 Okay, let's get on the way. Welcome everybody and this is our usual weekly linguistics department seminar and we're very happy today to have one of our very own to be the speaker. Just realised a few minutes ago that it's our 13th anniversary this year, are you sure? So as you came to this office, in 2005 to do a PhD... Masters in PhD. Masters and then PhD. Masters and then the PhD, so postgraduate study. Which she finished in, she finished the postgraduate studies in about 2010. Her research topic was a study of Berber languages from North Africa. SOAS has actually been a centre for the study of Berber for quite a number of years. Actually going back way into prehistory, you probably know that it was... What was it? You were the fountain of old memory. Atel Yad. Jim, Jim, Jim. Yeah, he was a Berber. Exactly, we had quite a flourishing area. He was a big specialist in it and then it went quiet for a little while and then we've had Aisha here and Simona Mali, Lameen Soak. So it's wonderful to see that this group of Afro-Asiatic languages is well represented. So Aisha has been on, is a permanent member of staff and has been here since in that capacity for... In 2012, I know a number of you have already had classes with her or taken classes with her. It's a real stalwart for the teaching of particularly syntax and also introductory linguistics. So today she's going to be talking about her research and with a focus on the roles of motion verbs in expressing discourse functions. With a focus on her favourite language, which is Berber again. So over to you Aisha. Thank you. Yes, so I'm going to be talking about a particular grammaticalization path, one that takes motion verbs and transforms them into markers of narrative discourse. And I'm going to be focusing on the verb Uhal in Saqbalit, which is a Berber language and like all Berber languages, it's an Afro-Asiatic language. And this one is spoken in northeastern Algeria. So the verb Uhal in Saqbalit canonically means to return or to go back as illustrated in one. And in some context, it can also have a sequential meaning or it is translated by the speakers with a sequential meaning. So she cut his nevel string and then she bondaged the boy. So the verb Uhal is translated and then. So I'm going to be talking about that and this is really much research in progress. So I have a lot of questions and not a lot of definite answers yet. And everything that I'm going to be saying is based on studies of my narratives. I haven't done elicitation on the discourse function of this verb. So it's just based on the narratives. So my aims today are going to show that Uhal has grammaticalized into a narrative discourse marker. So a type of discourse marker whose main functions are to link sequential units of discourse in narratives, but also to show a speaker's opinion on how these units of discourse relate to each other and particularly to mark new units of discourse or units of discourse that contrast with previous units of discourse. And I will suggest a possible grammaticalization pathway for the verb intact by it, which has not been reported before, well, as far as I know. And this pathway I'm going to say is that the verb encodes initial emotion and then has grammaticalized into a marker of change of state. And then from the change of state meaning or the change of state construction, it has further grammaticalized into a marker of narrative discourse. So my talk is organized as follows. I will be giving a definition, very briefly, definition of grammaticalization. I'll describe some of the main characteristics of grammaticalization. Then I will talk about the grammaticalization or the development of discourse markers. I will give a very brief, again, typological overview of the motion to discourse marking grammaticalization path. And then I'll focus on the verb oral intact by it. I'll give an overview of the verb. I'll discuss the discourse and morphosyntactic properties of grammaticalized oral. And then I'll talk about some possible source constructions. So grammaticalization is defined by Hopper and Trogott and Trogott in various places as the process whereby lexical material in highly constrained pragmatic and morphosyntactic context is assigned grammatical function. And once grammatical is assigned increasingly grammatical operator like function. So grammaticalization, although we often we speak of grammaticalization of items, most of the time grammaticalization actually involve a particular morphosyntactic construction. So an item grammaticalizes from a particular morphosyntactic construction, sorry. Grammaticalization, so an item as it grammaticalizes, yes, may undergo a number of changes. One change that often occurs, almost always occur, is de-symenticization, also called semantic bleaching. So the item loses some of its lexical meaning and of course it gets a more grammatical meaning. An item may also undergo phonological reduction, also called erosion. It may undergo various degrees of de-categorization. So an item loses some of the characteristic properties of the category to which it belongs. It may be undergo context extension, so it may be used in more contexts. It may lose its syntactic freedom and the construction which involves the item undergoing grammaticalization may undergo reanalysis. So the construction may change underlyingly and these changes are not necessarily visible on the surface. So say that we have a source construction which involves two closes. This construction may be reanalyzed into one close, so two closes become one close. And of course it's not necessarily visible on the surface. So we can have an example of grammaticalization involving some of these changes. So the example is from Heinz 2014 and it involves the grammaticalization of the verb to use as in 3A. He used all the money into a habitual marker used to as in 3B. He used to come on Tuesdays. So grammaticalized used to has undergone de-cementicization. It's a very difficult word to say for me. So it has lost some of its lexical meaning. It has acquired a more grammatical meaning and aspectual one. It has undergone de-categorization. So it has lost the ability that all verbs in English have to be inflected. It's used to can be phonetically reduced and it's used in more contexts. So used to is now, no, it's not used in more contexts but it's used in other contexts. So used to takes verbal compliments rather than nominal compliments. Yes, so because you are is a motion verb and we're talking about the grammaticalization of motion verb. I thought I ought to say a little bit about the common grammaticalization pathways for motion verb. So it's very common, motion verbs are very common sources of grammaticalization cross linguistically. And they grammaticalize mostly into markers of tense, aspect, mood or modality. So very often we have verbs like go and come, developing temporal meanings. So they mark past, future, etc., change of state as well. And they can also very often grammaticalize to mark close level relations. So they grammaticalize into ad positions, marking path semantics, case markers and very often they also grammaticalize into purpose markers, especially go and come. And well more rarely cross linguistically on the cross linguistic scale, motion verbs can also grammaticalize into discourse markers. So a lot of the time it's the deictic motion verbs that undergo these grammaticalizations. So the verbs like go and come. Some other verbs like return, intact by elite, it also occurs in oceanic languages. And verbs meaning reach or proceed in some languages. These verbs grammaticalize to become narrative discourse markers. Or they are also sometimes called textual connectives, consecutive or discourse connectors. But I mean these terms mean exactly the same thing. This grammaticalization path is very common in African languages. So it's been reported in all linguistic fill out of the African continent. It's found more sporadically in other areas and in other linguistic families. So it's found in Somo-Sanic languages, a couple of Turkey, Dravidian and one Romans language. It's a Spanish language from South America. I can't remember which one. Let's briefly look at what discourse markers are. So discourse markers are generally adverbs. So things like indeed, well, after in English. They don't have to be just adverbs, but mostly they are. They are defined by Trogot as categories which have their own syntactic slots and their own intonational properties. Their function is to mark the speaker's evaluation of relationships between sequential units of discourse, particularly between the current utterance and some prior discourse. The current utterance is minimally one sentence but it can be a longer stretch of text. And the prior discourse can consist of an actual utterance or can be something that is contextually reconstructed. So an inference or a presupposition. Narrative discourse markers are a subtype of discourse markers. They are found mostly in narratives, hence the name. They mainly signal, they have a particular function, mainly they signal that there's a new event coming. So they modify an event that is new in the story. And Hein 2000 says that we can paraphrase their semantics as something like, watch out, now something new is going to happen that is relevant for what follows. So this is their main function. They can also additionally mark other meaning. So they can mark that the new event is unexpected, that the new event is sudden, or that it's simply a new event that is a logical consequence of a prior event. So I have two examples coming from Chadic language from Kera. Kera has two narrative discourse markers, both grammaticalized from motion verb. The first one is illustrated in 5A. It's grammaticalized from the verb to come. And well, it marks, it's used by speakers to signal new events that are logical consequences of previous events. So the prior context of the previous events described is they set out to find a trap, they then set it up in the beans. So apparently, I don't understand what this example, but apparently setting up the beans is a logical consequence of setting up to find the trap. Well, and in B we have an example of the second discourse marker, narrative discourse marker. This one is grammaticalized from the verb to go, and it marks that a new event is unexpected or occurs suddenly. So the context is Hyena hung around chatting, suddenly she saw that the fish was no longer there. Again, it's very, it's unexpected that the fish is not there apparently. Right, so now that the background is done, I can focus on Ural in Tagbalit. So as I said at the beginning of the talk, Ural canonically encodes motion and it's a written shaped, it's a motion with a written shaped path as illustrated with a different example in six. I went back to that place where I was staying. It's most often, it most often encodes a motion that has a written shaped path but it doesn't have to encode a written shaped path. So in some context it's just encode simple motion. So in seven, a little bit after that came one boy on a bicycle. So this is an extract from one of my pair stories and it's the first time the little boy turns up in the story but the speaker uses comeback. The verb Ural when it encodes motion has the property, most syntactic properties typical of other verbs in Tagbalit. So it usually carries an agreement marker and this agreement marker has to agree with the subject NP if it is overt. So Tagbalit doesn't have to have an overt subject but if the subject is overt then the agreement marker on the verb has to match with the subject that is overt. The verb like a lot of the motion verbs in Tagbalit can also host a ventive critic. So the role of the critic is to direct the motion event to the direction of the speaker. So if the event is directed towards the speaker, a verb like Ural can be modified by the ventive critic. The verb is intransitive and it is optionally followed by a locative or a directional complement and this is what we have in 6. I went back to that place where I would say staying. So to that place where I would say staying is a directional complement. And the verbs when it encodes motion can inflect for different aspects. So in the previous example it was in the perfective. We have an example in 8 where the verb is in the imperfective. She would come back to where all the women were grouped. Now the verb Ural is also used in a certain context with change of state semantics in which case it's translated as to become. So we have an example in 9. He arrived, the poor man, he became yellow. So the verb again obligatorily agrees with a subject NP if there is an overt subject NP like the other Ural. However it doesn't normally occur in different aspects. It usually occurs only in the perfective aspect and because I haven't done elicitation I don't know whether it's ungrammatical to have the verb in other aspects but I just know that I don't find it and I've never heard it. The verb never occurs with the votive marker and it's obligatorily followed by a copular clause. So a nonverbal clause. In 9 it's this thing introduced by the copular D and the word for yellow. So this is something that could be used as a clause on its own and it would mean he is yellow or it is yellow. So literally what the second line of 9 means is he became he is yellow. And now grammaticalize Ural, the Ural which I think is a narrative discourse marker. In 10 we have the examples that I presented at the beginning of the lecture. She cut the navel string and then she bandaged the boy. So grammaticalize Ural does not encode event semantics as opposed to the motion Ural and the change of state Ural. It co-occurrs obligatorily with another verb. This other verb is the main verb of the clause. It encodes event semantics and it carries the semantic load of the clause. Ural does not inflect for aspect, it's a bad way of saying it. It only inflects for perfective. So it only occurs in the perfective stem and it does not, it cannot take the votive. It inflects for agreement and normally the agreement marker agrees with the subject of the main verb but it doesn't have to agree with the subject of the main verb. So in some context I find that the verb agrees with something else and usually that something else is either a participant that was mentioned in the prior context in the previous events. And maybe a thematic participant of the main verb but not its subject. So this is what we have in 11. So in 11 the verb Ural here has the first person singular agreement marker but the main verb has a subject agreement marker which is the third person plural feminine. So they don't have the agreement doesn't match. The agreement matches the, oh there it is, the dative argument of the verb. And I think the speaker was mentioned, I mean she talked about herself in previous sentences just before this example as well. Yes, so as I said speakers translate grammaticized Ural as meaning as having sequential semantics so as meaning and then and after but they also describe other meanings and sources which have focus on this topic also describe other meanings for grammaticized Ural. So another one is in co-active aspect. So it is often described in this context as marking the beginning of a new event. So start doing something. It is also described as having terminative aspect semantics. So I'm translating from French. So Shaker translates it as meaning in French final which can be translated as finally but and it will be interesting I think in soon. But I think that finally in French has an unexpected meaning that finally in English doesn't have. So maybe that's why he translates it as finally final more. And it's also sometimes translated as meaning going back to doing something. So now here's why I think that Ural has grammaticized into a narrative discourse marker. Why I think it might be a narrative discourse marker. So first of all the verb doesn't serve to link two conjuncts. So it doesn't link just one single clause with another single clause or one event with just one other event. What it does most often is that it links sets of sentences which describe related sets of events to other sets of sentences that also describe their own sets of related events. I mean I call these thematic paragraphs to help me but it's a bit different to the thematic paragraph in given. So let's forget about that. So the sets of sentences they can be of any size so they can consist of just one sentence or they can consist of five, six sentences. And so I have an example of a long paragraph. So the prior context is this long paragraph and then it's this little sentence in 13 that is linked to that whole paragraph with the verb Ural. So the previous paragraph is so about Mustafa's birth. Yeah sorry, it's not very glamorous but I have a good example of Ural so I don't care. So the contraction started at six in the evening. My stomach was sore, it was sore, it would be sore, it would come down, sore for like a whole paragraph. Well the speaker is talking about contractions and then there's the I fell asleep, I slept until four in the morning. So it's a, we can say a new set of events, a bit unexpected, you know. And so what Ural is doing is linking that utterance or that sentence to the whole paragraph that comes before, not just to one sentence that comes before. And then in 14 we have a little bit of how the story continues so then she slept and then I awoke, my tummy, my tummy, blah blah blah blah blah. So I also think that Ural is a narrative discourse markers because it seems to have the discourse functions of narrative discourse markers. So it often introduces new series of events and these new events are often unexpected and sometimes it's hard to say that they are unexpected but for sure they differ from what could be the default inference. So what we would expect, well yeah so unexpected. But unexpected to me sounds very strong but anyway. They use Ural also to return to a topic of conversation after an aside. They use Ural to introduce a new event that's very rare in my corpus. So they use Ural to describe a new event that is the consequence or the culmination of previous events. And then there are contexts where it's hard to tell whether Ural is here because it's describing a new event or just because it has a different aspectual function. So in 15 we have a set of sentences, so a thematic paragraph that precedes a sentence introduced by Ural. So she took me and led me down on the ground. Yes so this is again something that happens after someone has given birth and I don't really understand what is the aim of these actions. But they used to be done straight after a woman had given birth and they go together. So she took me and led me down on the ground, she put me on my face, she walked all over me, she walked on my back. She walked all over me like this, one sat here, the other sat there. They bound me like this and then this is the next sentence after all this. After I got up, so she doesn't say it but it's inferred, I got up, I threw away all the waste, I cleaned everything, I cleaned everything, all this waste. So definitely a new event, so you know she's getting up after whatever, these events that we don't really know what the aim is, what their aim is. But it's also unexpected, she has just given birth and then she wakes up and then she starts to tidy up everything. So she's probably marking, you know, signalling that this is unexpected. There's another example in 17, it's from the past stories again. I don't know, I mean I don't know whether it marks unexpectedness here but it could. So this is just, she's describing the moment where the little boy steals the basket of pear in the pear story. So then he robbed one basket from him, one basket he robbed it, the other he left it empty. So again that's, you know, it's unexpected, you don't expect to, you know, have your basket of pear stolen. And in 18, I have an example that I took from another source. So it's an example from Shaker, he gave it to him and he translates Yuhal here as finally and he says between brackets and contrary to what he had said before. So again contrary to what was expected, finally he gave it to him. Okay, so there are examples from other people also showing that it may mark unexpectedness. Now I have very few examples as I said where the event, it's one event introduced by Yuhal and it's usually linked to just one previous event as we have in 19. And I think that here it marks that the event is a logical consequence of the previous event or it's the culmination of the previous event or state of affairs or whatever. So when I grew up they veiled me, they veiled me. I also find the verb when the speaker returns to the topic of conversation after they have stopped the topic of conversation to talk about something else. So I had a better example in my corpus but I chose this one because it's shorter and it fitted on the slide. So the speaker is talking about where she lived when she was younger with her parents and then she starts 20, we left. And then she explains to me the war with France started, so that's why we left. And then she goes back to the conversation we left, we went to Bria, etc. And yes, so there are contexts where I don't know whether Yuhal is used here just as a discourse marker because it marks a new event or because it's a spectral, it marks to start doing something. So these construction, where are they? They're a little bit different to the previous ones because the second verb, so the verb which follows Yuhal is not in the perfective in those examples. It's in the Auri stem preceded by this particle and usually this construction we find in infinitive clauses or subordinated clauses that encode purpose. So yes, so the translation of that event is I started to call Nunu Nunu, it's a boy, it's a boy. So it's repetitive, so she gave birth again. It's a different story but it's another story of giving birth and after that she called her eldest daughter to say it's a boy. But she repeated it several times that it was a boy because she was happy maybe. And I have another example involving exactly the same construction, so the particle here with the verb in the Auri stem. So the previous context is another boy went to him, another boy went to him, he lifted him, he brushed him, etc. And then after that even him started to pick his pairs. So again this is an extract from the pair stories. So a little boy, there's a lot of little boys in this particular example. So one little boy has fallen down and all his pairs have fallen down. Several boys are picking up the pairs and one little boy, that's the one in 20, is helping this other little boy cleaning his stuff. And then the little boy who's helping him and cleaning his stuff then starts to help him to pick up his pairs also. So there's a change, you know he was doing something, now he's doing something else, he's picking up the pairs. But I don't know whether it's because the event is repetitive, you're going to pick up the pairs repetitively. Or whether it's because it's a new event, it's not clear to see. Right, so these are the main discourse properties of Ural, Grammaticized Ural. Now I would like to describe some of its morphosyntactic properties. So first of all I'm asking the question and this is when I don't have a lot of answers but I have a lot of questions. So I'm asking the question of what kind of construction is the construction involving Ural plus the following verb. So are we dealing with a multi-closal construction or a monoclosal construction? And what is the status of Grammaticized Ural if we are dealing with monoclosal construction? So in some examples we have evidence that the construction is monoclosal. So in 23 we find the verb Ural with its discourse function so that verb we find it kind of like embedded after an adverbial clause. So when I grew up they veiled me. So the adverbial clause is modifying the verb to veil and it precedes the Ural. In Tagbalit and I think in other verbal languages when we have a multi-closal constructions which is formed from a verb plus another verb in the perfective, which is the kind of construction where we find Ural most often. In those constructions we may optionally include the complementisably which means that. So there's two examples of that in 24a and b. So 24a, I remember that Amirah left at 12 o'clock or I remember Amirah left at 12 o'clock. In the Ural plus verb in the perfective construction we can never find the complementisably. So 25b which is the version of 25a when I grew up they veiled me with the complementisor is wrong. Yeah actually 25a doesn't have when I grew up it's just they veiled me sorry. And 25b shows that we cannot include the complementisor there. So Ural plus verb in the perfective doesn't behave like multi-closal constructions involving a verb and another one in the perfective. Now there's also evidence that some constructions that involve grammaticized Ural are multi-closal. So in Tagbalit and in a lot of other verbal languages there are multi-closal constructions which involve a verb, a second verb in the Aorist form. So there's an example in 27. Us we went down to the living room to sleep so this is a multi-closal construction and as you can see the second verb is in the Aorist and it's preceded by the particle. Basically the canonical strategy to have two clauses in Tagbalit is to have this particle with the verb in the Aorist. There's another one in 28 but we don't have to look at it. And of course as I mentioned before there are contexts where we find Ural with a second verb that is in the Aorist form which is preceded by the particle ad and I started to call Nunu. It's a boy example again in 29. We've looked at this example before. So I mean these recall are the constructions that are not clearly discursive. They could have an aspectual meaning. So it is possible that we are actually dealing with two different constructions involving a grammatized Ural. One which is multi-closal as in little i and one which is monoclosal as in little two. The multi-closal one is when the second verb is in the Aorist and preceded by the particle ad and the monoclosal one is when the second verb is in the perfective. As I said before so they seem to have different meanings anyway. In the multi-closal construction Ural is partly grammaticalized so it has undergone some semantic bleaching so it's gone from motion to in-coative aspect. It's starting doing something repetitively. It's undergone some de-categorization so it can only occur in the perfective form but it has retained some verbal properties so it can have a closed-hole complement that's the ad plus V Aorist. And in the second construction Ural is more grammaticalized. So it's gone from motion to something else which I will discuss later and it ended up being a discourse marker. It's undergone de-categorization so it only occurs in perfective forms. It's got a less rigid agreement so the only time when I don't find the agreement matching the agreement on the second verb is when the second verb is in the perfective. So maybe Ural is more adverbial there, I don't know. And of course I mean the whole construction has undergone possibly some kind of re-analysis depending on what the source construction of that construction is. And I'll talk about that in a few minutes. So now I'd like to talk about the category of Ural when it has this discourse meaning and when it occurs in monoclosal construction. So motion verbs and other activity verbs in Tagbalit and a lot of verbal languages when they grammaticalize usually they become either particles or what people describe as auxiliaries. So I don't think that Ural has grammaticalized into an auxiliary. I don't think it's an auxiliary because it's the following verb that carries agreement and the tense aspect mood load of the clause, not Ural. And verbal languages, I mean technically verbal languages do not really have auxiliaries actually. And it's not either a tense aspect mood particle. So there's a lot of tense aspect mood particles in verbal languages that precede the verb but they have two properties that are illustrated in all the examples in 31. So the first one is that these particles they attract critics. So critics usually in Tagbalit and in other verbal languages they attach to a verb but if the verb is preceded by a tense aspect mood particle the critic goes on the tense aspect mood particle. So as you can see in 31 here and in 31 again. So the second example in 31 and also these particles can never be separated cannot be separated from the verb by an independent word. So they can be separated from the verb obviously by a critic but not by a verb that can stand on its own. The last example after I will tell you the story where the adverb after occurs between the particle and the verb is ungrammatical. Ural in the verb in the verb in the perfective construction Ural never attract critics. So this is what we have in 32. The critic has to stay on the main verb. And there are examples in my narratives very few but still there are examples where the verb Ural is separated from the second verb by the adverb meaning after. So after I started to throw away all the waste. So now finally the possible sources of grammaticalization so as I said the beginning at some point towards the beginning of the talk items grammaticalized from particular morphosyntactic constructions from particular morphosyntactic contexts. So in the literature on the motion to narrative discourse marking grammaticalization path there are three scenarios proposed or three morphosyntactic construction that are possible sources for this development. The first one is proposed by Heinz 2000. I don't know if I call it or if he calls it probably he calls it the iterative scenario. So according to him a verb so a verb starts as a motion verb so it expresses motion and then there's a construction in which a verb repeats the motion encoded by a previous verb so the motion verb becomes to express echoing motion so not really motion in the clause where it is used but motion that has been expressed in a previous clause and then from echoing motion it becomes a discourse connector or a discourse marker. The second scenario is the futureity scenario so it's proposed by a number of people so according to that scenario sorry we go from motion to purpose and then to future tense and then to discourse marking so discourse markers actually grammaticalized from motion verb that have themselves grammaticalized into future tense markers and the third scenario is the repetition scenario it's found in some oceanic languages and it's proposed by Moïse Fourier so and it's a path that is specifically proposed for verbs meaning return or go back so in this scenario we start from motion and then the motion verb grammaticalizes into a repetition marker so something marking again and then from again it grammaticalizes into a discourse marker so the discourse comes from the repetition or iteration meaning so there are I think only two possible scenarios the first one there's no echoing motion in verb or in tag-by-lit so it can't be the right path so there are two possible paths for tag-by-lit oral so the first one is the purpose future scenario it's possible given that we find a construction where oral is followed by an ad plus verb aorist construction and the ad plus verb aorist construction marks purpose a lot of the time it's possible but it doesn't seem likely it doesn't seem likely because I almost never I mean I never find in my corpus oral used in a purpose if construction never so I never find examples like return in order to and also oral does not encode future at all there's no verbal language I think where comeback oral has grammaticalized into a marker of future so it's not the right scenario plus discourse oral precedes well the canonical oral that marks discourse functions precedes most often perfective verbs than aorist verbs okay when it precedes aorist verbs we don't know whether it's I don't know whether it's really discourse or whether it could be aspectual and finally it's not a good scenario for tag-by-lit oral because so items that are grammaticalized in tag-by-lit and in verbal that are grammaticalized from the verb plus ad aorist construction they tend to incorporate the ad particle so the end form incorporates the ad somehow so an example is the future particle rad in tag-by-lit verb from morocco and so this form is a reduced form of he wants plus ad plus etc and you see the rad encode incorporates the ad yeah it's clear so the second scenario is the repetition scenario it's possible because some earlier sources describe a repetitive meaning for oral so I found one example in a dictionary from why is it coming like this from 1901 so he went back to sleep and so yes I wrote it like it was written in the dictionary so that's why it's written in a different way than I do so here this is ural here and ural means the the re-beginning of a state or of a new event I think yeah so it's possible but in my data I don't have these repetitive meanings of ural in my data in my narratives I don't find it so I don't know so what I would like to suggest is that the source for the discourse the narrative discourse marker ural is actually the change of state construction this is how do you say an intuition that I had a long time ago when I started translating my narratives and I had this intuition before I found all the data talking about how motion verbs grammaticalize into discourse markers so maybe I'm wrong but if that's the case then we the source of the discourse marker ural is the construction that I described at some point where ural means become where it expresses a change of state and it's followed sorry not by a nominal clause it's the wrong way of describing it by a copular clause and the copular clause expresses a state and then well possibly from there the verb ural still means change of state to become but then instead of being followed by a copular clause expressing a state it's followed by a verb expressing a state in the perfective so the reason I I'm thinking about that is because it's very common in Berber to express states or things that are expressed in English with adjectives it's very common in Berber to express them with verbs and to have the state of meaning you need to have the verb in the perfective so if you have the verb in the imperfective then it becomes a change of state and then from there so it's still expressing a new state ural with occurring with any verb in the perfective and instead of marking a new state you mark a new event I'm happy with that except that I don't have evidence for the most important the one in between it's a shame but well I haven't looked yes so possible evidence no actually I thought about one possible piece of evidence it's again the example that I found in the dictionary from 1901 so he went back to sleep, what he translates as he went back to sleep so this is the verb to sleep in the perfective aspect I think but the verb to sleep is a stative verb at least in the dialect that I work on it's a stative verb so it doesn't mean he's slept literally it means he is asleep so this is an example with a stative verb so I can't really support my hypothesis so yeah I'm done so some conclusions so there is evidence that the verb return or go back in TagBallet has grammaticalized into a narrative discourse marker its main function is to link one or more new events to a series of previous events there is evidence that some of the constructions in which it occurs are monoclosal Ural has lost most of its verbal properties but does not function as a particle or an auxiliary it's not clear what its category is the source construction for the grammaticalization in TagBallet is not clear either but the discourse function may possibly come from the change of state construction and I'm done with the other references if you're interested thank you sorry I didn't understand a lot about the effective verb and states it seems to me that being imperfect when expressing a state is wrong no it is so some stative verb can occur in the imperfective but they don't express a state they express a change of state so he is becoming this in other verbal languages it can have all sorts of meanings as well when you put the imperfective on stative verbs but I can't remember now the sleep example the verb actually means to be asleep but it's still in an effective form is that what you find weird? in Berber it's very common to have the perfective as not a change of state a stative so he is asleep I don't know what else he is wet so it means he has fallen asleep he has entered the is that what you mean? I don't think so some people say that maybe it's understood as a stative it has a change of state in it I don't think so one reason why I don't think so it's because of the ventive critique in Berber it doesn't just occur with verbs of motion it can occur with different kinds of verbs and usually these verbs they need to encode a change and all these verbs like to sleep that are translated as stative they cannot occur with the ventive when they occur with the ventive it means something else it means he slept somewhere else he went somewhere he slept he came back but I don't want to talk about that so I'm going to go find the example but it's taking ages yes? I actually have another question if you go to example 20 I think it was there I think the person another marking on Uval was but they are different that person plural yes but I couldn't see the third person plural in the context around there I think it comes before she's talking about her family alright so does it really always refer to some entity that was talked about? in minorities yes in minorities yes it could be some kind of performance but I don't think so it's repeated a lot from seeing this context I thought that it might be that she's talking about the parents or something and then she's in the first person plural to include herself in the family I don't remember what comes just before that maybe it's the example where when I grew up they veiled me blah blah blah we left they are in the they're mentioned so it can really go back a couple of sentences yes and she's in previous paragraphs she's talking a lot about her family what she needs to do for her parents her grandmother you know the adults around her she's to work on so they're really there in the story curious about how frequent these constructions are who you looked at I haven't done sort of like I haven't it depends on the speaker it's very frequent with older speakers I've got a speaker that uses them a lot she's monolingual and she's older and then I have speakers that are bilingual they speak Arabic and well they don't use them a lot and when they use them they use them in the very sort of you know it agrees with the following verb and it's just a new event but it's more frequent with older speakers that are monolingual and do you notice any substrate influence in their Arabic they would use this as well no no these are those that don't speak Arabic but you said you had bilingual oh yeah but they don't use they don't use it that much no I haven't no I haven't even considered influence from Arabic but so I guess in this example that we've got that has has a third person for all my friends or so does not it looks like in most other examples they have the same normal person it's I guess overall do they mostly share the same person number marking yes yes they do I mean I looked at other sources and there is a source also mentioning this mismatch in agreement so I found an example where Ural is has the third person singular agreement marker and the verb following is in the third person to rule feminine but then again in Taqbali third person singular can be the I mean it's also the default sometimes so a lot of verbs that grammaticalize you know then they kind of like become grammaticalized from the third person singular masculine so yeah as I said I didn't do any elicitation I wasn't working on Ural I wasn't interested in Ural at the time so I didn't ask a lot of questions but so it might be that the speaker is making some errors but I mean she would correct herself why she knows my language I mean at some point you have to trust her yeah but if your story was right and it's function is primarily a discourse marker then it's difficult to explore that elicitation you really need to discourse context to be able to figure out the semantics and people's intuitions will break down if you if you start asking if you just start pulling out individual have you got any conversational material I mean curious about whether it's no I'm not recording right I'm sort of curious about whether in turn taking it Ural can be used as a continuity or having said that and then what oh yes yes then I need to do some recording I'm just thinking of there's a Japanese construction you would say having said that so you can use that as a prompt to get somebody to say in the next thing do you have any reference for the Japanese? I can find some yeah that's with the vote to speak not with a motion yeah that's very interesting I'm just wondering whether you have any speculations as to why younger bilinguals do not use how as the discourse marker do they use linear discourse markers instead or no not very much if they I mean for sequentiality sometimes they use the Arabic but it's very rare and it doesn't I don't think that it has the same function the Arabic is really after but it's very rare no I don't know it your question is a bit similar to Peter's earlier question I don't know whether it's because it's influenced from Arabic or whether it's because I don't know whether they're younger and you know it's maybe young people don't use that maybe it's not cool people may actually create other kind of discourse markers yeah I haven't looked at yeah there's language shift going on so that younger speakers are less fluent in this one because one of the things that we often see across linguistically is that the first sign the first formal structural sign shift is actually at the discourse level so people you know the discourses become less well I don't know what the situation is like at the moment I haven't been to Nigeria for a while but I think young people are one of the strongest verbal languages and you know there's a big region in Nigeria where you know Takbalit is spoken as you know the everyday language young people I mean outside of cavity I think the young people speak Arabic more often and they speak Takbalit to speak to an older member of their family like a grandmother a mother that doesn't speak Arabic but amongst themselves yes I think they speak Arabic but outside of the cavity region in the cavity region I have a few friends and I hear them speak in Takbalit amongst themselves have you sent anything in any other verbal language that's parallel to this I haven't looked and I don't think there's a Ural I don't think I'm not sure we find it in this form in other languages but I mean no I looked at some sources like Cosman you know the grammars from Cosman so he works on some Moroccan verbal I didn't find anything but I didn't do like I think we should tell Madyum to keep her eyes out yeah oh yeah Peter can I just ask what you meant when you said it would shift and it would be the discourse what do you mean yeah so one of the indicators that you're getting shift going on is that people find it more difficult to to organize and structure discourse so what's the name we worked on Scottish Gaelic Dorian talks about stylistic shrinkage as being one of the first signs of shift so the range of possible construction types that you get tends to shrink and then the possible flexibility of the language alternative constructions start to reduce and you end up with more fixed kinds of expressions so if somebody's telling a story it'll be the Cinderella kind of story where we know what follows what and so on so Dorian suggests that the first kind of step in language loss language shift is stylistic shrinkage people not being able to have the full stylistic range of expression and then eventually getting down to where they only know fixed expressions and their lexicon collapses and their grammatical system reduces so then it would make sense that his speaker is older than then and of course then you also have contact effects so that things like and I think you see this in a number of other areas so that stuff like discourse connectives adverbial things get pulled in especially where if languages have say non-finite they use non-finite expressions of linkage and the contact language has more finite instructions with linkers and then after that having done that consequently they can get borrowed in or copied in so that you get a copying of the patterns so kind of native discourse organization can then get affected by the influence of the contact language and Dorian is a good source if I want to read about it it would be successful well she's the first one who identifies stylistic shrinkage as an example sorry sorry sorry sorry sorry sorry sorry actually it was interesting I'm glad that just to give you an example some of the languages I've worked on in Australia you could find speakers who are fluent in the sense that you ask them how do I say that they can translate perfectly for lonely and they have conversations they talk to each other in conversation but if you sit them down and ask them to tell a story like that they can't do it because the genre of storytelling has kind of disappeared so the notion of having you know the style of storytelling has dropped out of usage and the next step in that is then you become the word or it becomes reader yeah actually yeah I thought you were going to make a methodological point which is to really capture to understand this you said there's many mysteries yet but to attempt to do that you need discourse material you need narrative you need conversations simply going for translation or doing elicitation or grammar checking won't actually lead you very far no it's methodologically from the point of view of studying these things yeah it's good to have narratives and not just elicitation even if it is pair stories you know what to do with this separation yeah actually I really like this language and discourse thing yeah it's yes because I see that in Chiang the sort of old French and now it's really switches to English French and I see that you know they speak like it's so simple so that's why I'm curious about what you're saying the way they tell stories they talk so that's nice yeah that's good okay well as is our usual practice we'll head over to the UCL Institute of Education if you'd like to join us join Asia and have an informal conversation about things please come along so let's thank her for