Over on Language Log, Mark Liberman inaugurated the new year with a discussion of how to pronounce the name of the year:
All around the English-speaking world, pundits are wondering in print about how to pronounce the year 2010. Is it “twenty ten”, or “two thousand ten”, or “two thousand and ten”, or what?
(The choice of variant also comes up in written English, on occasions when year names are “written out” as phrases.)
Mark surveyed some opinion on this burning issue, and his posting elicited a huge number of comments (over a hundred the last I looked), some arguing for one usage over the others (with a variety of justifications), some offering other patterns for pronouncing year names (“twenty oh/aught nine” for 2009, for instance), plus discussions of the patterns in other languages (a topic Mark invited) and of ways of pronouncing other numerical expressions (as in “2010 bottles of beer” and in street addresses).
Almost all the writing on the pronunciation of 2010 is aimed at picking one variant as the “correct” one. That is, the writers generally subscribe to the One Right Way principle (probably without having formulated it), in roughly the following form: if no meaning difference can be discerned between expressions, then variation must be prohibited, with one expression designated as the one right way to express this meaning.
Eventually, a few of the commenters on Mark’s posting interrupted the passionate pursuit of One Right Way by objecting to the assumption. Why not, they asked, just tolerate variation? People should be free to choose whichever variant they like, in fact to choose different variants on different occasions. (It’s not like people have trouble understanding the variants.)
Then we get a comment from Mackenzie about the choice between “two thousand ten” and “two thousand and ten”:
That there’s no “and” should be obvious. “And” is only used in numbers when a decimal is involved.
(For “decimal”, read “fraction”, since the putative generalization here is supposed to cover things like 2 1/2, read as “two and a half” and not as “two a half”.)
There are three notable features of Mackenzie’s comment. First, it assumes One Right Way with regard to the and/zero alternation in numerical expressions, including year names. Second, it makes the choice of this one way according to what I’ve called “rigid complementarity”, building on the fact that and is obligatory in certain contexts to insist that it is permissible only in those contexts (the “only” in Mackenzie’s comment is significant), with zero obligatory elsewhere. And third, Mackenzie offers the “only used in numbers when a decimal is involved” generalization as a bald assertion.
My rigid-complementity posting discusses the first two points and has further detail on the and/zero alternation. One relevant detail is that British speakers tend to prefer the and variant, American speakers the zero variant — the American preference probably resulting from the advice given by some American schoolteachers (serious usage handbooks make no recommendation for either variant, so far as I can tell).
When people make bald assertions like Mackenzie’s, if I have the opportunity to ask them how they know these things, I usually get one of two answers:
(a) [often in bafflement] Everybody knows that!
(b) That’s what I was taught in school.
These people usually have no interest in hearing about actual practice, even of elite writers and speakers. The Language Lords have spoken.
As for the original three-way choice for 2010 as a year name, David Crystal has suggested that “twenty ten” is likely to be the dominant choice, since it’s the shortest of the three. (But, of course, there’s no reason to insist that the dominant choice be the only acceptable choice; see my discussion of the tyranny of the majority as a principle of usage.)