Ergative English

As National Grammar Day approaches, I’ve been thinking about one way in which the grammar of some languages can be mind-bendingly different from the grammar of English. Specifically, I’ve been wondering what it would be like if English were an ergative language.

Imagine this. Imagine that in a sentence like She kissed me or I kissed her, the agent has the nominative she/I case form, and the patient has the accusative me/her case form.

Hmm. All right, I guess, you don’t have to imagine that, because that’s how English is already. But now also imagine this. Let’s take a sentence like She smiles. The way English is now, the subject of this intransitive verb has the same form, she/I, as the agent of the transitive verb kiss, and the patient of the transitive verb kissed is the odd one out, with its her/me case forms. I’ll sum it up in a list, with the matching case forms having the same color:

1. 1. She kisses me.
   2. I kiss her.
2. 1. She smiles.
   2. I smile.

Now suppose that instead, the subject of this intransitive verb has the same form as the patient of the transitive verb kiss, and the agent of the transitive verb kissed is the odd one out. It might look something like this:

1. 1. She kisses me.
   2. I kiss her.
2. 1. Her smiles.
   2. Me smile.

In an arrangement like this, the she/I forms would be called ergative case forms, and the her/me forms would be called absolutives. So that’s how English might look if it were an ergative language.

Eh. That’s not so mind-bendingly different. But it’s also not as ergative as we can make it. So far, we’ve grouped intransitive subjects and transitive patients together in having the same case forms, but what if we also made the verbs always agree with the absolutive noun phrase? Here are our four sample sentences again, this time with the verbs agreeing with the absolutes. To bring it out more prominently to the eye, I’m underlining the verbs and what they agree with:

1. 1. She kiss me.
   2. I kisses her.
2. 1. Her smiles.
   2. Me smile.

Now this is starting to look pretty weird. But it’s still not as ergative as it could be. At this point, we’ve just made English “morphosyntactically ergative”. When we replaced the nominative/accusative case forms with ergative/absolutive ones, that was just a change in the morphology of the pronouns. When we also changed the verb agreement rules, we started to get the syntax involved, and hence the term morphosyntactic. However, we could let this ergative/absolutive way of thinking infect the syntax even more deeply, and turn English into a thoroughly “syntactically ergative” language. The main way that this shows up (at least, the way it’s demonstrated in the sources I’ve read) is in coordination.

In ordinary English, you can combine sentences with like subjects the way I’ve done with our examples below:

She kisses me; she smiles. –> She kisses me and smiles.
I kiss her; I smile. –> I kiss her and smile.

The single, factored-out subjects She and I function as both the agent of kiss and the subject of smile. In ordinary English, that’s no problem, because both those things are considered to be subjects. What you can’t do is try to factor out a noun that’s the patient of a transitive verb and the subject of an intransitive. In other words, sentences like these next ones can’t be shortened, where the person getting kissed is the one who smiles:

She kisses me; I smile. [No way to factor out the me/I]
I kiss her; she smiles. [No way to factor out the her/she]

In syntactically ergative English, what you can and can’t do get reversed. If the same person is doing the kissing and the smiling, you can’t shorten the sentence:

She kiss me; her smiles. [No way to factor out the she/her]
I kisses her; me smile. [No way to factor out the I/me]

On the other hand, if the same person is getting kissed and smiling, you can factor that patient/subject out:

She kiss me; me smile. –> She kiss me and smile. [I’m the one smiling!]
I kisses her; her smiles. –> I kisses her and smiles. [She’s the one smiling!]

And these aren’t even all the possibilities. This presentation is based on information I got from Bernard Comrie’s Language Universals and Linguistic Typology, 2ed (1989), which also describes languages that are syntactically ergative while still maintaining nominative/accusative case forms, languages in which some verbs work ergatively and others don’t, and various other combinations and degrees of ergativity. But if you’re like me, just the stuff I wrote up here should have been enough to get your mind thoroughly bent out of shape.

Still, you may be wondering something. English does have a way to make patients of intransitive verbs serve as subjects: the passive voice. So do ergative languages have some analogous workaround for when you don’t want the patient of a transitive verb to serve as the thing that the verb agrees with? Tune in on National Grammar Day to learn about … the “anti-passive”!