Back on August 7, E. Ward Gilman (of MWDEU fame) e-mailed me about Burchfield’s Fowler (billed as the third edition, 1998, of Fowler’s Dictionary of Modern English Usage, 1926):
Quite a few years ago, when I was hoping to be able to write a review of Burchfield’s Fowler, I ran across another curiosity. It was the pronominal use of “various”, as in “I spoke with various of them”. This use was condemned in the original Fowler, with citations, and in Gower’s edition, with additional citations, but is omitted entirely from Burchfield. I don’t know whether it was omitted inadvertently or whether Burchfield left it out on purpose.
I replied that it was hard to tell: maybe Burchfield just decided the usage was now acceptable; unfortunately, we can’t ask him (he died in 2004).
Now some further discussion, reproducing some material posted to the American Dialect Society Mailing List back in August.
Both NOAD2 and AHD4 list pronoun uses of various, but with a usage note in each case. NOAD2 calls the pronoun use “colloquial American”, noting that “some traditionalists” insist that various is only an adjective. I see this opinion as denying the possibility of language change, in particular a change in which various developing an pronominal use parallel to several.
(In some postings i’ve called the operative assumption here Originalism — the very silly idea that the original use of a word is its only acceptable use now.)
[Digression 1 about part-of-speech classifications: these sources seem to use pronoun as the part-of-speech classification for quantifier words used as determiners with of. I’ll play along with that here, with a proviso to come shortly.]
[Digression 2 on this topic: OED2 does not in fact label such uses of several — in several of — as pronominal. Instead, these uses are treated as elliptical uses of the adjective several. OED2 does something similar for few in few of. But the June 2009 draft revision for many classifies it as a pronoun in many of. I assume that when revisions get to several and few, they will be similarly reclassified.]
[Digression 3 on this topic: on ADS-L, Randy Alexander noted the treatment of these items in CGEL as “fused head” structures: fused
determinative-head (in few, many, several), and fused modifier-head (in various). I am in fact a proponent of this analysis, but chose to formulate my discussion in more traditional terms so as to connect with the dictionary treatments.]
Back to usage advice. AHD4 reports very substantial hostility to pronominal various on its usage panel, with somewhat less hostility towards it with inanimate NPs. The usage note concludes:
It is not clear why this usage should be regarded as an error, since it is analogous to the use of quantifiers such as few, many, and several.
Well, it is indeed analogous, but few, many, and several developed pronominal uses well before various did. The offense of various is that it came late to the party; the doors were closed, and no new items are to be allowed in. Then commenters like Fowler fixed on it as an innovation, and others piled on, as often happens; it became part of the peevelore, to the extent that you felt
that MWDEU had to warn readers that some people view the usage as a
straightforward error, so that you might want to be cautious about using it. (Crazies win, as I said in a Language Log posting a while back.
The case is interesting, because it’s not clear what the fix is supposed to be. The usefulness of pronominal various is that various of can combine with definite NPs, as in various of them and various of these commenters. So if you proscribe pronominal various, then some other partitive construction must be used (several of, a number of), or an indefinite construction must be used (various people, various commenters), or a noun or pronoun head must be supplied for the adjective various (various ones of them, various commenters of these). Each of these work-arounds has its defects, but the first is probably the best; its defect is that the association with variety is absent in the alternative partitives.
