Code-Switching in Relation to Other Language-Contact Phenomena: A Theoretical Account

ABSTRACT


INTRODUCTION
Bi-/Multilingualism can be considered as a normal linguistic situation where two or more languages are in contact. This field is an interdisciplinary and complex area of study, and it generally concerns itself with the study of the interaction and use of two or more languages in terms of speech production, language processing, and comprehension. A number of scholars have been interested and studied bi-/multilingualism from different perspectives. The pioneering phase of bi-/multilingualism started in the late half of the 20 th century with the works of Weinreich (1953), Haugen (1953), and Mackey (1967). Research on bi-/multilingualism has recently become a norm, especially that the majority of the world countries are bi-/multilingual. According to the statistics with regard to the world's languages, there are 7,388 living languages 1 comparing it to the number of countries in the word, which are approximately one hundred and ninety-five (195) sovereign states, according to the U.N. Here, we can deduce that there are more languages than countries. Therefore, bi-/multilingualism is and will be widely spread. This fact that there are more languages than countries makes of the field of bi-multilingualism a richer and more interesting ground for linguistic studies. More than that, recent studies have shown that bi-/multilingualism is healthy as it "brings opportunities not only to the individual but also to the society as a whole" (Li, Dewaele, & Housen, 2002, p. 3). A number of linguistic behaviors have come out of bi-/multilingual situations such as code-switching, code-mixing, borrowing, diglossia, and interference. CS has been celebrated most in literature. It can be said that CS has been the central issue among other language contact phenomena in the literature on bi-/multilingual research. Milroy and Muysken (1995:7) state that CS is probably the "central issue" in bi-/multilingualism. This is confirmed by Riehl (2005) who advocates that most of the research done on bilingualism centers around the phenomenon of CS. In the same vein, Bullock and Toribio (2009) stated that "of all of the contact phenomena of interest to researchers and students of bilingualism, code-switching has arguably dominated the field" (p. 1). In comparison with other language-contact phenomena, some linguists consider CS as a subfield of bilingualism (e.g. Avaliable at: www.ijssers.org Dabane, 1995, p. 125). In this regard, Gardner-Chloros (1995, p. 68) argues that CS is such a "broad blanket term" for a number of intra-lingual phenomena. In the same vein, Milroy and Muysken (1995, p. 7) state that CS is the focal point of research on bilingualism; they depict it as a cover term that subsumes different forms of bilingual linguistic behavior. Therefore, a definition of CS is required to get more insights into its nature and position in bi-/multilingual studies. Then, an exploration of CS and its relationship with other languagecontact phenomena is proceeded.

II. DEFINITION OF CODE-SWITCHING
Numerous attempts have been made to come up with a comprehensive and precise definition of CS, but this task has proven to be intricately difficult. There is no definite definition of the phenomenon of CS; rather there are several definitions that tap on different facets of CS phenomenon. Poplack (1995) defines CS as the "juxtaposition of sentences or sentence fragments, each of which is internally consistent with the morphological and syntactic (and optionally phonological) rules of its lexifier language" (p. 200). According to Myers-Scotton (1993, p. 4), CS is defined as the "selection of bilinguals or multilinguals of forms from an embedded language (or languages) in utterances of a matrix language during the same conversation". Hence, for Myers-Scotton, CS involves a matrix or dominant language and an embedded one. Furthermore, she emphasizes on the aspect of proficiency in both languages. She states that CS is a "type of skilled performance with communicative intent" (1995, p. 7). In the same vein, Rasekh et al (2008, p. 552) argue that CS takes place when competent bilingual interlocutors share knowledge of the two languages well enough to distinguish terms from either language at any moment during the interaction (Gimode, 2015, p. 23). According to Fishman (1999, p. 147), CS is defined as the "alternate use of two or more languages in the same utterance of conversation". Before that, Gumperz (1982, p. 59) describes CS as a conversational act. He defines CS as the "juxtaposition within the same speech exchange of passages of speech belonging to different grammatical systems or subsystems". This concept of conversational CS overlaps with other definitions commonly used to refer to CS, but Gumperz deals with CS more as a product of social settings or situations that guide the language use. Auer (1995) agrees with Gumperz (1981) that CS is socially driven, and he suggests the notion of contextualization as a model to account for conversational CS. In the same line, Giles (1984) also contributes to the understanding of social meaning of CS through his theoretical framework of speech accommodation theory, later dubbed as communication accommodation theory. For Giles' theory, during the social interaction, interlocutors "are motivated to adjust (or accommodate) their speech styles as a means of evoking listeners" social approval, attaining communicational efficiency between interactants, and maintaining positive social identities" (Beebe & Giles, 1984, p. 7). From this basis, Giles sees that the accommodation theory is highly applicable to account for the social motivations of CS. As far as Wardhaugh (2010) is concerned, he sees that CS "can occur in conversation between speakers" turns or within a single speaker"s turn. In the latter case it can occur between sentences (inter-sententially) or within a single sentence (intra-sententially)" (p. 98). In the same line, Muysken (2000) also sees that it is noteworthy to distinguish between these two different types of CS, namely inter-sentential and intrasentential CS. For Muysken, the intra-sentential CS is referred to as code-mixing. Annamalai (1989) advocates that CS takes place during a unit of discourse, but CM "is not normally done with full sentences from another language with its grammar" (p. 48). According to Bentahila and Davis (1983), CS "is the use of two languages within a single conversation, exchange or utterance"; they argue that the "act of choosing one code rather than another must be distinguished from the act of mixing the two codes together to produce something which might itself be called a third code" (p. 302). Based on what is aforementioned, one can conclude that CS is a language phenomenon that is natural in bi-/multilingual contexts. It is the outcome of a situation in which two, or more, language varieties are in contact and used alternately back and forth in a given speech or utterance. This alternation between language varieties can take place at the level of a discourse, turn, utterance, constituent or a marker. Moreover, CS can be studied mainly from two perspectives, a linguistic perspective and a social one. The former tries to identify the linguistic principles and structural constraints that govern the production of code-switched utterances while the latter deals with the social motivations and functions that lead to the output of CS. Therefore, CS is a cover term that may encompass other linguistic phenomena of language contact such as come mixing. In what follows, we will discover CS in its relation to other language contact phenomena.

III. CODE-SWITCHING AND OTHER LANGUAGE CONTACT PHENOMENA
A. Code-switching versus code-mixing CS and CM have been treated differently by researchers. Some of the latter use them interchangeably while others insist on making a distinction between the terms. According to Bokamba (1989 In the same line, Tay (1989) distinguishes between the two linguistic phenomena. The difference, for Tay, resides in that CS is carried out across the sentence boundaries, while CM is conducted within the same sentence and the same speech situation. Gardners-Claros (2009) agrees that the distinction between CS and CM is based on the "spot" at which the switch takes place. For her, CS takes place intersententially when "a bilingual speaker uses more than one language in a single utterance above the clause to appropriately convey his/her intents" (p. 68). On the other hand, CM is an intrasentential alternation which involves more than one language used "below the clause level within one social situation" (Gardners-Claros, 2009, p. 69). Thus, CS occurs between utterances while CM takes place within an utterance. In the same vein, Mashiri (2002) differentiates between CS and CM in the sense that in the former, the languages involved in the switching keep their morphological and phonological attributes while in the latter (CM), the embedded language items occur in the matrix language sentence, obeying the placement rules of that matrix language.
In distinguishing CS and CM, Wardhaugh (1986) sees that CM is conducted when the participants use and exchange both languages in the course of a single utterance. For him, CM taps on various linguistic levels such as lexical items and morphology, without changing of the topic. The same was argued by Annamalia (1989) who proclaims that the difference between CS and CM lies not only on grammatical aspects of the exchanges because no new grammar is created beyond the grammars of the languages involved in the switching sites. According to him, in CM, there is no variation in topic and participants, and all of the latter share knowledge of both languages. He also argued that CM is a linguistic or discourse strategy that necessitates a kind of language competence of the speakers. The same was argued by Moradi (2014) who advocates that CM involves the alternation of two languages within a sentence. This language alternation is fluent, rapid, and unhesitant; it reflects the output of a bilingual who is competent in both languages; therefore, CM shows the ability of the speaker to use elements of each language involved alternately within a sentence. As far as Sridhar and Sridhar (1980) are concerned, CM can be distinguished from CS in two ways; the first is that CM is not accompanied by a shift or change in the speech situation, and the second is that it takes place at the intra-sentential level. Tay (1989) agrees with Sridhar and Sridhar in the sense that CM takes place intra-sententially within the same speech event or situation while CS occurs across sentence boundaries.
Another perspective of distinguishing CS and CM is regarded with formality of the situation. It is generally argued that CM is conducted in less formal situations, whereas CS likely occurs in more formal ones. In this regard, Kachru (1978) distinguishes between the terms on the basis of function or motivation. He argues that CS takes place from a 'standard' variety to a dialect one; it serves the function of showing solidarity or disapproval and so on. Conversely, CM takes place informally, and it often involves the exchange between a local variety and a more socially accepted and prestigious one, like English in order to show the ability to use that prestigious variety as a form of education. He argues that CM involves the transference of linguistic units from a language to another, which may result in a new code of interaction, such as Westernized Hindi.
For Mazraani (1997, pp. 8-9), the difference between CM and CS is that the latter has a discourse function; she states that CS is a phenomenon in which "sections in one code are followed by sections in another in the same conversation". On the other hand, CM involves "the mixing of different varieties within a single utterance". For her, CM does not affect all linguistic levels such as syntax and phonology. Likewise, Ugot (2010) sees that CS and CM are generally expected because they respond to communicative needs and call for adaptability of languages to respond to these needs in social interaction. For him, CS involves "the lifting of phrasal, clausal or sentential structures" (p. 29). Thus, syntactically, CS happens when diverse utterances from distinct languages establish one discourse, whereas CM marks the infusion of single items from the donor language into the first language or mother tongue (L1) construction (Ugot, 2010).
According to Muysken (2000, p. 109), the term CM can be used to refer to all cases when items of a language are used while the speaker(s) use another. In his book "Bilingual Speech: A Typology of Code Mixing", Muysken (2000) sees that CS can be equated with one of three proposed types of CM, namely alternation. For him, CS is not an inclusive and appropriate term as it separates other language contact phenomena such as borrowing. Hence, he avoids using CS; instead he prefers to use CM as a neutral term. He advocates that CM can refer to "all cases where lexical items and grammatical features from two languages appear in one sentence" (Muysken, 2000, p. 1). However, in their work later, Rene and Muysken (2005) admit that there is a distinction between CS and CM and that the latter can be counted as an intra-sentential form of CS; this latter is seen as an overall process of alternation of codes (cited in Gimode, 2015, p. 24).
On the other hand, Myers-Scotton (1993) does not see a necessity in trying to distinguish between CS and CM because this distinction creates an "unnecessary confusion". She states that: A number of researchers associated with Braj Kachru[...], but also some others, prefer to label as 'code-mixing' alternations which are intrasentential, although it is not entirely clear whether this applies to all intrasentential CS (codeswitching). While I grant that intrasentential CS puts different psycholinguistic 'stresses' on the language-production system from intersentential (code switching) CS (a valid reason to differentiate the two), the two types of CS may have similar socio-psychological motivations. For this reason, I prefer 'CS' as a cover term; the two types can be differentiated by the labels 'intersentential' and 'intrasentential' when structural constraints are considered. (Myers-Scotton, 1993, p. 1) All in all, one can conclude that considering CS and CM as different language processes or phenomena is needless and nonessential because CM can be considered as part of CS. The definitions of the terms by different researchers render CS as a more inclusive term which covers also CM. Therefore, CS indicates both intra-sentential and intersentential switching of codes along with between longer stretches of texts, but it does not include borrowing to refer to foreign words that have been integrated into the system of another language. In what follows, the term borrowing is further discussed and its relation to CS.

B. Code-switching versus borrowing
The notion of borrowing and its relation to CS has generated an intense debate and resulted into controversial standpoints. This controversy has risen to whether CS and borrowing should be treated as different or the same entities. This problem can be linked to what Weinreich, Labov, and Herzog (1968) call the transition problem: since languages change diachronically, we cannot locate at what particular point a lexical item becomes borrowing or loanword in the recipient language. However, a number of researchers have attempted to put certain criteria that would help in distinguishing the two terms.
A group of researchers associated with Poplack (1981Poplack ( , 1987Poplack ( , & 1990 2 argue that CS and borrowing are two different mechanisms. They propose three types of criteria to determine whether a lexical item or foreign word can be classified as borrowing or CS. These criteria are morphological integration, syntactic integration, and phonological integration. According to them, a lexical item cannot be counted as borrowing unless it is morphosyntactically and phonologically integrated into the recipient language. According to this approach, if the integration of a lexical item taps only one or two of its criteria (morphology, syntax and phonology), that lexical item is considered as an instance of CS, not borrowing. Therefore, for Poplack and Meechan (1998), CS does not require any morphological or syntactic integration, code-switched items still keep their 2 i.e. Poplack, Wheeler, and Westwood (1987), Sankoff and Poplack (1981), and Sankoff, Poplack, and Vanniarajan (1990). For abbreviation morphosyntactic and phonological form as well as grammatical patterns when they are used, whereas borrowings require morphological and syntactic as well as phonological adaptations of the grammar of the recipient language (Poplack and Meechan, 1998, p. 132). As far as Halmari (1997, p. 173) is concerned, phonological assimilation can be seen as the main, if not the only, determining feature of borrowing.
Determining borrowing from CS based on morphosyntactic and phonological adaptation is not freed from criticisms. According to Myers-Scotton (1992), "while most established forms may be well phonologically integrated to ML by no means do all borrowed forms show such integration" (p. 31). In this regard, Romaine (1995, p. 601) uses the example of chips that has become a borrowing in Punjabi, and it has not undergone any morphological or phonological assimilation. The same point was raised by Pickeles (1999) who argues that there is a cross-linguistic influence in the way /r/ is pronounced by Maghribi students in France, which is not case of borrowing; otherwise, every French lexical item which contains the sound /r/ would be deemed as borrowing (cited in Aabi, 1999). Meanwhile, not all morpho-syntactically integrated lexical items are borrowed forms. For instance, the syntax of expressions like bon appétit remains intact when borrowed from French to English. Bentahila and Davis (1991, p. 384) raised some instances of morpho-syntactically integrated French words into MA system, but they fall within CS phenomena. For example, "Taymshiw l-la toilette bash ypissi-w''. Here, even if the word pisser is adapted to MA morphological system; still it is not considered as borrowing. Hence, this three-fold criteria reported by Poplack and others is very significant, but not enough for setting a clear-cut boundary between CS and borrowing.
As pointed out above, the mophosyntactic and phonological integration seem to be very essential for distinguishing borrowing from CS; however, other factors may come into play in this regard. Myers-Scotton (1992& 1993 argues that a categorical distinction between borrowing and CS is not necessary, but then she sees that frequency can be the best single criterion to associate borrowed forms to the recipient language (cited in Boztepe, E., 2005). Therefore, frequency is another essential factor for making distinction between borrowing and CS. Borrowed forms are used more frequently in the community and, thus, become established in the host language, while CS is considered as a momentary use of language by a bilingual speaker. Hence, CS is typically an idiosyncratic alteration of languages by bilinguals. Conversely, borrowing is more established forms in the host language that can be used not only by bilinguals but also purposes, these authors will be referred to as "Poplack and her associates." monolinguals of the recipient language. Besides, Myers-Scotton (2002) sees that the distinction between the two terms should not be the key and critical concern in analyzing bilingual speech, and she states that the sole sense in which a difference is to be made between them is regarded to their status in the mental lexicon: "Lemmas underlying code switching forms are only tagged for the embedded language, while borrowing forms have lemmas for both the donor and the recipient language" (Myers-Scotton, 2002, p. 153).
Unlike Bentahila and Davis (1983), Sridhar and Sridhar (1980) and others who consider CS and borrowing as distinct phenomena, Myers-Scotton (1992, 1993) and Matras (2009) see that the concepts as a universally related processes and part of a single continuum. Myers-Scotton also disagrees with the researchers mentioned above who claim that the main characteristics of borrowed forms is to fill lexical gaps in the recipient language. For her, not all borrowings take place owing to the gap or absence of an equivalent term in the recipient language. This was supported by Haugen (1953) who stated that "borrowing always goes beyond the actual 'needs' of language" (p. 373). Consequently, Myers-Scotton (1993) came up with the idea of distinguishing between core borrowings and cultural borrowings. Core borrowings are those lexical items which have their "viable" counterparts in the recipient language. Cultural borrowings, on the other hand, refer to the lexical items which are new and unfamiliar to the recipient language. Another classification of borrowings worth discussing is nonce versus established borrowings.
According to Poplack & Sankoff (1984), there are two types of borrowings: established loanwords and nonce borrowing. The former refers to lexical items which are fully integrated and frequently used; they have become part of the recipient language. Grosjean (1982) defines them as elements from a donor language which are "integrated phonologically and morphologically into the base language" (p. 127). On the other hand, none borrowing refer to the use of lexical items of a language which have not yet integrated and become an established part of the host language. This concept was first introduced by Haugen (1950) and, then, has been taken up by Poplack and Sankoff (1984) and Poplack and Meechan (1995). For Muysken (1995, p. 190), nonce borrowings are items which are "borrowed on the spur of the moment", without receiving any status in the host language yet. Riehl (2005Riehl ( , p. 1947) sees that nonce borrowing can be equated with idiosyncratic loans. Some researchers such as Bentahila and Davis (1991), Bokamba (1988), and Myers-Scotton (1993) disagree with the notion of nonce borrowing. They generally argue that any lexical item used from a language and it is not yet "established" and integrated in the system of the host language should be treated as a CS element, not a borrowed form. For them, nonce borrowing is simply equivalent to CS; it is just extraneous and does not add any explanatory value to the study of language contact phenomena.
As a conclusion, the distinction between CS and borrowing is not straightforward and easy. However, one can deduce that borrowed items refer to elements taken from a language and have been morpho-syntactically and phonologically integrated into a host language. These elements can be used by both bilinguals and monolinguals of the host language. On the contrary, CS refers to the incorporation of foreign language elements in a base language without necessarily being morpho-syntactically and phonologically adapted to the host language system, so code-switched items are not conditioned by the grammar of the recipient language. While the validity of morphosyntactic and phonological integration criteria has been argued among researchers, still these criteria are crucial for distinguishing the two concepts, yet they are not always sufficient. Therefore, other criteria may be evoked for this distinction such as the language use, especially in terms of frequency and speech community acceptance of the loanwords. Other researchers such as Myers-Scotton (1993) and Heath (2001) see that it is not necessary to search for any differences between CS and borrowing as they are related phenomena and part of the same "developmental continuum". Moreover, this distinction is not crucial for the study of bilingual speech. In this regard, Saib (1989, p. 48) states "En effet MC (mélange de codes), dans le contexte marocain, a trait ä un continuum allant de 1'emprunat PC (permutation de codes)" (Aabi, 1999, p. 10). This continuum process is not only notable in the alternation between Arabic and French, but also between Amazigh and Arabic and between the Arabic varieties, e.g. MA and SA; this kind of switching is called diglossic switching. The following point discusses CS in relation to diglossia.

C. Code-switching versus diglossia
CS and diglossia seem to account for different phenomena of language contact. However, there are some common meeting points of the concepts. The term diglossia is more attached to Ferguson even if the term diglossie was used before Ferguson by the French linguist Marcais (Fasold, 1984, p. 34). According to Ferguson, diglossia is defined as: …a relatively stable language situation in which, in addition to the primary dialects of the language (which may include a standard or regional standards), there is a very divergent, highly codified (often grammatically more complex) superposed variety, the vehicle of a large and respected body of written literature, either of an earlier period or in another speech community, which is learned largely by formal education and is used for most written and formal spoken purposes but is not used by any sector of the community for ordinary conversation. (Ferguson 1972(Ferguson [1959, p. 345) Based on the definition above, one can deduce that diglossia involves two varieties of a language; a 'high' variety and a 'low' one. The former refers to the standard form of the language in question. This standard variety is typically the official language which is learnt in formal education; then, it is limited to formal contexts by the educated elites; whereas the colloquial form of the language ('low' variety) is mostly used in informal domains for everyday conversations by ordinary people. Ferguson (1972) asserted that it is essential to treat the two varieties in diglossic situation as differently allotted functions within the speech community (Fasold, 1995, p. 35). The gist here is that there is a strict distinction of the domains of language use in diglossic situation.
Ferguson's definition has not been welcomed without criticisms. A number of questions are raised as a reaction to Ferguson's definition such as how close together and how far apart should the 'low' and 'high' varieties be for a language situation to be called "diglossia". This question was raised by Fasold (1995, p. 50) who sees that there are no "absolute measures" to specify the distance between the 'high' and the 'low' varieties in a diglossic community. Another question one may pose in this regard is: Can we talk about only one 'high' or 'low' variety in a speech community? Ferguson disregarded that there may exist more than one 'high' or 'low' variety of a language in a given speech community. For example, in Arabic countries, where diglossia shines fully, there are CA and MSA. Both are used as high varieties in formal contexts. More than that, another variety emerged in between SA and colloquial form of Arabic; it is named as Middle Arabic or Educated Spoken Arabic. The question here is where can we position this "Middle Arabic" in a diglossic situation?
The classic example of diglossia in bilingual studies is the case of Hemnesberget speech community in Norway, which was the basis of the pioneering study on CS by Blom and Gumperz (1972). According to the two linguists, Hemnesberget speech community is characterized by the use of two different linguistic varieties in different situations or contexts. The low variety is Ranamal, a local variety, which is basically used in everyday interactions and in 'low' social language functions. On the other side, the high variety is called Bokmal which is the standard Norwegian variety; it is used in 'high' social functions. The case of this speech community is described as a situational CS form by Blom and Gumperz (1972). However, Fasold (1984, p. 194) terms it as "broad diglossia". In this regard, Brown and Colin (1979, p. 47) treat diglossia as a subset of CS, describing it as "a particularly tidy case of much more general phenomenon, CS", and which depends on communication situations. The latter involves the role of participants and features of the scene (Gimode, 2015, p. 27).
A clear manifestation of diglossia is the Arab world. According to Ferguson (1959), Arabic represents a good example of diglossic situation because of the co-existence of a standard literary variety called "al-luɣa al-fusћa" and the colloquial variety named "al-luɣa al-ʕammija". For Fishman (1985, p. 40), the linguistic situation of Arabic language reflects a case of classic diglossia where two varieties of the same "genetically related" language are used in different contexts. However, the diglossic situation of Arabic is more complex than the simple high-low dichotomy (Ennaji, 2002;Kaye, 2002). Therefore, Fishman (1972) suggested an extended version of diglossia. The latter covers any linguistic situation in which two different language systems are used side by side. Hence, it is not limited to the varieties of the same language. For him, the language use domains are what governs the linguistic choice of the participants as "proper usage dictates that only one of the theoretically co-available languages or varieties will be chosen by particular classes of interlocutors on particular kinds of occasions to discuss particular topics" (Fishman, 1972, p. 244).
One could argue that CS and diglossia seem to express different language contact phenomena in the sense that diglossia reflects one-to-one relationship while CS binds the utterance or the conversation together. In other words, diglossia employs two different codes in different contexts separately, whereas CS involves the alternate use of two languages or varieties within the same speech exchange. However, both linguistic phenomena share the fact that participants should use two codes. Besides, diglossia can be studied within the framework of CS as the latter can take place not only between distinct languages but also between different varieties of the same language (Bassiouney, 2009, p. 31). In this regard, Mejdell (2006) argues that CS "should be understood in a broad context to encompass both varieties and different languages" (cited in Bassiouney, 2009). Hence, diglossia can, sometimes, be considered as a form of CS. This form can be called diglossic switching. The latter refers to the alternate use of a high (standard variety) and low (its colloquial form) variety of the same language within the same speech event or conversation, not necessarily in two different contexts or situations. Here exists a knot between CS and diglossia. The latter can be treated as part of CS, especially in its diglossic switching phenomenon.

D. Code-switching versus interference
The distinction between CS and interference has not drawn as much attention and concern as the distinction between CS and other language contact phenomena such as borrowing and CM. However, a number of linguists attempted to tackle this distinction from different perspectives. The phenomenon of interference has been explored more in psycholinguistic and interlanguage studies (Weinreich, 1953;Sharwood-Smith & Kellerman, 1986;Poplack, 1987;Grosjean, 2001). Other terms are used more or less interchangeably with interference such as language transfer and cross-linguistic influence. One of the pioneers of contact linguistics is Uriel Weinreich. The latter uses the term interference as a cover term that includes a range of language contact phenomena to refer to "instances of deviations from the norms of either language, which occur in the speech of bilinguals as a result of their familiarity with more than one language, i.e. as a result of their contact" (Weinreich, 1953, p. 1). Based on this definition, though inexplicitly formulated, CS can be considered as an instantiation of interference. Other researchers use other related terms as umbrella terms which include CS, borrowing, and transfer. On the other hand, there are other researchers who see CS and interference as separate notions that cannot be included under a single cover term. According to Poplack and Meechan (1995), CS is defined as "the juxtaposition of sentences or sentence fragments from two languages, each of which is internally consistent with the morphological and syntactic (and optionally phonological) rules of its lexifier language" (p. 200). Therefore, CS involves the activation of two different languages or language varieties. Concerning interference, it is used in a strictly monolingual context where only one language is operational or activated. In this regard, Grosjean (1995) sees that interference takes place when "a speaker-specific deviation from language being spoken due to the influence of the other 'deactivated' language" (p. 262). Thus, interference indicates the influence that a language has on the way a person uses another. In the same vein, McArthur (2012) argues that interference takes place in the speech of bi-/multilinguals, and it impacts different levels of language: accent, pronunciation, syntax, morphology, vocabulary, and idioms.
De Bot (1992) states that it is difficult to distinguish the different language contact phenomena such as CS and interference. For him, "many instances of cross-linguistic influences are related to code switching and cannot be simply separated from this on theoretical and empirical grounds" (De Bot, 1992, p. 19). In this regard, Pardis (1998) argues that cross-linguistic influence cannot be distinguished unambiguously from CS phenomenon in terms of processing ( cited in De Bot, 2002, p. 291). As for Poplack (1990, p. 39) "each of the mechanisms for combining material from two grammars within a single utterance result from two different processes and is governed by different constraints". Whether or not CS and interference involve the same processing, the key issue in this context that needs further investigation is the ability of speakers to control their engagement in CS and interference. Generally, it seems that speakers can take the decisions to code-switch or not, but it is less clear if speakers can decide and control interference phenomenon in their speech in the same ways. This issue of speech control indicates that there are differences between CS and interference in terms of psycholinguistic processing (Treffers-Daller, 2009, p. 6).
In relation to the issue of control, Poplack (1987) differentiates between smooth and flagged switching. The former is effortless and fluid while the latter draws attention to itself; it is often marked by hesitation, repetition and so on. De Bot (2002) distinguishes between motivated and performance switching. The first indicates the cases of CS in which the speaker switches to the other language intentionally, whereas the second, performance switching, is unintentional. Whether CS is intentional or unintentional raises a long debate among researchers. In the same line, interference of linguistic elements is also considered unintentional and spontaneous. However, Treffers-Daller (2009) questions if and when interference is used as a strategy or when elements transferred have entered the borrowing language permanently. Here comes in line another language contact phenomenon, borrowing.
In this context, Grosjean (2001) distinguishes between dynamic and static interference. Dynamic interferences take place when an element in a language appears accidently, without intention, in a sequence of another language. They are "ephemeral deviations due to the influence of the deactivated language". Static interferences refer to elements which have become part of the implicit grammar of a person. In this regard, Treffers-Daller (2009) wonders to what extent CS can be considered as static or dynamic. The distinction between static and dynamic interferences is not often used by researchers, especially in SLA (Treffers-Daller, 2009). This may be traced to the fact that researchers in psycholinguistics prefer to use the term transfer than interference as the latter carries negative and suspect connotations.
Researchers also distinguish between negative and positive transfer. The latter refers to the cases in which the knowledge of a language helps in learning another while the former type occurs when learning new things is interfered and obstructed by previous learning experiences. Generally, when the languages involved are similar, the linguistic interference is positive, and when they are dissimilar, more obstacles and errors are made; hence, negative transfer. Other types of transfer have been discussed such as linguistic (formal and semantic) versus conceptual transfer (Jarvis & Pavlenko, 2008). Within the linguistic transfer, there are the formal transfer which is related to inadvertent borrowing and the semantic transfer is associated with the influence of another language in the use of a target-language word. On the contrary, conceptual transfer refers to the "ways in which conceptual representations are structured and mapped to language" (Javis & Pavlenko, 2008, p. 112). In the same line, Silva-Corvalán (1994, p. 4) differs between direct and indirect transfer. The former involves importation of a new form from one language into another (e.g. lonche "lunch" in Spanish) while the latter involves the use of a form which corresponds to a structure of the contact language.
The distinction between language-contact phenomena has witnessed an ongoing and indefinite consensus over the terminologies to be used. However, some general conclusions should be made to advance this discussion. For CS and interference, it has been marked that a number of researchers (e.g. Poplack, 1990;De Bot, 1992;Grosjean, 1995) see that it is difficult to clearly distinguish between them as both require the same processing and involve the impact of one language over another. Other researchers (e.g. McClure, 1977; Poplack and Meechan, 1995) stress on the separation of the two concepts as CS takes place in a bilingual setting while interference occurs in monolingual contexts. However, McArthur (2012) sees that interference can also occur in bi-/multilingual settings. For example, it can take place among children who learn two languages simultaneously; here, interference is not unidirectional as usually argued, but it can be bi-directional. In this regard, interference can be seen more relative to CS than distinct. The key issue that calls for new investigation is to what extent CS and interference are controlled by the participants. In other words, are they used intentionally as discourse strategies or do they take place accidently and spontaneously? This investigation will prove much more about the similarity and the differences between CS and interference.

V. CONCLUSION
To sum up, this paper deals with the different concepts related to language contact phenomena. As indicated before, it is difficult to draw definite conclusions from the range of studies and definitions provided by linguists as some of them use different terminologies for apparently the same phenomenon while others see the opposite. The key issue is not only to investigate whether linguists use various labels for essentially similar phenomena, but also to explore if the linguistic phenomena under the investigation are fundamentally distinct either at their manifestations at the surface level or in terms of mechanisms and processes which result into these surface forms. The terminological confusion also comes from using different labels in different neighboring disciplines. For example, with regard to the notion of interference, SLA researchers avoid using interference; instead they prefer transfer or cross-linguistic influence while psycholinguists still use the term interference, and scholars of language-induced contact use convergence or intersystemic influence. The same goes for CS, psycholinguists use language switching. The latter is hardly ever mentioned by researchers on CS studies in general. CS has been the most popular linguistic phenomenon in language-contact studies. A lot of studies have been conducted from different perspectives in an attempt to investigate this phenomenon comprehensively. The following section attempts to go through and explore more some of these popular studies on CS.