"an academical village" is at least unambiguously stupid
Teo has put up an interesting post discussing the use of a/an and how it evolved along with pronunciation (complete with bonus UVA tie-in!). Misuse of a/an is a pet peeve of mine — when I come across "an university" or a similar formulation I immediately know that what I'm reading is going to become terrible, if it isn't already. It's a great tip-off that the author hasn't made a connection between writing and speaking, and that spells trouble — a stream of "whilst"s is rarely far behind.
On the other end of the spectrum, of course, is Yglesias's neverending cavalcade of homophones — a case where speaking and writing are so tightly connected that all kinds of amusing minor typos result (see also: my constant overuse of colons, semicolons, emdashes and commas). I suppose you can err too far in either direction, but Matt's is by far the more readable option.
Of course, none of this explains the dismayingly frequent "an historic" formulation. I can't account for that one at all, except to say that I think our species is probably doomed.
Comments
Dude, "an historic" is totally correct. "Haitch," as the English say, isn't really a consonant. I think the use of "an" with historic is justified by your speech-writing connection notion; the alternative sounds stupid.
That said, A Stream of Whilsts is a marvelous title for something. An indie rock song maybe.
I guess it may depend on where you grew up and how hard your H's are. If you pronounce it with a very soft one ("it's 'istoric, guv'nah") I can see throwing in an "an". But I say it with a harder one, like "hysterical" or "hysterectomy". The use of "an" there sounds really weird to me.
"The Hysterical Hysterectomies" would also be an marvelous title for a indie band.
The answers to these questions can be found in Garner's A Dictionary of Modern American Usage.
Although I don't have a copy handy, my recollection of the "a" vs. "an" dilemma when preceding "historic" depends on pronunciation and dialect. The question turns on whether the Standard English speaker/writer aspirates the "h". Because the vast majority of English speakers aspirate the "h" in "historic", "a historic" is the answer for Standard Written English. And that is why "an historic" looks and sounds strange unless you have a certain British accent (as illustrated in your comment above).
Found a copy, Garner says sound rather than letter governs which indefinite article is proper.
From Garner:
People worry about whether the correct article is a or an with historian, historic, and a few other words. Most authorities have supported a over an. . . .
The theory behind using an in such a context is that the h- is weak when the accent is on the second rather than the first syllable (giving rise, by analogy, to an habitual offender, an hallucinatory image, and an hysterical crowd). . . .
Anyone who sounds the h- in words of the type here discussed should avoid pretense and use a. An humanitarian is, judged by even the most tolerant standards, a pretentious humanitarian.
I'm not one to judge Garner, but the example he gives in his kicker is weak. Not only is the stress in the first syllable, but the h-vowel combination is not like the other examples he's given: hümanitarian versus hysterical. The first is much harder. It's much easier to say "an habitual" than "a habitual" while it's much harder to say "an humanitarian" than "a humanitarian". There are two standards 'cause too make sense.
But don't get me started on people from Houston—they don't pronounce Hs at all.
Oh, Jesus: "two".
Well, it might all be relative, Kriston. IOt's only easier to say "an habitual" if you drop the H -- anabitual. Otherwise I think it flows smoothly into "Ahab", which is perfectly pronounceable.
Hey guys. The deal with the h words, as we discussed at Unfogged a while back, is that the h is dropped in some dialects, making the words vowel-initial, and even in other dialects weakened enough when unstressed for "an" to be used. As Kriston points out, though, this doesn't work for "humanitarian" because h is one of the contexts where the palatalization rule I discussed in my post applies, meaning that even if the h drops out the word is still not vowel-initial but begins with [y] instead.