 know. So I'm Joey Love-Strand. I'm a postdoctoral fellow at SOAS and I'm happy to be able to host these webinars today. So today we have with us Makoto Furumoto who is a research associate at the Institute for Languages and Cultures of Asian Africa. He's part of Tokyo University Foreign Studies or Tufts where he's been working on Swahili dialects in Tanzania. He's going to be presenting to us today some of the results of his research. So Makoto, thank you for be willing to join us even though it's late in the evening there in Japan, even though you're missing getting to watch part of the Japan-Germany World Cup game that's happening right now. Sorry, we didn't coordinate the schedules a little better in the beginning, but we look forward to hearing what you have to do to share, but your research and be able to discuss that with you. So thank you very much for being here. Okay, Joey, thank you for the introduction and let me share my screen first. And as I checked before, I hope you are not seeing my presenter view. Looks good. Okay, perfect. Okay, let me start. And as I said, I'm Makoto Furumoto as Joey introduced, and I belong to the Institute called Ilka, and Joey knew what this actually name of the Institute, but I can't explain it in English. I belong to some Institute of Tokyo University of Foreign Studies, and I'm very happy to give a talk here today, and I'm very happy for you to come here. And maybe if I understand correctly, maybe this is my first time to see everybody, every participant, if I understand correctly. Okay, and as Joey introduced, I'm doing my field work in Zanzibar, Tanzania, and working on the description of local Swahili dialects. And this Kappa photo which I took Zanzibar is a photo of a local festival, and this festival is probably derived from the Persian New Year celebration, I think, and it's which called and no rules. And in this festival, as you can see, people fight with each other using stems of banana leaves. Okay, anyway, so I'm starting my talk about Swahili dialects of Zanzibar. Okay. So, first of all, I would like to share some basic information about the Swahili dialects I'm working on. And my field Zanzibar is a character in Tanzania, including two main islands. One is called Unguja, and the other one is called Pemba, and this is Unguja, and this is Pemba. And a local dialect of Unguja Island called Qunguja was used as a language of trade in the 80s and 90th century and spread widely across East Africa, and this dialect forms a basis of Southern Swahili. And in this talk, I would like to focus on not on this dialect, but on other Swahili dialects, which called Kimakun Ruchi and Kitumbatu. And these dialects also spoken on Unguja Island, this island. And in some previous studies, these two dialects called as a rural dialect of Zanzibar. So I also following this, as I said, terminology. And on Pemba, on this island, there is a different dialect called Tipenba, and Tipenba is not focused on this talk, but I will show some examples of Tipenba as well. And Kimakun Ruchi is spoken in the Southeast part of the Unguja Island, so like these areas. And the Tipenba is primarily spoken on the small island called Timbada, this is Timbada Island. And the Timakun Ruchi and Tipenba, Tipenba, are called Hikai and Hikai respectively by the speakers. And in some literature, it's been noted that the word Kae, the Makun Ruchi's name, is derived from the word Kae, which means archaic. But I consider this etymology, the etymology is not correct. And probably, as some authors have suggested, Kae and Kae can be traced back to a Proto-Bantu Kae, which means home or village. And here I'm showing some lexical differences between the Unguja and the two rural dialects of Sohiri. And basically, I'm using the orthography of standard Sohiri, but some IPA letters are also used to transcribe some peculiar sounds. For example, the initial letter of the Tipenba, which means come out. So its initial letter represents a paratilateral approximate. And this sound is peculiar to Tipenba, and as far as I know, it's not being described as proprietary by other authors. And the Makun Ruchi and Tipenba, the verbs in A, lacks their cognate in the Unguja, but they can be traced back to Proto-Bantu. So it's likely that Timakun Ruchi and Tipenba share part of their vocabulary, not with the Unguja, but with other Bantu languages. And as for words in B, I would like to use the difference in terms of length, that the Unguja words are dissyllabic, while most of the Timakun Ruchi and Tipenba correspondents are monosyllabic. More generally, the Unguja differs from Timakun Ruchi and Tipenba in that it lacks monosyllabic independent words. And probably this restriction of length of the Unguja words can be attributed to stress, which is obligatory assigned on the penultimate syllable. And if, as this generalization is correct, it can be expected that Timakun Ruchi and Tipenba do lack penultimate stress. And for Timakun Ruchi in particular, this expectation is likely appropriate because it is well compatible with the results of acoustic analysis, which I carried out together with my colleague, Yasunori Takahashi. And the prosodic feature of Timbantu is still unclear, and I'm wondering how I can correct it for acoustic analysis in Timbantu. And because of the sociolinguistic situation in Timbantu, it's more difficult to collect data in Timbantu than in Makun Ruchi. And as you already confirmed, Timakun Ruchi and Tipenba appears to be grouped together and can be differentiated from the Unguja. More generally, assuming a broad division of the coastal Swahili dialects into northern and southern dialects, Timakun Ruchi and Tipenba do are categorized as southern Swahili. And according to Derek Nass, the Unguja is also categorized as southern Swahili, but it resembles northern Swahili, which is spoken like coastal areas of northern Kenya or Somalia. And probably the ambivalent classificatory status of the Unguja is probably attributed double to immigrants who came to Zanzibar from Lam in the 19th century. And conceivably, Timakun Ruchi and Tipenba do emerged on Unguja much earlier than Tipenba. And at present, Timakun Ruchi and Tipenba do are under some pressure from the Unguja. However, it should be noted that Timakun Ruchi is still actively used across the generations. And in contrast, the situation in Timbatu is more complicated. And in my impression, in Timbatu, men tend to mix the local variety with the Unguja, the more prestigious variety, while some women only speak Timbatu. And so this is applicable to the women and men in this picture as well. And this gap between men and women is probably due to the difference of mobility. On Timbatu, men often go far to the sea for fishing or trade, and most women are engaged in farming and stay in the island. And in this talk, I would like to introduce two kinds of my findings. First, I will talk about language contact between social dialects. And second, I will show two traces of grammaticalization which appear in Timakun Ruchi. And the man in this picture wearing the yellow shirt is a Timakun Ruchi speaker who provides the Timakun Ruchi data I will show today. Okay, so firstly, I would like to talk about language contact or dialect contact within Swahili. So concretely, I would like to talk about the observation which suggests the blowing of a tense aspect marker into Timbatu. So I prepared some examples to introduce basic information of tense aspect marking in Timbatu. These examples are all from tales narrated by a Timakun Ruchi native speaker, and they show that Timbatu has several ways to express anterior situations. For example, past tense can be encoded with the prefix li as in example one. And while this prefix has been borrowed from Yunguja, this is not what I want to talk about today. And example two and three show that for perfect and completed, the prefix is ma and sha used, respectively. And when perfective is encoded, babes lacks term prefix and the same vowel as the penultimate syllable of the stem appears as a final vowel. And this morphological pattern is called vowel copy, and this is shown in example four. Yes, it's four. And in this example, the penultimate vowel a is copied as the final vowel. So in general, babes end with the final vowel as in one and two, but when vowel copy is applied, babes can be end with other vowels like a in four. And I realized this generalization is not applicable to babes like for harmony in three. So this is because this bab is borrowed from Arabic and some borrowed babs is not subject to any as a final vowel alternation. Here, I would like you to recall that Tumbatu is a small island. And on Tumbatu, there are two villages. One is called Gomani and the other is called Djongore. And two villages are linked by a tiny road which can be seen in this picture. And while the languages in Gomani and Djongore are very similar, but there is a variation. And for example, in Djongore, the prefix N, which is the homo organic syllabic nasal is used for anterior situation in addition to the other strategies which I showed in the previous slide. So in five, six, so the prefix N appears. And maybe this prefix is used for to encode perfect if I think. And remarkably, the syllabic nasal is used in a similar way in Kippenba. And this is shown in eight. And more precisely, Kippenba uses a prefix N, or perfective, and this prefix can be reduced to the syllabic nasal in certain environments. So in some, my findings is that use of the syllabic nasal in Tumbatu and Kippenba, and I consider that the Tumbatu has borrowed this prefix, which only appears in the variety of Djongore from Kippenba. And I'm wondering this borrowing correspond to the movement of Tumbatu people, especially men who often go outside the iron. And I know a little about a similar case of the similar borrowing with the exception, but with the exception of the prefix male, which came into Timakunduji from Northern Swahili. And now I would like to move to the next topic. And for grammaticalization, I would like to talk about two traces which is observed in Timakunduji. One is a tense marker derived from the verb meaning come. And the other one is a bound pronoun derived from the demonstrative. And in Timakunduji, the verb ja, which means come, can encode future tense in some context. And nine a is a typical, where expression, and it's a typical example in which jar marks future together with the imperfect prefix now. So, and as you can be seeing nine B, the future marker chart can appear in the same context. So, how's it, this, so in other words, so nine B confirms the, I'll say, suggest that jar in nine a is used for future reference. And what is remarkable is that this verb can encode a future in isolation. And based on the cross-flaring stick observation, it's been proposed that an imperfective component is essential for the grammatization of motion verbs into future markers. However, as can be seen in 10 or 10 B, in Timakunduji, jar in isolation can encode the future. So, Timakunduji just appears to cast out on the hypothesis about the grammaticalization of motion verbs into future markers. And what is more remarkable is that Timakunduji jar covers past in addition to future. Past marking by jar can be seen, example 11. And it should be noted that when jar is related to encode past tense, it can be followed by an unprefixed stem of other verbs as in 11 B. And additionally, it can be followed by the constitutive form of other verbs as in 11 B prime. And the formal difference of the following verbs correspond to the aspectual difference. When followed by an unprefixed stem, past perfective is encoded. And when followed by a consecutive verb, past imperfective is encoded. And the last example in which jar is disjointed from a register verb is not worthy because it's used for imperfective situations even though it's accompanied by the perfective prefix. And as the perfective prefix never marks the perfective aspect in this construction, it appears that the perfective marker has lost the original aspectual function and undergone functional change to the past tense marker with jar. So, Timakunduji jar, which means come maybe typologically rare because it's de-grammaticalized to cover both future and past. So in my understanding at risk, there is no similar cases in other languages. And so I think this finding may be significant, but I was more interested in what I wanted to know more about whether two different functions are synchronically or diacronically related to each other. And I think a key observation is that jar makes future reference without recourse to any support of co-occurring term markers such as imperfective now or conditional marker or subjunctive marker. And in contrast to this, when it works to mark past, it obligatory requires perfective markers. And based on these observations, I concluded that future and past uses of jar can be located at different stages of the same grammaticalization processes which is summarized here. Precisely, I consider that the first step is just the change from a motion bar to a future marker which marks future in isolation. And after this step, jar has acquired a necessary but not sufficient feature for past reference as a result of reanalysis of its future denoting function. And because of this how I said, insufficient function, it requires the support of perfective markers. And in the final stage, jar and the co-occurring perfective marker lose their semantic compositionality and then they are used. And because of this lose, they are used for past imperfect situations. Okay, now this is, I'd like to move to the last topic. And the last topic is demonstrative in Kimakunduchi. And as can be seen this table, Kimakunduchi exhibits a three, four distinctions of demonstrative. That is, there are protissimal, medial, and distal demonstratives. And the protissimal and the medial have monosyllabic contracted forms in addition to distalabic basic forms. And Kimakunduchi contracted demonstratives differ from uncontracted basic forms in that they can only be used pro nominary, but not add nominary. So in other words, contracted demonstratives cannot modify nouns, cannot modify nouns. So this is shown in 13. So basic forms like you know, can appear after the head noun and modify noun. But contracted demonstrative like you cannot appear in the same context. And notably, contracted demonstratives can be co-referential with playable noun phrases and contracted demonstratives can index any functions and semantic rules. So maybe, oh, yeah, yeah. So, for example, so in this example, the contracted demonstratives correspond to the playable object, which means that story. And yes, and what is important is that the reference of contracted demonstratives are obligatory at topics. And in most BAMP2 languages, verbs are accompanied by agreement markers referring to the subject or object. And the way of object indexing varies among BAMP2 languages. In some languages, such as Kimakunduchi, verbs can force the object marker only in the pre-stem position, while other languages allow it to appear after the verb. Example 15 shows that in the Shimawari dialect of Komuryan, the recipient and the tame object of the diterrancy verb, which means div, can be indexed through the pre- and post-stem markers, respectively. And notably, such a post-stem object marker appears similar to Kimakunduchi contracted demonstratives. And, furthermore, it's been proposed that demonstratives can gradually develop into anaphoric pronouns referring to topics and eventually become pronominal or agreement markers. And against this background, I was wondering that Kimakunduchi contracted demonstratives can be located at an initial stage of the development into a post-stem object marker. And Kimakunduchi is showing the kind of missing in between stages of the grammaticalization of post-stem object markers. Okay, let me conclude my talk. So, in this talk, I introduced part of the linguistic features of Kimakunduchi and Tsukunbaru, which are being summarized as a well-directed Zanzibar. And I'm not sure if you enjoyed my talk, but I'm happy if you found that there is some interesting linguistic variation in Zanzibar. And as for the grammaticalization I showed today, I already written papers. So, if you want to know more about them, please get in touch with me. So, that's it. Thank you for your patience. That's great. Thank you very much, Makoto. We've got plenty of time for some discussions and questions and responses. I know we have a few, so I hear these speakers among the attendees today. So, we look forward to hearing your comments as well. If you want to ask a question, you can either use the raise hand function in Zoom and I'll give you access to the microphone to ask your question. Or you can put a question in the chat and I can read it out for you. There was already one question that popped up from Nina Watson while you were discussing and she's wondering if you could say more about the different locations where you're collecting data. You mentioned it was more difficult to collect data on the island of Tombattu. Could you say a bit more about that situation and what made it difficult? Actually, there are several reasons and one reason is that, how to say, it's one reason is related, how can I say, so mindset of the Tombattu people. So, how to say, actually, Tombattu is a very closed place. So, how to say, and it's very difficult for people from outside to communicate with some local people, especially women. So, this is one reason. And secondly, so, for example, in Makundoch people can, how to say, switch the, so this Makundoch people can speak both to Makundoch and to Yunguja and they can switch these two varieties consciously. In contrast, on Tombattu, it's difficult for people to switch languages. So, and if I ask them to pronounce some nouns or verbs, they tend to pronounce the words in a more standard manner, not in the natural way. This is the second reason. I hope you got the reason. Yeah, so just some complex diglossia and language, making it difficult to elicit anything. Lutz Martin is here. I would like to ask a question. Do you want to go ahead with your question, Lutz? Thank you. Brilliant. Thank you. Thank you, Makoto. Really nice talk. Really interesting data. I really enjoyed it. And I mean, one of the things which is really quite striking about it is what you have on your last slide as well, the link between language contacts and grammaticalization. And I was really curious, you know, what your experience is about trying to find out what the relationship between the two is, because it's so difficult sometimes to disentangle what is so similar. So, I mean, one example, you had with it, yeah. Because you're right, it's quite striking that it's both past and future reference, but of course, the past reference is more widely spread. So, first of all, I have to note that maybe the past reference you are thinking is different from the past reference which appear in Makundoichi. So, in Makundoichi, in addition to the past use I showed today, it's used as a kind of perfect marker in a similar way to, like, for example, to Yunguja. So, maybe so, I think Makundoichi, the situation of Makundoichi should be different from, I would say, the situation in other dialects of Swahili, I think. And so, were you able to find, you know, particular motivations for claiming that something was grammaticalized internal to the language versus being borrowed through language contact? Were there clear evidence that you used to make that distinction, or did you have to kind of guess based on other patterns? Actually, so, as for the at risk currently, I'm thinking that the development of our functional change over here in Makundoichi is independent of language contact. So, it's not influenced by any other language. That's my opinion, but I'm not sure. Yeah, okay. Good. Thank you. Thank you, Lutz, as well, for the question. There's another question in the chat from Amina Kamis, I think, looking for some clarification about one of the examples. Do you have an example with the word Kemenapika? And you're wondering about tense markers me and na in this word. Do you know what that example is? Which one? I cannot see the comment. It should be in the chat. And it's about Kemenapika, is the word, and says, looks like there's a tense marker, me and na. Kemenapika, so, I can imagine the example, but I think this example, I didn't show this example in this, right? Yeah, Amina, maybe if you know the slide, that would be helpful to find out on the slide, but I can say something about this, because, how can I say so? Actually, so, in both Makundoji and Githun Bhattu, the prefix, the perfect prefix, ma or me, can be combined with the prefix, imperfect prefix, na. And I'm, what I should say, and I'm thinking that, so, I'm not sure how I can, how to say, analyze what segment these two prefixes semantically, but I think they are, how to say, they very look like, the English construction, which consists of a present perfect and progressive, like I've been working on, so, and I'm thinking there are some semantic link between Swahili dialect and English, that's what I'm thinking. So they're imperfective and perfect, but they are perfective and they combine for some kind of complex tense aspect. We've got some time for more questions, so feel free to raise your hand and put more questions in the chat. In the meantime, can I ask you about something on slide 19, where you showed your path of grammaticalization, and maybe I just missed this, but you say at the end that there's a set about stage four, that there's some evidence that these are now, that the perfective market plus jaw has become non-compositional. Could you speak more about what the evidence is that that's become? Ah, yeah, yeah, yeah. Actually, so this stage is, so, so in this talk, I didn't show any, how to say, an example of which suggests the imperfective use of, or past imperfective use of the deconstruction, but actually, when a perfective marker and jaw is followed by the consecutive verb as in this, how to say, 11 b prime, it expresses imperfective situations, but formally, it's accompanied by the perfective prefix like, so there is a gap between form and function, or form and aspect, and I'm thinking this construction can be located at the final stage, stage four. So in this stage, so the perfective prefix has completely lost its original aspects or aspectual function, that's what I'm thinking. Okay, so, but then in terms of the phenology, morphology, there's not any difference yet, there hasn't been any change in this one. So apparently, it appears the same as a stage three and a stage four, but semantically, I think there are a difference. And you wouldn't try to analyze jaw as like a non-present marker to try to do some sort of complex, you know, semantic analysis saying it's either past or future, but not present. Actually, so when I made these examples of the grammaticalization of this verb, I was wondering in a similar way to you, but I found, and I found for future reference, jaw works in isolation, and then I changed my analysis and actually, so non-present function can be applicable only to past reference as in stage three. I think I haven't answered your question. Well, yeah, I mean, I think it's, I'm not sure, I'm not sure what the evidence would be for for either analysis, you know, analyzing. Ah, okay, so one point, an important point is that when jaw is used for past reference, it's always co-occur with perfected markers, but not with other tense marker, tense aspect markers, that's why I'm thinking like this. So I saw, and because of this, I'm thinking so the level of non-speech time may be not appropriate, but I'm thinking, how's it, I'm thinking jaw works together with co-occurring perfected markers. That's how I said central point of my proposal, my analysis. Okay, so then jaw does have this abstract meaning and is the perfected marker that clarifies when it's past. I have another question from Amina. Amina, would you like to ask your question if I unmute you? Okay, thanks. Thanks, Amina. Okay, so I, I know this example as well, but and wait. So, so just for, for those who are watching, maybe haven't seen the chat, Camina mentions that in another language, Kimakunduchi, they use the word hebu in the negative word. As an example, hebu usunkeavo don't run away. Yeah, yeah, yeah. If you could comment on this. For this comment, I'm also seeing and I know this use of hebu, but And for those who don't know, what's the etymology of hebu, or what does this word mean elsewhere? For etymology, it's, okay, so I was thinking about this, but I forgot to I was thinking at the moment, so wait. Actually, so in Makunduchi Watumbato, in other dialect, maybe, there is the verb which is formally similar to this negative word, but I'm not sure if there is a etymological link between them. So, and I can't say anything more about this. Well, I don't see any more questions or comments, so maybe we'll leave it there for now. And of course, Makoto already shared his email address for anybody who wants to reach out and discuss more or see more of the data. It'd be great to continue that discussion in another platform. Thank you to everyone for joining us in any part of the discussion. Thanks again, Makoto, for this research and sharing it with us today. Glad you're able to join us, and I hope you enjoy the rest of the World Cup game. Yes, thank you. Thank you very much. Okay, thank you everyone.