What is ergativity in linguistics? (And other follow-up questions)

292 views

I’m not a bonafide linguist, but I’ve studied a dozen languages—some formally, some informally, some only on a surface level, some deep enough that I’ve needed to dip my toes into proper linguistics to go further. While I’m sure the iceberg goes deeper than I’ll ever know, the one topic I still haven’t gotten even a basic grasp on is ergativity, which is wild because *apparently* it’s a main feature of many languages. I would say I get the gist of it, but that’s not even true. At best, I know which parts of speech it involves, and I won’t even say that out loud because I may well be wrong.

So, my basic question is: What is ergativity?

From there:

– What are some examples of it in different languages?

– And what is its linguistic counterpart called, as in, if a language does not display ergativity, what *does* it have?

You don’t have to explain it like I’m five, more like I’m that guy at work who’s approaching retirement age that saw you scroll TikTok once and now wants you to explain what “griddying” is. I’m 27, but I think that method might be best.

Thanks!

In: 13

6 Answers

Anonymous 0 Comments

I’ll go ahead and explain it as simply as I can at first for anyone else who’s interested.

There are two *main* types of verbs: those that take one noun, and those that take two. The first are intransitives and they take subjects, S: I ran, I slept, I jumped, I sang. The second are transitives, and they take agents A and patients P: I(A) read a book(P), I(A) cooked some pasta(P), I(A) killed a mosquito(P), I(A) answered you(P)¹.

Fundamentally, most languages treat the intransitive S as the “default” role, either nominative or absolutive². Then, one of the two transitive roles – A or P – behaves exacts like it, while the other role acts in a unique way. In nominative-accusative (or just accusative) languages, the accusative acts a special way. In ergative-absolutive languages, the ergative acts a special way. Typically this is by requiring a special marker on the noun to show it’s the special role, or by failing to agree with the verb the way the default role does (oftentimes by still agreeing, but with a second set of affixes).

So in an accusative language like English, the intransitive S and the transitive A act alike: I ate, I ate X. This is the nominative role. The transitive P acts differently, X ate me, receiving its own unique pronoun form, the accusative. It also behaves this way in verb agreement: he eats, he eats them, both receiving the *-s* suffix because “he” the nominative of both, while as an object, they eat him, the verb ignores what role the 3rd person masculine is taking because it’s no longer in the default, nominative role that agrees with the verb.

In an ergative language, instead the intransitive S and transitive *P* act alike. Take Tzeltal, a Mayan language. The verb agrees with both persons, both the A and P. The 1st person markers are either *k-* or *-on*. When we have a 1st person S we have *bejen-on* “I walked.” When we have a 1st person P we also have *-on*, *a-maj-on* “you hit/beat me.” However if the the 1st person is A, it swaps to *k-maj-at* “I hit you.” The S and P roles match, both using *-on*, while the A takes its own special form, *k-*. That’s “opposite” English and most European languages, where A agrees with whatever S does, and P follows special rules.

In addition, the shared form tends to be *unmarked* and that makes it *privileged*. Typically it’s always present on the verb in agreement, and/or the form that takes no special suffix to show its role in the sentence. It’s the default. And as a result, it’s allowed to do things in the language the special, marked form can’t. For example, in English we can ask both “who saw it?” and “it saw who?”, but some languages are less permissive. You’re only allowed to ask the *default* role. If English, or any nominative-accusative language, had such a restriction, you could only ask “who saw it” or “who was seen by it?”, using a passive voice to rearrange how the sentence is structured in order to get the “who” into either S or A roles, the defaults ones. In an ergative-absolutive language with such a restriction, you could only ask “it saw who?” because P is the default role. Languages with such restrictions also tend to have an “antipassive” voice that turns a sentence like “I see it” into “I see to it,” which shifts “I” from the special A role of a transitive to the default S role of a (derived) intransitive.

You can get other things sometimes too. In a nominative-accusative language like English, “he saw her and ran” can *only* mean that *he* ran, because the A of “saw” is the same, default role (nominative) as the unstated S of “ran.” You couldn’t have the unstated nominative match up with the *accusative* in the transitive. In ergativity, though, “he saw her and ran” could *only* mean that *she* ran, because the P of “saw” is the same default role (absolutive) as the unstated S of “ran.”

It’s worth adding that most languages are only ergative in parts of the language. Most often, it’s just in case-marking on nouns to show their role in the sentence, and verb agreement and all the underlying syntax (like which roles can be questioned and who runs in “he saw her and ran”) is still nominative-accusative, matching English. Other times you get splits based on tense (past tenses are ergative, present/futures are nominative) or person (1st/2nd persons or all pronouns are nominative, 3rd persons or non-pronouns are ergative). Even among “very” ergative languages, they can differ in just how far it seeps into everything, and it might not *fully* exist in any language, as there’s a couple places that always have S and A correlate³.

¹ Strictly speaking, S A P are different from subject agent patient, SAP are talking about the syntax (the structure) and agent and patient are talking about semantics (how they’re actually acting relative to each other). It works in most cases, but run into problems in cases like “I like him,” which has an experiencer A and theme P, but no agent (something that effects change) *or* patient (something that is changed by the effect). There’s a reason these types of verbs tend to vary a lot between languages as to how they’re structured, as with German *du gefällst mir* or Spanish *me gustas (tú)*.

² You get other arrangements as well, where there is no default in one way or another, or where there is a default but it’s not used by the S role, but they make up a pretty small minority of languages (plus like half of Austronesian, including most of the languages of Taiwan+Philippines+Indonesia, which are often incorrectly thrown in the ergative-absolutive)

³ One that’s pretty easy to explain is imperatives. All languages have imperatives like “eat!” and “eat it!” having the unstated subject (if unstated) commanding the S or A role, there’s no “ergative imperative” where the command of “eat!” means “you, eat!” but “eat it!” means “it, eat you!” where the S and P roles are selected

You are viewing 1 out of 6 answers, click here to view all answers.