While I’m on blogging hiatus, here’s some research

You might have noticed that I’m on a bit of hiatus this month. I’m working on my dissertation and preparing applications for post-grad-school jobs, but luckily something I’d done a little while ago has come through the pipeline for me to share.

Back in June, I presented a portion of my dissertation research at the NAACL-HLT conference, but all I had at the time was a computationally-dense paper to show you.  Well, the conference has uploaded videos of all the presentations, so if you’re interested in what I actually do academically, you can find out.  In short, this portion of my research is about the improvements in word segmentation that happen when you combine multiple types of information instead of using a single type.  It’s a computational model of how infants could use additional information to learn words better, as well as learning the likely stress patterns for words in the language they’re learning.

The video [20 min, plus questions that are unfortunately hard to hear]

I tried to make the talk approachable to the non-specialist, so take a gander if you want to see some of my dissertation research (which, of course, is pretty far afield from the discussions on this blog).  There will be math, too, in case you are a specialist, and if you want the whole story, you can see the paper that accompanies the talk.

In other news, I’ll be giving a talk on some of my new research showing how Twitter can be used to map the range of dialectal syntactic variants (e.g., double modals like might could and the needs done construction) at the LSA annual meeting in Minneapolis on January 3.  Check out the abstract here, and maybe I’ll see you there!

Rejecting a form that’s too true? A new complaint against “one of the only”.

People pop in fairly regularly to complain about “one of the only”, which I’m just really not that interested in. Usually the complaints are in response to my argument a few years ago that it was perfectly grammatical and interpretable (specifically rebutting Richard Lederer’s silly claim that only is equivalent to one and therefore is inappropriate for referring to multiple items). I haven’t gotten as many only=one complaints lately, but I’ve now received a new objection, presented as part of a comment by Derek Schmidt:

When [only] precedes a noun used in plural, it implies that there are no other similar items that belong to the list. “The only kinds of writing utensils on my desk are pencils and pens and highlighters.” […] But I have many of those pens, so if someone asked if they could borrow a pen, and I said, “No, that’s one of the only writing utensils on my desk!” that would be a little disingenuous and if someone was standing at my desk and saw the number of writing utensils, they would be baffled and think me a fool. Rightly so. Because they would understand it (logically, even) as meaning “that’s one of the few”, which is very false. So… “one of the only” means about as much as “one of them”.

To buttress his point, he referred me to a grammar column in the Oklahoman, which I never grow tired of noting was once called the “Worst Newspaper in America” by the Columbia Journalism Review. That was 14 years ago now, and I sometimes wonder if it is fair to keep bringing this up. Then I read Gene Owens’s grammar column in it and I wish the CJR had been harsher.*

About one example of “one of the only”, Owens writes:

“Now I can understand if he were the only English speaker or if he were only one of a few English speakers,” Jerry said, “but I don’t know how he could be one of the only English speakers.” That’s easy, Jerry. If he was any English speaker at all, he was one of the only English speakers in the area. In fact, he was one of the only English speakers in the world. […] The TV commentator probably meant “one of the few English speakers in the area.” But even if the colonel was “one of the many English speakers in the area,” he still was one of the only ones.

It continues on in this vein for a while, and but his point seems to be approximately the same as Schmidt’s, boiling down to the following statements:

• It is grammatical to say “one of the only”.
• It is used regularly in place of “one of the few”.
• Examining it literally, one could say “one of the only” to describe something that there are many of.
• This would be a strange situation to use it in.
• Therefore “one of the only” oughtn’t be used in the case where it wouldn’t be strange.

Up till the last sentence, I agree. In fact, I don’t think any of those points are controversial.** But the last sentence is a big leap, and one that we demonstrably don’t make in language. Would it be silly of me to say:

(1) I have three hairs on my head.

Thankfully I’m still young and hirsute enough to have many more than three hairs on my head, and I think we’d all agree it would be a silly statement. But, parsing it literally, it is true: I do have three hairs on my head, though in addition I have another hundred thousand. In case this is such a weird setting that you don’t agree it’s literally true, here’s another example:

(2) Some of the tomatoes I purchased are red.

If I show you the bin of cherry tomatoes I just bought, and they’re all red, am I lying? No, not literally. But I am being pragmatically inappropriate — you expect “some” to mean “some but not all”, just as you expect “three” to generally mean “three and no more”. These are examples of what’s known as a scalar implicature: we expect people to use the most restrictive form available (given their knowledge of the world), even though less restrictive forms may be consistent too.***

To return to Schmidt’s example, it may be truthful but absurd to protest that one of 30 pens on my desk is “one of my only pens”. But just because the truth value is the same when I protest that one of two pens on my desk is “one of my only pens”, this doesn’t mean that the pragmatic appropriateness doesn’t change either. Upon hearing “one of the only”, the listener knows, having never really heard this used to mean “one of many”, that pragmatically it will mean “one of the (relatively) few”.

There is, perhaps, nothing in the semantics to block its other meanings, but no one ever uses it as such, just as no one ever says they have three hairs when they have thousands. This is a strong constraint on the construction, one that people on both sides of the argument can agree on. I guess the difference is whether you view this usage restriction as evidence of people’s implicit linguistic knowledge (as I do) or as evidence of people failing to understand their native language (as Schmidt & Owens do).

Finally, and now I’m really splitting hairs, I’m not convinced that “one of the only” can always be replaced by “one of the few”, as the literalists suggest. If we’re being very literal, at what point do we have to switch off of few? I wouldn’t have a problem with saying “one of the only places where you can buy Cherikee Red“, even if there are hundreds of such stores, because relative to the number of stores that don’t sell it, they’re few. But saying “one of the few” when there’s hundreds? It doesn’t bother me, but I’d think it’d be worse to a literalist than using “one of the only”, whose only problem is that it is too true.

Summary: If a sentence could theoretically be used to describe a situation but is never used to describe such a situation, that doesn’t mean that the sentence is inappropriate or ungrammatical. It means that people have strong pragmatic constraints blocking the usage, exactly the sort of thing that we need to be aware of in a complete understanding of a language.

—
*: I am being unfair. Owens’s column is at least imaginative, and has an entire town mythos built up over the course of his very short columns. But I never understand what grammatical point he’s trying to make in them, and as far as I can tell, I’d disagree with it if I did. As for the “worst newspaper” claim, this was largely a result of the ownership of the paper by the Gaylord family, who thankfully sold it in 2011, though the CJR notes it’s still not great.

**: Well, it might be pragmatically appropriate to use “one of the few” in cases where the number of objects is large in absolute number but small relative to the total, such as speaking about a subset of rocks on the beach or something.  I’m not finding a clear example of this, but I don’t want to rule it out.

***: Scalar implicatures were first brought to my attention when one of my fellow grad students (now a post-doc at Yale), Kate Davidson, was investigating them in American Sign Language. Here’s an (I hope fairly accessible and interesting) example of her research in ASL scalar implicature.

My research at NAACL 2013

You may have noticed that I’ve been being quite bad about updating the blog the last couple months. I’m sorry for my negligence, and I’m hoping summer will leave me with a bit more time to keep up the blog. But the reason I’ve been remiss is that it’s time to really batten down the dissertation hatches, and boy, that doesn’t leave the time or energy for much else.

Tomorrow morning, the battening of said hatches pays off a little bit, because I’ll be presenting a portion of my dissertation research at the North American Chapter of the Association for Computational Linguistics (NAACL)’s conference. I don’t imagine too many of you are attending the conference, but if you are, I’ll be presenting in Session C tomorrow (i.e., Monday) morning at 11:55, so swing on by.

If you’re not down here in balmy Atlanta, you can always read the paper in the comfort of wherever you are. It looks at how an infant learning a language can combine syllable identities and stress patterns to segment words within the language they’re learning. I’ll warn you, it’s a lot less accessible than the stuff I write here, but it’s the actual computational psycholinguistic research that I earn my keep with, so I hope you’ll give it a look if you’re interested in such things.

The subjunctive might be dying, if you ignore where it’s going strong

If you believe the grammar doomsayers, the English subjunctive is dying out. But if this is the end of the grammatical world, I feel fine — and I say that even though I often mark the subjunctive myself.

The most talked about use of the subjunctive is in counterfactuals:

(1) Even if I were available, I’d still skip his party.

For many people, marking the subjunctive here is not required; either they never mark it, using the past indicative form was instead, or they (like me) sometimes mark it with were, and sometimes leave it unmarked with was. For this latter group, the choice often depends on the formality of the setting. I’m calling this “not marking” the subjunctive, rather than “not using” it, because it seems less like people making a choice between two moods for the verb and more like a choice between two orthographic/phonemic forms for it.

It’s similar to the alternation for many people (incl. me) of marking or not marking who(m) in the accusative case, discussed by Arnold Zwicky here and here, and Stan Carey here. That said, I believe that (at least some) people who never use were in (1) do not have a grammatical rule saying that counterfactuals trigger the past subjunctive, and I’m not worried about that either.

For being such a foolish war, World War I did generate some artistic propaganda.

This blitheness about the subjunctive does not go unmourned. I recently found myself being Twitter-followed by someone whose account just corrects people who fail to use the subjunctive in sentences like (1).* And Philip Corbett, associate managing editor for standards at the New York Times, annually rants about people failing to mark the subjunctive. Consider one of Corbett’s calls to man the ramparts, which he begins by quoting, in its entirety, a 90-year-old letter complaining that the subjunctive must be saved from impending destruction.** Corbett continues:

“[…] despite my repeated efforts to rally support for [the subjunctive] the crisis has only grown. For those few still unaware of the stakes, here is a reminder from The Times’s stylebook”

What are the stakes? What would we lose without the subjunctive? Corbett cites sentences such as these:

The mayor wishes the commissioner were retiring this year.
If the commissioner were rich, she could retire.
If the bill were going to pass, Secretary Kuzu would know by now.

If these were the stakes, I’d ditch the subjunctive. Corbett points out that in each of these we’re referring to a counterfactual condition, which should trigger the subjunctive. But note that using the indicative/unmarked was doesn’t make that any less clear. There is nothing to be gained from using the subjunctive in these cases but a sense of superiority and formality. (Not that I’m against either of those.)

But here’s the weird thing: all this defense of the subjunctive, all these worries — they’re all only about the past subjunctive. And the past subjunctive is weird, because it’s only marked on be, and it’s just a matter of using were for singular as well as plural. For everyone worrying that this is some crucial distinction, please note these sentences where it is insouciantly the same as teh indicative form:

(2a) The mayor wishes the commissioners retired last year.
(2b) If the commissioner wanted to, she could retire.
(2c) If the bills were going to pass, Sec. Kuzu would know by now.

If anything, the loss of past subjunctive were strikes me as regularization of English, the loss of the last remaining vestige of what was once a regular and widespread marking system. Losing the past subjunctive makes English more sensible. I don’t see that as a bad thing.

And anyway, the subjunctive probably isn’t going to disappear, not even the past subjunctive. The past subjunctive is, to my knowledge, necessarily marked in Subject-Auxiliary Inversion constructions:

(3) Were/*Was I a betting man, I’d say the subjunctive survives.

A quick look at Google Books N-grams makes it look like were subjunctive marking has been relatively constant over the last 40 years in written American English, so maybe this is all just a tempest in a teacup.

Plus all of this worry about the subjunctive ignores that the present subjunctive is going strong.*** I’ve written about sentences where the present subjunctive changes the meaning (though I wrote with a dimmer view of the subjunctive’s long-term prospects), and Mike Pope supplied an excellent example:

(4a) I insist that he be there.
(4b) I insist that he is there.

In cases where marking the subjunctive is important, it’s sticking around. In cases where it isn’t important, and the subjunctive follows a strange paradigm, identical to the indicative for all but one verb, it may be disappearing. This is no crisis.

Summary: People who write “if I was” instead of “if I were” aren’t necessarily pallbearers of the English subjunctive. It may be regularization of the last remaining irregular part of the past subjunctive, with the present subjunctive remaining unscathed. And if the past subjunctive disappears, there will be, as far as I can tell, no loss to English. Go ahead and use it if you want (I often do), but to worry that other people aren’t is wrinkling your brow for nothing.

—
*: I do respect the tweeter’s restraint in seemingly only correcting people who’re already talking about grammar.

**: That this destruction has been impending for 90 years has somehow not convinced the ranters that their panic may be misplaced. Also, Corbett keeps titling his posts “Subjunctivitis”, which I think sounds great, but not in the same way he probably does. -itis usually means an unwelcome inflammation of the root word, and I can’t help but see all this as an unhelpful inflammation of passions over the subjunctive.

***: In fact, and I think this is pretty cool, (Master!) Jonathon Owen directed me to a classmate’s corpus work suggesting that for at least some verbs, marked subjunctive usage is increasing.