 My student colleague and friend, Tucker Childs, I remember him from the olden days when I was a beginning PhD student and I gave him my first romantic sketch and he almost fell over backwards because the English was so bad. I hope my English hasn't improved a little since then. So we go back a long time since I started out as a beginning Africanist. Tucker comes from Portland State University in Portland. He is one of the most indefatigable field linguists I know of. I said I'm pretty fat like it though. He doesn't look it. So he has worked on a number of Atlantic languages for all of which he has put his first descriptions. He is also a leading scholar in the domain of Idiophone Research and has pioneered a new approach in the 90s which is now, you know, papers on this issue of non-classics, I believe. And he started recently to establish yet another field side, working on a new language in a multilingual context. And so I'm very happy to have him here today to tell us about you techniques for measuring multilingualism. Thank you very much for the kind introduction. It's always so invigorating to come here to see colleagues and people I've seen, you know, here and there over the years. Oh, hi. Speaking of what? We are his students, by the way. We hold the gate. Yeah, I just had it. Anyway, we can talk later. Anyway, it's just great to be here because it's so invigorating and such a stimulating place. You know, I sit back in my little hole in Portland, Oregon and no one, I can't talk with anyone about African linguistics or about contact phenomena, multilingualism, idiophones, pitches and creoles. There's just no one there. This is like a total holiday for me. Anyway, thank you very much for being here today and I hope everything will work out. Thanks for setting that up. I think the links are all good now. Anyway, Jeff, thank you for bringing me here. So these two projects that you that are interacting today are tremendously exciting. You know, I feel kind of funny even talking about my own work when there are these massive projects with cast of millions, cast of thousands, cost of millions. You know, my little documentation project is so tiny. But anyway, that's what I'm going to talk about today. Right. So pretty much everything that I'm going to talk about has already been talked about today. And I'm sorry for those who have been in my company because I'm going to repeat a lot of the stuff that you talk about. But I think it'll be all right. Okay, so let's see to get clear which. Okay, so multilingualism has now been recognized as the norm rather than as the exception. And I don't think that's anybody that that's anything that comes as much of a surprise to the people here. But it's certainly something that needs to be described and theorized. The projects that aren't going here are certainly wonderful in that regard. So instead of worrying about the ideal speaker here, we're going to talk about the actual speaker here, one that's typically multilingual. So I've been involved with African multilingualism since 1970 when I was a Peace Corps volunteer. And I was at it what at the time I thought was a tremendously exciting place. It was at the intersection of the three borders of Guinea, Sierra Leone, and Liberia. And at that intersection, you had, it was a Kisi area, but no one spoke Kisi in the towns. They spoke Mandinka, Malinke, or Fula. There were a lot of Fulas there. They controlled the cattle and the transportation. There were even some Lebanese merchants. So you heard some Arabic people from Mauritania, from Cote d'Ivoire. And there were three major market towns. One in Liberia, one in Sierra Leone-Quendou, and then Becadou, which was over the border in Guinea. In all three of these places, you had just terrific markets, really, really exciting markets. And the swirl of languages there was tremendously exciting, especially to me, a little white kid from a small Midwestern town. The white town was so small, you saw the population change from 350 to 400. That's a total, not thousands. So that was my first exposure to it. But all the towns were in the Kisi-speaking area, and the Kisi had been there, but the Kisi had been dominated for, well at least the trade in the Kisi area had been dominated for centuries by the Mandi expansion, which began in the first millennium, and gradually grew to be what it was at the time I was there. Other prominent languages were, of course, the pigeons that it had arisen in reaction to, or maybe in defense from the colonizers. And my thesis, in fact, my dissertation was on the substratal influence on the three extant pigeons in the area. So you had a sort of French, which I call Guinea French, Creole in Sierra Leone, and then Liberian English. And what I proposed to do was to see how the substrate would influence the structures of these three pigeons. And there was sort of a noble project at the time. People had just shifted their attention to the substrate, the African substrate in pigeon and Creole studies. So it was an exciting topic. And there were two other people that were sort of doing the same thing, people who knew about African languages at the same time. And so we had a nice little conversation going about it. But when I got there, I found out, when I got there to do my research, this is After Peace Corps, I found out that I didn't know enough about Kisi to really do the work. I could speak it, but the only things I could do were buy food in the market and talk to some girl or order a beer. But to do the sort of normal everyday stuff that you really need to know how to do was impossible for me. Of course, I was now a little bit of a linguist, and I knew that I knew nothing about the structure of the language, absolutely nothing. I ended up doing my thesis on the phonology and the morphophonology of the language of the mountain class system, but it just changed from a prefixing to a suffixing language, so there's a lot of interesting stuff to talk about. Okay, all three of these countries had been dominated by colonial powers. France and England directly had dominated Sierra Leone and Guinea. And Liberia had been dominated to some extent by the Americans. Firestone rubber, these big concessions, firestone rubber, the extractive industries like iron ore and timber companies, and they, of course, had a great influence on the economies of the country. And in fact, they were much responsible for the pigeons that arose. Okay, because I didn't know anything about what I was doing, I think this illustrates, well, the sort of hubris that a lot of researchers have when they go to Africa and get involved in a multilingual situation. You really don't know what you're doing. And I'm sure your laughter sort of supports me that you have to be there on the ground and you know a lot less than you really think you know once you get there. And hopefully what I'm going to talk about today will help and other people in avoiding some of the same errors that I committed when I was going. It's not only competence in doing the analysis and the fieldwork, but it's also the whole process of setting up a research project, you know, finding your indispensable collaborators and sort of being aware of how things can go wrong. We are just talking about it over there. You can expect everything to go wrong and so just be prepared. So the reason I'm telling you all this is to illustrate how difficult it is to characterize a multilingual situation. It's an elusive beast. You don't really know what you're doing until you get there and it's unlikely that you're going to know enough to do the work that really needs to be done. And so you divide shortcuts, you divide different mechanisms for getting at what you want to describe. And that's what I'm going to talk about today. Okay, so here's a brief overview of the talk. Okay, my focus is not on the pristine and disappearing situations of rural small scale multilingualism as Federico was talking about this morning, but rather on the more common situation in places I've worked. I've done field work in East Africa and Southern Africa, but my major work has been in Sierra Leone, Liberia, and Guinea. And I've seen many different kinds of multilingual situations. And you know, or at least I know on the basis of that experience, that there could be many different types of multilingualism. There can be situations where you have an urban variety arising. There can be pigeons. There can be this sort of flat monoclassic multilingualism. That's the new word we invented today. Or seriously, multilingualism in South Africa, I saw how the manipulation of language and the set of propagation or the enforcement of multilingualism by the apartheid government furthered the government's objectives. So it was a political tool as well. So you see all these different functions of multilingualism and that points to another difficulty, not only the technical difficulty of analyzing a linguistic object, but also of seeing it in the proper sociocultural context. That's just as much a challenge as anything else. Okay, this study forms part of a larger study aimed at representing multilingualism and collaboration with several colleagues at my home university in the departments of geography and urban studies, namely mapping multilingualism in Portland. Well, some of you might say, well, Portland's all white people. And to a certain extent that's true, but it's changing rapidly. And I think it'll be easier to handle than anything that exists in Africa. But nonetheless, they have one of the best urban studies departments in the country and the geographers there are quite good too. And I'm really looking forward to getting more involved. I won't talk about that much today, but that's the context in which you should see what I'm talking about. Okay, so the challenges to an empirical characterization of the situation is formidable. Just capturing the data is difficult enough. The analysis may prove even greater challenges. Measuring multilingualism, except in a very crude way, may be impossible, though I will suggest in the paper some recent innovations that may facilitate the progress, the process, and we've certainly seen some today. I have a rather lengthy review of the literature that I don't think I'm going to go through too much, but I do want to say something about the framework. So, all this comes from Vivian Cook and opposes a monolingual perspective to a bilingual perspective, what he calls multi-competence. And I'm glad I found out he was a male before I came here and could use that pronoun because I wasn't sure before I came here. But this is a perspective, the bilingual perspective, and it should be multilingual perspective, is one that I think is familiar to most everybody, so I won't spend too much time talking about that. But one thing that the papers in this edited volume, it's another one of these great big handbooks, you know, the cost of a million dollars and none of what it's going to ever buy, talks about the multi-competence. And they're mostly interested in, unfortunately, in SLA, Second Language Acquisition, and so you hear about how the learning of second language influences L1. To me, that's not a terribly important issue, but that's what the papers are mostly about. Nonetheless, the perspective is a good one, and I think one that's worthwhile for our purposes. One point that they make in one of these papers is talking about the variation that can exist in the relationship between languages. And there were some nice, well, I guess only the middle one's a real Venn diagram, but below separation, you had two circles, language A, language B, separate. And under interconnection, you had overlapping circles with, you know, a shared area, like a Venn diagram. And then the third one was integration with a coincidence, you know, the two languages are the same. So there are just many different ways languages can relate to each other, and that's true not just of the languages themselves, but also for the different parts of the grammar. And so this just points to one of the, points out to one of the difficulties in characterizing multilingualism. Another place I looked for encouragement or approaches to characterizing multilingualism were these two sanctioning institutions, the European one, C-E-F-R, and then Actful in the United States. And I'm going to talk about Actful because I was actually involved in developing the guidelines for measuring how well one spoke an African language. And it was kind of a fun experience because the people from Actful knew nothing about African languages, nothing whatsoever, but you had some of the world's leading experts, American experts on African languages. So Will Laban was there, David Dwyer was there, Rush Xu and so on. And so for me as a graduate student, it was a really heady time hanging out with these people who knew so much. Anyway, so one of the things we decided in one of the many workshops we had was that learning how to use earphones was considered to be one of these top, top abilities. So if you look at this, I call it a spike, I don't know, what do you call this shape? Got an idea? So you look down at the bottom and presumably, you know, there are very few language skills that you have when you're a low novice, whatever that is, that means saying bonjour, maybe nothing else. And then going up you acquire more and more competencies in the languages until you reach distinguished. We put earphones in the category of advanced high. So that's the second or third to the top. Distinguished is probably like a native speaker. Superior is pretty close. You maybe speak with accents but you can talk about pretty much anything. What these organizations are good for is sort of calibrating these levels and sort of differentiating between speakers just to let you know what one thing I didn't like about the whole process was that to test your ability in different languages and I spoke several different languages at the time so I was tested in a bunch of different languages. What they do is they push you as far as they could until you broke down. So you just felt like crying at the end of one of these sessions. It was horrible. They ask you questions and try to get you to use the subjunctive and French and to all these sort of crazy things. Anyway, so they at least have ways of evaluating proficiencies. Notice, however, that this is still the monolingual approach where you're testing someone in their ability to speak a language like a native speaker. You're not looking at the person's language ability. The next place I was looking around was at various missionary techniques, primarily SIL, but there are a bunch of other ones. Is that me? Oh, okay. A lot more romantic. I've worked with missionaries for many, many years in the field in West Africa and in Southern Africa as well. People working on click languages down there. You know, I have mixed feelings about them as everyone does. I've worked with missionaries for many, many years in the field in West Africa and I have mixed feelings about them as everyone does. But they've been very helpful and they've been pioneers in developing field techniques and programs that I've used. For example, I use Flex, so I like Flex a lot. It's good for developing a grammar and building a lexicon and those kinds of things. But of course, we always put an A-line. Bring it over here. Anyway, so I looked at some of their techniques and I didn't find any of them really very happy. They were useful, but you have to remember what their purpose is. What they're trying to do is to figure out if they need to provide translations of a Bible for a certain group. You have to figure out what's the most important dialect or how many different Bibles do we need? How many different teams do we have to send in? So that's very different from what I think we're doing. And there's also a tendency for them to split rather than lump. And that's, I think, because they give them sort of job security. They will always have Bibles to translate if they keep translating them into different languages, what they call languages. And I found that particularly true in my little group of languages, the Mel languages. They're not Atlantic anymore, they're Mel. And it's really, I'd say, two major languages with a lot of dialects that have diverged quite a bit. And they've divided them up into all different languages. So as I say, with my little group, I think they've overdone it. Okay, so I wasn't terribly happy with their techniques. I also looked at their literature in social psychology, domain analysis, code switching. I wasn't in language attitude studies, all different kinds of stuff. I can't talk about that, I think, anymore. So the next slide. I have some of the questions that I'm thinking about. So actually, what is it? And talking about your typology is one way to get there, I think. Is Africa special? This is a question that I'm often asked. I think, yes, Africa is special. And you can ask me why. And the question is if you want. Oh, here I go again. Is it different from multilingualism and other parts of the world? What are the governing ideologies to Africa? And then what are the social factors? What is the role of social factors in determining the extent to which an individual is multilingual, particularly with regard to gender and age? This is one place I think that research can do some real good. Local identity, ethnicity, language and nationalism, and then the orientation towards the city or country. Yeah, and so on. Okay, with these questions in mind, what I thought I'd do is see if I could understand better what multilingualism meant in this small town where I was working. Thank you, ELDP. Supporting this work. Can you see this map? Okay, is that too small? It's okay? Okay, good. All right, so if you look at this map, you see the yellow and green, and then there's crew down here. It looks like the Atlantic languages are really important and vibrant and stuff like that. But if you take away Wolof, Timney, Fula, and probably Seder, you're left with a lot of little languages spoken all down here. Yeah, if you look at Federica's area, there are a lot of little languages all over the place. And that's particularly true down in this neck of the woods where I operate. So this is Liberia. I don't know where are we? Yeah, this is Sierra Leone. I work around here and on the border. Anyway, so let's look at the next map. So this is a map of a so-called Bulham. Bulham is one of the groups that I think is only one language, but they're divided up into three. I did work on what they called CRIM, and that's so crazy because the language has no phonemic R. And Bulham is the other language right here, and then they show Sherbro being spoken all along here, which is also ridiculous. And then here's another one. They call it Mani here, but no one calls it Mani. They call it Mani. They do have M's in the language. And this is a dialect of Bulham that got separated by the Timmy at enough time depth so the language really is a different language now, I think. But it really is Bulham. There's one language, Bulham, and Kim, a long time ago. But the missionaries are still at it. I just talked to a woman. I said, well, did you read any of my stuff? Because I said, you know, it's really one language. She said, yeah, but I don't really think you're right. I said, have you been there? And Denny worked, no. So anyway, they're going to go on. So this is where the research site is located. They call it a peninsula. I don't know if it's a peninsula, or Shingi is the town that I'm located in. And this is a terrific fishing area. It's unfortunately being fished out. These big trawler ships are offshore and they're just scraping the bottom and taking away all the fish. Anyway, this is where we are. Okay, so if the world were a perfect place, this is what you'd want a multilingualism metric to look like. You'd want everybody in the community to be actively participating in the project, helping you design things, helping you achieve outcomes and do something that everyone wanted. It would also be easy and quick to administer. I'm not supposed to laugh at that. You want results, of course, that are reliable and valid. And I would like to have extrapolation possible on the basis of a representative sample, sociolinguistically sensitive and informed, takes into consideration reining attitudes and ideologies and no literacy required. I work with people who don't know how to read and write, generally speaking. And it should be conducted in the volunteers' language of choice, the volunteer or the person that you're getting the linguistic data from. And fun and non-antibodating. It should be no test. It should be easy to analyze. Results should be comparable across languages. And of course, that's the way it ended up, just perfect in every way. Yeah, so, you know, it's inevitably a sequence of compromises you get towards a particular goal. Okay, so people say, well, geez, why are you looking at ideophones if you want to find out about multilingualism? Well, you've already had one reason. And that comes from the, you know, the actful proficiency guidelines where ideophones are considered at advanced level of proficiency. And so if you control ideophones, it's assumed that you know the language. And you also know about sociocultural stuff. It's not just functional instrumental kind of stuff, but really a little bit beyond that because the ideophones are so socially implicated. Second reason was that they're found in all the languages in the area. The three major languages that I was dealing with in this study are Sherbro. This is the target language of that documentation. I meant the language everybody's switching to and Creole, which is the resident pigeon. Creole in Sierra Leone. Another big advantage to ideophones is that they're perceptually salient. You'll hear in just a sec. And also, I devised a way. This is my field methods instrument. This is all I had to do. I just carried in my pocket all the time. Any time I came across someone who was appropriate, I'd say, hey, you want to help me out? And people usually would. It would take 10 minutes to collect the data. Okay. Let me talk about the stimuli now. In this table, I don't think you can read this. I couldn't blow up the font, but there was a look at the numbers screwed up through. Anyway, these are all ideophones in a thesis done at Forta Bay College in Freetown. What she did was she listed the first column, ideophones, then the so-called popular context, like the definition. But really, ideophones are hard to define. Next thing was in a Mende sentence and then the translation. Okay, so from this list of... What was the total, 300 and something, ideophones? Something like that. No, 286. I recorded 66 and then selected 20 out of those for the study. And just to illustrate to you what... The person I recorded producing the ideophones was Momou Tazif Karoma, who was a professor of lecture and linguistics at Forta Bay College. He had supervised this thesis and also was Mr. Mende. He knew everything about Mende there was ever to know. I brought him along as a fellow lecturer to... You know, the inn, what's it called? It's like a summer institute of documenting endangered languages in the state, infield. Colang is its name now. I brought him along and he served... He and I taught a course... Oh, you were in that class, weren't you, Jeff? He and I taught a course on ethics, which was kind of interesting. And he also served as an informant. The problem was he didn't really know how to be an informant, so he'd always tell the answers to the students. And I said, no, you just have to be a non-linguist and not his buddy. He did it the whole time and sort of ruined the course. But anyway, here's him talking to me, explaining how to... What did he film me? Round of the... I don't know what I'm talking about. Look at that. Look at the color of the film. Color of the film. Color of the film. He also studies that. Go round quickly. Yeah, go round quickly and completely... Go round quickly and completely. Citizens here? Citizens. People who are full, come quickly. People who are full are... a Fendi. Okay, so... Fendi is caught in... is a... thread on that to make it close. Okay, well, you can get a feel for what it was like. So from our discussion of each one of the ediphones, and I use that discussion to inform my scoring of the individual ediphones, I cut out... Well, I cut out a couple of productions of the ediphone by itself and then ediphones and sentences. He usually produces his own sentences. He didn't like the ones that a student come up with. And then I put those in a bunch of files. Oh, yeah. So how did I choose the ediphones to use? Well, I wanted to get a range of iconicity. So... I choose... I chose some ones that were onomatopoeic all the way up to, you know, size, sound symbolism ones. Ones that weren't quite so clearly iconic. So in between those were something like the ringing of a bell, there's water falling, and then the movement of air. I tried to mix up the formal features as much as possible. Nailiness versus breadth of meaning. Some ediphones can mean only one thing. They use it only... They have severe selection of restrictions and a very specific and narrow context of use across the senses. I didn't find any smell ones, though. Smell ediphones aren't that common. Yeah, so all the different domains. I tried to pick ones that I thought were familiar. All right, so they were... Each subject was... could listen to three practice ediphones and then you had the ediphone produced in isolation from a written list and then it was pronounced twice by itself. Then it was used in one or more sentences, usually two. Here, let me give you an example. So this is an ediphone meaning state of being thin and small, and I don't think this is the sentence he produces, but it'll be used like talking about the monkey's arms. You see what I mean about perceptually salient, right? It's pretty obvious which word is the ediphone. And here's one that I bet everyone can guess. You weren't supposed to pick. Bad, money and bad, money and bad. He loved producing some of these. All right, so here's one that... Oh, so everybody got this one right. You know, there was the sound of a sheet. And this is probably the most clearly iconically ma'am of you. And this one, practically no one got it. But this one, people did get... No. Yeah, you don't. He's kind of a ham too, so... Anyway, you got a feel for what the stimuli were like. They were given 20 of these and they were asked to explain what the ediphone meant. And so far they've been evaluated on a scale of one to three. I think I might change that, but anyway, for right now, that's where I evaluate them on. Okay, so here's who the subjects were. I tried to get a mix of Sherbro and Mindy speakers. So some were all Sherbro, no Mindy. Sherbro and some Mindy. Mindy and no Sherbro. And then Mindy and some Sherbro. And everyone spoke Creole and some spoke English. Not really, nobody spoke English. The subjects were all pretty well educated. They'd gone through maybe 12 years of education. There were nine females and six males. Here are the instructions. We don't have to look at those, I don't think. Okay, let me move on to the findings now. Well, yeah, so what they had to do was they had to explain. And they could use whatever language they wanted in their response. And I did the interviews, I did a few in Creole, but it was mostly assistants who did it in either Mindy or Sherbro or Creole. Okay, so as one might expect, the judge forms something of a bipolar distribution subject to either identifying them correctly or not. And what I was hoping for a lot of things, but I wanted to see what speakers of Bullum who also knew Mindy, if they would identify it earphones. They were able to identify it earphones. And surprisingly, the old people were, but the young people were not. Not quite so surprisingly, but it was really quite dramatic. One problem was that I couldn't find many people, many speakers of Mindy who also spoke Bullum. They typically wouldn't learn Bullum. And the one person who did, I don't think really knew it very much. So young people knew them, not as well as adults. The data is still really insufficient. As I say, this is a pilot study. But I think, I can say that earphone knowledge seemed, I'm sorry, I was doing this to the microphone. Earphone knowledge did correlate with expected competence in the language on the basis of people's histories and the basis of what they said. So that's something good. There are a few things about earphones. We don't have to worry about this either. You want to see it? I already said about the goat or sheep was 100% of the time falling objects. This one surprised me. Bing was identified as something falling. I forgot what it was. I don't think it was in the water. The heavy rain, people got a lot of that. And then for the birds in flight, feel, feel, but not the fasca, fasca one. Usually labial dentals and labials have to do with moving rapidly or air movement or something like that. So eventually what I'd like to do is to use earphones from the two other languages in the multilingual area. Creole and Sherbro and C, how that will work out with a much more varied sample than I had here. And of course I'd like to do a lot more in terms of language surveys and ethnographies to find out what the multilingual situation is on the basis of that information. The real future, I think, but I'm sure some of you have heard of Mark Dingaman say he has, with regard to earphones, done a lot of videotaping without him being there. Just leave the machine on and take pictures of people talking. I think they're cracking palm nuts or making palm oil or something like that just so they were seated and they were right there and he could have the camera steady on them the whole day. And he got like three earphones over two days or something like that. It wasn't a real productive thing but completely naturalistic. That, to me, is a day set around him and because there's so much storage capacity now we can do it. I mean, Jeff and Frutica have talked about this in relation to their project. And I've sort of done this in a preliminary where we're talking about using ASR to identify languages in over an extended discourse. I was just blown away by how much stuff you have transcribed. I mean, I know you've got a big team but still it's a lot of transcription. With automatic speech recognition you can at least identify the languages very reliably, very quickly and that's something that will be useful in this sort of work. So I think that's another place that research should go. So I guess this is pretty close to the conclusion. I'm not going to go over all the mistakes I made but there are a lot. So what methodology is best? This is kind of dull stuff. A combination of different approaches. I think keeping the idea of multi-lingual or multi-components models in mind get native speakers involved and think about the social linguistic and social cultural context. So this is basically a paper about methodology and what I wanted to do is to find some instrument that would provide me with insights as to the competence of multi-linguals in the various languages in the Schengen area and I think this is going to work. I mean, it's got some kinks to iron out but I think it's going to be useful. So a method for everyone to try. Thanks very much. I'm trying to find your underlying hypothesis regarding this. And so this is a question so not having found that key in the talk I was trying to construe one myself and I thought, okay, if I'm in a multi-lingual environment like this, so one, of course, knowing that multi-linguals are not several stacked on monolinguals, what information would that give me on the type of multi-linguals? And I thought, okay, it has often been claimed that multi-linguals in Africa, for instance in Africa, nobody has a mother tongue. They have no native language. The concept is superfluous. Or people have more than one native language or they have more than one native language competence. And of course both can be true or false depending on the context. So some people may really have no native language so they speak no language with this fully-fledged repertoire that he associates with the native language because they have been so mobile that they have grown up in such a diverse context where no attention has been paid to socializing them into this kind of jeep with a multi-lingual setting. Or some people may have several monolinguals-style sort of prototypical monolingual. So is that something that motivates? Yeah, I think what I'd like to do is to find out, okay, let's go on the assumption that you've got everybody maximally competent in all three of the languages. And then is that true or not? And if it's not, why is it not true? And then look at the social factors that go along with these different competencies. And you can do it with life histories, you can do it with social factors and I think get a pretty good picture about what the multi-lingual situation is. That's what I meant. So it's a kind of triage task to identify different types of speakers and then dig deeper. Yeah, okay. As a psycho linguist, of course I'm about method and what you're actually testing here. So if you give speakers a prompt and then you ask them what do you think this means, this has not much to do with their proficiency, competency, but it has something to do with what they can derive out of the sound. And what they think it is. So it's basically, it's a meta-judgment and it's not much related to the competency. And we know also there's an asymmetry that you can understand loads before you start speaking. So there's an asymmetry between perception, between comprehension and production. So we have an asymmetry there. So first of all, if you ask them what do you think this means, this doesn't, has nothing to do with what they understand or how they can produce it. So their language competency is not tested. Well, yeah, competency is a word that's used really kind of strangely in this literature and I just picked it up. But what I'm saying is that it has not much to do with what they can produce and what they would actually do in a different situation where they might be using it. So that's one of the methodological issues about this setup of playing an isolated sentence with something and asking them for an interpretation. Because you can ask me what that means and I don't know if Linda's shareable or any of it and I can come up because of the sound symbolism. So you're tapping into a different source here and into a different kind of interpretation process that's on a meta level. So that's one of the issues. The other one is the assumption that, so I'm wondering, and I don't know the literature on idiophones in the cycle, the linguistic literature so well, but Noriko has done work on this, is about when are idiophones acquired? So is it really something that is on a very high level or is it something that comes in? So when you learn it, when you learn languages, you learn things in different speeds. You learn certain things that are easy, certain things come. So one of the question is, is there any evidence that the use of idiophones is a very high level skill set? Because from Japanese, from her work, it's intermediate and the highly proficient stop using them, actually. So they might avoid them, right? Yeah. So one of the other questions is the basic assumption that is underlined that this is a high proficiency item named the using idiophone in a production which is different from guessing what it could mean. Yeah. This is using it in a comprehension test. So what I'm saying is that the methodology, I think, is not tapping into multilingual competency in the way that the work that you're doing on the ground or that Jeff is doing or that Shridharika is doing. So if choosing a different method in terms of testing competency or testing it will give you better results than this because it's tapping into a meta-linguistic interpretation. That might be interesting and it reveals attitudes and ideologies and ideas about idiophones and something about how easy iconicity is to map certain kind of things and how difficult it is. We would need to, for example, compare the idiophone you chose with the lexical item in the strongest language to see if there's any mapping of a B sound or whatever the sound was to see if they have something to rely on that comes from their own language to be able to get a clean set of stimuli to then test that. But other measures would be helping you more to tap into actual competency in terms of recognition, so comprehension and in terms of production that would go rather into that reaction to that cognitive domain because the first slide that you showed are these two separate ones. Do they have overlaps or is it one system? The cycling with the literature is all about this idea there's language and language in the brain or there's language and language overlaps a little bit or now we think it's one system that goes into different ways. I think that's the really interesting thing to test, for example, for idiophones. We need to establish the acquisition history first to have an understanding what we're actually testing from a logical point of view. I understand your objections. First of all, I remember that there were different levels of iconicity. The ones I played with were the higher ends of non-arbitrary associations so you just have to get a feel for it. Another sort of issue that I mentioned is whether in a multilingual area will idiophones be transferred or be shared across linguistic boundaries or is there just one sort of confidence with regard to idiophones and this is way too little data to answer that question. I think that's an interesting question to ask what you think about. What was the... Oh, yeah. Of course, this is one possible way of getting at idiophone knowledge without having people produce them or having to do something like Mark has done with leaving a recorder on all day and doing with a bunch of different people. I was able to send people out and it seemed to work pretty well and people were not reluctant to be recorded. In other situations people have been reluctant to be recorded. So everything is compromise. You want a combined method. Yeah, that's what I was going for. You would want to have a similar set with things where you know there would be producing to have naturalistic produced idiophones with a similar set of whatever work they've done like very good and pretty good on positionals and cost placement events for example where you can elicit them and get them. That gives you an idea about what they use. You want to back that up with naturalistic data and then you can take that to use in a testing situation and that you correlate with the language proficiency that you have assessed whether someone speaks a little bit of this or a lot of this because again it's again this issue of competency someone might be competent in talking on a market, the other one might be competent in telling useful stories but they're using a lot of them. So it's again we need to have a dynamic notion of this competency thing and the ones that are used to telling lots of stories for example I assume they know those better than the ones that are doing only trade because that's the vocabulary they have for example. I was interacting with Peter and he was doing a Sylvester and Tweety word so I did something like that with a bunch of Zulu speakers and not Sylvester and Tweety No, just don't use Sylvester and Tweety it's what I'm saying it's something that is normal in the given situation we're using Sylvester and Tweety with the Japanese people in English not everybody really knows but it's cartoon and it's an action packed cartoons with bowling balls going down gutters and stuff like that and a lot of stuff that you expected to be characterized by idiophones and so I devised a lab experiment that I won't go into detail but what I was looking at was the synchrony of idiophones and that part turned out pretty well because the experiment was so poorly designed and it was so labor intensive I didn't carry it much further I mean maybe today you could do something like that maybe not Sylvester and Tweety but with a lot of the stimulus material that was developed for example for positionals and whatnot was a bottle standing like this and like that and like that knowing that bottles are recognized trees are recognized you know, adapting it to the cultural thing not to think about Sylvester and Tweety not at all, that's just insane yeah sorry, I didn't mind actually, my class first for Japanese nominated and made this did exactly that Sylvester and Tweety and there are some schemes that through elicit a lot of Japanese mimetics but what I found so I was an official of their testa so these participants are also tested for their proficiency level it was very interesting to find that the use of mimetics did not correlate with the level of proficiency that's so it correlated a little bit in that beginning level speakers can't possibly use mimetics it's very difficult and they don't have a knowledge I think that might a high level of proficiency and knowledge of mimetics might correlate but production wise it didn't correlate very well and in fact stereo level speakers coming from Korea who also has lots of mimetics do not use any mimetics whatsoever for rolling events for example Sylvester rolling down where Kita has discussed quite a bit because they are they tend to use a lot of Koroko but internet level speakers started to use mimetics and adversarial speakers used quite a bit of them but I thought maybe it has something to do with also the social or the register factors of mimetics mimetics is associated with colloquial language it has to do something with faithfulness as well obviously video recorded conversation with internet friends and so forth I think it's a very daily life setting but as soon as speakers are given in this setting where they have to describe that narrates the events for research purposes some of the non-native speakers actually native speakers are quite some of them are quite unmeted and use a lot of mimetics but non-native speakers still have this sense of being tested for their proficiency even though they were talking to age peers so early on I found that if I asked native speakers or Japanese to narrate these Sylvester and Tweetie to me they don't use mimetics but once they start talking to Asia another kind student they started using a lot of them the animated description of the scenes but these very high level speakers square level speakers coming from Korea they want to be very good so mimetics are kind of I'm Indian once maybe an interesting vocabulary in that you cannot pinpoint the exact words that describe the sensation or describe the motion and even native speakers would have different choices and the book is a book but then what kind of walking it is might be interpreted differently of native speakers and non-native speakers would find difficult to link the mimetics that they know they have a knowledge of to they have come up with different possibilities and so native speakers who may not care so much about being perfect might be willing to describe them in a way that they would interpret I think your basic point is right on that idifons, the use of idifons is tied to many social factors to many cultural factors and also to some pure linguistic factors teasing that apart is something that a lot of people are worried as a matter of fact there's a conference in Japan and Tokyo in Christmas on idifons and mimetics and a few Africans are going to be there I'll be there too see you there but to go back to something I didn't answer and that was when idifons are acquired and this is a I know of only one study and it was in South Africa and she was looking at whether kids controlled idifons and they didn't until fairly late in the game so that goes along with what you were saying as well and I think that's right some of the more complex social cultural linguistic behaviors so this is a very interesting set of vocabulary for sure so I'm really interested in your work yeah I mean yes I think it's really a fascinating topic that really requires more research so maybe not to use it as a tool to tell competencies but I'd rather look at idifons you know look at it look at that as part of idifons how it's shaped and what it's shaped going back to this question of where it comes in the acquisition and whether it is actually something that is correlated with high proficiency I think it's a very interesting question particularly in this area because I know the study on the Umbara test idifons have been studied but to my knowledge both Manda languages and Atlantic languages are idifons as a systematic part of their vocabulary so they all have it to modify color words at least the three basic color terms that they all share they all have it for manner of movement and they all have it for instance to describe different types of forming and these are the ones that are most interesting from that perspective because they are not automatically or at least not directly iconic so I think the more iconic ones are really not interesting because you can also not really know to what extent they are actually part of language you know or just you know this could just sound so if they are really robust feature of languages in this area in a particular set of domains so that might mean that actually they might be acquired in a different way than in South Africa for instance so I think it's an empirical question to look at them and see what role they play in the repertoire when they are acquired in which genres and register also don't really believe that this is a real urban contrast I was trying to remember to make that claim that idiophons get lost in urban context was it you? I was never buying into that claim I'm sorry to tell you it's a more nuanced claim but I think it's a party I think it holds to the extent that a particular intirectional context goes away when people move to time so a formative context so storytelling etc they are stigmatized so what you have is even DH idiophonization in urban context so in Zulu there is an idiophon that means something like soft and light that's no longer an idiophon in urban variety but do you think that it's related to the standardization process that Zulu undergoes well it's a change in Zulu it's a change in Zulu I mean you have a whole range of Zulu's in my city so it's a tiny Zulu so nice Zulu I don't know maybe that's more of a standard one so look at the intirectional context and then collect these genres and then compare speakers that would be wonderful it's a complex topic and it's not easy to research for a lot of reasons that are undone raised nonetheless I think it's still you know equivalent we have in our era for slaying of pieces three questions I think it's the topic you brought it more towards the end but do you know much of this in this area about idiophones and context I mean do you have any sense about whether they're barred I mean so they exist how many are the sounds I mean are any of these and just to make it a little more interesting partly why I asked is there's one language I mean it's interesting anyway I have a specific data issue like one language where we are in lower Fungville we know some of the languages have idiophones I don't know if we know about all of them or not but there's one that's probably a more recent entrance to the area and it's the one I've worked on most and I've tried hard I couldn't never elicit one but you know I wouldn't say it's an idiophone like a really rich I mean it's not an area where you're stepping on them all the time but some languages have a decent sense and at least one I know I mean one time the thunder was rumbling I'm like what's that sound they're like thunder you know but they're like come on you have to have a way I can come up with an idiophone in English for thunder you know but they wouldn't do it for me right and so I'm wondering just to what extent do we have some sort of area like how arial idiophones are and what kind of content is this indicative of a newer entrance to an area if they left them which we believe for other reasons is the case do you have any sense from this? Yeah I do I should think about that the short answer is they're usually not transferred across linguistic boundaries the function can be and in that people wrote out and but in certain cases you can and I'm thinking of one particular one there are people in Southern Africa called the Tonga who are controlled by their wives and these are gross generalizations and when they come into Hansburg to work they hang out with the zoos because the zoos are considered tough and macho and they control the household and what they do is to sound more zoo-like when they go back to the Tonga villages is use zoo idiophones so that's one case of them being crossing a linguistic boundary and identical and that's I think I'll find out in the largest studies whether I think there's a continuum of a sort of areaality in the same way there is the iconicity and I haven't found well I have found some area ones so I'm sure other people have worked in West Africa when something goes on for a long time you say she walked down and there's another one so like that high tone I can I can I think that's what I'm hoping to turn out here there is something like that but the general answer is no what about the idiophones in Creole because I'm sure they didn't come in that way in the study they must have invented them or picked them up from the study how we invented them there are a lot of ways idiophones can be derived from what is it I don't know I don't know I don't know so that means it means something I think cow is really white and the word cow is an idiophone but it's also a word for snow but all nouns have a non-consistence of it has morphology but when it appears as an idiophone it's stripped of all the morphology and it gets an extra high tone as well so there are internal ways of creating idiophones in the language I don't know enough about the idiophones of Creole to tell you that so do you find them in Creole like they they've got a full set of idiophones I feel red is black it's like this is no common Portuguese and they have like dozens of these things so somebody must have the need to come up with idiophones so maybe they brought them in this paper I wrote about idiophones in Pigeons and Creoles it seemed to me that the reason they existed in the Pigeons in the rest of Pigeons was that these were local languages and they were idiophones in my research and Bill Sandlin's research and someone else's we found that the more sort of local a person was the more likely the person was to use idiophones and when the Pigeon marked someone's identity or membership in some sort of individual unit that became a symbol of that social so so maybe actually if you phrase your research not use them as a measurement of competency but a measurement of competency as well it's wrong with kind of claiming a language as a kind of identity language as a kind of language that's maybe but I think we have loads of topics for discussion over the years and so we all free talker and move over to the IOE as usual for drinks and that's quite good again