Gert Webelhuth (webelhut@gibbs.oit.unc.edu)
Tue, 18 Jul 1995 10:44:11 -0400 (EDT)
On Tue, 18 Jul 1995, Kai Lebeth wrote: > maybe I was a bit too lazy in explaining, why I still think, that > in a construction like: > > [Ein Held zu sein] macht Spass > a hero to be is fun > > the Dp "ein Held" is just an object in a (identity) > copula-construction, and therefore receives nominative, like every > other copula's object. To make this clear, assume the following > subcategorization frame of the identity copula. (Maybe it is easier to > assume that the identity copula is not a raising verb) > > HEAD: ( infin & > SUBJ<[[1]-index] & > PROM_ARG : [1]-index) & > SC<DP[[1]-index]> > > > As I mentioned in my first mail, I assume a very simple case-principle > for German, which says that every DP with a structural-case that > stands in an agreement-relation with a finite verb, gets nominative, > while DPs that do not stand in an agreement-relation to a finite verb > and have structural case receive accusative. The agreement-relation > can be represented by putting the respective INST of the verbs > argument under a feature like, e.g, PROM_ARG (most prominent > argument). In the example above I assume that the copula's subject is > the PROM_ARG. As the copula's semantics always coindex the index of > its subject and object, the object (ein Held) is a PROM_ARG of the > verb-projection, too and therefore receives nominative. > I don't understand how your theory is supposed to get the right result for [[Ein Held zu sein] macht Spass]: since [sein] is non-finite, it will not assign nomininate to its subject and hence also not nominative to its "object." Then, unless there is a principle or rule that assigns nominative to the subjects of infinitives that are not structure-shared with accusative objects of a higher verb, how do you guarantee that the "object" of the copular in these constructions appears in nominative case, since the accusative is ungrammatical: *[[Einen Held zu sein] macht Spass]. Perhaps I am missing something, but I think your theory needs such an additional principle as much as Jeff's theory, or any other for that matter. That the solution to this problem cannot just lie in the case that the copular assigns to its "object" can also be shown with the German "als" construction. In many instances, the complement of "als" must agree in case with its logical subject: Als Vorsitzender/*Vorsitzenden darf er das Kommittee ernennen Ich habe ihm als Vorsitzendem/*Vorsitzender im Namen aller Anwesenden zu seiner Wiederwahl gratuliert Note that if such an "als" phrase appears in a non-ECM infinitive, then the complement of "als" must appear in the nominative if its antecedent is the logical subject: Er hatte geplant, als Vorsitzender/*Vorsitzenden das Kommittee zu ernennen Ich bat ihn, also Vorsitzender/*Vorsitzenden die Sitzung zu eroeffnen The ECM case pattern differently again. Here accusative is easily possible: Sie lassen ihn als Vorsitzenden immer so lange reden wie er will. The question is: is nominative possible here as well. It seems to me that it is but I am not quite sure. What does everybody think? In any event, I see no way around postulating that the subject of an infinitive must be nominative unless it is structure-shared with an NP in a higher clause that requires a different case. I looked at the Andrews article one more time and Andrews does indeed reference Jakobson 1936 and implements the idea that nominative case is the absence of case in Icelandic. See the discussion around page 470. Best, Gert
This archive was generated by hypermail 2.0b3 on Fri Dec 18 1998 - 20:33:03 PST