#sloppy interpretation

Inour recent video about pronouns, we discussed how we figure out just what they mean, since on its own, “he” or “himself” doesn’t seem to mean anything. At first glance, it looks as if reflexive pronouns like “himself” always have to refer back to some other nearby noun phrase — their antecedent. More precisely, reflexive pronouns and their antecedents co-refer, which means they both pick out the same thing in the world.

     (1) Mark believes in himself.

But as we pointed out in this episode, that idea doesn’t really hold up when it comes to quantifying phrases, like “everybody” or “no soldier.” In the sentence below, the subject doesn’t really point to anything, but the pronoun still meaningfully connects up to it.

     (2) None of Mark’s brothersunderstandthemselves as well as he does.

A more complete picture of reflexives, then, is one where they’re always bound to some nearby noun phrase, whether it’s a name, a quantifying expression, or even another pronoun. They don’t necessarily have to refer to anything, but that connection always has to be there. In other words, they have to be co-indexed with their antecedent, even if they don’t have that reference link.

In contrast to reflexives, regular pronouns like “her” and “he” don’t always need to be bound to something nearby, or even something far away. They can either skip over intervening noun phrases to connect up to a distant subject, as in (3), or else refer to that same person by way of the speaker physically pointing them out, like in (4).

     (3) Delphine is worried about whether or not Cosima will forgive her.

     (4) She’s worried about Cosima.

Of course, interpreting some sentences requires that ordinary pronouns act a lot like reflexives; assuming “they” in the sentence below is understood to be picking out a portion of Sarah’s sisters, we’re forced to suppose that it’s co-indexed with — and so bound to — the quantified phrase “most of Sarah’s sisters.”

     (5) Most of Sarah’s sisters don’t even know they’re related.

The point is just that non-reflexive pronouns have a degree of freedom that reflexive pronouns don’t, and can maneuver in ways reflexives can’t. But this ends up raising an interesting question about certain kinds of sentences, like the one found in (6).

     (6) Alisonknewshe was in trouble.

In principle, the sentence above has two separate paths to arrive at the interpretation where “she” refers to Alison: either “she” is co-indexed with the noun phrase “Alison,” or else “she” is open to picking out whoever is poking out the most in the conversation at that point in time, which by coincidence happens to be Alison. That is, “she” is either bound to the subject, and so indirectly referencing Alison, or else freely referring to the person named Alison directly, by way of the surrounding context.

That might seem like a distinction without a difference, and in this case it very nearly is, since there’s no detectable change in meaning (i.e., in either case, “she” winds up referring to Alison). We might even wonder whether it’s necessary to eversuppose that regular pronouns are bound, save when following quantifying words. Maybe, in all other cases, non-reflexive pronouns are simply free, choosing sometimes to co-refer with some other part of the sentence, and sometimes not to. After all, it seems a bit redundant to have two equally valid ways of arriving at the exact same result.

But as it turns out, there’s actually something we can do to tease these subtly separate meanings apart from each other, despite how closely tied together they might appear. To see how, consider the sentence in (7).

     (7) Rachel went into hiding, because she had to.

Notice that the second clause is fairly easily understood to mean “because she had to go into hiding,” even though the verb phrase is nowhere to be seen. That’s because this represents an instance of ellipsis— the omission of part of a sentence when it’s clear what’s being cut out. In particular, when we encounter ellipsis, we always understand the missing material to be identicalto some nearby, suitably related string of words. After all, the second half of (7) can’t mean “because she had to stay safe,” as much as that might make sense. This identity condition on ellipsis helps us to quickly recover what’s left unsaid.

And though it might at first come across as counterintuitive, ellipsis is a surprisingly powerful way to get at a better understanding of the behaviour of pronouns. In fact, by making them disappear, we can actually end up clarifying how they work! Take a close look at this next sentence.

     (8) Dr. Leekie went to his office, and Dr. Nealon did too.

Beginning with the assumption that “his” refers to Dr. Leekie, and not some third party that hasn’t been mentioned, the first half of the sentence is understood to mean “Dr. Leekie went to Dr. Leekie’s office.” Now, if we had arrived at this meaning by having the possessive pronoun “his” freely pick out Dr. Leekie, simply because it was convenient, we’d expect the missing part of second half of the sentence to match this choice word-for-word. That is, we’d expect the whole sentence to mean “Dr. Leekie went to Dr. Leekie’s office, and Dr. Nealon went to Dr. Leekie’s office.”

But that isn’t the only meaning we get. The sentence can also mean “Dr. Leekie went to Dr. Leekie’s office, and Dr. Nealon went to Dr. Nealon’s office.” The whole thing is ambiguous, because the missing material can be interpreted in two different ways. And if we assume that the absent verb phrase must be identical with the one in the first half of the sentence, as we did in (7), that initial VP must be capable of carrying two slightly different interpretations: one where “x went to Dr. Leekie’s office,” and one where “x went to x’s own office.” So, the pronoun “his” can either act like a ‘true’ non-reflexive pronoun, and directly pick out Dr. Leekie, or it can act more like a reflexive pronoun, binding it to the nearest available subject.

The fact that the second clause (and so the sentence as a whole) has two detectable meanings lets us peer into the inner workings of that first clause. It tells us that the first half of the sentence has two subtly distinct interpretations, and that two unique mechanisms are at play, all because pronouns have the option of being either bound or free.

So a completely natural phenomenon like ellipsis can be co-opted and used as a tool, to help shed light onto the otherwise sticky semantics of pronouns, and provide us with even more evidence that reflexives and non-reflexives both fit into the same basic category — with a few differences, for sure, but also more in common than you might have thought.