 Good afternoon everyone. My great pleasure, on behalf of the Department of Statistics to chair this session which is a talk given by Dr. Carolina Gerich. Gerich. Yes. So, Carolina is probably familiar to many of us because she has recently completed her PhD study from this very Department of Linguistics at Sowers. And she's now doing a post-doctoral research in acting as a post-doctoral research associate still at this department. Exactly. So, her talk today is, which is a long title, who has the right to know the interaction and how to find that out. Lessons from... Right. Right. Thank you very much. Hello everyone. Thank you very much for being here. I know this sounds like quite a mouthful and it might be a little bit intimidating but I'll do my best to sort of unpack this talk as we go. You know, it's actually very exciting to be here because I remember being here as a student coming to these seminars and a lot of things going over my head. So, I was trying to do this talk in such a way that you would kind of all be able to take something away from it. If you have any questions then there should be enough time later after I've talked to kind of figured it out. And maybe without further ado, there's some weird things going with the display. So, I'll try to sort of adjust that. Exactly. So, then you can't see parts of it anyways. So, what I will be talking about today has to do with Amazonian Kichwa, the language I've been doing my PhD on but not only that. First, I'll give you a little bit of an introduction into both the language and the project that I've done for my PhD because essentially this talk is sort of a very brief recap of what I've been doing for the last, say, six years and trying to see what else can be done in that field, potentially what you could do for your own research because there is a lot to be done. So, we'll start with evidential systems in Kichwa languages. From that, I'll move to evidentiality within the epistemic domain, sort of broadening the research questions to see how evidentiality fits with our phenomena. Then we will talk about epistemic authority. I'll explain what it is in a second as well so it should be fairly clear when we actually come to that. And then we'll end with some conclusions as usual. So, starting with Amazonian Kichwa. It might not look like that from the orthography that I've been using but actually Amazonian Kichwa is a member of a Ketchwa language family that you are probably, most of you might be familiar with. It's the largest indigenous language family in South America and this is a map showing you kind of where it spans. So, starting in Northern Chile and Argentina, going all the way up to South and Columbia. The language family has about 8 million speakers as far as we know because it's hard to get accurate data. And you probably know and you can also see on this map that it's mostly spoken along the Andes. So, the language I've been working on is in Ecuador. That's also why it's called Kichwa, not Quechua because they decided to call it Kichwa on the basis of just the three-way vowel distinction rather than the five-way vowel distinction introduced by contact with Spanish. And this form with a K and a C-H and a W, it's actually a recent orthography development so I've been using that because now it's a preferred orthography. And this language is quite extraordinary in terms of the language family because it's spoken in the Amazon. So, mostly Ketchwa culture and languages are associated with the Andes and this is a very different cultural milieu and also very different linguistic influences and cultural influences which are hard to unpack because we know quite little about the history and the linguistic history of the Amazon. Right, so moving swiftly on from this big map, a little bit more about Kichwa in Ecuador. It's one of the 14 indigenous languages spoken in the country. It's a relatively small country by South American standards especially. There are several varieties spoken all over the country, about seven officially and three of them are spoken in the Amazon. So, I'm saying Amazonian Kichwa even though you might have seen some references to my previous research when I was saying Tena Kichwa. And this is because I was referring to Tena, that's the capital of the province where I've been doing my fieldwork, but there is multiple problems with that because there is also a river by the same name. Which is actually spoken 50 kilometers away from the capital and there are other dialects that are small and spoken near. So, me and some of my colleagues working in the same area have sort of decided that we are just going to say that we work on the Amazonian Kichwa varieties. I'll show you a map which shows that in more detail in a second. So, just to keep that in mind that actually Amazonian Kichwa as I referred to it here is a dialectal grouping. And I've just been doing work on one of the dialects within that grouping, which for the purpose of this talk I referred to as the Upper Naples or Tena Kichwa. So, that's not an ideal name, but we have to go with something. And according to the data that we have, this particular language variety has anything between 20 and 40,000 speakers. But it's really hard to get accurate data again because the census that's been done, the most recent one in Ecuador in 2010, it's not actually specified, it's not asked the people who filled it in to specify what language they speak. So, the only way that you can determine which dialect of Kichwa they speak is on the basis of the residents, which obviously is not ideal. So, there are issues with the data to which we have access. So, Kichwa as a whole, so these seven dialects are regarded as one language within the constitution. And it has been given the status of an official language of intercultural relations. So, Spanish is the official language and then two of those 14 indigenous languages, Kichwa and Xhuar, are the official languages of intercultural relations, which means that they have sort of semi-official status in the territories where people who speak them live, which is not all over the country. There is pressure on the dialects of Kichwa, both from Spanish, the dominant language of the country and the continent in fact, and the unified Kichwa, which is an official version of the language that has been introduced in the 1980s. So, there has been a sort of unification reform, orthography has been unified, and that has been implemented as the official variety of this minority language. So, as it were, it's having, according to what I've seen, a negative effect on the vitality of the language on the local level. And there is language shift in progress. So, in my field sites, the oldest people to communicate with everyone in Kichwa would be in their sort of late 20s. And the kids, the teenagers have passive knowledge of it. They know it quite well, but they wouldn't use it amongst themselves. And the children mostly would use Spanish, talking to one another. It might be a little bit different in villages which are further away from the main roads, because mine was sort of relatively well communicated with the outside world, but it's still not looking great. Right, and this is the map I mentioned a while ago. Those are the different Kichwa varieties spoken in Ecuador. So, you've got sort of five of them here within those yellow lines. This is roughly the Andes. And then you've got two shades of green. There is another one which I'm not sure that you can see sort of to the right up there, very bright green. So, those are the Ecuadorian, sorry, the jungle varieties. And my field work was around this place. So, around the town of Tena, and we will zoom into there now just so that you can see how this looks. Right, this place is not great. I can see a little bit more than you do on the right. Right, this might be better. It's not really going to let me do this. So, the village where I've been working mostly is roughly there on the river. And you can see that this is the Andes, right, and we are descending into the Amazon. So, from there on, it's basically jungle all the way down to the Atlantic coast. So, we're only just entering the Amazon, and there is... This map doesn't really show that great, but along to the rivers you can see that the vegetation is not as dense, because there is a lot of cultivation going on, and the landscape is being dynamically transformed both by herding of cattle and by all exploration and all kinds of pressure for land. It has to be said that the population of Ecuador has increased fourfold in the last 60 years. So, it has gone from 4 million to 16 million, and that's generating a whole lot of issues which also impact language vitality but that's a matter for a whole different talk. Right, so in terms of how this language looks, because I've already told you that it's a little bit difficult, it's quite different from Andean varieties in terms of culture of the speakers, but actually linguistically it's also quite distinct. So, as any other Quechuan language, Amazonian Quechua is suffixing, so no prefixes, maybe apart from one that I found over the course of my work, agglutinative, the dominant word order is SOV, however in the data that I've been looking at I found that actually now the SVO, which is sort of Spanish-influenced word order, is equally permissible to most speakers' ears at least, and there are also discourse-related factors that influence the word order, so there is that to be thought about, even though if you look at typological literature probably you'll find the information that Quechuan languages generally are SOV, so that's still dominant but nevertheless so. Then there are evidential and clitics, so evidential particles which attach to the end of the words, actually to the end of phrases, they can occur on any phrasal category, hence and clitics, so you might have heard of evidentiality as part of the ten-suspect mood morphology, that's not quite how it works in all these languages, they are basically independent markers which are not integrated with the ten-suspect mood system, and we'll be talking more about this later, they're basically singled out because that was the topic of my dissertation, and the cognates are there, but when I started looking at them quite a few important differences with this sort of established evidential paradigm also came up, so I'll talk you through this in a bit. In terms of what's really different in Amazonian Quechua from Peruvian Quechua or Andean Quechua in general, actually in Ecuadorian Highlands there are also no adjectives, so only pulmonic consonants, we've got both voiced and unvoiced stops, and that's particular to the Amazon, whereas in the Andean varieties you only have the voiced, sorry, the unvoiced stops, and that's also the unified Quechua orthography, it presents it in this way, so people who have been exposed to this orthography tend to lose this distinction but it's actually there. There is no exclusive, inclusive first person distinction, so no pronouns for me and you, but not you, and all of us together, that's not existent in these varieties. There is only residual object marking, so in Peruvian Quechua, Andean Quechua, Quechua that's morphologically more complex, you would get obligatory marking on the verb for both subject and object. That's not the case in the Amazon, there is only, so subject marking is still obligatory, but there is only one object suffix that remains, and that's for the first person plural objects, and there are quite marked lexical differences, and that's probably, we did little about this, but partially probably because of the fact that it's a very different environment that we live in, so different animals, plants, conditions, and also different influences from a totally separate group of contact languages. That the Andean Quechua has not been in touch with. So that's where we stand on the language, and my PhD project basically consists, that's the outline of it. It was an ELDP funded project, so I was fortunate enough to be funded for three years to do it. Because it was an ELDP funded project, documentation of the language was a requirement. So that meant I actually had to go and do fieldwork and create a corpus of the language apart from working on the issues that interested me from the theoretical point of view. I've done about ten months of fieldwork, two longer trips, six months and four months. Because when I went to the field, the sort of anthropological documentation project was finishing there, I stepped into the anthropologist's shoes and I was fortunate enough to inherit a collaborative project. So I had people who could already operate alone quite well. They were computer literate, and I've been working with a team of five, sometimes, and seven, we grew up to seven people at certain points, doing transcription, translation, segmentation for me as well, and actually we did some training as well so they went out and recorded stuff on audio and video. That's actually quite important. We managed to create a 13 hour corpus that's all audio and video, and this doesn't include elicitation. So I've been doing my elicitation sort of on the side, but the 13 hours it's all videotaped, all taped, right, and it's 11 hours of naturalistic discourse plus two hours of sort of elicited discourse where I would present them with stimuli to have a relatively controlled discourse setting, to know what people were talking about and to kind of keep it simple. But it's all monolingual, so it's very, very nice data to work on because people were there on the ground from day one to help me do this. And for some documentation purposes, maybe because some people didn't want to see me. They would be intimidated if I was around, so basically the quichua researchers went by themselves and recorded these people, went to events that weren't accessible for me. And all these 13 hours have been transcribed in a line, translated into Spanish, and basically revised two times over. So I looked over them and then I came back to the transcriber, we looked at it again, looked at our differences, and then kind of reconciled them. The shorter part of the corpus, which is the two hours, it's parsed and glossed on the morphological level. So that was the basis for my PhD research. And the initial focus was on the analysis of evidential markers. And I'll explain now why it was the initial focus because when I went to the field expecting to find something, I found something quite distinct, in fact. So just by means of a reminder, what is evidentiality, right? Evidentiality indicates the evidence the speaker has to support their statements, that's Palmer's take on it, and that's actually the revised Palmer's book on modern modality. In the previous version, the year 1986, he was looking at evidentiality as part of epistemic modality. And in the new version, epistemic modality and evidentiality are sort of both within propositional modality, but we'll come to that in the second. Linguistic marking of the source of information, that's the most well-established definition from a big monograph on evidentiality by Eichenwald. And the more recent one is the coding of mod of axis. We don't really need to dwell on this for very long, but what's interesting about that is that people who have come up with this last definition claim that the source of information is sort of always the same. So sun is shining, that's the source of information. However, what matters for the marking of evidentiality is how I access this information. So it's a more precise definition to their years and actually when you go on the surface, when you've just sort of heard about it, it might not make much of a difference, but when you start looking at what individual languages do and how speakers use it, it actually makes a lot of sense. So this is all quite abstract. So just to sort of show you what the literature has been saying about what kind of things can be marked by evidential markers. We've got the sort of two main types on the left-hand side, direct evidentiality and indirect evidentiality. Direct is also attested, so that means that we've experienced something through the senses. More often than not, that would, for language marking systems, evidential marking systems that would incorporate all kinds of sensory evidence, but there are some languages which would also distinguish all the things that you can see sort of on the right-hand side. And indirect evidence, that's anything that's nonsensory, and here we've got sort of two main groupings. One is reported evidence, so hearsay, which can be second-hand or third-hand, and then there is folklore, which is different because it's hearsay that's sort of replicated by the community. Again, not all languages make that distinction, but a lot of them do. And then we've got inferred evidentiality, which, and the inference, can be based either on results of an action or on reasoning. So if we, I don't know, come over to John's house and he's not there, but we know at the time that we came to visit, he normally plays football. We say John must be playing football, and that's based on our general experience, inference from reasoning. However, if we come to our house and we see that there is a vase that's broken and our dog is sort of nowhere to be seen, we see the results of the dog breaking the vase and we see the dog must have broken the vase. So that's inference from results in very simplified terms. And now in catch-one languages, what we traditionally have is the freeway distinction between direct, reported, and inferred. So there is one marker for all kinds of direct evidentiality, and then we've got the verbal reports and then inference, with other markers. This is how it looks in practice. So a very simple example from Martina Fowler's thesis on Cusco Quechua, so that's actually a Peruvian variety. We've got three markers at the end of an inflected verb, right? Parashan mi. It's raining, I can see that. Parashan cha. Maybe I'm in the house, but I can hear the rain banging on the roof. So that would already come as inference because I don't actually see the rain. And reportative, someone told me it's raining. However, so that's basically when I went to the field to Ecuador, I was expecting to find a neat freeway paradigm like this. It didn't happen. So I found out that the reportative marker was not there at all to start with. And instead, people were actually using construction which would incorporate a verb of speech and mi, so the direct evidential and the verb of saying, which is not uncommon cross-linguistically, but it's not how it was supposed to be working according to the data from other varieties that I've seen prior to going to the field. And then I started digging deeper, and I will just show you what I found on the example of mi, so the direct evidence particle, because that's the one that I sort of had the most time to analyze, and I think that's the one that I managed to also analyze in most detail. Can you see that okay? Is it sort of right? So we've got what I found out is that actually in Tena, Kichwa, Amazonian Kichwa, we cannot find a one-to-one correspondence with the types of evidence that I showed you a while ago and how this marker would occur. So we've got mi with direct evidence. It's raining, I can see that, that's fine. But we also have it with inference. So there's a similar situation to what I just described with John going to play football. Cesar Mingama Mirishka. I'm pretty sure Cesar went to go and work in the field with other people because I know him, he's not at home, so I infer, and I can still mark it with this direct evidential marker. I am in the house, I hear steps outside, I was expecting my father to come. So I say, because I inferred it, but again, it's okay to use me in this context. And then the sort of most striking one is the guesswork one, and I need to give you a little bit more context for this. It says, means the seed goes to the left, to the left. And that was an experiment that I've done where I had people watch the free-shell game with the cups and the seed, and I cut the clip so that first they just saw the magician performing the trick, but they didn't see where the seed has ended. And then the second time they watched the whole thing to be able to see whether they actually guessed correctly. So these guys that said that they've been watching this thing for like, I think, five or six cases of this game, and they've never guessed correctly. And this was the sixth guess or so, and still they were using the me. So it's either guesswork or inference, but definitely it's not the direct evidence. And one thing I should also comment on you can see on top that it says direct visual BPG. So BPG stands for Best Possible Ground, and that's taken from the work of Martina Fowler on Kichwa, sorry, on Cusco Kichwa, where she said that basically we need to expand the category of direct evidence because some facts in the word are just not accessible to us. So for things that pertain to our sort of live experience, the best evidence is indeed the direct evidence. However, if, for example, we've never left Europe and someone who has been to Africa tells us there are elephants in Africa, if we know that this person is trustworthy, we can just accept this sort of a face value and incorporate it into what we know about the world because it's something that's not accessible to us. We can treat this here, say then, as best possible evidence. So there is a difference between what Fowler calls personal information and encyclopedic information. And she was saying that in Cusco Kichwa me could basically mark both of them. But that's, you know, as you can see, this is a sort of broader array of cases than just direct and best possible evidence. Right, so having found this, I wasn't really sure what to do with that because I was familiar with evidentiality and epistemic modality, as I told you, I read up on things. And those are the things that I was expecting that might be, maybe, marked there somewhere. So propositional modality. But then this marker wasn't doing that, that didn't really fit. So it's not exactly the way I'm telling this, it's not the order in which it happened. I think you probably can't see that, which is sort of semi-on purpose, but I'll talk you through that. And so I came up this sort of newer research, which was saying that actually, epistemic modality and evidentiality are two aspects of epistemicity. It's the white thing over there, probably, you can't see that very well. We could also switch off the slide if I knew how to do that without switching off all the others. So if someone does know that, then please go ahead. And so, let's see. Anyway, so, oh, that's better, isn't it? Not really. Okay, so that's the only thing that's in white, so it doesn't really matter, I can just talk you through that. I don't really mind this up to you. I know what it says. Right, so basically this newer research, instead of thinking in terms of modality and how evidentiality is a part of modality and where it belongs, Casper Boy said, okay, this is something that's called epistemicity. And it is so because basically these two things, epistemic modality which has to do with sort of support for factual status of propositions, so in simpler words, whether I judge something as possible or not, and whether I know that something happened because I've seen it, or because I've heard it, or because I've inferred it, are kind of two kinds of epistemic phenomena, so epistemic as related to knowledge. Evidentiality has to do with epistemic justification, as he calls it, because it's how we justify what we know. And epistemic modality has to do with epistemic support. So how much do we support what we know, right? Are we willing to vouch for it because we think it's certain, or are we maybe not so willing to vouch for it because we're not so sure whether it's certain or we might think it didn't happen at all. So when I kind of came across this way of thinking about it, I started to think, okay, that makes sense, but that's probably not the only two dimensions of knowing that are out there, and what else can we sort of relate that to? And that's when I sort of opened the Pandora's Box in terms of my PhD. So basically the current look at this, and epistemicity is sort of a term that I've been using, it has had different definitions in the literature. So Boyd has used it to refer to just evidentiality and epistemic modality, but Bergfist, for example, in his 2017 recent lingua article, has been referring to it as a broader phenomenon, and that's sort of the outlook that I've adopted. So epistemicity here is all the sorts of phenomena in discourse and grammar that are somehow related to how we know things in interaction. And I'll talk you through these different facets of it that we know about so far, but just to, because I said, you know, different dimensions of knowing. What does that mean exactly? So within conversation analysis, there has been quite a lot of work on this already, but we might not be very familiar with that as language documenters because that's not the things that we really do, right? We concentrate on different level phenomena. So I came across this, which is effectively different dimensions of knowledge in interaction, which was done by conversation analysts who work on sort of more familiar languages, if you will. So you can see we've got epistemic access, primacy and responsibility. And if you look at what it says, knowing versus not knowing types of evidence, degrees of certainty, this is already covering evidentiality and epistemic modality, right? Certainty, possibility, types of evidence. And this is but one of these dimensions, that they have found to be manifest in conversations and languages such as English, French, even Japanese. Then we've got epistemic primacy, which is related to the right towards knowledge. So how much can I know, possibly? How much can I know of respect to you? Who can claim this knowledge in conversation, in interaction? And then the third one is about obligations or rights to have certain information. So that tends to arise somehow by virtue of context, but you could imagine what this would mean that, for example, if you are in a situation where I don't know if you approach a professor, the professor has the obligation towards knowing certain knowledge, right? So it arises by virtue of who you are socially. And that's all socially constructed. So it was really when I came across this that I started thinking, okay, maybe this has something to do with what I've been seeing with this me in Kichwa. But it turned out that there is, so right, so we've got epistemicity, I'm getting ahead of myself. And then I started looking at a sort of broader array of phenomena that have already been described for different languages that have to do with the dimensions of knowledge and interaction. And the one that's sort of relatively well-established, apart from evidentiality and epistemic modality, is mirativity. Those of you who do, Tabeto Berman might have already heard about this because that's where it sort of originated. And this has to do with marking of unexpected information. And some people also refer to it as marking of the fact that the mind of the speaker of the address you also sometimes is unprepared for receiving certain information. So I realized that something, right? It's a periphrastic strategy in English, but you do mark that as well discursively that you haven't known something before. And I'll just show you an example from Chechen, actually, that's taken from work by Martina Bruy, who was citing someone else. Ah, let's see, whatever, I can go a little bit. Right, that's better. So you can see how this works. We've got Zara has come. I knew she was gonna come, but then you mark it differently if you weren't expecting that. So something maybe along the lines of like, wow, Zara has come or oh, she's come when you haven't been expecting that. That's actually marked by a dedicated verbal suffix in Chechen. So that's one epistemic system that we already know of, and that's been that originated with the works of Scott Delancey, so beginning of the 1990s. So it's been around for quite a bit. All right. Then we've got Egophoricity, which really is a fancy name for conjunct district marking systems, which have been first analyzed also in the 80s and were initially regarded as related to person marking. And this monograph that's about to come out now, about Egophoricity, or maybe it's already come out, says it's related to primary knowerhood. So that's sort of about, again, rights and obligations to know. It's very fuzzy, but actually that's kind of where we're at. We're trying to figure out in this field what these markers are doing. A simple example of that is from Tabeto Berman, from Navarre. And it kind of shows you why it was thought to be related to person marking, because we've got, I went with A, which is egophoric, and A law stands for allophoric. So the A marker basically implicates personal involvement. And then we've got, you went and she went, because I'm not involved in these actions. So it's not that it's so much a person distinction, but it's a distinction between whether this action is somehow involving me or is it everybody else. And it can also be exploited in discourse. For example, the non-egophoric marking, so the allophoric marking, can be used with first person actions when you do something unintentionally to sort of show you that, show the interlocutor that you weren't involved with that. So related somehow to person at first glance when you look at it, but actually it's not. So that's egophoricity for you. Right, and this is where it's starting to get really complicated, because this is all recent work. And I personally am still to figure out how this relates to the kind of overall domain of epistemicity that I told you about, because I think we are getting dangerously near rejuplicating it. So complexity epistemic perspective is a concept coming from, again, the work of Henrik Bergfist. He says what it has to do with is asymmetries of access to information. And then engagement, which is another notion in the development of which Henrik Bergfist, the same author has also been involved. Essentially, I think it gives more fancy label to the same phenomenon. That's why I included it here. I don't really want to show you examples, because again we could spend the entire talk talking just about that. So relative mental directedness of speaker and address towards an entity or state of affairs. It takes a lot to unpack. What it means essentially is what you see above, which is how engaged we are in what we're talking about, and does their take on it, they subsume a lot of different phenomena which, to my mind, are more precise about how you can know things, how you relate to people, but essentially what's important about this is that this is intersubjective, right? So this has to do not only with how I relate to the proposition, but also with how I relate to my interlocutor. So essentially there is that. This is different from the other things we've seen. The mirativity is about whether I am surprised. Egoforeicity is about whether I am involved. And this has to do with a third dimension of there is more than just one person talking and we have to deal with that. And as you see, both these references are basically one is not even out yet and the other one has just sort of come out. So this is all new and that's why I'm showing you this space. Right. And then there is epistemic authority and that's kind of what the rest of the talk is going to be about because when I started looking at how this was defined in the literature and what this Enclitic me in Tenakichua was doing, I figured that that's probably the closest to its meaning that the epistemic domain has sort of come. So epistemic authority has to do with the relative rights to know about some state of affairs as well as relative rights to tell, inform, assert or assess something. In simpler terms, again work from conversation analysis, primary right to evaluate matters assessed. And one thing that I forgot to put on this slide which I found very enlightening was work by a Japanese linguist Akio Kamio from 1997 about territory of information. So his work basically says that there are different kinds of things in the world in our sort of awareness about which we can know to a different, to a certain extent. So I, you know, I basically speak here today from an expert position on epistemic marking in Tenakichua because you know less about than I do. However if I was talking to a native Kechua speaker, we wouldn't be in the same situation so it's a dynamic thing. And the territory of information of any given person comprises of information that's sort of close to them. So it can be information related to their personal experience, profession, family members, etc, etc. And essentially what Kamio was saying is that in conversation you negotiate some sort of a confluence between different territories of information. So again it would be weird for me to for example, I don't know, comment to Peter on Australia because I know nothing about Australia and he was born there. So I cannot act as an expert because Australia is within his territory of information not mine. And interesting things start to happen when you have people with competing claims to information. Right, and so epistemic authority is not that new in terms of linguistic inquiry but it's only recently that we started looking at it within language documentation. There is conversation analysis work on it in English where how it's marked linguistically is through terrain design. So through structure of conversation. And this isn't really very clear I think this thing below but I extracted it from the PDF because I wanted you to see all the transcription conventions for conversation analysis. This is Veronica and I think Jenny talking about Veronica's grandchildren. Right now the heritage say the first assessor, the person who knows most is likely to speak first and that's Veronica speaking not very fondly of her grandchildren. So I think it's a nice little devil I don't think it's nasty you see with Jillian she can be a nasty little bitch. So that's a grandmother talking about her grandchild okay alright and her friend is well you're saying something is that it's a shame isn't it? So she is trying it's kind of a shocking claim right you just heard someone say that about her granddaughter but she's not competing she's sort of aligning with that position trying to position herself with respect to that and that's what English doing in this regard. This is sort of a default thing where you've got first assessment which is an assertion but you could also frame it as if it was a question and then it has to do with how you negate the thought of epistemic right. I know this is a lot but I just wanted to show you this to kind of so that you can see that there has been work done on this in English. Now a slightly easier example I think to gloss the first line should all be in italics this is from Jammin Junk in Gallibur and here I'm not talking as an expert because obviously Candide is an expert on that language so I'm quoting a work from Eva Schütze-Bernet who says that there are epistemic authority and critics there which derive transparently from person marking, especially this one that's first person inclusive marking me and you. So in the first one then Negardi that's sort of assertion of individual epistemic authority I was there you weren't therefore you haven't seen my granddaughter doing things with the grass trailer and the second one is this boy has fallen over he looked like as if he was dead but we were both there and I'm inviting you of using the Mindy to comment on this so we're constructing a sort of discursive reality where we both know the same amount and that's also relatively recent work so we've got that already we've seen that in English this is a language of Australia and there is also a work in Japanese that is done on that so in Japanese you've got two discourse particles that have to do with what Hayano calls epistemic primacy which is essentially very similar to epistemic authority but primacy implies that there is an imbalance so one person would know more than the other and here we've got yo which is information not shared with the recipient so it's just the speaker's knowledge and we've got ne which is shared information and again I don't know why this literature is always about grand children actually another example T is talking to S and S is the grandparent so the first person says well how that boy walked ne shared information I'm sort of being polite and I'm letting you in on this information because it's your grandchild so I don't want to be mean and kind of say that he is really cool and really you have the right to assess him and you can see this picked up in the second turn when the person it's actually a combination of devices right but she says he does yo we must make him walk so first you can see that there is a combination of discourse marking but also with tense right the first turn comments only on one particular instance of an action which we've witnessed and the second one is a general claim about the boy and that's where the epistemic primacy also comes in I know more about his walking than you do and then there is immediately referral to an action which is sort of pertinent to the grandparent and not to the neighbor who just happened to see the child we must make him walk so this is relatively well described for Japanese and again it has to do with epistemic authority right and basically this with these empty arrows it just shows you that there is probably other phenomena that we're still not yet aware of that might be marked in languages that haven't been described but you know that's this kind of what we have for now and having discussed all this I will just move to Amazonian kitchen again to show you in more detail why I think this me marker is a marker of epistemic authority and I contrast it with one other in clinic so if you look at this now having heard everything that we've heard about epistemic authority as a domain again let me just put this out a little bit not sure whether it's made it any better bigger? okay right so the first one is straightforward right it's I see that it's raining direct evidence doesn't it can be kind of subsumed within epistemic authority so you are likely to have authority over something if you've seen it directly so it's not kind of an exclusive thing so that's a straightforward example we can analyze it as an epistemic authority case because someone has seen it that's fine inference conjecture when Cesar went to the MINGA we assert that with me if we know Cesar and we know about his routines then effectively we could also claim authority over this knowledge right more context would be needed there but that's you know that sort of our knowledge that's already established so easy to imagine that we can claim authority here and same thing with your favor coming you might know just as you know people's voices you might know the sound of their steps or how they move or something like that so it could be explained that I know when my favor was supposed to come it's kind of in line with my general experience of what he does I can assert with authority that he's coming I don't have to have direct evidence and the last one it might just be a purely persuasive thing when I think I've really guessed it I know where it is now that kind of thing so I'm trying to convince you that my guess is correct because the two of us are trying to figure it out and therefore I'm asserting authority over this guess for which I don't have basis in direct evidence alright what's it doing so that was sort of anecdotal explanation with epistemic authority hmm but these examples actually show you that it does have to do with authority of knowledge and actually that it also has to do with exclusivity of knowledge so me as well as Aver and Clitix in Tenakichua also have sort of information structural associations so me is associated with focus it occurs on focal constituents but it's not always there it only occurs in about 6% of all turns in my corpus so it's not enough to say that it's a focus marker and generally you can use it or you can omit it however there is one context where really people wouldn't agree to the sentence you provide to them if it didn't have a me and that's in that instance is corrective focus constructions so we've got example of two of those here Manda Nukaushichu, Nuka Armini she is not my daughter she is not my wife and I cannot say it just with the bear now so I'm correcting you you have an assumption I'm showing you this assumption is wrong and with what I say next I assert my epistemic authority over that so you don't know this I know it I've sort of shown you explicitly and I stick a me there just to make sure that it got across the other one Manda Atarikaniichu, Tiano Kadiami she didn't stand up they just sat there so again you were maybe saying that they stood up and went I tell you they did not stand up they were just sitting there I'm correcting your false assumptions same mechanism really and again this cannot be said without me and this contrasts interestingly with another enclitic which is in the same morphosyntactic slot which is Ta and in terms of focus marking this can be said to be associated with so with focus that reinforces the true value of a proposition so this is taken from an interview with a midwife and the interviewer is asking sort of for confirmation they're talking about what to do with a placenta when the baby is born and she says does the midwife have to bury it and the midwife essentially repeats the same turn saying sorry it's a little bit out of line but I hope you can see that and so same thing repeated information with the Ra so yes she has to bury the placenta we share this information because I already told you but you're also let on on it so it's the two of us sharing this information here and in this case people would maybe tell you that you can actually say it without the final enclitic but they would never say that in interaction and me is absolutely out you cannot say that with me so basically this is a situation where this enclitic would be obligatory where we share this information confirming it to you but it's a thing that it's shared between the both of us right shared versus exclusive knowledge in interaction so how does it work it's really another example confirming what I said before and bringing it together so on the left hand side it's kind of two sets of dialogues I'm not sure if that's clear so A and B here and then we've got another one there this actually came about when I was we were talking about making chicha so traditional drink and I said I can't do that and my good friend in the village said yes you can do it so she was trying to make me kind of define my own expectations about myself it's a little bit of a strange case right because she is asserting epistemic authority over something that's my domain of experience but she is the cultural expert here she knows about chicha about making it and she knows me so she probably knows that I have all the necessary skills in order to make that drink however if we had a different context so if I was already trying to make chicha and she just wanted to encourage me then it would be appropriate for her to say you can do this because I'm already trying so I'm assuming and trying that I can it wouldn't make sense for me to start doing it if I didn't think I could so she's just reinforcing this feeling in me that I can from the position of authority saying yes you can do this from this keep going and actually me I haven't preferred examples of this but me is also quite often used in in directive or prohibitive context so when you're trying to say don't go there don't climb the tree or rip your trousers things like that and that's following this analysis because there is something that we know is going to happen but the person who is performing the action and we don't want them to do this so we invoke this me as a claim to our epistemic authority to stop them effectively from proceeding on the course of action on which they are set so that also confirms this authority thing and I thought it was going to take me much longer to go through this but you're probably already saturated so we're going to start wrapping up slowly so this was really just a snippet of data to give you a flavor of what's going on in Amazonian Kichwa and so what I found mainly about these markers and there is me and Ta and there is actually 10 more of them so I haven't had the time to explore all of them I remember the day when Irina told me that I should probably look at all of them 15 rather than 3 that was quite shocking in the end I didn't look at the semantics of all of them but I was able to establish that they all occur in the same morphosyntactic so effectively they can be considered as a paradigm and they are cognates of evidential markers in other varieties but they are semantically and functionally distinct from them so that's kind of the main finding and it's already pretty obvious from what I told you however the system that you've seen of just the free evidential markers that Mark direct inferred and reported is not universal in the whole casual language family there is some new research coming from Peru which suggests that in the dialects over there you get systems with 5 or 6 markers even where there is a distinction between mutual and shared knowledge so actually that was kind of hopeful when this came out because it means that there is something to it in the way that the systems are evolving because we also see here that there is a distinction between knowledge that we share and knowledge that is exclusive so it's not exactly the same system but same type of sort of pragmatic semantic discourse evolution these markers again contrary to what you might find when you look at typological overviews of evidential literature where Quechua comes up as one of the languages of grammatical evidentiality are not grammatically obligatory they are more dependent on you know on how to structure the turn of pragmatic needs of participants of discourse and that's not really that distinct from what we know about Quechua in Peru it's just or other varieties that have been described which mostly happen to be in Peru where authors have actually said that they are in the obligatory but because the work has mostly focused on the cases where they do occur there wasn't so much talk about the conditions under which they occur and under which they do not occur so it's not that this is a new thing and they've sort of been demoted from obligatory markers in Peru to suddenly occurring in 2 to 7% of turns in the Equatorian Amazon is just that we haven't really in our research before looked at what it is that makes them occur but so that's something that should probably be looked at in more detail in the future these markers at least the two that I've shown you are intersubjective because they don't only have to do with how I as a speaker know things they also have to do with how my knowledge relates me to my interlocutor so there is kind of a triangle so it's not just me it's also you and there is also the text external word that we both relate to in using these markers and because of all of this their analysis actually needs to go beyond morphosyntax and semantics so that's I think where the real challenge keeps in the context of languages like Kichwa because you've seen that we started with an English example and that was already conversation analysis because we don't even have to provide glosses for English however if you want to study these things in endangered languages or minority languages you all know too well that there is a lot of ground that we need to cover before we get to this so it means that within a PhD thesis I could only barely scratch the surface of this and that's why I think it's really important to commit to studying these things and not to regard them as something secondary when you have pragmatic strategies and discourse you cannot account for these elements unless you know something about it and just because the grammar doesn't tell you what they are evidentiality 30 years ago a lot of people were saying those are just stylistic particles there for the language to be pretty we have to walk away from that and see how we can work to be able to account for this so maybe team up with someone who does conversation analysis maybe record a lot of natural text so then when we understand the language actually we can come back to it and figure out what it does but I think we also from the point of view of creating language documentation that's representative we have to think about this as well it's not just enough that we cover semantics we also have to think about discourse so because that's the first thing that's going to disappear effectively especially if these markers are not ingrained in the grammatical system or are not part of the tense aspect mood and general conclusions all the Amazonian Kichwa evidential analytics could in fact be analyzed as discourse markers so discourse markers that's coming back to Schifrin in 1987 who analyzed expressions such as you know in English syntactically non-obligatory elements that bracket units of talk and operate on the discourse level and that's effectively what these particles do so there has been almost what 40 30 years now of studying that within English and I've talked to some people who have done this kind of research for English-Spanish well-described languages there might be a lot of interest to kind of incorporate that but there hasn't yet been much talk about all these model particles or weird words that we discover in language documentation and if we look into more standard linguistics we might find that a lot of these phenomena have been described and we might draw on that right the meanings they express are related to ownership of knowledge and interaction rather to source of evidence so that follows from the fact that they are not evidential morphosyctactic marking of the ownership of knowledge is by far as you have seen not unique to Kichwa and when there is no morphosyctactic marking of these kinds of phenomena as you have seen on the basis of the English example that doesn't mean that it's not there it's just expressed by other means so for me it was quite an interesting thing to come around to because when I first started doing research I thought you know evidentiality is a grammatical phenomenon therefore languages that don't have evidential markers we shouldn't really consider them and I think I've come sort of full circle because I think we have a lot to learn from ways of expressing the semantic functional categories for other means so to me it's important not to get fixated on you know whether or not we have this as a particular morpheme because there might well be other means of expressing that right the study of these phenomena is still an early stages of development so there you go maybe that's something you might want to consider doing and in fieldwork context there is a challenge ahead of us in terms of devising a methodology that's actually going to work for that right, thank you very much I'm done here and those are the references I don't want to make life more complicated please do I think there's a whole another domain that you didn't talk about which is Dan Spurba's work on ground and that's fed into relevance theory and the stuff that David Wilson and other colleagues have developed and I mean you didn't refer to that but that's a whole another domain in terms of intersubjectivity right the problem that common ground semantics or pragmatics ran into is infinite regress so what do I believe that you believe that you believe about that situation because you talked about it as intersubjectivity and making estimations of what the other person knows yeah I did actually let me just go back a couple of these weird slides so that we can go to this example so again I have trouble sort of delimiting where one thing stops and the next one begins but in my analysis especially of me and actually Ta as well what I think they do if you look in terms of common ground so common ground is a set of beliefs that's mutually believed to be shared between participants of interaction and in this relevance theoretical sort of question under discussion think the way that communication works is that we pose ourselves questions which we then resolve and that the answer to the question that we've been posing comes as part of the common ground so what I think these markers do in terms of common ground is that the me it's different sides of the same thing but what it does it tells you okay I know you might not believe me because you were expecting something different but actually this is what is the case and I'm urging you I could just say that without the marker but I'm telling you let's just accept it and move on to the next thing because I tell you this is how this works so it's sort of and then there is another marker Cha about which my analysis is not very sound so I didn't really want to talk about it today but that's a cognate of the indirect inferred evidentials and what that does is sort of the contrary of this where it says I think that this is the case but it's not on the sort of epistemic modality level uncertainty or not but like just sort of presenting it out there but I don't have enough information for this to become part of common ground can you elaborate on that and then we can solve this and start this in our set of mutual beliefs there's a lot more to that as well anyone else? oh yeah but that's not yes I guess that's also my concern about concern about this whole epistemicity thing right because okay dimensions of knowledge it sounds appealing it's cool you know it's intellectual to do this but when we actually start pinning it down we're again we're not really philosophers of language so my issue with domains such as engagement say I can see clearly where egophorecity comes in or what it means to be surprised I can sort of quantify that but if we assume that there is a semantic factional category which has to do with everything that has to do with the imbalance of knowledge I think effectively that is too broad to be descriptively helpful so I think the whole field is stumbling into that and also there is a whole tradition of information structure which is not talking to this and it's related to the same phenomena because it's looking at how the organization within a sentence so not discourse but a sentence level accounts for what we know and what we don't know so common grounds as far as I know what it's originated within information structure or it's very it didn't alright so but it was fed into there and then it came back here and I don't really have an answer to that other than 20 years of research and then we see what we've come to I don't know if you have any take on how this could be resolved no I'm just saying that which was that you need to look interactionally and and to look I don't I don't like this term natural discourse but you do need to collect instances of real language use to try and make sense of things and treating it as a grammatical phenomenon in a paradigm is not actually going to lead you very far with you on that thank you I'm interested in just your methodology and your knowledge of kitchenware as well how fluent are you in kitchenware and how much do you know about the interpersonal relationship of everybody in your field where you're working and how how basically how did your analysis come out of that the knowledge that you have so what I've mostly and so my kitchenware now I think is intermediate I can pretty much read my entire corpus without looking at the translation but the way I picked it up I tried to learn some Peruvian before Peruvian kitchenware before I went there but then basically we just started recording I did the elicitation so I learned a little bit from that and from just infinitely reading through this corpus because recordings were coming up pretty early on and I had to look at the transcription so the first ones were really hopeless when I was saying oh I think it says this and they said no there is no such word and then you would pick it up but it was just days and days on end trying to figure that out and when I last went there in May I was able sort of with a kind of draft to deliver a discourse in Kichwa to kind of show them my appreciation for the whole project that we've done and we've had an hour long meeting that only took place in Kichwa and I kind of was able to follow so that's where I met with the language in terms of the interpersonal relationships that's tough I think I know some about it it's still not too much but that's why I also want to keep working in the same place but that was also the point of dividing the corpus so I had this one where there were pair stories and kind of narrative descriptions of simple images where I could sort of know what people knew and then I looked at what these and clinics were doing in that part of the corpus and extrapolated sort of interim results from that and then went to real instances of language use which I agree is a better term than natural discourse and so ever there was kind of the same pattern going on or whether there was something different and when there was something different I was sort of trying to think about where it might have come from so that was, it's patchy and probably if you get a native speaker to comment on that they would have much more insight kind of straight away without need to do five years of research but we don't unfortunately have native speakers of the language and kind of meta-linguistic awareness yet so this will have to do for now Historically and why is it different from the Cusco Caccio right where the Cusco Caccio stands for other epistemic types and maybe what's the origin of the epistemic there's one article that I referenced which was Heinz and Heinz they have proposed a development for these markers in terms of subjectivity intersubjectivity kind of thing so basically things becoming more and more subjective and then more and more intersubjective and that would kind of align but the details of their analysis were foreseeing quite different things to what I found so my PhD has only looked at the synchronic kind of state of things in the last chapter on diachrony but I wouldn't want to comment on that just because I don't yet know enough I have thought about turning it into a research project the issue that we run into there is that this is very different from a lot of sources that you can access that have a kind of big historical depth because not many people have been doing this research on the basis of spoken discourse so effectively and in written interaction it's a very different thing of how you project yourself and how you use these things and in fact in teaching of unified Quechua there is a prescription that the me should always go on the head of the first constituent of the sentence so I'm worried that in terms of bigger time of it it might be hard to do that but it would be interesting to see how it compares to Peruvian Quechua date and stuff but the research that I know about is mostly based on a more rigid instances of language use so that would have to be factored in as well no, they don't that's why I'm saying it's actually a paradigm so the me, there is nine in clinics which do not co-care so they're in the same slot I should have actually prepared that so that you could see this and then there is five more which co-care with some of them it's a little bit of a complex picture of how they can co-care but three of them that I've been talking about today never co-care so and I try to elicit that but it seems that they're mutually exclusive and it's pretty unacceptable to use them together I'm just wondering about your experience as an L2 speaker of the language whether you use these in clinics yourself naturally, whether you got a reaction to that or whether you got a reaction to to misusing them and did that tell you anything I was thinking about the Japanese yor which you mentioned that as an L2 Japanese speaker I think I'd shy away from that sometimes because I worry it can sound kind of arrogant all too strong for an L2 speaker was there anything the equivalent to that that taught you anything I don't think I have sort of good enough command language to deliberately mess up in a way that I could then use to tell me something right say ok if they said this in this context this is what it means I'm not there yet however I've been using quite a lot of the me and chat and the topic sort of topic marker as well and there is a good response to that always so and if they thought that I was arrogant for using too much me I haven't found out about that yet but there are people who use more me than others I can see that with my consultants as well at least in sort of meta-awareness level where I haven't quantified it but some people do seem to voice that much more the relationship of the markers with corrective focus but I was going to ask if you've seen anything to do with the topic yeah I did it's not here I think I got over you because I was too anxious to answer the question so a little bit of topic and I'm not sure I find it interesting and I'm not surprised that it sort of correlates with information structure analysis so there is this Enclity Ga which has been analyzed again for Peruvian as a topic marker in Tena Kichua or Amazonian Kichua it's not frequent enough again to run this analysis because it's around the same percentage it's about 5% of turns it occurs on topical constituents again not all of them but it can also go on secondary topics and generally presuppose the information I actually have an article about this in the working papers I think from 2016 published here at SOAS of the top of my head I don't think I can tell you in much more detail that was an interesting crossover actually when I saw that in the real language use corpus and then I went back to the elicitation I've been doing of a sort of naturalistic discourse where I found when they were describing pictures a lot of times they would go like Kaiga and then proceed to describe the demonstrative Ga the topic marker and then proceed to describe the picture with another topic marker in it which never happened in natural discourse you wouldn't get the stocking of the topic markers so it's somehow related to the presupposed information it doesn't seem however from what I've been able to figure out to have this epistemic correlation I mean it probably does because if it's again you know common ground issue if it refers to something that's presupposed then it is epistemic like everything is epistemic in the bottom line but it's not strictly a topic marker it covers much more to find the last analysis it took her a good 20 years to finally feel that she's practiced enough to write the paper about it now the menu I want to ask you the question I thought it was going to be can you show us the Japanese example sure is that the end of the question that's pretty easy alright there you go and the question is what about prosody of course if you have something like yoku anoko arita ne arita ne and aruki yo something very different from aruki yo you need to look at international contours there this is the case it's about who holds global authority right so I think you've expanded the ground absolutely but must also incorporate prosody international my answer to this is more funding for collaborative research two years ago there was when they would show but I totally agree this is much needed as much as kind of looking at the position of this not only sort of within the clause but also within the term there is new interesting research being done on how discourse markers and languages such as Spanish are influencing their meaning by their position in the term and the people doing that are actually also doing prosody analysis so there is a lot that I would need to teach myself to do that but this is definitely the next level kind of trying to figure out what it is that they actually do it's also a huge idea because you know too much work that's being done in this area is based on second hand sources which are purely textual and you cannot analyse this adequately I find only on the basis of of actually interesting I was giving a talk about similar about corpus design to people who actually only work on Spanish and their corpus is all sort of real language but it's only audio and they were asking me so what is there to be gained from video and we cannot expect especially in my setting when it was native speakers who are computer literate but not really computer savvy no one could really do that to transcribe a party conversation without video that's in plain impossible and also if it has to do with the exist like some people analyse evidentiality as they exist so that has to do with the word that surrounds you might be pointing that's involved and a lot of cues that are given by the image that simply would lead you to misinterpret these markers in their absence so and I found that actually at least in my context people weren't hugely intimidated by the presence of the video camera because also that's given them something to then watch back right they really enjoyed that when I went back with a CD and they could see the interview and the most successful thing I've done over the whole time I've been working with them is produce this video where they were talking about ethnobotanical plants everybody absolutely loved that so I think there is a big case to be made for always just going for video as a default to be surprised I have learned about something which attempts to work out of the surprise by saying that we can stop using the concept of mutual knowledge instead we use something like mutual metaphexness so it becomes instead of saying I know you know I know you know I know you know the fact you can start thinking about saying manifest to Yen that it is manifest to Peter that it is manifest to Yen so seems even possible that we can find some discourse markers that talk about metaphexness and even mutual metaphysics in addition to something like knowledge there is a whole set of markers I think again Henrik Bergfist was writing about it in Kogi where there was a marker which indicated you should know this it's but we talked about it right so I'm reiterating this to you but really it surprises me that you shouldn't remember that this is already it should already be manifest to you I wonder though whether this mutual manifest is not running into the same issue as common ground would run into what does it have that really makes it all that much different and it should be based on facts that you know about rather than just being a theoretical knowledge designer okay I'll have to read out on that thank you alright and I also have no issue of sharing the slide so if that's something interesting I can leave it available to you and they can also be uploaded on the website okay brilliant great okay thank you very much