the evolutionary history of the plant that happened long ago and (probably ) far away . in the words of Page and Whitham . changes color during the summer . fruit by changing color . and the reason I do not is that the events that reasons( here -and-now beliefs and desires) cause are not the behavior that reasons (there-and-then content ) explains . the Scarlet Gilia . This explanation of the plant ' s behavior appeals to thereand -then facts. and botanists are interested in explaining why plants do the things they do. Early in the season hummingbirds are the chief pollinators . Plants don ' t have thoughts and desires. is that ). the one given by Page and Whitham (1985 the botanists from whom I take this example. something the plant does. whether thesefacts have to be here-and -now facts. This is . response A plant . to indicate the general line of .Externalism and Supervenience 155 experience. Later in the seasonthe hummingbirds migrate and hawkmoths become the principal pollinators . ' Explaining the plant s behavior in this way is not .change from red to white in the middle of July each year? One explanation . the plant does this in order to attract pollinators . sometimes very interesting things . experiences and feelings.

Natural selection does not causethe chemical changesin the plant that . . changescause The seasonalchangescausechemical activity in the plant that produces a pigment change. What ( here are seasonal changes causes the plant to change color " " (longer sunlight .>Chemical >Color Change] Activity The brackets in (B) are meant to indicate that what natural selection causes is not the warm days that cause color change. as it were. Giving the adaptive purpose . in -and-now ) turn .>Chemical . " for which the plant changes color is not describing a more distant member of a causal chain that has chemical activity in the plant as proximal cause and pigment change as its effect.>Color change Activity but . What natural selection brings about is a change in the plant s chemical constitution .Chapter 5 156 ' not at all like mentioning Jimmy s fall in explaining his stutter . This is not what the seasonal . the ringing of an internal biological alarm clock. N . or whatever ). (B) Natural selection. causeswarm days to causecolor change. Jimmy ' s fall produced a neurological condition . but why one thing that is happening here and now (seasonal changes causes another here) and-now event (change of color ). Natural selection is. warmer seasonalevents. that " causes irl turn . so to speak. rather. The causal pattern is not (A ) Natural Selection >Warm Days .> [ WarmDays---. . What the evolutionary process explains is plant behavior why it changes color. Jimmy ' s stutter. the causeof one thing (warm days) causing another (color change)'. and natural selection is not the remote causeof these here-and. rather. What natural selection explains is not . It . bring about a pigment change. what is happening here-and now.

why it changes color . The twin evolved in quite a different environment . . You could study these two plants under a microscope for years. Consider. Twin Plant changescolor in order to repel beetles. one that is a molecular twin of the Scarlet Gilia . As a result of this selectional pressure. the explanation of why it behaves this way is quite different . and never realize that there were quite different " reasons for their " behaving the way they do. a slow change occurred : the plant evolved into a form in which it changed color. Although the change in color is a purely physical event.I emphasize not. in the midst of every flowering season The . of descriptions of those . from red to white . That is much different from causing the change of color. This " retreat" to there-andthen facts to explain the plant ' s here-and-now behavior is not . Scarlet Gilia changes color to attract Hawkmoths . arriv ~ that were attracted by red blossoms. beetleshate white blossoms.because we (or botanists ) hanker .does not consist of descriptions of these internal causes It consists instead. in the midst of its flowering season rapacious beetles . Although Twin Plant is physically identical to Scarlet Gilia . the explanation of the plant ' s behavior . why it changescolor in the summer. They avoided Twin Plant and it flourished . a physically indistinguishable plant . to something that happened in the history of the plant (or the history of this kind of plant ). and although it therefore behaves in exactly the same way.Extpma~ and Supervenience 1i m 157 a condition that explains why the plant behaves the way it does. now. events and process that made these internal events cause es that external change. an environment in which . know everything there was to know about their present physical constitution . This requires going outside the plant . and although it is produced by well -understood chemical changes inside the plant .

and though . does not have the same reasons ' for behaving this way . For both the change in Scarlet Gilia and Twin Plant are produced by exactly the same chemical . This is not to say there . when we are looking for why a system. you do not necessarily know why the pl ~ t changescolor. the adaptive explanation does not provide remote causesof current events. including a plant . I do something . What it provides is an explanation of current behavior. We are. behaving the same way as these other two plants . Swamp Plant materialized (conveniently potted ) when lightning struck a plant nursery. Swamp Plant. indistinguishable from the first two . Swamp Plant is the botanical analog of an animal doing something mindlessly . It doesn t have any reasons for behaving this way . For. We can. instead. What we want to know . Swamp Plant. I do something just as Swamp Plant does something ' when it changes color in July. as a result. In the middle of July Gust like Scarlet Gilia and Twin Plant) Swamp Plant changescolor. we already know that ). behaves the way it does. Why ? Why is Swamp Plant turning white ? Is it doing so in order to repel beetles? To attract hawkmoths ? No .Chapter 5 158 after remote causes of current events. once why it is. imagine a third plant . Even if you know what physical events inside the plant cause the change in color. why the current events are causing changesin the ' plant s color. that has no reason for changing color . Often. changes Yet. if we like . trying to find out why the surface change is being caused. they did not change color for the same reason. We know that A is causing B. we are looking. When I snore or hiccough. But Swamp Plant and I don t have reasons for behaving this way . not for the internal cause of surface change (in the case of the Scarlet Gilia . though physically identical to the other two plants .

and intentions . Nonetheless. do not have reasonsin the sense in which we have reasons. In Swamp Plant s case the behavior has no adaptive purpose . . In explaining a system' s behavior. The plants. If its change of color repels a stray beetle. I have taken liberties in describing these plants . Yet. that made C causeE. no psychological . there would be no purpose in Swamp Plant behaving this way . no teleological explanation . For dramatic effect I have spoken of these plants as having reasons for behaving the way they did .Externalism and Supervenience 159 are not causal explanations for our behavior.Scarlet Gilia or Twin Plant behaved. we are often looking . not inside for the physical cause (C) of external change (E). does not test for the right thing . That is not why it does it . . Nonetheless . a coincidence. but of C' s causing E. ' . explanation . Swamp Plant could pass a botanical Turing Test. It does not tell us what we want to know : whether the behavior (in the caseof plants ) is for some purpose or (in the caseof animals) is explained by beliefs and desires. not triggering. no purposive . in my case there is no rational . that is dumb luck . causesof behavior. The plants do not have thoughts . of course. I used this way of speaking to make a point about the explanation of behavior. just as the original Turing Test. structuring. Swamp Plant has no reason to change color. or attracts some wandering hawkmoths .but outside for the events and circumstances that shaped internal structures. We are looking for what I elsewhere (Dretske 1988 called ) .the cause not of E. They are not minded . Swamp Plant is a botanical zombie. The botanical analogue of the Turing Test. I described their behavior as purposive . though not purposive . It would behave exactly the " " way an intelligent plant . unlike these other two plants. It is only to say that.that is the business of chemistry (in the case of plants ) and neurobiology (in the case of animals ). desires.

experience causessalivary glands to secrete It doesn t do so in other dogs. but for a structuring cause. So why is this dog salivating ? Why is it behaving this way ? This . Why isn ' t it . an explanation for why the auditory ' . say. ' in saliva' s being secretedinto the dog s mouth . but in . The bell rings (R). Perhaps. causesthe dog to salivate? Well . cause saliva to be secretedin the dog ' s mouth (M ). by causing the dog to have dog a certain auditory experience triggers a process that results . These sensory events . What we want to know is why the dog is salivating . If salivation is thought of as something the dog does (not simply as a glandular event occurring to the dog or in the dog ) if . . not in the stimulus that elicits the behavior. though not voluntary . and this produces a certain auditory experience (E) in the dog . Some (not trained at all ) don ' t do much of anything . and we can seek an explanation for it . What . is behavior. jumping ? Other ( differently trained ) dogs jump when they hear the bell . Why ? And once again. tell us why the dog is doing what it is doing . At least the bell . in what learning theorists describe as the contingencies (correlations between the ringing bell and the arrival of food ) to which the dog was exposed during training . It is a request for the cause of one thing causing another. The ringing ' bell is the triggering cause of behavior. The dog hears the bell ring . not for a triggering cause of the dog s behavior. as a result of conditioning .only why it is doing it now. it seemsthe answer to this question lies in the past .Chapter 5 160 A bell rings and a classically conditioned dog behaves the way it was conditioned to behave: it salivates . The behavior. in other words . it is . given the conditioning processthe dog cannot help salivating when it hears the bell .then the causal explanation for it thought of as behavior resides. Yes but that doesn t . to salivate. then. in one sense the ringing bell causes the . clearly. is a ' request.

And . But if what the mental is asked to explain ' is not why saliva is being secreted in the dog s mouth . when the dog hears a bell . then the mental need not supervene in order to be causally relevant. therefore . on what the mental is supposed to causally explain .how.. and experiences are supposed to explain bodily movements and ' . one would expect the content of thought and the quality of experience to supervene not on the physical constitution of the animal but "s on the historical events and proces5e that shaped that ani- . then it is. On the contrary. If thoughts . it salivates. on what one takes to be the behavior of the system. but with the es causal process that result in bodily movement and change. if this is what the mental is supposed to explain . If behavior is identi fied . but why . for example. The internal cause of glandular activity . it is easy to imagine Fred and Twin Fred moving their arms with much . then it is easy to imagirie Dog and Twin Dog salivating for much different reasons. not with bodily movements and change. the neurophysiological processes. A question about the causal efficacy of the mental depends critically . hard to seehow there could bedifferent causes for the behavior of physically indistinguishable animals . desires. if this is what behavior is. would be the same in both .Externalism and Supervenience 161 facts about the dog ' s past experience} 9 Dog and Twin Dog might have much different reasons for salivating just as Scarlet Gilia and Twin Plant had much different reasons for changing color. for exactly the same reasons. therefore. different purposeS20 Whether or not the mental has to supervene on the physical constitution of a system in order to be causally relevant to that system' s behavior depends. admittedly . Dog and Twin Dog could have different reasons for salivating . changes events like the secretion of saliva in a dog s mouth .

derived from the perceptual system the states are states of A state is a conscious experience . Rather.Chapter 5 162 mal ' s current control circuits . in each succeeding generation. Since states acqwre their systemic functions through an evolutionary process. of F (thus making its possessorsensuously aware of F) if the state has the natural (systemic) function of providing information about the Fness of objects standing in the appropriate contextual relation (C) to the system. in any generation .already assumed to be natural selection . for appropriately related objects. over many generations. But natural selection does not . for some determinable property F.natural selection is being identified as the source and creator of conscious experience. Statesthus become conscious by acquiring .exactly RepresentationalThesis says it supervenes. the function of indicating . then we can expect to see more giraffes with longer necks in . 4 Evolutionary Origins where the The Representational Thesis identifies conscious states with representational states. and each generation comes with its stray longer -than -average-necks. their determinate value of F. very long necks indeed. In the case of perceptual experiences . . states that have a certain information providing function one form or another. this function is systemic. increases the proportion of longer necks ) . It merely selects among traits that are . It will surely be objected that natural selection does not create anything . natural selection can produce . succeeding generations . it makes more long necks (i . If individual giraffes enjoy a competitive edge by having longer necks. And if the advantage persists over many generations. make necks longer .

the process by means of which a variable resistor becomesa volume control . make everyone conscious. H someoneis conscious. If nobody is conscious. for example . H it is given a few to start with . in some small way. In my trip at volume controls . I' m building an as a volume amplifier and I want a variable resistor to use to the electronic parts store. then nobody enjoys the advantage of being conscious. not how the story goes. Natural RepresentationalThesis assigns to . This is why natural selection cannot create conscious . Instead. selection is not supposed to select for consciousness That is . upon selection . Once I have made my choice. H we want to know where consciousnesscomes from . But the fact that it . even toa ~mall degree. the role that the The objection misidentifies natural selection. Consciousnessis no different . that it was. becomes consciousness . Given . natural selection can increase the number of consciousbeings in succeeding generations. already a volume control . why it cannot be the origin of consciousness ancestorsdid not already give creatures something their have. electrical components that are not yet (though they may become) volume . Selection makes something a volume control . It selects for something else. once I have chosen a controls variable resistor and installed it in my amplifier . If nobody enjoys the advantage of being conscious. it becomes a volume control . Consider . it can.Externalism and Supervenience 163 Natural selection is not the creator of long necks. as it were. it uses long necks as the raw material out of which to make more of them. becomes a volume control . But it cannot make someone conscious. I' m not selecting among volume controls . It cannot beings. consciousness cannot be selected for . by being selected . does not mean that it started life that way. I am selecting among variable resistors. It cannot evolve. we will have to look someplaceelse for it . I do not look control . something that .

Chapter 5 164 the right raw material . representations represent. They need sunlight and water. through this process of selection. What natural selection starts with as raw material are organisms with assorted needs and variable resources for satiSfying these needs. it comes into being . Information about external (and internal ) affairs is necessaryin order to coordinate behavior with the circumstances in which it has need-fulfilling consequences . What natural selection does with this raw material is to develop and harness information -carrying systems to the effector mechanisms capable of using information to satisfy needs by appropriately directed and timed behavior. and smell things . being Nor do we have to start with consciousness to under stand how . then. We don ' t have to start with volume controls in order to understand how. Despite my practice. You don ' t have to be conscious to h ~ve needs. the benefits of information are clear. throughout these lectures. it can create volume controls . Once an indicator system is selected to provide the needed information it has the function of providing it . As a result. they came into . Even plants have needs. They see hear. For creatures capable of behaving in need-satisfying ways. No point in hiding if nothing is chasing. An animal has a need for N if N is what it has to have to survive and reproduce . No point in chasing if there is nothing to catch. Through a process of selection they have become perceptually conscious of what is going on around them} 1 A word of caution. The states these systems produce by way of performing their iri ormational duties then become representations of the conditions they have the (systemic) function of informing about. of comparing (for illustrative purposes ) natural . the organisms in which these states occur are aware of the objects and properties their internal . through a process of natural selection.

confers a function on a system . Through ignorance or carelessness I can select a variable capacitor to be my . but natural selection cannot . It has to do X becausethe way it gets selected is by having its performance of X contribute in some way to the survival and reproductive successof the animals in which it occurs. in the complete absence of any . volume control . and stupid choices. Artificial selection and conscious design are function -conferring activities that do not depend on objects beingableto do what they are designed or selected to do. for example. Natural selection is quite different . its delivery of information could not have been selected for. If a system didn ' t actually deliver information .Externalism and Supervenience 165 and conventional functions and. The reason it is important to keep these processes distinct is that artificial selection can occur. wire it to my amplifier . when it is selectedfor. an item cannot be naturally selected to do X unless it22 actually does X. choose (or design) something to be a volume control (thus giving it that function ) even when it is utterly incapable of controlling volume . It is this contribution to reproductive success that.23It could not . what they have the function of indicating . This is why it is possible to have conventional representations that are incapable of doing their job . representations it is of the utmost importance for understanding this account of the origin of perceptual consciousness that one not confuse artificial selection with natural selection . even under optimal conditions . Unlike artificial selection . and the function it confers is doing what the system did .incapable of indicating . faulty manufacture and installation . . Such are the consequences of bad design. therefore. natural and conventional . effective performance on the part of the item being selected I can. and wonder why it won ' t do the job I gave it . (in this case provided information ) that increased fitness.

5 Chapter


therefore, have acquired the function of delivering information . For a variable resistor to evolve, in some natural way, into a volume control , for natural selection to create a volume control , the resistor (i .e., some ancestral token of this resistor type ) would actually have had to control volume . I can select a component to do a job that it cannot do , but nature cannot. Once a type of device has the function of controlling volume , current tokens of it need not- ever again!control volume to retain their function . But it could not have acquired that function in any natural way unless it (i .e., earlier instantiations of it ) actually did the job that it (i .e., later instantiations of it ) have the function of performing . For an information - delivery system to acquire the natural function of delivering information , for it to produce natural , representations then, the information it delivers must actually do something . It must make a positive causal contribution to fitness. It must be useful to, and actually used by (or have been used by ), the organisms to which it is delivered . If it isn' t , there can be no selection for the system that provides this information . We can build a system that doesn' t do anything with the information it has the (conventional ) function of delivering (measuring instruments are like this ), but nature cannot. If nothing useful is done with the information , nothing in nature can acquire the function of providing it } 4 As an act of blatant nepotism , I can hire my shiftless cousin , Wally , to perform a useless information -gathering task. I can pay him to keep track of the number of flies that land on my desk each day. Wally thereby has an information - collecting job . It is a conventional job , a job I give him . There are no natural jobs of this sort . There is nothing in nature capable of giving anyone, or anything , a job like this. This is why there are no Wally -sensesor Wally - experiences ,


and Supervenience


nothing in nature that has the function of supplying useless information .25 There are none becauseit is only by supplying information the organism needs- or finds useful - that a system acquires the function of supplying it . It is for this reason that gauges and instruments are not good analogies for sensory representations . Gauges and instrument do not do anything with the information they provide . Even if we could imagine automobiles reproducing and transmitting genetic material to their vehicular offspring , it would be hard to seewhy they would evolve into their present form . What would their gauges and instruments be for? Since these gauges perform no useful chores for the automobile , there would be no need for this information . Only if the information these instruments supplied was used in some way - to shift gears or apply the brakes behaviors that contributed in some way to the reproductive " successof the car- could we imagine these " sensors being selectedfor and, thus, developing the function of supplying information to the car. Only then would the car start sensing its o~ speed, oil pressure, and gas level. Since things acquire their functions gradually , over many generations , does this mean that consciousness comes in ? degrees Does it mean that creatures gradually become conscious ? Yes and No . Whether the answer is " yes" or " no " depends on whether one is talking about types or tokens, kinds or individuals . Yes the process by means of which , we---humans - became conscious is a gradual process. It did not - it cannot- happen overnight . I am conscious though I have distant ancestors who were not , ' and the evolutionary -by process responsible for this change is a gradual , a step individual , level, though , the step, process. At the token, the answer is No . Individual human beings - you and l- do not become conscious in some gradual way . There is, of course,

5 Chapter


a sensein which we speak of ourselves as more or less conscious , of slowly regaining consciousness of gradually losing , consciousness and so on, but our status as conscious , , beings, our capacity to be conscious is not a status or capacity we come by gradually . In this (dispositional ) sense individuals , are either conscious or they are not . It' s like being pregnant . But if I am conscious and my very distant ancestorswere not , when did my less-distant ancestors start being conscious ? At the same time a poor man becomes rich as you keep giving him pennies.

J. 1961 Short term memory in vision. Philosophical . 159 182 . 1988A Cognitive . Oxford : Clarendon Press. Theory Consciousness of Yte . . K . and Coriell. Bennett . HumanitiesPress ArmstrongD . University Press Bemecker. A. Malcolm 1984 . Consciousness and Causality : A Arm strongD Debate on the Nature of Mind . Bach. Bach. . . K 1987 Thoughtand Reference Baker. .. Cambrid : CambridYte Baars B. 1988 What conceptof consciousness Marcel and Bisiach 1988pp. Ithaca NY: Cornell University Press . . Oxford : Basil Blackwell .References . -273 pp 258 . Linguistic Behavior. forthcoming . 1993 A bat without qualitjes . 1976. In Davies and Humphreys 1993 .. Akins . New York: . ? In Allport . England : Cambridge . Arm strongD . . " . Technical . S. and N . R.112 . M . 1991. S. Propositional attitudes and self knowledge .. M. . Bell . . 1986 Thought and object: de re representations and relations. Dretske on the explanatory role of belief . . L. Studies 63: 100. . M. 1968 A Materialist Theoryof the Mind. Averbach E. Cambridge . K. . 1980 TheNatureof Mind and OtherEssays . Svstem ]ournal : -328 . In Brand and Hamish 1986 . . A. 40309 . Press University .

and S. : . and . . no. . 4: 670 672 . 1990aInvertedEarth In Tom . 4: 596 . of Brown D. 13 and . Block N. Blakemore C.. . ExplainedIn The . . and Biro. 1981 Readings thePhilosophy Psychology . 1993 Reviewof Dan Dennetts Consciousness . Philosophical .. . Oxford: BasilBlackwell . 1977Beliefdereo . .References 190 ? . 3: 327 348 . 1: 233 . 113 134 . and relateddisordersIn Milner and Rugg1992pp. . Topics . Bisiach E. Block N. BoghossianP 1989 Contentand self knowledge Philosophical . 1993app.. 14 ' . J. in Block N. . 1874Psychologie Empirischen . J. . vom . 1987 Mindwaves Thoughtson . Standpunkt . Block . . Fodor 1972 What psychological states are not . . 1992 Can externalismbe reconciledwith self-knowledge -267 Philosophical .... 113 137 . 1991 Consciousness subjectivityIn Villanueva 1991 pp.. and J. berlin 1990pp. . no. neglect : . and Biro. 17 no. . N . . eds. . . 1993 Consciousness objectivity In Daviesand Humphreys . Cambridge . ed. forthcoming On a confusionabouta function of consciousness . Bisiach 1988pp. BrentanoF. of MA: Harvard. . Bilgrami. Leipzig. . . . and Behavioral BrainSciences . . Behavioral Brain . Sciences . 1990b Consciousness -598 Sciences . Greenfield. . 1988 The haunted brain and consciousness Marcel and . 53 80.. Topics . IdentityandConsciousness Intelligence .. . 81: . 178 196 . 1993Swampman La Mancha Canadian .. Block N. Journal Philosophy of . I.26. 1991 Evidenceagainstepiphenomenalism . Block N. . 20 no.. . Review 159 181 . 1992 Understanding consciousnessclues from unilateral . 1: 5. and accessibility Behavioral Brain Block. Journal Philosophy of . 74: . A. 23 no. In . . Bisiach E. J. N. I. . . Journal Philosophy 338 362 of Burge T. 101 120 .

-310 . 1992 Consciousness conceptsII . In Marceland Bisiach 1988pp. P M. shrews and misrepresentation TheJournalof .. -304 . . Burge T.. Burge T. Church land. . . .. . in Studies Philosophy . . Clark. . 1988Individualismand selfknowledgeJournal Philosophy of Carruthers. 1992a The Contentsof ExperienceEssays Perception . 1993b Mice. C. 273 . A. UniversityPress Christensen. 11 37 . B. 1984 Matterand Consciousness . : . . 136 157 . and Rugg1992pp. . 1993aSensory . 5: . .References 191 .. . 1957 Perceiving Philosophical . . 1994 SelfKnowledgeOxford : OxfordUniversityPress . T. Study Ithaca NY: Cornell . . Crane T. Church . P M. PressA BradfordBook . AristotelianSociety and ~ . Carruthers P. . . : on . 1979 Scientific Realismand"the Plasticity of Mind. . : Cambridge . Research . no. Crane.P Minnesota . 1992bIntroduction In Crane1992pp. 1. 53 Phenomenological Churchland . In Milner . ed. . . pp Proceedings . M. . 41 59. . 85. 90 Philosophy . and PetraStoerig 1992 Reflections blindsight. . . Cambridge MA: MIT . . :A . 1989A Neurocomputational land . Chisholm R. . A. eds Minneapolis University of in . no. Cassam Q . . ed. Press QualitiesOxford: Clarendon Clark. Frenchet aI.. / . Crane T. pp. 1989 Brute experience Journalof Philosophy86. A. no. . . Press . Perspective MIT Press . . Church land. P. 1982Otherbodies Thought ObjectEssays Intentionality Woodfield ed. : . . P S. Oxford: OarendonPress . Burge T.. and : on . 1988 Reductionand the neurobiologicalbasisof consciousness . 1979Individualismand the mental Midwest W: Studies Metaphysics. A. P. 1993 Sense subject and horizon . . Philosophyand . on Cower. Cambridge UniversityPress . . 1993cThe nonconceptual contentof experienceIn Crane1992a .. . CambridgeCambridge UniversityPress . 4: 749 779 .17 . . -269 . . 258 . CambridgeMA: . . 6: 290 .

C. . 4. 42 77 . . . . . 15 . Dretske F. 100 no. . MinneapolisUniversityof Minnesota : in . Dretske F. Freeman Co. 9. 1992Perceptual contentand localsupervenience . 1969Content Consciousness and . . Brain New York W. . and M. Little.. t . 1993Living on the edge Inquiry 36 nos 1 and 2: 135 160 . Savageed. 1984Firstpersonauthority Dialectic 38. DavieSM. 1991b Postscript Reflections Instrumentalismreconsidered . Dennett D.References 192 . C. Philosophical . . a. . W. In Rosentha11991 . . : . London Routledge Kegan : and Dennett D.39. Humphreys 1993b Introduction. C. Centennial Issue 461 484 . Explained : : .. ' s own mind. Damasio A. 1%9. : .. of . . . . and H. . . 1987Knowing one and es . Brown . Kinsbourne 1992 Time and the observer The . . 72: analysisJournal Philosophy 741 765 of . . . and : Perception Cognitionvol. Behavioral Sciences . In Davies and . 1978 The role of the perceptin visual cognition In Minnesota . C. Wade in Studies thePhilosophy Science . Dordrechet . . : . 1987The . . eds. . : and . and G. . and Brain in where and when of consciousness the brain.. 1979Simpleseeing In BodyMind andMethodEssays Honor . Dennett D. 1988Quining qualia In Marceland Bisiach1988pp. 92: Society 21 45. Blackwell . Proceedings DaViesM. 1991aConsciousness . . Davies M. . 60. . . Dennett D. . Humphreys1993app. Seeing KnowingChicagoUniversityof Chicago . 2: 183 . 1991 Individualism and perceptualcontent Mind. Paul. : Press . 1975Functional . Dennett D. Cummins R. . . Humphreys. -247 . . CambridgeMA: MIT Press . Davidson D. R. W. . D. Dennett D. . F. 1988Replyto Burge Journal Philosophy 664 665 . 85: . Tapscott eds. . Oxford: Davies M. . .. . H.. Davidson D. . Proceedings Address of theAmerican Association . .. . of theAristotelian . . 1. 1993a Consciousness . . Gustafsonand B. of .. Davidson D. Damasio1993 Brain and language In Mind and . C. Intentional Stance Dennett D. no. and G. of Virgil Aldrich. . . Press and DretskeF. Holland: Reidel .

VisualCognition Action D. . Dretske. Oxford: . . . CambridgeMA: MIT Press . . . 1988Explaining . and 2 Science . 201 216 : -283 . 1990b Doesmeaning matter? In Information Semantics . Psychologist .. In An Invitation to .. F. eds Cognitive . . . 1981 Knowledge theFlowof InformationCambridge MA: . of . . Behavior Dretske F. and Dretske F. . . . Topics . DretskeF. and . 121 136 Studies71: . . 6: 784 . . 1992 What isn' t wrong with folk psysmologyMetaphilosophy . 406 263 . . Hollerbacheds CambridgeMA: MIT Press and . Oxford: Blackwell Epistemology ' -221 . 1991b How beliefs explain behavior : Reply to Baker. . A. Dretske F. . forthcoming Differences makeno differencePhilosophical . . H. 1994Modesof perceptual . P 1987Neuroethology releasing . In . . 10 - . . Dretske F. . Dretske F. . that Dretske F. of in : prey mechanisms -catching of . . Dretske F. In . Ewert J. 1986aMisrepresentation BeliefRadu Bogdan ed. Dretske F. and J. ..ilosophy the . representation Ph . . rvIITPressA BradfordBook / . He'll eds. 1993aMentaleventsasstructuringcauses behavior InMele. Erdelyi. 1986b Aspects of cognitive representation In Brand and . . 63: . . .References 193 . Studies 113 117 . 1982Varieties Reference . . 102 no. experience and . . M. Philosophical . . Oxford: Clarendon . Kosslyn andJ. Dretske F. 1 and 2: 1. OxfordUniversityPress . Cognitive . Evans G. . 1991aDretskes replies In Mclaughlin 1991180 . Osherson S. Oxford: Oxford University Press . . Dretske F.. . . 1990a Seeing believing and knowing . . Press . . EnriqueVillanueva ed. Volume . RobertoCasati Barry Smith and GrahamWhite. : Vienna Holder-PichlerTempsky . Dretske F. . Sciences . toads Behavioral BrainSciences : 337 368 . .13 . Mind. DretskeF. . . . 1993b Can intelligencebe artificial? Philosophical . 47 no. 1992 Psychodynamicsand the unconscious American -787 . 1993cConscious . Harnish1986 . . . 23 nos. . Mental Causation . . .

. 3. 1983Modularity Mind. . Uehling Jr. . 1993How do we know our minds: the illusion of first person of . E. 4: 533 . and D. . Corballis 1976 Negativeaftereffects visual .. 1980Methodological . . . French P. 1990 TheBlackwell Dictionary Cognitive of Psychology Oxford: Blackwell . H.1: Fodor J.14 . T. Olson eds. D. of no. Fodor J. Reconsidered . . Wettstein eds 1986 Midwest . : and . . EysenckM. mind : From cognitive connections to mental representations. as solipsismconsidered a research strategy in cognitivepsychologyIn TheBehavioral BrainSciences 63-72 . and M. . C. ed. . .14 and . FlanaganO. . . 1990 TheOrganization LearningCambridge MA: MIT . . . Astington P Harris. 19 Journal Philosophy . in . . Minneapolis in . UniversityPress Fodor J. of no. . . . Studies Philosophy . Press Universityof Minnesota Gallistel C. 1987Psychosemantics . . IndianapolisIN: Hackett . 16 knowledge intentionality The . A. 1994ElmandTheExpertMentalese its Semantics .References 194 . Cambridge . . . 1992 Mind and Brain. GodfreySmith P 1994A modernhistory theoryof functions Nous GoodmanN. . 1992Consciousness . . . Theories theMind. pp. and H. . and 1. 1990Visual . CambridgeMA: MIT Press . .. 10 Studies thePhilosophy Mind. Behavioral BrainSciences : 1. E. . . . . Cambridge~ : MIT Press . of . of Fodor J. Reprintedin Mind andBrain New York: W. . 1988 The developmentof children' s knowledgeabout the Flavell J . 1969Husserls notion of noema The . . Developing of New York: Cambridge . . . . . . . J. . perceptionReprintedin Rock1990pp.. CambridgeMA: Mit Press . 66 Journal Philosophy . . and . . . 48 -57. GodfreySmith P 1989Misinformation Canadian . . . 1976LanguagesArt. of . 3: . Freeman Co. Favreau O. CambridgeMA: MIT Press . H. . -550 . 20: 680 687 . MA: MIT Press . ' FellesdalD. Vol : in : of . K. W. .. R. .. 25-36 Farah M. . . In J. . Agnosia Fischbach G. Gopnik A. Press . ScientificAmerican 267 no. .

.. Hall. Mind. 1990 Gibsonianrepresentations Research 243 252 . 1979The sprandrels SanMarcoand the Panglossian paradigm : A critique of the adaptationist programme. . Horgan T. IndianapolisIN: Hackett . O. matchingentail a lossof information In M. for : prospects unification Psychological . Grice. . 1992 Color perceptionand neural encoding doesmetameric . 1988Privilegedaccess : . P. Forbes . 1984Jackson physicalinformationand qualia. C. Representation perception and . . D.References 195 of . processing " : and cognition connection in Hatfield. Rt Jmelhart . Stich ist affordancesPhilosophy Connectionist Theory . 1992 TheNatureo TrueMinds Cambridge CambridgeUniversity / . Hilbert . Quarterly34: 147 151 . Erlbaum and D. I991. In berlin 1990 . . no.. forthcoming Is consciousness perception what passes . 1989 Studiesin the Wayof Words Cambridge MA : Harvard . 15 . 12 . C. Hillsdale NJ: Lawrence . Hebb D. G. and D. and connectionistsymbol . J. S. for . . G. Association of .. eds. GiizeldereG. no. and R Lewontin. R. D. 1990TheIntrinsicQuality of Experience Tom . KaszniakA. G. MO: Times . Hull . ) (London 5205 581 598 Society of ProceedingstheRoyal . 2. and Behavioral BrainSciences . Harman G. Hardin. D. Hilbert. in one for . Press . 1986Color Philosophers . Hatfield. . Philosophical . 1981 . J. R. Grier . . J. W.. color vision and the objectivityof color . Hatfield. 68: Studies 351 370 . Mirror / Mosby " of the . 52: . 1984 Biology of Animal Behavior St. Hell. Hameroff A. . 1969TheMind's Eye Psychology . . . . . . Hilbert. . 1987 Color and Color Perception Stanford : Stanford University/ CSU. . Gould. 19922 volumes EastLansing MI: Philosophy Science . Hell.. " ' s own mind? In Towards a ScientificBasis Consciousness . . . L. 1992aComparative -39. 1981DiscriminationIn McFarland . eds : . . . . 47 238 251 . . Scott eds CambridgeMA: MIT Press . S. . . PSA . 1: 38 . R 1992bWhatis colorvision? Philosophical on . Louis . Today : . . J. . UniversityPress . RamseyS. . : . W. G.

Dame IN: Universityof Notre DamePresspp. Actionsreasons theexplanatory of contentIn . 1986pp. . Philosophy Science. . -397 . . H. .. Freeman . . 1974 Vision in a monkey without striate cortex: a case . . 6. 53: .. . 1992 A History of the Mind: Evolution and the Birth of Consciousness York: Simon& Schuster . . . . .. -256 . . and HorganT. of in . pp. and . 108 128 . . Land. . 1991 1991 -101 Mclaughlin . . : Kluwer Academic . Humphrey N. in . . . JacksonF. Press Kan WaG . . lienson eds 1991Connectionism thePhilosophy of Mind Dordrecht . In Hwnphrey. . . 237 no. Uehling Jr. 1977 The retinex theory of color vision. . American . Cambridge . Rock ed. . and . Brain . pp Reprinted Rosentha11991b . 377 382 .. . 392 394 .References 196 . MA: ~ PressA BradfordBook . and H. Wettsteineds Notre of . E. . . H. . and . Greenfield1987pp.P .. 3: study Perception 241 255 . . 18 . . G. HorganT. and .. . Kinsbourne M. New York: W. Humphrey. Cambridge . 155 163 . 1970 What the frog s eye tells the monkeys brain. . : Jacobs H. . Loewer 1986 SolipsisticSemanticsIn French et al. Kitcher P 1993 Function and design MidwestStudies Philosophy . N. Humphrey N. . 1981Comparative VisionNew York Academic . and B. . contours Reprintedfrom Scientific American in ThePerceptual World I. andJ. 379 . Publishers ' ' .. / . New . le Pore E. -337 . Color . 1977Perception UniversityPress ' t know. 239 . 1976Subjective . . .. 595 614 . role . 83: . 1986What Mary didn Journal Philosophy 291 295 . 1992Qualia awareness SellarsNous . . and Co. Hump~ ey N. . Beh 0013: 324 . 73 . . JackendoffR 1989 Conscousness theComputational Mind. . . . . FrenchT. pp. JacksonF. . 1987 The inner eye of consciousness Blakemoreand . . In and Bisiach1988pp. 1972Seeing nothingnessNewScientist 682 684 . N. Scientific . . 1988 Integratedfield theory of consciousness Marcel .. LeedsS.

1990Phenomenal . Subject Thought and Context. . . 1986 Singularthought and the extent of inner space In P . Oxford: OxfordUniversityPress . . Hill . statesIn Tom . . . 1991The . . . Minds CambridgeMA: Mit Press . T. 47 no. experience . 1988 Biologicalfunctionsand perceptualcontent Journal . Marr. McCulloch and W. J. Czyzewska acquisitionof . Lycan W. G. Lettvin. the computational . Cambridge MA: MIT PressA Bradford / . lloyd . volume 1. 121 158 . .. and E. Papers . 1989Simple : . McGinn C. and M. 1989MentalContentOxford: Blackwell . . D. 47: Engineers 19401951 . 1987 Consciousness . Marcel A. Lycan W. . . . 1982VISion Francisco H. . . PressA BradfordBook / . . ~ . . Presspp. .. . 1981TheOxfordCompanion AnimalBehavior . Y. Lewis D. G. Petit and J. 1983 Philosophical . D.. . of . consciousness lloyd . . the frog' s eye tells the frog' s brain. Freeman Co. . Bisiach 1988MentalRepresentation Consciousness . H. . McGinn C. Oxford: Blackwell . . Proceedings the Institute of Radio of . Maturana W. Psychologist . of Matthen M. : Dordrecht Kluwer. . McDowell J. 6: 796 801 . . and functionalism In Marcel . Press . D.. Oxford: Oxford University . . OxfordUniversityPress . In Horganand 'fienson1991pp. Problem Consciousness McGinn C. 81 108 .Cambridge MA: MIT . S. 137 168 . 1991Leapingto conclusionsconnectionism . and . of Lyons W. .27. 1: 5. berlin 1990pp. 1990What is the Subjectivity of the mental In Tom . of . Oxford: to . 444 459 . eds. D. Oxford: Oarendon Science in Contemporary and Bis ch 1988Consciousness . no. 1986 TheDisappearanceIntrospection . 1982TheCharacter Mind. Oxford : . information American . J. Marcel A. Book " " berlin ? . ...References 197 . San . . 1988 Phenomenal . Pitts 1959What . R. and : W. . Loar B. 85 Philosophy . McDowell . ed. mind. OarendonPresspp. . 1990 . 1992 Nonconscious Lewicki P. and . J. . H. Mcfarland.

McLaughlin B. Psychologist 90614 . . A. U. In Philosophical . 33: . American . effects the conceptual : . . on . Nagel T. 1991 Critical Essays thePhilosophy FredDretske of Oxford: Blackwell . . Millikan .References 198 . Milner and Rugg1992pp. NatsoulasT. T.. R. 4: .302 . Millikan . L. . . for Millikan . of Consciousness ." Australasian . 1979 Intermodalmatchingby human . . A . 1992 Disordersof perceptualawareness commentary In . Millikan . 4: 13 . Journalof Philosophy 86. . . Millar . 1978Consciousness . 1989a Biosemantics. . . 's analyst . of Neander K. . Miller . Review Questions . R. 1986 Thought without laws: cognitive science with co~. Review83 no. neonatesNature282 403 404 : . Foundations Realism Cambridge . Milner. 1989b In defense of proper functions . A . Cambridge . Philosophical . London Academic : Press . defencePhilosophy Science . . Concepts Natsoulas1983 ofconsciousness ofMind Behavior . Milner . G. A. . MA : MIT Press .. Philosophical Review 95. . . R. 1974 What is it like to be a bat? Philosophical 435 450 . N. 1984 LanguageThought and Other BiologicalCategoriesNew . Cognitive . G. 1922 The refutation of idealism.. Philosophyof Science . E. 1992 The Neuropsychology . . G. -59 Neander K. . Journal and . . R G. . Crofts Nemirow 1980 . 1991 Reasons ExperienceOxford : Clarendon Press .. G. New : . . 89 . 1991b The teleolo~ cal notion of " function. 1991aFunctionsas selected . and . Review Nagel Mortal of 's . . and MD . : . Studies : London Routledge KeganPaul and . W. I. Psychology YorkAppleton . Moore. . D. . MA : MIT Press .. and Borton R. ed. . 56: 288. Ruggeds . .. 1984 Husserl. 69 Journal ofPhilosophy -Century Neisser 1967 . D. Meltzoff. tent. 58. 139 158 .

Oxford: Blackwell and . FunctionNew York: Macmillan . . 1975 The meaning of Meaning . Peacocke 1992aScenarios . 1970PainperceptionPhilosophical . . . 7. Press Review 368 . W. W. " Papineau0 . Roberts1959 Speech BrainMechanisms . . C. 252 . . 79: -93. no. Whitham. 1987Reality Representation . . . 105 135 . Noonan H. ? but deepcontingency MentalCausation. 91 of challengeJournal Philosophy . Rasmussen . no. and L. published in Paige K. and . and . Perner J. Minneapolis Universityof Minnesota . A. Peacocke 1992bA Study Concepts . 1993DirectReference . 1993 Object dependentthoughts: a caseof superficial . C. and Peacocke 1983Sense Content . C. 1985 Report of research American . 1978 When is attribtuion of beliefs justified ? The . Scientific Science . 1991 Understanding Representational Cambridge MA: . 283 . . NJ: Princeton UniversityPress .. Nikolinakos 0 . . . J 'll necessity -308 Presspp. . N. Oxford: Clarendon . . Knowledge Minnesota Studies in the Philosophyof Sciencevol . RecanatiF. Press . 2: 88. . A Clinical Studyof Localization of . and T. Oxford: Clarendon . Language Mind and : . Z. and . the . . Pylyshyn . . eds . Oxford: Blackwell . and Behavioral BrainSciences 592 593 . PapineauO~1993Philosophical . Oxford: Blackwell Naturalism . . " " . Penfield W.. . . Putnam. . PrincetonNJ: Princeton . K. . 1994Generalanesthesia . . Princeton . MIT Press . . . CambridgeMA: MIT Press . 1977 In defenseof introspection Philosophical 77 88. . Press : Gundersoned. 4: 74. 2: Exchange Quinton.. . 1983 Sensing World Indianapolis IN: HackettPublishing .. . . Pitcher G. . G. . of 1957 Penfield W. the Perkins M. . Pitcher G. . He andAMele . H. of . . Comparly Mind. .104 . . 1: . . 1971A Theory Perception . University . consciousness the skeptical . concepts perceptionIn Crane1992pp.References 199 . and T.

5 lAnguage . of in experience colors Conception theMind: Essays Honorof of A. 2: 155 166 Ro thai. . Humphyries . . 1988 A questionabout consciousness H. Oxford: Oxford UniversityPress . . 4: 489 . RosenthalD. no. N.References 200 . . Oxford: Blackwell . . . sentencesin M. 292 315 . The Review . . and ShepardR. . Studies . 94 no. Philosophical Studies27: . NJ: Lawrence Erlbaum Associates . Research . 1992aOn the physicalbasis linguistic representation conscious . no. Report no. 289 319 . . Philosophical . . Tuedio . Harman. . In . PerspectivesMind. 102 Philosophical . George Miller . . 1: 50 . of . -orderthoughts Philosophy . . Dordrecht Reidel on : . ShoemakerS 1975 Functionalilsmand qualia. 15 36 . . . 1: 9. and quality . . . -278 . Consciousness . ShepardR. -513 . 1992 Conscious unconscious . Rey G. 1993aHigher .. switched. 1992The . 1986Two concepts consciousness . . . M.6 Psychology. process in language . of RosenthalD. and es and Rugg M. 3: 329 . Studies 68: Rey G. . In Villanueva1991pp. . 4: 911 918 . . . Rey G. Groupon Mind and Brain ZiP Universityof Bielefeld RosenthalD. G. and P Merikle 1990 On the inter-relatedness theory and measurement in the study of unconscious process Mind and es. ~ andPhenomenological Research . 263 . 1992 Sensationalsentences . and individuation in Marr ' s theory of . .. no. 53 . Otto and J. Seager 1991Metaphysics Consciousness of SearleJ. . eds. RediscoveryMind. London Routledge : . . Philosophical consciousness .28 . of . 1990 A theory of consciousness . D. 15 space Behavioral BrainSciences . the . 40. . ed. -359 Rosenthal D.. CambridgeMA: MIT Press of L. . Hillsdale . 1991aTheindependence consciousness sensory . -orderthoughtsand the appendage . . Nature Mind. no. D. Philosophical . Shapiro vision. 1991bThe . memory commentary In Milner and Rugg1992pp.. 1993bMultiple drafts and higher . RosenthalD. W. 1993 Content kinds. 1992bWhat in the world determines structureof color ? and -51. 1993 Sensational . theory of . Davies and G. of Reingold E. . . N.. . . eds. no.

1983 From Folk Psychologyto Cognitive Science The CaseAgainst . and . Sperling.. and Phenomenological . . Wettstein 1986 . M . 11. R. S. eds. Cambridge . Anderson and J. 1960 The infonnation available in brief visual presentations. . From French. no. Stalnaker . no. Stein.910. Philosophy . . . JE . T. ShoemakerS . no . 100 no 4. MA : MIT Press . B. and M . . vol . . 1993a. Sober E. StampeD . Studiesin SemanticsMinneapolis : University of Minnesota Press . C. . ed . Stoerig. Sterelny. 1991 Qualia and consciousness Mind . 1990b Qualities and Qualia : What's in the mind . G. Uehling and H . Cowey 1992 Wavelength processing and colour experience and Brain Sciences15. Philosophy Research . SO . 1993b. Oxford : Blackwell . Propositional Attitudes . no. S. 1992 Does the perception of temporal sequence throw light . . Atascadero. and A . 249. 1990. K.314. 4: 905. Meredith 1993 TheMerging of the Senses . 1990 TheRepresentational : . " " ShoemakerS 1994.228. of Philosophy Mind . Supplement : 109. In P. 1990a First -Person Access In Tomberlin 1990 . Narrow content . . Self knowledge and Inner Sense. Wettstein . 1977 Towards a causal theory of linguistic representation.2: 225. 53 Oglcal ShoeJnaker. Behavioral . Philosophical Perspectives 4: Action Theory and . 2. 1990. 1986 Introspection and the self.References 201 . A .theory . Treisman. Behavioral . Research . Midwest Studies in Philosophy . ShoemakerS . A .kerS . The Royce 2: 249. and Phenomenological Lectures. . MA : MIT Press . Special access lies down with theory . Issue: SO7524. P.. Centennial . . on consciousness Behavioral . ShoemakerS . and Brain Sciences15. Tom berlin .79. Phenomenol Research . Psychological Monographs 74. . . Theoryof the Mind . Stanford: CSU . . French.131. 1: 78. Lovely and suspect ideas . 1984 TheNature of SelectionCambridge . Owens. CA : Ridgeview . and Brain Sciences16.1: 53. no. . Philosophy and Shoema . ShoemakerS . Uehling.314. Stich. no. Belie Cambridge : MIT Press }: . 54 . E.

. Attention in Vision London: . J. In Bisiach -1988pp. VelmansM. . . . 17 211 230 Topics : van Gulick.. . . urn and consiousness . ed. 1991 Introduction Dissociated . yishi. . UniversityPress . ? conscious Behavioral . Consciousness in . andBrainSciences . . H. Weiskrantz . 158 176 .. 1989 TheMetaphysics Mind. Bisiach eds Oxford: Clarendon . 1986Thesubjective qualitiesof experience .References 202 . . Tye M. Oxford: Clarendon . . . . Tye M. . 1990TheChilds Theory theMind. 1983AnimalThoughtLondon Routledge KeganPaul . M. of Van der Heijden. ' Wellman H. : . 1991Consciousness . Cambridge England Cambridge of . 1988. 1992 Selective . 1994Qualia content and the invertedspectrumNous Umilt . R. . 1992The . . -356 . CA: Ridgeview . Press WeiskrantzL. 1. ed. IssuesIn Milner and Rugg . : and . 95: 1. . 1991pp. ValbergJ. 14 Villanueva E. . Tye M. 4: 651 668 . Science J. of PressA BradfordBook . A . 1988b Thought . 1993Understanding phenomenal . Philosophical . 1991The Imagery . . . . : . 1988a Some contributions of neuropsychology of vision . . . 183 199 Weiskrantz L. duh. 1989. Tye M. the mind: arewe all just ? armadillos In Daviesand Humphries1993pp. . . C. pp. Contemporary .. Routledge van Gulick .17 . Press . 1991 Is human informationprocessing . What difference does consciousness make? . 1964AttentionOxford: BasilBlackwell . . R. Atascadero . Marceland E. . 334 . Tye M. Wilkes K. R. . . . without Language . Mind. Oxford: OarendonPress . V. 1992VISual qualiaand visual content In Crane1992pp. no. . .10 . CambridgeMA: Mit Press .. Puzzle Experience . Walker S. Debate . . A. . L. . CambridgeMA: M1T . 137 154 . ... C. / White A. 1988 The control operationsof consciousness Marcel and . and memory to the problem of consciousness In Marcel and Bisiach 1988 .

W. 1976Teleological . . C. BerkeleyUniversityof California . pp . Puhl . Berlin : de Gruyter . . Zeki. Review 139 168 . W. Wright.. . 82: . 3: 69 76 Reprintedin Mind andBrain New York: W. 126 147. 69 90. L. . . 1: 29 48. L. Meaning Scepticism K. 27 39. privacy Wright . no. - . Scientific . and EH . Wittgenstein . . de Haan 1992Facerecognitionand awareness . philosophy of mind : sensation.. Mind andr. .Qnguage no. 1992Thevisual imagein mind and brain. Young A. pp. 1974 PhIlosophical ' s later . . L. 1991 Wittgenstein and intention . ed. Investigations Oxford : Basil Blackwell . F. H. . S. Explanations Wright.. and . afterbrain injury. de Haan1990Impairments visual awareness .References 203 . 1973FunctionsPhilosophical : . 5. Press . In Milner and Rugg1992pp.. of Young A. and EH . F. Freeman Co. . American267 ..

. 10 . Bemecker.. . . ... objects10 11 actsvs. 152 Belief15 . Bilgrami . E. 7 . 58 .. . Baker R. Aspectual shape30. Coding 171n16 . 100 104 creature state 98. 107 . functionsof. Clark. BurgeT. remotevs. 186n20 . . . . proximal 153 . 39..Index Aboutness28 30 . . . . 119 Blindsight . Behavior .. Qrlsholm.. 73 . . R. . . 176n3 . . Qua1ia with of conscious and conscious . Olurch1andP M. 21 71 72 177n6179n21 . .. Cause151 159 161 . .. 105 BachK. . . 33 173n22 \ . A.. . -20 Cognition 16 18 . 132 135 176n3 ArmstrongD . K. Bisiach . . A.S .10 of properties 101 102 . 104 106 141 . 79 178n15 . . 28 . 24 . 105 116 minimal. .J . 98 See also . 116 122 of higherordertheories . . vs. Concepts9. . . conceptual sensory9. 95 Analog l7ln16 . B.. 20 136 138 belief 58 60 . 176n17 . Boghossian 40 Brentano . . .100 113 162 and . . . Consciousness . -17 -60 Bennett. Experience . 22 Color 88-93 171n10 . Connecting also Consciousness See . . 131132 . CarruthersP. . . . phenomenal22 66 69 83 . 155 . M.. . 24 54 55 125175n14 175n15 Calibration 15 18 20 49 . . 28 O ristensen C. 175n14 Biro J. . Appearances . . .32 Awareness . . .. P.. L. . Perception of abstract . objects97 98 . 170 181n7181n10 .. Awareness . 67 73 128 doxastic67 132 135 176n3 . 77 170171n8 . Closure 175n16 . . - BlockN. 22 . -21 Adaptation 20 Akins. 182n18 . .

24 .8. 35 125127 . explicit vs.7. . .. 113 116 119 . 29 101102 . .. . D. . . . 169n4 .. . 177n9 HatfieldG. Humphrey . states systems169nl vs.J . implicit. 113141170176n10 .2 . . 89 . EvansG. Intentionality 8. . . . . .. Images102 Indication 4. 176n18179n22181nll . 164 . 104116117 Hilbert D.. . 15 125 . 175n14 . L.. . . Historycausal 153154 . 127 130 149 See Perception also . 39 . . 170 . 17 .. . of children 110 111 .. systemic acquired 2. Go . 54 55 175n14 'll . D. . No . assigned6. . Digital 171n16 Directedness. M. .92 . 34 37 62 94 . 140 141 152 . Cummins . 154179n22 186n17 .. 180n5 HardinC. conscious unconscious vs. . Evolution 162 .34 . . 29 101102180n5 . . 172n19 .. 7. 183n1 .. . 179n22 . . J. 183n3 Davies . vs. 171n11 . . 59 110 . . 32 Discrimination -69 70 . 15 77 151183nl . 102 103 ... -75 Dreams . .. .Index 206 . . Phenomenal experience of animals 107 111112 . HorganT. 117170177n9 . of properties29 . . of the senses . . . 9. I . . 32 ..) (cont . 130131 FlavellJ. 28-34 Intemalist intuition 145149 150152 . He . 117120121173n2 Firstperson . . 72 vs. . T. 181n10 . -order theories consciousness (of Higher . D. 23 . . . Co . 21 178n13 indicatorfunction 2..10 12 15 17 19 . Godfrey 171n19 Goodman ..8. 24 .. -95 .S . 48. . . . Experience . . . Introspection39 48 105 106 inner sense62 106 .Gallistel Ro5 -SmithP. 68 . . . . . Information 2. 148175n14 . 93 172n19 . 24 Crane . . 185n15 . Context -26 . of . . . 7. relevance. ).. . . . .. InvertedSpectrum72 177n7 . . Davidson . DreamsHallucinations . 4.. 99 of perceptual . 24 .50 75 . . . 4. 83 132 145(see Qualia ) Externalism40 54 123 127 . 48 52 59 75 166 . . 16 Constancy Content . Introspective knowledge 39 41 53 56 63 135 140 149 . Follesdal . 7. . 183n1 Dennett . biological 15 126 169n4 conventional natural 6. . R.1 . 181n11 Hallucinations . . .19 169n3 Consciousness . N. -54 authority -40 53 . . . 176n20 . Function . 105 perceptual forms . . 14 48-50 75 . Fodor . GopnikAo59 Giizeldere . . . 138 . Enc B. G. also . 19 170 173nl. 7. . Harman . . . . quality of. ..

97 l8Onl . . absent141 151 . PernerJ.. Appearances of LooksSee asa species consciousness . Pointof view. 48 . 174n13 .. . . . 43 59 110 . 7 . . 24 43 177n4 sense cognitive 118 vs. 170 Jackson. . 174n8 Perception109 110 also . epistemic non. PaleySyndrome146 147 . 176n3 Putnam H..F Reference -25 28 . Knowledge55 .. Noonan H. Pylyshyn Z. 67 95 151170 .. . 85 86. 8.. 141 . 12 22 129 Marcel J. 187n24 . knowledge displaced40. Natural selection -50 156 . 3 Representation . . . R. 173n21 .epistemic 17On8 Learning .Index 207 .. 24 . -30 . . PaleyW. 175n16 . 4. Privilegedaccess attitudes xiv 176n20 . 179n22 Peacoke 172n18176n3 Kitcher P. 179n22 . LoarB. 45 81 . E.. NagelT. 141 170 181n7 . 177n9 Propositional .. 17 67 68 172n18 . 173n3 secondary Lloyd D. 89 Protoknowledge Millar A. Quinton A. . . 4. 36 106177n9 Perkins M. Qualia xiv. 7.I McGiImC. 120 . 179n22 . . . LycanW. . 141151 -26 . M. . 65 ~ . See Qua1ia Matthen .. .. . .. 85 . Proprioceptionxv. Miller I. 7S 81 McDowell . W. . . 186n22 . 176n19 Recanati. . . 24 . 32 33 . . A. . 108126 . -56 . . 111139 141 of . 7 . 126 Replacement argument .. 20 Modularity MooreG. 43 Millikan R. 175n14 . See Introspective . . 185n16 KinsboumeM..1 . 99 118 .. experience . . . . . . . . 102 103 Meaning .. 67 81 82 . 28 72 123 .. . 23 .F . .. Metameric matches -91 . .. awareness . . 50 77 125 . 179n24 . properties 132 135 also Materialism . 65 . 5 Phenomenal . 174n13 PitcherG. 27 28 67 169n1 . . . 91 172n19 . . . 41 44 58 63 vs. 17ln13 assigned . K. 49 . problemof. . 25 . 78 80 Natsoulas . 22 65 72 73 78 83 85 94 123 124 . ... . . . 39 . inverted 70 73 123 . 169n3 nonnal misperceptions LewisD... . 101 ... . .. 183n1 . xiii Neander . 176n20 Phenomenal . PapineauD. T. . Nemirow L. LyonsW. . .. 177n10 . . .. 125 126 . Marr D. . . Pains 102 103 ... Jackendoff . 7. 162168169n3 Naturalism . . 170 173n2118On4 179n22 . . 1. Otherminds. .. .

. cause Triggering . 170n5177n8179n19 179n22182n14182n16 . . vs. RosenthalD. . 172n19 . . 126 .. Subjectivity33 65 . M. . . Schroeder . W. 40 . misrepresentation. 175n14 Zombie144148185n13 . . 179n2 ... 43 .. . . 179n22 . Sensory systems20 . Rey G. 53 55 . .. 175n14177n9177n10179n22 .. 5 . 4-5 Teleology . N. T. Representational 48 57 72 73 81 124 126 154 . 10 124128130132 Treisman .. 7. H. Turing . conceptual sensory9. 143159 Twinearth126183n2 . Structuringcause159 161 . 125 . 182n14 . . metarepresentation . 126 152 161 . 25-26 45 states systems8. . . . WrightL. 92 .xv. .. 183n1 SearleJ.. 36 176n20 181n7 . . . .6. . . 162 . ShepardR. . . . A. .Index 208 . . 182n17 . 17On4 . . 174n7 of properties2. 176n20180n2182n15 . Seager . vehicle 35 . StampeD. 121 Wellman M. TyeM. Swampman144 145 148 183n3 (cont Representation . . pure and hybrid. . 3. 108 109 . 182n1 vanGulickR. . 5 . Weiskrantz . 98 106 111112 . . 27 28 67 174n13 of objects23 . . contentvs. . 172n19 . 181n11 StalnakerR. SwampPlant 158 . SterelnyK. 72n20 .12 42 45 72 107 118 119 . 7. . . Secondary properties88 Sensation12 16 18 22 121 . .. . Shapiro L. 172n20 . Supervenience . . . . .. . .8 vs. . Sense23 . 159161 test .. . . .) vs. Representational Thesisxiii. Thoughts .19 vs. . systemic acquired 12 19 Naturalism xiii .. . . . . . 59 110 ... 183n1 Stich S. 177n9 . L. . Wittgeristein . . 31 . . 8-9. . . 175n14 . -37 conventional natural 6. . 24 25 mental 5.4 . .44. . 127 130 180n5 data . -27 pictorial. 23-27 94 95 . 182n14 Velmans . . . ShoemakerS 72 77 173n1 . Sense . de re. 19 17On6 . xv .

Sign up to vote on this title
UsefulNot useful