 OK, hefyd rhai! Rhaid i chi i chi'n gwybod, dyfodol o'r cyfnod i'r ffyrdd yma am ymgyrchu, i'n gwybod i'r ffordd yw'r Gwydiannau Gifson, Dr Hannah Gibson. Hannah wedi gyda hwnna MAPHD-ynghylchu'n Lengwysig Ysbryd, i chi'n gweithio'r cwmwysig Ysbryd. Rwy'n gweithio'r cyfnodau cyfnod, o'r llynwys ymgyrch a'r cyfnodau dynamig, Over the course of her PhD research, she's become more and more interested and focused on the importance of language change and language contact in her field area, which has led to a new project, which we're going to hear about in a minute. Great. Thank you very much for the introduction and the invitation to come and speak. So, yes, as Rachel said, I'm going to be talking about contact and change, and I'm going to be asking some kind of broader questions about language contact and language change, but then going to look at some specific case studies which relate to a project that I'm currently working on. So, there's a sort of general consensus that languages can change and that language change can result from external factors, so in that context I'm talking about contact with other languages, but also internal processes, so again talking about grammaticalisation processes. But there is little consensus on what exactly constrains these types of processes, the changes that can happen, nor on the interaction between the two types of change, so the interaction between internal processes of change and external factors of change. In terms of agreement, I think people, there's quite a lot of agreement that there are no absolute constraints on structural borrowing, so what can and cannot be borrowed. But despite that, there is a kind of, it appears that borrowing is not entirely arbitrary, so the types of borrowing that take place follow some kind of patterns perhaps. This is referred to in the work by Thomason Kaufman as a borrowability scale, so you might find that lexical items are borrowed more easily than grammatical structures very briefly put. However, again, even within that context, if we observe sort of those patterns, we will still want to ask why is that the case, so why are certain classes more easily borrowed than others, and what kind of factors play into that. I'm going to be looking at a part of Central Tanzania, so here highlighted with a sort of orange oval, and the Riff Valley area of Central Tanzania is unique in that it is the only area on the African continent where languages from all of the four African language filers are found. So there's some slight dispute about one of them, which are the Koisan languages which are there in blue, but otherwise you can see that you have all of the other three language families are found. So you have Nylosaharan, you have Afrosiatic, Niger Congo, and the Koisan languages are found there. Against that backdrop, there is a sustained history of language contact between these language families, high levels of bilingualism, multilingualism and continuing language shift. In terms of what I'm going to be talking about today, I'm going to be examining language contact and change through the lens of a subset of languages which are spoken in this area of East Africa. So just to sort of give you a taste of what's to come, the Tanzanian Bantu language rangi has an unusual word order in which the auxiliary appears postverbally. Those of you who knew me when I was doing my PhD research here will know that I was working on rangi and this is what I started looking at. So you have a construction as in one where you have mother will collect water and you have the infinitival verb collect comes before an inflected auxiliary. And this is unusual in the context of East African Bantu languages where the opposite order, so auxiliary verb predominates. And it's also unusual for SVO languages to which rangi belongs, which again more commonly exhibit auxiliary verb ordering. So this kind of causes rangi to stand out from both a comparative and a typological perspective. Rangi has spoken an area of high linguistic diversity will look more exactly what I mean by that in a minute. And it has proposed that this feature is the result of contact with non-Bantu languages so that it's a result of this contact situation. An alternative proposal has also been forwarded, which is that this actually might be better understood as a feature resulting from language internal processes, so grammaticalisation processes. Having said that, rangi's not alone amongst Bantu languages in exhibiting this post-verbal auxiliary placement. So by way of background to the project when I started my PhD I thought that rangi was the only language to exhibit post-verbal auxiliary placement. Pretty soon after I started my research I found Mbugwe also did that, which is a language which is spoken nearby. Toward the end of my doctoral research I found Goosey and Courier and then just before Christmas I found that there were two more languages, Angarema and Super Symbiti. So as I said earlier on this project which started as one and then started as four has now become a project looking at six languages. And I think I've got them here so we've got Rangi and Mbugwe which are spoken in this at the central area of Tanzania or north central area of Tanzania highlighted in the blue here. And then so there form one cluster, I'm going to be calling this the Kondoa cluster and we'll look at that later on. And then the other four languages are spoken up here near Lake Victoria, I'm going to be calling those the Lake cluster. And as you can see here, so this is a map of Tanzania, some of those languages also found on the Kenyan side of the border so they also cross over into Kenya. So that's the second set as you were, Goosey, Courier, Angarema and Super Symbiti. OK, so in terms of what I'll be doing today I want to develop an account of this verb auxiliary ordering in this subset of languages. And I suppose the sort of origin of that is to say well what do these languages have in common that might be able to account for the presence of this marked structure. So as you saw on the previous map just to go back there although we can say these languages are close to each other and those languages spoken in an area close to each other they're not actually in contact with each other so we don't have contact between all of the six languages. So we're looking for something in common and I also want to look at the questions I started with so to what extent might this be considered to be the result of contact specifically with non-Bantu languages spoken in the area and to what extent could it be considered to result from grammaticalisation processes so language internal processes. And I have a few other sort of smaller questions relating to that. I'm going to be looking at what the differences are between these six languages specifically with respect to this construction. Then I'm going to be looking at other contact features or other possible contact features to sort of be able to get a larger picture of what's going on in terms of interference between the languages. Looking at possible pathways of change so possible proposals that could be used to account for this structure and then what other kind of information might we be looking for to support either account or indeed a combination of the two. So that's sort of a little bit of what we're doing today. The bits of an introduction I'll then be looking at, it'll be quite sort of data heavy sections so looking at this particular ordering. A bit on the sociolinguistic and the historical backdrop which I think is important to sort of try and understand what we're talking about when we talk about change. Proposals to account for it so again as I said a kind of pathway, possible contact features and then what I'm going to do so this is the beginning of a project but what other kind of questions I want to be looking at. Okay so on to the data, I think everyone's got a handout, there's quite a lot of data but hopefully that will all be clear on that. And in terms of just background, tense aspect, mood, distinctions in Bantu languages can be encoded through simple verb forms or complex verb forms. So we have examples here from Swahili, we have something like I am going to school where we have a present progressive marker na in the verb form. I am going to school, a simple verb form and the example in three, similarly we cooked food, a simple verb form and all we have is just a single past tense marker here li. So that would be what I'm calling for today's purpose is a simple verb form, typically simplex with tense and or aspect marked on the verb form. We also have complex verb forms, we also have complex verb forms so we have some examples from three other languages where you have a combination of an auxiliary and a main verb. So the example in four from here, another Tanzanian language, we have an auxiliary followed by the verb and something like we will have bought. So now you're encoding a slightly more complex tense aspect distinction. Sorry about the alignment in five, we are buying another Tanzanian language in Gindo and we have again auxiliary followed by some kind of main verb. So here the auxiliary is a kind of do verb, buy and we are buying kind of present progressive reading. And finally we have an example, this is from the PC. Finally we have an example from Tswati where we actually have interestingly so more than one auxiliary form. So here we have the boys might smoke pot again and we are going to have three, we are going to have two blue circles above the words. So I had a little technical problem before I started here but basically in terms of the ordering is the thing that we are looking at. We are going to have an auxiliary and another auxiliary and a verb. And if you just look at these examples you can see that there is already slight variation. So in Heihei you have some subject information on the auxiliary and on the verb so we have got that two on both of those. In Gindo you have subject information on the auxiliary but the verb appears as an infinitive so just there as a ku. And in Tswati we have the inflection on the auxiliary and the verb here so we have got ba the whole way along. So just sort of looking at that we have slight variation in subject marking. We have variation in terms of the number but also the form of the auxiliary. Despite that the order that dominates across the language family is auxiliary verb. The exception being the six languages that I am going to be looking at today. So the first language is rangi and just to give you some sort of broader background this is our auxiliary constructions, our complex constructions. In seven we have that ill person has died and we have an auxiliary form. Let's see where the circle goes. Yes we have an auxiliary form rhi there which is inflected also for subject information and past tense information. But then we have the verb coming after that so this ill person has died. So these are past perfect examples. We will have exactly the same in eight. We have our auxiliary followed by the main verb. So this sort of shows you that in rangi we do have this order. We do have auxiliary verb. We do have what we have here and that would be more common across the language family across Bantu language and across East Africa. However in two tenses in the immediate future tense and in the general future tense this order is inverted. So in nine you have we will plant banana plants and you have the verb plant comes before the auxiliary. So we will plant banana plants. This is some kind of immediate future. Some kind of incipient action we will plant banana plants just now or very soon. And then a general future which is much more general not specific in terms of time. You will open these beehives and again we have the verb first open followed by an inflected auxiliary. There. And just to show you that an attempt at changing that order is not acceptable. We have example eleven with an intended reading of something like I will cook food using our auxiliary. And attempting to put the verb afterwards and it results in ungrammaticality. So this is actually this is a grammatical encoding. This is how the future tense constructions are formed in these languages in this language. A very similar case in Umbugwe. So we have a simple tenses like we saw at the very beginning from Swahili where you just have past tense marker. So in the example in twelve I cultivated my farm in this case last week. We have a past tense marker and similarly in thirteen I will not talk to you. We have a future tense marker just that ja there. So this is a simple construction simple verbal tense. And then you have the ones that are formed through auxiliaries. So in fourteen we have a present progressive construction. The visitor is ill. So the English translation of that doesn't necessarily sound present progressive but this is something like the illness of the visitor is biting. So you have the verb followed by the auxiliary. So that's how you form the present progressive in Umbugwe. So a combination, a complex construction of a verb followed by auxiliary. We have exactly the same in the recent past progressive. He used to refuse beer, he used to refuse to drink beer. We have our verb here refuse followed by an auxiliary, a different one here we have rare. And finally in the present habitual do you eat fish or do you eat fish? We have our main verb here eat followed by the auxiliary. And just to point on Umbugwe there are actually other tenses that show this ordering but each one uses a different auxiliary. So the present progressive is specifically formed with the auxiliary again today. The recent past progressive specifically with that auxiliary and some inflection and habitual always employs this auxiliary. So far we have got our Rangi and Umbugwe examples. So those are the languages which are in the central area which were coded originally in blue. And then we also find the same verb auxiliary ordering in the other languages up near the lake. As I'm sort of going through the languages you'll see that the state of description and the data that I have access to sort of varies. So I worked on Rangi, there's someone else and quite a lot of work has been done on Umbugwe. As we get onto Goosey there's less, so my examples are sort of there's less rich source of examples. And similarly as we go on you'll sort of see some of that in the data. So here we have what's been described as an untimed fact or occasional habit. And again we have our main verb followed by an auxiliary. And similarly in a present continuous example in Goosey we have a main verb followed by an auxiliary. The only difference here in these two forms actually in the case of Goosey is that there's this ra marker on byte here. So this is the only distinction that you actually see between these two and there's also some change in the tone marking. So we have something like 18, I am unfolding the blankets, a present continuous event. So again we've got this unusual word order, this verb followed by an auxiliary. And actually in Goosey in a number of tenses so we also have it in the example in 19. You were biting, this is a hadernal pass so you were biting earlier on but earlier on today. And we have our verb followed by an auxiliary. And similarly the last one here we have is a past habitual construction and it's the verb followed by the auxiliary. But this is a different auxiliary now so the habitual is formed similarly like we saw in Bugwe, a different auxiliary used for different tenses. Goosey there's not a complete one-to-one mapping but this is a different auxiliary from the one we saw up there. So that's like 20 I was harvesting, some kind of past habitual action. So we've seen so far Rangi and Bugwe, this is Goosey so this is the first language on our kind of lake cluster. A neighbouring language spoken in that area is Curya and I've identified this ordering in at least two tenses, possibly more. Again this kind of now comes to a state of description so people saying oh I can't go into all of the auxiliary constructions, it's too complicated. So what I have here is what I know so far. So present progressive we have our main verb again followed by the auxiliary. So they read, they are reading and similarly with a simple present he or she is becoming injured. So here again we have a tonal distinction on the final vowel there. Now these examples from Goosey but also from Curya also have a little marker at the front so here it's glossed as focus. And I'm going to talk a little bit about that later on in terms of what that might mean for information structure. But this presumably comes from some kind of structure which was like a cleft construction. So this would have been it is reading I am or something like that but it seems that that's fossilised now. And again I don't have the specifics of if you can have that without a focus marker is that then a focused verb they are reading not something else. But those are the kind of questions I'd like to look at in more detail. So that's Curya and then on to one more language. So if you're following on the data then this is on page 5, 2.6. We have a number of tenses, we have a present progressive tense construction here. I am singing verb followed by the auxiliary. So we have sing again we have that little focus marker we have an infinitive marker. And this auxiliary here I've actually glossed as a locative auxiliary. So this is the way that you would also say I am at school or I am in the room or something like that. So it's almost like singing I am at. So that's a sort of what I'm calling here so far anyway a locative auxiliary which contrasts slightly with what we have in the present progressive. Again our verb with an infinitive marker a focus marker but we have this auxiliary re. So these two have two different auxiliaries are used one with some past tense information and one some kind of locative auxiliary. Now the language that I presented so far I showed actually that there was a rang example which was ungrammatical if you changed the order around. And it seems that in gwrome at least in the past progressive there's actually some flexibility of word order. So that's another interesting area for further inquiry. So here we have what we are not allowed in the other languages and what we're not allowed in the other tenses which is the auxiliary coming before the verb. And in this in terms of what we have as a translation he or she was waiting and I was singing. We don't actually have much to distinguish between those and they're both considered to be past progressive examples. So that's sort of interesting in terms of in other tenses those were completely ruled out. And in this tense at least it seems that those two structures are coexisting. So I can talk a little bit about that later on. And our final example, a final language here is subas in beauty where you have this ordering in the present progressive. So 26 something like we are digging with a verb infinitive and our focus marker or historical focus marker and our auxiliary. And we also have a present habitual construction verb here. Interestingly there's a habitual marker on the end. So if you think of an infinitive typically we think of them or we might think of them as not hosting tense aspect information. We certainly might think of them as not hosting subject information. But here you have infinitive with a habitual marker on it as well. But again this consistent verb auxiliary order as something like we are dancing, we dance, we are usually dancing. Okay so a bit of an overview kind of interim summary of what we've got so far. Rangi this ordering is found in two tenses and the two tenses use different auxiliaries. So we have two tenses, two auxiliaries and they are ri and ise. Mbugwe, we have six tenses which exhibit this order. Six different auxiliaries and there's a one-to-one match so one different auxiliaries used in each of the tenses. In Gwsi we have five tenses where this has been identified but two auxiliaries. So actually I think in four of the tenses it was it's re and in one of them alone it's this rengo one. So you can also ask about the relationship between those two. In Cwria I have so far two but I put a question mark because there might be more. One auxiliary is consistently this re. Ngoreme, two tenses, two auxiliaries and that's the form. And then subasinbiti two tenses and the same auxiliaries used in both forms. So we've got even within this sort of construction we've got a little bit of variation in terms of the number of tenses, in terms of the auxiliary forms that are used and the overlaps. Some of them have a one-to-one match, some of them don't. And on here I could also shown that for example in Ymbugwe all auxiliary constructions have this ordering. In Rangi not all auxiliary constructions have this ordering. So there are also other auxiliary constructions which don't have verb auxiliary. Okay so that I think is the data side of things. Now I want to talk a little bit about the socio-linguistic background and the historical backdrop. So I said that these are languages spoken in two areas. So what I'm calling the Kondoa cluster, this little giant really red triangle is Kondoa, so a town in central Tanzania. And the first circle that comes up here is Rangi and the second one is Bugwe. And in terms of this map, so this just shows you a little bit about the language families as well, the Bantu languages here are coded in green. The Krushitika are a very sort of light purple, so some around here. Nylosika this much darker, stronger purple that you have up there. And what have sort of disputedly Koisan languages we have Sundawe and Hadza there. So a kind of lighter yellow, I know that doesn't show up brilliantly on there. So this is our Kondoa cluster. We've got Rangi and Bugwe spoken here. Now in terms of the historical picture, this area of central Tanzania has a long history of language contact. So the Krushitika languages that are present there today have been estimated to have entered the area some 3,000 years ago. Bantu languages something like 1,000 years later. But like Datogo, which is a nylotic language actually much more recently. So you've got a sustained sort of history of language contact between languages from different families, very different language types in terms of syntax, in terms of all sorts of domains. But this kind of long history of contact. And that's kind of played out in this quote we have here which says that all linguistic observations of Bantu and Krushitika languages in the area indicate that there has been significant interaction and indeed interference between the language groups. So now we're sort of looking at well what is the nature of that, what kind of change are we looking at, what kind of contact are we looking at. Just to bring you up to date with the sort of present day situation. So Condola town where I had that red triangle and the surrounding areas. There are some 40 languages spoken in that area. Presently they include so we have Krushitika languages, Iraku, Burungae, Gorwa and Alagwa. We have nylotic languages, Datogo and Maasai. Koisang language, Sandawe. And then Bantu languages, Rangi and Bugwe, Gorwa and Chaga. Now there are lots of other languages that are spoken by people who live there. Swahili is also a dominant language throughout East Africa. It has official language status in Tanzania so it's spoken on a day-to-day basis. But these are sort of languages in terms of, well I'll talk a little bit more about the actual interaction between the language communities. But these are the kind of languages which have a strong foothold in the area and have done for some time. So to just narrow in on this Condola cluster, we have Rangi, which I think is yet, which is this sort of spoken in this whole area. So Condola town is here. Dodoma, which is the capital of Tanzania, but sort of administrative capital of Tanzania. And then Arusha, which is a big city, sort of closer city to Kenya, so it's also very important in terms of trade, right at the top there. And there is the Dodoma Arusha Road, which connects them. And Condola is a sort of town almost halfway between the two. From about 60 kilometres south of Condola to about 60 kilometres north of Condola, the villages are Rangi-speaking villages. But the south side, you have Rangi and Burunga speakers. So on the south side of Condola, the villages are mixed. You have households where people speak Rangi and Burunga at home. You have neighbours who speak the different languages. So these are the two sort of primary languages which are in contact there up to today. And then as soon as you go north of Condola, Alagwa, another Kushitic language has a stronger sort of foothold. And then you find these mixed villages, if you will, being Rangi and Alagwa as you go further north. In terms of Umbugwa, and then you have another Kushitic language here, which is Gawawa, which actually spreads the much wider area, but that's sort of up the north there. And then you have Umbugwa, the much smaller speech community, spoken around the town of Babati. And this is in contact up to today with Irakul, so one of the large Kushitic languages there. Also Masai, but that's actually this whole sort of area. And I sort of pointed to it earlier on, but as you can see, particularly or historically, the Umbugwae and Rangi-speaking communities are separated by the Gawawa and Irakul-speaking communities. So these two Bantu languages are actually entirely surrounded by non-Bantu languages. Having said that, Rangi and Umbugwae are presumed to share a common predecessor language, so some kind of proto-Rangi-Umbugwae that they're supposed to derive from a common source in terms of the sort of more ethnographical and anthropological perspective these communities are supposed to have moved together and then split either one account says after the battle with the Gawawa, so after the battle with this community, or at a later stage when they travel south in search of water and ended up, the Rangi ended up in the town here, so Kolo and Halby are the centre of the Rangi-speaking communities. The Lake Cluster, so this is what's on this map anyway, called Lake Victoria. We have this whole area and also, sorry, onto the Kenyan side as well. I'm not going to talk too much about that, but it is a kind of factor in the picture. And what you have here is actually a really high concentration of Bantu languages, so actually all of these little lines here are Bantu languages. You have Nyloftic languages, so that's the dark purple, but in contrast to down here, you don't have Cushitic languages. So again, even if we're looking at contact in two separate instances, they're not even with the same language groups. Having said that, high concentration of Bantu languages in a small area along the lake. A sustained history of language contact, particularly with the Nyloftic languages Lluo, Maasai and Atoga. And then a strong Nyloftic presence specifically in Kenya as well, and I just said that, there's no strong Cushitic presence in the region. In terms of the contact languages, this picture is slightly different from the Rangi and Bwgwe story because what I've done here is the languages in bold are contact between unrelated language families. So Gusi is in contact with Lluo. Gusi is a Bantu language, Lluo is Nyloftic. The same down here, Curia is also in contact with Lluo, Ngoreme with Atoga and Soba Sambiti with Lluo. However, the Bantu languages, these four Bantu languages are also in contact with each other. So as you can see here, Gusi is in contact with the other three in the study. Curia is also in contact with Gusi obviously, conversely. So whereas with the Rangi and Bwgwe, we were looking at perhaps two separate, perhaps a common language, a common contact language, here we could also be looking at a sort of different situation where it's gone from Nyloftic language and then between the Bantu languages. So this is, this just kind of shows you where they are and that they're in a close sort of, they're also in contact with each other. And here we do have the Kenyan side, so here we have just to show you this Nyloftic languages here as well. So that would be the Tanzania area, this is Lake Victoria again. So this is Gusi and Curia spoken on this side of the border as well. So they basically sort of form a chain these languages one next to each other. Okay, so it's all very well to say languages are in sustained contact, there's lots of multilingualism, there's lots of bilingualism, but that's not enough to then be able to account for the presence of a particular structure or a feature that you consider to be contact feature. You actually want to see, well what could be the source for this, what could be the source of borrowing or interference. So I'm going to talk a little bit about what I started with, so what the initial focus of the research was, this order, but I'm also going to look at other possible contact features. So the verb auxiliary order in rangi and umbugwe, so in those Kondoa cluster languages. The two Cushitic languages spoken there, Iraku and Goa, both have a rigid SOV word order and they have a periferastic future construction in which a verbal noun proceeds an auxiliary, a Goa auxiliary, to form a future tense. So when I was looking for sources of verb auxiliary order in rangi, especially when I was only looking in rangi, I thought this was great, this was rangis in contact with Gorwa and okay this is data from Iraku, but this seemed like a really good candidate. So you have an example in 28, the animals will drink water and you have a verbal noun for drink, so drinking, and then you have that followed, so that's exciting for my purposes, followed by the auxiliary. So here we've got verb auxiliary order in a contact language, in a Cushitic language that's spoken in the area. Great. Similarly in 29, tomorrow we will not go to work, we have, so actually this is doing, a verbal noun and then followed by an auxiliary. So these seem to be possible candidates for sort of a contact induced account of that verb auxiliary order in rangi, their contact languages and they have this order. So the question is, is this a possible source of transfer for a verb plus auxiliary in rangi and or in bugwe? And the answer is maybe, but still not sure. Martin Maus, who this data came from, noted that these examples are actually really marginal, even in Iraq. There were other ways of forming a future tense, there are other constructions. So not that that in itself means, okay, because they're marginal they can't have been a source, but it does sort of put some question marks over that. You would want a construction, you would imagine a construction to be more widely used, pervasive throughout the language, especially if it's going to then be a sort of candidate for contact induced change. And we also know that more cross-linguistically, movement verbs such as go are often involved in grammaticalisation processes and encoding future tense. So maybe this is actually a case of grammaticalisation in Iraq and that the other forms are more common. And, yeah, it's not also clear in rangi whether, for example, the origin of the auxiliaries, whether any of them are something like go, which might add to this if you want to think of it as kind of grammaticalisation or even aerial grammaticalisation. So there's some kind of question marks over that and as yet I don't have other possible sources for this ordering as such. But there are other features in the languages which I consider to be good candidates for contact features. So one of them is inclusivity distinction in rangi. So rangi first person, plural possessive pronouns, have a distinction between inclusive and exclusive. So you have an example like 30. Today I angered our grandfather and when I'm saying are, I mean mine and my siblings and my other family members and I don't mean yours who I'm talking to. So this is not yours. So this is an exclusive possession. So today I angered ours not including yours. Contrast that to 31. So sorry, that's that one. Contrast that to 31. The death of our relative in Condeau. And now I'm talking about are and it's also the person that I'm talking to. So this is an inclusive possession. The death of both mine and your relative in Condeau. So the reason I sort of touched on this is that including a inclusivity distinction sorry, is that including inclusivity distinction such as this is not a test in any of the other neighbouring Bantu languages that I know of. Nor is it found in the Bantu language more widely. It's only found in these possessive pronouns as well so you don't find it any kind of marking on the verb. So it's really restricted use and it's only in the first person sorry, it's only in this and it's not in the absolute personal pronouns so there's no distinction in like we, it's only possession. So is this a contact feature? Not found in Bantu languages, that sort of points in that direction. Now it's been proposed by Philonova in a book on inclusivity that Iroku is one of the few languages in Africa that has this inclusivity distinction so that's one of the languages that would be a possible contact feature. However I haven't found data supporting the presence of that in Iroku so whether she's got that from a source that I've not been able to access or that sort of that's a red herring I hope it's not but this seems like a good candidate but again haven't found the source of ok this is a contact language with this kind of distinction and that's where Rangi has got it from so again the picture is still quite murky in that regard. Another story that I hope is a bit more robust is cause final negation so again looking at possible candidates for contact features so negation in Rangi is verbal negation in Rangi is marked through a negative marker C and then tool call which appears either after the verb or clause finally so we have in 32 today our friends did not come and we just have our C negative marker there and then we have that one comes immediately after the verb so today our friends did not come and in 33 the chicken will not give us eggs we have a negative marker and then we actually have the negative marker completely at the end so not immediately after the verb ok so this is the case in Rangi it's also the case in Mbugwe slightly different from structure in Mbugwe but they have a negative prefix te which appears at the beginning of the verbal complex and then negative marker tool call which optionally appears at the very end so 34 is something like we did not run to the hospital at all it adds a sort of emphatic reading you can also have it without tool call in Mbugwe that would be grammatical and it would just be we did not run to the hospital so Mbugwe has this optional clause final marker in Rangi it's obligatory and in some joint work with Vera Wilhelmson who's working on Mbugwe we've proposed that these negative markers tool call and tool call in Rangi and Mbugwe are actually instances of lexical borrowing which have been incorporated into these two Bantu languages so although postverbal negation itself in Bantu is not unusual the actual form of these negative markers is suggest that there are examples of lexical borrowing that have been become part of the grammatical structure for negation and we propose two candidates one is the Cushitic language Borungae spoken in the area in which Tuku means entirely or wholly so presumably you have some kind of negation where you did not run and then you have entirely or wholly or some kind of intensifier at the end which is again a very common pathway for grammatolisation of negative strategies or also a lagua, a Cushitic language spoken in the area where you have Tuku or Tuk which again expresses the completeness of an action so we think that this is a good example for interference between Rangi and Mbugwe so the Bantu languages and neighbouring Cushitic languages and that it shows borrowing of a lexical item even if it's not actually borrowing of an entire negation strategy but that's been incorporated into these languages so just to show you that as well as what I've started off by talking about this verbal auxiliary ordering there seem to be other areas in the grammar of the languages where some kind of contact has played a role also on negation in Korea there are two negation strategies which at the moment I can't seem to find any difference in terms of what they mean or what they do so you have in 35 they have not read today and you have a negative marker just a simple ta they have not read today and then in 36 you have a double negation strategy also here at least translated as they have not read today with a negative marker at the beginning and then again a post verbal negative marker unbound on its own there so I would sort of in this context ask well does this reflect the fact that there are these two negation strategies that coexist that you find across the tense aspect paradigm so you find exactly this distinction in different tenses does this reflect different stages in a grammaticalisation process or processes could we propose that this is also the result of contact induced change in kind of vein as we found or we proposed in Rangi and Umbugwe so that's also about sort of more first syntax and about negation one of the other languages that I'm looking at in Gwremae has a vowel system in which you have seven vowels in nouns but only five vowels in verbs so you have minimal or near minimal pairs of nouns in which there is a distinction between the mid vowels but you don't have a distinction between those vowels in verbs and Tim Roth who I've been in communication about this with has a corpus or something like or 1700 verbs and he hasn't got any examples of seven vowels in the verbs so the question so this is a sort of question that I'll go well is this perhaps contact related and that's I would ask that because I'm looking for contacts and Bila which is a Bantu language spoken in Democratic Republic of Congo seems to have acquired two additional vowels in the verbal but not the nominal system and it's been proposed that this is a result of contact so again this is a sort of a different area but just to show that there are other proposals that have been made that are sort of distinction in verbs and nouns in the vocaliic system have been proposed and have been proposed to result from contacts and then coming back to this specific case Gwremae is in contact with the Nylotic language Datoga and very nicely for my story Datoga has 7 vowels in verbs and only 5 vowels in nouns so that seems to kind of at least that one on the surface of it wraps up quite nicely as again another possible example of contact so we've got sort of what it appears as perhaps grammatical sort of morphosyntax possible interference here in the sort of phonology so in the vocaliic inventory so the question is what other kind of things do I need to look at what other questions do I want to ask to complete this picture to sort of fill it out a bit more and some of those questions I've already kind of given rise to through data that I've already got so in rangi for example I showed that order at the beginning you have verb auxiliary order that's only the case in declarative main clauses in the future tense construction so remember in rangi it's just in the immediate future and the general future in future tense constructions this order changes round so in future tense constructions you get the auxiliary before the verb which in sorry in these context I'm going to talk about in a second so this is the case when it comes after a WH element so when it's preceded by a WH element when it's part of sentential negation which as I mentioned is formed using this kind of C in tukul brace if it's part of a relative clause part of a clef construction or if it's preceded by the subordinators jolly and corny so they're a very sort of specific context so in the main clauses in declarative main clauses auxiliary verb order is simply ungrammatical I showed the very example very early at the beginning we will cook and tempting to do auxiliary verb and it's not accepted in these context the order changes not in is accepted but it's obligatory so the question that I asked then here is well if you look at these contexts to what extent actually is this order the result or related to at least historically information structure so is there something to do with focus going on here we've got WH questions clef constructions again some kind of contrastive focus would be usually conveyed through that sentential negation some of the other ones are not quite so clear but again what role my information structure play in this and I pointed out earlier on that in the languages of the lake so in those four languages you also have this focus marker so I was harvesting and it was harvesting or it it is harvesting I was something like that so again this is a kind of in terms of information structure information packaging that might suggest to you that okay well this is actually much more to do with that than to do with contact so that might put you towards a grammaticalisation account because we want to say well actually this order resulted from it being a clef construction it's nothing to do with Crescitic it's nothing to do with Nylotic it's actually a language internal process that's going on so I would want to look at the other languages so I know that's the case in Rangi I don't yet know about Nbugwe and I don't yet know about three of the other languages ordering changes in negation but I don't know about clef constructions I don't know about relative clauses I don't know about WH questions so really you would want to fill up the picture with all of those what I have for Rangi as so-called inversion constructions for all of the languages and then also say well maybe there are other contexts just because I have identified those you might find other contexts in which that order changes and then another sort of question is well what direction or directions is this change taking place in so in Rangi and Nbugwe for example those the Condeau cluster do we want to say that just focusing for a moment on the verb auxiliary order is that the result of contact with one common contact language or is it the result of contact with two contact languages independently but they just so happen that the two languages then develop these these orderings independently or is it perhaps much older than we think or older than that would suggest and the contact between a predecessor of Rangi and Nbugwe and the predecessor of some kind of Southern Crescitic language and then Rangi and Nbugwe kind of developed independently and maintained that in the two languages and the same question you would want to ask for the languages of the lake as I pointed out earlier on slightly more complicated because do you want to say okay this is transfer from Goosey but then from Goosey into the other Banti languages or is it transfer from Lluo in all instances or a different Nylotic language in different cases so the two kind of separate clusters are interesting in that they have some things in common the reason that I'm looking at them as a sort of subset at all is ordering but the social linguistics are also sort of slightly different and so that would be on the contact side and then because I have this kind of ongoing grammaticalisation language internal questions as well does actually the presence of a so-called unusual word order in six languages actually suggest well it's not so unusual tool okay so it's unusual if you take all the languages of the Bantu family and you only find it in six but still six is not one it's still something that is going on in these languages to account for so again looking for something that's common in these and maybe it's a more acceptable or more common pathway of change than we thought so in terms of my sort of continued research as I said the state of the description of the languages varies so I want to look also to see if there are other tenses I pointed to the fact that in Korea I know it's in two but I think it might be in others as well are there other inversion contexts so the uh questions, negation, relative clauses that do sort of interesting things in other languages, in all of them and are there additional inversion context so are there other contexts in which that word has changed and I'd be looking also for other domains so these other kind of contact features that I've pointed to a few of lexical borrowing would be again if you talk about a sort of borrowability, if you talk about a scale people say well actually by the time you've borrowed a grammatical structure such as verb auxiliary order you would also expect lots of other types of change to have happened in the language you'd also expect lots of other instances of interference I talked a little about the phonology and then also I'd like to look more at the sort of morphosyntax so what about other sort of temporal distinctions in some of the languages it seems that there's a distinction between past and non past versus degrees of past which might be more common so you want to look at all of these kind of things to see whether those are also contact features sort of aerial picture confuses things a little bit more so some of these might be instances of aerial grammaticalisation and then I've presented some kind of sociolinguistic and a little bit of the historical background but you really want to also have a much deeper understanding of how those things are playing out in the present day and have played out in the past so you want to look at what exactly characterises these patterns of bilingualism if these kind of changes are happening you expect high levels of bilingualism you expect it not just perhaps for one generation but you expect it on going there has been a shift from for example Alagwa and Burungae so the two Cushitic languages Cushitic contact languages for for Rangi meaning that lots of Rangi speakers today are ethnic Alagwa and Burungae or people are increasingly speaking Rangi as a second language although then perhaps their children would speak it as a first language so you've got those kind of patterns going on and then also in the context of Tanzania I mentioned this a little bit and in the context of East Africa more widely you also have increasing and ongoing pressure from Swahili so not only do you have all of these languages in the area but you also have a kind of much larger language of wider communication kind of mapped across the top of it as well which definitely adds to the picture and just sort of to bring it back to the questions that I was looking at at the beginning just to sort of say that all of this is done essentially with a view to better understand the process of language change and to understand the interplay between language contacts so contact and juice change and grammaticalisation and by no means am I suggesting that those two can't go hand in hand so you might find that grammaticalisation along this particular pathway has been catalyzed by the fact that there is a lot of language contact that there's high levels of bilingualism multilingualism and also again in terms of larger questions looking at the way in which structural transfer more generally but with a focus on these languages is delimited by universal constraints on language change and grammaticalisation so is it the sort of anything goes model versus no these things happen in this and these things can't happen so what kind of changes can take place and what kind of changes can't take place and I think that's me done Thanks Hannah very much for such an engaging talk on what appears to be quite a complex complicated field of research no doubt there will be some questions Thank you There's one big question which came straight to me which you didn't refer to at all and that is having established any kind of evidence for a genetic relationship between 200 clusters which would imply that all these processes if there were that all these processes might have taken place sometime in history in other words not be relevant to the particular groups which are there but to some predecessors of them so is there any genetic evidence of these clusters being part of one cluster at some time in history I think so you're absolutely right I didn't mention that so probably before all of that I should have said that the reason that I was showing the examples from Swahili and some of the wider Bantu languages is to point you in the direction that I don't think that this is a sort of genuine inheritance from Bantu inheritance is not a kind of longer standing inheritance from proto Bantu in terms of the link between the two what I called the two clusters a long long time ago yes but then if I was to some kind of map like this you have I don't know so the languages where I have the languages of the lake so the little ones up there say there's 12 languages there and I found it in four and even the neighbouring most closely related languages to superson beating and goremi do not have it on the Bantu side so what I've sort of done is said well it doesn't look like that because it's only the ones that seem to be in contact with ok that's slightly more complicated but in contact with a non-Bantu language or in contact with a Bantu language which is immediately in contact with a non-Bantu language in terms of the relationship between the two clusters themselves not really I mean their Bantu languages but there I think I don't know if Lutz has a comment to would they be close on the tree on the phylogenetic tree to each other or would they be quite different they would be quite far apart but they would not be as far apart as rangy and syswati or something so I don't think I don't think that's that's the story but no that's a really good that's what the whole thing hinges on really otherwise yes they've got it from somewhere else and all the other languages have lost it Phil so on imperatives I haven't tested but in the languages that I do know about so rangy and bugwe you wouldn't form an imperative with a compound construction so it would be sit and that would just be a single verb or incline or something as soon as it becomes a future tense or habitual tense that question wouldn't play out anymore and there are so there's a growing kind of account of lexical borrowing so there is a list of instances of presumed words of Cushitic origin in rangy Cushitic words in bugwe and also up in the great lakes area I think the thing that struck me most was that what I've done today is present the things that are unusual but in so many regards all these languages are so Bantu they're so pro-certificly of the language family and yes you do have instances of lexical borrowing you do have words especially ones that look like very old borrowing but you don't have many other things you don't have, I don't know, borrowing of prepositions and sort of things like that so that you would expect in between you've got some lexical stuff and then you've got what seems to be some really grammatical stuff Thank you for your presentation If pursuing the language context to explain this class of languages I wouldn't consider also another context, another level of context with the internal variation within the same language so that that came to my attention because I know within some languages which have been dominant languages in Masai and Llu you might have one that, for example, if I would see the southern dialect of Mogi language is apparently a lot more intimate by Masai or Llu than the northern one and in the case of Suba I think I think you have a certain Suba symbete which is the Tanzanian one and it's different from the Suba spoken in Kenya It's particular there, it's the Kenyan one but the two of them are inhibited by Llu Now, there will be other instances where you'll find, for example the north end of Cuyu has this particular feature of using the verb God to express the future to grammaticalize it into the future tense where the southern dialect does not and presumably the northern dialect has probably taken these features from the land of Masai so you may want to go on to proceedings looking at whether the data or the sample that you have to which of the variants will actually be good land that's certainly a really good point I mean, even with the Gusi and Curya so, for example, I had on the handout on Gusi half a million Gusi speakers have Curya as their first language and of Curya it's also spoken by about a quarter million people in Kenya and then again, as I sort of showed, Leon that would be in contact with different languages and I already know that there are dialectal variations within those so I think the point about actually making sure that I also know which dialect the data are from is certainly really important and then ideally I would like to get data from different dialects within the languages as well because in terms of the contact, that's vital because, well, this is actually much more in contact with language X I call it Curya we'll see in some of the mediating language because we can transfer from between a lot to get real contact I think it would also consider Lluia because within that zone I think it would be much more mediating than any other language than anyone else would for this numerical science The Lluia thing is also important and I think with the Maasai it's interesting because when I talked about, for example, Rangin and Bugwe, we're talking about people living in the same houses or neighbouring houses with Datoga and perhaps to a lesser extent Maasai these are also, the Datoga particularly are continually their nomadic people so the type of contact will also be different it's not necessarily seeing the same kind of I don't know family structures, the same kind of transfer to Sheldon and that would be interesting to look at up at the lake as well definitely Thank you I have a question about the Mlora about the valves about the valves with the valves it's worth around $7 and then you might not need to look at that as a case study I mean I don't know about the regions and quite a lot of the language but most of them have valves which is pretty difficult but I'm just wondering if you have anything to say about that Yeah, so I mean unfortunately you're absolutely right so that could be another one of my contact features potentially gone because the idea is that it was, these languages started out as seven valve systems and then some of them have lost them and have become five valve systems so actually maybe you want to say Ngurrame is a language which has that existence and I think people do research about nouns and verbs and actually that distinction is also not so odd in terms of what we like to change and what we don't like to change in which case, yes, that's not a contact feature the fact that it just so happens to be the case in Detogger which is also spoken again is chance and doesn't really give you evidence in support or against Yeah No, thank you It's really Lutz So in that sense I think it's interesting because it looks for an explanation because it seems very disharmonious in terms of basic word orders and then the other thing I was wondering about in that context is because this really seems to be playing with the basic head compliment relations so is there other evidence somewhere else in the language where this plays a role so you know that it's a language and it's a language where this plays a role so genitives and have nouns or adjectives or the monstratives all these sort of things because in principle it's a really nice case that you can see how how basic headedness the language can change and it has both the context and the grammatical information structure which then seems to result in a structure which doesn't have a rough trace which is weird Yeah, so in terms of head finality as far as I know no, like as far as I know the order of the noun demonstrative is as expected and possessive and all sorts of other things that I've come across it's certainly a good area to keep looking especially if I'm even looking for whether I'm looking for these constructions or I'm looking for the opposite because you want to see basically things behaving oddly and yeah, I mean in terms of the verb and the objects and the relationship of the two just because I know more about rangu in rangu nothing can intervene between the verb and the auxiliary with the exception of some kind of marker which shows direction so some kind of directional particle which actually seems to be more part of the verb than the sort of anything else in which case it does sort of show a strong phrasing together of the verb and the auxiliary and again as you've said the object being oddly left it would be interesting to see if in other languages anything can intervene between those especially because you already see on the verbs slightly different behaviours so you see on the verbs some kind of habitual marker one of them had some other kind of continuous marker whether those ones actually behave slightly differently from some of the language where it's literally an infinitive not even necessarily with any other kind of marking followed simply by the auxiliary but I know certainly look for more of those relationships Phil so in rangu there would definitely be a change in the word order in guisi there would also be a change in the word order in negative so the verb orcs the verb orcs so the verb orcs changes the order do you mean a different auxiliary being used or do you mean the order of the so the sequence the ordering of that changes in rangu changes in guisi and the other ones I don't yet know have you ever just mentioned that in any presentation the mixing is up a bit just to show that these are the factors we want because they're all the terms there's no exception and the last one focus must be with the only example that language with the verb orcs or the after Christmas six what happened when you missed them or were they discovered after Christmas had I missed them description has developed so I was reading a comparison of five of the Mara languages so I spoke another language and someone said oh it's really interesting one of these languages has an odd infinitive auxiliary order I don't have time to go into it but it's sort of stated description and then also looking so I was originally looking for other languages near rangu and then I found some up there well I hope I don't pick up any more really but yeah so my knowledge I'd missed them other people had not published what they'd written yet combination but I hope I don't think they've changed their word of the since in the last five years I hope not anyway it won't be a Christmas topic didn't feel like it though so thank you for coming and let's take thanks again to Hannah thank you very much