No one seems to doubt that language, both in its communicative functions, and in its systems of semantic classification, plays a major role in identity development. I have mainly spoken so far about the communicative-interactional function, but what is most specific about the contribution of language arises from its categorial semantics. Natural language construes relations among types and between types and instances of types. It makes meaning by creating differences of kind, far more than by creating differences of degree. Yet we also know that material phenomena of all kinds differ significantly by degree, and that differences of degree may become resources for the emergence of new kinds or types (and conversely, averaging over insignificant differences of kind at larger scales where those differences are less relevant allows average quantitative variation to become the significant bearer of information; for examples see Lemke 2000a). Mathematics begins and develops over much of its history as that very extension of the semantics of natural language toward more precise description of differences of degree (Lemke 1999, in press-a).
The semantics of natural language is complemented by resources for meaning-by-degree such as vocal intensity and coloring, bodily and facial gesture and posture, the pacing of speech and activity, and drawing and visual representations of many kinds, with their resources for variations of degree in color, shape, shading, etc. Identities may be constituted and expressed as much in how we draw as in what we write, as much by our movement and vocal styles as by the semantic content of our talk. But natural language semantics lends itself far better to categorical distinction than to subtleties of degree. It is language that allows us to dichotomously contrast male and female, straight and gay, child and adult, black and white, Dutch and Deutsch, when we really know better that none of these are singular categories; all are constructed from variation along many potentially independent dimensions, and most variation along each of these is far more nearly quasi-continuous than categorial. Logocentric cultures and modes of academic analysis bias us to classify identities too categorially, to marginalize the normal hybridity of identities across our artificial categories, and to pay too much attention to the verbal-typological construction and performance of identity, and not enough to the integral nonverbal-continuous modes by which we make meaning and meaningful identity.
Do we construct different identities for ourselves when speaking different languages or different dialects of ‘the same’ language? In the ways I have defined identity we surely do, and the moreso as we consistently distribute our use of language varieties among different activity types and typical settings (cf. code-switching, diglossia). The more also as in speaking we also perform differences in styles of vocalization, in speech rhythms and intonation patterns, and in facial, gestural and movement styles. Identity differences will also be greater insofar as we come to speak different registers and use different genres in the two language varieties, as we come to use different registers of politeness and constructions that index and constitute different modes of interpersonal relationships (formality, intimacy, aggression, conciliation, affiliation, empathy), and as we come to enact different social roles through the varieties and take up different social statuses and positions. At the same time, given the patterns of redundancy that make culture, we are likely to also express different beliefs, attitudes, and values, even if only to the degree that such matters are untranslatable between the semantic systems of different languages and dialects.
At the later stages of socialization, particularly in academic contexts, we use language to take up professional or pre-professional identities. There are substantial differences in values, attitudes, and semantic orientations (Hasan 1996), indexed and constituted in part through language use, not only between differently gendered speakers, or those with different class habitus, but also among historians, literary critics, biologists, and mathematicians (not to mention lawyers, librarians, and administrators!). Professional subcommunities regulate the range and types of identities that are considered acceptable or ‘promising’ and may well do so with evident class and gender biases, but also with subtler distinctions that make them seem more compatible with acceptable identities in various ethnic cultural traditions as well. Much of this, in logocentric academia, is regulated by specifying the pragmatics of register use. It is not enough to master the vocabulary, the preferred grammatical constructions, the semantic systems of classification and collocation; you have to learn how and when to make which kinds of jokes, how to mix the formal register with informal talk, to appear to ‘think’ and ‘come from’ the preferred attitudinal stance of physicists or economists or computer programmers.
It is a wonderful mystery how language socialization and specialist enculturation take place in their normal contexts. But it is a more poignant and humanly crucial question how they so often fail to occur in educational settings. Modern democratic systems of mass education seek, in their own eyes, to make the opportunities of middle-class lifestyles available to all. More cynically, we may say that they seek to produce as many well-trained workers for middle-class occupations as the employment economy needs, and that unwittingly (for the most part) they also serve to justify the lesser life chances of all the rest as the result of individual lack of talent or willingness to study. A few supporters of critical educational models seek to empower all students to understand social inequities and the belief and attitude systems which sustain them. On a larger scale, however, the very structural artifacts of our educational system may well serve to maximize the disparity in achievement between those who come to school most ready to learn what and how the school teaches, and all the rest.