Pragmatics & Cognition - Volume 8, Issue 2, 2000

The conditions for verbal irony comprehension implicitly or directly claimed as necessary by all of the recent philosophic, linguistic and psycholinguistic theories of verbal irony (Clark and Gerrig 1984; Kreuz and Glucksberg 1989; Kumon-Nakamura, Glucksberg and Brown 1995; Sperber and Wilson 1981, 1986) were experimentally tested. Allusion to a violation of expectations, predictions, desires, preferences, social norms, etc., was confirmed as a necessary condition, but pragmatic insincerity was not. Pragmatically sincere comments can be comprehended ironically. A revised set of conditions was proposed, involving intentional violation of Gricean conversational maxims and the portrayal of a contrast between expectations and reality. A cautionary note was made, however, regarding the viability of a single account of verbal irony comprehension.

Current research on the syntax-semantics interface demonstrates the dramatic extent to which syntactic structures constitute transparent reflections of well-defined semantic regularities. As this paper shows, the empirical results accumulated within this framework strongly suggest that a theoretical distinction should be made between two distinct levels of meaning representation: A level of conceptual meaning on the one hand, and a uniquely linguistic level of meaning — Linguistic Semantics — on the other. The semantic notions and regularities which turn out to determine major syntactic phenomena are best interpreted as belonging to the level of Linguistic Semantics, rather than to the level of conceptual meaning. This view helps characterize language as a unique and functional system — a cognitive system whose function is defined at the level of Linguistic Semantics. It explains the fact, most recently highlighted by Levinson (1997), that the expressive power of language, as a tool for the communication of meanings, is constrained in non-trivial ways.

In this study I examine some uses of connectives, and in particular co-ordinate conjunction, from a critical discourse perspective; these uses, in my view, cannot find a satisfactory explanation within current frameworks. It is suggested that we need to identify a conceptual level at which connectives function as hypo-textual signals, activating systematic law-like conditional statements (IF-THEN), which form default specifications of consistent structured knowledge frames. I argue that an account of connectives at the conceptual level of their function that does not take into consideration such tightly structured background schemata, representing both general knowledge and ideologies, cannot afford any generality. As a result, ‘deviant’ or ‘subversive’ uses of these connectives can neither be identified as such nor find an adequately general explication within existing accounts, whereas in the proposed framework such uses find a ready explanation of sufficient generality. This framework lies at the intersection of disciplines: Linguistic pragmatics (empirical pragmatics, critical discourse analysis), on the one hand, and cognitive science, on the other. Consequently, this proposal, too, can be regarded as a plea for crossing boundaries and joining forces.