Needs specimen

[Arnold Zwicky]

From: "James E. Clapp" <jeclapp at WANS.NET>:

>Arnold Zwicky wrote:

>> in central ohio, it is indeed frequent, and many locals do not use
>> NEED/WANT V-presprt.

>It's a technicality, I suppose, but surely the verb in the phrase
>"needs cleaning" (or whatever) is not a present participle, but
>rather a gerund.

this is even more of a technicality, but i'm using the term "present
participle" as the name of a morphological form.  it could just as
well be labeled "form 4".  some english grammarians call it the" -ing
form".  rodney huddleston calls it the "gerund participle"!

the point is that english has a single verb form, with an affix -ing
or -in (the realization depending on nongrammatical factors), and with
a huge number of uses (two dozen or so), some of them clearly verbal
in character ("washing" in "I was washing the car"), some clearly
nominal ("washing" in "My washing of the car took three hours"), some
clearly adjectival "having" in "Anyone having a hat on will be
expelled"), some not clear at all ("having" in "my having a hat on"
famously has some verbal and some nominal properties).

>It's the direct object.

no, it's not.  it shows none of the characteristic properties of
direct objects - no passivization (*Draining is wanted by these
iris beds), no "tough-movement" (*Draining is easy for these iris
beds to want), etc.

yes, it follows a tensed verb, the way direct objects do.  but so
do what are sometimes called "verbal complements", as in the
progressive or in aspectual verb constructions ("The bell stopped
ringing").

(some complications are introduced into the analysis of such data
by the fact that NEED and WANT can *also* have direct objects,
as in "I need/want an ice cream cone" or "Your essay needs/wants
more attention".  the problem is that clearly nominal occurrences
of V-ing can be direct objects in such cases: "Your essay needs/
wants immediate rewriting."  but cf. "Your essay needs/wants
rewriting immediately.")

>The fact that this grammatical construction is so basic and
>straightforward ("The bed needs making"; "I need your loving":
>subject-verb-object) makes it seem much more "correct" to those whose
>dialect uses it than "The bed needs made" or "I need loved by you(?)"
>which is harder to analyze.

the V-ing version isn't as basic and straightforward as you think.
(you're just used to it.  it's the one you have in your dialect.)
*just* like the V-en version, the V-ing construction is interpreted
"passively" - that is, the subject of NEED/WANT is interpreted
as the *object* of the following V.

>I assume it is a shortening of "to be made," "to be loved"...

this is an entirely gratuitous assumption.  just because there
is a longer paraphrase doesn't mean that the shorter construction
is, diachronically or synchronically, an elliptical version of the
longer one.

i believe that historically the verbal-complement V-ing construction
is a reanalysis of the direct-object construction with a V-ing object;
"This shirt needs cleaning" could be understood either way.  the V-en
construction is then an entirely natural reshaping of the V-ing
construction (probably with a substrate influence - i seem to recall
that the V-en construction is one of those scots-irish things in
american english - although it could have arisen independently); in
the reshaping, passive (V-en) morphology is supplied for passivoid
clauses like "This shirt needs cleaning immediately".  (the same
instinct inclines speakers of languages without tough-movement to
reshape this construction with the passive: "The cat is hard to be
seen" rather than "The cat is hard to see.")  if you step back
from the brute facts of your native dialect, you'll see that V-en
actually makes more sense in this construction.  "needs cleaned"
speakers are neither speaking illogically nor just whimsically
leaving out words.

arnold (zwicky at csli.stanford.edu)