Colloque : Syntactic Cartography and African Languages
Conférence du 18 juin 2024 : How Language Contact Accounts for Mixed Typologies
Intervenant : Salikoko S. Mufwene, Université de Chicago, professeur invité du Collège de France
Retrouvez les enregistrements audios et vidéos du cycle :
https://www.college-de-france.fr/fr/agenda/colloque/syntactic-cartography-and-african-languages
Chaire Linguistique générale
Professeur : Luigi Rizzi
Tous les enseignements du Pr Luigi Rizzi :
https://www.college-de-france.fr/fr/chaire/luigi-rizzi-linguistique-generale-chaire-statutaire
Le Collège de France est une institution de recherche fondamentale dans tous les domaines de la connaissance et un lieu de diffusion du « savoir en train de se faire » ouvert à tous.
Les cours, séminaires, colloques sont enregistrés puis mis à disposition du public sur le site internet du Collège de France.
Découvrez toutes les ressources du Collège de France :
https://www.college-de-france.fr
Soutenir le Collège de France :
https://www.fondation-cdf.fr/faire-un-don
Suivez-nous sur :
Facebook : https://www.facebook.com/College.de.France
Instagram : https://www.instagram.com/collegedefrance
X (ex-Twitter) : https://x.com/cdf1530
LinkedIn : https://fr.linkedin.com/company/collègedefrance
[Music] [Music] U thank you Luigi for inviting me to participate in this colloquium um my Approach is going some to be somewhat different from the papers that we have attended this uh morning in the sense that I will um assume a dionic orientation and then focus more on typological changes in the emersion of Crees then the kind of detail analysis of uh the cartography that we uh have attended uh so far um the revitalization of the scholarship on Creo language varieties has brought to Center Stage the role of language contact in language change and language speciation and my Approach is uniformitarian in the sense that I assume that the same restructuring processes can apply to any any language under similar ecological conditions and by ecological conditions I mean anything outside language that affects the behavior of speakers or signers and then can lead them to interact specific ways and something that is Central to this is uh what is known in population studies as population structure and that means for instance whether or not the population is integrated or segregated residentially and to what extent segregation if that is in place affects how people uh interact with each other and who uh whether the the power relations are egalitarian or uh uh subjugating and who are the people that are likely to uh uh make accommodations to which other group and uh which population is majority and how do you get to that majority and so forth I will not be getting into all these details but this is just in order to give you an idea uh usually when I talk about this aspect of Creos I also invoke economic uh dynamics that are very critical in um uh the the ways that a population is structured especially in a contact setting so a great deal of attention has been directed both to how structures of cre and pigion differ from those of the languages that we call lexif or the languages that the subordinate group had to learn in the contact setting and also to causes of the structural changes and that’s where I invoke population structure or ecology St influences claim also sent a stage in competition with the role of common processes of second language acquisition and that of universal grammar through the agency of children um in the immersions of Creoles and the role of children that was really a promoted by Derek bierton who gave us who claimed that um cre started as pigeons and then they were the pigeons were grammar and the children by some process of what I call kudeta sarcastically um would say our parents were too stupid to you know make up a language for themselves so we’ll make one for them um but otherwise a lot of people have recognized that adults contributed to the immersions of Creoles all pigs and uh everybody that communicated was really involved in order to understand this you have to give up the tradition of thinking that language a language is transmitted vertically from the uh caretakers to the children uh that is an EP phenomenon language is just transmitted horizontally uh so uh regardless of age and the ra populations in which children are actually mothers for their parents and parents are learning a number of things from children that happened to me in the United States where uh I attended school really based on what I can call Scholastic English and then my child was born and I would go to several birthday parties or host birthday parties and hear children speak in a different way and I would learn um vernacular variety of English really to be able to socialize with everybody else that was not an academic so that not worry and nowadays creis have been polarized between cre exceptionalist and uniformitarian exceptionalist in the sense that since the 19th century people have generally assumed that um cre are some sort of De ation from the normal between quotation marks in which languages have usually evolved and the critical Factor was contact and so putatively contact didn’t play an important role in the speciation of Latin into the romance languages and English which is also an obvious case came from somewhere and you don’t have to take into account the fact that the uh colonizers of England the Germanic colonizers of England spoke different languages and out of that we know just just by the name English that the angles must have prevailed but not without influence from the other languages that the the other population that the angles came in contact with so these are the kinds of things that we have to remember and uniform is that’s what I explained earlier as a uniformitarian is I focus here not her sorry I focus here on the nature of some changes some of the changes to show how qus can contribute to the theme of this colloquium syntactic cartography okay so I start with adjectives what’s the matter with adjectives predicative adjectives uh in English cre so you can have I just have made up things here that can apply to the English Crees that I know including Jamaican Creole about which we heard something today um I have worked mostly on Gala on the coast of South Carolina and this representation that I you see here is what is called ey dialect the adaptation of English you have to use the English pronunciation to read them uh correctly with adjustments to pronunciation like the very tall boy okay in Jamaica you hear a boy okay but what is interesting here in this attributive Construction in the noun phrase the word order and the syntactic or lexical categories are exactly like in English nothing really has changed but when we we get to two we have the boy very tall or the boy tall okay without a copular and the challenge is to explain why there is no copular the boy the best uh the the boy is the best um the boy taller than uh than me or the B the boy tall past me uh that varies from one cre to another uh but you cannot say the girl very fast sorry the girl fast run or the girl very fast run you cannot say that so in this respect the verbs have remain verbs the question is whether um adjectives become verbs in a predicative function as has been traditionally claimed in the literature and I I you against that particular position in saying actually what happens here is that adjectives have remained adjectives they have the behavior of adjectives in the noun within the noun phrase except that there is no copular here and why is there no copular except for the absence of the copular they have maintain properties of adjectives in a noun phrase like in the lexify so the syntax of the predicative phrase has changed so typically you know the way I learn syntax I’m a little bit conservative um the sentence can be expanded into a non phrase and a verb phrase okay that applies well to the lexify the European lexifier by the way I’m us talking about Crees here in a restricted sense about those language variety is that have a European lexify the term has been extrapolated and the more it has been extrapolated the more confused we are about what a creole is so I’m going to try to keep things a little bit clean here so my system here what I say is then the predicative phrase can expand I think I went further than I wanted to be yeah can expand into a verb optional NP optional NP or optional preposition phrase or it can expand simply into an adjective phrase or it can expand into a preposition phrase okay so um that means in the same way you can say the boy tall you can also say the boy in the house also without a preposition phrase and without a popular but when you say um the the boy is a teacher you have to say the boy the teacher in Gala or the boy a teacher in Jamaican Creole okay and that’s interesting so the noun phrase is not involved here okay and the reasons well we we notice in the morphology of uh the Creole that the verbal inflections have disappeared okay tens and aspect markers are fre morphes which are not dedicated to verbs only okay and what I want to argue is the reason why there is a cular in the lexifier because there are constructions in which you need something to carry the tense or aspect or mood inflections but when you don’t have that kind of restriction anymore then nothing really keeps you from using the same TM markers for tense aspect mood markers uh for adjectives too or for prepositions and and and and the like okay and example here for T the girl been run very us okay and Bin is what we call an anterior marker because the t t system itself has changed into u i mean from an absolute relative system in lexifier to just a relative tense system uh in the uh in the Creo and the boy being very tall and that’s fine the teacher been in at the house and that is what you get the behave in similar ways okay and what we see here is that because there are no inflections verbal inflections and the mark of anterior is just a periphrastic morim and the mark of aspect there the in Gala but the in Jamaican cre U is also can be attached to any one of those for instance if I said the boy the tall does not mean the boy is tall it means the boy is emerging now as tall okay if you think that cre have simplified thing things think twice there are nuances in cre that you don’t find in the lexify and that is quite interesting okay so there are more reasons and they come from substrate languages in which most of which is not if not all of them um the category adjective is either non-existent or underdeveloped by underdeveloped is that I mean you find a small set of items and later I’m going to say the syntactic category is indeterminate that you can really call adjectives but many of the cases things work differently things that are expressed by adjectives in the lexif are expressed by verbs very often inflected in the Progressive or very also also in the per perfective should say perfect not perfective in in the perfect so K is one of my mother Tes the one that has suffered with less attrition than my ethnic language tiansi so I chose the examples [Music] from Maria Bella so it you translate idiomatically in English as Maria is sick but what people are saying is that Maria is sickening sickening is having the experience of being sick okay and that’s in the Progressive or you can have a commutative construction Maria or Mar Mal and now I have gloss it as a connective because it’s something similar to of in English you can find it in a variety of constructions there is no specific meaning associated with it other elements in the sentence help you determine uh how to interpret the construction okay on the other end you can have yayak mingi Ando means strong and mingi means much and this is something that is similar to a construction in the lexifier with a copular okay but um you can also say that means my elder brother has a lot of strength okay so Ango is kind of you know depending on how you want to look at it it’s probably a noun it’s probably an adjective or it’s something in between it’s ambivalent okay the membership of the items that can be considered as adjectives is typically very small in the substrate languages and based on the examples presented here one can say they fall in the indeterminate category okay and the substrate influence shouldn’t be ignored um regarding the grammatical change in Creos even though what is interesting the Africans who didn’t have adjectives in the languages have acknowledged something in the lexify that is an adjective but when the adjective has a predicative function there is less pressure on them to use use a copular and have a verb phrase so here we go into an isolating a morphosyntax uh uh of the kind that you can find in Chinese for instance some of these properties where in Chinese the adjective has a predicate phrase without a copular where a preposition can have a pred a predicate phrase without a copular and and and the like we are dealing here with a case of typological real alignment from inflectional to isolating morphosyntax which accepts categories um other than verb than than the verb uh to had a predicate phrase like adjectives and preposition English itself is also a contributor to this because in small Clauses you say they CAU him hence in the bag so hence in the bag is a small clause and there you don’t have a copular in this particular case Okay more than the restructuring of the predicate phrase loss of verbal inflections made it difficult to distinguish between the gerant and the infinitive because there are no more markers as a matter of fact the finite non finite distinction has erased from has been erased from CRE grammas uh if you think that the finite nonfinite distinction is based on the morphosyntax and not on the semantics or anything okay and that’s what I assume though an alternative uh one was uh adopted between closely subordinated with four to or the complement say that you find only uh in uh English Crees and there is no counterpart of that for say in uh French Creoles and my reason the reason I assume is that in English when you actually listen to discourse in non-standard English a lot of people like the quotative reported speech and he tell and he and he told him said and he asked him say and you hear these things in non-standard English and that has uh influenced this uh in French I don’t say you don’t hear that and that explains why French cre don’t have that the counterpart of that construction but what is interesting here too so you you hear Jes tell me f come or in some other cre come or F come depending on the Creole and James tells tell me say come so and James told me come or James tell me again come without say and in 14 what you get is a Serial verb construction okay and serial verb construction is recognized not So Much by opposition versus subordination versus nons subordination is more in terms of how verbs are sequence without a grammatical mhe between them okay although English has svz SVC like serial verb construction like uh in for instance go get the book or come get the book svcs have increased and diversified in cre and you found things such as Jim go hun J you see the translation there and uh I meant Jane but anyway Jan swim cross the river and Jan across the river in France you have alternative ways atra or Al is it acceptable thank you um because I had an argument with my wife about the construction said okay and Lori take knife cut the meat Lori cut the meat with a knife please bring your car come carry me to the hospital this I heard it in Gala somebody was speaking okay for please come and drive me to the hospital okay Peter Peter in at the house the paint the wall you see the translation there and in French you get okay and the conclusion here is the same information can be packaged morphosyntactically in different ways and different languages develop different strategies and this is where you find the cultural dimension of language Evolution okay these are interesting instances of typological realignment rather than simplification of the predicate this is how a lot of people have wanted to explain the immersions and prevalence of Serial V constructions in Crees it’s not a matter of simplification it’s a matter of which strategy you think works and and the initial predicate doesn’t specify Teta roles as people have claimed uh because that would be relative compared to English grammar or compared to French grammar um okay the information is simply packaged in a different way the notion of what counts as an event varies from one language to another and in the usual definition of Serial web construction people say what is Express in English as one event is broken down to different subcomponents in another language the question is you know who knows what the size of the minimal what the minimal size of an event is it’s a matter of perception and in chongo kba once um uh a Serial verb construction is used instead of comitative construction or a subordinate construction is usually in order to dramatize the events what is happening is it has a pragmatic function it’s not just a matter of saying well subordination is too complicated or whatever is that in the way in which you identify different components of what is going on you are dramatizing your narrative in that way in the qu languages it may not be that way but this is the alternative that’s in bant we have and this happens in The Narrative tense only not when you have this all the inflections on the ver okay an important reason for this structural change is the significance of Serial ver constructions in Congo especially in languages without verbal T inflections and so that’s how BTO category can the banto group can join the group because in the narrative tense you don’t have this inflections on the verb either so recategorization within the non phrase 2 uh more specifically in the determinant phrase what happens there the clearest example for me comes from haian CRE where shut for the cat um gets the uh the determinance to the right side so you get sha for the cat sha for the cats and the uh macro plural has been dissociated with the dissociated from the article or the the determiner sh for those cats sh uh cut plural for cats for indefinite reference and then you get that shut without any thing a beown for cats in case of genetic reference but there is something funny here that is happening curious y shut and not sh okay and so the question is is y a fossil of the article or is your a numeral quantifier in this particular language is it possible to just have a determinant system that focuses on things higher than one and then for the rest you stick to another typological option and that of Y as a numeral quantifiers there are no consistent with the systems of many almost subsaharan African languages that’s something that we have it’s difficult for us to learn how to use the article in European languages because you know you don’t know you want to remember when you don’t have to use an article at all and very often we generalize in having an article everywhere these are some of the subtle things that are just difficult and if you compare the fr system with the English system you’ll also find differences like uh count water but in English not mass not count Mass water in English but in French look in the definite okay uh M uh also mass in English which is uh a bare noun whereas in French it labu uh in uh definite article so these things don’t really match and that’s where we have to be careful and there are no more articles in cre um and so hence the reason why I ask the question is it is y not a Quantified U because it appears in the position where other numeral quantifi occur day for two for three and so forth okay and they all preed the now and French itself case has a case of ambiguity in right but when you say Ang it can be the quantifier one but it is also interpreted as an indefinite article that’s easy to have that confusion because the other articles occur also phenomenally and so to complete the Paradigm you have to consider the alternative of saying that H is an article the question here is whether you count on stress to know where is a quantifier you can say I saw one of them or V I saw one of them in case the uh reference is feminine okay and so but you cannot say Jan okay we suggest that is behave you more like a quantify because you can also say and and so forth these are the kinds of things that we have to bear in mind in general quantifies preced the noun in Haan cre as in a p okay in English cre the article has been maintained in the same prenominal position while it’s clear that they have a definite article it is less clear that there is an indefinite one why you can say the boy or boy or The Boy The Boy them for the plural but you say one boy okay I think it’s only G that uses uh but the other cre that I know use one instead of the uh okay wouldn’t that be a reason to think that there is just something of a mix there in the interpretation of the article and um that um somehow it looks like Kos went back to the past history of English where one was used instead of the indefinite article substrate influence appears to be um the explanation again for this peculiarity especially since the substrate languages as I know don’t have articles okay there is there are markers for U definiteness and there are determiners of other kinds but not the article as it is known in the European lexif cre grammar um I I was doing this in a hurry cre grammar are outcomes of hybridization from the contact feature pools where what happens is that from different languages you have these different features they seem to be competing for the same functions and there is something that is going out there that can help us explain how cre grammar emerge and this is when you have to go deeper into the Ecology of Lang anguage contact to understand the Dynamics and of course see for instance where there are factors such as sence such as regularity or whatever such as convenience to the people that are learning the new language you get a lot of these diversions when the group of speakers Heritage speakers of the lexifier is very small there is segregation and the other population because they are multilingual have found it convenient to communicate among themselves in the lexifier but in the process they have modified it okay a unique or exceptional in having typologically mixed system in a way I’m indirectly referring here to the work of Jose and Greenberg in the 1960s when he pointed out that languages are not usually typologically consistent okay so where you have a determin a preceding a non phrase you would also expect the inflections for tense aspect and move to preced the verb but that is not always the case that is quite interesting or you can have like in the case of um um agglutination um a different constituent order in the agglutinated verb than in the larger sentence where you have for instance SV o okay for the larger sentence but in agglutination you may have s o v or you may have a different kind of order and these are things that we have to reckon with and this you find it more for instance in austronesian languages where in the agglutination system you have um the uh s SV o but for the larger sentence you may have vso or a different kind of word order and you know how do you account for these kinds of things so substrate um I’m sorry um and what in checking the literature I noticed that both both Celtic and Germanic languages have articles but Latin didn’t have articles okay where did the article come from in French and in Spanish and so forth in Spanish it’s interesting because the form of the article is close to that of article the article in Arabic okay uh but the u u Latin had a longer history with the Celtic languages before uh the Franks came to colonize go and brought in their own article system but what is also interesting is that when you compare the article system in French to the article system um uh in German U German has three uh genders uh whereas French has only two what what else happened in in the process so there are so many things that you have to take into account okay so and there are similarity is that between French and German and a lot of you here some of you here speak German and not German better than I do but this is something to really look into in any case what I want to show here is that uh the artical system is not from Latin but from languages other than Latin and what Happ in French is just one of those cases of hybridization through competition and selection from the fure pool just like in the case of Creos substrate and asrate influences appear to be the explanation especially if the article system seems to have been anal feature in this particular part of Europe including the British Isles though the function is not identical from one language to another but there are such AAL features and one of the things that Enoch has pointed out in his work about Creoles and earlier than him it was Charles Gilman now dead who thought that AAL features have had a higher chance of becoming part of the structures of the features of Crees in the case of Atlantic Crees then features that you find only individual Lang and what Enoch has made clear for instance for the So-Cal gu languages is that although they are not mutually intelligible they are typologically very similar and so they were very often very the speakers of uh well languages were often the majority proportionally speaking uh in the slave population on the plantations at the time when um uh Crees were emerging and when you bring in the founder principle that gives a better chance of features of the earlier population to continue rather features of other people that come and increase the population only gradually then you can see why a particular part of Africa could have that kind of influence moreover some of the features of where the where languages you can find them in the bant languages too if one goes uh with me I mean if one agrees with me that language contact is an important actuator of system change then one must also accept the reality of mixed systems or hybrid grammars as and calls them as normal and non exceptional evidence of mixed typologies can be found in several if not all other languages we should be able to trace them in the process of linguistic hybridization back to layers of population movements and language contacts in the history of mankind so if you really unfortunately some of these things are not tractable it’s difficult uh but the evidence shows us that there is no modern languages today that is close to what could be primordial languages in the emergence of language in mankind and since the Exodus Out of Africa when you pay attention to history for the last U uh 10,000 years or so you notice that we have been colonizing each other over time and and so for it’s not just the Europeans that came to colonize the world this is the last layer one of the last layers the B had done it already since 5,000 years ago in you know getting into the territories once inhabited by the pigmies the koi and Sun populations and we know that these languages have influence the structures of banto languages and you cannot explain the diversification of banto languages uh unless you recognize the role of these other populations that the B population came in contact with and then subsequent contacts among the B people themselves and for induran it’s a similar story uh the Indo Europeans or Proto Indo Europeans didn’t disperse all over Europe in uh an inhabited territories you know history shows us that they displace other populations and we learn more about it by paying attention to the Roman Empire for instance what happened and what is the consequence of that particular expansion and so this should be I think the default rather than the exception that contact population movements and contacts you know are the drivers of some of these changes and in many cases then we are going to find we can account for typological nonuniformity within the morphos syntaxes of these different languages and I this is my what I want would like to to consider claim as my contribution to uh the craphy of syntax and so forth and thank you very much for your [Applause] [Music] attention e