Monday, January 15, 2007

They are just wrong

OK, we all know the story on singular they. Those evil grammar nazis are trying to force everyone to write their sentences with his or her and not use they to refer to words like everyone, noone, etc. But fearless descriptivists stand up to this nonsense shouting' "let my people be!"

Did you know that using they to refer to singular antecedents is, like 400 years old. Man, if it's good enough for Shakespeare then it's good enough for everyone. Whoo-hoo let's get the language party started!

Except, I don't really like that narrative. The whole prescriptivist argument is based on the claim that a pronoun must, MUST I TELL YOU, agree with it's antecedent.

A pronoun must agree with its antecedent in number and gender. If the antecedent is singular, the pronoun must be singular; if the antecedent is plural, the pronoun must be plural. If the antecedent is masculine, the pronoun must be masculine; if the antecedent is feminine, the pronoun must be feminine.

So it sounds like the descriptivists are just tossing the rules out the window. The real question is, is it really true that pronouns must agree with antecedents? Is it true for all pronouns in all languages or just the English ones? Or what? And what does it mean if it's not true?

Hey I know, let's take a different language at random, like, um how about Swedish? Must Swedish pronouns agree with their antecedents?

Turns out only sometimes. Here's what you need to know--in Swedish, nouns are marked for both number (singular and plural) and grammatical gender. There two grammatical genders in Swedish--common and neuter. (I know, shouldn't that be masculine and feminine? Apparently not.)

Also in Swedish, if the antecedent is a subject of a sentence and the pronoun is within that sentence, then there is a special pronoun you have to use, sig. Sig doesn't change form for different antecedents. It's the same whether the noun is common, neuter, or plural.

Peter slog sig i huvudet.
'Peter hit himself in the head.'

Lotta slog sig i huvudet.
'Lotta hit herself in the head.'

Barnen slog sig i huvudet.
'The children hit themselves in the head.'

But, sig comes in many cases, including a genitive (possesive) case which does change form based on the gender or number, sin/sitt/sina, for common, neuter, and plural nouns respectively. But, the funny thing is, it doesn't agree with the antecedent. It agrees with the noun that it is the possessor of!


Max älsker sin mamma.
'Max loves his mother.'

Nora sover i sitt rum.
'Nora is sleeping in her room.'

Martin läser sina böcker.
'Martin is reading his books.'

So there you have it. Pronouns don't ALWAYS agree with their antecedents. At least not in every language. So it' not crazy to think the same may be true in English.

But clearly, sometimes they do. I mean "Mary loves himself" is not English. And that is why I don't like the normal narrative of freewheelin' descriptivism versus dour prescriptivism. Because there is something interesting to learn about pronouns, antecedents, and agreement in the world's languages and English in particular that just gets papered over if we just kick back and get the language party started.


parsnipgirl said...

As a semanticist I feel I have to point out the obvious: the whole ruckus is over agreement with *quantified* antecedents. Interestingly, this probably relates back to our earlier discussion of number agreement with group nouns -- the rules are just not that clear and maybe that's part of the reason people are willing to use "they/their/them" with an allegedly singular antecedent. If we say something about "every student", we're making a statement also about the group of students, hence "they/their/them" seems ok. Maybe. But wouldn't want the grammar nazis to have to think about anything complicated and interesting like quantification and how it relates to thinking about the domain of the quantifier as somehow being a "group".

liberal elite said...

Actually, Pinker makes that point in The Language Instinct. His analysis is that because the pronoun is quantified over it's a different relationship than pronoun antecedent.

Do you know of any natural language that actually uses two different pronouns for these two relationships?

parsnipgirl said...

No, I have never heard of any language doing that.

alienvoord said...

Language parties are the best kind of parties.

Site meter

Search This Blog