Nominal predicates in copular sentences are peculiar because in certain languages, they acquire case other than accusative case. Even English was so, e.g. "it is I (NOM)" vs. "it is me (ACC)." Perhaps this is because there is no actual verb that could assign accusative case to them.

How do post-copular DPs receive case if there is no actual verb? I've heard that the pre-copular DP gets theta-marked by the post-copular one but I'm not sure where I got that from.

Are nominal/adjectival/prepositional predicates considered arguments?

Could anyone direct me to literature related to theta-marking and case-marking in copular sentences?

  • "Post-copular" DPs are predicative so yes, they govern "pre-copular" DPs. But note that in the case of copular inversion, the pre/post terminology makes no sense. Overall, one of the DPs serves as a predicate which subcategorises for a subject.
    – Atamiri
    Commented Jul 18, 2017 at 10:22
  • @Atamiri I wanted to call them internal/external arguments, but I am unsure of whether they are arguments of the copula. I totally forgot about copular inversion. Commented Jul 19, 2017 at 9:59
  • I only wanted to point it out in order to avoid confusion. The “complement” isn’t generally considered to be an argument. In fact it’s not a complement. In most (lexicalist) syntactic theories as well as in pragmatics the noun or adjective (together with the copula if there’s one, note that in some language there’s no copula) is a predicate assigning case (though then again, there’s the case of nonsubject copular agreement).
    – Atamiri
    Commented Jul 19, 2017 at 11:09

1 Answer 1


Post-copular DPs do not need Case because they are not arguments of copular be. Such DPs, as Atamiri says, are (one-place) 'predicates' or constituents of 'predicates' (see Bowers' alternative analysis infra), their only argument is, indeed, theta-marked internally to the DP, at some stage occupies the Spec DP position, and, eventually, raises into Spec T/Infl and becomes the subject of the copular sentence.

Note, however, that this applies only to cases of copular be, not to those in which be expresses identity (i.e., 'A is B' : 'A = B') as in That tall woman is John's wife / John's wife is that tall woman, where be is a two-place verb with two 'reversible' arguments, one in complement position and the other - under the VP-Internal Subject Hypothesis - in Spec VP (= Spec be). Whichever of the two DPs functions as complement of 'identificational' be does need Case, of course, but that causes no problem, because dyadic be is transitive-like and assigns (accusative?) Case to its internal argument.

Copular be, on the contrary, selects just a non-verbal one-place predicate as its complement, i.e., minimally a 'predicative' DP in the cases you are interested in. The hedge 'minimally' is not gratuitous, though, because there is fairly compelling syntactic evidence that the complements of 'copular be' are not directly 'predicative' DPs, but Predication Phrases headed by a 'functional' Pred head that selects a DP as its complement (i.e., structures like [__ Pred [DP]]) and assigns Case to it. (See John Bowers' 'The Syntax of Predication', Linguistic Inquiry 24/4 (1993), pp. 591-656, or his more recent summary 'Predication', in The Handbook of Contemporary Syntactic Theory, Blackwell 2001, pp. 298-333 for classic statements of this alternative view).

  • What evidence is there to support the claim that equative 'be' is a 2-place verb? Could you direct me to literature on PredP and its ability to assign case to its DP complement? Commented Jul 19, 2017 at 9:37
  • The evidence that 'equative be' is dyadic is basically semantic: if it is semantically equivalent to the identity function '=', which obviously takes two functionally interchageable arguments (a = b) = (b =a), it cannot but be itself dyadic. As to whether the DP complement of Pred needs Case, above I have assumed that it does, and that such DPs are not inherently predicative and only acquire that character as complements of Pred, but that might not be so, and, in that case, they would not need Case at all. Read Bowers' and Edwin Williams' early papers on predication for further detail.
    – user6814
    Commented Jul 19, 2017 at 14:56

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.