Tuesday, 11 September 2012
DCAL
Deaf Children Development Conference – Part 2
GARY: We are ready to start again. So just as a little taster, we've got two more presentations, brief presentations, before lunch and then after lunch it's your turn. We've got question time, interaction, your view on things. So we are taking this in turns, we are giving you some information now, but hopefully you are going to give us back lots of information after lunch. We are thinking about food for thought, thought for food at lunchtime. I'm going to pass over to Gabriella Vigliocco, who is going to talk about how people learn sign language and how they understand sign languages and this is really relevant for people who are perhaps learning sign language as an adult or perhaps children in school who need to learn sign language. Over to you Gabriella. GABRIELLA: Thank you, can you hear me okay? So first of all I would like to thank Lilli for organising this because it's a great opportunity, also for someone like me who used to work mostly with academics, to indeed see to what extent what we are doing might have a relevance in the real world. So what I'm going to do is of course after I have introduced the people with whom this work has been done, that is the group I'm working with at DCAL, which includes some more senior people like Dave Vinson, Robin Thompson and Pamela Perniss, who is here today, and more junior people, Rob Skinner and Neil Fox, who is working with us currently. What I would like to do is to talk about 2 aspects of language and communication that are common between sign languages and what I would like to do is to ask whether and to what extent we can learn something about spoken language from sign language and to ask to what extent does the implications from looking at sign language in the way we theorise about language, the way in which we think about it, important issues like language revolution, language development and how the brain processes language. I will do this focusing on these 2 areas. The first one is iconicity, in terms of to what extent properties of a form, of a sign language, of a word in spoken language and sentences do resemble, have some visual link with properties of reference in the world, objects and actions in the world, and then I'll look at integrating to articulators, the hands and the mouth in language and what that tells us about sign language and what are the implications for the spoken language. So starting from iconicity; now sign languages have plenty of instances of iconicity, this is true in single signs and I'm giving you some examples of the sign for camera, of the sign for eating and also for more somewhat abstract domains like the sign for thinking in BSL and we see this across sign languages and I'm also illustrating here some examples in ASL. Now it is usually believed and usually argued that spoken language is iconicity challenged, but really that may not be completely the case and in fact in languages we do have cases where we can -- where the actual form of the word resembles the actual noise made by the animal for example.
But turning some languages -- going with the languages we know best like English and maybe Italian in my case -- but looking at Japanese for example there are more transparent links between some properties of the linguistic form of the word and some properties of the reference in the word and I'm giving you here the example for one word in Japanese that seems to capture, if you look up the definition, a very specific type of sensory sensation. In Japanese you have a very, very large repertoire of words that have this more kind of sensory feel to it. But then there's another fundamental property ‑‑a couple of other fundamental properties of spoken languages that appear to be a lot more iconic. These are the fact that people when they speak they gesture as well and these oftentimes reflect the properties of what they are talking about and also what is called in literature prosody which is the musicality in the language, how is the information ‑‑the envelope, the actual pattern ‑‑that is there. That also may reflect more closely and more vividly aspects of the visual experience, of the acoustic experience and so forth. So in a sense, although this has been neglected, iconicity is there quite present in spoken languages too. Nonetheless, because I regarded most theories of language has looked at spoken languages only, these are phenomena that are less common in spoken language, they are more clear in sign language.
Now, what we did in our work was then to begin to look at whether indeed ‑‑so there is iconicity across language: does it matter? Does it have an impact in processing? Does it have an impact in development? Well, before I get to that, the other thing that I wanted to mention is that there are elements of iconicity that of course leaps along arbitrariness and it's not the case that in different languages you would see the same word or the same sign for a given reference. So, for example, Italian and English speakers use remarkably different onomatopoeic words to refer to the animals up there and likewise, if you look across sign languages ‑‑and here I am illustrating it in ASL and BSL ‑‑here you have the sign ‑‑although iconic the sign for "cat" is very different in those two languages as well, nonetheless it is iconic.
Now, how did we start looking at the processing? Well we looked in one experiment I would like to present here at how fast the signers produce signs for given pictures. So they were presented in this case very simply with a picture, they were asked to keep their finger on a computer keyboard until they were ready to sign and then produce the sign corresponding to the picture and we measured how long it was taking them to produce the sign and the idea behind the study here is that signers would have been faster at producing signs that were more iconic, where there was more of this transparency for words that were more iconic. In the study we also looked at people who started signing earlier and people who started signing later. In the results that I will present, when I talk about iconicity, I will refer to ratings, judgments, that native signers have provided us on a scale from 1 to 7 where one is, it doesn't resemble any property of the reference at all, to 7 this implies, yes, it does resemble a lot of properties of the objections and of the action. So the basic results here ‑‑well, first of all, the basic result with respect to iconicity is that what we did observe is that signers were faster at producing iconic signs than non‑iconic signs. So what we have here is that increasing how iconic the sign was, people were actually faster in producing that sign. And that suggests that indeed it is somewhat ‑‑there is something about iconicity that makes the sign easier to retrieve.
We further found that ‑‑and probably intuitively, this is precisely what you would expect ‑‑that people who learn sign language, people who learn it as their native language, were faster than people who learned BSL later. What is interesting, however, is that iconicity facilitated, helped, for both groups. There was no difference. It is not the case that only those who learn BSL later would show the effect of iconicity or vice versa: both groups did.
So next what we asked this ‑‑sorry, I'm going to skip through this slide because of technical issues. Next, what we asked was, okay, we are seeing here that it seems like to help a little bit the production, but is this a case that indeed has any impact on development whatsoever and the way we asked this question is, well, if iconicity is having an impact on language processing, well perhaps then what we should observe is that children who are acquiring BSL as their native tongue, their native language, they would learn iconic signs earlier than non‑iconic signs. In order to address this question, what we did was to use really wonderful data that was collected some time ago by a number of colleagues, some of whom are here, that consisted of questionnaires that were given to parents and where parents were asked to tick a box for a list of signs as to whether their children comprehended, understood, those signs and produced those signs at different ages.
Here, I'm just giving you some examples of the signs that were included in there.
Now, what we would expect is that if, indeed, iconicity is playing a role in development, as I said, then we should observe that children are understanding and producing iconic signs earlier than less iconic signs. And here is the results for this work. Let me walk you through. We have in the first graph corresponded to the comprehension, understood signs, and the second one corresponds to production, to produced signs, and what we are looking at is the proportion of signs that are understood or comprehended depending upon the iconicity. What we found particularly interesting is that indeed children tend to understand and produce iconic signs earlier than non‑iconic ones and, moreover, there is a trend such that the younger ‑‑this is true especially for the older children, what we see here at "late", which is the dotted line, which seemed to show more of an effect of iconicity. So indeed we have that more iconic signs are understood and produced earlier than non‑iconic signs and this is particularly so for the older children.
Now, what are the overall implications of these results from our perspective? Well, I mean, it seems like overall this data ‑‑and we do have a number of other studies that have looked at BSL as well as ASL ‑‑indeed showed that iconicity makes signs easier to produce and understand and to learn. One possible implication is that indeed perhaps we could take more advantage of iconicity in the learning of both sign languages and spoken languages as a L1 and L2 whereby for spoken languages I believe the very critical implication is that we should look at these aspects like the musicality and the gesture much more closely than what we have done so far with respect to, indeed, whether these other aspects that have been classified and considered as linguistic only may play a crucial role in processing and may help both the development as well as the rehabilitation perhaps.
From my perspective, another fundamental implication of this work is that it forces us to re‑think our theories of language. It has been considered up until now that arbitrariness, the fact that language and the word are only arbitrarily linked one to the other has been considered a fundamental principle of language and the only fundamental principle of language, but now perhaps what we see here is that we came to this conclusion because we looked at some specific spoken languages. Perhaps if we would have looked more broadly to other spoken languages as well as sign languages, we would have seen that perhaps iconicity also may be considered as a fundamental principle that may help in making language meaningful, in linking language to our activities and to the world.
Okay, let me move now in my very little time remaining to the second area that I would like to discuss today with you, which is the integration between two articulators: the hand and the mouth in sign language and their implications.
So when we produce signs, what we do have is the hand pattern are combined with English mouthings and I cannot show it because of technical difficulties, but when signers are producing a sentence like this one, they would integrate the different signs with producing on the mouth at different points aspects of the phonology of English, which is a very interesting phenomena and sometimes it is necessary to disambiguate what is being talked about but nonetheless it is being done all the same.
So what we asked in this study is, what is the impact of this mouthing and, in particular, to what extent they are part of the sign lexicon, they are an integral part of the sign itself or to what extent they are in a sense showing us how signers are bilinguals, they are bringing together their two languages, from one articulator, a mouth and with the other articulator, a hand pattern. Now how do we go about trying to investigate this issue? Well what we did was to get signers to slip, to make mistakes, as we would do for speakers as well, when we want to see the processes that are engaged in production. In particular, what we were very much interested here was to see whether, when signers were making mistakes, so presented with a picture of a flower, say, they would produce the sign for "free", whether the mistake involved both the hand or the mouth, could involve only the mouth or could involve only the hand, and the idea here was that if they always go together, the hand and the mouth, well then it would appear that indeed that the hand and the mouth, although the mouthing are English patterns, they are integrated with the manual production within the mental representation for the sign. If no, well then mouthing and hand pattern might be taken as simultaneous bilingual productions. Here I'm presenting the critical results for this study. Let me walk you through because it's not immediate how to look at this. So what we have here is what was produced in the hand and what we have here is what was produced in the mouth in this experiment. So what is critical is when there were errors, the same error in both hand and mouth, and these are relatively few cases, if you look there are many more cases, 72 here, where there was an error in the hand but no error in the mouth. Now, even more striking is that well these errors were really, really few. Yes, of course, all language producers are very skilled and these errors, although we design experiments that try to induce them, occur only rarely, so that in fact for the vast majority of cases people were just correct across the board. Of course there were also a very large number of cases in which the manual pattern was not accompanied by mouthing, which may well be just because the test was actually remarkably boring. What is critical here is that they don't go together. It's not the case that when there is an error on the hand there is an error on the mouth. So it seemed to be clearly suggesting that what signers are doing is that they are combining their 2 languages, the English on the mouth and BSL on the hand. So indeed they reflect bilingual production rather than being an integral part of the sign and then indeed I guess one implication of this finding is that learning BSL in childhood does not interfere with learning English; they are independent. Okay so I'm going to stop here and I just want to thank you for listening to my presentation. Thank you. (Applause) GARY: I just want to emphasise, that trying to do this for Gabriella is like running a marathon, there is lots and lots of information that Gabriella has got across to you in a few minutes. I want to congratulate you, you did a great job. I'll open the floor for questions. Any questions, clarifications? NEW SPEAKER: I have read somewhere that you if you get a group of speakers together and they speak different languages it's going to take a lot longer for them to come up with a kind of inter language, whereas if it was 2 groups of signers with very different languages they would take something like 3 days to come up with some kind of inter language. I'm wondering if the process they go through is grouped to this. GABRIELLA: Very interesting, very nice question. Deaf people are much more common of poor signing, more used to having to cope with communication in a variety of ways in the sense they are recruiting their thinking abilities in a different way than speakers who would be in a sense more fixed on what their language should be. An example of this, the speakers take some time to adjust to people who have a different accent in a way that I don't think would be the case for signers. I think signers have the ability to adjust more quickly. Now, you may wonder to what extent is this related to their bilingual status, as bilinguals perhaps are more used to having to deal with different languages and so this helps in the adaptability. That's what they have to cope with on a daily basis or to what extent iconicity may play a role in this. You may ask, well if we were to look at Japanese and Korean people where both languages have a lot more in terms of onomatopoeia and words that more sound symbolic, more iconic, then perhaps for them it would be easier, but it's a great question.