Ergativity for Novices

[bradrn]

I’m sorry if this has already been addressed in this thread, but I couldn’t find mention of it. Does anyone know if the objects of adpositions in ergative-absolutive languages tend to be in the ergative case or the absolutive case? I would expect the absolutive to be used, since it is the default and includes the accusative, or perhaps for it to depend on the semantics of the adposition. This would only be apparent, I would expect, in languages that use affixes to mark case. Georgian and Basque seem to mostly use additional cases, but I’m curious about what is usual in languages that use the ergative or absolutive case.

Huh, interesting question! I don’t know the answer to this, and I don’t really have time to research this properly just at the moment, but here’s a couple of pertinent observations in the meantime:

• Adpositional complements tend to be marked; the more marked case in an ergative–absolutive system is the ergative, suggesting that they would be marked with the ergative.
• Ergative case is very often syncretic with peripheral cases such as instrumental and ablative.

So, although I don’t have any concrete examples, I strongly suspect that adpositions would be marked with the ergative case in such languages.

(Also, you say that the absolutive ‘includes the accusative’, but that isn’t quite correct; the absolutive case tends to correspond to the nominative case instead, since both are unmarked.)

[willm]

Huh, interesting question! I don’t know the answer to this, and I don’t really have time to research this properly just at the moment, but here’s a couple of pertinent observations in the meantime:

• Adpositional complements tend to be marked; the more marked case in an ergative–absolutive system is the ergative, suggesting that they would be marked with the ergative.
• Ergative case is very often syncretic with peripheral cases such as instrumental and ablative.

So, although I don’t have any concrete examples, I strongly suspect that adpositions would be marked with the ergative case in such languages.

(Also, you say that the absolutive ‘includes the accusative’, but that isn’t quite correct; the absolutive case tends to correspond to the nominative case instead, since both are unmarked.)

Thank you for your response! All of that makes sense. In particular, I didn't realize that adpositional complements were usually marked per se, and that explains why they're often in the accusative or another oblique case better than my conception of their being patients of some sort.

[bradrn]

In particular, I didn't realize that adpositional complements were usually marked per se …

I should probably add that I’m actually not very sure about this myself; I’m just generalising (i.e. guessing) based on what little I know of how they behave, in particular that they usually take some sort of marked non-core case. (As I said, I don’t have much time to research just right now, so my last post was mostly guesses, hopefully correct ones.)

… and that explains why they're often in the accusative or another oblique case better than my conception of their being patients of some sort.

The prototypical patient is something which is being acted upon and undergoes a change of state, as opposed to the prototypical agent, which is an actor which is in control of an action. From this point of view, it is difficult to define peripheral arguments as being prototypical patients or agents: often, they are as much actors as they are acted upon. So no, peripheral arguments are not patients in any straightforward sense. (Indeed, some peripheral cases are agentive enough that they have a notable propensity to turn into ergatives! Hmm, another reason why adpositions would take the ergative case? But that argument falls down when you note that they also take the accusative…)

[willm]

Your explanation makes a lot of sense, including the ambiguity. I appreciate your taking the time to answer!

[Vardelm]

(Indeed, some peripheral cases are agentive enough that they have a notable propensity to turn into ergatives! Hmm, another reason why adpositions would take the ergative case? But that argument falls down when you note that they also take the accusative…)

That should just be a matter of where everything developed from. Since an ergative case can come from a genitive or instrumental case, I could see those cases expanding to be more of a general "peripheral" case: genitives especially. If adpositions developed from verbs ("to be at", "to be on", etc.) and the patient of those verbs was accusative, then the adpositions would develoe to use the accusative.

[bradrn]

(Indeed, some peripheral cases are agentive enough that they have a notable propensity to turn into ergatives! Hmm, another reason why adpositions would take the ergative case? But that argument falls down when you note that they also take the accusative…)

That should just be a matter of where everything developed from. Since an ergative case can come from a genitive or instrumental case, I could see those cases expanding to be more of a general "peripheral" case: genitives especially. If adpositions developed from verbs ("to be at", "to be on", etc.) and the patient of those verbs was accusative, then the adpositions would develoe to use the accusative.

No, that’s not quite what I was talking about… my argument was that some peripheral cases are more agentive due to their semantics, and this makes them more likely to be marked by the ergative than by the absolutive. But this argument doesn’t work, since they are also more likely to be accusative than nominative, which is inconsistent with labelling such arguments as ‘agentive’. So it looks like synchronically their marking has more to do with markedness considerations than with semantics.

[Vardelm]

No, that’s not quite what I was talking about… my argument was that some peripheral cases are more agentive due to their semantics, and this makes them more likely to be marked by the ergative than by the absolutive. But this argument doesn’t work, since they are also more likely to be accusative than nominative, which is inconsistent with labelling such arguments as ‘agentive’. So it looks like synchronically their marking has more to do with markedness considerations than with semantics.

Gocha. That makes sense. Marking is more abstract than semantics (as I see it) and so analogy might be used more liberally.