The question that I have is a simple one.

Of every neuter noun, pronoun, or adjective, in any IE language (as far as I know), the nominative and accusative cases agree in all numbers. Why is this?

In the IE languages, case syncretism seems to be a pretty common phenomenon. The dative/ablative plural in Latin is a well-known example, and there are further examples in Hittite and Sanskrit. However, as far as I know, these particular examples are usually (partly) explained in terms of usage overlap. That doesn't seem to be very likely here.

I tried to come up with a naive guess. My first (and only) idea was that neuter nouns often stand for inanimate objects, and these wouldn't ordinarily be construed as agents of voluntary or intentional actions. So it makes sense that, at least in certain contexts, nominatives of neuter nouns simply wouldn't come up. And browsing through Homer, I must say that neuter nouns are indeed rarely the subject governing a (transitive) verb, although examples of this happening do occur:

ὡς οὖν δεινὰ πέλωρα θεῶν εἰσῆλθ᾽ ἑκατόμβας

So when the terrible portents (neut.) had thus interrupted the hecatombs of the gods (Iliad, 2.321)

Having tried my luck in the literature, the only relevant observation I could find was in section 13.2.10 of Comparative Indo-European Linguistics: An Introduction by Beekes and De Vaan, who connect the phenomenon with the PIE (or pre-PIE?) ergative. Not being a linguist, I found their explanation too brief to be convincing. So if this theory represents a majority opinion of any kind, I would greatly appreciate some additional explanation.

  • I think your "naive guess" is probly the reason why it doesn't cause any problems, though not necessarily a reason why it was adopted in the first place. The fact that the syncretic suffix is -a in so many cases is evidence that it was present very early in I-E evolution. – jlawler Sep 29 '14 at 16:18
up vote 12 down vote accepted

I was going to write you an e-mail but I'll write my answer here instead ;)

First, most Indo-European scholars disregard the ergative hypothesis. However, I do not know any other reason for the equation of the neuter nominative and accusative. So, I'd like to present the reason why I believe that a prestage of PIE was an ergative language.

In most Indo-European languages the neuter nominative and accusative are identical. The Anatolian languages, however, are exceptional in this regard. In these languages we find that neuter nouns take a special ending -ants when they are the subject of a transitive sentence. This ending has been called the 'ergative'. However, there are different viewpoints on whether this actually is a case in the Anatolian languages or it is a 'personifying' suffix -ant- together with a nominative masculine/feminine ending -s. To answer this question one needs to dive deep into Anatolian philology. In my opinion, the answer is that there was in Proto-Anatolian a semantical suffix -ant-, that developed into a proper case ending -ants in the separate Anatolian languages.

The next question then is what the Proto-Anatolian case system looked like. In my opinion (I have not seen an account of someone exploring this perspective, but feel free to provide literature!) Anatolian would not have needed to create a separate ergative case for neuter nouns if there had been a case to give to neuter nouns in the Agent position (= subjects of transitive sentences). This is also what we find in older stages of Anatolian languages. Here the suffix -ant- has a semantic value, but there are no cases of neuter Agents in the 'nominative-accusative' case.

Keeping in mind that the first split in the Indo-European language family is commonly considered that between Anatolian and 'Common Indo-European' (the rest), we can reconstruct the following (MF = nonneuter, N = neuter, A = agent, S = subject (of an intransitive sentence), P = Patiens (object of a transitive sentence)):

  • Anatolian:
    • MF: A = S = -s, P = -m;
    • N: A = X, S = P = -m.
  • CIE:
    • MF: A = S = -s, P = -m;
    • N: A = S = P = -m.

So what do we reconstruct for PIE? There are two possibilities: either A = -m is original and Anatolian lost it, or PIE had a gap in the system which CIE restored. To my mind, the second possibility is more likely, since it is a very common development to fill a gap in a grammatical system, whereas it is very strange to create a gap in the system for seemingly no reason. One possibility to create the gap is that in PIE there was a close connection between morphological neuters and semantical inanimates. However, this relationship was not perfect in PIE (for example kweklos 'wheel'), and we find no comparable behaviour in nonneuter nouns.

So (at least in my opinion, which, again, does not represent the majority of scholars) in PIE neuter nouns did not have a form in the A-position. We have to explain why this is the case. In my opinion, the cleanest way to give an explanation why a nominative-accusative system would even make a distinction between A and S is because it developed from an earlier ergative-absolutive system.

In the case of Indo-European, we can even sketch a quite detailed scenario for this development. We start from a Pre-PIE ergative language with ergative marker -s (which would have agreed with the genitive, as the Hittite evidence shows) and an absolutive marker -m. There were animate and inanimate nouns, but there was no morphological distinction between the two. Then at one point the system changed to a nominative-accusative alignment pattern. For the accusative all nouns took the old absolutive ending -m. The nominative was more complicated. Animate nouns generally occured as the subject of transitive sentences, so they took their nominative from there, and they became the masculine words in PIE. Inanimate nouns, however, could by their nature not occur in the A position, and as such took their 'nominative' from the absolutive -m. It was, however, not a true nominative as it could not be used in the A position. This generalisation was only mate after Anatolian split off.

This is my view on how PIE developed: the 'nom = acc' for neuter nouns in Common IE and the Anatolian ergative are both solutions to fix a gap in the PIE declension system, that came into being as an original ergative system partly changed into a nominative system. This is reasonably close to the Leiden point of view, but other schools of Indo-European linguistics might differ. I'd be interested to hear any other points of view.

  • +1. Thanks a lot Milan! Indeed it resembles the picture sketched in the Beekes/De Vaan book, but your explanation is much more thorough and clear. :) – RP_ Sep 30 '14 at 8:20
  • +1. You should develop on that further and publish it, looks decently believable and you might actually be on to something – Darkgamma Sep 30 '14 at 19:25
  • @Milan This is fascinating. Could this be linked to the way causative verbs are formed in PIE? – geodude Nov 20 '14 at 13:42

Your Answer

 

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.