 Well, and Susan will introduce him, but first I just want to explain a few things about the lecture series and about the discussion that follows. And while I'm talking, I'm going to pass around this bunch of papers. And if you want notes, receive notes, just put your name in the dress. I've been a little remiss. I have not yet sent out the notes from the last lecture. They will be coming to you next week. And if you want some of those notes and did not make a request last time or did not attend the lecture, you might want to make a special note on this list, also requesting the notes from Carol Padden's talk. As most of you know, we've been putting this ASL lecture series on so that we can bring to you current thinking and research on American Sign Language. And we want people to come away with new ideas, ask questions of the researchers, and basically continue on in their, I think there's a lot of excitement about American Sign Language. And we want people to continue feeling excited and curious and really understanding the full implications of this language. The coordinators of this series are Keith Cagle, who's sitting here in the blue sweater. And Susan Fisher, who's here in her rusty red blouse. And I. And in black. And Aaron Brace is interpreting for me. Susan will explain all the reverse interpreting and so on. We've set it up so that everyone can benefit from interpreters, both hearing and deaf members of the audience. Let's see what else. I have extra brochures if you want them. This is going to be basically two lectures today. The first will be given in English and the second in American Sign Language. And we're trying this for the first time this year. So when we send out the evaluation forms, be honest about what you think of that. It was a suggestion, actually many people last year suggested that and that's why we're doing it this year. After the second lecture, starting about 3.15 probably, we're going to be going to room 3120. And if you don't know where it is, stick with me or Susan or Keith. And there'll be tea and coffee and cookies and a discussion with Scott, which can be just sort of relaxed. I think it'll be far ranging and relaxed and doesn't have to be very serious. So come along and we'll talk about his ideas or any other questions you might have. Susan, would you like to introduce Scott? Okay. No, you have no choice. You want to interpret for me? Sure. Okay. Really? It is my very great pleasure to introduce Scott Liddell to you. Scott, Meg, Scott, Erin. Scott comes here immediately from Gallaudet College, which I found out at lunch today is this is the first time in his life he's lived on the East Coast. I hope he enjoys it and will continue doing so so that he will be close to where we are. Scott got his PhD in Linguistics from UC San Diego, where he did very important and groundbreaking work on American Sign Language syntax. Worked on relative clauses and on generally non-manual markers for syntax. In the last few years, however, he has been concentrating on ASL phonology, the sublexical structure of ASL, and has developed an important new tool for linguists, researchers, and I hope sign language users that can give us insights into ASL that maybe other systems cannot. And it's precisely about that that he is going to talk to us today, among other things. So please welcome Scott. Thank you. I don't know how many months ago it was that Susan called me. She'd like me to talk about something here. And we talked for a while about what, and eventually settled on this topic, which is what is the relationship between transcription systems and analysis and or theory. And so, since that's a topic I've been struggling with for about the last five years, it seemed like a good one to talk about. It's not an easy one, but it's kind of fun, and especially these days it's real exciting. So what I'd like to do is talk about the way transcription systems have evolved, what kind of transcription system I'm using right now and maybe get some feedback from you or bounce some ideas off you as to how that affects the way you analyze signs. First, let me just talk a little bit about why anybody would want to analyze American Sign Language. There are lots of people in linguistics who will analyze any language, just because it's an interesting thing to do. It's a real challenge, and it builds up a storehouse of knowledge about what people know about languages of the world. But in the case of American Sign Language, it's had some more practical benefits. For example, when I first started working on American Sign Language, which was about 1974, people were waiting for an announcement to come from somebody that American Sign Language was really a language. I remember people would be writing letters to the lab saying, when is an announcement going to be made that American Sign Language really is a language? Fortunately, we've come past those days, but there's been a real nice effect from that, and that is the demonstration by linguists that American Sign Language functions, has the structural properties of spoken languages. It's a demonstration that deaf people who know American Sign Language already have a language, and that if they learn English, that's a second language, in addition to the first language. So it causes educators to rethink their ideas about what deaf people know and the use of language in the classroom and so on. Another reason people are interested in sign languages is they want to know what are the universals about sign language. And as long as only spoken languages were studied, universals were universals of spoken languages, but as soon as sign languages were brought into the picture, then it's able to, one of the goals is to be able to separate out what things are universal to language as opposed to byproducts of using the vocal tract or byproducts of using the hands and body to produce a language. So anyway, this effort at describing signs has led to the problem of how do you write them down? You can't take pictures of all of them. When you talk about a sign, you can't just hold it up and say this one. It has to be some way of writing that down. And the first person to do that was Bill Stokey, and a book he published in 1960 called Sign Language Structure. And in that book he developed a notation system, and really it's more than a notation system, it's a theory of sign structure. And in that system he talks about all signs having three parts. And he called them tab, des, and sig, and interpreted into English that means location, handshape, and movement. Okay, so he divided every sign into a location, a handshape, and a movement, and with no segmental structure. And what I mean by segmental structure are in spoken languages words are put together through combinations of consonants and vowels, and each individual consonant or vowel is a segment. And you put three segments together like C, A, T in English, those three segments in that order produce the English word cat. And Bill Stokey's model of sign structure did not have any segments in it. It had three parts, but the parts were not segmentally organized, they were organized simultaneously. So every sign had simply a bundle of three parts. And that's one on the handout. I guess I can also put it on the overhead to find out how to turn it on. Each sign has a tab, which is a location, a des, which is a handshape, and a sig. I translated that into the box on the right here to show that another way of writing that is just put all that stuff in a box and say every sign has location, handshape, and movement. The fact that they're written one after the other here is of no importance at all. This is the convention of writing one for a second and third, but there's no ordering there. Okay, so let's go right straight to number two on the handout then, where I transcribed the sign to A. And that U shaped symbol is Stokey's symbol for the chin. So what he says is the location for that sign is the chin. This part that I'll underline here, I'll point to it right there. This part is a B hand with the palm facing towards the bottom. And then this last symbol right up here means move away from the bottom. So if you put that all together, you've got chin as the location, hand all the way down including the elbow as the thing that is going to act. And then moving away from the chin, you get the sign good. So that's how he transcribed the sign good. Okay, and it consisted of those three parts. And the chin doesn't come first and the hand comes last. All those three parts are organized together. The only thing that he designated as part of his system is this symbol right there that looks like a T. It puts the palm towards the chin. But it does not distinguish between this and that system. It could be either one of those or anything in between. But basically what it does distinguish is between that and that. So anyway, that's a real quick description of how his system describes signs. He had a specific list of locations. He said these are the locations at which a sign can be made. A specific list of hand shapes. These are the hand shapes that can be used in the production of a sign. And a specific list of movements. And all signs then in that system are a combination of one item on that list of locations, an item from the hand shape, item from the movement. That's not exactly true, but I mean that's what the theory says. There are some exceptional ways that that's actually done in practice. So anyway, that is the basis for how the system works. And what I'd like to do is go through some basically three issues and talk about how those issues can be dealt with using Stokey's model of sign structure. Then introduce a segmental model of sign structure and talk about the same issues are dealt with in that in a phonetic description system. And then finally introduce a morphonemic description system. Talk about how the same three issues are dealt with in that system of transcription. And you get very different results depending on how you transcribe the signs. Okay, the three issues that I've decided to talk about are first of all, what is a sign? It's a fairly basic issue if you're talking about a sign language. You have to be able to tell signs from non-signs. What sort of process produces compound signs? In other words, ASL makes compound signs like, for example, we'll get to it on number two on the handout, but you have the sign good like this, the sign night, and the compound good night. And one issue I want to talk about is what sort of process produces compounds from those separate signs. And the third issue I want to talk about is may not even seem like an issue, but it will become an issue I think as we go along, and that is what sort of process produces this sign? 300. You may not have thought about a question like that before, but that's the kind of thing that I enjoy. That's why I brought it here today. So those are the three issues. What is a sign? How are compounds formed? And where does 300 come from? Okay. First, let me say some things in general about Stokey's system of sign representation. It was actually published in 1960, developed something prior to that. And for more than 20 years had basically universal acceptance. I think it's fair to say that it was adopted all over the world. People interested in sign language, they might have to modify it, they might have to add new symbols and so on, but the basic conception of signs being organized that way was not disputed. And what's interesting to me is that in spite of that, in spite of the fact of how universally it was accepted, when you looked at people's research, who were working on like, for example, problems in morphology or problems in phonology, and you looked at that research and see to what extent the notation system was made you saw, you find that it was almost totally ignored. It was extremely difficult to find any of that notation in any papers dealing, for example, with ASL morphology. And the reason, I'll get to a reason, but it seemed to me that the reason is that although it was easy to find those three parts for each signs, the system was not particularly helpful at describing grammatical processes. The way signs change, the way signs get formed and so on. In fact, if you look at the Clima and Balooji's book on the signs of language, which has extensive descriptions of compounding, all kinds of inflectional processes, verb agreement and so on, I've looked through the entire book trying to find a single Stoke notation. And I don't believe, unless I miss something, I don't believe there is even a single sign notated in this notation system in the whole book. And so, you know, on the one hand it was very easy to learn and easy to describe signs using this system, but on the other hand, when it got down to describing grammatical processes, the system didn't lend itself to being able to describe those kind of processes. Okay, so let's get down to those three questions. The first one, what is a sign? Well, in the Stoke model, a sign is anything that has those three parts. If something is going to be a sign, it has to consist of one of those movements, one of those hand shapes, and one of those locations on those lists. Okay, so there are lots of things I would like to call signs, and I think most everybody else would like to call signs that couldn't be called signs in that Stoke system. For example, the number three. If somebody says, how many things are there, or the answer is three, and I do that, to answer the question three, then following the Stoke system of notation, I did not produce a sign. And the reason I did not produce a sign is because there was no movement. Okay, there was a hand shape, which is one of his hand shapes, and the hand shape was at a place, but it didn't do anything. And the fact that it didn't do anything made it be treated as a non-sign. Okay, so basically the signs one through nine, since they don't do anything all the way up to nine, were not signs. And then ASL incorporated a sign ten, because it does something. So ten is a sign because it does something, eleven is a sign because it does something, twelve is a sign because it does something. And you have a similar situation with finger spelling. A through I were not signs because they didn't do anything. J was a sign because it moved. K through Y were not signs, but Z was a sign because it moved. Okay, now let me take your question. Curious, the way some people sign six, seven, eight, nine, they move their fingers. Now that would be a sign. If it moved, it would be a sign. If it didn't move, it wouldn't be a sign. And that is according to Stoke's conception of what a sign is. According to common sense, everyone would agree that all of those are signs, but the system imposed that restriction and said if it doesn't move, it's not a sign. So there were things that I would like to call signs, like the manual alphabet. I'd like to call every one of those a sign. I'd like to call A a sign, B a sign, C a sign, D. And all the numbers. I'd like to call all of those signs too, in spite of the fact that they don't move. Okay. But the system says no, you can't do that. It doesn't move. Okay. The next question is a question of compounds. And that's what you've seen as number two on the handout or number two on the overhead. I've transcribed. We've gone through the transcription of good. The transcription of light shows two B hands. There was the first B hand, the second B hand, and then that X there means contact. They have two B hands and then a contact makes it. Now, let me make the two signs separately for you and then the compound and see if all of Stoke's parts are still there. Okay. So the sign good looks like this. The sign night like that. Now, good night looks like that. And all the parts are still there. Still a B hand at the chin that moves away. There's still another B hand there that this other hand moves towards and makes contact. All the parts are still there in this system. So in Stoke's description of how compounds are formed, it's a lot like the way English compounds are formed. You just put two words together and pronounce them as one. But all of Stoke's parts are still there so there's nothing missing. It comes out looking like the way English compounds are formed where you have one word and another word and just put them together. That forms an English compound. But if you look for the sign good, in other words, let me sign good night and when I do that you try to find the sign good. I don't think any of you saw the sign good in there. The way your hand moved, your wrist moved when you signed good night. Okay. But I didn't do this. Good night. I went straight from the chin to the other hand. Okay. But the notation system, here's Stoke's transcription of the compound of good night over here. You have to write down everything because it's all there. The B hand is still there, the chin is still there. All the parts are still there. So you have to write it down as if those two parts were just the same what side by side. But when you look at a compound as opposed to the individual signs, you notice there's a big difference between those signs. And again, let me talk a little bit about clean and belugies work on compounds because they did a lot of work on compounding and they noticed all those kind of changes. I just showed you in this sign like good night. But since they had adopted Stoke's notation system, there was no way to describe those differences using the notation system. So basically what they had to say without using any notation at all was say, well, these signs change. They shorten. They get smoother. And describe it in terms like that. But unable using that notation system to talk about any difference in the structure of the compound as opposed to the structure of the signs as they exist outside of that compound. Okay. So anyway, in Stoke's system, compounding is simply putting two signs together. And that's the way the transcription system treats it. It's the way the transcription system has to treat it. Okay. The third problem is the third question is 300. And that one is not easy to deal with in Stoke's system because as you see the note I put under 3A here, 3 wasn't even a sign in Stoke's system. So presumably 300 comes from the sign 3 on something else, but when you don't even have a sign 3, it's hard to talk about how you produce the sign 300 when you don't have anything to produce it from. So it's problematic dealing with this kind of a system notation when by definition, one of the parts you're trying to make a sign out of is not even a sign. So I will deal with 300 as I get into these others. Okay. Which is what it's time to do now. Let me talk briefly now about a segmental system of sign notation. This started in the idea that signs were segmented is something I first talked about in 82 at a linguistic site of America meeting. And that is the idea that signs can be broken up into segments just the same way that spoken words can be broken up into segments. And the basic idea is that the segmentation is done on the basis of the way the hand moves or does not move. So the two basic units, the two basic kinds of segments that signs are broken into are movements and holds. So going back and talking about a sign like, like a sign like think, the sign think first moves towards the forehead and then once it gets to the forehead it stops. On the moving part I call a movement, the stopping part I call a hold. So the sign think then consists of a movement and a hold. Similar to the sign know consists of a movement followed by a hold. And initially there was some discussion about well if it didn't stop it would go right through the skull. Of course it has to stop. It has nothing to do with segmentation. It has to stop because there's a barrier there. But you can make signs without touching the body. In fact you can make the sign know without touching the body. And when you do that it moves and it stops. Just like this. So even if it doesn't touch the body it still moves and then it holds. So the idea is that signs move and stop because that's part of their structure. And that those sequenced things, those segments are what build signs. So some more examples. A sign like power is hold movement hold. Initially here then a movement and then a hold. A sign like tie is movement hold movement hold. So it moves, holds, moves, holds. I think a little more of a pause between your description and your demonstration. So the sign tie moves, holds, moves, holds. And an even longer sign in terms of segments is a sign jump. Which is hold, movement, hold, movement, hold. So there's a hold here. Then a movement, hold, movement, hold. And now the big difference now between Stokey's description and this phonetic segmental model of sign structure is that each segment consists of parts like Stokey was talking about. Each segment has a hand shape. Each segment has a location. Each segment has some kind of movement or hold. Each segment orients the hand in a particular way. So it's not like brand new parts were created. It's more of a reorganization of the parts that Stokey developed in talking about structure of sign. But that particular, when I first introduced the idea of segments, the segments were all one bundle of features. And that idea simply won't work. And let me explain with a quick example why that won't work. And again, I'll use another compound to talk about why that idea won't work. The sign sleep, if you're going to talk about the sign sleep, I would describe it as a movement hold like this. First the hand moves, then it holds. Now the hand shape that it begins with looks like this. And the hand shape that it ends with looks like that. So if you were saying that all those features belonged in either one segment or the other, then you'd have to say that that hand shape belonged in the first segment, that hand shape belonged in the second segment. Now the interesting thing about that is that this sign becomes part of a compound in oversleep. And that hold, the final hold in sleep disappears. It does not do this. The hold disappears. But that hand shape and that location and that configuration don't disappear. They're still part of that compound. The hand still comes to that configuration at that place. In spite of the fact that there is no hold there. So this required a change in the way of describing the signs. Because you can't delete a segment and have part of it stay there. If you're going to delete something, it's gone. Okay, well what this required was the borrowing from spoken language, the theory of auto-segmental phenology. And let me go into example four here, which is an example from a spoken language. This is Thai to talk about a similar kind of phenomenon and how that is treated in auto-segmental phenology. Okay, Thai has five tones. Low, mid, high, rising and falling. Okay, if there are no marks on the word, then it's a mid-tone. So this is mid-tone, high-tone, mid-tone. And this is the way you ask the question, are you going or not? You would say, by the young, by the young. So by is a mid-tone, high-tone, young, mid-tone, by the young. Okay, now what's interesting about this is, you can eliminate that word for. In rapid conversation, the word or can be totally lost. And what you say instead is by young. Okay, and the interesting thing about that example is, that this, even though the word or is lost, the rising tone isn't. So instead of saying with a mid-tone on both by young, you say by young. The voice inflection rises during this word, even though normally it's produced with a mid-tone. So what's happened is the word, the segmental part of the word is gone, completely gone, but the tone isn't. The tone stays behind. And so the idea is that you have segments, and then you have things that attach to segments. And even though you delete the segment, the things that were attached to that segment stay behind. Okay, so in this case, you delete the two segments here, but the thing that was attached to those segments stays behind. And that's a similar situation of what happens in oversleep. You delete the whole segment, but the hand shape, the location, the placement of the hand is not deleted. That stays behind even in spite of the fact that the segment was deleted. So if you look at example number five, that gives you an example of, these are the segments in the sign good, old movement hold. And then all the information about the hand shape, where the hand is, how it's oriented and so on, is in a separate bundle of features. So this is where the hand is at the beginning of the sign. The hand shape, where it is, how it's oriented at the end of the sign. And another set of bundles for facial expressions. If you have a difference between facial expression or non-manual signals at the beginning of the sign versus the end of the sign, those features would go down there. Okay, that's a real quick introduction to that. That may be something we can talk about later if it's still not clear. Okay, let's get to those three problems. So now what about what is a sign? Can you include numbers now? The answer is yes, you can include numbers because it's just a hold with a set of features like hand shape, orientation, location. So it fits the model. The model says, yeah, it's all right to have that set of features to be a sign. What about fingerspelling or letters of the alphabet which do not move? No problem, again, they're just a whole segment, a single hold segment with a particular hand shape at a particular location. And so it allows more things to be included into what is a sign. Are there still problematical things? Yes. A problematical sign would be something like, it's a sign Wayne Smith showed me in Taiwan Sign Language, which if I remember right means watermelon. And the sign is made something like this, where the hands don't do anything at all, but the head is performing some action with respect to the hands. The old piece of it. Yeah, I remember that. And the question is, if it's a sign, how do you write it down? What is its structure? What is the structure of that sign if you're going to call it a sign? And it's kind of problematic in this kind of a model because the hands are doing a single hold. And yet you don't want to say there isn't any motion there because there clearly is motion of the head in that sign. There's no real easy way to talk about how you account for the fact that the head is doing something in spite of the fact that the hands aren't doing anything. So in this particular view, there's still problem areas in terms of what can be included as a legitimate sign and what things that you're just not able to transcribe. Okay, I think it's time for the next one here. Okay, the next question is, what sort of process produces a compound? Okay, and here, the notation system makes a huge difference when you're talking about what sort of process produces a compound. Because as you can see from number seven here, the sign, good, has three segments, hold, movement, hold. So what it looks like happens in the compounding process is that segments are lost. So this movement and that hold are simply lost in the compound formation process. The initial hold is kept, but those other segments are lost. And one of the repetitions of night is lost. And so what the result is, you begin the sign with the initial hold of good and then you do the movement hold of night and that makes the compound. So the way the analysis goes is tremendously affected by the kind of notation system that you're using. And in this particular case it shows that it's not simply putting two signs together, the way it has to be treated in Stoke's model. There are some changes that occur to those signs, some segmental changes, and you can describe those segmental changes in terms of deletions. Okay, so that's a fairly major difference there in the analysis depending on whether you're talking about a simultaneous representation or a segmental representation. The third question was, what about the sign 300? And that one's still a killer for this phonetic system because presumably the sign 300 is something that's built from the sign 3. The A there just represents all the set of features that puts the hand in that place with that configuration. And it's just a shorthand for saying a three-hand shape at that location, palm facing out, base of the hand facing down. I just used a symbol A. I could have used any other symbol, but it's just an abbreviation. Okay, the problem is 300 has a hold movement hold, but that set of features that produces this is very much like the set of features that produces this. The difference is that the fingers are bent and the thumb is bent. But you can't just simply have some set of features B for all of these because otherwise, 500 would look like this. If the B was the same, 500 would look like that. But 500 doesn't look like that. 500 looks like that. So all the features that go into making up this 100 here have to be different for 100, for 200, for 300, for 400, and for 500. There's a difference in every single one of those because the final hand shape is different. The final hand shape in 100 is this one. The final hand shape in 200 is that, 300, 400, 500. So since all those final hand shapes are different, that means you need five different things here to make 100 out of 100, 200, 300, 400, 500. And that's kind of a messy way to treat this process. Okay, so that, what has happened is now a further change in this model has taken place. And it has, what has happened to all those features that used to be in one bundle, and I'll put my, I'll just talk about them right here. All those features that used to be in one bundle are now in separate bundles. So everything that used to, here is, these are two fingers extended. This says that the two fingers are straight. The thumb is opposed. The thumb is straight. Whoops, this is a mistake here. That should say extended. The base of the hand is at this place. The back of the hand is facing towards the body. The base of the hand is facing down. So basically all these features now in separate tiers. And then you say, all right, now if that separation is made, how do you describe 300? And now the answer is consistent. It's the same thing for 100, 200, 300, 400, 500. And that is what it says is, there's a movement and a hole. On this tier, on the level that describes the shape of the fingers, the fingers have to be bent. And on this tier that describes the shape of the thumb, the thumb has to be bent. And so all together, just interpreting this diagram for you, it says the sign begins with this set of features on the left. And the thing after it, it stays that way throughout the sign. So the fact that there are two fingers extended does not change. There are two fingers extended throughout the whole sign. The fact that the thumb is opposed to those fingers as opposed to back here does not change throughout the sign. The fact that the hand is oriented in a certain way does not change throughout the sign. So all those things stay the same except for these features that change the configuration of the fingers and the configuration of the thumb. So by taking those same features and reorganizing them, then what happens is what causes the sign, what produces the sign 300, then becomes a simple question. And that is, it's a suffix. This is just a suffix. ASL is not supposed to have suffixes other than this one. But the answer turns out to be it's a suffix using this particular model of sign structure. In Stoke's model, you'd have to say it's a movement change because the hand shape is there, the location is there, what happens is you change the movement. But when you put it into a segmental model you get a completely different analysis of what kind of grammatical process it is. Here in this process you say there is this thing right here is a suffix, and its suffix is on two numbers. One, two, three, four, five. So the point I'm trying to get at is here it makes a fairly major difference in terms of the analysis depending on how it is you transcribe a sign and how you think signs are structured. Okay, so I've just taken these three examples. There are lots more examples we could talk about later if you want to but those three examples taken for a start show that it makes a real big difference in analysis depending on how it is that you think a sign is structured, whether you think it's structured in terms of a simultaneous unit or whether you think it's structured in terms of segments. So for just a real quick summary here the Stoke model of sign is some simultaneous bundle of features in a segmental model is some combinations of holds and movements and depending on which segmental model it's either a single segment, a segment broken into a couple of parts or a segment broken into a whole bunch of parts and whichever one of those you pick has a large outcome in terms of the way your analysis finally ends up in terms of what kind of grammatical process is responsible for that. All this has come basically during the last five years or so and it's been real interesting and exciting times working on this and it's hardly over. I think it's fair to say that there's not a consensus in terms of the way that signs ought to be treated and I expect over the next several years there's going to be a lot other kinds of models and a lot of different proposals about exactly how signs really should be structured but this will at least give you an idea of the kind of things that are happening with people who are working on sign structure and representation of signs and that it's not just a notational issue that these things really do bear on the way problems get analyzed. So why don't we stop here and see what kind of questions you have. I understand what you're saying about what a sign is. I'd be curious about the phenomenon of gesture. Sometimes I have a hard time deciding the exact difference between a gesture and a sign. Some of the students have asked me that one to know and I'm not sure you know how to explain it to people especially from the hearing world. That is a loaded question particularly when you said that is only a gesture. All signs are gestures in that they're produced by moving your hands and body in certain ways but when somebody says when they say but that's only a gesture How could you separate those two perhaps into two classifications? But when somebody says that's only a gesture of what they're saying is that it's not a sign and in the case of spoken languages that's fairly easy distinction to make because when you move your arms and body nobody would ever accuse you of speaking. But in the case of a sign language the division between gesture and sign is much more difficult and I can't tell you exactly I wish I knew the answer to that question it's something that comes up often with the students at Gallaudet. I thought perhaps we could answer that in terms of sign language is something taken from the sign language that deaf people use a gesture might be taken from a broader population gestures you'd frequently see from spoken languages used in combination with spoken languages. That will work in a lot of cases in other words simply making a division that way will work in lots and lots of cases but there are hard cases in the case of ASL where the things that deaf people do you're not sure if you want to call it a sign or not there's something I don't do very well but in response if someone asks a question and the response is yes I think you know what I'm talking about and you want to just use your face to answer yes yes with expression and there's a little twitch right here alongside the nose I don't know how you decide if that's a sign or a gesture it's never used as part of any sentences that I know of it's only used as a response but it's something that deaf people do in regular ASL conversations so it's a tough question and I admit and I don't know the answer to that one that's the gesture but I think you'd have to say that in Japanese sign language could I raise another question? Could you talk just a little bit if you don't have a whole lot of time about generalization to make about grammatical processes in your system that you can't make in other kinds of literature? Let me try to do this in a couple of minutes it's not easy but I know what you're getting at the typical thing that happens with grammatical processes especially with respect to verbs let me give you some examples of what happens and then talk about the way a segmental treatment of that would look signs like the difference between weight and this form of weight or look and this form of look basically what happens is all of this information from here on down stays the same, same handshape same locations involved, same orientations all those things stay the same what happens is this part up here the movement part is different so grammatically what happens is something that in spoken language is almost unheard of there may be some examples of spoken language but if there are it's extremely rare and that is to take the segmental information and replace it with new segmental information you keep the things that's not part of the segment take the actual segments replace them in other words instead of having a straight outward movement you have a circular movement like this but everything else stays the same the handshape stays the same the location stays the same all those things here stay the same what happens out of segment the segments are replaced basically what that is is a real unusual sort of grammatical process that takes the original segments and replaces them with another set of segments so that's a very unusual kind of grammatical process for a spoken language but it seems to be one of the major ways that particularly ASL verbs are inflected there are particular movement patterns associated with each inflection and those would be what the inflection would do would be take what the verb came with remove the segments and remove them add in whatever the correct segments are for that particular inflection so it's a very unusual sort of process but that's the way it has to come out of this analysis it seems like it's still focused very much on hands and shapes and movements of hands the problem with the Thai watermelon before in comparison it seems to be existing in other places in American Sign Language like a relative motion someone's got poles going by the poles are going by and the person might still be looking inside is that a really small problem? I'm not sure it's a problem anymore let me tell you why the reason it had to have been a problem in the model in which all of these were a single bundle was that you'd have to repeat everything all over again in spite of the fact that the hands weren't doing anything but now with all of these things separate all you need to do is non-manually put the non-manual features down here and those are enough to account for a movement so having all of those features separate and not having a movement depending on two sets of handshapes and two sets of locations or whatever now having a movement simply dependent on two anything getting from A to B any way whether it's non-manually or manually that doesn't seem to be a problem anymore so I forgot to mention that I was ready to do that but it slipped the I don't think that's really a significant problem anymore because now what defines a movement in this system is having two sets of features attached to it in this case there are two features here that attach to that movement two features there that attach to that movement so it's a movement and if there's only one of everything then it's something then it's a whole and the problem is that thing it seems like it ought to be a movement right? but it can be now not because there are two sets of handshake features involved but two sets of non-manual features yeah thank you I'm curious about your choices over the suffix I assume you'd call forth and try to describe that also similarly in the suffix think about those not necessarily not necessarily we're getting into a whole other thing but there is a class of signs that simply replace handshake like one week, two weeks, three weeks four weeks, five weeks where it's not built from the sign one it's built from the sign week and you put the correct handshake in that sign and this might be one of those four fifths where it is they're an established thing meaning place first place, second place or first, second, third and what you do is you don't take the sign one and change it what you do is you take that configuration and put a handshake into it and so my guess is that on that particular example it's the kind of thing that you put a handshake into rather than that you built from a number there are a whole bunch of those this year my students are telling me that first place is first place, second place, third place now two years ago they were telling me it's first place, second place, third place so that's one that seems to be changing but the point is that that is not and from my opinion that's not built from the sign one it's built from a movement from one place to another in a certain configuration what you do is you put in the right handshake so I think for your particular example that's one that where everything is there except the handshake and you simply add the handshake into that sign