John Frow, Social Semiotics, 2001, 11, 3, 275-287.
Abstract: Since the beginning of this century, psychiatrists and linguists, assuming a correlation between disordered talk and disordered cognition, have sought to devise language tests with diagnostic efficiency for mental ‘illnesses’. Schizophrenia in particular has been assumed to be characterized by disorders of cohesion, of reference, and of symbolization. Much of this work is flawed by its a priori assumptions about the reality of the category of schizophrenia and about the relation between ‘normal’ (non-figurative, ‘logical’) and ‘deviant’ (figurative, ‘magical’) uses of language, as well as by particular methodological problems such as the failure to control for experimental context and for the effects of psychotropic drugs. Nevertheless, the debates within psychiatry and linguistics over communicative disorders have a good deal to tell us about the ‘normal’ uses of figurative language in social interaction. In particular, they raise complex questions about the metacommunicative functions of metaphor: How does figuratively coded language work to convey multiple simultaneous and sometimes contradictory messages? What kinds of discursive relations does it thereby establish or maintain or disrupt? How does it contribute to narrative cohesion, and are there tensions between figure and story? On what basis, if any, is it possible to distinguish between ‘appropriate’ and ‘inappropriate’ uses of metaphor?