I'm done my homework

[Arnold M. Zwicky]

On Dec 10, 2004, at 6:03 AM, David Bowie wrote:

> From:    "Arnold M. Zwicky" <zwicky at CSLI.STANFORD.EDU>
>
> : it seems likely to me that BE DONE/FINISHED NP is an extension of BE
> : DONE/FINISHED VPing (I'm done washing the dishes, Are you finished
> : doing your homework?), with a VPing complement.  this would be on
> : analogy with other aspectual verbs, which occur with both VPing
> : complements and ordinary NP objects (I began/continued/finished
> : singing, I began/continued/finished my song, etc.).
>
> Makes sense to me.
>
> You're listing these as *an* idiom, though, and i'm not sure they
> really
> are. BE DONE NP sounds <ObIntuitionDisclaimer> more well-formed to me
> than
> BE FINISHED NP, for example, though i don't think i actually use
> either of
> them.

well, that's what i get for using terminology loosely.  what we're
dealing with here is a (conventionalized) formal pattern associated
with meaning.  some of these patterns allow for great freedom with
respect to what can occur in specific slots; these we traditionally
call "constructions".  other patterns are highly constrained in this
respect; these we traditionally call "idioms".   but very few idioms
are completely fixed in every respect, so the line between idioms and
constructions is unclear.  really, the things we call "idioms" are just
small-scale constructions.  these days i think there's no point in
trying to draw the distinction (i used to think otherwise, and said so
in print), so that we need a term that covers them both.  in my 2002
NWAV paper "Seeds of Variation and Change" (the handout for which can
be viewed at http://www-csli.stanford.edu/~zwicky/seedhnd.pdf) i used
the rather ugly umbrella term "complex sign types" ("comsits" for
short).  (simplex sign types are variously known as words, lexemes, or
lexical items.)

all this is relevant because comsits are subject to considerable
variation as to what material can fill particular slots.  people differ
as to which verbs participate in the Dative Movement construction (I
gave/?donated the church a lot of money), the Subject-to-Object Raising
construction (I believed/?imagined Carlos to be a spy), the
Quasi-Serial Verb construction, or idiom (Go/?Run see who's at the
door), the GoToGo construction, or idiom (She's going/?coming to San
Francisco and give a talk on firewalls), etc.  As the constructions
become more or more constrained, with very small numbers of fillers for
particular slots, linguists are more and more inclined to treat the
fillers separately: the literature on passivoid V + VPpastparticiple,
for instance, usually looks at the head Vs "need" (My shirt needs
washed) and "want" (My shirt wants washed) separately.  (side question:
are there people with other verbs, like "demand" and "require", filling
this slot?)  but there's a single pattern here, so it makes sense to
refer to *a* comsit.

that's what's going on with "be Ved my homework" (for which the pattern
is BE + Vpastparticiple + NPdirobj, with completive meaning).  there is
*a* comsit (which we might call an idiom or a construction, according
to taste), but what's interesting to variationists is the details of
this comsit with V "do" versus V "finish".  so we talk about BE DONE NP
and BE FINISHED NP separately.  but, strictly speaking, they're just
instances of one comsit (which i referred to as an "idiom", though
absolutely nothing hinges on that wording).

arnold (zwicky at csli.stanford.edu)