Couple.

[Laurence Horn]

At 7:42 AM -0500 12/7/00, Paul McFedries wrote:
>And another thing...
>
>Since the informal couple (sans of) usage means "two or a few," could it
>also be related to the expression "couple, two, three," as seen in the
>following cite from something called "Denver Westword":
>
>"He admitted to having had sex with her a 'couple, two, three times,' the
>first in August 1998."
>
>This would also seem to mean "two or a few," so perhaps in addition to
>Dennis' phonological argument, the informal couple is a shortening of this
>longer and puzzlingly redundant phrase. This would require that "couple,
>two, three" have had widespread usage in the past, so perhaps I'm making
>things too complicated again.
>
Interestingly, the OED cites this sense:

b. (With 'of' omitted) = couple of (cf. coupla; U.S. colloq.).

        1925 S. Lewis Martin Arrowsmith xvi. 188 A couple months in Italy.

        1934 D. Hammett Thin Man xxii. 173 She touched me for a couple
hundred to blow town.

as well as the above-mentioned "coupla":

coupla, orig. U.S. Colloquial form of  'couple of'.

        1908 H. Green Actors' Boarding House 254 A coupla parties is
come for rooms!

        1934 D. L. Sayers Nine Tailors iii. ii. 276 He'd had nothing
to eat..for a coupla days.

But in neither case does it "allow" for the meaning to be looser, in
a cardinal sense, from that of 'couple'.  The OED does acknowledge
degrees of looseness in whether the two items in a couple have to be
a real yoked pair, rather than just two items of the same kind:
[COUPLE, 7a]
'Two individuals (persons, animals, or things) of the same sort taken
together; properly used of such as are paired or associated by some
common function or relation; but often
loosely, as a mere synonym for two'

My own intuition is that "a couple of Xs" (and not just "a couple
Xs") can be used when the speaker is allowing for there being two or
(somewhat) more Xs.  Of course, this is true for "two" as well, and
for all scalar terms, but my sense is that "couple" is inherently
vaguer in its cardinality than "two".  This only applies to cases
where it serves as a quantifier of another head noun; I can't say of
a congenial m'enage a' trois that they're a nice couple.   [Notice
also that the OED's definition of head-noun "couple" is as
heterosexist as its definition of "love":  'a man and woman united by
love or marriage'.  Jesse, can we change that?  Or has it already
been done?]

larry