Changes in word order

[The Hanged Man]

How can a natural language change its word order (SVO, SOV, VSO, etc.) during its evolution?

[Yng]

Hmm. Good question. I know of a number of languages that've done this - English being (sort of) one of them, Egyptian another. And presumably celtic at some point.

Probably the easiest way is that a language with multiple possible word orders (due to heavy constituent marking or whatever) settles into a default word order due to loss of case endings or simple frequent stressing of a certain element.

In English's case, it apparently changed from V2 word order (similar to German) to simple SVO (with a few remnants of V2). This was probably simplification based on the sort of 'base' word order (it was underlyingly SVO) and possibly conforming to French norms, although quite how a language's word order could transfer to another without heavy bilingualism I'm not quite sure.

[Airship Hobbit]

although quite how a language's word order could transfer to another without heavy bilingualism I'm not quite sure.

Considering the Norman conquest of England, there was some heavy bilingualism. The upper class was practically defined by being descended from the French. I read somewhere that Richard the I couldn't even speak English.

[Niedokonany]

One could imagine for example a cleft construction being utilized more and more to the point where it replaces the 'normal' sentences. Or some other marked topicalized or focused pattern spreading in the same way. Or sentences with a dangling, additional constituent like "he's already told me, the director" change into a standard pattern - isn't something like that happening in French right now?

Still another possibility is, say: nom-acc, passive voice, SVO > ergative alignment and OVS.

[Yng]

Considering the Norman conquest of England, there was some heavy bilingualism. The upper class was practically defined by being descended from the French. I read somewhere that Richard the I couldn't even speak English.

Yeah, but there wasn't really heavy bilingualism - the upper classes mostly spoke French, the lower classes English. There wasn't really a well-developed middle class at that point - but the upper classes weren't that wide, were they? Still, I suppose the influence of their incorrect English could've spread.

[Radius Solis]

I think most often, a "word order change" is not so much a restructuring of the unmarked clause formula, as it is a change in which existing clause formula is least marked. Most languages permit three or four or five of the basic word orders - these may or may not involve extra grammatical morphemes, but will be valid orders in which the S, V, and O elements can appear.

There are two main motivators for an alternative clause construction to be reanalyzed as basic. One is language contact: for example, Breton changed its basic word order when, under the influence of French, its speakers began using the existing SVO cleft construction to the exclusion of what had previously been the default VSO order.

The other is that other grammatical changes can cause problems for a word order, leading speakers to select a different one. In particular, SOV languages that lose their case systems are prone to switching to a verb-medial structure, as the verb position helps keep subject and object distinct. This happened in both the Romance languages and the Germanic languages. It is not hard to imagine other such scenarios.

[The Hanged Man]

After some time, I'm working on word order change again. And these are changes that I invented recently, but first I'll explain some terms:

O.GEN - direct object of verb, being in genitive case
V.Noun - verbal noun, action of doing something
Part. - a particle, of course
Remaining terms should be clear .

First, change SVO>SOV, where O.GEN V.Noun is action nominal. It should appear in a language before, because as far as I know, action nominals involving genitive should be rather in V.Noun O.Gen order, because of standard genitive and possesed noun order in an SVO language.

SVO>S Aux. O V.Noun> S-Aux O V.Noun> S O V

Second change is one of kind SVO>VSO. It involves right dislocation and grammaticalization of subject and object pronouns, like in colloquial French

SVO>Pron. V Pron., S, O>Pron.-V-Pron., S, O>V S O

Another change, from VSO to SVO, by topicalization of subject and grammaticalization of subject pronoun:

VSO> S(topic), V Pron. O> S V-Pron O> SVO

Second change VSO>SVO, now involving evolution of an auxilary

VSO>Aux S V.Noun O>Part. S V O>SVO

The last change I have invented so far, is one from SOV to SVO. Like in the first change, it needs a change of order of genitive and possesed noun:

SOV>S Aux V.Noun O.GEN.>S-Aux V O>SVO

What do you think about my changes? Do they look plausible?

[TaylorS]

Alpic started out as A-R-P-V (A = Agent, R = Recipient/Direct Object, P = Patient, V = Verb) and purely head-final, Adjectives, Genitives, and relative clauses came before the noun they modified. Then under Hesperic and then Celtic influence it became V-A-P-R, head-initial, and Adjectives and Relative Clauses switched to come after the noun they modified. Then when the Patientive case ending eroded word order switched to A-V-P-R.

[The Hanged Man]

@TaylorS, switching word order by influence of other language is much easier... But it's not what I'm looking for. I want to create whole language family, descendant from single proto-language, and I want some of it's members to change their word order because of internal causes...

[Yng]

@TaylorS, switching word order by influence of other language is much easier... But it's not what I'm looking for. I want to create whole language family, descendant from single proto-language, and I want some of it's members to change their word order because of internal causes...

It's much easier if it comes from an originally freer word order, which is the case in most IE languages, for example. A common change, from what little I remember of reading about this, is SOV to SVO as a language stops marking case, particularly where ambiguity can appear.

Of course, this isn't universal, but in a lot of languages no ambiguity appears... If you look at Welsh or Irish, in which argument marking with non-periphrastic constructions is basically dead, you still don't have much ambiguity because of the way genitive constructions work (only one definite article, etc):

Rhedai'r (that is, y) dyn y fferm - The owner used to run the farm
Rhedai meistr y fferm - The master of the farm used to run

The only ambiguity would be in extremely contrived sentences, like 'an owner used to run the farm', which occur sufficiently rarely to maintain understanding.

[TaylorS]

@TaylorS, switching word order by influence of other language is much easier... But it's not what I'm looking for. I want to create whole language family, descendant from single proto-language, and I want some of it's members to change their word order because of internal causes...

One common way of changing the word order is to do what French is doing:

NOM.PRONOUN ACC. PRONOUN VERB *pause* SUBJECT OBJECT

"Je t'aime, moi toi."

Now, say the pronouns start becoming verbal inflections, which is occurring in French right now. Then say the pause marking the phrase boundary disappears, this will shift the language to VSO.

NOM-ACC-VERB SUBJECT OBJECT

/ʃ-t-em mwa twa/

[The Hanged Man]

One common way of changing the word order is to do what French is doing:

Yes, this is one of my ideas, too . You say it's common: where does it happen, except French?

[Shihali]

IRL, Arabic has shifted from VSO to SVO. I believe the pattern was VSO with SVO as a marked variant and case-marking, then the case-marking final vowels fell along with bits of the verbal inflection system, and then SVO became the usual word order.

Akkadian and Amharic shifted from hypothesized-VSO to SOV, but that was under substrate influence.