so I have a Syntax II final Friday and am really confused about one of the study guide questions: "Why can't semantics be the input to Syntax? Illustrate with examples". Could anyone please shed some light on this, or direct me to a source? Thanks!

  • 4
    What I think they mean is that many syntactic rules ignore semantic impossibilities, such that you can make a syntactically legitimate sentence even while violating semantic rules, like Chomsky's colourless green ideas sleep furiously. // Note, however, that semantic rules do have some influence on some syntactic rules, such as when determining whether to use it or he or she to refer back to a noun in an earlier clause: London/Lauren is beautiful; you would like it/her.
    – Cerberus
    May 9 '13 at 2:40
  • Thanks, Cerberus! Anyone else have any other suggestions or places where I could go to get more information? May 9 '13 at 4:57

I find "semantics is the input to syntax" to be rather obscure in meaning, but perhaps it means that everything about the syntax of a phrase can be predicted from the semantics of that phrase. If so, all you need to do to show the thesis is false is to give some fact about the syntax of a phrase which is due to its meaning rather than tradition. That sounds pretty easy. Can you you predict whether a language has prepositions or postpositions from the meaning of pre-/post-positional phrases, for instance?


"Why can't semantics be the input to Syntax? Illustrate with examples".

Since the question is about "the input", I suppose the meaning is "the only input". Take e.g. word order. It's a syntactic phenomenon, and if semantics were the only input, there would be no reason to wind up will all possible permutations of {S,V,O} in the world's languages (as it is widely agreed upon that semantics is universal). The very existence of the permutations is a witness of some other factors besides semantics contributing to syntax. The permutations are contingent on something but if they were contingent only on semantics, word order would be the same in all languages

  • I don't follow this argument. Word order is the result of a language-specific essentially arbitrary rule sawing that in English, the verb is between the subject and object; in Arabic, the verb is before the subject and object; and so on. The question is not "can all of syntax be performed using only semantic properties", the question is about the "base". Syntax can use the semantic input to create non-semantic structural properties.
    – user6726
    Sep 16 '18 at 19:35
  • @user6726 I'm not sure whether the question is about "the main input" or "the only input" but "the input" suggests the latter interpretation. Furthermore, if it's about "the main input", then it's too vague and probably not worth answering ("the main" in what sense or for whom? quantitatively? how do you quantify this? and if semantics is not "the main input", then what is?)
    – jaam
    Sep 16 '18 at 20:49

"Everyone in the room knows at least two languages" is a sentence that is true in two quite different situations, one where everyone knows two languages, say English and French, and the other where for anyone you take you get two, possibly quite different languages, so for Jane you get French and Dutch, for John you get Dutch and German, for Peter you get Swahili and Malay and for Didi you get Greek and Arabic... This is sorted out in linguistic theory by disambiguating the phrase quoted in the beginning syntactically, so that for the first situation the DP " at least two languages" precedes and commands the subject DP "everyone" and for the other reading, the two DPs are in the order in which they appear at surface. Syntax must hence disambiguate scope before semantic interpretation can get started.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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