> A ream of paper is a singular and specific object although reams do
> vary in size according to the nature of the paper being measured.
> This may well be forgotten by those who have begun to measure paper
> by the S.I. metric methods. Would you argue with:
>   "There are many buckets of water"
>    (given that buckets may vary in size?)

The sentence above is of course unarguable, except that the use of
"many" in the affirmative sounds a bit stilted.  For that reason, a lot
of people use "a lot", and not many people use "many".  "A lot" has the
further benefit that it doesn't distinguish between count and mass (and
therefore singular/plural):

   A lot of cheeks were turned.    A lot of cheek was displayed.
   (Not) many cheeks were turned.  (Not) much cheek was displayed.

which is often helpful when you don't want to have to make that
distinction because you're concentrating on something else, especially
in conversation.

That said, the question remains about the category status of "lot" in
"a lot of".  It *looks* like a noun phrase, so that would make "lot" a
noun.  And it's a singular noun, so how does the agreement work?  Isn't
there something irregular about all this?

Of course there is.  The noun phrase "a lot" has entered the Twilight
Zone; it how has a function instead of a meaning, and has been demoted
to modifier (in fact, quantifier) status, and it has lost its right to
be the subject and agree with the verb.

OK, we're all agreed, I suspect, on the star below:

    A lot of papers *is/are blowing around the room.

(except on a nominal meaning for "lot", roughly = "bunch").
Notice now that a variant of "a lot of" is "lots of", and that
it has identical agreement properties, i.e, none.

   Lots of evidence has/*have accumulated.

So, although one looks singular and the other looks plural,
looks can be deceiving; neither one has a number at all.

Now let's put these into existential constructions with "there",
which we can do with some (but by no means all) verbs:

   There are a lot/lots of papers blowing around.
   There is a lot/lots of evidence in this trial.

Same pattern; "lot(s)" doesn't govern agreement.  Though the placement
of the singular-sounding "a lot" right after the plural "are" will
bother the fastidious, because the construction is just too close for
comfort.  Ditto for the plural-sounding "lots" right after the
singular "is".

OK, now let's try a different word: "reams".  This is a classifier for
paper, thus by extension for anything writeable on paper, thus just
about any abstract noun involving words.  "Evidence" is clearly one
such.  As a classifier, "ream(s)" can participate in the same
construction, but as a junior-grade apprentice at best; it's simply
not as well-established as "lot(s)", and is therefore much more

    There is/are a ream of evidence in this case.
    There is/are reams of evidence in this case.

    There is/are a ream of funny jokes in that newsgroup.
    There is/are reams of funny jokes in that newsgroup.

I haven't awarded asterisks here.  Frankly, it strikes me more as an
issue of personal style than of grammaticality, let alone "correctness".
The degree to which anyone analogizes "ream(s)" with "lot(s)" will
vary enormously from person to person (as I think we have witnessed in
this thread), and the style one might want to adopt in a particular
writing job will vary independently as well.   That's far too many
uncontrolled variables to prescribe for with a simple fix.

One suggestion to those still troubled by this: consider using modal
auxiliary verbs. They also don't distinguish singular from plural, so
they can be used with confidence in cases where you're not sure which is
appropriate.  Besides, a "can" or "might" or "should" often adds a
thoughtful qualification to a remark and may prevent one from sounding
too authoritative (though I fear they won't save me here :-)

--- Followup (on the same topic, though not the same thread)

>Another of my bones of contention with supposed speakers of English is the
>corporate world forcing incorrect usage of the language on the already
>deficient consumers.  The prime example of this is "less calories," which
>is actually printed on all manner of diet and other foodstuffs.  I expect
>to be seeing "fewer fat" any time now.

Well, I hope your bone is a real contender.  As for me, I'm interested
in (though -- shamelessly -- not angry about) your observation for
different reasons.  If "less calories" is common, and it sounds to me
like it could be (in contrast to "fewer fat", though "fewer fats" with
the same meaning might make it), then it means that "calories" has
been reified into a mass noun, with the resultant use of "few" (and
"much", "there is", etc, which should follow in due time).

That's interesting to me because it reflects on how English speakers
(principally American, but only in the first instance; the whole world
is the customer for American linguistic engineering :-) are
understanding the sense of "calorie".  Originally a unit of scientific
measurement, (note the odd spelling; I originally spelt it "calory"
and reconsidered) it's been used far more widely as a relative measure
of the "food value" of various foods.  Those schooled in the physics
may well use it in the plural with some meaning because they may have
a sense of what it means, experientially.

This is a pretty learned bunch, and I'm sure many if not most of us
could cite the scientific definition and many of us may work with it
daily.  But let me suggest that we are not the intended audience of
this ad. 

The intended audience is swayed far more by esthetic than scientific
reality, and the fact that "less calories" has a much better syllabic
structure than "fewer calories" does renders it (imho) far more
suitable as the kind of slogan that advertising deals in.  In other
words, Somebody Up There thinks it's more likely to stick in our minds.

Besides, there's the precedent of 'data', for much the same reasons.
Only scientists deal with single datums; everybody else encounters
it en masse.  And plenty of others.  And a long train of abuses and
usurpations of grammatical authority in flagrantly switching mass
and count reference:

    The Waters of the Pacific       [exact count, please?]
    ...a lot of car for your money  [387 milliHenrys/Dollar?]
    There was bug all over the windshield.

etc.  Relax and enjoy it.  The language can get along without us.

 -John Lawler                     More grammar
  Linguistics Program   University of Michigan
 "..and, who knows? Maybe the horse will sing."