Polysemy and Maturity

Screen Shot 2016-02-18 at 4.03.36 PM

Polysemous pachyderms (Tontan Travel)

Harvard is considering whether the long-established term “House Master” should be changed, because a group of Latino students has protested that it reminds them of the tradition of slave masters.

Professor Steven Pinker, of the same parish, made a Twitter comment on the controversy that has been retweeted hundreds of times already and deserves to be:

We should be teaching students: 1 All words have >1 meaning.. 2. Mature adults resist taking pointless offense.

I assume the second part scarcely needs to be defended. Potential occasions for taking offense occur all the time: We get jostled in a crowded store, a vehicle cuts into our lane ahead of us on the freeway, people misstate our rank, or fail to recall our name, or neglect to cite us when they could have, or look at us funny. Only a hot-tempered psycho bristles at all of the slings and arrows of outrageous fortune, every single time. Flying into high dudgeon over distantly related but clearly unintended meanings of familiar phrases is at best silly, and at worst unhinged.

But what about Pinker’s first proposition, that every word has more than one meaning? At first it seems incautious. Universal polysemy? Surely we can find at least one word that has just a single sense. But the more I think about it, the more I think his generalization is much closer to being true than to being false.

You might think that the dinky little grammatical words of English have unitary meanings. Not so. They generally have meanings that call for technical notations that you probably don’t want to look at, and they’re not monosemous. One elementary Russellian proposal for the meaning of the, a fairly standard one given in a set of lecture notes on formal semantics for philosophers, is this:

λP<et> . [λQ<et> . |P | = 1 ∧ PQ ]

(I did warn you that you might not want to see it.) And it does not work for all uses: In The moon has no atmosphere, the definite article has the above meaning, but not in The more the merrier.

Something similar holds for most other little words. All contributes a universal quantifier to the meaning of All operators are busy, so the sentence translates into the first-order predicate calculus thus:

(∀x) [ operator(x) → busy(x) ]

But that treatment doesn’t extend to the uses in That’s all the better for us, or The wall was all covered in graffiti, or You’re getting ketchup all over the floor. (It would be ridiculous pedantry to say that the latter can never be true unless every square millimeter of the floor is receiving a ketchup coating. All isn’t always a totality.)

No, the little grammaticized words are not the place to look for monosemy. It occurred to me at first that basic terms for what philosophers call natural kinds would yield better results; elephant, I thought, surely means only one thing. But no, that’s not true. First, there’s the ambiguity between Loxodonta africana (African) and Elephas maximus (Asian). And there are the totally different meanings in That’s the elephant in the room and The stadium is now a white elephant. Similar points apply to lion, tiger, monkey, and so on.

The more ordinary nouns and verbs and adjectives that we use each day are often polysemous to a degree that boggles the mind. What does see mean? At least 10 things: (1) understand (I see your point); (2) judge (See how it sounds); (3) experience (That’s not the way I see it); (4) find out (Let’s see what this violin sounds like); (5) date (Sam’s been seeing Maxine); (6) consult (You should see a doctor); (7) visit (Let’s go see Maria); (8) ensure (See that you don’t break it); (9) escort (Let me see you home); and (10) send away (See this person off the premises).

You can do the same sort of exercise yourself with other common words. You should be able to find a dozen meanings for post (try it, and then look here), or for charge (try it, and then look here). Likewise set, jack, count, bolt, and thousands of others.

I’m sure Pinker’s generalization cannot be totally exceptionless; he is exaggerating just a little. [Added later: A commenter below suggests that monosemous might be monosemous. That's a pretty good bet. Look for monosemy in rare technical terms introduced for just a single purpose and not sullied by general currency. Technetium, for example. But the moment a word gets out and is passed around a bit, new meanings start sticking to it like burrs to corduroy pants. Look at how people use exponential, for example.] So replace his “all” by “most” if you wish. But don’t overlook the fact that polysemy seems to be the natural state for most common words in English.

And don’t overlook his bottom-line point: It is not mature or sensible to take a highly polysemous word like master, and find an echo of slavery in it, and treat that echo as a microaggression, and take (pointless) offense, and call for Harvard to change an administrative technical term that has been in use for generations. He’s right about that.

Return to Top