Rejecting Chomskian grammar for more meaningful & useful models

complex network of nodes

Haga clic aquí para la traducción al español (Google Translate)

Since the 1950s, linguistic theory has been dominated by the Chomskian model of grammar, in which syntax is treated as an autonomous system & the sentence is taken as the primary unit of analysis. While this abstraction proved useful for formalising aspects of grammatical structure, it has also imposed arbitrary constraints on how we understand language & language learning; larger & inseparable dimensions of meaning, context, & discourse have largely been relegated to separate domains, leaving pragmatics as a problematic “add-on” rather than a key feature of language, thereby obfuscating our understanding of language development & competence.

In contrast, Construction Grammar (CxG), rather than treating syntax as independent of meaning, views language as a network of conventionalised form-meaning pairings, AKA constructions, at every level of analysis, from morphemes & words to phrases, sentences, & extended discourse. This provides a unified account of language and language development, in which learners acquire increasingly larger & more complex constructions rather than abstract syntactic rules detached from meaningful, pragmatic communicative function. Although expressed & organised in different terms, CxG is also wholly compatible with Systemic Functional Linguistics (SFL).

The cognitive dimension of this model is equally important. CxG relates linguistic forms to semantic frames; conventionalised, structured patterns of meaning that speakers use to interpret & construct experience. For example, concessive expressions such as althougheven though, instantiate productive form-meaning schemas associated with the broader semantic frame of concession, making them both systematically describable & pedagogically accessible.

This broader perspective also aligns with research in English for Academic Purposes, particularly focusing on genre-based instruction (GBI), informed by SFL. Functional linguists such as John Swales & Ken Hyland have shown that academic meaning is organised through rhetorical moves & discourse patterns that frequently extend across entire paragraphs; a paragraph is not merely a collection of independent sentences but a coherent unit of reasoning whose meaning emerges from the relationships between its constituent parts. Restricting grammatical analysis to isolated sentences therefore obscures much of what writers & readers actually do.

Rejecting the arbitrary constraints inherited from Chomskian grammar & looking instead to cognitive & functional models of language, i.e. CxG & GBI, allows language to be understood & taught as an integrated system of meaning-making operating across multiple levels of analysis. Such a framework is not only theoretically more coherent & cohesive but also more useful for researchers, more teachable for educators, & ultimately more accessible for students learning to write effectively in academic contexts.