LINGUIST List 3.438

Tue 26 May 1992

Disc: Adjuncts

Editor for this issue: <>

Directory

1. Eric Schiller, Re: 3.431 Adjuncts
2. bert peeters, French prepositions governing non-finite clauses

Message 1: Re: 3.431 Adjuncts

Yes, the anonymous Englishman (or at least UK dweller) was correct,
in his/her/its post of Vol-3-431. My tongue was firmly in cheek,
and I agree completely with the lexical semantic links between
time expressions and use of 'when' as a complementizer, similarly
locational phrases and 'where', etc. I do take these links very
seriously, even though in an unpublished 1985 paper I referred to
these properties of time and space as TARDIS features (and still do).
The question that arises is whether these facts are grammaticalized
to the point where the features should participate in the formal
syntactic and semantic analyses. Similar questions arise with
regard to animacy hierachies and other features which seem to
bridge the chasm between real world semantics and grammatical
theory.

Exactly how to link such notions to descriptive and analytical
frameworks means dirtying one's hands in the data, something
which some members of our community seem to object to. Within
the Autolexical community, such questions of grammaticality can
be asked and answered, and perhaps soon one of us will do so.

Eric Schiller
University of Chicago
schillersapir.uchicago.edu

> Date: Fri, 22 May 92 01:15:02 BST
> From: WAB2phx.cam.ac.uk

> Lloyd Holliday (3048/4) is right in saying that what makes the grammaticality
> of
> I remember the first time when we played golf
>
> suspicious is the contiguity of "time" and "when".

> It is not a phonetic problem. The problem is one of semantic overlap, as also
> between "place" and "where" in
>
> I remember the first place where we played golf.

If the first sample sentence is "suspicious" (I must confess it doesn't shock
my non-native ears - or eyes, and I wonder whether for native speakers it could
 be something they say quite often in careless speech), the second sample
sentence is definitely far better (if not entirely acceptable - how else could
one express the idea?). This leads me to the assumption that semantic overlap
is not the entire story either. Nicolas Ruwet has a delightful study in his
recent *Syntax and Human Experience* about meteorological expressions such
as "la pluie tombe"/"rain is falling" or "le vent souffle"/"the wind is
blowing". Now, these are OK, yet there is a significant semantic overlap.
What else is rain, he argues, but a downward movement of water particles?
What else is wind, one could add, but a blow or displacement of air?

Not that I have any answer to the English problem at hand - I just wanted to
point out that semantic overlap is not necessarily the reason behind the
"suspiciousness" of the first of Lloyd Holliday's sample sentences.

> Lloyd Holliday is also right to conclude that "day"
> >is less basically a word like "time"
> concluding that this made it easier to assign unquestioning grammaticality to
>
> I remember the first day when I played golf.

What are the criteria permitting to distinguish between less and more basic
words? What makes "day" less basically a word like "time"? "Place" and "time"
seem to be at the same level of generality (at least in Indo-European based
languages). Does this mean that when "place" is being replaced with something
"less basic" (proceeding by analogy, I suggest as a working example "univers-
ity campus"), the second sample sentence becomes better? Is

	I remember the first university campus where I played golf

any better than

	I remember the first place where I played golf ?

I'm finally getting to what I wanted to talk about in the first place...

> A preposition in French must govern a noun, and such a noun must be overt in a
> finite clause. Yet, like English, a preposition can govern a clause, without
> the intercession of a noun, if that clause is non-finite
>
> Jean est venu A [retrouver son ami]
> John managed [to find his friend]

A word of warning is necessary: not all prepositions in French can govern
a clause (even a non-finite one). The most common ones are "de" and "a`".
On the other hand, not all instances of "de" introducing a clause preceded
by a verb are prepositions. If "de" were a preposition in

	Je vous propose de me suivre
	I suggest you follow me

pronominalization of the non-finite clause would result in

	*Je vous en propose

whereas the only correct way of pronominalization is

	Je vous le propose
	I you it suggest
	This is my suggestion to you