Newmeyer, Frederick J.. 1999. Bridges between generative and cognitive linguistics. Book Article. Berlin; New York: Mouton de Gruyter. pp. 3-22
Notes
Many concepts have equivalents e.g. iconicityNot all generative linguistics presuppose model-theoretical semantics e.g. Chomsky or Jackendoff (example of autonomous syntax and nonobjectivist semantics going well together)GG and CG are not "notational variants" but are "considerably closer than is Generally thought"GL and CL are united in "cognitive commitmet" but often cite opposing research from other cognitive sciences which are "riddled with disputes"Even the autonomy of syntax may not be as contentious as it might appear since it does not entail a) there can be no commonalities and b) there is a high degree of arbitrariness between form and meaning. Also showing that there is motivation and relatedness does not proove that autonomy is false.Also suggests a modular treatment of Lakoff's analysis of perceptual deicticp. 14 Compares current state in CL to early transformational grammars: "Early generative grammarians were so enamoured of transformational rules that they typically ignored obvious nonsysntactic...explanations for unacceptability." Maybe in CL, a similar "over-appeal to cognition-based explanations" exists "with a downplaying of the role of principles grounded in discourse and communication"Removing distinction between grammar and lexicon more point of emphasis than theory. Similarly distinction between pragmatics and semantics has also been rejected by Jackendoff.Further convergence between GL and CL can be seen with the Minimalist Program where "all central grammatical principles apply directly on the surface or at the level where grammar interfaces with meaning; language particular variation is restricted to the lexicon. The de-emphasizing of the computational system in favor of the lexicon is a step in the direction of cognitive linguistics." (p. 15)His other two proposals of potential similarity need further investigation, although there is something to them:p. 4 "Generative and cognitive linguistics are both true, describing different aspects of cognitive functioning. This could be effected in two ways: 1. The child acquires the structures and principles of both theories 'side-by-side'. 2. Generative and cognitive linguistics describe different levels of the same reality."Option 1 has some support from developmental psychology while Option 2 is appealing but it is difficult to see how those levels would be defined since we cannot even agree on what that 'same reality' would be. Unlike, for instance, physics - where Newton and Einstein and Quantum Mechanics can be treated in the same way; (|Comparing Pinker and Aitchinson or Lakoff would show both the deep divide and similarities; also it seems that few on either side of the debate have taken up his call for mutual understanding - Jackendoff and Talmy being the exceptions - as well as many researchers dealing with particulars rather than theoretical problems)