 good afternoon everyone and it's my great pleasure to welcome you to this instance of the linguistics department assembly. I'm trying to come to terms with restructuring and I'm not sure all of that matter either my name is Lutz Martin I'm at the moment the interim dean of the faculty of arts and humanities languages and cultures and I'm half the linguistics section half in the Africa section in the school of languages cultures and linguistics and it's my great pleasure to introduce our speaker this afternoon who is Daiz Kishinagawa who is visiting us from Tokyo University of Foreign Studies. He's here as a visiting professor for six months to March I think. And he works on African languages and Bantu languages. He is one of the world's leading experts on the Bantu languages spoken around Monti Majaro in Northern Tanzania. Daiz Kishinagawa is a PhD, let me just check that. In 2008 in Nagoya University and then taught for a couple of years at Kagawa University in the north of Japan and then went to Tokyo to Tokyo University of Foreign Studies two years ago to continue the long tradition of studies in African linguistics and Bantu linguistics. His topic this afternoon is on some topological characteristics and the group internal variation in Kiwimanjaro Bantu languages. Thank you very much. Thank you very much Lutz and first of all I would like to thank Lutz and all the members staff of the Department of Linguistics and I'm very proud to be here and I'm very proud to... I guess I'm the first person of the speaker from my university I think so in this part it's I'm very proud and also I'm very proud to be listed in the previous Bantuist speakers so yeah I mean this is what I want to say for the first. Now so I'm going to talk about their group internal linguistic diversity found in Kiwimanjaro Bantu languages from a micro typological viewpoint. First I will briefly introduce an overview of Kiwimanjaro Bantu languages and micro parametric approach to Bantu languages. Then I will show you some characteristic features from three major subgroups of Kiwimanjaro Bantu languages namely Rombo, West Kiwimanjaro and Central Kiwimanjaro. In section one it's a little bit hard to see but in section one mainly from the data of Rombo I will introduce some phonetic features which are relatively rare in Kiwimanjaro Bantu as well as in Bantu in general. In section two which will be the most lengthy part in this presentation I will talk about two topics. One is about an inflectional suffix which is aga which is traced back to Ploto Bantu stage seemingly dated far back in some five thousand years ago. I will explain a historical process which leads to its synchronic distribution and some typological correlations regarding its presence and absence. The other topic is on vowel lengthening suffixes which may tell us about historical contact with pre-existing languages spoken before present Bantu group came into the region. The last section deals with the Ulu language which belong to Central Kiwimanjaro languages which are most studied language group in Kiwimanjaro Bantu but still we can find some unrevealed phenomena which may shed light on historical processes in Kiwimanjaro Bantu and on close Bantu typological issues. Okay let's get into the introduction section. Kiwimanjaro Bantu languages which is spoken in this area and I just mentioned it as KB for short hereafter. A language cluster spoken around Kiwimanjaro which is in border area between northern Tanzania and the Kenyan and of course there and this is a Kiwimanjaro area and you can approach from Dar es Sala maybe you will take I don't know more than 10 hours by bus or something and from Nairobi as well you can approach there as it will take something like four or five hours and and of course the Bantu languages are affiliated in Benue Kongo branch of Niger Kongo language phylum and the KB is in its north eastern group. According to Gathri 1917's Zonal classification which is shown in the left hand side KBs are grouped into E60 and named Chaga group. Though the unitary name Chaga is standard ethnonym for the people living in the region, since the publication of Gerald Phillipson and the Marie-Laure Montaukou's comprehensive introduction the name of the language group became to be called Kiwimanjaro Bantu languages because some languages of clear linguistic affiliation with this group are also spoken by ethnically known Chaga people such as Roa which is spoken by Meru people living in the slopes of Mount Meru or Gueno on Paray mountains which is highly endangered and Dabida language which is spoken in Kenyan side. The languages mainly dealt in this presentation is highlighted in yellow Roa and Uru in central Kiwimanjaro and the Ku variation Blombo. And where is it? Yeah this is a geographical distribution as Derek Nass who is a pioneer of linguistic study of Kiwimanjaro Bantu languages along with Gerald Phillipson points out that and it is shown in the bottom side. The KBs show much wider linguistic variation than one would expect from a unitary name Chaga but unfortunately most of the published data of Chaga languages are from very limited languages or practically from one only one language which is called Wunjo it is of central Kiwimanjaro language. So in this talk I will present the data from other languages than Wunjo in order to show you I hope at least parts of such linguistic diversity yet to be known wisely. Before going to details discussion of languages specific data I would like to show a brief linguistic overview of Kiwimanjaro Bantu languages. Though the KB shows a wide linguistic variation there are two basic characteristics which would apply to all the varieties one is morphological complexity of verbal constituents and the other is tonal complexity. Since the former is a main topic of the following discussion I just show you a short example illustrating part of tonal complexity unlike tonal languages of Asia whose function of tone is mostly lexical like Chinese I mean the Chinese dialects that is the tone shows semantic distinction words in those languages. It is usual in African languages including Bantu that the tone shows grammatical function. Let's see the video clip which was taken with my teacher of Hulu language. Sorry for the noises. Did you find the difference? I asked you know so the first one is I see myself and the other is I see you right? It's like something the same but he said well you know there is of course the difference and the difference is marked by tone so the first one I see myself is something like that and this is I see you. Anyone knows the difference okay. I see myself is high low high high high where the second one is high low low high so you know the tone makes a difference. So yeah so this case show that the tonal difference denotes the difference of person of object that is one is reflexive myself and the other is second person singular and it can be said that it is still a kind of lexical contrast that is reflex marker cool has a lexical height on G cool on it so yeah actually this cool itself has a height on but this height on is realized on the following level so G cool and G cool warning is so anyway so the cool of this you know reflexive has a height on and the second one doesn't have a height on so this is kind of the lexical contrast as well however there are many more examples where a tone works to denote purely grammatical concepts such as subcategorization of past tense so I mean the farthest and nearest can be marked by tone solely or polarity I mean yes or no or close type marking is also what we are marked by some other Chagakimangya languages okay so and also I would like to mention briefly about the I mean the you know methods of my study. Next I want to believe mention the importance of my type of approach to ban two languages in the sense of general typology it is the large-scale typology or macro level typology that contributes to our knowledge of language inversals and one of the outstanding results would be for example where that was language structure was edited by hasperma dire gale and de bernard comley but isn't shown in this slide as Daniel 2011 claims there are important areas which are not sufficiently covered by macro typology in other words macro typology by definition is not aiming for deeper analysis of structural variation within genetic languages groups which must be equally essential to understand the language diversity and universal processes in this context there are two major approaches which aim to cover such fields namely intro a genetic typology which can be paraphrase and micro variation typology and aerial typology in source there is a great project working along this line of trend tied to the morphos intact variation in ban two typology context and change led by roots mutton as shown in the project aims this project has already set up 142 micro parameters in order to capture the structural variation within ban two and based on the parameters linguistic data including those from under described languages are collected and integrated in order to investigate the issues including ban two internal linguistic diversity and the process processes of historical change including those induced by language contact my research in Kimanjoban two languages shares their basic aims with this project okay so let's get into the detailed discussion the first time this is from Lombo let's move on to the sections of specific languages as I said in the introduction I'm intending to show the data which illustrates internal linguistic diversity found in three languages from each major subgroup of Kimanjoban two the first one is from Lombo and is about phonetic characteristics characteristics I just briefly present two relatively characteristic phenomena the first one is about the this one interdental lateral approximate which is something like a test is not only in Lombo but in Uru as well this sound seems the one the one we put it in em adhesion paper as labial lingual which is typologically rare in the world's languages listen listen to the sound but it's you know the sound quality is not good the recording is not clear so but in typical pronunciation the tip of the tongue slightly touches the upper lip, like this, and both Wombo and Uru, and this sound is realized as alphan of L sound, I mean phoneme L, followed by high-back bowels. The next one is postnatal trilling. Yes, in Wombo, dental and bilateral voice stops are rather regularly realized as trills, maybe you can easily pronounce the sound. Trill. And yeah, so let's listen carefully, you know, the words marked in words. So the second line, Veyanra, and the second one, akampra. Did you listen the sound? The first one is a little bit weak, Veyanra, but the second one is a lot clearer, akampra, is something like that. So yes, this sound is also reported in another Bantu language, which is called basa, but this language is spoken in Cameroon and seemingly no direct relation to Kirimanjoban II, and basa has only bilateral trills, while Wombo has bilateral and alveolar. So this is also, you know, somewhat rare characteristic of the phonetic level. Yes, and within Kirimanjoban II, this phenomena is only found in Wombo, and no other varieties have this kind of the sound. Okay, so this is the first part, and the second part is the data from Roa, which is from West Kirimanjoban. This is maybe a little bit lengthy, but yes, let's move on. Let's get on to the section 2. Here I will talk about morphosyntactic phenomena, limitedly found in West Kirimanjoban, mainly mentioning the Roa language. The first topic is about a verbal inflectional suffix, which is from aga, but in this language it's realized as a, which is reconstructed in Porto Bantu as aga, and its function is said to be various imperfective, typically habitual. Shown on this slide is one of the parameters set up by morphosyntactic variation in Bantu project, and it is about future tense marking. It says future tense reference, how is future time reference formally divided. The answer is different from language to language in Kirimanjoban II, but as shown in the table, West Kirimanjoban languages tend to have one only future, which is marked by aga, this suffix, while the others have two or more future distinctions. These basic facts have already been mentioned in the literature. For example, in nurse 2003, it points out that systematic difference of future category between Central Kirimanjoban and Lombo versus West Kirimanjoban. I mean, the Central Kirimanjoban-Lombo has two clear different futures, while Western Kirimanjoban only have one single future. While Phlipson and Mola who describes aga as a kind of isogloss, which divides WK and the rest. That means Western Kirimanjoban only have this aga suffix, and the others doesn't have. The question here is whether there is a logical correlation between presence or absence of the suffix aga and the number of subcategories of future. And my tentative answer is, I think it's possible. I mean, there must be a kind of logical relation. The key is simple. A is an element which is affixed in the final vowel slot. I mean, for example, in the first one, it is the first N is a kind of focus marker and LOO is a subject marker and MAN is a stem. And the following it is A. This is a suffix. But this suffix can be divided further into two parts, R plus R. First R is which we call final vowel, but it's not good name, but it's kind of the inflectional suffix. And second R is what we call post-final. So there are two elements. I mean, yeah, I wrote it R as a single unit, but this can be divided into two parts, R plus R. And yes. Yeah. And R is an element which is affixed in the final vowel slot. So that means that R is now, I mean, the R itself is slotted in the final vowel slot or inflectional suffix slot. And the slot is strictly closed for the limitist number of inflectional suffixes, such as default indicative subjunctive A and perfect past marker ILE and so on. And all of which are dated back to port ban 2 stage. This means that there is little possibility to introduce a newly grammaticalized morphine in the position. So the position, the final position is closed. So there's little possibility to get into new grammaticalized morphines. While in languages without R, the future tense is marked in prefix slot. For example, for example, for Luichi kappa, in this sentence, Eichi marks the future. And the right hand sound, Luichi kappa, Luichi kappa, in this sentence A, I mean, the prefix, I mean, just before the stem A marks the future. Yes. Whereas the latter open, I mean, the tense aspect marker slot, which is the second from the subject marker slot, was rather open for newly grammaticalized elements. This means that those languages have structural advantage to introduce newly grammaticalized tense markers. Hence, though it is needed to be verified by descriptive data from other languages, we would tentatively hypothesize that the correlation between the presence or absence of A in one hand and the number of subcategories of future on the other hand. Right. It's a little bit complicated, I think. But do you have any up to here if you have any questions or can I go ahead? Yeah. Okay. So I will go into more complicated place. Well, so another point I want to mention on Aga is a historical process which leads to the present distribution and to historical morphological split occurred in war. The first question that is why Aga is limited, limitedly found in rescue Manjaro is rather simple as mentioned in Mahiu on Bunjo. Next one. Yeah, yeah, yeah. And in Nelson Pipson 77 on Old Moshi, both of which are CK, Central Key Manjaro languages, it can be seen that those languages have lost the vowel length contrast in their phonological system. Hence, the long vowel R cannot be realized as it is. The same thing is applied in Rombo where the vowel length contrasts are neutralized in proposal position as illustrated in seven. So, for example, in the first one is Finua, which means trip off imperative form. But morphologically this is Finu plus R. But surface realization is not Finua, it's Finua. So, you know, the length is neutralized. But as clearly seen in seven B, if you put another element sentence finally, this U A part or the you know stem final vowel can be pronounced. I mean Finua, not Finua. So, this is, I mean, the kind of the neutralization of vowel lengths in terms of their position. Okay. So, while in Rombo, I mean while in Luar and other with Key Manjaro languages, the vowel length contrast is kept in phonological system as shown in this slide. Hence, suffix R can be realized. So, the lacking of vowel length contrast can be at least one of direct factors of Aga's disappearance in central Key Manjaro and Rombo. Then, the next process is about the systems of languages with R. Philips and Montreux make a general comment on R, which is read as R is a single morpheme and its function is to denote habitual in the future as is written in the, you know, upper side. But at least the fact in Luar is different in that A of future and A of habitual is clearly distinguished by tone. We can confirm like this. This is the future, this is the final part of the vowel is, you know, following the tone. But habitual, the tone is flat in the final position. Very subtle difference, but it's a future. But habitual, it's a very much more clearer if you see the difference of the following noun, I mean the initial syllable of the following noun. The first syllable is low, but the habitual sentence, the initial syllable of the noun is realized as a high tone. Yes, and there was that. On the other hand, you cover, but you know, so Philips and Montreux explains that this, I mean they are not aware, I mean at least, you know, the language they are working on doesn't have the difference. And also, on the other hand, Yukawa, 1989, claims on R of Mashami, which is also Western Kilimanjaro language, that there are two distinct R's, which is the same situation as in Roi, and they may have different origin. He said, if the ending R has originated in aga, the form dealt with here, which is present habitual, is semantically the true descendant of the form which had aga as ending, because aga input bound to stage has a kind of the habitual meaning. The author is not sure how the forms in two to one, two to three, I mean the future have come to have the meaning stated here, the ending R of those forms might have another origin. So Yukawa said, you know, the two different R has two different origins, but my explanation, explanation on this issue is slightly different from Yukawa's. The origin of two different R is still aga, or prot band 2, but it must have split in a certain point in history. Let me briefly explain. Yes, aga in proto Kilimanjaro band 2 stage was thought to be a kind of progressive marker, just as found in contemporary kivosho, which is shown here, lusoma, which means I'm reading, this is the kind of the progressive meaning. And I suppose that this form, or this meaning is of the something like proto Kilimanjaro band 2 stage, and the form changed from aga to aga, that is intervocalic G was developed, which is rather regular sound change in Western Kilimanjaro. So this is like this aga, it means continuous or progressive meaning, this is stage one. And this continuous progressive aga was shifting its semantic field to habitual in the future due to the introduction of a newly grammaticalized continuous marker, k, which is from a vowel stem meaning sit or stay. And as Haspelmat claims that this kind of semantic shift, I mean from present to, I mean from progressive to future, progressive habitual is some kind of university well-attested according to Haspelmat. So the stage two is invasion of k, which is newly grammaticalized marker, which means progressive. And because of this, aga was moving to outside the area, or in other words sifting to other semantic field. And stage three is the morphological split of aga, but how did it happen? Yeah. In Roa, there are two types of past marking template. One is for default active verbs, or verbs with perfective aspect. In this template, the verb can be inflicted three parted pasts, I mean the past one and past two and past three. This is a normal or default template or past marking. Yes. And in this template, the verb can be inflicted two parted pasts and there is no addition of final elements. While in the case of stative verbs, which are inflicted by suffix e, the first one, or other imperfective forms, the past tense is marked by prefix e plus lengthening of final vowel with a high tone. Please be noted that this template applies not only to predicates with verbal stem, but also to non-verbal stem as in existential construction, which is in term b. Ne4 is I am, I'm here, where its past form is Ne4, Ne4. And the stem is e4, but this is not verbal stem. This is a demonstrative stem. That is only, yes, that is only R of habitual was included in the paradigm of stative conjugation. And as a result, the tonological shape was changed to R. So there is only single R, but there was two different meanings. The future of R is placed in this conjugation pattern. And habitual, R of habitual is shifted into this paradigm. And it makes that future is nothing changed, but the R of habitual is when it's conjugated to past, it's got to be R, R, R, something like that. And the output goes like R, R. And this is, I mean, this is newly formed as a habitual. And this form is replicated to present as well. This is my explanation. So the process can be summarized as in this figure. First, the semantic field of R was shifted to future and habitual. And only the latter was shifted to a different morphological template for stative verbs. And the two were distinctively, let's say, morphologized. So this is my explanation. So this is a little bit complicated anyway. So the next one is the last one of this part, which is vowel copy suffix and the final vowel lengthening. The other topic of vowel inflection found limitedly in rescue manjaro is about the lengthened final vowel. As already mentioned, this marker occurs as a past marker of stative predicates in RWA and also found in SIHA and in another rescue manjaro languages. Though its phonological form is a lengthened vowel, morphologically it can be seen as a copied vowel of the inflection of suffix, which is nearly equivalent to what is called vowel copy suffix in band 2. Sorry. Yeah, this one. The vowel copy suffix is normally functioning as past marking element in other band 2 languages. But in past marking elements in band 2 general, what is most popular is TA marker R, prefix R, as shown in this slide. 1 and 2 times R is 78% of the languages in the database have a form of R for making past. And the second one is also the same R. And also it is used ILE. This is a second popular form for marking past. Okay. Yes. And vowel copy suffix is that this is the least frequent element as past marker in band 2. And this geographic distribution is splatic. And if you closely look at the examples, there is a clear difference between typical vowel copy suffix and the lengthened final vowel in rescue manjaro. So this is a kind of typical example of the vowel copy suffix. C laowa means I came. The stem is laowa means to come. If it conjugated into past, the final vowel is going to be R, which is the same as the stem vowel. And the second one is R and he or she went or has gone. The final vowel is the same one as the stem vowel. This is a typical copy vowel suffix. And this is also found in herero, for example. But this is also the same way. I mean, is it right? We taught. But it is the imperfective past or something. And this is also the same thing. And the final all is copied vowel to stem. Okay. Yeah. So that means that these are different from the case in rescue manjaro in that the target of vowel lengthening is final vowel and the lengthened vowel is placed in post-final position. However, the last example from Billa, which is so-called the fullest band 2 language, shows exactly the same structure as those found in rescue manjaro. So this is literally, I mean, almost the same structure as was structure. So the final vowel lengthings happens in the world final position. So the lengthened vowel is final vowel itself. So in other words, the copied vowel is in the post-final slot. So when it comes to Billa, so this is basically the same construction of rescue manjaro, especially Roa. And when it comes to Billa, this language is spoken not only by Bantu people of the Billa area, but also by so-called pygmy people living in the Congo forest. This is a list of the languages spoken by pygmy people. And one of which is buti sua. And also this is buti sua people are speaking Billa language. Yes. And this means that, yeah, so that means that, you know, pygmy people are speaking this Billa language. And this means that the pygmy people lost their original language and shift to use the language of their neighboring people, Billa. And at least to me it's a little bit interesting. In folk tales in Chaga people and in historical literature, like by Thomas Spear, it is said that the pygmy people who is called in Wakaningo in Chaga lived in the Kyumanjaro region before present Chagas came into the region. So, but this very small and slight evidence, but and also this is nothing but speculation so far, but there might be small traces of the language of the people, I mean of the language of pygmy people, which is now almost completely vanished. So that's a kind of speculation, but there may be a slight possibility to know about, you know, old pygmy, all the languages of pygmy is like. Okay. So the final part, I think. In Shui, I will talk about Uru language. And in this language, we can find some kind of archaic features in Kyumanjaro Bantu languages. The first one is about Aga, which I talked before. So finally, I will briefly pick up some features only limitedly found in Uru central Kyumanjaro language. All the features can be seen as an archaic traces. I mean, I would pick up three features, but all the features can be seen as an archaic traces which are almost lost in other central Kyumanjaro languages. The first one is an alleged residue of Aga, which we discussed in previous sections. I repeat again that as reported in Philip Sondland's Molehill, it is said that the relics of Aga are only found in West Kyumanjaro, but further research on the system of Uru shows there seems a residue of Aga, which is high-toned suffix A, and it is only used for present habitual, which is shown in the example in Leven, Lukaapa ikirili is meaning is we hit the tree, we hit this tree or something like that. I mean, we habitually hit the tree, but Luiecikapa is the past form. So past habitual doesn't have a kind of the height on the R, and actually H part shows the past habitual, but in present habitual, this is the final R shows a kind of habitual, and this may be residue of habitual R, which is not reported yet. Okay, so this is the one trace of, I mean, the Agaic trace of Central Kyumanjaro found in Uru. The second one is inflected subject marker. The next phenomena is what seems to be an inflection of subject marker, inflection of subject marker, which is verbal prefix that is bound morpheme. In 12a, kopila is me, I mean nim re mi, nim re mi. It's a very typical kopila sentence, someone needs a farmer or something. So there is no subject shown in the sentence, and it's possible, grammatical, nim re mi, someone needs a farmer. So this can mean that he is a farmer, or she is a farmer, or maybe I'm a farmer, it's also possible, nim re mi. It's a very typical and neutral kopila. Right, in 12a kopila is me, and it is used as present kopila for any class number or person of subject, or even with a subject NP, it is itself, NP itself, as in this example. But the second example, cool in 12b, is only used for second person singular, and this is actually homomorphic with its subject marker. That is in uru, subject marker can be used as an independent kopila, but this is not necessarily unusual in kb, and in the other brand too. For example, similar example found in archaic swahili varieties, so I read it a little long, but 12b is nim kun re mi. This is also the kopila plus farmer, but this one means you are a farmer, and this is not used for other person or other numbers, nim kun re mi. And this first cool is from, actually so I glossed under the cool, it is a subject marker, second singular, the cool itself is homomorphic with subject marker, second singular. So this means that subject marker itself can be used as a kopila in this language. But this itself is not so unfamiliar, right? So in Kimangera it is possible, but also I guess in some varieties of swahili it is possible to use subject marker as a kopila, but however what is a bit surprising is example 12c where the subject marker which is used as a kopila seems to be inflected by 10 smirking prefix. So the last example is kun re mi. Kun re mi is something familiar to me, I mean or something familiar to other bandwidths because you know the subject marker can be used as a kopila, but the southern is something strange I think. So kun is a subject marker, but e is a kind of the, it's a prefix, 10 smirking prefix. And it seems that you know this example is a kind of you know subject marker is inflected by the prefix, you know. So in this point this is a little bit surprising. This phenomenon has not been known at least in other major Kimangera ban 2 languages, but we can find similar examples in a grammatical sketch of Gueno which is a little known language of Kimangera ban 2 and classified outside the three major subgroups. So it is linguistically significant in that it will tell us some archaic features of Kimangera ban 2 which may have been lost in main body of the languages. So as Philippine Nass said I mean the Gueno itself is a little known ban 2 language of northern Tanzania and the Gueno is spoken by several thousand people in the northern Parle mountains which is not I mean different from the Kimangera mountains and all Gueno and probably have long been bilingual in ASU although the reverse is definitely not true. So the language itself is socially very weak and all of the others Gassry or Nassan Phippsen and Winter classify Gueno as a dialect of the Chaga group but it is of interest partly because it has long been geographically separate from the rest of Chaga. So Gueno has a kind of the very rare kind of phenomena and this Gueno language in this language as shown in Satin A and Satin B the first form Phuleke is analyzed as subject marker Phu which is first person plural and past marking prefix le and locative copula stem ke. So this is and the meaning is Phuleke numba is a we were at home so Phuleke part is meaning we were we plus copula plus past and plus some locative notion plus numba it's very clear and also Phuleke fwagwa it means we were buying so Phuleke part can be also used as a kind of the auxiliary introducing the progressive form. However in Satin C and Satin D copula ke is dropped but still well formed as a cooperative predicate inflected SM in Uru is exactly the same construction so C and D is in C is aghe so if it's aghe it's very clear and it's very you know natural sentence with copula but it's okay I mean in C you know the ke is omitted and it still it is well formed so this is a kind of what I said inflected SM in Uru so this is a basically the same construction so you know the what I said as an infected inflected SM which is very rare in Kimanjaro Bantu or even in Central Kimanjaro but this is you know found in the Gueno language and the last one negative particle the similar example that is situation where archaic feature in Gueno is only attested in Uru within three major groups is the system of negative particles most of current languages take an invariant negative particle as shown in this table so for example raw languages the negative particle is invariant all the time you know in D in D in D in D and for example in Rombo the same thing happens I mean the negative particle is all the way cool cool cool cool but in Uru it's a little bit strange and it's very interesting that you know basically four can be used all the person and number but partly you know partly it's possible to you know alternate other from for example first person singular it is possible to use four but also possible to use knee which is subject marker and third person singular as well four and all is possible the all is a kind of the it's a subject person subject marker it's not subject marker but it's a kind of the we call it self-standing pronoun but it's a kind of pronoun of the third person singular and the third person plural is the same war is a third person I mean the yeah it's a kind of the self-standing pronoun yes so however Uru I mean so Uru partially retains the seeming original system I mean if you see the Gueno system it's very you know systematic negative negative particle is I mean let's say agreeing with the subject person and the number so if the subject of the sentence is first person singular the negative particle is knee and if the subject is second person singular the negative particle is four something like that and Uru system is a kind of the you know Uru retains the kind of original system which is attested in Gueno where negation particle agrees with the person and number of sentence subject these features in addition to the case of aga seem to tell us that Uru retains some archaic features which seemingly had shared but currently be lost in other central key manager languages this means that the detailed investigation of Uru would shed light for bridging phenomena between central Kimanjaro and other groups which in turn may bring novel findings both for historical processes historical processes and typological micro variation the summary in summary what I intended to do in this presentation is to show how diverse linguistic features are found in Kimanjaro one two which in other words this language group is very suitable to investigate by the approach of microparametric typology which is developed by by Lutz and his project members and I hope I could demonstrate at least partly that the microparametric approach is also beneficial for issues on historical changes and finally I hope fine grained description of specific relatively rare phenomena makes the close bond to typology more precise and possible thank you very much for listening and maybe if I may start can you go back to the reflected subject marks it's something here yes because it's sort of curious what is inflected at you know it's a question of stem and aphix really because for the first ones you could think that that this is you know it's a copula which is inflected so in 12b you know I mean even a copula is sort of auto-inflect but it's better to inflect a copula than the subject mark I think and then I'm not sure about the c but if you go to the next next slide yes these the 13c and d not d 13c that's interesting because that then what now follows is inflected the word um ah yeah yeah yeah yes yes you know there's examples like that in southern Bantan guni languages where you have historically it's a bit like it was almost like john was dancing but it was an inflect at the dance inflect so it's a john he was he danced and then you did something like it's uh it's I think it's I am drinking I'm about to drink I'm not sure what the exact inflection is but so the in g is the subject marker first singular and the in g also like the subject looks like a subject marker for it so actually it should be in g and g not so not in g and g not so in g and g not so in g subject marker are the historic you know auxiliary whatever that has come to the inflected drink so in g not so it's just drink so what you have was I was or I am I drink and the whole thing gets together and then you have essentially you know this so that you can see it there's different stages the modern form is just in ganasa but you can see in older forms how it all was done together so this looks a bit like that but then the question is what is the year because the R is the subject marker the R is the subject marker and what you gloss gloss that 13c is past perfect or defective perfected I think so so if the analogy with Zula is right this ought to be an auto auxiliary yeah yeah yeah so certain fees yes actually so our gay parties function like an OZM but you know what is important is this certain D yeah so as you said you know the case is that you see we can see aga is a kind of the auxiliary and maybe gay can be re-analyzed as a stem of the auxiliary but so this one knee gay is you know knee is a cop yeah this one is subject mark and then also it is used as a copilot where you know to me it's a kindly inflected by yeah well you know as you compare the you know the for example certain B under the certain A I think to make a numerically straight copilot and at least according to Nassan Philipson and our Philipson Nass they think that you know fully this case this one knee gay is a kindly shortened form or form which is dropping their K could be a the gay and the lay be the same thing but there are some phonological change I'm just asking because in Kenyans there are knee inflected languages you get ory so you have rey which comes as an auxiliary form to be you get ory or you get a barry or you get agreement right right I mean the history how do what do you think so I mean historically that the lay in this case and the lay in Kenyans where he lay is maybe the same one I think but this I mean in this case it's not the same I mean the lay is a past marker and the Lee is actually you know the Lee of them past I mean the plot one to Lee can be used as or can be you know reconstructed as a copilot but that one is I think different thank you thank your master students wow okay this is great I think what we can play next time you come to a talk rather than just listening you listen with the view of what is my question going to be yeah and then at the end of the talk you have a question it's a useful way to engage with the talk and then it's more exciting as well so one question if I can move maybe to methodology you showed us where you had the interviews with you know you looked for tones can you talk a little bit more about these data coming from is this elicited are these texts is it from the literature do you find variation between what you found and what you found in the literature so so I mean it's very data rich but it's also I mean you have very detailed data how did you get that well so this is actually this is a not a recording session I mean this is a kind of the you know video shooting of the I mean the usual daily research setting where and but this is basically the station I mean there so how do you see how do you say I see something and how do you say I see myself how do you say something like that so this is basically the station and also to me it's a very big task to do very good research or be you know non-irritation way so to me you know it's I'm still struggling with how can I do to get a very good data of natural conversation yeah yeah yeah how does this thing from so actually so it depends on the it depends from direct to direct so I mean the in maybe in other languages or in other varieties you know so this may be blocked by the hiatus breaker but I think Swahili is I think or maybe Zanzibar Swahili takes such kind of you know hiatus breaker something like L I think if I you know apply this to the Zanzibar Swahili I think this may be finu yeah or something like that y or l to be inserted but in this language you know the hiatus breaker is not inserted it's not so what was the question I think yeah I mean this is a neutralization of length I think yeah yeah yeah length length neutralization can you go back to the four the negative particle in negative particle maybe the last one I think here here right yeah so here you have the choice from the first singular and the third singular and the third plural between two negative particles before correct yes and I guess the four is the one which was inherited in the language and then the needle and the both were enomated or were they borrowed I don't know if they were borrowed could it be from Wenno are they losing out so that but as far as I know Wenno is very beautiful it's it's an endangered language so it seems hard to think that we would have borrowed and such a thing I mean how how would you explain that how would you explain the fact that it's restricted to only third persons the first and the first singular and the first plural and and how is it is it freely interchangeable is one taking over is the free self-standing pronoun taking over the four one for these three persons and can you yeah yeah yeah thank you yeah actually so what I'm thinking is that you know the original system would be that of Wenno I think so all the negative particles are you know agreeing with certain subjects so this is a kind of Uru case or Uru system is the kind of the let's say the residue of the original system where I don't know but yeah yeah of course you can say that you know this is a result of the language contact but this is you know what is I mean in the same geological and the genealogical group so this I mean I need to say this is result of the contact I think it's rather you know they're sharing the same origin and you know Wenno is following the original one but Uru is you know retaining something but it's not complete system and actually what about the before I say it was inherited but it's probably a grammatization process yeah yeah yeah do you know what's the origin I mean the original whole yeah how did you this is the point I mean so yeah actually so this is this building a bit I mean in the paper of Modbigo's and of the era and they mentioned about this phenomena and they're not they don't have clear answer you know what is the original for where what was my answer I think yeah actually this one in about Wenno they are wondering you know for where the voice coming from so one possibility is from class 17 locative and the other possibility is second person singular cool and then you know they're wondering which one is correct but I guess at least in Uru for maybe from second person singular I think and because you know if you closely look at the system or paradigm of the pronouns I think you know that tells you the fall maybe from second person singular but what I'm thinking and what as long as I know this case Wenno's case maybe from class 17 so there is you know discrepancy between the two so to me this is what I know I'm wondering if there's anything kind of just personally going on with the markers I mean would there be any way would there be any way to see I mean in Gueno I mean they're they're they're very much like they're much less regular it would be interesting to see I mean is Gueno do they use it more and whereas like a language like raw and cool would use these would use these markers less do you know where I'm going with this word where you know that these these final markers could be kind of on their way out and the negative marking actually on the left over before left over you know rather than I guess in comparison to Gueno where you actually have this this kind of full I suppose inflected set yes you know so so on the on the verb itself do you do you get do you get negative markers at the only marking ah yes actually so yes in kiban joban 2 I think there is no language which has pre-initial negative marker so but it has a kind of the post post-initial negative marker but they are used only for I mean no main closes so in main closes you know the negative marking is only by the you know sentence final particles this is all languages yes you know it seems to be that the locative is much more plausible one so I'm a second person singular is a grammaticalized form for negation doesn't I mean I don't think I've come across it was with the location one that's it like the French I mean step but there's no market yeah there's a space element in so I think that's that you know to me that seems more plausible I don't need other any other questions and in that case I think we are probably going to have drinks at the Institute of Education in about 10 15 minutes time I think that five would be and I guess please anybody and then we maybe have some folks