Aspects of the Theory of Syntax

The book's central claim is that linguistics should aim to construct a 'generative grammar'—a perfectly explicit system of rules that generates all and only the grammatical sentences of a language—as a model of the speaker's tacit competence rather than a description of actual speech behavior. This explicitly rejects the taxonomic structuralist program of Zellig Harris and Leonard Bloomfield, which treated linguistic analysis as a discovery procedure for classifying observable utterances. Chomsky argues that a focus on performance data misses the central fact that speakers possess abstract, unconscious knowledge of rules they have never been taught and cannot articulate, and that only a competence-based theory can explain the creative aspect of language use.

Aspects introduced the competence/performance distinction and the architecture of the 'Standard Theory' model of generative grammar. Competence refers to the speaker-hearer's tacit knowledge of language; performance refers to the actual use of language in concrete situations, subject to memory limitations, distractions, and errors. The Standard Theory held that a grammar consists of three components: a syntactic component generating deep structures, a semantic component interpreting deep structures, and a phonological component interpreting surface structures derived from deep structures by transformations. This tripartite architecture, particularly the claim that deep structure is the sole input to semantic interpretation, became the target of the 'linguistics wars' of the late 1960s and 1970s.

Syntactic Structures (1957) focused narrowly on syntax and treated transformations as the primary source of recursive power, with only a peripheral treatment of semantics. Aspects expands the theory into a general framework encompassing phonology and semantics, assigns recursiveness to the phrase-structure rules of the base component, and introduces the Katz-Postal hypothesis that transformations do not change meaning. The most important departure is the book's explicit turn toward language acquisition and universal grammar: Chomsky argues that the theory must be constrained by learnability considerations, particularly the poverty of the stimulus, a concern largely absent from the earlier work. The 1965 book thus represents the first major revision of the research program, which Chomsky later called the Standard Theory.

The book was immediately influential but also generated significant dissent. The claim that deep structure alone determines semantic interpretation was challenged by generative semanticists such as George Lakoff, Paul Postal, and John Robert Ross, who argued that transformations can change meaning and that semantic representation is the deeper level of grammar. This dispute, often called the 'linguistics wars,' led to the development of two competing frameworks: generative semantics and Chomsky's revised Extended Standard Theory. The evaluation metric for choosing among grammars, which Chomsky proposed in Aspects, was criticized as empirically untestable and eventually abandoned. Even sympathetic reviewers noted that the abstract formalism was applied almost exclusively to English, raising questions about the universality of the proposed principles.

The competence/performance distinction remains foundational to most generative work in linguistics and has been widely adopted in cognitive science, even by researchers who reject Chomsky's specific syntactic proposals. However, the Standard Theory's tripartite architecture and the deep structure/surface structure distinction were substantially revised, first by the Extended Standard Theory, then by Government and Binding, and eventually by the Minimalist Program. The poverty of the stimulus argument continues to be debated, with some researchers pointing to statistical learning mechanisms and rich environmental input as partial explanations for acquisition. The book's emphasis on mentalism and the cognitive turn it helped catalyze in linguistics have proven more durable than its specific formal apparatus.

A frequent misconception is that Aspects claims linguistics should ignore actual language use entirely. In fact, Chomsky treats performance data as essential evidence—it is the primary source from which the linguist infers competence—but insists that the data must be idealized to filter out memory lapses, slips of the tongue, and other performance factors. Another misreading treats the evaluation metric as a psychological theory of how children actually select grammars, whereas Chomsky presents it as a formal device for comparing the simplicity of competing grammatical theories. The book is also sometimes read as a fully worked-out theory of Universal Grammar, when it is better understood as proposing a research program and specifying the constraints any adequate theory must meet.