Review Essay: Hubert Cuyckens, René Dirven, & John Taylor’s (2003) 'Cognitive Approaches to Lexical Semantics'

Description
Review Essay: Hubert Cuyckens, René Dirven, & John Taylor’s (2003) 'Cognitive Approaches to Lexical Semantics'

Please download to get full document.

View again

of 10
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information
Category:

Crafts

Publish on:

Views: 0 | Pages: 10

Extension: PDF | Download: 0

Share
Tags
Transcript
  Sweetser, Eve E.1988 Grammaticalization and semantic bleaching.  Proceedings of the 14th Annual Meeting of the Berkeley Linguistics Society , 389–404. Berkeley: BerkeleyLinguistics Society.Tomasello, Michael (ed.)1998  The New Psychology of Language: Cognitive and Functional Approachesto Language Structure , Volume 1. Mahwah, N.J.: Lawrence ErlbaumAssociates.2002  The New Psychology of Language: Cognitive and Functional Approachesto Language Structure , Volume 2. Mahwah, N.J.: Lawrence ErlbaumAssociates.Ullmann, Stephen1957 [1951]  The Principles of Semantics . Oxford: Basil Blackwell; Glasgow: Jackson,Sons and CO.Yarowsky, David1995 Unsupervised word-sense disambiguation rivalling supervised methods.  Pro-ceedings of the Annual Meeting of the Association of Computational Linguis-tics  (1995), 189–196. Hubert Cuyckens, Rene´ Dirven, and John R. Taylor (eds.),  CognitiveApproaches to Lexical Semantics . ( Cognitive Linguistics Research 23 .)Berlin / New York: Mouton de Gruyter, 2003. xiii þ 504 pp., ISBN 3-11-017709-9.Reviewed by Cory D. Wright, Departments of Philosophy & Cogni-tive Science, University of California, San Diego, U.S.A. E-mail:  3 cory@mechanism.ucsd.edu 4 Cognitive Approaches to Lexical Semantics  is introduced by John Taylor,Hubert Cuyckens, and Rene´ Dirven (eds.) in their intriguing article ‘Newdirections in cognitive lexical semantic research’. One of their main pointsis that the 13 research reports to follow register a sharp break from tradi-tional assumptions in both cognitive lexical semantics, and in cognitivelinguistics more generally. While the editors are right that the constituentarticles jointly reveal that certain issues (e.g., polysemy, prototypicality)are less settled than commonly supposed, it is slightly misleading to ad-vertise them as ‘new directions’. Not only are many of the reports not ter-ribly new (roughly half srcinated from presentations at a 1996 workshopby the same name at the 16th Scandinavian Linguistics Conference inTurku, Finland, others were presented elsewhere, some are revised disser-tation chapters or extracts thereof, and some draw heavily on or other-wise recycle parts of that author’s previous work), but it is also di‰cult570  Book reviews Cognitive Linguistics 18–4 (2007)  to locate any such radical departure from now-traditional assumptions(indeed, most papers conform to the well-worn methods and types of rea-soning that have been familiar to cognitive linguists for a quarter centurynow; in particular, many draw heavily on, and attempt to work out, ideasfrom Langacker’s  Foundations of Cognitive Grammar ). As an example,the editors encourage the abandonment of certain ways of thinking aboutword meaning by arguing that many problems in the field of lexicalsemantics can be traced back to two, ostensibly inept metaphors. The firstis the conduit metaphor: words are containers for meaning, where mean-ings are objects with fixed and determinate properties that are coded anddecoded by the linguistic expressions of speakers and hearers such thatcommunication is the transfer of thoughts from one mind to another.The conduit metaphor is certainly dubious, and it is a struggle to beunsympathetic with the editors’ criticism here; nevertheless such criticismhardly constitutes a ‘new direction’ (Reddy 1979).While inept metaphors such as this one can plainly have pernicious ef-fects, there are plenty of ‘‘enlightened’’ theorists working today who arenot seduced into false or misleading understandings of the nature of ver-bal communication by the conduit metaphor. Wherefore the problem,then, in bracketing the conduit metaphor as being theoretically benign— say, like taking the sun to rise or time to be a flowing river? Such meta-phors are problematic only when their problematic entailments are takentoo seriously. As such, the second, building-block metaphor of syntag-matic composition is a much more interesting and subtle target of criti-cism, especially insofar as it deeply influences actual theoretical principlespertaining to semantic compositionality. (Of course, criticism of this met-aphor goes back even further than the first.)The building-block metaphor suggests that complex linguistic expres-sions are constructed by assembling smaller units according to certainprinciples. According to the editors, the conduit and building-block meta-phors jointly endow what they call the  standard view , which is the viewthat words have fixed and determinate meanings that combine to contrib-ute to the meanings of the complex expressions in which they occur suchthat a language user can ascertain the meaning of the whole from themeaning of the parts. Taylor et al. disparage the suppositions compris-ing it: that the component parts of complex expressions have fixed anddeterminate meanings which make fixed and determinate contributionsto the meaning of complex expressions when combined; that the seman-tic properties of those parts are fully maintained in those complex ex-pressions; and that there is no surplus meaning accruing to a complexexpression which is not attributable to its parts and the manner of theircombination. Book reviews Cognitive Linguistics 18–4 (2007)  571  Plainly, whatever problems the standard view has, those problems arenot about compositionality per se (cf., even cognitive linguists invokecomponent / composite relations); rather, the problems of the standardview derive from additional commitments to semantics being fully com-positional and to an emphasis on semantic fixity and determinateness.But then, what do such additional commitments and suppositions—falsethough they may be—have to do with the building-block metaphor perse, which—pace Taylor et al.—merely concerns composition simpliciter?That is, why think that the building-block metaphor, by itself, entailsthat semantics must be fully compositional or logically entails strongviews about semantic fixity and determinateness? So, despite being sym-pathetic to their criticism of the standard view, I am uncertain that theeditors have yet made a decisive case against the second metaphor, sinceit is unclear why the two are necessarily beholden to each other. (At thevery least, more discussion would have helped justify why these furthersuppositions are through-and-through metaphorical; but again, it seemsthat only if the metaphor is antecedently defined in terms of theseproblematic suppositions does the metaphor itself become theoreticallyproblematic.)Unlike many anthologies, the 13 research reports constituting  CognitiveApproaches to Lexical Semantics  do exhibit a reasonable amount of thematic convergence. As alluded to, two notable themes pertain to the(dis)advantages of monosemy / polysemy approaches to lexical semantics,and to the prevalence of prototypicality and its e¤ects on lexical catego-ries. The now-traditional assumption in cognitive linguistics that poly-semy is the norm is challenged in several reports—e.g., Theo Janssen’s‘Monosemy versus polysemy’ and Jordan Zlatev’s ‘Polysemy or general-ity? Mu’, and particularly in Jens Allwood’s ‘Meaning potentials and con-text’. Janssen and Zlatev argue that these debates about cardinalities of lexical sense are founded on a false dichotomy. Janssen begins by raisingmany di‰cult questions about the relations between concepts and lexicalstructure (some readers will likely be frustrated by the lack of answers,though) on his way toward advocating a solution of mutual consistency.For Janssen, monosemous approaches can account for lexical senses as‘undivided, highly abstract meanings’ whereupon they are supplementedwith pragmatic principles for attributing extensive creativity to speakersas they solve problems of linguistic coding in the context of diverse usageevents. Supporting examples come from the analysis of possessive locu-tions. For his part, Zlatev claims that the answer to monosemy / polysemyis ‘Mu’. One is led to infer that it somehow means that the debates them-selves are confused, but Zlatev doesn’t tip his hand in alerting readers towhat he had in mind with this black-box of a concept. Zlatev alludes to572  Book reviews Cognitive Linguistics 18–4 (2007)  an analysis of lexical meaning in terms of use-potential, and attempts toally this perspective with connectionist models of polysemous lexicalorganization; but the most fascinating aspect of the report is the quirkymetaphysical framework developed around what Zlatev calls  minimal dif- ferentiated language games . These might be described as representationsof relations between sentence- and situation-types situated against sets of background practices, though it is di‰cult to be certain since they arebasically an amalgam of theoretical elements drawn from Merleau-Ponty,Wittgenstein, Searle, Dreyfus, Langacker, Talmy, and many others.Jens Allwood’s article, ‘Meaning potentials and context’, advances arobustly monosemous perspective on the nature of word meaning interms of what he calls  meaning potential  . For Allwood, a word’s meaningpotential is the sum total of all information that has ever been conveyedin using it relative to a set of speakers, or in his words, the ‘union of indi-vidually or collectively remembered uses’ (p. 43). Combinations of suchinformation result in determinables activated by lexical items under con-textually determining conditions of use and convention (e.g.,  morphology has a single meaning potential which is di¤erentially activated dependingon whether the context is biology or linguistics). It’s not entirely unfairto say that Allwood’s program is to cognitive linguistics what Cappelenand LePore’s (2004) semantic minimalism is to philosophy of language.Allwood’s contribution is stimulating and certainly seems to solve theproblem of specifying the number of senses a word may have (i.e., anygiven word has exactly one meaning potential). I write ‘seems’ becausewhether polysemy exists and to what extent becomes a mere terminologi-cal matter between theoretical frameworks, if I understand him correctly.For Allwood, contexts create the conditions that determine how any givenpotential is activated, such that di¤erent ways of activating meaningpotentials (i.e.,  determinations ) yield di¤erent answers to cardinality ques-tions of lexical sense. This solution is claimed to avoid the (putative)problem of reified polysemy (whether for decontextualized-types or forcontextualized-tokens—presumably, because words no longer are takento have multiple, related senses!); but, it really only presents theoristswith a repackaged phenomenon. For rather than di¤erent senses in di¤er-ent contexts of use, we seem to now just have di¤erent determinations of meaning potential in di¤erent contexts of use.In some sense, Allwood’s concept  meaning potential  calls to mindLangacker’s concept  base  (i.e., the conceptual content evoked by a lexicalexpression). However, the main drawback of the former seems to bethat Allwoodian meaning potentials don’t constitute a practically viablesolution. For if lexical senses are mapped one-to-one into monosemousmeaning potentials as just characterized, then specifying the meaning of  Book reviews Cognitive Linguistics 18–4 (2007)  573  any given word is just to specify the union of all information which hasever been conveyed in using it; and since no language user or linguist hasaccess to all such information, the meaning potential of an expression canat most be approximated but never actually be specified. Hence, postulat-ing meaning potential solves the problem of underspecificity of meaningsimply by overspecifying meaning potential. And since it is unclear howAllwood intends to fix the identity conditions of meaning-potentials(reductions of high-dimensional state space vectors?), one may wonderwhether this construct is theoretically unmanageable and unwieldy.In ‘The Nawatl verb stem  kı¯ sa : A case study in polysemy’, DavidTuggy presents a nuanced case study of polysemy using  kı¯ sa  (‘emerge’)in Orizaba Nahuatl, which e¤ectively challenges the monosemous inclina-tions of Allwood, Janssen, and Zlatev. In a similar case study, ‘Imageschemas and category coherence’, Augusto Soares da Silva identifies sev-enteen distinct senses of   deixar  in Portuguese, which he catalogues undertwo groupings—  deixar 1  ‘to leave’ and  deixar 2  ‘to let’—and then fur-ther according to e.g., variable construal of the landmark entity and theactivity / passivity of participant. Prototypical uses of   kı¯ sa  involve theconceptualization of something moving from an enclosure or sector (e.g.,  goes out of the house ), as do those of   deixar 1  to some extent (e.g.,  Thierryleft the program ,  the wound left a scar ), and are schematically diagrammedin figure 1. Among the scores of additional senses of   kı¯ sa  astutely notedby Tuggy are those where the landmark from which the trajector movesout of may be construed vaguely delineated (e.g.,  comes from the desert , emerging from the darkness ); those where the trajector and landmark re-main indistinct (e.g.,  a scar appeared from the wound  ); metaphorical uses(e.g.,  emerged from illness ); those where material is hewn or transformed(e.g.,  a great sculpture emerged from the clay ); event or character pro- Figure 1.  Prototypical sense of   kı¯sa  in Nahuatl  574  Book reviews Cognitive Linguistics 18–4 (2007)
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks