Talking heads: Language, metalanguage, and the semiotics of subjectivity By Benjamin Lee (review)

186LANGUAGE, VOLUME 76, NUMBER 1 (2000) Van Schooneveld, Cornelis H. 1977. The place ofgender in the semantic structure ofthe Russian language. Scando-Slavica 23.129-38. Waugh, Linda, and Stephen Rudy (eds.) 1991 . New vistas in grammar: Invariance versus variation. Amsterdam : Benjamins. Zubin, David. 1977. The semantic basis of case alternation in German. Studies in language variation, ed. by Robert Fasold and Roger Shuy, 88-99, Washington DC: Georgetown University Press. -----. 1979. Discourse function in morphology. In Givón, 169-20. Department of Foreign Literatures Ben-Gurion University PO Box 653 84 105 Be'er Sheva Israel [yishai@bhumail.bgu.ac.il] Talking heads: Language, metalanguage, and the semiotics of subjectivity. By Benjamin Lee. Durham, NC: Duke University Press, 1997. Pp. ix, 376. Reviewed by William Frawley, University of Delaware Benjamin Lee's terrific book, Talking heads, mostly concerns the tension between speech action (talking) and abstract, detached mind (heads) (see Wertsch 1998). But it is also about 'talking heads' in the popularized sense: those interchangeable speaking-minds who flit from one marketed crisis to another. Here at the turn of the millennium, the idealized unified self of rational tradition runs smack into the commodified, 'mass-mediated' (345) self of infinite choices and lifestyle. L's main argument is that subjectivity is made out of performativity (speaking and thinking as conventionalized actions), indexicality (reference to properties of context), and metaindexicality (reference to indexicals). In Ch. 1, 'The foundations of performativity: Austin and Frege', he traces the influence of these notions to Frege and Austin, two unlikely bedfellows at first glance. What does a logician convinced of the ideality of thoughts have in common with an ordinarylanguage philosopher who believed in meaning as contextualized action? In splitting truth from its mode of presentation, Frege opened the way for the differentiation of content from context and hence for Austin, for whom the mode of truth was essential to meaning itself. Austin thus needed Frege (and translated him into English!) to propose truth as only one form of the assessment of statements. Performativity, indexicality, and metaindexicality feed a number of other strains in modem theorizing about meaning and subjectivity. L spends the next three chapters discussing the applicability of these ideas to Derrida, the Russell-Donnellan-Kripke-Putnam disputes on reference, and Peirce. These are unavoidable choices in some way—after all, who else is there?—but some unusual lessons fall out of the exposition. For example, in Ch. 2, 'Deconstructing performativity', L uses some of Davidson's (1986) ideas to sketch out a genuine alternative to the proposal that we carry around no conventions in our heads (Derrida) or lots of them (Searle). Shared meanings require only passingtheories—on site constructions ofregularities triggered by the metalinguistic properties of speech acts themselves. Communication occurs when passing theories coincide. This is a very interesting idea, especially with the increasing pressure of movements that strip inner minds of their richness and offload the regularities into the environment (see, e.g. Clark 1997). Passing theories cued by the indexical and metaindexical features of speech performance might be a way to reconcile abstract intemalism with an informationally rich environment. These ideas drive a reconsideration of the great proper name/natural kind debates. In Ch. 3, 'Reconstructing performativity', L stresses the influence (and sometimes the disregard) of indexicality in accounts ofthe determination of reference. He comes out as a kind ofperformative Putnamite, casting naming on a scale of indexicality and metaindexicality: from deictics, where properties of the speech event are maximized, to natural kinds, where they are minimized (SiIverstein 1981). In Ch. 4, 'Peirce's semiotic', he underscores Peirce's unwillingness to erect REVIEWS187 barriers between logic and indexicality. Peirce rums out to be a proto-dynamicist: signs are inference triggers. Thus while Peirce and Frege both made important discoveries about that darling of static logical analysis, quantification, Peirce viewed quantifiers much the way Blakemore (1987) views pragmatic and metapragmatic markers generally—as instructions to users and hearers. Interestingly enough, in logical semantics, quantifiers are also instructions, but disembodied, computational operations—functions. This makes for two kinds of dynamic meaning: one where speakers and hearers matter (talking) and one where they don't (heads). The following four chapters...