the cooccurrence of two genitives

[Matthew S Dryer]

I sent the following message to the LINGTYP list last week, or at least
attempted to, but it apparently didn't make it.  Its central point is
similar to one made in a recent message by Alan King.

Edith,

I confess that I have only been skim-reading previous discussion of the
double-genitive problem on Lingtyp, but your recent summary is so clear
and well-thought out (in typical Edith Moravcsik fashion) that I found it
well worth reading closely.

My conclusion on reading your summary is that it seems clear to me what
the correct solution is, and the arguments you offer against it seem
flawed.  Namely, the correct solution is your syntactic solution 2.1.c:

>>c/ Third, there may be a direct constraint against two genitives in
>>a single phrase.

My claim is that essentially the same constraint applies in Hungarian that
applies in English.  You argue that

>>as we have seen above, in Hungarian, unlike in English, the
>>   cooccurrence of two different types of genitives is also excluded,
>>   one preposed the other postposed.

However, there is a fundamental difference between Hungarian and English:
the preposed and postposed genitive constructions in English are DIFFERENT
constructions, while the preposed and postposed genitives in Hungarian are
the SAME construction, differing only in position.  In other words, there
is no reason, internal to the grammar of English, to say that the
prenominal genitive and the postnominal genitive are instances of the same
phenomenon: it is only semantic or typological considerations which lead
us to assign the term 'genitive construction' to both of them.  In
Hungarian, however, they are apparently instances of the same
construction, characterized by the same genitive suffix.  While one
precedes the noun and the other follows, this apparently reflects the fact
that Hungarian lacks a grammatical rule requiring one order.  In that
sense the grammar of Hungarian does not involve two different genitive
constructions.

Matthew