New CBB

I'm in the beginning stages of creating a conlang, and at this point, I'm playing around with different ideas to see how well they work. While I've decided not to apply grammatical genders to things that don't have actual genders, cases intrigue me. Aspects of my current conlang are based off Latin, and so I've run into cases quite a lot in my research. Thing is, as much as I like the diversity cases offer, I can't see what purpose they serve if you can communicate more or less the same information via context and syntax. What are the advantages/disadvantages to using cases in language?

Using cases frees the syntax to be used for other functions.

_________________
| | | (Nahuatl) | (Sanskrit & Pali) (Yaghan) | (Antregabua)
CrazyEttin's Lair

Gender and case need not have anything to do with eachother. Sure you can (and many european languages do) have different conjugations for different genders, but if you want to make it less european (and easier to remember) you could go with a completely regular case system.


, for example topicalization, interrogativity, conjunctivity, animacy.

_________________
My name is Christian Munk, and I may be a chipmunk afterall.

Native: | Fluent: | Less than fluent: , , | Beginner: , Volapük | To do list: , Lojban
Creating: Mhmmz

Mama dała dziecku zabawkę? | Did the mother give the child a toy?
Nie, mama dała dziecku coś innego. | No, the mother gave the child something else.

Mama dała zabawkę dziecku? | Did the mother give a toy to the child?
Nie, mama dała zabawkę komuś innemu. | No, the mother gave a toy to somebody else.

Dziecku dała zabawkę mama? | That's the mother who gave the child a toy?
Nie, dziecku dał zabawkę ktoś inny. | No, that's somebody else who gave the child a toy.


That is what you can do in Polish using syntax for focus-marking and Nominative, Accusative, Dative...

_________________
(native)
(learning)

--
Języki sztuczne i lingwistyka, po polsku.

Some languages do have very free syntax even without case marking (and with very weak other strategies of disambiguating) - a language can rely fairly well on things like the meaning of a verb (a clause along the line of 'a taxman stole the bike' is quite unlikely to, in any serious setting, have the bike stealing the taxman - and many languages simply just use this knowledge.)

Cases don't really offer much different from what adpositions offer, except maybe the option of introducing congruence (but e.g. many Russian verbs with prefixes have something along the lines of adpositional congruence with the verb). Congruence of course increases the amount of redundance, which is useful. (Or else languages probably wouldn't have systematic ways of encoding redundant bits of information every here and there). Most Biblical Hebrew prepositions don't differ much from case, in that they affix to the noun.

Cases likewise lack the option of using a case marker intransitively (unless you use some kind of dummy pronoun or whatever to carry intransitive case). Of course, not all grammarians recognize the term "intransitive adposition", but I think the term makes sense. Read http://www.abc.net.au/radionational/pro ... on/3521076 - although one can disagree on whether what he found properly is an intransitive adposition (or something more like an omitted adposition), the idea of calling adpositions without explicit objects intransitive makes a great deal of sense. Such do lend the language a new dimension of reuse.

Some languages permit coordination of adpositions:
around and beyond X
with or without Y

This is not permitted, normally, with cases, although I bet some language that normally is described as having cases has come up with a clever way around that (maybe something like house.at and.in?, house.at dummypronoun.in, whatever)

Many languages permit coordination where different surface cases are demanded, e.g. some languages that have verbs that require objects in other cases than accusative can permit two verbs that require different cases from each other to take one common object according to some resolution strategy. (same goes for adpositions that govern some case.)

In the case of adpositions and verbs governing some case, this does increase the redundance a bit - if there's ten verbs that demand an object in the allative, and you hear an object in the allative - but you only hear half the verb, it's likely your mind can puzzle together what was said without you ever knowing your mind in fact did exactly that. OTOH, adpositions and other things may also contribute to such redundance. (same redundance thing goes with prepositions - if all prepositions denoting direction use accusative and all denoting location use locative, and you miss the preposition but hear the case, you got pretty much of the message already!)

Of course, adpositions or verbs governing cases is a kind of congruence as well, although less obviously than e.g. nouns and their adjectives taking the same case.

So what can we do with case or adpositions? We can encode information about participants in a verb phrase - does the subject act of its own volition, is the object wholly affected, is the object/subject definite, what kinds of participants are there (here, English kind of has syntactic case - subject, indirect object and direct object; other languages slice this up differently, some entirely even merging direct and indirect objects, some having the transitive subject distinct from a merged intransitive subject and transitive object.)

Differential object marking is quite a common phenomenon (wikipedia says 300+ languages)- e.g. Finnish, Turkish, Spanish, Biblical Hebrew and Malayalam have it. Here, case marking tells us something salient - in Finnish, it gives away the telicity of the verb, in Turkish and Hebrew the definiteness of the object, in Spanish it gives away that the object is specific and human. Of course, in the case of Spanish, this is accomplished with a preposition. This also is the case in Hebrew (and one that actually doesn't merge as often as, say, be- or le-).

I think you mean declensions. Grammatical gender and biological gender needn't have anything to do with each other either. The most common two-way distinction is animate vs. inanimate. Also, in all gender systems there is some degree of arbitrariness, since not all words fit neatly into the existing gender categories of the language. If you only want masculine vs. feminine for male vs. female, I wouln't add any gender distinction at all, as it is not strictly necessary. Different lexical items for some common kinship terms would suffice.

As for case, which is actually the topic of this thread, I agree with the other posts in that it frees up the word order so that the syntax can handle other things. But you probably already know that, since you study Latin. If you want to "fake" case you can use postpositions that attach to the noun, which would be pretty much the same thing.
One advantage of case which I don't thing has been mentioned, is that it can make the language more compact. The Ablativus Absolutus construction in Latin, which consists of a noun + a predicate, both in the ablative, would need to be translated into a subclause in English: Vere ineunte 'As spring begun'. But this also has to do with the Latin verbal
system ...

_________________
| | | | |

Wow, hugely informative answers. Thanks!
I'll definitely have to take a closer look at the possibility of adpositions vs. cases.

Statistically, pervasive head-marking seems more closely associated with free word-order, than pervasive dependent-marking is. (I don't know why, though.)

That means e.g. having the verb agree with every participant, and the possessum agree with the possessor, etc., would free up word-order more than case-marking every noun and having all adjectives agree with their head-nouns etc.

(I don't know why, though.)

(But obviously pervasive double-marking would allow free word-order.)

_________________
I am not responsible for the accuracy of my sources; they're responsible for their own mistakes, if any, and also responsible for defending their own statements if you disagree with them.

Yeah, case marking seems still to require the phrases to be left somewhat intact - e.g. you can't much move a possessor far from the possessum, etc. Some level of flexibility is of course possible (heck, full flexibility basically is), but the parsing probably is less straightforward, and such freedom tends to be somewhat limited afaict, e.g. the possessor might be movable only to certain spots, etc.

Maybe that case-marking tend to develop more often in certain types of languages (head-final) rather than other (head-first or non-configurational languages).

OV rather than VO -> therefore case-marking,

Rather than:

Case-marking -> therefore OV rather than VO.

_________________
constructedlanguages.net

I have a feeling this in part also is because linguists by tradition have preferred to call suffixes case, but consider prefixes closer to prepositions.

Well, see this thread and search for "free".

Also, verb-final languages tend to be dependent-marking and verb-initial languages tend to be head-marking; or, at least, head-marking within the clause.

Also, although languages in general tend to prefer suffixes to prefixes about three-to-one, verb-final languages prefer suffixes to prefixes about five-to-one, verb-medial languages prefer suffixes to prefixes about two-to-one, and verb-initial languages tend to be about even-handed between suffixes and prefixes.

It's been suggested that VO languages tend to have head-initial noun-phrases and be postpositional and have suffixing nominal morphology, but to have prefixing verbal morphology and put adverbs before verbs; while OV languages tend to have head-final noun-phrases and have prefixing nominal morphology and prepositions, but to have suffixing verbal morphology and to put adverbs after the verb.

I don't know which of the above ideas contradict other ideas above; and if two are contradictory, I don't know which one is right (if either is).

I also don't know that the above contradict anything the last two posters have said. Actually I don't think it does contradict them. But I could be wrong.

See the following:
http://wals.info/feature/combined/23A/26A
http://wals.info/feature/combined/23A/81A
http://wals.info/feature/combined/23A/85A
http://wals.info/feature/combined/26A/81A
http://wals.info/feature/combined/26A/85A
http://wals.info/feature/combined/81A/85A

http://wals.info/feature/combined/23A/95A
http://wals.info/feature/combined/26A/95A

http://wals.info/feature/combined/51A/69A
http://wals.info/feature/combined/51A/95A
http://wals.info/feature/combined/69A/95A

_________________
I am not responsible for the accuracy of my sources; they're responsible for their own mistakes, if any, and also responsible for defending their own statements if you disagree with them.

Freer word order. No need for as many prepositions. But ultimately I think your question is misphrased: it makes it sound like not having cases is the default and languages which do have them are kind of deviant. A speaker of a language like Latin could just as easily ask: "What's the point in having fixed word order and the preposition "of" when you can just as well have cases?" Case and syntactic marking of grammatical relations are basically two ways of solving the same problem - neither is obviously more basic than the other, I don't think.

_________________
Twitter: @jsbaker750

There's no reason to think either is more basic than the other; but according to WALS.info a lack of cases is more common than a case-system.

_________________
I am not responsible for the accuracy of my sources; they're responsible for their own mistakes, if any, and also responsible for defending their own statements if you disagree with them.