Possession/modification by simple juxtaposition

[Östen Dahl]

I feel a little responsible for the Västerbotten data, since I seem to have been
the one who suggested them to David Gil in the first place. ("The Västerbotten
dialect of Swedish" is a misleading label since the dialects in the province of
Västerbotten are quite heterogeneous, and sometimes not very "Swedish".) It
should be emphasized that in both cases these are just one of several options
and at least in the case of possessor constructions not even the most common one
(although nobody has done a frequency count to my knowledge). I also think that
"simple juxtaposition" is not a wholly adequate label for these constructions,
which are rather to be seen as a kind of incorporation involving among other
things "compound" prosody. 

Östen Dahl

On 2008-11-22, at 13:17, Wolfgang Schulze wrote:
> Personally, I wonder whether the rigid pre-modifying strategy, namely 
> Por-Pum / Attr-N is documented with juxtaposition at all. In WALS, David 
> Gil mentions the "Västerbotten dialect of Swedish, in which alienable 
> possessors and colour properties may both be expressed by means of a 
> compound construction in which the modifier precedes the head":
>
> Pelle-äpple
> Pelle-apple
> 'Pelle's apple'
>
> rö-äpple
> red-apple
> 'red apple'
>
> But do we have languages in which both juxtaposition and 
> pre-modification occur as a general (!) pattern, or is the 
> post-modifying Pum-Por - N-Attr correlation (with yuxtaposition) a more 
> general option, competing with the heterogeneous Por-Pum / N-Attr 
> strategy (again with yuxtaposition)? Summarizing some of the data given 
> in this thread, we might state for yuxtaposition:
>
>                  Attr-N                 N-Attr
> Por-Pum     e.g. Swedish       e.g. Guaraní      
> Pum-Por     e.g. Welsh          ? 
>
> [Both Swedish and Welsh 'partial']
>