# Introduction ilingualism is a tradition in Africa considering that language contact phenomena are omnipresent everywhere in all African linguistic communities. As a result, we can observe linguistic contact phenomena of code-switching and code-mixing among students and other bilingual speakers of French, who B Author ? ? : Université des Lettres et des Sciences Humaines de Bamako. e-mails: ibrabsangho@yahoo.fr, mohamedminkailou@yahoo.fr alternate or mix elements of French language together with their mother tongue in their daily conversations. Mali is a multilingual country with French as the official language. Several languages have been identified, but only a handful has received recognition from law. According to I. Skattum (2009c), approximately twenty languages among which thirteen have received the status of national languages: Bamanankan, Bomu, Bozo, Dogoso Fulfulde, Hasaniya, Maninkakan Mamara, So?ay, Soninke, Syenara, Tamasheq and Xaasongaxanon. The relationship between the ex-colonial European languages and the local languages in Africa leads to many phenomena of language change and variation. The local languages receive and adapt new terms from the European ones that express concepts or ideas the native speakers do not have in their language. From this point of view, E. Annamalai (1989, p.48) notes that "in many situations of languages in contact, constituents of one language can be found with the constituents of another language in a number of linguistic phenomena, namely lexical borrowing, transferring, interference, calquing, diffusion, relexification, code-switching and code-mixing, etc." Indeed, the present study concerns the nature of the linguistic contact phenomenon of code-switching and code-mixing which is the use of two different languages in the same speech, as one of the impacts of French language on French speaking Africa. It describes and classifies this kind of language practice to see if the structures being switched or mixed conform to the existing theories and constraints on Code-switching and Code-mixing. Considering the Malian context, literature on Code-switching is quasi-inexistent. Recent research works on the topic include M. Minkailou and I. Abdoulaye (2016) who demonstrated how Bamanankan speakers insert French constructions into Bamanankan and/or juxtapose Bamanankan and French constructions. A further related study still by the same authors analysed the phonological processes of adaptation that the switches by Songhay speakers of French undergo (I. Abdoulaye and M. Minkailou, 2017). A last work in this sense, still by M. Minkailou and I. Abdoulaye (2018) pertains to the types of switches made by Fulfulde speakers of French. A concluding remark about all these studies remains the behaviour of the switched French verbs: in fact, the past participle of the switched French verbs is adopted by speakers of Bamanankan, Fulfulde and Songhay as the base form of the verb and is used as such in all their constructions. Regarding Songhay in particular, there is a need for confirmation since the study that was conducted was about the phonology of code-switching. And that is how this paper becomes highly significant. In fact, while the phonological processes involved in Songhay-French Code-switching have been analysed and determined (I. Abdoulaye and M. Minkailou, 2017), the nature and type of switches have not been specifically determined. Therefore, this study aims to contribute and to fill this gap. The paper specifically aims at determining the types of switches made by Songhay speakers of French, the nature of those switches and the behaviour of the past participle of the switched French verbs. To reach these aims, the following research questions will need to be answered: What are the types of switches made by Songhay speakers of French? What is the nature of those switches? And how does the past participle of the switched French verbs behave? The significance of the study lies in that it will serve to reinforce the few studies conducted on the languages of Mali. The Malian sociolinguistic landscape needs to be adequately portrayed and made known to the public and to researchers. The paper will further serve to comprehend how far the theories developed so far are applicable to all instances of code-switching. The study will finally confirm or refute the behaviour of the past participle of the switched French verbs by Songhay speakers, i.e. the idea that participants adopt the past participle of the switched French verbs as the base form and use it as such in all instances. Theoretically and conceptually viewed, codeswitching and code-mixing generally appear as a basic language contact phenomenon with key concepts such as code-switching (hereafter CS) and code-mixing (hereafter CM) which are not always explicitly defined in the same ways by researchers. Many distinctive definitions of the two concepts have been proposed. According to L. Naseh (1997), the earliest definition of code-switching dates back to U. Weinreich (1953), who defines bilingual people as "individuals who switch from one language to the other according to appropriate changes in speech situation". But, in recent literature, there has been some variation in defining the term 'code-switching' in comparison to 'code-mixing'. For instance, P. Muysken (2000, p.1) refers to codeswitching as "the rapid succession of several languages in a single speech event", however, code-mixing refers to "all cases where lexical items and grammatical features from two languages appear in one sentence". As for A. Bentahila and E. Davies (1983, p.302) codeswitching refers to "the use of two languages within a single conversation, exchange or utterance", while J. J. Gumperz (1982, p.59) perceives the concept as "the juxtaposition within the same speech exchange passages of speech belonging to two different grammatical systems or subsystems". According to R. A. Hudson (1980, p.53) "codemixing represents the case when a fluent bilingual talking to another fluent bilingual changes language without any change at all in the situation". He also calls it a kind of language cocktail. In the eyes of P. Trudgill (1992), it is a process in which language users indulge in code-switching between languages of such rapidity and density, even within sentences and phrases, that it is not really possible to say at any given time the language they are speaking. Milroy and P. Muysken (1995, p.7) consider CS as "the alternative use by bilinguals of two or more languages in the same conversation". S. Poplack (1993) cited in P. Muysken, (1994, p.157) defines code-switching as "the juxtaposition of sentences or sentence fragments, each of which is internally consistent with the morphological and syntactic (and optimally, phonological) rules of the languages of its provenance." We note that this definition is in line with J. J. Gumperz's (1982, p.59). Aligning the subject matter of our study with these definitions, we adopt, as a working definition of code-switching and code-mixing the approach of B. B. Kachru (1983), R. Singh (1985), and S. Sridhar and K. K. Sridhar (1980) which refers to CS as inter-sentential switches and CM as intra-sentential switches. This approach of CS and CM, we think, is the sum of S. Poplack's (1993) definition of CS as "the juxtaposition of sentences or sentence fragments, each of which is internally consistent with the morphological and syntactic (and optimally, phonological) rules of the languages of its provenance" and C. Myer-Scotton's (1993, p.4) definition of CS as "the selection by bilinguals or multilinguals of forms from an embedded language (or languages) in utterances of a matrix language during the same conversation". As a result, both terms show a systematic combination of elements of two languages in the same speech act by people speaking more than one language in a conversation. In other words, to be closer to the African context of switching as stated by P. Muysken (1994), we adopt this definition as an umbrella term to cover interchangeably the linguistic phenomena of both code-switching and code-mixing that will occur in the insertions or alternations of words or groups of words by Songhay speakers of French in verbal interactions. # II. # Research Methodology Following the methodologies adopted by variationists such as W. Labov, C. Myers-Scotton, S. Poplack and R. Redouane when doing research on language variation and change phenomena in terms of field methods, the design of this study is a descriptive survey guided under an interpretivist approach of research philosophy which involves the description and examination of the processes and mechanisms of codeswitching and code-mixing in the linguistic contact between French and Songhay language, and the impact of this language contact situation. It also involves seeking to categorise and describe the Songhay-French code-switching and code-mixing phenomena. The study has adopted participant observation, which appears particularly useful to this investigation because, according to C. Bowern (2008), it is a source of potential information and a very powerful tool for investigating language contact. C. Bowern (2008) considers also that it is a core tool in sociolinguistics that permits the researcher to do an exploratory work to find out how people actually behave in language use. The advantage of this instrument to our study is that it provides spontaneous data through verbal interactions between the participants. More, it gives the opportunity to record information as it happens in the setting, with the researcher being either active or passive in the activities. The population of the study consists of all Songhay speakers of French, and the sample constitutes of 90 participants selected out of five groups called "grins" 1 III. # Results and Discussion in Gao, Timbuktu and Bamako, three principal urban settings where these selected bilingual speakers reside. As instruments, participant observation and conversational interview have been used to collect the data. Spontaneous and fresh conversations have been recorded, transcribed and translated into English in an oral corpus. The cell phone has been used for recording the data. The data are qualitative and have therefore been qualitatively analysed. In the classification of the instances of codeswitching and code-mixing of Songhay speakers of French, the analysis of the data focuses on the structural dimension of code-switching and code-mixing practices. As far as the structural dimension of codeswitching and code-mixing is concerned, we have classified our data into two separate segments, following E. Boztepe's (2005) point of view stating that the intra-sentential and inter-sentential distinction can distinguish the two types of switches equally well. In this vein, considering the structural constraints of the switch points, the first segment will group instances of codeswitching, and the second segment will group instances of code-mixing. Then, we will categorise the different 1 "Grin" is the name for a kind of commonplace where friends usually meet after work to chat and while away time. It is also the place where people of the same age group meet to while away time or to converse. switches found in each segment, and identify the morpho-syntactic structures encountered. The segmentation of our findings follows the definition of E. Boztepe (2005) according to which intersentential switching, sometimes called extra-sentential switching, is the type of switching which takes place at sentence or clause boundaries, or outside the sentence or the clause level; while intra-sentential switching is the type of switching which takes place within a sentence or a clause. This classification and categorisation of our findings are based on the two dominant theories to the problem of syntactic constraints on code-switching: the alternation and the insertion approach, as P. Muysken, (1994, p.156) pointed out. Thus, the theoretical background of the classification of the instances of intersentential switching found in our data is based on S. Poplack's (1980) word-order-equivalence which states that switches occur at sentence or clause boundaries. The theoretical background of the instances of intrasentential switching found in our data is based on C. Myers-Scotton's (1993a, 1993b) linguistic model for intra-sentential codeswitching known as the Matrix Language Frame Model. The first model shows that switches occur between sentences or clauses; the second model consists of the insertion of a foreign lexical or phrasal category into a given system. In her study of bilingual Puerto Rican in New York City, S. Poplack distinguishes three types of codeswitching: inter-sentential, tag-switching and intrasentential, as can be seen in figure 1 Figure 1 shows the three types of codeswitching proposed by S. Poplack (1980, p.615). She (S. Poplack, 1980, p.605) opined that tag-switching "requires least bilingual proficiency and minimal knowledge in L2". She stated that "inter-sentential switching requires more knowledge in L2 and intrasentential switching requires a high level bilingual proficiency because the speaker needs to know enough of the grammar of both L1 and L2 in order to be able to produce grammatically correct utterances" (S. Poplack 1980, p.605). Though S. Poplack affirms in her study that inter-sentential and intra-sentential are of equal use, Romaine (1995, p.123) points out that "all the three types of switching can be encountered within one discourse." Contrary to S. Poplack's (1980) classification of the types of switches, in our study the data are classified according to two types of switching, as proposed by E. Boztepe (2005): inter-sentential switching and intrasentential switching. We think that tag-switching can be grouped with intra-sentential switching because both involve inserting words from one language into another. From this point of view, we propose three representations of the types of Songhay-French codeswitching: inter-sentential switching, intra-sentential switching and reverse direction of intra-sentential switching. Thus, figure 2, as seen below, proposes to represent the inter-sentential type of code-switching of Songhay speakers of French. The two circles indicate the two languages involved in CS. The circles are of equal size, this is to fit the "equivalence Constraint" of S. S. Poplack (1980). The circles run parallel to each other; this indicates that the two languages involved in the switching are parallel constructions. The opposing direction of the two parallel arrows indicates that the inter-sentential switching is bidirectional. On the basis of this parallel construction of two different codes, it can also be stated that the production of inter-sentential constructions by bilingual speakers requires competence in both languages. Following this definition, P. Muysken (1994, p.156) explained that the phenomenon of switching in the alternation model resembles the kind under which codes are switched between turns or utterances. In this section, the kinds of sentences and clauses encountered in the instances of the inter-sentential switching of Songhay speakers of French extracted from the oral corpus of our data are described. The descriptions are based on S. Poplack's (1980Poplack's ( , 1981) ) claim for the word-order equivalence between the languages involved arguing that switching occurs at specific switch points in discourse. In other words, speakers switch codes between sentences or clauses. # b) Instances of inter-sentential switching at sentence level In the extracts below, speakers place sentences of Songhay and French side by side. They do this without violating the internal structures of the two languages. This point of view goes with S. Poplack's (1980Poplack's ( , 1981) ) word-order equivalence between the languages involved in code-switching. We can notice that the juxtaposition of the sentences in Songhay and French varies according to the situations and the speakers: In these three extracts speaker : Uhh hu! (.) Parle maintenant. Maa n'a har? "Come on! (.) Speak up now. What did he say?" 1B2 ? juxtaposes Songhay and French sentences. In the first extract, the juxtaposition consists of a declarative simple sentence in French followed with a direct question in Songhay introduced by an interrogative pronoun 'maa' (what, which). While in the second extract, the juxtaposition consists of two declarative sentences in Songhay followed with two sentences in French, a declarative complex sentence and an interrogative sentence. In the third extract, we have an interjection followed with an imperative sentence in French and an interrogative one in Songhay. The following are examples of inter-sentential switching in which speakers from Bamako, Gao and Timbuktu juxtapose sentences in Songhay and French: 1B5 ? : Ay ?i baa ye faajikaaray. (.) C'est tout! (0.2) Je peux disposer? "I don't want to chat. That's all! (.) May I leave?" T5 ? : Woodi ti nee da k'a har yer se. (.) Yer na guna ga. (0.2) On n'a pas vu ça. "It's that he told us.We didn't see it. We didn't see it." T11 As a matter of fact, it can be stated that the practice of inter-sentential switching by Songhay speakers of French is general because we have found examples of this kind of code-switching in all the five selected "grins" of our sample. We can also say that in all the examples, there appears a kind of parallel use of Songhay and French codes through sentences in both languages. This supports the claim that in the intersentential switching, the switching phenomenon is bidirectional. : Je voulais résumer, c'est tout! (0.2) Ay har, Ablo wane ciinidi ga ? "I wanted to summarize, that's all! I said, concerning what Ablo has just said?" # c) Instances of inter-sentential switching at clause level In the extracts below, speakers juxtapose in the same sentence clauses in Songhay and French in a way that the internal structure of the two languages is not violated. This juxtaposition shows clauses in Songhay and French within the same sentence where we have either the main clauses juxtaposed to the subordinate ones, or independent clauses juxtaposed to each other. For example, in the following extract, speaker 1B6 ? juxtaposes two clauses in the same sentence where the main clause is in French while the subordinate clause is in Songhay. The word 'telephone' is not considered here as a switched word, but as an integrated word in Songhay: # 1B6 In the extracts below, the juxtaposition of clauses in Songhay and French language varies according to the situations and the speakers. We can also notice that in the inter-sentential switching at clause level, like that at sentence level, the clauses are used in parallels in both codes. So, the switching phenomenon in this kind of switches is bidirectional, too: : Moi, je me demande mise ka ra boro ma huray ni doo hal a ma jaw ni telephonedi ni wane bomodi cire. "I wonder how someone could get into your house and take your phone under your head." ? 1B2 : [...] (0.4) Saadi ka a wii kur, il a mis l'as carreau 4 ? . "As soon as he won, he dropped the ace of diamonds. 1B6 ? : Ah, ay ta bay nin ka c'était trop fort, quand même. "Ah, really, ((the clause in Songhay means literally "I know that")) it was too funny, honestly." 1B6 4 One of the playing cards : Ay bara tarey yah, donnes-moi ma situation-là. "I was outside, tell me my situation." In this segment, we describe and categorise the morphological features of the instances of intrasentential switching encountered in the 18 conversations of our oral corpus. Our descriptions follow C. Myers-Scotton's (1993a, 1993b) Matrix Language Frame Model, according to which there is an asymmetrical relation between the matrix and the embedded language in the switching situation. In other words, this model states that the two languages involved in a conversation do not have the same status. One of these two languages must be the base code, the "matrix language" as C. Myers-Scotton may call it, that is, the medium of the conversation where alien words or phrases are inserted. In C. Myers-Scotton's Model, content morphemes can be inserted into mixed constituents only when congruent with the matrix language categories, while function morphemes cannot, that is to say that the grammatical morphemes must be from the base language. This kind of switching is largely the most frequent among our participants. In the following extracts, speakers insert different categories of words from French into Songhay language. The procedure looks like what P. Muysken (1994, p.156) called placing 'an alien lexical or phrasal category' into a given language. In this case, the Songhay language is the base code, that is, the medium of the conversations, and French language plays the role of embedded language, that is, the language that provides the repertoire of the 'alien' words. This proves that the intra-sentential switching is unidirectional because the conversation is directed only in one of the two codes involved. But, the reverse direction is possible when the speaker changes the direction and uses the other code instead as the medium of the conversation, as will be seen further in this article. We propose to distinguish the different categories of the 'alien' words 'imported' from French into the Songhay language in the following extracts. # e) Verbs found in the intra-sentential switching of Songhay speakers of French In the extracts below, we have noticed that the speakers insert in a Songhay-based code the past participle of the verbs in French. Consider the verb forms in the following: ? 1B3 ? : Bari, war gagné wala? "Bari, did you win?" 2B1 : Le programme-là ka damndi, (h), à la mairie de Goundam 5 # ? -là, i reconduit ga. "The programme that was planned, (h), at the town hall in Goundam, they renewed it." 2B8 Speaker : Non! A na fini deh! (0.2) "No! It's not finished deh! " 1B3 has inserted in his speech the past participle of the verb 'gagner'. The action of the verb is stated in a past tense corresponding to the 'passé composé'. We notice that the full form of the 'passé composé' which should be 'avez gagné' is not used. The first part of the verb which is 'avez' is omitted. This is because the use of 'avez' is not congruent with the grammatical structure of Songhay language which requires placing the infinitive form of a verb before the subject to express past actions, for example: Ay kaa "I have come" or "I came" Ali koy "Ali has gone" or "Ali went". Speaker 2B1 also uses the same form of the verb, the past participle of the verb 'reconduire' at the end of his speech, 'i reconduit ga'"they renewed it", in French "Ils ont reconduit cela". Speaker 2B8 , like 2B1 and 1B3 ? , uses the same form of the verb, the past participle of 'finir'. However, it appears that the verbs used in these three examples are from different groups of verbs in French. 'Gagner' is a verb of the first group, 'reconduire' is a verb of the third group, and 'finir' is a verb of the second group. So, we think that, whatever the group of the verb is, the embedded verb takes the form of the past participle in all the instances of our study where a single word is embedded as verb. But, when we have more than two verbs embedded, only the first verb form is in the past participle, the other forms follow the grammatical rules of the embedded language, like 'commencé à voler' in this extract: the intra-sentential switching, we have personally observed the speech of language teachers and communicators to find out if other forms of the following verbs like 'gagner', 'reclamer', 'coller', 'finir', 'fournir', 'reconduire', 'repondre', 'atteindre' (found in our data) can be used in an intra-sentential switching of a Songhay speaker of French instead of the past participle. From this observation, it can be stated that the past participle of these verbs is generally embedded in an intra-sentential switching of a Songhay speaker of French. Other forms of these verbs do not work in this kind of switching because they will not be congruent with the grammaticality of the Songhay language used as base code in switching practices with French. For example, in this extract: ? 2B1 If we substitute the subject 'war' (second person plural) with other subjects like 'ay' (first person singular) or 'yer' ( first person plural) or 'Ali nda Moussa', the switching will always work with the same verb form 'décollé': : Han ka war décollé? "The day you leave?" -Han ka ay décollé -Han ka yer décollé -Han ka Ali nda Moussa décollé. But, if we substitute the verb form 'décollé' with other verb forms like 'decolle'( first or third person singular) or 'decollons' ( first person plural) or 'decollent' (third person plural), the switching will not work with the grammatical system of the Songhay language: -Han ka ay décolle (is not appropriate in Songhay-French switching) -Han ka yer décollons (is not appropriate in Songhay-French switching) -Han ka Ali nda Moussa décollent (is not appropriate in Songhay-French switching). The same way, if we substitute 'décollé' with other verbs like fournir or atteindre, the switching will always work with their respective past participle forms, and other verb forms will not be congruent in this kind of switching: -Han ka war fourni. From this angle, if we consider the verb form in the following extrcat: ? 2B2 The embedded verb collé is preceded by a Songhay veb 'kaa' which is conjugated in a perfect tense with the Songhay modal verb 'go', here in the contracted form 'o'. In French, when two verbs precede each other, only the first is conjugated and the second must be in the infinitive form. This rule is violated in the Songhay-French code-switching. 'Collé', instead of the infinitive form, takes here the past participle form, because of the switching phenomenon. The infinitive is not congruent with Songhay internal system in the switching. This phenomenon is not a serial verb construction. Research needs to be done to find out if it is a case of vowel harmony. This is obvious when we substitute 'coller' with 'reconduire' or 'fournir', two verbs of second and third group: : Hey, wa koti a ra nga baadi war ma noo ga a se, yer'o kaa collé ga. "Hey, you cut his share from the banknote and give it to him, we will stick it after." ? : Le programme-là ka damndi, (h), à la mairie de Goundam-là, i reconduit ga. "The programme that was planned, (h), at the town hall in Goundam, they renewed it." 3B1 - ? : [?] (0.2) Ay ta nda ciimi, ay si hin ka har ye repondu ga quoi. (0.4) Beero, tu peux ? comme beero go nee, yaama a ma may hayka a har yer se a ra. "[...] Really, I cannot say that I am able to answer it. Beero, can you ? as beero is here, maybe he can tell us something about it." G3 : # ? [?[ (.) Deh ni ma ni objectifoo atteint, c'est ça quoi! "[...] Anyhow, you have to achieve your objective, that's right!" G14 : [...] Mais Sam, nga mo a nka dixième repris, takaa woo nda ir na Backoo 7 -Embedded verb forms with Songhay affixation Some French verbs follow the process of derivational affixation in the Songhay language. These embedded verbs are formed by adding to the French verb the Songhay derivational affix '-ndi'. In Songhay language, this derivational affix is added to a verb to form another verb, as we can notice in the speech of tee nda cere. [...] Waatidin, Sam nda ? nda Ousmane Issoufiwoo, nda Albaa?ayan ? "[...] But Sam also repeated tenth form, that's why we did the Baccalaureate together with him. [...] That time, Sam and ? and Ousmane Issoufi, and Albaasha and others ?" T11 ? where the speaker uses pure Songhay (bayndi, "to make know") compared to the speeches of 2B8, T9, T11 where the speakers embed the French verb forms with the Songhay affix '-ndi': 2B8 ? : Savoirnd'ay! Maa n'o hinse yane? "Make me know! What have you done for me?" T9 7 Short for baccalaureate : [?] Sanda woodi ti nin ? sanda woodi ti nin, nda yer'o wir ka passerndi message quoi! (.) [...] " That is ? that is in case we need to pass a message!" ? T11 f) Nouns found in the intra-sentential switching of Songhay speakers of French : [...] Mais, je crois que, njorfudi, nda i hiraw misedi ka borodiyo kaa nda, hayadi'o organiséndi! [...] "[...] But, I think if we pay correctly the money as it should be, the thing will be organised! [...]" In the extracts below, speakers insert different kinds of nouns from French into the Songhay language: compound nouns, common nouns, proper nouns, abstract nouns, collective nouns. These nouns are used either as subject or as object. However, we have found that these embedded nouns sometimes keep the Songhay inflectional affixesdi/-o/-wo (for singular nouns), -diyo/-wey (for plural nouns) which indicate the Songhay definite nouns. As for the Songhay indefinite nouns, the embedded words do not undergo any change for singular nouns, but the plural nouns take the inflectional affixes -yo/-yey (-yo in Timbuktu dialect, -yey in Gao dialect). This angle confirms C. Myers-Scotton's (1993a, 1993b) Matrix Language Frame Model, according to which the grammatical morphemes must be from the base language and 'content morphemes can be inserted into mixed constituents only when congruent with the matrix language categories'. Nouns are content words, and they adapt here the Songhay grammaticality: -Definite form of nouns, singular (-di/-o/-wo) and plural (-diyo/-wey) ? 1B2 ? : Policediyo kur ka dam constadi ka ra i har ( ) [...] "All the policemen that did the report in which they said ( )" G1 -Indefinite form of nouns, singular (bare form) and plural (-yo/-yey) : A ga Biblo feuillewey kortu deh a g'i bibiri. "He tore out the sheets of his Bible and rolled them up in pieces." # ? 2B1 ? : Ciji, yer nda ngi borodiyo bara débat timmente ra meh. "Last night, we held a long debate with them, you know." : Joueuryo ka ja ngi ?aa hay gi hal more i'g'i goro banc de touche ? "These players, since they have been playing till now, they are still sitting on the substitutes'bench ?" The adjectives encountered in the Songhay-French intra-sentential switching of our data are largely numerals and adjectives of quality. Other kinds of adjectives are used, but less frequent than numerals and adjectives of quality. Like nouns, some of these adjectives take the inflectional affixes -di/-o/-wo (for singular nouns), -diyo/-wey (for plural nouns), confirming C. Myers-Scotton's (1993a, 1993b) point of view in her Matrix Language Frame Model, according to which content morphemes, like nouns here, adapt the grammar rule of the base language. -Examples of embedded numerals ? [A too fort ra nong'ka], wala a way ka ?indi war, nda a gay hal a g'o ciini ra, a'a cii hal a ma bendi nga moo wooyo si fer. A go daabu gi nin ka cii. "He is so smart that when he wants to fool you, he feigns to close his eyes from the beginning till the end of his speech. He closes them and goes on speaking." G11 ? : Ya nka si baa ya troisième leger tee, donc, troisième fort no ay ga baa. "I don't like to make light tea, I like strong tea." G15 -Example of the embedded distributive adjective 'chaque' (each): : Il y a des gens, un moment-là, ils sont pas ? i ?ii disponible. (0.2) Waati fooyan goo no, i ga regretté ? "There are some people, at a certain point, they are not ? they are not available. After sometimes they regret ?" ? T14 h) Adverbs found the in intra-sentential switching of Songhay speakers of French : Jour di ti chaque jeudi wala? "About the day, is it on Thurdays?" Contrary to verb forms, nouns and adjectives, the adverbs we found in this Songhay-French intrasentential switching are fixed words. All these adverbs do not undergo any change. # ? 1B4 ? : Hey, kaa din ay motoodi; après kur ay nda ni go joué. "Hey, come and take my motobyke; after, we play together." T13 ? : Mais, c'est-à-dire, après hawme handudi, ma? ga ti a maa, bor'ka prêt, akoydi'o hin ka hajje ? "So, that is, after Lent, you know, any preson who is ready can ?" T12 ? : N'o may combien? "How much do you get?" FG2: ? Woo no ma bay ka? staryey ti ir. (0.2) Ay nee tellement que staryey ti ir, a si hin ka selen ya ne. "It is a fact that we are stars. I say we are stars, that's the reason why she couldn't talk with me" T7 : Ay har a boori! (0.2) Franchement, a na ti Ablo wane idée woo foo, puisque war mara woo jaatindi In the conversations of Songhay speakers of French, speakers sometimes reverse the direction of the conversation using alternately both languages as base code in their discourse. This does not challenge, anyhow, the unidirectional character of the intrasentential switching because the conversation is directed only in one of the two codes involved. The speaker only changes the direction and uses the other code instead as the medium of the conversation. In our data, we found some examples of intra-sentential switching where the base code of the discourse is not Songhay language as seen in the extracts above, but French instead. In these examples, the Songhay embedded words generally express emphasis, excitement, surprise, joy, anger, agreement or disagreement, time, and other exclamations. # Conclusion : Ay ta, je suis quelqu'un qui me déplace beaucoup. (.) [...] "As for me, I am particularly always on move. (.)" The issue of code-switching and code-mixing is a language phenomenon facing all languages in contact situation. In Mali, it is a consequence of the multilingualism that characterises the Malian linguistic landscape and the Malian society as well. The data have revealed that the practice of code-switching and codemixing by Songhay speakers of French is becoming general. Table 7 shows the final results of all the participants involved in the study with 1169 total amount of utterances or speeches grouping 539 speeches of code-switching and code-mixing representing 46%, 443 speeches in native language only representing 35% and 220 speeches in French-only representing 19%. The use of code-switching and code-mixing is higher than the use of the native language, and French-only is less used. The table shows also that the use of intrasentential switching is considerably higher (82%) than the use of inter-sentential switching (18%). As a matter of fact, this predominance of intra-sentential switching confirms the "size of constituent constraint" of S. Poplack (1980Poplack ( , 1981) ) and supports the claim that smaller language units like nouns or verbs tend to be more switched than larger ones like sentences or clauses. # 46% Theories on code-switching and code-mixing claimed by S. Poplack and C. Myers-Scotton have been revealed to be applicable to the different forms of the two types of switching in the study of Songhay speakers of French. Among these types, the intra-sentential one abounds much, representing 82% of the total number of the speeches of our data, as shown in table 7. This type of switching, based on C. Myers-Scotton's (1993a) insertion model is, according to P. Muysken (1994, pp.156-162) "typical of the African code-switching materials". The predominance of intra-sentential switches is evidence that confirms S. Poplack's (1980Poplack's ( , 1981) "Size of Constituent Constraint" according to which smaller language units like nouns, determiners, verbs, adverbs, and adjectives are more frequently switched than larger size of constituents such as sentences and clauses. It also supports the works of other researchers like S. Berk-Seligson's (1986) study on Hebrew-Spanish, C. W. Pfaff's (1979) S. Poplack's (1980Poplack's ( , 1981) ) and L. A. Timm's (1975) studies on English-Spanish where smaller size of switches like nouns, verbs, adverbs, adjectives, etc. were had been found to be more frequently used than larger ones. The study has found verbs, nouns, adjectives, adverbs, conjunctions and exclamations in French language that are inserted in Songhay-based conversations. We have also noticed that most of the inserted verb forms are past participles, and this concerns all the three French verb groups. Among the categories of the intra-sentential switching, the predominance of nouns is also noticeable throughout our data, though we did not consider any statistics to show their exact number and suggest future research to do so. However, the most striking finding in our study is the frequent use of the past participle of the verbal forms of switches. But, this claim should be taken with care, for more research needs to be done in order to investigate the grammatical categories of the Songhay-French intra-sentential switching using other data collection procedures. We recommend that future research studies be done to generate new data and new findings on the problem of code-switching and code-mixing of Songhay speakers of French for further refinements. It will also be interesting to undertake more research work on codeswitching and code-mixing issues on other languages in the Malian linguistic landscape such as code-switching of Songhay speakers of Bamanankan, Bambara speakers of French, Fulfulde speakers of French, Fulani speakers of Songhay, Fulani speakers of Bambara, Songhay speakers of Kel-Tamasheq and codeswitching between French and other Malian national languages. Though various efforts have been made and are being made, the government must support any action aiming at protecting, preserving and promoting the national languages. 1![Figure 1: The types of code-switching and the degree of switching in them (S. Poplack 1980, p.615)](image-2.png "Figure 1 :") 2![Figure 2: Inter-sentential Songhay-French code-switching](image-3.png "Figure 2 :") 3![Figure3proposes to represent the intrasentential type of code-switching of Songhay speakers of French. The two languages involved in CS are represented in two different forms. These forms are not of equal size. The larger one indicates the base language or matrix language. The smaller one indicates the language that provides the inserted words. This is to fit the "Matrix Language Frame Model" of C. Myers-Scotton (1993a, b). The direction of the arrow indicates that words from French are inserted in a Songhay-based code. The switching is so unidirectional.](image-4.png "Figure 3") 3![Figure 3: Intra-sentential Songhay-French code-switching a) Instances of inter-sentential switching of Songhay speakers of French This segment deals with the structural description of some instances of inter-sentential switching encountered through the 18 conversations of our oral corpus. It follows S. Poplack's (1993) definition of code-switching according to which code-switching is 'the juxtaposition of sentences or sentence fragments, each of which is internally consistent with the morphological and syntactic (and optimally, phonological) rules of the languages of its provenance'.Following this definition, P. Muysken (1994, p.156) explained that the phenomenon of switching in the alternation model resembles the kind under which codes are switched between turns or utterances. In this section, the kinds of sentences and clauses encountered in the instances of the inter-sentential switching of Songhay speakers of French extracted from the oral corpus of our data are described. The](image-5.png "Figure 3 :") ![Journals? 1B2 : On joue à 'Cent2](image-6.png "") ![A boori! Ay ta si cii ni doo mo. (0.2) Je garde seulement le silence. "Fine! I am in no position to comment. I have to keep silent." 2B3 ? : Non! I si hin ka ?indi yer. (.) I si hin ka ?indi yer. (0.2) Nous, on vient de commencer. (0.2) "No! They cannot fool us. They cannot fool us. We have just begun." 3B20 ? : Attendez! (.) C'est pour cela que j'ai dit qu'on laisse faire. (0.2) Maa se ne ay har yer'o nan ga? [?] Donc, en ce moment-là on doit ? on doit quand même parler de ça. "Wait! That's why I proposed to let things take their course. Why did I propose so? [...] So, from now on we must ? we must really talk about it." 3B23 ? : [...] On peut décider de dire que, bon voilà: 'yer sanba war se woo, war ma faaba ga nda bomo' [?] " We can decide to say the following: 'we send you this to assist you'." G9 : ?iiya! ?iiya! Il est toujours chef D.E.R3 3 ? , prof., section tronc commun à l'E.N.A. (.) [?] "Stop! Stop! He is still the Head of Department of the common-core syllabus section at E.N.A ((Ecole Nationale d'Administration)). G1 2 One of the belote games 3 Department of Teaching and Research : Non! Avant, avant! (.) Man'ti sohô da g'ay go. "No! Before, before! It's not now I mean." ? FG8](image-7.png "?:") 4![Ay nee, je peux disposer? "I say, may I leave?" G12? : Bara boro fooyan ga ni wow ka? ?i hin ka ni wow, parce que tu as accepté d'être devant ? devant les gens. (.) "And some people insult you who may not insult you, just because you have accepted to be their leader. (.)" G13? : C'est pour dire que cawyan sii nda ? haywana ? "That's to say that for studies, there is no ?" T12d) Instances of intra-sentential switching of Songhay speakers of French: Bon! (.) C'est pas forcément bara akoydi ma bana njerfu. [...] "Well! It's not compulsory that the person should pay a fine."](image-8.png "4 )") ![To support our claim that Songhay speakers of French use the past participle of the embedded verb in : Non! I si hin ka ?indi yer. (.) I si hin ka ?indi yer. (0.2) Nous, on vient de commencer (0.2). Yer na ti kala un petit oiseau (0.2). Mise ka yer g'o yer'o sinti ka commencé à voler (0.2) [?] "No! They cannot fool us. They cannot fool us. We have just begun. We are like a little bird. We are just trying to start flying."](image-9.png "2B3") ![.. -Han ka war atteint. .. But, not: -Han ka war fournissez... -Han ka war fournir... -Han ka war atteingnez... -Han ka war atteindre...](image-10.png "") 6![Wani n'ka joué hô? "Has Wani played today?" G1? : I nee Ha-Ko ka gagné un à zéro. "They said that Ha-Ko won one to zero." G1? : Ayyo! Baa hano ka? ngi nda Mazaa borey joué là, i yenje. "Yes! Even the day when they played against Mazaa people, they quarrelled." FT4 ? : Sévaré, i g'i préparé gi ka koy yenje. "In Sevare, they are getting prepared to go to fight." FT2 ? : Eh! Wala ? wala RFI cindi ka félicité gi hô, ma na mom partiedi? "Eh! Even ? even RFI ((Radio France Internationale)) has congratulated them today, didn't you listen to the related part?" T12 : Ay ta, wallaahi, Dra traumatisé ay ? wallaahi, Dra traumati ? "You know, truly, Dra has traumatized me ? truly, Dra's traumati ?" T14 6 Short for Abderhamane : Walidi hasara ay ga da. (0.2) Ye baa ay ma racheté. (0.2) La vie c'est comme ça. "Honestly, I lost the game. I want to redeem my honor. Life is like that." ? Examples of embedded verbs of second group like 'finir' (to finish) ? 2B8 ? : Non, a na fini deh! (0.2) "It's not finished deh!" G14 ? : [?] (0.2) Jina, nda a sinti ka couroo tee ma ne, nda ni si effort fourni deh a ga ni couroo kayandi. "[...]. Moreover, if he begins your course, and you don't make any effort, he stops your course." T11 Examples of embedded verbs of third group like 'reconduire' (to renew) : [?] (0.4) Woo mo, éclaici ga borodiyo se. (.) Est-ce que ni faham? (0.4) [...] "You have to clarify that also to everybody. Do you hear me? ?" ? 2B1](image-11.png "FG2?: 6 ?-") ![Woo no ma bay ka? staryey ti ir. (0.2) Ay nee tellement que staryey ti ir, a si hin ka ?elen yane. "It is a fact that we are stars. I say we are stars, that's the reason why she couldn't talk with me."FG2 ? : Eh, Fifi goo nda forme! "Eh, Fifi has a graceful figure!" T7 -Other examples of embedded definite form of nouns, singular (-di/-o/-wo) and plural (-diyo/-wey)](image-12.png "FG2?:") ![Examples of embedded nouns used with French determiners : Biifo, ay guna a ga telephonedi, a go cii kuna ? "The day before yesterday, I saw the telephone with him, he was calling ..." ? 1B1 Koy kuboy yane ay wane ijemayroyo deh à l'école publique (0.4). Ni nda woo ma koy nda cere, le petit-là. "Go and take my children at the public school. You go with this one, that little boy." 2B3 ? : Non! I si hin ka ?indi yer. (.) I si hin ka ?indi yer. (0.2) Nous, on vient de commencer. (0.2) Yer na ti kala un petit oiseau. (0.2) [...] "No! They cannot fool us. They cannot fool us. We have just begun. We are like a little bird." FT2: g) Adjectives found the in intra-sentential switching of Songhay speakers of French [...] Guna la différence. [...] "You see the difference."](image-13.png "FT2-?:") '. (.) Maa n'o wii ga nda?"We're playing 'Cent'. Which card should youplay?"?1B2? Yer'o kaa reconduire ga. (is not appropriate inSonghay-French switching)-Yer'o kaa fournir ga. (is not appropriate in Songhay-French switching)But,-Yer'o kaa reconduit ga.-? Examples of embedded verbs of first group like'gagner' (to win)? 2B8: Muusa na cii (0.2) Muusa na cii, ay gareclamé! "Moussa didn't say anything. Moussadidn't say anything, but I claimed!"?G1: Baa zaarikayo a cindi ka connecté. "Even thismorning she was connected."?G9: Sanda, boro no a na dam a ra ka? g'a géré."Well, he engaged someone who managed it."?G1: Ah, donc, ni n'k'a renforcé deh! "Ah, so, youjust reinforced it!"? 1Speeches / utterancesNumberCityCS and CM Intersentential IntrasententialOnly Songhay Only FrenchTotal1Bamako692472561827542Gao158974141923Timbuktu121078024223Total96539443410 410220 2201169 1169Percentage18%82%35%19%100% © 2019 Global Journals ka ye guna, c'est vraiment très intéressant! (.) [?] "I say, well done! Honestly, in addition to Ablo's idea, the meeting itself, it is really very interesting!" ? T12 : Bon! (.) C'est pas forcément bara akoydi ma bana njerfu. (0.2) Laissons ça! "Well! It's not compulsory that the person should pay a fine. Forget it!" ((Here, speaker T12 i) Conjunctions found in the intra-sentential switching mistakenly uses this adverb instead of the adjective forcé)) The conjunctions encountered in the Songhay-French intra-sentential switching are fixed words, and like the adverbs, these embedded conjunctions also do not undergo any change in their initial form. ? : Yer kondey woo, yer ti talkayo kondey, mais boro kur ka wirci dam a ga, wala bun, wala ijemaa, yer kur'o koy. "Our association is an association of poor people, but any member who got sick, or a death in his family, or naming ceremony in his family, we all go to his place." Jigiba, that is, Jigiba's voice, when he is speaking ?" : ?ine nda Jigiba, i kur mana baa cere, c'est-àdire quoi, Jigiba jindo, nda a ?elen mane quoi ? "There is no much difference between Shin and ? G13 j) Exclamations found in the intra-sentential switching : Mazaa borey là, i ga yenje parce que, (h) n' ga bay, c'est des villages contigus () "Mazaa people are quarrelsome because, you know, they are adjoining villages ( )" We have found many interjections used to express abrupt, emphatic or excited utterances: * Understanding the Phonological Processes at Work in Language Contact Situations: the Case of Songhay-French Code-Switching, IJELLH, Volume V, Issue I IbrahimaAbdoulaye MohamedMinkailou 2017 * The language factor in codemixing EAnnamalai International Journal of Sociology of Language 74 1989 * The syntax of Arabic-French code-switching AbdelâliBentahila EirlysEDavies Lingua 59 1983 * Linguistic constraints on intra-sentential code-switching: a study of Spanish/Hebrew bilingualism SusanBerk-Seligson Language in Society 15 1986 * Linguistic Fieldwork: A Practical Guide ClareBowern 2008 Palgrave Basingtoke, U K * Conversational codeswitching JohnJGumperz Discourse Stategies 1982 Cambridge University Press * RichardAHudson Sociolinguistics 1980 Cambridge University press * The Indianization of English: The English language in India BrajBKachru New Dehli Oxford University Press 1983 On mixing * Investigation into Bamanankan-French Code Switching among University Students in Bamako, IJELLH MohamedMinkailou IbrahimaAbdoulaye Issue VIII 2016 IV * Investigating Fulfulde-French Code-switching among Fulani University Students in Bamako MohamedMinkailou IbrahimaAbdoulaye 2018 6 * Bilingual speech: A typology of code-mixing PieterMuysken 2000 Cambridge University Press Cambridge * Duelling Languages: Grammatical structure in codeswitching Myers-Scotton Carol 1993b Clarendon Press * Common and uncommon ground: Social and structural factors in codeswitching Myers-Scotton Language in Society 22 1993a Carol * Codeswitching between Persian and Swedish LeilaNaseh Eurosla 7 1997 * Constraints on language mixing. Language CarolWPfaff 1979 55 * Variation Theory and Language Contact ShanaPoplack American Dialect Research D. Preston 1993 Benjamins * SuzanneRomaine Bilingualism 2 1995 nd ed. * Grammatical constraints on code-switching: Evidence from Hindi-English RajendraSingh Canadian Journal of Linguistics 30 1985 * The syntax and psycholinguistics of bilingual codemixing ShikaripurNSridhar KamalKSridhar Canadian Journal of Psychology 34 1980 * Spanish-English Code Switching: El porque y how-not-to LenoraATimm Romance Philology 28 4 1975 * Introducing Language and Society PeterTrudgill 1992 Clays Ltd St Ives * UlrichWeinreich Languages in Contact:Findings and Problems. The Hagues: Mouton 1953 * Issues in Code-Switching: Competing Theories and Models. Teachers College, Colombia University Working Papers in TESOL & Applied Linguistics EBoztepe Retrieved March 3 rd 2005. dec. 2005. 2011from 3 * Deuxième Congrès mondial de linguistique francaise ChantalLyche IngseSkattum LeDans Le Francais Du Mali: Une Étude Perceptive I: FNeveu ;VMuni Toke ;Jacques ;TDurand ;LKlingler &SMondada Prévost T EDP Sciences December red. * The Role of a Dominant Research Experience in Shaping one's Vision, Ljouwert PieterMuysken Fryske Akademy 1994. March 3 rd 2011from * Sometimes I'll start a sentence in Spanish y termino en español: toward a typology of code-switching ShanaPoplack Linguistics 18 7/8 1980. March 3 rd 2011from * Syntactic structure and social function of code-switching ShanaPoplack Latino Language and Communicative Behavior RDuran New Jersey Ablex Publication Corp 1981. March 3 rd 2011from