Cases in agglutinative languages

Discussion in 'Etymology, History of languages, and Linguistics (EHL)' started by francisgranada, Apr 24, 2013.

  1. francisgranada Senior Member


    In the IE languages the term "case" is more or less clear. There is a relatively small number of cases that are used with or without prepositions. Moreover, the case endigs have different forms according to paradigms and not to the fuctionality (eg. -a is of "muž", but also of "víno" in Slovak).

    In the agglutinative languages like Hungarian, Turkish ... there are no prepositions, but a relatively high number (about 20-28 in Hungarian) of "endings" (or better morphemes attached to the noun, adjective...). A particular "ending" has always the same meaning/function (regardles of paradigms). Besides of these "endings", there is a plenty of postpositions that are written (and pronounced) separately from the preceding noun, otherwise they behave similarly as those "endings".

    My question: is it linguistically correct to speak about cases in this kind of languages? If yes, then how many cases are there in Hungarian: 28 (according to the "endings") or only a subset of those 28 or also constructions with postpositions are to be consedered separate cases? In other words, what is the criterion for defining a "construction" as grammatical case?

    Thanks in advance for your opinions!
    Last edited: Apr 24, 2013
  2. myšlenka Senior Member

    it's difficult to find syntactic criteria to distinguish between postpositions and grammatical case, but you can use phonological criteria to distinguish between the two. Grammatical case is considered to be more tightly connected to the noun than the postposition. In the languages you mention, Hungarian and Turkish, there will be vowel harmony between the noun and the case ending. This does not apply when you have a postposition. For other languages, there will perhaps be other various sandhi phenomena, stress assignment rules and other word internal/external processes that give you the answer.
  3. francisgranada Senior Member

    I agree with you, but is it enough? … From a strictly morphological point of view we could say that when it's written together and the vowel harmony takes place (in Hungarian and Turkish, as examples), than it is a "case". But ...

    1. There are also "case markers" (at least in Hungarian) that have only one form, e.g. -ig (terminative). I.e. some "case markers" seem to be immune to the vowel harmony (whatever be the reason for it, phonetical or etymological or something else).

    2. Some (maybe all in the very past) "case markers" were originally postpositions. A documented example: the word útra ("onto road") contains -ra which is the marker of sublative, and út is "road". The first written occurrence is utu rea from 1055. Now, is it correct to say that the sublative case did not exist in Hungarian about a thousand years ago?

    3. Some cases are considered "not proper cases" by some grammarians, for example -szor (multiplicative), even if they are written together with the noun (or number) and they are affected by the vowel harmony (egyszer=once, háromszor=three times). What is the true criterion for declaring them "not proper cases"? Or, in other words, which are the proper cases?
    Last edited: Apr 24, 2013
  4. Ёж! Senior Member

    I think this all depends on why do you talk of cases in the first place. If one wants to point at some parallel between usage of nouns in fusional and agglutinative languages, then why not use the word 'case' (after having clearly defined the meaning of the word either explicitly or implicitly, of course); if one wants to draw distinction, then again he could use some word different than 'case' after having defined it — why not. Why exactly strive for the same terminology in all… er… cases? ;)
  5. berndf Moderator

    German (Germany)
    I know next to nothing about Hungarian, so I can only report what I've read: Some grammarians seems to distinguish between "proper cases" and other postpositions. "Proper cases" are produced by postpositions (including the null-marker for nominative) that play a role in marking arguments of the preducate as distinct from its adjuncts in a valence-theoretical context.

    To give you an example in English: He waited for me in a bar. Here for me is a prepositional object, i.e. an argument, and in a bar is an adjunct. This is so, because the verb of valence one, the intransitive to wait, is considered to constitute a different predicate than the verb to wait for somebody of valence two, while in a bar is additional information added to the predicate that does not modify the predicate itself.

    I don't know, if this makes sense to you, if you try to apply it to Hungarian.
  6. myšlenka Senior Member

    It doesn't really matter what the spelling conventions are. They are after all mere conventions so they could in principle be changed into Chinese signs. Hungarian would still be a case language.
    I have no knowledge about Hungarian, but is it possible to add a postposition after -ig?
    This is possible, but I am not in a position to know that for sure.
    Can you add -szor to something else than numerals? Nouns for instance?
  7. francisgranada Senior Member

    I agree.
    No. The general rule is: or one "case marker" or one postposition (they are equivalent from this point of view).
    This seems to be a contradiction to what you have stated before ("Hungarian would still be a case language ....")
    A good question. Not only to numerals, but not to whatever word. Possible words are all the numerals (of course), but also: many, more, less, few, ... etc. (maybe tomorrow I'll find also some nouns :) ...)
  8. Gavril Senior Member

    English, USA

    It's often said that English doesn't have cases, but that seems clearly wrong if "case" is defined as "phonological marking of grammatical role". What English really lacks is fusional case forms of the kind seen in Latin, Greek, Icelandic and so on. The same (or a similar) argument could be applied to agglutinative languages such as Finnish, Turkish etc.
  9. myšlenka Senior Member

    In coordinated structures, do you have to repeat -ig for both nouns? Or is one -ig enough for both of them?
    You can change the spelling conventions of Hungarian tomorrow, but it would still be a language with case. I don't see how me not knowing whether the sublative was a case 1000 years ago contradicts that.
    So, basically it attaches to quantifiers and not to nouns? Well, if that's the case, there's your answer :)
  10. francisgranada Senior Member

    Yes, it makes, also with the multiplicative I've used as example. By the way, I’ve read somewhere an other explanation for this particular case, i.e. that this -szor is a “formant” (suffix), not a true “case marker”. But this seems to me rather an etymological point of view and not functional …
    In this particular case may be. But this is not true for all the so called "improper" cases. I have chosen this example only to show that it is written together and is also affected by the vowel harmony. (My question is not especially about the "improper" cases, this is rather a “subquestion”).
    I'll answer with examples:
    Három szép nagy és új házig - Up to three nice big and new houses.
    (the case marker is only on "house", unlike in Latin, Slavic etc., where three, nice, big and new would be declined as well).
    Három házig és öt kertig - Up to three houses and five gardens.

    Ok. (You’ve said that “This is possible …”, i.e. “…to say that the sublative case did not exist in Hungarian about a thousand years ago”).
  11. myšlenka Senior Member

    Case is first and foremost a nominal category so if you have suffixes that don't attach to nominal stems, it would be an "improper" case.
    And again, case is mostly a nominal category so the spreading of case to the whole noun phrase in Slavic and Latin is not really relevant for Hungarian. Your second example reveals that you have to repeat the -ig for both nouns, i.e. you can't use only one -ig to cover them both. Consequently, -ig is a case suffix, not a postposition.
  12. francisgranada Senior Member

    A postoposition (from the Hungarian point of view) is a priori written separately, so the question might rather be: why postpositional structures are not considered “cases”? But thanks to your interesting posts, I think we have the answer: the postposition does not need to be repeated (at least in Hungarian):

    Három ház és öt kert mögött - Behind three houses and five gardens.

    (at the moment I don’t know if this is allways valid, but it seems to be a “good” criterion)
  13. myšlenka Senior Member

    In Hungarian postpositions and case suffixes are post-nominal but they have different syntactic and phonological properties. They don't behave the same way. Consequently, postpositions are not considered cases because they don't behave like cases :)
  14. francisgranada Senior Member

    Yes, indeed that's what I have not "noticed" before. From this point of view you are right about utu rea (post 6#): without more informations we are not able to decide whether rea was "already" a case "suffix" (even if written separately) or "still" a postposition ... Thanks for your answers :).
  15. myšlenka Senior Member

    Glad to be of help in a language I don't speak :)

Share This Page