Subject Object Verb orderings for visual languages

[Sunny]

A colleague pointed out a recent paper in the PNAS journal:
"The natural order of events: How speakers of different languages
represent events nonverbally.", http://www.ncbi.nlm.nih.gov/pubmed/18599445

I think that close analysis of OSV and SOV orderings are interesting
and important, I don't agree with the conclusion in their paper.
They are only drawing from four example languages which doesn't
statistically support their conclusion. Further, they are testing
with a very limited set of concepts. While it is true that several
natural sign languages often use SOV ordering (ASL for example), they
also commonly use other orderings depending upon the subject matter
and context. Also, not all signed languages use a primarily SOV
ordering. For their conclusions, I would argue that there is a
'natural order' that humans use when representing specific subject
matter, that this natural order is not context free, but critically
depends on the type of information being conveyed. Further, I would
suggest that the reason that the ordering for each language was
different was due to the language formalization, the ordering being
imposed culturally whereas the free-form gestures are able to better
match the subject matter being communicated.

[Ceruti, Marion SPAWAR]

Interestingly enough some languages are SOV where pronouns are concerned
but SVO for other nouns, at least in some cases. Per esempio, in
italiano

Lei mi ha aiutato. = "She me has helped." meaning "She has helped me."
At least this is how you would say it in Lombardia.

I would like to know if other cultural sign languages follow the OSV/SOV
order of the language and cultures they are supposed to represent or if
the SL is totatly independent from the spoken language, i.e. a separate
subculture that does not share the same rules about word order. Ideas?

[Sunny James Fugate]

As a single example, American Sign Language doesn't match spoken
English in respect to part-of-speech ordering. ASL is generally OSV.
e.g. "I park go" I mis-spoke in my original post about ASL being SOV..

If you are talking about SEE (Signed Exact English), then, of course
it will be the exact literal English translation and follow OSV. I
suspect that depending upon how independent the signed and spoken
languages are they will either be coupled or not. Either way, the
study referenced below was looking at gesture and not formal signs.
It would be interesting to follow up with a study to determine if part
of speech order is dependent on subject matter and whether the gesture
POS order is the same or different than formal sign order.

-Sunny

[neil...@emaki.net]

We actually just read that paper (and the longer paper first done only
with English) for a discussion group based on my urgings. I just want to
clarify a few points:

The paper was not arguing for Subject/Object/Verb orderings, but of
ordering of semantic information, Actor-Patient-Act, that just happens
to correlate prototypically with grammatical functions. Syntax and
semantics are two different things that relate to each other. Their
conclusions were essentially supporting this: no matter which
grammatical order is default in a language, people make similar
judgements about ordering for non-verbal *semantic* information.
Ultimately, they are talking about aspects of the syntax-semantics
interface (again, on this point I recommend Jackendoff's "Foundations of
Language").

Their choice of sample languages were made largely because Goldin-Meadow
(whose lab I used to frequent) has colleagues in those other countries
who also do research on Homesign — gesture systems spontaneously
created by deaf children born to hearing parents who don't otherwise
learn a sign language. The evidence from homesign systems is that
regardless of culture, they have a preference for Actor-Patient-Act.
This is not SOV because there is no syntax in homesign.

Sign languages have no correlation to the verbal languages of the
culture they belong to. They are wholly independent systems that have no
bearing on each other. If I recall, I don't think that they were arguing
that sign languages are predominantly SOV anyways — their argument is
that for these transitive concepts, there are semantic preferences for
orderings that are independent of language orderings.

I recommend also reading the all English study that this was based on,
that has a lot more detail:

Gershoff-Stowe, L., & Goldin-Meadow, S. (2002). Is there a natural order
for expressing semantic relations? Cognitive Psychology, 45, 375-412.

Best,

Neil