Page 1 of 1

VO -> OV, prepositions -> postpositions?

Posted: Wed May 19, 2010 9:49 am
by Tengado
Say a language had relatively free word order - unmarked SVO, but also OSV, SOV (object can be fronted to topicalise/emphasise it, or put verb final for another special reason) and prepositions. Later the word order shifts to fixed verb final order. What would happen to the prepositions? Verb final with prepostions is supposedly against a strong universal. How would the situation resolve? Would/could the prepostion reanalyse as a postposition to the word before? Eg S na-O V -> S-na O V ? Any other ideas?

Is this change actually attested? If so, what happened?

Posted: Wed May 19, 2010 10:00 am
by TomHChappell
Read one or more of the books that come up in this search.

My particular preference is Word order and word order change, Volume 1974 By Charles N. Li, if only because I've already read it.

But to answer your questions, IMHG (in my humble guess) probably one or both of Word order correlations and word order change: an "applied-typological ... By Jasmine Tragut (because it includes "word-order correlates") or English Historical Linguistics 2006: Syntax and morphology By Maurizio Gotti, Marina Dossena, Richard Dury (because that's a well-studied actual historic example of such a change in a well-known natlang) would be better.

---
Hope that helps.

Posted: Wed May 19, 2010 10:45 am
by Radius Solis
In general I would expect the prepositions to stay prepositions. I think you've got the universal a bit backwards: strong postpositional systems are very restricted to OV languages, but strong prepositional systems (and mixed systems) are found in every kind of language. But it's certainly possible for a preposition system to gradually decay after a language changes to OV, as a result of new postposition creation. And it's probably possible for prepositions to become postpositions of the same nouns by moving from the beginning to the end of the noun phrase.

But in general, it is relatively rare for grammatical marker to switch from attaching to the word on one side to attaching to the word on the other, without itself changing position. And even then, when it does happen, it's when the marker in question is located between two types of word that are normally always found together in the same order, which can make it unimportant to know whether the marker attaches to the left or the right.

I can think of one possible exception that would work that way: a preposition marking accusative case on the direct object, which in SOV could conceivably turn into an ergative-marking postposition on the subject. But the whole prepositional system doing this in unison would be bizarre.

Posted: Wed May 19, 2010 11:31 am
by brandrinn
There are plenty of examples, after all. Most early attested IE languages had SOV dominant order, and almost a unanimous preference for prepositions. They all had suffixed noun case, of course, but the point is still valid.

Re: VO -> OV, prepositions -> postpositions?

Posted: Wed May 19, 2010 2:02 pm
by Basilius
Tengado wrote:Say a language had relatively free word order - unmarked SVO, but also OSV, SOV (object can be fronted to topicalise/emphasise it, or put verb final for another special reason) and prepositions. Later the word order shifts to fixed verb final order.

OK. Essentially, you want to alter the markedness, then get rid of the marked options. No problem, except that you don't seem to have any idea why such a change may happen. "Just because".
Tengado wrote:What would happen to the prepositions? Verb final with prepostions is supposedly against a strong universal.
Forget universals. Either you understand why they are there, or you can't rely on them.
Tengado wrote:How would the situation resolve? Would/could the prepostion reanalyse as a postposition to the word before? Eg S na-O V -> S-na O V ?
No. Your na marks the role of O. It won't mark the role of S. S and O even don't form a constituent, clauses don't all fit in a template of just three slots, your protolanguage probably had other prepositions besides na, etc. I don't see an easy way to re-interpret prepositions as anything else.
Tengado wrote:Any other ideas?
As Radius said, just leave prepositions where they are. Unless you really want a headache.

Posted: Wed May 19, 2010 2:29 pm
by Xonen
FWIW, Baltic Finnic has kind of had the opposite change in word order: from predominantly SOV to predominantly SVO (word order is still relatively free, though). And yes, this has apparently affected the placement of adpositions to some degree; the system still favors postpositions, but is no longer exclusively postpositional. Several adpositions can even be placed either before or after the noun they modify, and it may well be that some which formerly tended to go after now tend to come before (ie. have effectively changed from postpositions into prepositions). Other prepositions are new ones that simply weren't there before the word order shift; they seem to have been derived from adverbial expressions and whatnot (and a couple have been borrowed from neighboring IE langs).

So I don't see why that couldn't work the other way around as well. The system would probably stay mainly prepositional, at least for some time, but gradually come to allow more and more postpositions as well.

Posted: Wed May 19, 2010 3:22 pm
by WeepingElf
I have just finished reading Ian Roberts, Diachronic Syntax, and he gives an example of a VSO >SOV change (citing Greenberg), namely in Ethiopian Semitic. Ge'ez, the liturgical language of the Ethiopian Christians, is a representative of the state before the change: it is as VSO as its sisters Hebrew and Arabic. Then, the Ethiopian Semitic languages shifted to verb-final order under the influence of neigbouring non-Semitic SOV languages; later, other word-order parameters changed, one after another, to harmonize with the SOV order. Several languages represent the stages of the change:

Ge'ez: VSO, NA, NGen, Prep
Tigre: SOV, NA, NGen, Prep
Tigrinya: SOV, AN, NGen, Prep
Amharic (14th cent.): SOV, AN, GenN, Prep
Harari: SOV, AN, GenN, Postp

(Tigre and Tigrinya are spoken more northward of Amharic, Harari to the south; the change propagated from south to north.)

Posted: Wed May 19, 2010 3:37 pm
by TaylorS
brandrinn wrote:There are plenty of examples, after all. Most early attested IE languages had SOV dominant order, and almost a unanimous preference for prepositions. They all had suffixed noun case, of course, but the point is still valid.
My understanding is that PIE did not have adpositions, but had adverbs that could fill a adpositional role, a system partially preserved in Germanic separable verbs (including English phrasal verbs), which would be why they would have become prepositions, if they would have come after the Object they would have been confused with a adverb modifying the verb.

Posted: Wed May 19, 2010 3:56 pm
by Travis B.
TaylorS wrote:
brandrinn wrote:There are plenty of examples, after all. Most early attested IE languages had SOV dominant order, and almost a unanimous preference for prepositions. They all had suffixed noun case, of course, but the point is still valid.
My understanding is that PIE did not have adpositions, but had adverbs that could fill a adpositional role, a system partially preserved in Germanic separable verbs (including English phrasal verbs), which would be why they would have become prepositions, if they would have come after the Object they would have been confused with a adverb modifying the verb.
Also, it seems that at least in earlier times amongst the Germanic languages adposition-noun order was more flexible than it is today; while such seems to have by the time most of the old Germanic languages were attested settled into a mostly adposition-noun order, the noun-adposition order is still attested and is still preserved in frozen forms in continental West Germanic languages and formal/archaic English.

Posted: Thu May 20, 2010 1:59 am
by Radius Solis
Elf: beautiful example, thank you!

Travis B. wrote:the noun-adposition order is still attested and is still preserved in frozen forms in ... formal/archaic English.
Could you provide an example, please? Preferably one that is neither 1. an actual postposition (we do have a few floating around) nor 2. a noun fronted within its PP for pragmatic effect.

Posted: Thu May 20, 2010 2:06 am
by Travis B.
Radius Solis wrote:
Travis B. wrote:the noun-adposition order is still attested and is still preserved in frozen forms in ... formal/archaic English.
Could you provide an example, please? Preferably one that is neither 1. an actual postposition (we do have a few floating around) nor 2. a noun fronted within its PP for pragmatic effect.
Forms with general form of there/where/here/hence/whence/hither/whither followed by an adposition, such as thereof, wherein, hereafter, and so forth. These kinds of forms are definitely much more productive and far more heavily used in continental West Germanic than in English today, such in the case of the the Standard German davon, worin, and so forth.

Posted: Thu May 20, 2010 8:48 am
by hwhatting
Travis B. wrote: Forms with general form of there/where/here/hence/whence/hither/whither followed by an adposition, such as thereof, wherein, hereafter, and so forth. These kinds of forms are definitely much more productive and far more heavily used in continental West Germanic than in English today, such in the case of the the Standard German davon, worin, and so forth.
The German situation is complicated by the fact that German has varying preferred constituent orders depending on sentence types - Verb-second in statement main clauses, verb final in dependent clauses, verb-first in yes-no questions.

Posted: Thu May 20, 2010 10:10 am
by brandrinn
Just for the record, I did not say PIE; as an unattested languages, it would be a fairly pointless example of a phenomenon in natlangs. We could conclude through internal reconstruction that *en was not a true preposition in PIE, perhaps. But there's no doubt that it's a preposition in Latin, Greek, and other languages that had few-to-no postpositions and primary SOV word order.

Posted: Fri May 21, 2010 11:56 pm
by Neek
brandrinn wrote:Just for the record, I did not say PIE; as an unattested languages, it would be a fairly pointless example of a phenomenon in natlangs. We could conclude through internal reconstruction that *en was not a true preposition in PIE, perhaps. But there's no doubt that it's a preposition in Latin, Greek, and other languages that had few-to-no postpositions and primary SOV word order.
Despite that the majority of Indo-European languages developed prepositions, there were no such things in the reconstructed PIE; even in Homeric Greek, what we would consider prepositions in a matter of centuries still carried more adverbial force.

Despite a lot of ancient IE languages being predominantly SOV, prepositions were still favored. Perhaps because adverbs were placed before the verb? And by analogy, when these adverbs were used to modify nouns, they still went before? It's difficult to explain why PIE did such a thing, but we can only look so far back.

Posted: Sat May 22, 2010 3:05 pm
by Niedokonany
hwhatting wrote: The German situation is complicated by the fact that German has varying preferred constituent orders depending on sentence types - Verb-second in statement main clauses, verb final in dependent clauses, verb-first in yes-no questions.
On a side note, wouldn't German syntax be even more interesting with postpositions/adjectives before nouns in dependent clauses and prepositions/adjectives after nouns in main clauses? (Or perhaps AdjN & Prep in affirmative main clauses but NAdj & Prep in interrogative clauses?)

Posted: Tue May 25, 2010 9:33 am
by hwhatting
Piotr wrote: On a side note, wouldn't German syntax be even more interesting with postpositions/adjectives before nouns in dependent clauses and prepositions/adjectives after nouns in main clauses? (Or perhaps AdjN & Prep in affirmative main clauses but NAdj & Prep in interrogative clauses?)
Sure it would :wink: . As far as it gets, I've seen the fact that some adpositions (wegen, entlang) can be used both as pre- and as postpositions explained by the varying constituent order, although more in a sense of enabling the use of postpositions in general, not as a trigger (you can use them pre- and postpositional independent of the constituent order in the sentence where they are used).