Please see the following:

We start with a sentence/clause like -

Mr Wilkins is the oldest person in the village.

It seems like we can "transform" the clause using certain "grammatical rules":

Mr Wilkins is not the oldest person in the village. (Transformation: Insert the word "not" after the linking verb. Result/meaning achieved: Negation)

Is Mr Wilkins the oldest person in the village? (Transformation: Bring linking verb in front of the sentence, switch to question mark. Result/meaning achieved: Interrogative mood)

Mr Wilkins must be the oldest person in the village. (Transformation: Add modal verb "must". Result/meaning achieved: Speaker's certainty towards the statement.)

What's more, these rules can often be stated in very clear, unambiguous terms.

(1) Is there a formal term for this process in linguistics? (2a) Has anyone systematically studied the possibilities from this process, and more importantly, the meanings that might be achieved? (2b) Does the NSM have provisions for the sort of what can be said to be the "grammatical/structural meanings" that might be achieved from these transformations? I would envision such primitives as NEGATION, MUST-MODALITY, CAN-MODALITY, NUMBER-PLURAL etc. to be in this language. If no, has there been any effort to come up with such a metalanguage or anything similar. (3) Do studies on grammatical features deal with what I mentioned in (2a) and (2b)?

Looking for leads on the appropriate monographs, journal articles etc.

1 Answer 1


You are describing the theory of transformation grammar practiced up to some point in the 70's, as exemplified by Chomsky Aspects of the theory of syntax and Burt From deep to surface structure: An introduction to transformational syntax. The question of "meaning preservation" was prominent in that framework, where there was a fundamental dispute over whether semantic interpretation was determined based on the output of transformations, or whether semantic properties are fixed in deep structure. The Katz-Postal hypothesis (1964) holds that all semantic interpretation applies to deep-structure, before the application of transformations. As applied to negation, one possibility is that there is an optional transformation inserting "not" in the appropriate place, and the interpretation of the sentence is based on whether or not negation has been inserted. The alternative (a consequence of the Katz-Postal hypothesis) is that there is an element NEG optionally selected in deep structure (by the phrase structure rules), which is then present for (initial) semantic interpretation or not – syntactic positioning of NEG is obligatory.

There were a number of specific cases where it seemed that some transformation "changed" the meaning of sentences, but that interpretation depends on establishing exactly what is present in deep structure. The theories which maintain that transformations do not change meaning (abstract syntax / generative semantics) posited a richer initial structure, where e.g. "Floyd" broke the glass" would have 8 underlying clauses. The alternative, "interpretive semantics", was to have a closer relationship between surface strings and underlying strings, and more "rules of semantics". Interestingly, contemporary nano-syntax views bear a striking resemblance to the structures posited by generative semantics.

In other words, we've been there and done that.

This handout by Partee very briefly summarizes the history of syntax-semantics relations

  • 1
    The Partee handout is useful; for more on GS, Logical Structure, and its extension into syntactic structure, see also the Logic guide, the Verb Phrase guide, and a list of several hundred English transformations.
    – jlawler
    May 31, 2020 at 16:57
  • 2
    I had not seen Ross's list of transformations: it's quite impressive, and especially useful for getting a grasp on all of the names.
    – user6726
    May 31, 2020 at 17:18
  • This line seems to have typos: The theories which maintains that transformations do not change meaning (abstract syntax / generative semantics) posited a richer initial structure, where e.g. "Floyd" broke the glass" would have 8 underlying clauses". I can't make an edit of less than 6 characters, so I'm informing you in a comment. Jun 6, 2020 at 20:42

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