LINGUIST List 4.179

Sat 13 Mar 1993

Disc: Pro-Drop

Editor for this issue: <>

Directory

1. Patrick Farrell, Re: 4.175 Pro-drop
2. Richard Ogden, Pro-drop
3. Almeida Jacqueline Toribio, Re: 4.175 Pro-drop

Message 1: Re: 4.175 Pro-drop

Does anyone know of any arguments against an analysis according
to which colloquial English is a language that sanctions
null subject pronouns.
Sure, the conditions on their occurrence in English are
quite unlike those found in Italian (root clauses only, auxiliaries can
sometimes disappear with them, etc.). But so what?
English, Welsh, and Italian all have quite different conditions on
the occurrence of overt pronouns as well; but I don't think
anyone would want to say that one or another of these
is a language that does not sanction overt pronouns or that what appear
to be overt pronouns in one of them really are not.

One conclusion that seems to have emerged from the discussion here
of subject 'pro-drop' is that languages vary considerably
with respect to the conditions they place on it.
Is this phenomenon in English pro-drop or isn't it?
Well ... what is pro-drop?
Suppose we define it as the phenomenon whereby a syntactic slot normally
occupied by an NP is left empty, yielding a pronominal interpretation.
Then, English has subject pro-drop.
Suppose we define it as the same kind of phenomenon, except that in
order to qualify it must be able to occur in embedded clauses or with
high frequency, or some such.
Then English does not have subject pro-drop.

Does it matter much whether one says English has pro-drop or not?

Is there an otherwise useful and sufficiently precise
definition of pro-drop such that one could decide anyway?

I know certain assumptions have been made in this regard,
but has it actually been established that
subject pro-drop vs not subject pro-drop is a typologically
interesting parameter of variation among languages?
Is there even any suggestive evidence (based on
more than a few Indo-European languages) that it might be?

Patrick Farrell

Joe Stemberger raises important questions about the phonetics
of 'pro-drop' sentences. He gives the example:

 Forgotten what that was about.

where I at least would pronounce this with a long initial fricative
which starts with very low amplitude friction (reminiscent of what
I do for 'voiced' fricatives in my speech). I don't know that I agree
with his conclusion that 'the auxiliary is absent, along with the subject',
because it seems to me that the reason why this is [f:] and not [f]
is that the way I do the phonetics of pieces like "I've" could be
just the kind of thing I have just described, and this may be the
phonetic exponency of auxiliaries or pronouns+auxiliaries.
What counts as 'deletion' or 'prodrop' at the phonetic level? We need
to know what phonetic features are dictinctive at any one place.

In the example:

 'M eatin'.

for "I'm eating", there is for me at least no homophony with the
word "meeting". There are several differences, any or all of which
may be significant. The nasal portion is somewhat longer than in
"meeting" and the tongue body is not as far forward (as [i]-like)
as in "meeting": more like a [m] in the context of "I'm", in fact.
Again: what is 'deleted'? It seems to me there are 'traces' or
'remnants' of "I" in the [m] which I do here. Similar comments apply
to the expression:

 'M gonna go.

These differences in duration and resonance may be distinctive
at this place in structure, they may distinguish "meeting" from
"'m eating".

I think this sort of thing may argue for a more constant, static
syntax than some of the more liberal 'prodrop' accounts being
suggested. I also think that the best account may be one of phonetic
variability rather than phonological or syntactic variability; but
with the proviso that the phonetic variability is constrained by
syntactic considerations. By this I mean that the portions between
astrisks in "*I'm g*onna..." and "*I've f*orgotten" etc can have a
different range of phonetic variability from "*m*eeting" etc.

Richard Ogden
Experimental phonetics laboratory,
University of York
England

Message 3: Re: 4.175 Pro-drop

I would like to respond to the query regarding the possible correlation between
 null subjects and dative subjects. Wayne Harbert and I suggest that such a co
rrelation does hold; in fact, we claim that experiencer subject constructions a
re possible only in those languages which license a null expletive. The propos
al appears in a paper entitled "Nominative Objects" in Cornell Working Papers i
n Linguistics 9 (Fall 1991). Those interested in obtaining a copy of the volum
e should contact me directly at: cwtjcornella.bitnet.

Jacqueline Toribio