Academic journal article SKASE Journal of Theoretical Linguistics

The Non-Modular Nature of Cognitive Grammar

Academic journal article SKASE Journal of Theoretical Linguistics

The Non-Modular Nature of Cognitive Grammar

Article excerpt

1. Introduction

The non-modular linguistic program called cognitive grammar (CG), proposed and pursued by Ronald Langacker in the general theory of cognitive linguistics (cf. Langacker 1987, 1988a, 1988b, 1990, 1991a, 1991b, 1999, 2000, 2005, 2008),1 can be best characterized in opposition to Noam Chomsky's generative grammar (TG), a modular approach to language. According to Langacker (1987: 35), TG theory represents

   a conception of grammatical structure [which] emphasiz[es] discrete
   components [and which] naturally encourages the investigator to
   focus his attention on phenomena consistent with this type of
   organization. He concentrates primarily on prototypical instances
   from each component, where the distinctions seem readily apparent,
   and tends to overlook any data that do not fit neatly into the
   pre-established boxes. [...] A case in point is the putative
   distinction between syntax and lexicon. In the classic conception
   (now considerably modified), syntax was thought to deal with novel,
   multiword expressions (phrases, clauses, and sentences) assembled
   in accordance with general rules. Lexicon was the province of fixed
   expressions, most no larger than single words; not predictable by
   rules of any generality, they had to be listed individually. The
   two classes of phenomena thus stood sharply opposed with respect to
   novelty, generality and size.

The sharp opposition between lexicon and syntax has led to a situation where, as Langacker notes, "a large body of data fitting neither category would be mostly ignored" (ibid.). A case in point is idiomatic language, which includes various stock phrases, collocations and formulaic expressions. Because in each language one can find a huge number of such conventional expressions, "knowing them is essential to speaking it well" (ibid.). This is why

a seemingly perfect knowledge of the grammar of a language (in the narrow sense) does not guarantee fluency; learning its full complement of conventional expressions is probably by far the largest task involved in mastering it (ibid.).

Naturally, "the knowledge of conventional expressions" can hardly be wholly accounted for by means of rules, which are a sine qua non of the generalization postulate embraced by generative grammar. Yet, the generalization postulate itself should not be treated as a dogma, for, according to cognitivists, this postulate is based on the misguided belief--on the so-called rule-list fallacy. Langacker (1987: 41) comments:

   The reluctance of generative grammarians to concern themselves
   seriously with conventional expressions is largely inspired by
   their abhorrence of lists. [.] It would be fallacious, however, to
   invoke the principle of economy to argue that conventional
   expressions should not be listed in a grammar--one could just as
   well argue that phonology should be excluded from a linguistic
   description because a grammar containing a phonological component
   is more complex than a grammar without one.

Because, as Langacker observes, the principal goal of generative grammar was to achieve the economy of description and thus to capture "significant linguistic generalizations," it was hoped that, by capturing these generalizations, a particular statement could be eliminated from the grammar in favor of a much smaller number of rules. However, this line of thought, which illustrates the rule/list fallacy, is misguided because, in Langacker's view (p. 41),

   one is forced to choose between rules and lists: the options are
   posed as rules alone vs. lists alone. If these are the only
   options, it can be argued that the rules must be chosen, for lists
   by themselves do not express generalizations. There is in reality a
   third choice, however, both rules and lists.

In what follows we shall demonstrate how a non-modular approach to language structure such as cognitive grammar captures generalizations without falling into the trap of the rule-list fallacy. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.