Word Order and cases, help?

[theweevil]

So, I'm curious as to what interesting things I can use word order for if its been freed up by cases and such. Anybody have some good resources concerning this?

[Gulliver]

You could use it for tense.

She like me = Present
Like she me = Past
She me like = Future

Or negation, or questions, or mood or emphasis any number of verby business.

She like me = neutral
Like she me = she really likes me
She me like = she likes me

Alternatively, you could use it to separate adjective+nouns from compound nouns (we do this with word stress in English mostly) like "bird black" for a black bird but "black bird" for a blackbird. Or have a taboo that does not permit women to start a sentence, or make verb-first sentences sound overly formal while verb-final sentences sound childish and cutesy.

[Yng]

You could use it for tense.

She like me = Present
Like she me = Past
She me like = Future

and how pray would one distinguish present and future in intransitive sentences

Or have a taboo that does not permit women to start a sentence, or make verb-first sentences sound overly formal while verb-final sentences sound childish and cutesy.

More interesting but general common-sensy linguistic stuff.

Generally speaking, word order - whether in a language with cases or not - is used to control information flow. English permits scrambling to a far greater degree than most people assume (probably because scrambling tends to happen in context and when people think up sentences out of context they sound more awkward); Dutch and Persian far more. Emphasis and topicality are pretty important as far as this is concerned. Unfortunately I don't know where to point you for free resources on this anymore, but if you google 'linguistics information flow' or something you'll probably get a good deal of results. Alternatively, if you have the resources, there are plenty of expensive (maybe available in good libraries though) books that you can read.

[Miekko]

You could use word order to distinguish objects from subjects, and use case to distinguish different types of objects (say, definite and indefinite objects; telic and atelic objects, ...) ...

[clawgrip]

You could use it for tense.

She like me = Present
Like she me = Past
She me like = Future

and how pray would one distinguish present and future in intransitive sentences

You could have a dummy pronoun that acts as a tense marker only in intransitive verbs.

She sleep = present
sleep she = past
she she sleep = future

[Rui]

You could use it for tense.

She like me = Present
Like she me = Past
She me like = Future

and how pray would one distinguish present and future in intransitive sentences

Delicious ambiguity that can be figured out from context. Note that there's a whole class of English verbs that aren't distinguished in the present and past (except in 3.sing, obviously). Yes, I'm looking at you, monosyllabic words that end in /t/: put, hit, beat, etc etc etc.

[Gulliver]

You could use it for tense.

She like me = Present
Like she me = Past
She me like = Future

and how pray would one distinguish present and future in intransitive sentences

Adverbs would do the trick.

Today, we frolic!
Tomorrow, we frolic!

As Chibi said, ambiguity is not exactly uncommon in natural languages.

Yng's right about shoving information further up the sentence to give it more importance. German does it, and I think Latin might (have) as well.

[Chuma]

Yes, moving information seems like the most obvious choice. As I understand it, there are natlangs that do topic fronting...
food ate dog = as for the food, the dog ate it
dog ate food = as for the dog, it ate the food
...and there are those that do emphasis fronting...
food ate dog = the food is what the dog ate
dog ate food = the dog is what ate the food
...which is sort of weird, since those are basically opposites.

You might be also able to somehow use it for nominalisation.
ate dog food = the dog ate the food
dog ate food = the dog which ate the food
food dog ate = the food which the dog ate
Actually the last one kind of works in English too.

Or if you want to be really crazy, you could just assign completely different meanings to them.
ate dog food = the dog ate the food
dog ate food = the dog liked the food
food dog ate = the dog wanted the food

[linguoboy]

Yng's right about shoving information further up the sentence to give it more importance. German does it, and I think Latin might (have) as well.

Not really. It's more a matter of topicalisation, which means that the greater the news value of an item, the later it tends to come in the sentence. Time expressions, for instance, tend to come either (a) in initial position or (b) immediately after the finite verb, and they are usually not the most important element in the sentence. This is also the preferred location for subjects, which tend to double as topics. Object pronouns (representing items already mentioned) come immediately after the finite verb, before even sentence adverbs, whereas nominal complements generally appear after any adjuncts right before the final (nonfinite) verb, if present.

For instance, "Ihn hab ich gestern gesehen" (lit. "Him have I yesterday seen") places more emphasis on "yesterday" than it does on "him". In fact, preverb position is so low on news value that it often gets filled with a dummy subject, e.g. "Es ist mir schon warm." ("It is to-me already warm", i.e. "I'm quite warm".) "Mir ist schon warm" is about the same--it doesn't put any particular emphasis on "me"--but fronting warm indicates topicality, e.g. "Dir ist warm? Warm ist mir auch." ("To-you is warm? Warm is to-me too.") Similarly existential verbs (e.g. "Es sind zwei Bäume im Weg") where the logical subject is new information and is therefore kept out of the topic position.

[Melteor]

Slavic languages have loose word order, and Polish word order is very free, so that could be a place to start. Informally, I think English word order can be somewhat loose, but obviously would not be attested in texts so often as it would be spoken.

[linguoboy]

Slavic languages have loose word order, and Polish word order is very free, so that could be a place to start. Informally, I think English word order can be somewhat loose, but obviously would not be attested in texts so often as it would be spoken.

And, apart from personal pronouns, English doesn't even have any cases. Chinese also has a fair bit of freedom (rather more than English, actually) within an overall topic-comment framework.

On the other hand, German word order is extremely strict despite mandatory case marking on determiners and modifiers. So although there is clearly some relationship between case marking and freer word order, I don't think it's as close as some people may be assuming.

[clawgrip]

Latin, Ancient Greek and Sanskrit all have pretty free word order, don't they?

[theweevil]

The tenses idea I like. I would also agree that Englush does have for freedom of word order than most people accredit to it. I think the thing that I have the hardest time getting my head wrapped around is the whole topicalization and focus shift deal; I mean, I understand it you could say, but it just is still weird to me for reasons I don't completely understand.

[clawgrip]

Can you elaborate at all on what it is you find weird about topicalization?

[theweevil]

Can you elaborate at all on what it is you find weird about topicalization?

I guess what I meant by that was: I don't like using word order to topicalize or change a sentence's focus. In my experience, when I see topicalization examples the most common method employed is word order. For example:

I like that cat vs that cat I like.

I understand that cat is being brought into focus but I would rather use a different method were I to topicalize something in a conlang. I would prefer to do something like this for example:

I gave Suzy a gift vs I gave Suzy a necklace vs I gave Suzy gift, a necklace
|---less specific----|--------neutral---------|--focus on gift, specified as a necklace----|

Hopefully that seems somewhat intelligible.

[Buran]

If you're going to use word order for tense, maybe throw mood and/or aspect into it. Also, consider forming the past and the future tenses with entirely different paradigms (word order for past vs. say, auxiliary verbs for future).

[Xephyr]

Theweevil, I'd like to ask a question about the motivation behind your design choices. If you are so unsure and lost as to how to handle word order in a case-marking language-- so much as to make a thread about it-- why are you using case-marking and free word order in the first place? In general one ought to first decide what one wants a language to do, and then decide how it does it, not the other way around.

EDIT To elaborate a bit, I (and most people probably) believe that conlangs should be based on real world examples-- not copied from other languages, obviously, but based on features with a grounding in reality. A person who knows a lot about synthetic languages would probably be better off making a synthetic conlang rather than an isolating one. If one wants to make a conlang with X feature, but one doesn't know very much about X feature, then they would rightly first try to learn about X feature. Which is what you're doing, but....

The problem with wording a question like "I have a conlang with free word order-- what can I do with it?" is that since it's a question about conlangs (which are make-believe), it invites people to suggest all sorts of outlandish crap, such as marking tense, which is something I've never heard of any natlang doing. (I have, however, on at least two other occasions seen people suggest doing it in a conlang...) If you were to pose the question in a different way, like "What are some things languages use word order for?", then you wouldn't get chaotic answers such as these. You WOULD, on the other hand, probably end up learning a great deal more.

Consider this: if a person came along and asked how polysynthesis worked, then people would (hopefully) respond by telling them a bit about how polysynthesis works. If, however, they came along and said "I have a polysynthetic conlang. Now how does polysynthesis work?" then a likely response would be "Why do you have a polysynthetic conlang if you don't know how polysynthesis works?"

[theweevil]

Theweevil, I'd like to ask a question about the motivation behind your design choices. If you are so unsure and lost as to how to handle word order in a case-marking language....why are you using case-marking and free word order in the first place? In general one ought to first decide what one wants a language to do, and then decide how it does it, not the other way around.

I made this thread with the intention of learning. I want to make a language that uses case marking, so that it may have a more flexible word order, but I do not know how I can go about it or what I can do with it, hence my question.

EDIT To elaborate a bit, I (and most people probably) believe that conlangs should be based on real world examples-- not copied from other languages, obviously, but based on features with a grounding in reality. A person who knows a lot about synthetic languages would probably be better off making a synthetic conlang rather than an isolating one. If one wants to make a conlang with X feature, but one doesn't know very much about X feature, then they would rightly first try to learn about X feature. Which is what you're doing, but....

I would agree with you that the best example for something is to look at natlangs that feature X. But, sometimes when trying to find information on X, I may have trouble finding it for what ever reason, so I come here, as it has proven to be a large wealth of knowledge and creativity.

The problem with wording a question like "I have a conlang with free word order-- what can I do with it?" is that since it's a question about conlangs (which are make-believe), it invites people to suggest all sorts of outlandish crap, such as marking tense, which is something I've never heard of any natlang doing. (I have, however, on at least two other occasions seen people suggest doing it in a conlang...)

I, for one, enjoy the outlandish ideas. I think that a conlang, being make-believe (a generally individual venture/work of art), can break rules where and when it wants. isn't it really dependent on the goals of its creator? Naturalistic conlangs are awesome, but so are unaturalistic-weird-wut-why did you do that languages. I mean, a conlang only has 0.5 speakers at best anyway right? (excluding, course, the conlangs that, you know, have more than that many)

"What are some things languages use word order for?"

This is also partially what I wanted my question to come out as.

[Vuvuzela]

This thing about Nahuatl non-configurationality might be useful . You can skim through the bits where it talks about polysynthesis, and just focus on the word order parts, which are enlightening.