LINGUIST List 4.1054

Wed 15 Dec 1993

Sum: German grammaticality, Unbounded coordination

Editor for this issue: <>

Directory

1. , German/grammaticality
2. , Summary: Unbounded (=flat) coordination

 First of all, let me give 'vielen Dank' to the 38 native speakers of
German who answered my plea for grammaticality judgements. In addition, since
some of you requested the results, as did some non-native speakers of German,
I've included them here. What I did was tally the first 30 responses. The
asterisk (*) was worth 2 points and the question mark (?) was worth 1. Hence,
those sentences with tallies nearing 60 are the most ill-formed, those around
30 are odd, and those near 0 are well-formed.

 Lots of discussion was engendered by the sentences in (2). It seems
that with other PPs and NPs 'geben' would be okay. 'In Kuhlschrank gibt es
einen Apfelkuchen.' was one and another was something like 'In Berlin gibt es
viele Leute . . . '

 Sentence (3c) also got a lot of comments since the pragmatic
circumstances that might lead to its use are hard to imagine. This
construction places contrastive stress on the location and implies that the
NP changes. Someone suggested a sentence like this: 'Auf dem TISCH, ist es
ein Buch, aber auf dem BODEN, ist es ein 'doorstop'. (Turhalt??? -can't find
'doorstop' in my dictionary.)



1. a. Auf dem Tisch liegt ein Buch.
 b. Es liegt ein Buch auf dem Tisch.
 c. Ein Buch liegt auf dem Tisch.

2. a. Auf dem Tisch gibt ein Buch. 55
 b. Auf dem Tisch es gibt ein Buch. 57
 c. Auf dem Tisch gibt es ein Buch. 17
 d. Ein Buch gibt auf dem Tisch. 58
 e. Es gibt ein Buch auf dem Tisch. 17

3. a. Auf dem Tisch ist ein Buch. 2
 b. Auf dem Tisch es ist ein Buch. 52
 c. Auf dem Tisch ist es ein Buch. 34
 d. Ein Buch ist auf dem Tisch. 2
 e. Es ist ein Buch auf dem Tisch. 4


 Thanks again.

 Mark Balhorn
 mbalhornuwspmail.uwsp.edu

I would like to thank the many respondents to my query about
the origins of unbounded coordination in TG. While no one
provided quite the information I was looking for, I was led
in the right direction and believe that it is accurate to
say that this was first introduced in Lila Gleitman's Lang
article (where she thanks Chomsky for suggesting it). Where
Chomsky got the idea is not entirely clear, but there are
references in writings of Lees and Postal which suggest that
the realization that there have to be unbounded structures
came to people in TG from a 1960 article in Language by
Longacre in which he criticizes IC analysis and favors string
analysis. However, Longacre's examples were series of
adjectives as in "tall, slim, well-dressed women" rather
than the usual kind of coordinate structure that usually
gets mentioned in this context ("John, Bill, and Harry").

I have also stumbled on the pedagogical TG of English
from the early '60's by one Paul Roberts which mentions
coordinate structures like "John, Bill, and Harry" but
do not offer an actual statement of a transformation
to produce these. Since Roberts claims that Chomsky
read his ms. and commented on it in detail (and since
Chomsky wrote a foreword), it may well be that Roberts
had some role in calling these constructions to Chomsky's
attention or more likely reminding him of something which
he had already noticed.

In any case, early versions of TG did not allow for unbounded
but only for binary coordination, although the lack of formalization
of TG it is impossible to tell whether this was just a lack
of insight about how to analyze certain NL constructions or
a defect in the very theory of TG as it existed at the time.