The prescriptivists are on my last nerve.  Some of them really believe that there is something wrong with this sentence:

(1) One of the only things I liked about living in Ottawa was the strong film community.

Reasonable readers, can you find the error in (1)?  The construction that “doesn’t convey any information”, the one that Richard Lederer calls a “strange and illogical expression”, the one Robert Hartwell Fiske cites as “further evidence that people scarcely know what their words mean”? Give up? It’s one of the only!

Oh, you didn’t find that to be illogical?  You thought you got some information out of those words?  Well then, congratulations; you’re a normal speaker of English.  Honestly, I couldn’t see what could the problem with one of the only possibly be.  Well, let’s look at Lederer’s argument against it:

“This strange and illogical expression began showing up a few years ago, and English took a step backward when it did.  The expression has been defended on the basis that it is no worse than only two, because only means ‘one’ and only two is oxymoronic. A specious argument! It’s like saying that robbing a bank is okay because it’s no worse than robbing a jewelry store.  Moreover, only in the sense of ‘only two’ does not mean ‘one’; it means ‘no more than.’  There is no meaning of only that fits with one of the only.

Well, that’s a kick in the gut of the facts — three kicks, in fact.  Kick the first is the claim that one of the only started showing up a few years ago.  Google Books reports it in two books around 1770, in The Dramatic Censor and The Sale of Authors, and reports hundreds of uses throughout the nineteenth century.  It’s more than a few years old, that’s for sure.

Kick the second is the idea that any reasonable person defends one of the only by noting that only two is oxymoronic.  I sure don’t, and I don’t understand who would.  There is nothing oxymoronic, nothing contradictory about the construction.  Only two is completely clear, comprehensible, standard, and logical — hundreds of pre-1800 usages of only two in Google Books attest to this.

Kick the third is Lederer’s definition of onlyOnly two does not mean “no more than two” in standard usage.  If it meant “no more than two”, then (2) would be a totally acceptable sentence.

(2) *The cyclops has only two eyes.

With Lederer’s definition (2) is fine, because a cyclops has only one eye, and one is no more than two.  But a quick poll of the only two people in the apartment at the moment revealed that (2) is utterly unacceptable; clearly Lederer’s definition is insufficient.  The real definition of only in only two is something along the lines of “exactly”, but with the crucial additional implicature that this is a smaller number than expected.  Violating this implicature makes a sentence sound weird, as with (3b):

(3a) I was sad when only two people showed up at my cats’ wedding.
(3b) #I was sad when only one thousand people showed up at my cats’ wedding.

Now, the fact that one gets this implicature, that only two sounds so much better than only one thousand, ought to suggest that there is logic underlying the construction. This, coupled with Lederer’s crummy definition of only, should lead a reader to be skeptical of his claim that no meaning of only can fit in one of the only. I am curious as to what Lederer thinks the definitions of only are.

So what does one of the only mean?  What happens if we follow one critic’s request to “parse it if you will, and see what you get”?  Let’s look at the example in (1).  The only things I liked about living in Ottawa is a noun phrase, identifying the set of things the speaker liked about living in Ottawa, noting that this set is the complete set, and implying that it’s an awfully small set.  That’s what the quantifier only means, that’s what it’s meant for hundreds of years.  One of modifies a noun phrase, selecting one member of that set.  The two combined, as they are in (1), pick out a single member of the set of all things the speaker liked about living in Ottawa. So what exactly were we supposed to see when we parsed this?  That it works?  I’m fine with that.

There’re a lot more arguments that one of the only makes sense, and Jan Freeman has a wonderful column with a few of them.  Notably, Freeman points out that one of the only is attested cross-linguistically, further destroying the notion that one of the only is somehow illogical.  So in the end, I have to ask this of the prescriptivists: Do you really have nothing better to do in your lives than to ignore the well-known meanings of words so that you get to call other people stupid?  Are you really unable to think of a better pastime than claiming that a reasonable, well-worn construction is illogical and incomprehensible?  Are you really so committed to those goals that you’re unwilling to comprehend an easily comprehensible construction?

Or as I screamed into my computer after reading this junk: Why are you spending more effort trying to misunderstand someone than trying to understand them?

Summary: Prescriptivists insistently grouse that people don’t think enough when they write, but prescriptivists seem just as likely not to think when writing.  Case in point: the arguments against one of the only are positively absurd, based off of a wanton misinterpretation of what only means, and completely independent of historical usage in English and other languages.  Of course one of the only is fine, a fact that has been known since 1770.