Wierzbicka, Anna. 1996. Semantics: primes and universals. PDF

Preview

Summary

180 LANGUAGE, VOLUME 74, NUMBER 1 (1998) manicc Verb Second phenomenon man phenomenon.Work Working papers in ngpapers n Scand Scandinavian nav ansyn ax 50 syntax 50.1-24. 1 24 Lund Lund:Un University vers y oof Lund Depar men oof L Department Linguistics. ngu s cs RIZZI,LUIGI.1990. Re RIZZILUIGI1990...

Description

180

LANGUAGE, VOLUME 74, NUMBER 1 (1998)

n Scandinavian 50 1 24 Lund: Lund University Scand nav ansyntax man c Verb Second phenomenon. manic Un vers y syn ax 50.1-24. Work ngpapers papers in phenomenon Working o Linguistics. o Lund Department of L ngu s cs Depar men of MA MIT Press. Press Re a v zed minimality. LUIGI.1990. Relativized RIZZILUIGI1990 RIZZI, Cambr dge MA: m n ma y Cambridge, Ans van Kemenade Un vers e Vr e Universiteit Vrije o English Eng sh Depar men of Department Boe e aan 1105 de Boelelaan Ams erdam 1081 HV Amsterdam Ne her ands The Netherlands kemenade@ e vu n [[email protected]]

ANNAWIERZBICKA York Oxford Ox ord Ox ord & New York: Pr mes and universals. un versa s By ANNA WIERZBICKA. Oxford Seman cs Primes Semantics:

500 $105.00. Press 1996. 1996 Pp. $105 00 x 500. Pp xii, Un vers ty Press. University Tasman a Rev ewed by BERTPEETERS, Reviewed PEETERSUniversity Un vers y of o Tasmania ca s he exploration o what wha she so aptly W erzb cka scon r bu ontoo semantics, Anna Wierzbicka's contribution seman cs too the ap y calls exp ora onof con nues too be immense. mmense o the he universe of meaning' he dazzling un verseo 'the 233 has been and continues mean ng (233), dazz ngbeau beautyy of o linguists nde b e mark markwh ch generations e on the he semantic which W has left seman c enterprise genera onsof ngu s s too come en erpr sean indelible a the me of o he time w thankfully Th s iss yet ano herco ec on of o her papers, will collection papers and at hank u yacknow edge This ye another acknowledge. c ear that ha it would he last. as wou d by no means be the wr ng it was clear writing he others), un ke all a the Chs 2 and 3, Ch 1 ('Introduction', Prev ous y unpublished unpub shed(like ke Chs. 3 and unlike o hers Ch. In roduc on Previously W s theoretical heore ca framework ramework he newcomer too natural na ura semantics seman cs (which 3 34 presents presen s the wh ch iss how W's 3-34) he necessary res of w h the ack e the he rest o the he book. book When I has come too be known) known with necessarybackground backgroundtoo tackle n this h s format'. orma Having rans a edextracts ex rac sfrom romW ear er W'ss earlier Hav ng translated prev ous yunpub shed I mean 'in say 'previously unpublished', n o French for o the seman cpr work into or an issue ssue of he journal devo ed too semantic m ves ourna Languefranfaise ran a se devoted primitives ma er a inserted nser edhere w hou major of material here without chunkso Ch Pee ers 1993), 1993 I recognize recogn ze large arge chunks ma orrewr ng Ch. (Peeters rewriting. dea s with w h the he importance o d means but bu with he help 1 deals o a w h the of analyzing not by any old mpor anceo ana yz ngmean ngno he p of meaning se of o which set o semantic seman cpr a of wh ch are lexical ex ca universals un versa s(although he opposite no true) m ves all a houghthe oppos e iss not rue primitives, of syntax nna e concepts; a so deals dea s with w h questions he combinability and presumably presumab yinnate concep s it also ques onso syn ax (e.g. e g the comb nab y o primitives, he issue ssue of o verb valency, he readeriss provided of h a short shor with pr m ves the va ency etc.). e c Furthermore, Fur hermorethe prov dedw o natural na ura semantic seman c metalanguage w h an account accoun of o how the he theory h s ory of me a anguage(NSM) NSM and with heory has history w h phenomena such as polysemy dea with men onbu ew On p. come too deal phenomenasuch po ysemy and allolexy, a o exy too mention p 28, 28 but a few. W falls a s prey, a er so many others, he belief be e that ha the he phrase prey after o hers too the phraseou ou se tient en iss too be found ound ou tout n Saussure. sn (cf. in Saussure It I isn't Pee ers 1985, c e.g. e g Peeters 1985 1991a). 1991a Ch 2 ('A o semantic seman cpr Ch. ons dealing w h sections A survey of m ves 35-111) 35 111 provides prov des separate separa esec dea ng with primitives', o d primitives old he test n Goddard/Wierzbicka he list s put es in Goddard W erzb cka1994) andnew new primitives pr m ves(i.e. e the pu too the 1994 and pr m ves(added added n the as couple in he NSM iss too be applauded, he last o years). con nuedexpans of the coup e of years The continued ono app auded especially espec a y expansion s nce it goes hand in n hand since handw h extensive ex ens ve empirical with researcha med at a avoiding he inclusion aimed emp r ca research avo d ng the nc us ono of wh ch do not no meet mee NSM's NSM s rigorous s andards There iss no comprehensive ab e of r gorous standards. concep s which comprehens ve table o concepts seman cpr semantic n the he volume. m vesg ven anywherein vo ume I will w redressthat ha lack ack below; be ow the he descriptors descr p ors primitives given n the he left e column in co umn are cover terms erms for or sets se s of o interrelated n erre a edconcepts concep s and are not no themselves hemse ves part par o the of he metalanguage. n italic. are in me a anguage The new primitives pr m vesare a c subs an ves substantives I YOU, YOU SOMEONE, SOMEONE SOMETHING, SOMETHING PEOPLE I, de erm ners determiners THIS THE SAME, SAME OTHER, OTHER SOME THIS, quan ers ONE TWO, TWO MANY (MUCH), MUCH ALL quantifiers ONE, MORE augmen or augmentor men a predicates mental pred ca es THINK KNOW, KNOW WANT, WANT FEEL, FEEL SEE, SEE HEAR THINK, nonmen a predicates nonmental pred ca es MOVE THERE THEREIS BE ALIVE MOVE, IS, (BE) SAY WORD speech and words SAY, ac ons and events actions even s DO HAPPEN DO, eva ua ors evaluators GOOD BAD GOOD,

REVIEWS

181

partonomy par onomyand taxonomy axonomy metapredicates me apred ca es interclausal n erc ausa linkers nkers

BIG, SMALL BIG LONG TIME,A SHORT WHEN, BEFORE BEFORE,AFTER AFTER,A LONGTIME WHEN TIME,NOW TIME INNEAR, SIDE SIDE, IN FAR, NEAR WHERE, UNDER UNDER, ABOVE ABOVE, FAR WHERE SIDE, HERE SIDE PART (OF), OF KIND (OF) OF NOT, CAN NOT CAN, VERY IF, BECAUSE BECAUSE, LIKE IF

imagination possibility mag na on and poss b y

IF ... WOULD WOULD, MAYBE

descriptors descr p ors time me space

The legitimacy discussed scussed aat length. would d have pre eg macy oof each pr primitive m ve iss d eng h Persona Personally, preferred y I wou erred 'IN' IN to o 'INSIDE' INSIDE as a spacepr Peeters detailed ers 1997 for or de a edd discussion based on French scuss onbasedon c Pee French). space primitive m ve(cf. addition ALIVE' iss we The add onoof '(BE) welcome. come Some time me ago n a letter e erto h s rev heFrench o this French reviewer, ewer the BE ALIVE ago, in Picoche coche quer rom the he me will now feel ee lexicologist ex co og s Jacque Jacqueline queried neP ed itss absence from metalanguage; a anguage she w recentinclusion. vvindicated nd ca edby this h s recen nc us on Very interesting Ch. 2 are the he frequent references erences n eres ngthroughout hroughou Ch requen re to o the he literature on ch child era ureon d language n wh which ch amp evidence dence may be found ound ampleesuppor acquisition anguageacqu s onin supporting ngev for various or the he appropr ous inclusions exclusions. nc us ons and exc us ons appropriateness a enessoof var Because NSM iss a me not a me universal versa syn metalexicon, a ex con it has itss own un metalanguage a anguageand no syntax, ax discussed d scussed aat length n Ch Ch. 3 ('Universal universal semanticcpr versa seman Un versa grammar eng h in grammar:The syn syntax ax oof un primitives', m ves various n Ch ouspr Ch. 2 aretaken den edin akenup once aga he rcomb combinatona o 112-47). primitives again, 112 47 The var m vesidentified n andtheir are examined rrial a poss nedand and illustrated means of so us ra edby meanso so-called ca ed 'canonical canon ca sen sentences' which ences wh ch possibilities b es areexam said d in n aall the he languages world he wor d (e.g. You/IId did d some bad, Th Thiss presumably something presumab ycan be sa anguages oof the e g You h ngbad under/above aboveaall these At some time hese oother her things, now I felt mebe before thing something good, eetc.). orenow h ng iss under h ngs A e some h nggood c Some oof the he descr n Ch Ch. 3 are d different eren from rom those hose used in n Ch Ch. 2 (and descriptors and reproduced p orsused in in n the he table ab e above or instance, are groupedtogether underthe he above):'DO,' DO 'HAPPEN', HAPPEN and 'MOVE', MOVE for ns ance aregrouped oge herunder THEREIS IS' and ('BE) descriptor events, movement', whereas 'THERE ALIVE', prev previously descr p or'actions, ac ons even s and movemen BE ALIVE ous y with h 'MOVE' MOVE under the he descr nonmen a pred descriptor p or 'nonmental grouped together oge her w predicates', ca es now make up a he r own called ed 'existence ex s enceand and life'. e The group Ch. 3 iss probab n Ch he grouping adopted probably category own, ca ngadop edin ythe ca egory oof their better be er one nce 'nonmental nonmen a pred rather hervague abe and m ha there here one, ssince predicates' vague as a label might suggest that ca es iss ra gh sugges are no further ur herpred o follow-although here are (cf. predicates ca esto o ow a hough there c 'DO', DO 'HAPPEN', HAPPEN 'SAY'). SAY Ch. 4 ('Prototypes Ch remains nsas as de read now as it was in n itss delightful 148-69) gh u a readnow Pro o ypesand invariants', nvar an s 148 69 rema orma (published n 1990 underthe he titlee 'Prototypes warnsagainst he w widede original save'). W warnsaga or g na format pub shedin Pro o ypessave ns the belief e that ha the he cclassical ass ca and the he pro o ca are mu prototype categorization mutually spread be o ype approachesto egor za onare ua y exclusive. exc us ve She illustrates, meansof an investigation hemean hewordsboa wordsboat, bache by meanso meaning bachelor, us ra esby nves ga onoof the ngoof the or he idea dea oof pro prototypicality congratulate, bird, mother, game, how the o yp ca yhas been congra ua eb rd lie, e mo her furniture, urn ure toy oy and game hen uses co color or terms, or emo well as the he words cup bird rd misused, cup, unc uncle, emotions, m sused and then e b erms words for ons as we mb to o show how the notion he no on oof pro o (again), cabbage, app apple, prototype may be pu put to aga n tomato, oma o cabbage e and cclimb o ypemay a be better use. A All in eruse n aall, she de defends ends the he thesis hes s that ha there here iss a p n seman semanticcinvestigations or place nves ga onsfor ace in both bo h approachesandthat ha no o be ga o the hecon nothing gained-quite contrary-from prototypes h ngiss to ned qu e to rary rom turning urn ngpro o ypes into n o 'universal un versa thought-saving devices' ces aallowing o ge h inaccurate get awayw away with definitions hough sav ngdev ow ng linguists ngu s s to naccura ede n ons and to o lightly which ch do no not ma match ch the he de definitions as 'nonprototypical'. n onsas categorize specimens nonpro o yp ca gh y ca egor ze spec mens wh Ch. 5 ('Semantic Ch semanticcfields', rs pub n 1992 primitives 170-83; published 1992) iss as d disapSeman cpr m vesand seman e ds 170 83 first shedin sap Ch. 4 was en ews aattributed r bu edto o lexical ex ca field e d theorists enjoyable. rejects pointing heor s saat oyab e In it, W re ec s vviews po n ng as Ch ha the he mean her word in n the large entirely meaning arge (e.g. e g that re y on every oother he lexicon), ex con ng oof words depends en without w hou rea ha hard he rad radical ca pos realizing hardly anyone today acceptss the positions valiantly z ngthat yanyone oday accep ons she so va an yaattempts emp s to o underm undermine. ne There iss no ev evidence dence oof any rea real familiarity with vast literature h the he vas on lexical am ar yw era ureon ex ca fields e ds (for recent co collection ec on oof papers on lexical ex ca field e d theory, Lutzeier ze er 1993 1993). Tr Trier heory see Lu er and or a recen Lehrerare the Lehrerare he on e d theorists heor s smen mentioned onedin n the he text, only genuine nothing noteworthy y genu ne field ex as though houghno h ngno ewor hy had been wr written enbe between ween 1931 and 1974 er 1974 1974. IIt mus 1974, and aga again must be sa said, n aafter d though, hough that ha W iss no not the he on writer er to o make this h s par mistake s ake (cf. only particular Peeters 1986). y wr cu arm c Pee ers 1991b on Kovecses 1986 In Ch Ch. 6 ('Semantics rs pub "primitive" 184-210; published 1994), W po points Seman cs and "pr m ve" thought', hough 184 210 first shed in n 1994 ns out that ou ha the he my o wh which ch 'primitive myth according people' non-Western h oof 'primitive pr m vethought', hough accord ngto pr m vepeop e (i.e. e non Wes ern tribal r ba soc abstract rac thinking, head again. societies) rearing ugly Abstract e es are incapable ncapab eoof abs h nk ng iss rear ngitss ug y headaga n Abs rac thought hough typically requires conceptss such as IF IF, BECAUSE BECAUSE, SOMEONE SOMEONE, ALL ALL, KNOW KNOW, and THINK yp ca y requ resconcep

182

LANGUAGE, VOLUME 74, NUMBER 1 (1998)

which ch are par he NSM NSM. IIt iss shown that ha these hese concep ex ca zed (185), 185 aall oof wh partoof the conceptss are eeffectively ec ve y lexicalized even in n the which he languages n wh ch they not to o occur he correspond occur; the anguages in previously hey were prev ous y thought hough no corresponding ng wordsarepo wordsare not proper understood ac no oodby those believe hose who be eve in n 'primitive polysemous, properly ysemous a fact yunders hough pr m vethought'. FranzBoas, who cclaimed a med in n Polysemy postulated Po ysemy iss NOT pos u a edlightly, gh y though. hough W approv approvingly quotes ng yquo es FranzBoas 1938 that ha 'devices dev ces to o deve deas are probab ha they develop generalized op genera zedideas probably yaalways ways presen presentand [that] hey are cultural used as soon as the he cu ura needs compe he na natives ves to o form orm them' hem (210). ha 210 The argumen compel the argumentiss that no language abstract rac thought, not ac anguage iss incapable ncapab e oof abs hough even though hough language anguage users may no actually ua y use their he r language or such purpose purpose. anguage for main The ma n argumen n Ch Ch. 7 ('Semantic rolee oof os ostension he ro ens on in n the he Seman ccomp argumentadvancedin complexity ex y and the rs pub n 1991 ha 'the he comp 211-33; 33 first 1991) iss that acquisition acqu s on oof concep concepts', published s 211 shed in complexity ex y oof a concep concept can be vviewed ewed as the he d distance s ance separa rom the he level eve oof indefinables' Several types nde nab es (212). 212 Severa separating ngit from ypes oof concep n an aattempt hen looked ooked aat, in o assess their he r comp nc ude 'abstract' abs rac conceptss are then emp to complexity; ex y they hey include or emo concrete' e concep emotions), ons 'relatively conceptss (exemplified concep exemp ed by means oof words for re a ve yssimple mp e concre conceptss or body par natural he na ura env environment, ronmen temperature (exemplified exemp ed by means oof words for parts, partss oof the s par empera ure he concep natural kinds' cu ura and na ura k nds (such terms, erms and the re and 'cultural water). such as bread and wa er In concept oof fire), abstract ha abs rac concep ess comp han concre concretee ones (but ar more bu far general, it appears that genera conceptss are less complex ex than would han traditional rad ona lexicographers d have us be na sec section, on W argues believe). eve In a final complex comp ex than ex cographerswou that semanticcrepresen ha the he seman sentences must proceed on a sstep-by-step encesmus basiss (one one concep representation a onoof sen concept ep by s epbas aat a time). or sen sentences W'ss ences as ssimple paraphrase,even for me There iss no such thing h ng as a gglobal oba paraphrase mp e as W own examp not po out that not app ha her argumen canonical o canon ca examplee I aatee an app apple. point argumentdoes no apply e W does no n ou y to sentences sen ences that ha cons consist semanticc pr s oof seman coincide nc de w with h primitives only; sentences, presumably, m veson y such sen ences presumab y co their he r paraphrases(in which ch case it may be awkwardto o use the he term ermparaphrasea paraphrase at aall). n wh he above (Part Genera issues', her aattention urnsher en onto o Having Hav ng covered aall oof the Par II, 'General ssues 11-233), 233 W turns lexical semantics ex ca seman cs (Part Chs. 88-12) semantics o the he seman cs oof grammar(Part II, 235 235-375; III, Par II 375 Chs 12 and to Par III Chs. 13 Partss II and III are bes best seen as illustrations us ra onsoof the he me 377-458; 13-15). methodology put 377 458 Chs 15 Par hodo ogypu forward Part II. I wou orward in n Par would d have we welcomed comed the he oppor discuss od scuss many po detail a in n opportunity points un yto n s oof de Chs. 88-15. Chs detailed 15 However a edcommen commentss wou would d have requ review ew much longer hanwas was However, de required redaa rev onger than must limit m myse o a list s oof chap o owed by a br brief e synops possible. myself to chapter synopsis: poss b e I mus ertitles, es each followed s definitions"' n ons" (Ch. not indeterminate, so-called ca ed 'Against "against 237-57): Meaning nde erm na e so Aga ns "aga ns de Ch 88, 237 57 Mean ng iss no untranslatable un rans a ab econcep definable and do have a seman nab eand semanticc invariant. nvar an conceptss are fully u y de 'Semantics and lexicography' Seman csand semanticcpr can be used as an primitives 258-86): m vescan ex cography (Ch. Ch 99, 258 86 How seman o improve instrument ns rumen to practice. mprove lexicographical ex cograph ca prac ce 'The The mean colour our terms and the ermsand he un universals versa soof see here seeing' 10, 287 287-334): Although meaning ng (Ch. Ch 10 334 A houghthere ngoof co are un universals versa soof see are no un hereare universals versa soof co colour. our seeing, ng there 'The The seman semantics natural cs oof na ura k kinds' nds (Ch. discussion scuss on oof how one m best 11, 335 335-50): might Ch 11 50 A d gh bes between ween lexicon ex con and encyc between weenmean and knowledge-with distinguish encyclopedia, meaning d s ngu shbe oped a be ngandknow edge w h illustraus ra tions ons from rom the he vas vast areao area of na kind nd terms. erms natural ura k 'Semantics and eethnobiology' Seman csand he use oof linguistic es s and linguistic 12, 351 351-75): ngu s c hnob o ogy (Ch. Ch 12 75 On the ngu s c tests evidence ev dence in n the he sstudy n genera categorization general and eethnobiological categorization udy oof human ca egor za onin hnob o og ca ca egor za onin n particular. par cu ar 'Semantic Seman c ru rules es in n grammar grammar'(Ch. 13, 379 379-401): partss oof grammarare grammarare sub subjected Ch 13 401 Large par ec ed to o semanticcru rules seman es wh which ch are abso whilee sstill be absolutely predictive being not un uniu e ypred c vewh ng language-specific, anguage spec c i.e. e no versal. versa 'A A seman semanticcbas basiss for or gramma and typology: grammatical description Transitivity and reflexives' cadescr p onand ypo ogy Trans v yandre ex ves (Ch. Ch he same gramma abe s across languages 14, 402 402-26): grammatical only precise 14 26 The use oof the ca labels anguagesiss justified us edon y if prec se definitions areprov de n onsare which ch po od differences erencesas as we well as to o a language-independent provided point semanticc dedwh n to anguage ndependenseman core. core across languages: semantics csoof ev 'Comparing grammatical categories evidentials' 15, Compar nggramma ca ca egor esacross anguages The seman den a s (Ch. Ch 15 heore ca demons on emp work done by oothers, 427-58): demonstration, building empirical universal 427 58 A theoretical ra onbu d ngon r ca workdone hers that ha un versa semanticc pr seman useful u in n the he sstudy primitives evidentiality m ves are more use udy oof ev den a y than han technical echn ca labels abe s such as INDIRECTetc.) WITNESSED, INFERRED INFERRED,INDIRECTe WITNESSED c W'ss latest W a es workbearsa work bearsall the he ha hallmarks marksoof the he W Wierzbickian enterprise. The format-a collecerzb ck anen erpr seThe orma a co ec tion o...