I have a good grasp on the idea of a generative component---clearly syntax is generative in this sense. But what is an "interpretative component" supposed to mean? Like, when a line is drawn between morphological theories that are interpretive and those that aren't, how would these differ in practice? Concrete examples are welcome.

  • 2
    Perhaps you should cite the source in which you saw this term. The source might clarify your question. May 29, 2023 at 12:04
  • It doesn't mean much any more. It was a name for early Chomskyan syntax and hasn't been used for 40 years or so. Think of it as early pseudocode for generative syntax.
    – jlawler
    May 30, 2023 at 20:36

1 Answer 1


The distinction is made in mainstream generative grammar, whereby only syntax "generates" and everything else "interprets". There are alternative theories, most notably Generative Semantics – it would take us too far afield to try to explain the difference.

In GG (Aspects version), the start symbol is S, then various other rules build (generate) full sentence structure. The insertion of lexical items into this frame is a bit problematic, but it does seem to just be another sub-type of structure generating. After the sentence is generated, it still needs to be "interpreted", that is turned into something else, actually two somethings else. First, you have to run that structure (now called "the underlying form") through phonological rules that ultimately control various muscles whereby sound comes out of the speaker's mouth. Second, it has to be converted into some other kind of representation, a semantic representation, which has some vague relation to "thought".

The basic division is between generating i.e. initial structure building, and converting or interpreting an existing structure. This completely leaves out morphology, because originally there was no separate component of morphology (and even today, there isn't, for some people). It also reflects just one view of language production.

The notion of semantics as an interpretive component and syntax as a generative one was turned on its head in Generative Semantics, which started with semantics as the basic generative engine. (Side note, jlawler's view on this will be better-informed than mine). Rather than going from an abstract structure of syntactic relations and figuring out a meaning to go with it, you can start with an intended meaning and find a syntactic way to express it. Slowly emerging from that debate was the view that rather that focusing no one thing building and another thing interpreting, you might build semantics, syntax and morphology separately, then somehow connect them so that a given generated syntactic structure might mesh with a particular morphological structure, and so on. The theory of HPSG (a competitor of GG) prominently leveled the playing field between semantics, syntax and morphology.

The original view stems from an underlying theory of rewrite rules, where "S" is rewritten as "NP VP" which we notate as "S→NP VP". There are no such rules in current Minimalist syntax, so the entire meaning of "generate" has changed. Instead (and if one were to write specific rules), two thing "merge" into one thing thus "NP VP→S". In current linguistic practice, the distinction between generative and interpretive is elusive to the point of being largely non-existent as a cross-theoretic meaningful concept.

Interpretive vs. generative morphology was a viable debate in GG for a period. This paper is an example of a that discussion, comparing a generative "words are atomic and built first" view vs. an interpretive "syntax comes first, morphology falls in line" view.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.