Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/17/84; site bcsaic.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!mhuxr!mhuxt!houxm!vax135!cornell!uw-beaver!fluke!ssc-vax!bcsaic!michaelm
From: michaelm@bcsaic.UUCP (michael b maxwell)
Newsgroups: net.nlang
Subject: Re: Past participles in English
Message-ID: <304@bcsaic.UUCP>
Date: Wed, 18-Sep-85 13:22:55 EDT
Article-I.D.: bcsaic.304
Posted: Wed Sep 18 13:22:55 1985
Date-Received: Sat, 21-Sep-85 06:10:36 EDT
References: <521@mmintl.UUCP> <520@scc.UUCP> <490@spar.UUCP> <546@scc.UUCP>
Reply-To: michaelm@bcsaic.UUCP (michael b maxwell)
Organization: Boeing Computer Services AI Center, Seattle
Lines: 27
Summary: 

In article <546@scc.UUCP> steiny@scc.UUCP (Don Steiny) writes:
>I proposed that, as a hypothesis, we do away with the PS rule:
>
>	I) NP => (DET) ADJ[0] N
>
>and instead say that the string (DET) ADJ[0] N is derived from an
>underlying S.
I think I may have said this before, but:
	the late Shah of Iran
=/=	the Shah of Iran who was late
There's a long literature on this topic; it was indeed one of the ideas
popular in earlier versions of transformational grammar (mid '60's-mid
'70's); there are good reasons why it was dropped.  One was the sort of
example I gave above.  If adjectives like "late" need to be generated by
your PS rule (I), then all adjectives *can*, thus eliminating the need for
a transformational derivation.  Then apply Occam's razor...
>	If anyone really makes it this far, are you familiar with
>"Generalized Phrase Structure Grammar?"   I just got a book on
>it, but it is obviously not one I can read on breaks and I am going
>to have to spend a serious few weekends ploughing through it.
Yes, I am familiar w/ GPSG (although like you, I haven't gotten around
to plowing through that particular book).  It does away with
transformations entirely, as do a number of other recent generative
linguistic theories (not including Government Binding!)
-- 
Mike Maxwell
	..uw-beaver!{uw-june,ssc-vax}!bcsaic!michaelm