Author Topic: Differences between signed languages and spoken languages  (Read 9345 times)

Offline Daniel

  • Administrator
  • Experienced Linguist
  • *****
  • Posts: 1421
  • Country: us
    • English
Re: Differences between signed languages and spoken languages
« Reply #15 on: January 04, 2014, 12:27:36 AM »
Quote
I don't know of any specific reference, but at least a few different South Asian signed languages do this extensively with causal dependency structures.
Like English "sign here and you'll own the car"? Or more grammaticalized than that? I'd be very interested in that data.
Quote
If you get beyond lexical conjunction, there are a lot of idiomatic verb conjunctions expressed spatially.
Serialization? Or something along those lines? I'm starting off with overt coordination (and other uses of "and") and then moving on to the other ones as well-- covert coordination, serialization, and a few other things like switch-reference.

Quote
Well, certainly it's not equivalent to English. I don't see why that should suggest that conjunction is not salient.
I'm specifically talking about the actual word "and", not the structure. It's arguably the case that English "and" is important in some way to be able to put things together, while it isn't in other languages-- they can use juxtaposition for example. Take Turkish as one case, where you can use a coordinator, but it just isn't as important or frequent as in English. Often in Turkish just juxtaposition is much more normal or even required for certain phrases-- take "annebaba" meaning mother-father 'mother and father'.
Welcome to Linguist Forum! If you have any questions, please ask.

Offline Corybobory

  • Global Moderator
  • Linguist
  • *****
  • Posts: 138
  • Country: gb
    • English
    • Coryographies: Handmade Creations by Cory
Re: Differences between signed languages and spoken languages
« Reply #16 on: January 04, 2014, 03:22:01 AM »
signed languages aren't equivalent to the spoken languages in their area, for example ASL and English or BSL and English.  ASL has a different word order I thought as well?

One interesting thing about ASL and BSL (British sign language), ASL developed from French sign language so while ASL and french sign language speakers have a sense of mutual intelligibility, BSL is completely different and unintelligible to other English speaking signers.
BA Linguistics, MSt Palaeoanthropology and Palaeolithic Archaeology, current PhD student (Archaeology, 1st year)

Blog: http://www.palaeolinguist.blogspot.com
My handmade book jewellery: http://www.coryographies.etsy.com

Offline Daniel

  • Administrator
  • Experienced Linguist
  • *****
  • Posts: 1421
  • Country: us
    • English
Re: Differences between signed languages and spoken languages
« Reply #17 on: January 04, 2014, 03:27:01 AM »
Quote
signed languages aren't equivalent to the spoken languages in their area, for example ASL and English or BSL and English.  ASL has a different word order I thought as well?
Right. I didn't mean to imply that they were, but that broadly signed and spoken languages might have some similiarities.
ASL is SVO or SOV-- I'm getting mixed results from Google. MalFet will know. I think it's flexible and perhaps varies by which structure is being used-- some results hint at that.

Quote
One interesting thing about ASL and BSL (British sign language), ASL developed from French sign language so while ASL and french sign language speakers have a sense of mutual intelligibility, BSL is completely different and unintelligible to other English speaking signers.
I saw a fun family tree of sign languages that descend from French Sign Language. It was very different than I'd expect having previously only seen trees for spoken families, yet still obviously the same kinds of relationships (dialects, subgroupings, etc.).
Welcome to Linguist Forum! If you have any questions, please ask.

Offline MalFet

  • Global Moderator
  • Serious Linguist
  • *****
  • Posts: 282
  • Country: us
Re: Differences between signed languages and spoken languages
« Reply #18 on: January 04, 2014, 07:21:34 AM »
Quote
If you get beyond lexical conjunction, there are a lot of idiomatic verb conjunctions expressed spatially.
Serialization? Or something along those lines? I'm starting off with overt coordination (and other uses of "and") and then moving on to the other ones as well-- covert coordination, serialization, and a few other things like switch-reference.

Serialization would be one expression, certainly.

Quote
Well, certainly it's not equivalent to English. I don't see why that should suggest that conjunction is not salient.
I'm specifically talking about the actual word "and", not the structure. It's arguably the case that English "and" is important in some way to be able to put things together, while it isn't in other languages-- they can use juxtaposition for example. Take Turkish as one case, where you can use a coordinator, but it just isn't as important or frequent as in English. Often in Turkish just juxtaposition is much more normal or even required for certain phrases-- take "annebaba" meaning mother-father 'mother and father'.

Signed languages don't have the English word "and" because they're not English. They have other forms that are, with various degrees of equivalence, glossed as "AND". If you're expecting isomorphic function between this one particular lexical item I posted and English conjunction, you're going to run into trouble. It's just a different system that works differently.

If you're interested in just that particular lexical item in ASL, that's fine but it's sort of an arbitrary choice. There's a lot more going on in conjunction.

Offline Daniel

  • Administrator
  • Experienced Linguist
  • *****
  • Posts: 1421
  • Country: us
    • English
Re: Differences between signed languages and spoken languages
« Reply #19 on: January 04, 2014, 12:32:24 PM »
Quote
Signed languages don't have the English word "and" because they're not English. They have other forms that are, with various degrees of equivalence, glossed as "AND". If you're expecting isomorphic function between this one particular lexical item I posted and English conjunction, you're going to run into trouble. It's just a different system that works differently.
I'm not claiming otherwise. But some spoken languages have words roughly equivalent to English "and". And I'm interested in their behavior, whether or not they're similar.
The fact that they vary is interesting (but not surprising), and I think it's relevant to understand how they vary.
My research questions some basic ideas like the very idea that "and" exists in the syntactic structure at all in English, specifically because other languages don't require such a word; I think it's an arbitrary strategy to help in the linearization process. But then what also interests me is how this word that appears to strictly mark coordination (that's what most of the literature says) then can be used for other purposes-- and I've found examples of fuzzy use of "and" in something like 12 language families around the world-- we simply have no reason to assume that a "coordinator" is a thing. So, in the end, I think we're saying the same thing, but I'm just starting by trying to gather data on words that are roughly like "and" in as many languages as possible-- including ASL.

Quote
If you're interested in just that particular lexical item in ASL, that's fine but it's sort of an arbitrary choice. There's a lot more going on in conjunction.
Typology relies on comparisons, so that's a starting point. I'd be interested in coordination broadly in ASL (or other signed languages) but I haven't been able to find any sources that discuss it in any detail. If you come across some, let me know.
Welcome to Linguist Forum! If you have any questions, please ask.

Offline MalFet

  • Global Moderator
  • Serious Linguist
  • *****
  • Posts: 282
  • Country: us
Re: Differences between signed languages and spoken languages
« Reply #20 on: January 04, 2014, 01:12:54 PM »
Quote
If you're interested in just that particular lexical item in ASL, that's fine but it's sort of an arbitrary choice. There's a lot more going on in conjunction.
Typology relies on comparisons, so that's a starting point. I'd be interested in coordination broadly in ASL (or other signed languages) but I haven't been able to find any sources that discuss it in any detail. If you come across some, let me know.

There is a lot here, but Sarah Taub's book might be a reasonable starting place.

The first thirty years of sign language linguistics was dedicated to establishing the continuities of sign and speech. This was an important political project at the time, as the common sense through most of the 60s, 70s, and 80s was that signing did not actually constitute "language" in any rigorous sense.

As is now extremely clear, that's a bogus position, but it's really only in the last ten years or so that it has been politically viable to argue that there might be significant structural differences between spoken and signed languages. And, over the last ten years, that's the direction the research has gone in force. It seems increasingly clear that signing is organized on some very different first principles than spoken languages are. As a result, it's very difficult to unify spoken and signed languages within a single set of typological archetypes.

I can't make this point strongly enough. I've spent half of my time over the last decade working on signed languages and the other half working on spoken languages. Most of my work in spoken languages is historical and typological. Nevertheless, I have a very, very hard time producing non-vacuous typological statements about any of the signed languages I work on. They just work on a fundamentally different set of principles.

Offline freknu

  • Forum Regulars
  • Serious Linguist
  • *
  • Posts: 397
  • Country: fi
    • Ostrobothnian (Norse)
Re: Differences between signed languages and spoken languages
« Reply #21 on: January 04, 2014, 01:14:45 PM »
Speaking of Pirahã (had a look at wikipedia), can anyone show me a video of whistled speech? Not a herder and his sheepdog, but actual language. I would really like to hear and "visualise" that somehow.

Offline Daniel

  • Administrator
  • Experienced Linguist
  • *****
  • Posts: 1421
  • Country: us
    • English
Re: Differences between signed languages and spoken languages
« Reply #22 on: January 04, 2014, 01:28:06 PM »
freknu, you should watch The Grammar of Happiness if you can find a copy.
Youtube has some short clips. This one has a bit of whistling, among other things:
https://www.youtube.com/watch?v=qsZlY43jI2I

Quote
There is a lot here, but Sarah Taub's book might be a reasonable starting place.
Ok, I'll look for a copy.

Quote
The first thirty years of sign language linguistics was dedicated to establishing the continuities of sign and speech. This was an important political project at the time, as the common sense through most of the 60s, 70s, and 80s was that signing did not actually constitute "language" in any rigorous sense.

As is now extremely clear, that's a bogus position, but it's really only in the last ten years or so that it has been politically viable to argue that there might be significant structural differences between spoken and signed languages. And, over the last ten years, that's the direction the research has gone in force. It seems increasingly clear that signing is organized on some very different first principles than spoken languages are. As a result, it's very difficult to unify spoken and signed languages within a single set of typological archetypes.

I can't make this point strongly enough. I've spent half of my time over the last decade working on signed languages and the other half working on spoken languages. Most of my work in spoken languages is historical and typological. Nevertheless, I have a very, very hard time producing non-vacuous typological statements about any of the signed languages I work on. They just work on a fundamentally different set of principles.
Interesting and helpful. And I'm certainly aware of diversity, but there are some underlying similarities in all languages, at least at the conceptual level. (For example, I think modification is a universal, not that it's surprising.)
I suppose I'm asking a hypothetical: what if I were able to support my thesis with evidence from signed languages as well? So I'm looking. If I don't find it, that's ok.
More broadly, somehow or other, "coordination" must be possible in signed languages, so I'm interested in what this could reveal about semantics and the structure of language, especially if there are significant differences compared to spoken languages.
I suppose the idea of "what could a language look like?" is usually central in my mind-- so exploring diversity (while, yes, starting by assuming/seeking similarities) is important!
Welcome to Linguist Forum! If you have any questions, please ask.

Offline MalFet

  • Global Moderator
  • Serious Linguist
  • *****
  • Posts: 282
  • Country: us
Re: Differences between signed languages and spoken languages
« Reply #23 on: January 04, 2014, 04:12:54 PM »
Interesting and helpful. And I'm certainly aware of diversity, but there are some underlying similarities in all languages, at least at the conceptual level. (For example, I think modification is a universal, not that it's surprising.)
I suppose I'm asking a hypothetical: what if I were able to support my thesis with evidence from signed languages as well? So I'm looking. If I don't find it, that's ok.
More broadly, somehow or other, "coordination" must be possible in signed languages, so I'm interested in what this could reveal about semantics and the structure of language, especially if there are significant differences compared to spoken languages.
I suppose the idea of "what could a language look like?" is usually central in my mind-- so exploring diversity (while, yes, starting by assuming/seeking similarities) is important!

Indeed, but we're not going to get there by begging the categories. You're appealing to something you call "conceptual" coordination, but I suspect that's not actually something that can be defined in coherent, positive terms. As a very dear and recently deceased mentor of mine used to say, "if your claim can be made equally well about an oil painting, you're no longer talking about language".

On the one hand, of course signed languages have some way accomplishing the work done by English coordinators. The remarkable thing about the human communicative faculty is that it does not appear to be rigidly bound to the constraints of its various formal systems. But, that fact of possibility doesn't actually tell us anything about a given language's structural organization, and most linguists are interested in precisely that question of structure.

That's the punchline here: in the absence of a proper (and compatible!) theory of constituency, there's no way to meaningfully debate whether or not signed languages have conjunctions in the morphosyntactic sense that we are familiar with for spoken languages. It's a modality of language built on a very different set of primitives.

Offline Daniel

  • Administrator
  • Experienced Linguist
  • *****
  • Posts: 1421
  • Country: us
    • English
Re: Differences between signed languages and spoken languages
« Reply #24 on: January 04, 2014, 04:48:41 PM »
By the way, which book are you talking about? Taub (2001): Language from the Body?
Unfortunately there's no mention of coordination/conjunctions in it.
As I said, this doesn't appear to have been studied much. I'd love to find the sources, and perhaps at some point I'll have a chance to investigate it myself.

Quote
Indeed, but we're not going to get there by begging the categories.
Perhaps, but what do you suggest, coincidence? Start with intuitive similarities and work outward.

Quote
You're appealing to something you call "conceptual" coordination, but I suspect that's not actually something that can be defined in coherent, positive terms.
That's exactly what I'd like to understand. And I don't know that I agree. My best formulation of "coordination" at the moment would be:
The instantiation of a set of constituents as a syntactic enumeration.
I think coordination is a type of quantification, and also sometimes a compression of multiple sentences, though they themselves are then, I believe, quantified under the set of what is being declared. Necessarily, I'm talking about canonical coordination, not things like pseudocoordination and whatever else, and this canonical coordination is conceptual (at the level of meanings), while real-language conjunctions and coordination structures may encode more (or sometimes just different) things.
But that may be getting off topic for this thread  ::)

Quote
As a very dear and recently deceased mentor of mine used to say, "if your claim can be made equally well about an oil painting, you're no longer talking about language".
I agree entirely and don't think that applies here. I'm not claiming I've solved anything-- I'm very much trying to figure it out, though.

Quote
On the one hand, of course signed languages have some way accomplishing the work done by English coordinators. The remarkable thing about the human communicative faculty is that it does not appear to be rigidly bound to the constraints of its various formal systems. But, that fact of possibility doesn't actually tell us anything about a given language's structural organization, and most linguists are interested in precisely that question of structure.
Yes, very interesting. I'm interested in the way that form maps to meaning, and that, as you say, it is remarkably underconstrained. I've chosen an arbitrary problem ("coordination") and I am exploring it . In the end I may conclude that "coordination" doesn't exist anyway, at least not as has been previously asserted. But I don't think there's anything at all wrong with that methodology-- it's flexible. I'll follow the data, but start from looking for similarities-- how can we know that an apple and an orange taste different if we don't first know that they're both edible and have taste?
We seem to be concerned with the same thing, yet you find it a reason to not ask the questions I'm asking, while I find it to be the very reason they should be asked-- I'm fine with answers in the form of yes or no, or even "that's undefined". (I don't even claim they're necessarily well-formed questions, but I do believe that if they are not there should be a way to determine why they are not.)

Quote
That's the punchline here: in the absence of a proper (and compatible!) theory of constituency, there's no way to meaningfully debate whether or not signed languages have conjunctions in the morphosyntactic sense that we are familiar with for spoken languages. It's a modality of language built on a very different set of primitives.
Which gets to the point: they're different, so that's why obvious "and" and "that" words (conjunctions) aren't easily identified. That's great. Further investigation should then reveal whether some roughly equivalent set of linguistic forms (or structures) exists or whether signed languages are fundamentally different. I'd be surprised by that latter possibility-- I'd imagine that being able to say "John and Mary" is a fairly universal requirement, but of course certain claims about Pirahã prove me wrong (although even there strategies involving multiple sentences exist).

That's the whole point: what is coordination? If it's something special, it should be found in all languages. If not, then it's a language-specific (if cross-linguistically widespread) strategy to encode something deeper-- probably multiple propositions or a set of propositions, which is exactly what I think it ends up being. Moreover, I'm interested in what these forms do-- if they're somehow privileged they shouldn't be too flexible, but they are, so it appears that coordination is just like so much else in language-- flexible, non-universal, and non-basic: coordination isn't an "answer" but a label applied to a complex form-meaning pairing that should be understood in smaller parts.


So I really appreciate your comments, but don't feel the need to argue to me that signed languages may differ. I take that as a given, but not as a reason to not ask these questions.
« Last Edit: January 04, 2014, 04:53:34 PM by djr33 »
Welcome to Linguist Forum! If you have any questions, please ask.

Offline MalFet

  • Global Moderator
  • Serious Linguist
  • *****
  • Posts: 282
  • Country: us
Re: Differences between signed languages and spoken languages
« Reply #25 on: January 04, 2014, 08:07:32 PM »
Daniel, my dear man, we are once again at a familiar point. I'm not telling you that you should not study this question. I am telling you that the way you've framed it -- as a yes or no question with the category itself taken for granted -- misconstrues the language data in a very fundamental and (to signed language linguists) very familiar way.

As I keep trying to say, we are a long, long way from understanding even the basic elements of constituency in signed languages, and consequently a definition stated "The instantiation of a set of constituents as a syntactic enumeration." is not meaningfully applied. There is good evidence to suggest that it will never be meaningfully applied, too. This is, for example, the broad point of Sarah Taub's book. If you came to the conclusion that it does not discuss coordination on the basis of an electronic search of some sort, I would encourage you to go back and actually read the text properly. From beginning to end, it is an account of why syntactic function in ASL (such as coordination) is not usefully understood as a distributed property of constituents.

You tell me that I don't need to convince you that signed languages may differ, but the scope of possibility you're allowing for this difference tells me that, in fact, I most certainly do need to. Signed languages exhibit some very unexpected properties that make them difficult to render in terms of spoken language properties.

Most signed languages (perhaps all?) possess a viable register of concatenated lexical forms, but they also have a register in which all "syntactic" information is expressed through elaborate iconic deixis: spatial diagrams that coordinate and colocate through the sometimes literal and sometimes metaphorical application of space and time to semantic distinctions. In ASL especially, signers shift fluidly between these registers, but it's not a free variation. Though it is very hard to describe the distribution coherently, signers generally have very strong intuitions about which is appropriate at any given moment.

The "lexical" forms themselves are not obviously lexical, either. Cory hinted at this earlier, but the scope of available iconic modification vastly exceeds simple contrasts of degree. There's a reason that most ASL dictionaries contain virtually no adverbs, which is that remarkably little of this semantic domain can be identified to a stable, citable form. To sign 'walk drunkenly' you sign 'walk'...but you do it more drunkenly. That scope of available iconicity permeates the language, and it presents some basic and very fundamental challenges to axioms of modern linguistic like morpheme compositionality.

If the implications here for something like conjunction and coordination aren't obvious, consider trying to categorically transcribe something like this clip:

http://www.youtube.com/watch?v=I2nX41KvnNY

If you can do it, you'll walk into tenure tomorrow. If you can't do it, let it give you an inkling of why most contemporary signed language linguists are abandoning questions about syntactic constituency as fundamentally incommensurate with natural signed language data. We may as well be asking whether vine-ripened tomatoes have yet given up their childhood dreams to play professional baseball. There is no correct answer because the question is wrong.

Offline Daniel

  • Administrator
  • Experienced Linguist
  • *****
  • Posts: 1421
  • Country: us
    • English
Re: Differences between signed languages and spoken languages
« Reply #26 on: January 04, 2014, 08:52:29 PM »
Quote
Daniel, my dear man, we are once again at a familiar point. I'm not telling you that you should not study this question. I am telling you that the way you've framed it -- as a yes or no question with the category itself taken for granted -- misconstrues the language data in a very fundamental and (to signed language linguists) very familiar way.
I don't mean to sound like I'm not taking what you say seriously. I'm just not sure how to approach it-- perhaps I should, for this project, only look at spoken languages, if, as you say, it's really so hard to compare them. But I'm not convinced. I also am not confident that I have nearly enough knowledge of signed languages to figure it out myself, so that might just be a practical reason, which is fine-- if I can't cite existing literature, doing my own field work isn't particularly possible at the moment.

If the question misconstrues the data, then I'd be happy to adjust it to ask the appropriate question. But if that means reinventing an entire way to describe signed languages, I can't possibly do that for my dissertation, so I'll just leave this as something on my to-do list of interesting questions.

Quote

As I keep trying to say, we are a long, long way from understanding even the basic elements of constituency in signed languages, and consequently a definition stated "The instantiation of a set of constituents as a syntactic enumeration." is not meaningfully applied.
Ok, that's fair.
But we can work out "constituent" in a more intuitive sense that MUST apply to signed languages also, even if this is not comprehensive or reliable.

That approach would be extensional:
Coordination of NPs in ASL would involve having two nouns (things, entities, whatever you want to call them-- referent entities anyway) and applying the same external relationship to both within (roughly) the same structure.
That would cover sentences like "dogs and cats eat" or "I like dogs and cats".
Most languages (again, there could be exceptions) give us some way of encoding those ideas-- create a list of things, apply them as an argument of a predicate. Something like that probably exists in at least some signed languages.
It might very well turn out that we can only identify the things then, and not the "structures" or "coordinators" involved. That's fine. I was curious to see if there are words like "and" that are obvious-- if not, it might be eye blinks or just timing or whatever else I can't imagine at the moment.

Coordination of VPs in ASL would involve two action words both asserted of roughly equal status in the structure, within a single structure. (I'm not a fan of distinguishing "serialization" as something special-- that's a form, not a meaning or underlying structure. It should be included in this, along with things like coordination of VPs in English.)
Obviously the line between multiple sentences and coordination is blurry, but it is in spoken languages too. In some spoken languages (but no signed languages, you say?) it's relatively easy to tell the difference.


Again, these definitions aren't perfect. But they should suffice to at least explore whether things seem to line up and if not to bring up some new questions to ask.


There's a very interesting strategy I read about in Quechua. It involves a list of nouns followed by a numeral (of the length of the list). It is effectively the same function as coordination. I'd be more than happy to consider it (broadly) a kind of coordination at a functional level. Then I'd investigate the differences. The data looks something like this:
John Mary Bill three ate sandwiches.

To me, that's fascinating-- it's doing the same thing as English coordination, but it's doing it by another kind of instruction to the parser. Again, as I see it, this is all about linearization.
(There are some older-- 1970s/1980s-- theories about coordination being a sort of mix of multiple layers of a sentence. That's roughly like what I'm thinking, though I don't know if I would maintain all of the details of those analyses. Essentially the idea there is that you line up several similar sentences then pronounce them as one.)

Anyway, I'm certain that if I knew ASL better (=at all) I would be able to at least figure out what data I'm trying to explain, though I'm equally certain I wouldn't be able to figure out every detail of it.

Quote
There is good evidence to suggest that it will never be meaningfully applied, too. This is, for example, the broad point of Sarah Taub's book. If you came to the conclusion that it does not discuss coordination on the basis of an electronic search of some sort, I would encourage you to go back and actually read the text properly. From beginning to end, it is an account of why syntactic function in ASL (such as coordination) is not usefully understood as a distributed property of constituents.
Ok. I will try to look through it in more detail at some point then. This is relevant. But it isn't helpful for actually immediately studying coordination in ASL. I have no objection to that book not mentioning it-- in fact, most descriptive grammars devote less than a single page (if anything) to coordination. I'm often working with limited data. But certain Taub, if she had wanted to, could have done a better job at least mentioning one of the relevant terms to explain why it isn't found. I'm not using that book for its intended purpose, so again I'm not claiming there's any fault with the book, but it certainly doesn't make it easier to access the material.

Quote
You tell me that I don't need to convince you that signed languages may differ, but the scope of possibility you're allowing for this difference tells me that, in fact, I most certainly do need to.
I'm asking hypothetically, even if I phrase it directly-- assuming that language X has coordination, what does it look like, and if it doesn't, then why doesn't it, while still leaving room for the fact that we might just not have documentation for it yet either.
I'm comfortable with my methodology and assumptions, but given evidence I'm happy to be flexible about it. I also won't continue looking at signed languages if they don't seem to be giving me relevant data. I'm just spending some time to see if they do.
Quote
Signed languages exhibit some very unexpected properties that make them difficult to render in terms of spoken language properties.
I certainly agree. This is why, I think, the terrible written glosses in books come across as meaningless to me. I can relatively easily imagine Dyirbal from a book and the data looks somehow convincing, but an ASL description just looks like it's pretending. Part of that, admittedly, is that it uses English words in all caps as if that's data. I don't have a better suggestion, though. I might actually prefer printed hand shapes, but that would leave out too much data.

Quote
Most signed languages (perhaps all?) possess a viable register of concatenated lexical forms, but they also have a register in which all "syntactic" information is expressed through elaborate iconic deixis: spatial diagrams that coordinate and colocate through the sometimes literal and sometimes metaphorical application of space and time to semantic distinctions. In ASL especially, signers shift fluidly between these registers, but it's not a free variation. Though it is very hard to describe the distribution coherently, signers generally have very strong intuitions about which is appropriate at any given moment.
This is fascinating. At the same time, speakers of spoken languages do this all the time (in different ways obviously). Reasonably, we'd pick one register to start and work from there. More registers just means more data-- more time to analyze, but better data in the end.

Quote
If the implications here for something like conjunction and coordination aren't obvious, consider trying to categorically transcribe something like this clip:

http://www.youtube.com/watch?v=I2nX41KvnNY

If you can do it, you'll walk into tenure tomorrow.
I can identify some parts, and I can't identify others.
Although ASL might be harder than English, I don't know that they're fundamentally different for the analyst: our job involves making conclusions from incomplete knowledge/understanding. I don't know every property of English but I can make some generalizations. The scale may be different, but in the end I'm pretty sure I can identify a few words in that and probably some phrases. In my limited understanding, my "theory" of constituency is not fully deterministic of course, but I could still make some relevant observations.

Your approach in general seems to be wanting to understand everything before analyzing anything-- I'm well aware that my conclusions might be wrong if I'm not aware of the relevant data. And certainly I can't personally do a sufficient analysis of ASL yet, and perhaps no one can. But I'm not convinced that it's impossible. Perhaps I should wait 20 years and check in with the field again.

The potential for failure just makes me more curious. I like trying things out. If it doesn't work, then it doesn't work. I'll try something else.

Quote
If you can't do it, let it give you an inkling of why most contemporary signed language linguists are abandoning questions about syntactic constituency as fundamentally incommensurate with natural signed language data. We may as well be asking whether vine-ripened tomatoes have yet given up their childhood dreams to play professional baseball. There is no correct answer because the question is wrong.
But surely the domains are closer than your example. My working hypothesis (which might be wrong) is that languages encode concepts and relationships between those concepts. I don't see why ASL would be an exception. How they do this in form is of course relevant, but I imagine that any form is fundamentally a means of encoding those concepts and those relationships between them. Even in ASL. Constituency may not be the only way to do this (in fact, broadly, I'd be surprised if it is), but I would think that underlyingly you could relate it to concepts and relationships in some sort of constituency themselves. But again, that's getting way beyond this thread in abstractness. I don't fully understand my own idea on that yet (related to spoken languages) but I'm working on it. I'll happily let you know if I ever "finish" (perhaps completing my dissertation, in a year and a half or so, counts as a step).


Now... trying to get back to something productive out of all of this:
You make a very relevant observation: constituency is not a coherent idea in signed languages. This would explain two things-- 1) why basic coordinators are not found in ASL or other signed languages, and 2) why linearization appears to be different in signed languages.
So, arguably, my questions were originally well-formed-- you answered them. The answer: constituency is not the same in signed languages. That would certainly mess with things like are being discussed here. Moreover, there appear to be additional properties of signed languages that allow for encoding of information differently than is possible in spoken languages (eg, spacial diagrams); this could explain why, even to the degree that the structures might be parallel*, there is no grammaticalization of conjunctions-- something else already fills that functional need-- conjunctions are a strategy of spoken languages due to what (limited) resources they have available.

*There's no reason we couldn't have a signed language parallel in structure to a spoken language. We could have signed English or signed Basque or whatever. Words, word orders, constituency, and syntax.
But we don't. That's what becomes interesting, and the reason for my broader questions earlier in the thread.



So, forgive me, if I do happen to spend some time asking tomatoes which professional sports they'd prefer to play. You might be right about that. But if so, I don't mind having asked the question, even if just to find out that there was a better question to ask instead.



(That video you linked to is amusing. I don't know that we need ASL to show us why rap music is crazy though :D )
Welcome to Linguist Forum! If you have any questions, please ask.

Offline MalFet

  • Global Moderator
  • Serious Linguist
  • *****
  • Posts: 282
  • Country: us
Re: Differences between signed languages and spoken languages
« Reply #27 on: January 04, 2014, 09:39:12 PM »
Quote
Your approach in general seems to be wanting to understand everything before analyzing anything--

Ha ha. I'm sorry to laugh, but this is asinine. Where on earth did I say anything even remotely like this? It's not a question of understanding "everything". It's a question of understanding *something*. It's a question of letting the data lead the theory and not the other way around. There's no need to reduce that to psychobabble.

I understand that you are attached to (little-g) generative categories of analysis. I'm very attached to them too in my work on spoken language typology. They've done wonders for the science of language. But, it's worth taking a historical view of the situation.

Signed language linguistics got its start roughly 50 years ago, about a decade after the first Generative work. Here is a list of things that have *not* been accomplished in that time:
* A viable transcription method.
* A non-vacuous typology.
* A phrase structure.
* A robust phonology of even one signed language.
* A set of functional word categories.
* A way of identifying minimal pairs
* A reliable definition of "word", "sentence", or "morpheme".

If this had been the scope of Chomsky's progress in spoken languages, none of us would have ever heard of him. You can call my reluctance to take too much for granted a desire to "understand everything before I analyze anything", but I'd argue that linguistics would be a heck of a lot more interesting if we didn't always assume from the outset that we understand how everything works. I don't mean to pick on you specifically here, but you'd be surprised by how many times a week I'm told by people who don't know any signed languages how signed languages "must be".

Offline Daniel

  • Administrator
  • Experienced Linguist
  • *****
  • Posts: 1421
  • Country: us
    • English
Re: Differences between signed languages and spoken languages
« Reply #28 on: January 04, 2014, 10:06:36 PM »
Quote
Ha ha. I'm sorry to laugh, but this is asinine. Where on earth did I say anything even remotely like this? It's not a question of understanding "everything". It's a question of understanding *something*. It's a question of letting the data lead the theory and not the other way around. There's no need to reduce that to psychobabble.
You have useful ideas and I often enjoy discussing them with you. But very often when you enter a conversation it jumps to the most abstract level of questioning whether the questions can be asked. I don't mean to suggest that I'd prefer you don't enter the conversations-- it just would be nice to be able to discuss a topic without redefining everything along the way.

Your position regarding signed languages seems to leave three possibilities:
1. You can't ask such questions. Don't bother.
2. First, revolutionize the field. Then, maybe, it will be possible to answer it.
3. It's just different, so we can't know.

Surely there's a simpler answer to this, such as an extensional definition of what seems to be normal coordination in ASL-- how do you have, for example, multiple nouns as the subject of a sentence? Variation is fine-- give as many (or as few) details as you'd like.

In the end, you seem very concerned that I will take a simpler answer as some kind of certain fact that will inherently misrepresent the real data. I assure you that's not what I'm doing.

aaanyway, let's not dwell on that. Discussing our discussing is not too interesting.

Quote
I understand that you are attached to (little-g) generative categories of analysis.
This is incorrect, actually. To clarify:
1. I have been educated in generative theories and I often have major objections to them.
2. I don't think language is deterministic or "generative", although I do believe we are able to generate sentences.
3. I think that anything along the lines of what Chomsky imagines exists at a conceptual level-- what I described above as concepts and relationships between those concepts. I don't think G/generative approaches can adequately explain linguistic form.
4. If anything, I'm limited by my inexperience with other theories. I should probably educate myself on others, but being an American Ph.D. student I'm required to submit to the major theories out there, though I eventually intend to move beyond them.
5. To a certain degree, I like Generativism, even though I don't necessarily think it's the right theory for language. I don't object to the theory, but the fact that it doesn't explain the data well ;) I'd actually like some kind of deterministic approach if it were possible, so I'm exploring to see just how much could be explained that way.

I hope that clears it up. I certainly don't mean to give the impression that I'm only open to generative explanations. (However, I do like explanations rather than hand-waving-- I have yet to find a theory outside of generativism, and perhaps this is due to my inexperience, that doesn't feel like a bunch of hand-waving. So I find myself between theories, without one I really agree with.)


Quote
Signed language linguistics got its start roughly 50 years ago, about a decade after the first Generative work. Here is a list of things that have *not* been accomplished in that time:
* A viable transcription method.
* A non-vacuous typology.
* A phrase structure.
* A robust phonology of even one signed language.
* A set of functional word categories.
* A way of identifying minimal pairs
* A reliable definition of "word", "sentence", or "morpheme".
Great information, thanks! This really does help.

Question: all of those items you mentioned above are logical extensions of the basic idea of "contrast". Are contrasts well understood in signed languages? If you take two utterances that similar but differ is it possible to identify the ways in which they differ? Near minimal pairs at the sentencial level seem easy enough. How you proceed with the analysis after that might be tricky when it's unclear what the unit of a "phoneme" or "morpheme" is, but at the very least we could start by gathering some basic data. For example, ASL probably has equivalents to "John is happy" and "John and Mary are happy". This is where I would start looking for coordination.

Quote
If this had been the scope of Chomsky's progress in spoken languages, none of us would have ever heard of him.
True. But also remember that it just appears that Chomsky has made progress. He's convincing. That doesn't mean he's right. I don't know what scientific consequence that has, but it's imaginable that someone could apply his ideas to signed languages. I guess the difference is that, probably, others (like you) would not believe him. Personally I feel about the same (as you do with signed languages) for the variety of spoken languages out there-- it's laughable to think that list of items you mentioned is solved for the diversity in the world, with languages like Pirahã and Dyirbal and Basque, and whatever else. Even English for that matter. You'd think, if Chomsky has succeeded, that we'd at least have a great theory for English by now, but in reality a lot of (the "small" details in) English is still a mystery. What does that mean about the extent of variation found elsewhere?
Certainly Chomsky has made some scientific progress and moved the discipline along. But to think that anything is more "solved" in spoken language than signed language seems less certain.

Quote
You can call my reluctance to take too much for granted a desire to "understand everything before I analyze anything", but I'd argue that linguistics would be a heck of a lot more interesting if we didn't always assume from the outset that we understand how everything works.
I never do, and that's the problem.
I make hypotheses, which I am fully prepared to find are wrong (and I modify them constantly), and I work from there. So I make assumptions only to test them. This is what I love (and what often frustrates me) about the field. If there's anything I've learned over the past about 6 years I've been doing this, it's that I really don't know anything at all. I have some strong intuitions, though-- that's where I start, and I do find I'm often wrong. (Although compared to existing theories, I'm not sure I'm any more wrong than them.)

I don't mind making errors while exploring. Let's assume coordination is coherent and work from there-- if at the end of the exploration we need to go back and revise the entire concept of coordination (or find out it doesn't even exist) that's fine too.

Quote
I don't mean to pick on you specifically here, but you'd be surprised by how many times a week I'm told by people who don't know any signed languages how signed languages "must be".
The difference (I'm now assuming something about these people) is that I'm not committed to my position if there's evidence against it. But I'd still like to explore it.
Epistemic modals like "must be" aren't really that unreasonable, if you think about it-- they're only certain within one's current knowledge-- if there's more to know or something wrong about our beliefs, suddenly that can change quickly. And I'm very open to that.


I think one of the problems is terminological: you seem to be under the impression that when I say "coordination" I imply some kind of analysis. I don't. In fact, I have absolutely no idea what coordination is in any technical sense; I merely have a perception of some phenomenon out there (at least people seem to talk about it a lot and extensionally I can identify lots of instances). So when I use any technical term, I really mean "the stuff that seems like X, whatever X actually may be underlyingly". I use terms as bookmarks-- they remind me what I'm talking about and that at some point I need to define them more precisely, when possible.

So again, if coordination turns out to be something very different by the end of our exploration, great. But why not explore?

I certainly won't be upset if you tell me that coordination (or "something kinda like coordination") in ASL appears to be done via some very unfamiliar structure for me. I'm not asking you to invent "and" in the language-- I'm wondering whether it's there and either way why.
« Last Edit: January 04, 2014, 10:08:27 PM by djr33 »
Welcome to Linguist Forum! If you have any questions, please ask.

Offline MalFet

  • Global Moderator
  • Serious Linguist
  • *****
  • Posts: 282
  • Country: us
Re: Differences between signed languages and spoken languages
« Reply #29 on: January 04, 2014, 10:46:21 PM »
You have useful ideas and I often enjoy discussing them with you. But very often when you enter a conversation it jumps to the most abstract level of questioning whether the questions can be asked. I don't mean to suggest that I'd prefer you don't enter the conversations-- it just would be nice to be able to discuss a topic without redefining everything along the way.

Your position regarding signed languages seems to leave three possibilities:
1. You can't ask such questions. Don't bother.
2. First, revolutionize the field. Then, maybe, it will be possible to answer it.
3. It's just different, so we can't know.

This is just silly. I've stated my position a few times now, and it is of course none of these things. If that's not apparent by now, I'm not sure what more I can do.

I don't doubt that you experience what I say as as "redefining everything along the way", but the irony is that it's all fully canonical in the literature. In other words, though it may contradict your intuitions, this is well researched, well documented stuff. If you'd like me to pass some references, I'd be happy to. In the meantime, I've made my point and repeated it a few times. I don't need to make it again. I hope you enjoy your ASL class, and I hope you stick with it long enough to grok the language.
« Last Edit: January 04, 2014, 10:53:01 PM by MalFet »