Re: New coinage?

Arnold M. Zwicky zwicky at CSLI.STANFORD.EDU
Tue Aug 9 17:44:24 UTC 2005

On Aug 9, 2005, at 7:33 AM, Ron Butters wrote:

> Ben Zimmer's comments are really helpful. Still, I think I can
> readily make
> up examples, though I also have the intuition that people would
> find them odd.
> 1 a. The politician has a passion for baby kissing.
>     b. The politicain baby-kissed his way across Kansas
>     c. The politican baby-kissed 37 infants in Kansas City....
> I supppose that one could argue that... "name-call(ing)" is an
> independent lexical item the meaning of which is greater than the
> sum of its parts. And there are obvious grammatical compexities in
> the differences in the ways that the three sets of sentences work.
> Still, would any speaker of English complain that the "c" sentences
> are morphosyntactically "ungrammatical"?
> In a message dated 8/9/05 12:25:46 AM, bgzimmer at RCI.RUTGERS.EDU
> writes:
>> "N-V" back-formations from deverbal compounds "N-Ving" or "N-Ver"
>> are not
>> uncommon, but I think they're notable enough to warrant lexicographic
>> recognition...

ben alludes here to the deverbal compounds of the form N-Ver and N-
Ving.  there's a fair literature on these; in the present context,
the crucial item is

Kiparsky, Paul.  1974.  Remarks on analogical change.  Anderson &
Jones (ed.), Historical Linguistics I, 257-75.  Amsterdam: North-

what kiparsky observes here is that the N-Ving compounds function not
just as nominals (their "gerund" use), but in essentially all types
of constructions involving the -ing forms of Vs (progressives,
postmodifiers, premodifiers, etc.); this point is hammered home in

Pullum, Geoffrey K. & Arnold M. Zwicky.  1991.  Condition
duplication, paradigm homonymy, and transconstructional constraints.
BLS 17.252-66.

kiparsky notes (and pullum and i stress) that, with certain
exceptions where a compound N-V verb has been backformed ("babysit"
and the like), this N-V verb doesn't occur *except* in the -ing
form.  so we all would predict that ron's (1a) is fine, but (1b) and
(1c) are not.  native speaker judgments generally agree with that

*but* there is constant pressure towards back-formation.  whenever a
N-Ving item becomes at all frequent, develops some specialized
meaning, etc., speakers are likely to posit a compound N-V verb,
usable in all forms.  these effects will work differently for
different people, depending on their experiences. a few more
citations of "baby-kiss" and "name-call", and i'll probably start
finding them more and more acceptable.  ("spear-fish" long ago went
that route for me, just because "spear-fishing" appears frequently in
the morphological literature!)

(ben makes the excellent point that "name-calling" isn't one of the
standard subtypes of N-Ving compounds, which might make it especially
notable when it's subject to back-formation.)

arnold (zwicky at

More information about the Ads-l mailing list