‘CAMBRIDGE ‘UNIVERSITY PRESS ‘Unverty Ping oat, Camtege CS, Usd igor “pbs the itd Sates mei by Cage Unveriy Pres New Yr ‘Cancige Unversity Pres pt fhe Unvenityof Camsig, earn the sive’ ison by eminem i te pars of Seca, esnng and esr Meee trata ees of nance srcanbie Tnfrmaes on hse wo abe org 71107527970, “Ti pulon sin opi. Subject taut exept toe prin of eer eolee lena agreements a repection of nf art take Pla witout the wren Ferman of Cambie Uae res rt pees 2005 Scone won 2083 epee 015 itd he United King by Blan a sow Acs rth pti ae fo he Bk ay ry Congres Caton Po te Den, dvi Aral Introd phnsog Dav Odden, Send Bion, Parc. camiepeintracor engung dngusts) ssa 7hs-toraatees id= SIN 070107427970 perdu) f camiay, Comparten stare Ponsa. 1 ie Ain resources fo th pletion at yw cambrieopoden Camtige Uivesiy Pest a repay er the peastnce a seco era ay mt te rin ein ‘Sn de a ares that ny content sch wees Wl ea Contents About this book page vi Acknowledgments ‘A note on languages Us of abbreviations 1 What i phonology? Phonetics - the manifestation of language sound 12 Phono: the smbole penpectve on sound Summary Exerciser Suggestion for futher reading 2. Allophonic relations English consonantal allophones 22. Mophonyin cher longus summary Exerses Suggestion fo further reading 3 Feature theory ‘Scientific questions about speech sounds 32. Disncve feature theory 133. Features and classes of segments ‘34 Possible phonemes and rues - an answer 135. The formulation of phonological rules 3.6. Changing the theory Summary Exercises Suggestions fr further reading 4 Underhjing representations ‘The importance of correct underlying forms ‘22 Refnng the concep a undying fom 43. Finding the uneryng form 444 Pratce at problem solving 445_Undetyng forms ond sentence evel phonology ‘46 Underhing forms and multiple columns in the paradigm Summary Brerses Suggestions for further reading 39 a5 61 67 n n 7 8 np 80 as 93 98, 98 107 107 ns38 © e209 topcaa wn pans ‘dy gamioo ‘bet pu rubbet qu ‘love? rw ‘500° kept theadmart key ‘head! gomba ‘oud eee wie’ Pou ‘haibor uo ‘cast pechah 12 Kirzan Armenian ln this language, certain safe vowels can be predicably derived fom other, Lndesvng vowels Discover what vowels nth language ae prey predicabie, ar ‘cow ings ‘aden aon Se ‘Srv fpedmax ta te” Boum oer ane Ci Soca a sine tnd pve the rule which deives the predcabe vowel wy met oF lov oak sing an ok seria Pabew port aw fee et for ge beh uk oxcons len Further ring “baby chicken’ oe “east ‘cont “dstncnon! “ngerat ‘a “rok Toute! ‘yelow ah od) we oe ‘3d wor! peak) ‘ive se os ‘pode! throat ‘nembie ‘vie Cahn 1995; Hole 1959; Hore 1994; Kahn 1976; Sai 1925, ean by. ie’ ee a kanant™ end aa Poe aver ‘sce ‘ti transator ‘conect ‘ercocat cot ‘cheese ‘post ofc! important "Yale a house! ‘bee led? “church val “riserable’ ‘ox ‘outs’ ‘adh “Rubiecese plant “senat bay ‘Seoxetorm ‘tier =» ‘water ‘wot CHAPTER 3 Reaves Feature theory PREVIEW. This chapter explores the theary for -epresenting language sounds as symbolic units. You wil: © see that sounds are defined in tems of a fixed set of universal features © learn the phonetic definitions of features, and how to assign feature values to segments based on phonetic properties ‘@ understand how phonological rules are formalized in terms of these features ‘see how these features make predictions about possible sounds and rules in human language40 “We have been casual about what sounds as cognitive units are made of, and jue treated them a8 lettre labeled by traditional articulatory doterp. tions. tis time now to raise a fundamental question: are segments further analyzed into "ports" that define them, or are they truly atomic - units ‘which ate not further divisible or analyzable? One of the scientific questions that need to be asked about language is what is posble speech sound? Humans can physically produce many ‘more kinds of sounds than are used in language. No language employs hhand:clapping, fingersnapping, or vibrations of air between the hand and cheek caused by release of air from the mouth when obstructed by ‘the palm of the hand {though such a sound can easly communicate an tirade). A got ofa scientific theory of language isto systematize such facts and exphin them: thus we have discovered one limitation on language sound and its modality ~ language sounds are produced exci sively within the mouth and nasal passages, in the aea between the lips and larynx Even staying within the vocal tract, languages also do not. for example, use whistles or inhalation to form speech sounds, nor is a Jabiolingual tl (aka. “the raspberry") a speech sound in any lam sage. It is important to understand that even though these various ced sounds are not language sounds, they may stil be used in commu nication. The “raspberry” in American culture communicates a com temptuous atttude; in parts of coastal East Aftica and Scandinavia, inhaling with the tongue in the position for schwa expresses agree- ‘ment. Such noses lie outside of language, and we never find plurality indicated with these sounds, nor are they surrounded by other sounds to form the word dag. General communication has no systematic limi tations short of anatomical ones, but in language, only a restricted range of sounds are used, ‘The issue of possible speech sounds is complicated by manual lan- _guages such as American Sign Language. ASL is technically not a coun- terexample to a claim about modality framed in terms of “speech sounds.” But i is arbitrary to declare manual language to be outside ‘the theory of language, and facts from such languages are relevant in principle. Unfortunately, Imowledge of the signed languages of the ‘world is very restricted, especially in phonology. Signed languages clearly have syntax: what isn't clear is what they have by way of ‘Phonologies. Researchers have only just begun to scratch the surface ‘of sign language phonologies, so unfortunately we can say nothing ‘more about them here. ‘The central question is: what is the basis for defining possible speech sounds? Do we use our “speech anatomy” in every imaginable way, or ‘only in certain welldefined ways? Feature theory 3.1.1 Possible differences in sounds ‘ne way to approach the question is to collect samples ofthe sounds of ll Df the languages in the world. This search (which has never been con- ducted) would reveal massive repetition, and would probably reveal that the segment [mi] in English is exactly the same as the segment [x] in French, German, Tabatilabal, Arabic, Swahili, Chinese, and innumerable ‘other languages, It would also reveal differences, some of them perhaps 2 bit surprising. Given the richness of our transcriptional resources for ‘otating phonetic dferences between segments, you might expect that sf collection of languages had the same vowels transcribed a [i] and jl, ‘then these vowels should sound the same, This is nots Varieties of phonetic [i] s. [Many languages Lave this pair of vowels: {for example, Matuumbi has fi] and (]- But the actal pronunciation offi vi difers between English and Matuumbi, Mateumb is higher than {in English, and Matuumbi isa bit ower than English to some people it almost sounds Tike fe] (out is clearly differen from fe), even the “pure” [e] found in Spanish), This might force us to introduce new symbols, so that we can accurately represent these distinctions, (This is done in publiations on Matuumbi, where the difference = notated 2 “extreme” versus “regular” 1) Before we embark on a program of adding new Symbols, we should be sure that we know how zany symbols to dd. It turns out that the pronunciation offi] and fi difers in many languages: these vowels exist in English, Kamba, Lomwve, Metuumbi, Bai, Kipsgis, Didings. and Sotho, and their actual pronunciation differs in each language ‘You do not have to go very far into exotic languages to find this phonetic Aifference, forthe difference between English [i] and German [i is also very noticeable, and is something tha a language learner must master to Aevelop a good German or English accent. Althougl:the differences may be dificult for the untrained ear to perceive at fst, they are consistent, physically measurable, and reproducible by speakers. If written symbols are to represent phonetic diferences between languages, totally accurate transcription should represent these differences. To represent just this range of vowel differences involving [land (i, over a dozen new symbols would need to be introduced, Yet we do not intreduce large numbers of rnew symbols to express these differences in pronunciations, because phonological symbols do not represent the precise phonetic properties of the sounds in a language, they only represent the estentil contrast between sounds. Other variants of sounds. Similar variation exists with other phonetic ‘categories. The retroflex consonants of Telugu, Hindi, and Koti are all pronounced differently, Hindi has what might be called “mild” retrflex- ion, where the tip ofthe tongue is placed just behind the alveolar ridge, ‘while in Telugu, the tip ofthe tongue is further back and contact is made ‘between the palate and the underside ofthe tongue (sublaminal) in Kot, ‘the tongue is placed further forward, but is also sublaminal. Finnish, a