A Dip into Syntax: Ergativity

One of the reasons that Kaqchikel is a particularly interesting language to study from a linguistics point of view is because it displays a relatively rare type of sentence construction known as syntactic ergativity.

Here's what that means:

Kaqchikel, like English, has both transitive and intransitive sentences:

Examples:

• The boy hugged the girl = Transitive
• The action in the sentence ('hugged') has two arguments
• There is both an agent - the boy - and an object - the girl
• Therefore, the sentence and the verb are transitive

• The mare gallops = Intransitive
• The action in the sentence ('gallops') only has one argument
• There is only a subject - the mare
• Therefore, the sentence and the verb are intransitive

The other thing that is happening in these sentences is agreement. One way to check agreement is by replacing the arguments (the boy, the girl, the horse) in the examples with pronouns (he, her, she.) You would get the following sentences:

• He hugged her
• She gallops

This shows what you probably already know. In English, there are two sets of pronouns. One set (he, she, we, they, etc) is used for subjects (S) of intransitive sentences and agents (A) of transitive sentences. The other set (him, her, us, they, etc) is used for objects (O) of transitive sentences. Therefore, when S aligns with A (which both have nominative case) against O (which has accusative case,) the system of the language is accusative.

English is a nominative-accusative language.

However, in Kaqchikel, this is not the case. In Kaqchikel, there are also two types of alignments (though this is not expressed through pronouns, but markers on the verb.) But here, S aligns with O (which both have absolutive case) against A (which has ergative case,) which makes the system of Kaqchikel ergative.

Kaqchikel is an ergative-absolutive language.

So, if we were to imagine that English was an ergative-absolutive language that displayed this through pronoun alignment, the above examples would look something like:

• He hugged her
• Her gallops

Here's a commonly used visual to help explain the above:

Why is ergativity useful to linguistics? Well, in accusative languages like English, the grammatical function of 'subject' of a sentence (the intransitive S & the transitive A) is not separable from the case of the arguments (since the S & A share the same case = nominative.) In an ergative language the grammatical subject (S & A) does have different case marking (S = absolutive; A = ergative.) This allows the functions of grammar and case to be teased apart and tested separately - which cannot be done in English. 

Okay, great! But we are not quite done... The above was an explanation of morphological ergativity. Kaqchikel has morphological ergativity but it also has syntactic ergativity. So, what's that? 

I should warn you - here is where my understanding (and the general academic understanding) of this phenomenon weakens. The literature is sparser, but subsequently more provocative. 

Anyway, in order to talk about syntactic ergativity, we have to discuss extraction. 

In English, extraction can be seen in several different constructions. Two such are wh- questions and relative clauses. Some examples:

• John gave the teacher an apple - (di)transitive
• Wh-question: What did John give the teacher?
• Relative clause: The apple that John gave the teacher... 
• In both the wh- and relative clause, 'the apple' has been extracted from the first sentence and has been fronted to the beginning of the sentence. This shows that this segment of the sentence 'the apple' is one unit capable of retaining meaning even when it is moved out of its original place.
• Similarly, we can do this for the other arguments in the sentence:
• Who did John give the apple to? (Extracts 'the teacher' via wh-question)
• It was John that gave the teacher the apple (Extracts 'John' via relative clause)
• John laughs out loud - intransitive
• Wh-questions: Who laughs out loud?
• Relative clause: John, who laughs out loud...

So, in English, we can extract the S, A, or O. However, in a syntactically ergative language, the ergative subject, or A, cannot be extracted. Kaqchikel is one such language. In Kaqchikel, direct extraction of the A is avoided by using sentence constructions known as agent focus and antipassivization (which I won't go into now.) 

And there you have it! That's syntactic ergativity. And why is it of such interest to linguists? Mostly, that's because it is difficult to explain. There are a few theories for why it exists - perhaps ergative extraction is more difficult for speakers to process than absolutive, nominative, and accusative extraction? Or maybe there is some internal grammar at work that prevents the ergative extraction? This is a contested area without any prominent explanation so far, which makes it interesting to explore!

Hopefully this has been somewhat helpful in elucidating this area of linguisic theory, how it relates to Kaqhikel, and how it relates to my interest in the language. 

(Side note: So far, all known syntactically ergative languages are also morphologically ergative, but the opposite is not true. Therefore, syntactic ergativity is a subset of morphological ergativity.)

Matyox! Ke ri' k'a!

Kaqchikel word of the day: Ke ri k'a (ke ree K'A) = See you later, goodbye.

- k' is an ejective velar stop, made by sharply clicking the tongue down from the back of the palate.