[matrix] MODIFIED.PERIPH

[Francis Bond]

G'day,

2009/6/6 Emily M. Bender <emily.m.bender at gmail.com>:
> The only thing I have is this comment in matrix.tdl:
>
> ; Three-valued sort for distinguishing unmodified signs from both
> ; left-modified and right-modified signs PERIPH indicates whether this
> ; modifier is left- or right-peripheral in its phrase - e.g., "the IBM
> ; temporary employees" but "*the IBM five employees"
>
> It looks like something that the matrix carried over only in part
> from the ERG.
>
> I make use of PERIPH in the Wambaya grammar, though not
> necessarily in the way it's used in the ERG:
>
> ;;; ERB 2007-12-29 I'd like to keep vocatives off of right conjuncts.  So
> ;;; I'm adding a new feature that adj-head phrases will get from their
> ;;; adj daughter, and which the bottom-coord rule for s coordination will
> ;;; look for.  Actually, I might be able to reuse PERIPH for this purpose.
> ;;; and indeed, adj-head-int-phrase already identifies the PERIPH value
> ;;; of the mother with the PERIPH value of the non-head daughter.  So
> ;;; it should suffice to give vocatives [ PERIPH + ] and have the
> ;;; s-coord rule look for [ PERIPH - ].
>
> Making this work also required propagating the PERIPH value in a bunch
> of phrases.

So just to make sure I understand -- PERIPH is being used to say I am
on the outer edge, don't modify me any more here?


-- 
Francis Bond <http://www2.nict.go.jp/x/x161/en/member/bond/>
NICT Language Infrastructure Group