Home / Articles / Volume 20 (2022) / An Analysis of English Usage Among Twitter Users in France
Document Actions

 

Abstract

This study analyzes how Twitter users geolocated in France interact online and what prompts their use of the English language, based on data retrieved through Twitter’s API. Six functions are analyzed using a discourse function approach that draws from Gumperz’s discourse functions of code-switching and Begum et al.’s functions of code-switching on Twitter: feelings, advertising, quotations, discourse markers, phatic expressions, and translations. The findings suggest that Twitter users in France resort to English most often when they are expressing or discussing emotions, while advertising represents the second most common function in the dataset. The study sheds new light on current linguistic practices on Twitter and how social media has influenced French digital discourse.

Social media is an important communication tool in modern society through which different cultures and languages come into contact. Twitter, in particular, has become a popular means of communication used in a variety of fields, such as politics, journalism, and academia. Twitter is also a multilingual platform where users have easy and rapid access to different languages. It is a widely used social media platform in France, ranked tenth with around 12 million monthly active users.1 French is used in both single-language and mixed-language tweets. In a large-scale study of code-mixing on Twitter conducted in 2017, the majority of the code-switching tweets had English as one of the languages (85.4%), but French was the next most popular mixed language (Rijhwani et al., 2017).

French was once the global language of diplomacy and international relations (Saugera, 2017). French began to lose its dominance and international prestige in the middle of the 20th century, however, when English started to replace it as the world’s new lingua franca. Currently, the use of English is widespread in France. The 20th century brought English words into the French lexicon due to the raised status of the United States in the economic, political, cultural, and technological spheres. Despite the increasing number of English words entering the French lexicon, the level of French-English bilingualism in France remains relatively low. In fact, the French have some of the weakest English-language proficiency skills in Europe.2 Yet French Twitter users often incorporate English words and phrases into their tweets. This study is focused on finding out what motivates French users to switch to English on Twitter.

Code-switching, or the alternating use of two languages in the same discourse and sometimes in the same utterance, is one way speakers adapt to contact with other languages (Auer, 2005). An example of code-switching from French to English is provided below. The English adjective cute is integrated into a French tweet:

@harrystxpid @jarreash Bah oui elle est cute la photo

‘@harrystxpid @jarreash Well yes the picture is cute’

In this study, I analyze how users geolocated in France interact on Twitter and what accounts for their different language choices. Specifically, I ask: What factors promote English usage and code-switching in tweets? I address this question by analyzing the discourse functions of tweets containing English words. Six discourse functions were found in the data: feelings, advertising, quotations, discourse markers, phatic expressions, and translations. The main finding is that French users tend to use English most often on Twitter when they are expressing or talking about feelings. In the process, they sometime adapt English words to French morphosyntax, resulting in innovative linguistic structures. These practices on French social media illustrates how the French language can adapt to other languages in contact.

The Internet was created in the United States and first became popular there. Logically, therefore, English initially dominated the Internet. Some linguists predicted that English, as an Internet lingua franca, would remain the dominant online language (Crystal, 1997; Fishman, 1998), and the primary online language remains English to this day.3 In comparison, French occupies a less prominent position on the Internet. “In 2012, only 4% of websites had French content. French was late to accept the Internet” (Cohen, 2012, p. 20).4 At the same time, the dominance of English has decreased since the mid-1990s. Between 1998 and 2012, web content in English fell from 80% to 55%. Indeed, the Internet is becoming increasingly multilingual, and other language users are continuously increasing their online presence (Lee, 2016). The publication of The Multilingual Internet (Danet & Herring, 2007) marked a significant phase in studies of online multilingualism. It was the first volume dedicated to studies about languages other than English online, and it underscored that most internet users are not native English speakers. Danet and Herring estimated in 2007 that two-thirds of the one billion Internet users communicated in other languages; today, that number is around three quarters.5

Social media changed the way people communicate, and thus, it offers an attractive research area in the field of linguistics (Bošnjak et al., 2012). The language of computer-mediated communication (CMC) tends to be quite informal and leans toward the vernacular and the colloquial. It can closely resemble everyday conversation (Herring, 2010; Mesthrie, 2011). CMC is in constant evolution due to continual technological advancements (Crystal, 2001), and the massive volumes of data available from social media platforms have provided an opportunity to investigate a broad range of language variations (Androutsopoulos, 2013). Indeed, a growing number of sociolinguistic studies are devoted to social media usage, such as those involving intercultural communication and cultural variation in computer-mediated discourse (Marcoccia, 2012); linguistic behavior in French and Morrocan forums online (Atifi & Marcoccia, 2006); the influence of English and the use of French vocabulary on Twitter (Renwick, 2018); and English borrowings among French gamers (Kraeber, 2021), to name but a few studies that explore the discourse of French CMC users.

Code-switching (CS) is the alternating use of two languages in the same discourse or a “juxtaposition within the same speech exchange of passages of speech belonging to two different grammatical systems or subsystems” (Gumperz, 1982, p. 59). CS is not a random mixing of languages; rather, it tends to follow certain well-established patterns and is subject to grammatical rules. Researchers commonly recognize three types of CS based on the degree of integration: tag-switching, inter-sentential switching, and intra-sentential switching (Poplack, 1980). CS occurs often in bilingual environments and is related to lexical borrowing, which happens when a word or a phrase is adapted in another language. Some scholars (Poplack, 1980; Poplack & Sankoff, 1987) make a distinction between cases of CS and lexical borrowing. For these scholars, if a lexical unit is integrated syntactically or phonologically or shows no integration at all into the base language, it constitutes a case of CS. However, if a lexical unit shows both morphosyntactical and phonological integration, then it is considered an instance of borrowing. Other scholars, such as Bentahila and Davies (1983) and Myers-Scotton (1993), consider the use of borrowed lexical items as another form of CS rather than a different process. They do not consider morphosyntactic integration as a criterion to distinguish CS from borrowing.

Following this latter practice, no distinction is made between CS and borrowing in the present study. Linguistic change in online environments occurs faster than in other written environments such as books or newspapers (Saugera, 2017). Because social media allows for language change at a rapid pace, the nature of language use in CMC, especially on Twitter, makes it difficult to determine whether a word or phrase is fully integrated in another language, much less its usage frequency. It is for this reason that a broad definition of CS is adopted in this study.

With the important amount of text available online, and with the increase in multilingual internet conversations that resemble speech-like conversational interactions, studies of CS online have begun to proliferate (Androutsopoulos & Hinnenkamp, 2001; Page et al., 2014; Paolillo, 1996). Recently, CS was also the topic for The First Workshop on Computational Approaches to Code Switching (2014), and it featured studies of language detection in CS online (Barman et al., 2014; Lin et al., 2014; Solorio et al., 2014), word-level identification (Chittaranjan et al., 2014), and predicting code-switching in multilingual communities (Papalexakis et al., 2014; Volk & Clematide, 2014). Many users on social media platforms code-switch from one language to another and sometimes use several languages at the same time. CS is much more prominent in social media than in formal texts (Das & Gambäck, 2015). The proliferation of CS online is to be expected, since CS also occurs frequently in bilingual/multilingual communities (Bhatia & Ritchie, 2013), and social media comprises many multilingual communities.

Pragmatic factors influencing why CS occurs have been identified in a number of studies and explored both in individual and interactional contexts (Gumperz, 1982), as well as in a broader social environment (Auer, 2005). Gumperz (1982) identified several discourse functions of CS in spoken conversations, such as the use of quotations; addressee specification; reiteration (when the speaker repeats a message in another language to clarify or emphasize a message); and interjections, which serve to mark sentence fillers and are similar to Poplack’s (1980) identification of the tag-switching type of CS.

Auer (1984) pioneered the investigation of CS within a Conversational Analysis framework. He argued that conversational CS is implicated in social situations and is linked to a broader social context related to speakers’ experiences and identity, and thus switching may be explained by a desire on the part of the speakers to express their identity. Communicating in another language can also be seen as a way to claim membership in or solidarity with a particular group (Auer, 1984).

Dewaele (2008, 2010) studied language choice and emotions among multilinguals and found that most multilinguals use their dominant language to express emotions, such as the use of swear words and emotional speech. Dewaele (2008, 2010) argues that multilinguals who use emotionally charged words in the dominant language feel stronger emotions and have a deeper resonance than in other languages. It bears noting that the dominant language does not always refer to the native language. It could also refer to the language which is used often or in which the multilingual speaker has a greater linguistic proficiency. Moreover, Pavlenko (2006) studied the perception of selves with the data gathered from the Bilingualism and Emotion Questionnaire (BEQ) (Dewaele & Pavlenko, 2001-2003; see also Dewaele, 2008, 2010) and found that the majority of the participants reported feeling different when using a foreign language. Although the above studies are based on spoken corpora, they are relevant in that language use on Twitter resembles that of everyday conversations.

Twitter has also been the focus of several recent large-scale studies of language choice. For instance, Nguyen et al. (2015) studied users in the Netherlands who tweeted in a minority language (Limburgish or Frisian) as well as in Dutch. The authors used an automatic language identification tool to classify the tweets according to languages. The results showed that users tended to adapt their language choice to suit their audiences. Most tweets were written in Dutch in order to reach a wider audience, but during conversations, users often switched to the minority language.

Rijhwani et al. (2017) conducted a large-scale quantitative analysis of code-switching patterns in 58 million Twitter tweets. They first collected 50 million tweets from across the world and identified seven languages (Dutch, English, French, German, Portuguese, Spanish, and Turkish) using a language detection device. They found that 96.5% of the tweets were monolingual, with a majority being in English (74%). Only 3.5% of all tweets were code-switched, mostly English-Spanish (21.5%), English-French (20.8%) and English-Portuguese (18.4%). The majority of the CS tweets have English as one of the languages (85.4%), making it the most popular mixed language on Twitter. They also found that French was the next most popular mixed language. In a second study, they analyzed eight million tweets from 24 cosmopolitan cities across Europe, North America, and South America and found that code-switching was more common in Istanbul, a non-English speaking country where 90% of people speak Turkish (12% of code-switched tweets), than in Houston (1% of code-switched tweets) where the majority of people speak English and a third of the population speaks Spanish.

Begum et al. (2016) studied CS in English-Hindi tweets among bilinguals. They established pragmatic functional categories, such as narrative-evaluative, reinforcement, quotations, abuse/negative opinion, reported speech, sarcasm, imperative, cause-effect, and translation. They found that users have a tendency to switch languages when they move from expressing facts to expressing opinions (22% of total switch points). Users also switch languages to reinforce an opinion (19% of total switch points).

The present study draws on methods used in previous work done on CS – in particular, Gumperz’s analysis (1982) of discourse functions of CS and Begum et al.’s study (2016) of pragmatic motivations for CS in tweets. The analysis provides new insights into language choice online by investigating the use of English in French tweets among Twitter users geo-tagged in France.

The data for this study were retrieved through Twitter’s API in RStudio, the integrated development environment (IDE) for R (R Core Team, 2018), using the rtweet package (Kearney, 2018). The Twitter API provides access to Twitter data and makes it possible to retrieve tweets at a large scale. In order to collect English words and phrases in French tweets from users geo-tagged in France, multiple steps were necessary. Computational research on large-scale code-switched data collection and analysis from Twitter is still at the experimental stage. Therefore, in order to be able to retrieve a large number of tweets, various codes were created with the following parameters. Tweets were first collected if (a) the language of the tweet was identified as English by Twitter and (b) the tweet was either geolocated in France or written by a user with one of the French regions identified in the profile. The goal was to be able to collect a large corpus of tweets, so that even if many tweets have to be discarded as non-French, a significant number of tweets would remain for analysis in the dataset. Retweets were not included in the corpus, since they were not authored by the users.

The time period for the data collection was six weeks, from March 7, 2018 to April 22, 2018, which produced a total of approximately 600,000 tweets. These tweets were then filtered according to the following parameters: tweets geo-tagged in France or with French region listed in profile and at least one tweet in the French language. This refinement resulted in 46,845 tweets. After this process, the data were still imperfect and needed to be filtered manually. For instance, many tweets were geo-tagged in France, but the location listed in the user profile was from another country. Those tweets were excluded form the dataset, and the manual filtering resulted in a set of 2212 tweets, all representing French tweets with instances of English words and phrases posted by Twitter users located and geotagged in France. The level of language proficiency of these Twitter users is unknown.

The analysis draws on Gumperz’s (1982) discourse functions of CS as well as Begum et al.’s (2016) functions of CS on Twitter, which provided a model for the analysis. To determine the discourse function, I looked at the context of the tweet and initially categorized the tweet into the following discourse functions of CS from Gumperz ‘s (1982) and Begum et al.’s (2016) studies: feelings, quotations, discourse markers, and translations. During the manual classification, two other discourse functions were found in the dataset: advertising and phatic expressions; resulting in a total of six categories. The frequency was counted for each category. The categorization posed certain challenges. Several tweets fit multiple categories (such as tweet (29) below, which was coded as both advertising and translation). Moreover, a number of tweets did not belong in any category of the discourse function analysis and were analyzed in a separate topic analysis study.6 Thus, from the 2212 collected tweets, 744 tweets were ultimately analyzed in this study. While this dataset may appear small, the sample is fairly representative of English use in the French-based tweets, since the study captured all French-English tweets geolocated in France over a six-week period. Below, an example as well as a definition is provided for each category. English words in the examples are in boldface.

Feelings: A user expresses or talks about emotions. In the example (1), the user expresses disbelief by using the English interjection wtf (‘what the fuck’). The use of wtf adds emotion to the sentence. It may also indicate exasperation on the part of the user.

(1)

wtf jamais de la vie https://t.co/7BD5vHUF2s7

‘wtf never in my life https://t.co/7BD5vHUF2s

Advertising: A user promotes a product, an event, a job, or a service. In example (2), the user switches to English to advertise darts.

(2)

2/21 pcs Flechettes Pointe Molle Safety Dart Arrows+ 100pcs Replacement Soft Combo https://t.co/TwTL9tquD9

‘2/21 pcs Darts Soft Tip Safety Dart Arrows+ 100pcs Replacement Soft Combo https://t.co/TwTL9tquD9

Quotations: A user uses a quote or direct speech set off by quotation marks. In example (3), the user switches to English to report the speech of someone else.

(3)

i’m not here to play dirty” mais gros tu vas te faire éliminer personne saura t’es qui! #Untucked #RPDR

‘“i’m not here to play dirty” but man you are going to get eliminated no one will know who you are! #Untucked #RPDR’

Discourse Markers: These are words used to organize or manage the flow of the discourse (Schiffrin, 1987). They differ from the feelings category in that they are not emotionally motivated. An example of a code-switched discourse marker is in example 4. (Note that bah is also a discourse marker in French, similarly to English ‘well’).

(4)

@AlexisJover_ Yep bah après les allergies ça peut venir comme partir à tout moment en fonction de comment tu vis ouais '^'

‘@AlexisJover_ Yep well allergies can come and go at any moment depending on how you live yeah '^' '

Phatic expressions: This refers to communication used for social purposes in everyday conversations (Malinowski, 1936). Phatic communication can include greetings, such as in example (5).

(5)

Hello les tweetos de la night

‘Hello tweetos of the night’

Translations: These are tweets written in one language and translated or paraphrased (as in example (32)) in the other language (French or English), as in example (6).

(6)

Quand je vais gagner au loto😂 When I win the lottery😂 https://t.co/bA2choLBPx

Table 1 shows the relative frequency of each discourse function. The distribution shows that French users on Twitter resort to English most often when they are expressing or talking about feelings, with a total of 489 tweets (55% of all discourse functions coded). Tweets with instances of English words or phrases used for advertising purposes are the next most common function, accounting for 144 tweets (16%), followed by Quotations, with 102 tweets (12%). English discourse Markers, Phatic expressions, and Translations appear least often in the data.

Table 1. Distribution of discourse functions

Dewaele (2008, 2010) argues that emotions increase the frequency of CS. Similarly, Begum et al. (2016) found that CS is used most frequently when expressing or discussing opinions and sentiments. In this study, users express or talk about their feelings in more than half of the tweets. In examples expressing feelings, the switch from French to English occurs in order to illustrate or punctuate the tweet with a feeling. This usage bears similarities to the use of emojis to add tone or clarity to a tweet (Roele et al., 2020). Among the most commonly used English words in this category, OMG is the most frequent, found in 43 tweets, followed by love (23 times), crush (21 times), yes – and its variants yeah, yep, and yup (15 times), WTF (13 times), and fuck/fucking (12 times).

OMG and WTF are among the most used English words in the data. These acronyms are also commonly used online (Chen, 2014). Isnard (2014) devotes two pages to the acronym WTF in his “Dictionnaire du nouveau français” and indicates that usage of this particular acronym is beginning to extend beyond the online sphere. WTF and mais WTF (‘but WTF’) are commonly used in text messages, tweets, and emails, not only among younger users, but among “des gens comme il faut” (‘proper folks’) as well.9 Moreover, WTF in French does not have the same degree of vulgarity as it does in English. It could be considered as a synonym for “C’est du grand n’importe quoi” (‘It’s a lot of nonsense’) in some tweets. It may also indicate surprise or disbelief (as in example 7), exasperation (as with the use of fucking in tweet 8), or heightened emotion on the part of the user. WTF is thus becoming a generalized intensifier, often used equivalently to OMG. Indeed, in tweet (9), OMG indicates heightened emotion.The emoji in the example appears to exaggerate the feeling of surprise and sadness. In tweet (10), the use of yes is also an indication of the intensity of an emotion. The user shows signs of extreme enthusiasm with the repetition of yes four times, followed by an emoticon as well.

(7)

@Kan4_CSGO @NarkussLol @GamersAssembly Ce faire voler son pc fixe.... Wtf ?

‘@Kan4_CSGO @NarkussLol @GamersAssembly Having your desktop stolen.... Wtf?’

(8)

Ça existe encore cette fucking écriture de Samsung? :'( https://t.co/TCdsTk2JVE

‘It still exists, this fucking spelling of Samsung? :'( https://t.co/TCdsTk2JVE

(9)

Je t'aime mei — Omg tu m’as appeller Mei? 😭 https://t.co/ajdi66SX8O

‘I love you mei — Omg you called me Mei? 😭 https://t.co/ajdi66SX8O

(10)

Yes yes yes yes Magloire est là :) #TPMP

‘Yes yes yes yes Magloire is here :) #TPMP’

Love is the second most common English word in the tweets. It bears noting that the expression love to love in tweet (11) is popular in the French language but does not exist in English. It could be classified as a false anglicism. This expression means “being in love” and is used as an adjective. It was probably invented because of the attractiveness of the English language. Saugera (2017) suggests that anglicims have a modern connotation and can be used as tools to grab attention. Another interesting example is tweet (12) with je vous love. From a syntactic point of view, the insertion of the verb love after the direct object pronoun obeys the grammatical rules of French, but not English; the placement of love is ungrammatical according to the word order constraints of English. Similarly, crush is used as a verb in tweets (13) and (14), and it obeys the grammatical rules of French. However, in English, crush is generally used as a noun to express strong feelings towards somebody, rather than as a verb.

(11)

Ah ouais gros j’suis love to love https://t.co/1QQ9O6tEQe

‘Oh yeah dude I’m really in love https://t.co/1QQ9O6tEQe

(12)

Je vous love 💖 #CQDLT @CQDLT8 @C8TV https://t.co/8GPqSpKhI1

‘I love you 💖 #CQDLT @CQDLT8 @C8TV https://t.co/8GPqSpKhI1

(13)

Je crush sur toi mais toi non... — Mais qui es-tu?? https://t.co/uGrK3P9IG0

‘I have a crush on you but you haven’t... — But who are you?? https://t.co/uGrK3P9IG0

(14)

@_hululu Mais moi je crush sur toi

‘@_hululu But I have a crush on you’

Advertising represents the second most common function in the dataset, with 144 French-English tweets. A number of advertising tweets are spam messages generated from a third-party source. Such tweets are identifiable because they have the same format and are posted several times by the same user. Items such as food (example 15), apparel (example 16), and electronic equipment (17) are occasionally advertised using direct translations in the posting. Nguyen et al.’s study (2015) of minority language Twitter users in the Netherlands found that most tweets were written in the dominant language, Dutch, to reach a wider audience. Although French is the dominant language in France, English is the global lingua franca. By using English, French tweets could reach a wider audience and users could optimize their promotion strategy. Moreover, English has an international appeal in advertising (Bhatia, 1992). Therefore, French Twitter users might use English to make their products or service appealing.

(15)

Diet Whey Isolate 97 - Category: Shakes Protéinés Vendor: Theproteinworks FR Price: 42.79 Condition: Protein Works FR Diet Whey Isolate 97 est très apprécié car, parmi toutes les whey protéines utilisées à l'heure actuelle, elle offre... - https://t.co/HknMb9N1j7 https://t.co/9CkPkjWt0G

‘Diet Whey Isolate 97 - Category: Protein Shakes Vendor: Theproteinworks FR Price: 42.79 Condition: Protein Works FR Diet Whey Isolate 97 is truly appreciated because, among all the whey proteins currently used, it offers... - https://t.co/HknMb9N1j7 https://t.co/9CkPkjWt0G

(16)

5.5CM Chunkly Heel Chelsea Boots Femmes Handsome Round Toe Seude Elastic Band Ankle Boots Casual Shoes 2017 Automne https://t.co/Zhy1n63664 [sic]

‘5.5CM Chunkly Heel Chelsea Boots Women Handsome Round Toe Seude Elastic Band Ankle Boots Casual Shoes 2017 Fall https://t.co/Zhy1n63664

(17)

Nouveautés High-tech #8: Coque iphone 7 plus, Coque iphone 8 plus,J Jecent [Fibre de Carbone] Silicone TPU Souple Bumper Case Cover de Protection Premium Non Slip Surface Housse Etui [Anti-Choc et Anti-Rayures] Coque pour iPhone 7 plus (2016) / iphone 8… https://t.co/muJlgRwfqc https://t.co/tLjor3NQCK

‘New High-tech #8: iphone 7 plus Case, iphone 8 plus Case,J Jecent [Carbon Fiber] Silicone TPU Soft Bumper Case Premium Cover Protection Non Slip Cover Case [Shockproof and Anti-Scratch] iPhone 7 plus (2016) Case / iphone 8… https://t.co/muJlgRwfqc https://t.co/tLjor3NQCK

In the job advertising category, Tweet posters sometimes switch to English to advertise a position in a technology field. Anglicisms, which are words or phrases borrowed from the English language and used in the French language here, are prevalent in the technology sector, since innovations continue to come primarily from overseas. According to Grigg (1997), the English terms which identify technological advancements are difficult to replace due to the frequency of their usage in the industry, the ease of global communication and standardization, or because of the failure to find a concise French equivalent.

A few French-English tweets in our corpus were generated from individuals or companies to advertise a job offer, as in examples (18) and (19). Users who are advertising for a job or different items might want to reach as wide an audience as possible. In tweet (19), the English acronym ASAP is used instead of its French equivalent dès que possible. Tweets are restricted to 280 characters on Twitter. Therefore, the use of the English acronym could be a strategy to conform to the character limit imposed by Twitter.

(18)

Poste en CDI #Montreuil - Tech Lead Data -  https://t.co/wzrfEplRn2

‘Job permanent position #Montreuil – Tech Lead Data -  https://t.co/wzrfEplRn2

(19)

Tu aimes les #médias ? les #startups ? Tu souhaites travailler dans la #communication ? J'ai le stage parfait (6 mois) pour toi, chez @ByMaddyness 👉 https://t.co/6e6lnZPvm9 A pourvoir ASAP ! #Stage #RTplease https://t.co/X2KDVEN43y

‘Do you like the #media ? #startups ? Do you want to work in #communication? I have the perfect internship (6 months) for you at @ByMaddyness 👉 https://t.co/6e6lnZPvm9 Position available ASAP ! #Internship #RTplease https://t.co/X2KDVEN43y

In the data sample, 102 CS tweets are quotes, representing 5% of the CS instances. Hoffman (1991) points out that sometimes people like to quote famous expressions in the original language of the quotation. Begum et al. (2016) also observed instances of quotations and direct speech in their study of CS. However, quotations and direct speech accounted for two distinct categories where quotations marks were not necessarily used. In this analysis, only instances of quotations and direct speech with quotations marks were found in the data.

Various instances of quotations in English can be found below. For instance, “enjoy la life” in tweet (20) is an instance of direct speech using English-French code-switching. It is similar to tweet (12) in that it obeys a grammatical rule of French, in this case with the insertion of the French feminine definite determiner la in front of the noun life. However, the insertion of a definite determiner in English (‘enjoy the life’) would be ungrammatical10 according to the grammatical constraints of English.

(20)

« Enjoy la Life » mdrrrrrr il me tue

‘« Enjoy life » lmao he’s killing me’

Another interesting example is tweet (21), which includes the English quote “Show and Tell.” Show and tell is a common classroom exercise in the Anglophone world that does not have a French equivalent. The use of quotation marks around the expression could be explained by the lack of a concise equivalent in the Francophone world.

(21)

salut salut, aujourd'hui en anglais on devait faire un exposé sur un sujet libre "show and tell" donc présenter quelque chose ect

‘Hi hi, today in English we had to do a presentation about an open subject "show and tell" so present something etc.’

Fifty-eight instances of CS that I categorized were discourse markers, where they serve to mark a pause or punctuate a sentence. Examples of discourse markers are shown below, including yes and its variants, and indeed. It bears mentioning that the word yes and its variants (22), (23), (24) are classified in this context as discourse markers because, as opposed to the use of yes in the feelings category, they are not used to express feelings, emotions, or enthusiasm; rather, they are used as discourse markers to indicate agreement or reinforcement. The anglicism yes is common in the French language. An interesting example is tweet (23) where the Twitter user uses the French equivalent oui in the first sentence and then switches to English with yup in the following sentence. The user appears to use the English equivalent here to avoid repetition and reinforce the statement.

(22)

Yes, 50 ans au service du rock progressif https://t.co/VF30BpLkCZ

‘Yes, 50 years working for progressive work https://t.co/VF30BpLkCZ’

(23)

@0swin Ah oui sur les 18:9 ! Yup mais bon ça touche bien moins de personnes que ceux qui tournent en 16:9 ou 21:9

@0swin Oh yes on the 18:9 ! Yup but it affects way fewer people than those who act in 6:9 or 21:9

(24)

Indeed c'est top pour un job — Yep je sais https://t.co/G5UGHrMEWY

Indeed it’s nice for a job — Yep I know https://t.co/G5UGHrMEWY

Phatic expressions are used in 49 tweets in the data. A majority of greetings are found in this category. Similar examples of greetings are found in Begum et al.’s study of English-Hindi tweets (2016). Hello is the most common phatic expression found in the present data, and it is a common English loan word in French. The form has been attested since 1895 and has been recorded in Rey Debove and Gagnon’s dictionary of anglicisms (Rey-Debove & Gagnon, 1980); it also appears in the French dictionary Larousse. Hello instead of its French equivalent bonjour or salut might have been used in tweet (25) below because of the appeal of anglicisms, which Bogaards (2008) argues can add exoticism and modernity to the original language.

(25)

@Lucasuyt Hello ! Tu as quelle config ?

‘@Lucasuyt Hello ! What configuration do you have?’

The Twitter data contain other phatic expressions as well, such as thanks and please. An interesting example is tweet (26), which uses the word thank instead of thanks as a translation of French merci. Here, the user either made a typographical error, or they might have chosen the wrong word form due to limited English proficiency.

(26)

@smokysex D'accord c'est bien, ce que je le disais, thank [sic]

@smokysex Ok it’s what I said, thanks

In the requests in tweets (27) and (28), the adverb please is used with exclamation marks. Here, the code switch could be interpreted as a means to further emphasize the request. It is also true that the French translation of please, s’il te/vous plait, is longer than the English equivalent. However, the abbreviations stp or svp exist in the French language, so the use of English is not the only strategy the user could have employed to conform to the reduced character limit imposed on Twitter users. Rather, the use of English in these examples appears to indicate reinforcement.

(27)

@_Yessouf Prend moi en un please !

@_Yessouf Bring me one please !

(28)

@SJallamion A retweeter please !!! Encore et encore !!!

@SJallamion Retweet please !!! Again and again !!

A total of 41 tweets were categorized as translations. In this category, the code switch occurs for the purpose of translation: One sentence is in one language, and the next sentence translates it into either French or English. This type of switch appears to be used primarily for advertising (29), news purposes (30), or for publicizing an event (31). It should be noted that Twitter allows users to see the translation of a tweet. However, the tweets analyzed in this dataset were not automatically translated by the platform.

(29)

La nouvelle collection été est en ligne ! Lien dans la bio. The new summer collection is online ! Link in bio..#mode #combishort #ete #ete18 #kaki #couleurs #femme #paris #france #shopping #vetement #womanclothing #womanstyle #frenchdesigner #frenchstyle #soleil #createu… https://t.co/ixLAdlXZsi

(30)

THERANEXUS : THERANEXUS ANNOUNCES THE FIRST EUROPEAN APPROVAL FOR ITS PHASE 2 CLINICAL TRIAL WITH THN 102 IN PARKINSON’S DISEASE PATIENTS https://t.co/q4dvqmqft2 THERANEXUS : THERANEXUS ANNONCE L’OBTENTION D’UNE PREMIERE AUTORISATION EUROPEENNE POUR L’ETUDE CLINIQUE DE PHASE 2 AVEC THN102 DANS LA MALADIE DE PARKINSON https://t.co/8wjtR2XszT

(31)

Rencontrez-nous au Le Cafe Soufflot pour le vin ou le cafe et discutez l'Homme Dans le Costume Noir _ 17h00 NYT Mercredi. Meet us at Cafe Soufflot for wine or coffee and discuss The Man in The Black Suit by @sylvainreynard at 5pm NYT Wednesday.æhttps://t.co/sWLM3U1GXT https://t.co/7BYubpl5u3

As noted earlier, several tweets fit multiple categories, and the categories are therefore are not mutually exclusive. According to Begum et al. (2016), translation from one language to another on Twitter could be a sign of reinforcement, for example, or a way to reach a wider audience with the translated tweet. The majority of tweets in the translation category in the present data are not addressed to specific users; rather, they are regular tweets that can be seen by other Twitter users. Therefore, the translation into English can be interpreted as a way to reach a wider audience.

Only two instances of translations are addressed to specific users. They are used to announce winners for a giveaway (32) and express feelings in English (33). It should be noted that tweet (32) is not translated literally. It is loosely paraphrased, and several new words are added in the third sentence, such as “lucky” and “announced as stated above.”

(32)

Le concours est terminé ! Merci à tous les participants. 😊 Les 2 gagnants sont @unemiraille et @jeonmat. Merci de nous MP au plus vite. 😉 Our giveaway is finished. Thanks to all contestants! The 2 lucky winners are announced as stated above. Please DM us as soon as possible.https://t.co/xiEEd3VPun

In tweet (33), the user translates the tweet into English to express feelings towards a French celebrity. The English translation is used to emphasize the user’s love for this person. This is consistent with the results of this study that show that English is used most frequently when expressing or discussing sentiments.

(33)

@olivierminne Je t'aime I love you https://t.co/EkU9DMtcnX

The findings of this research provide insight into the motivations behind French users’ switches to English in tweets in France. The findings show that French users resort to English most often when they are expressing or talking about their feelings. Words related to emotions such as WTF, OMG, crush, and love appear most commonly in the data. In the emotion category, OMG is the most common English acronym in the data, found in 43 tweets. In his 2014 dictionary, Isnards describes the usage of OMG in the following manner:

En France, tout le monde l'emploie pour sa concision. Surtout les jeunes, qui s'exclament à la moindre occasion. "OMG le fou rire !!!", s'écrie une internaute qui vient de voir sur YouTube la vidéo d'une fille qui se brûle les cheveux en voulant faire une démonstration de fer à lisser." OMG en effet! (Isnard, 2014)

‘In France, everyone uses it because it is so concise. Especially young people, who exclaim on any occasion. “OMG laughing hysterically,” screamed an online user who just watched a Youtube video of a girl who burned her hair while doing a hair straightening demonstration.” OMG indeed!’

OMG is one of the most popular English acronyms on the Internet (Chen, 2014). Like speakers of other languages, French Twitter users may use OMG to conform to the reduced character limit on Twitter: With only three letters, they can convey various emotions such as surprise, intensity of feeling, or exasperation. A similar acronym does not exist in the French language. Moreover, since most French people are not bilingual (Saugera, 2017), they might not be comfortable using the English language in a full sentence. In fact, according to the 2018 EF English Proficiency Index, France ranks 35 out of 88 countries with regard to English proficiency, a ranking considered to be in the moderate rather than the high range. In 2015, French proficiency in English ranked 37 out of 70 countries studied, a ranking considered to be low. While it thus appears that English proficiency in France is improving, this improvement still falls short of the standard used to define bilingualism (Dewaele, 2015). However, using an acronym in English does not require a high level of language proficiency. This consideration might encourage the use of English acronyms by Twitter users in France.

English swearwords such as WTF and fucking are also frequently used in the data. The use of swearwords in a second language conveys less emotional force (Dewaele, 2010). This can make foreign language users less inhibited than native speakers, since they do not perceive the swearwords to be as obscene. In particular, WTF does not have the same degree of vulgarity as it does in English. Its usage is common and popular online in France (Isnard, 2014). In the present data, it is used as a strategy to grab attention. Saugera (2017) suggests that English can be used to create humor, to attract attention, as well as to supply emphasis. Pennel (2016) argues that English words also appear in the digital world because equivalent French vocabulary is lacking. Acronyms such as OMG and WTF do not have French equivalents that are single words. The internet is a language contact situation that allows for linguistic change at a rapid pace (Crystal, 2005). Languages that do not have proper equivalences are thus seeing new instances of linguistic borrowings, especially from English, which remains the most-used language online and the global lingua franca.

Language choice is not only a powerful means of communication but also an expression of identity (Le Page & Tabouret-Keller, 1985). The virtual audience has an impact on the language used, and the online environment allows users to share the identity they want to project online (Georgakopoulou & Spilioti, 2016). Speaking a particular language can signal membership in a particular social group (Auer, 2005). Through the use of English, French users on Twitter can share their interest in the Anglophone culture by making it part of their online identity. Indeed, people may feel freer to express themselves as they choose in social media outlets, and this has arguably led to a rise in commonly-used expressions which help users portray various interests and identities (Graciyal & Viswam, 2018).

A number of works have been written about the influence of American culture in France and the infiltration of the English language (Adamson, 2007; Cohen, 2012; Étiemble, 1964; Grigg, 1997; Gueldry & Gott, 2009; Portes et al., 2006; Saugera, 2017; Siserman, 2013). These reflect the linguistic reality of contemporary French society. English is seen by many as an attractive language (Crystal, 2001), and the attraction of the English language is especially strong among young French people (Pennel, 2016). Thus, the use of language mixing in the data can be explained in part by French users’ fascination with the English language. It may also be a result of the influence of English as the online lingua franca. As Saugera points out:

The influence of English on other languages has been magnified significantly by the advent of the World Wide Web, and the concurrent widespread use of English as its lingua franca. More than previous forms of media, the mass media and associated technologies facilitate the rapid development and diffusion of borrowings. (Saugera, 2017, p. 7)

Finally, French-English CS is also found in commercial tweets to advertise different products or events. Many English words in France are related to the advertising world (Thody, 2000). English is the first linguistic choice of global advertisers and marketers (Bhatia & Richie, 2006). Twitter’s popularity has encouraged the formation of numerous communities, and this has attracted many marketers who are able to advertise their products on a large scale (Aladwani, 2015). The platform facilitates product promotion by making it easier to target and connect with a large pool of users. The use of English appears to be a necessity in the advertising field, and French users are using the lingua franca to reach a wider, invisible audience.

Social media platforms have an important impact on linguistic usage. One of this study’s contributions is a broader understanding of Twitter users’ linguistic choices, particularly as regards Twitter users in France. More generally, it sheds light on how globalization and social media’s mass appeal are influencing the French language. In this study, French users on Twitter resort to English most often when they are expressing or talking about their feelings. At the same time, through language choice, the Twitter users are sharing a multilinguistic and multicultural identity. Because language and culture are so intricately linked, people tend to see linguistic evolution as an erosion of identity, and this often leads to stereotypical and entrenched attitudes about right and wrong ways of speaking (Kinzler, 2020). It is therefore important to raise awareness about misconceptions about the role of language in a global society. Change in language usage is natural and a result of linguistic and social interactions.

A limitation of this study is that the data collected to identify French users does not provide a complete view of the users’ background. In particular, information about the users’ linguistic and English proficiency is not available, other than what can be inferred through their tweets. The language used in tweets provides just a limited view of the user, with limited social context, making it difficult to determine a user’s language proficiency. Another limitation of the study involves the sample size. Only 744 tweets are analysed in this study, although many duplicated tweets, spam tweets, and monolingual tweets were removed from the data. Analyzing CS tweets of French users is still at an exploratory stage. A possible continuation of the study would be to compare CS tweets with monolingual French tweets to investigate if French tweets express the same discourse functions as CS tweets. Expanding the scope of this study by including English usage from other platforms such as Facebook would also provide more extensive knowledge regarding linguistic choices online, as well as providing additional insight into the nature of code-switching in mediated digital discourse.

  1. https://www.bfmtv.com/economie/entreprises/utilisateurs-audiences-bizarreries-twitter-en-15-chiffres_AV-202204260218.html, accessed January 9, 2022.

  2. https://www.euronews.com/2018/11/03/which-eu-countries-are-the-best-and-worst-at-speaking-english, accessed January 9, 2022.

  3. As of December 1, 2022, the number of French websites still hovers around 3.5%.

  4. https://www.verbolabs.com/internet-users-by-languages-worldwide/, accessed December 4, 2022. https://en.wikipedia.org/wiki/Languages_used_on_the_Internet

  5. See note 1.

  6. Hadour (2021).

  7. Original tweets are in italics throughout this article, and English translations are provided under each tweet. The translation was carried out by the author and checked by a second bilingual scholar.

  8. A total of 744 tweets were analyzed in this study, but this number includes tweets that fit multiple categories.

  9. Quote from Didier Pourquery, a journalist from the newspaper Le Monde, who wrote an article about the acronym “WTF”: https://www.lemonde.fr/m-actu/article/2014/04/25/juste-un-mot-wtf_4406440_4497186.html

  10. Except in certain contexts where a certain kind of future life is understood from context, such as ‘Enjoy the (good) life (in your retirement).’

Adamson, R. (2007). The defence of French: A language in crisis? (Vol. 137). Multilingual matters.

Aladwani, A. M. (2015). Facilitators, characteristics, and impacts of Twitter use: Theoretical analysis and empirical illustration. International Journal of Information Management, 35(1), 15-25.

Androutsopoulos, J., & Hinnenkamp, V. (2001). Code-Switching in der bilingualen Chat- Kommunikation: Ein explorativer Blick auf #hellas und #turks. In M. Beißwenger (Ed.), Chat-Kommunikation (pp. 367–402). Ibidem.

Androutsopoulos, J. (2013). Online data collection. In C. Mallinson, B. Childs, & G. Van Herk (Eds.), Data collection in sociolinguistics: Methods and applications (pp. 236-249). Routledge.

Atifi, H., & Marcoccia, M. (2006). Communication médiatisée par ordinateur et variation culturelle: analyse contrastive de forums de discussion français et marocains. Les Carnets du Cediscor, 9, 59-73. https://doi.org/10.4000/cediscor.629

Auer, P. (1984). Bilingual conversation. John Benjamins.

Auer, P. (2005). A postscript: Code-switching and social identity. Journal of Pragmatics, 37(3), 403-410.

Barman, U., Das, A., Wagner, J., & Foster, J. (2014). Code mixing: A challenge for language identification in the language of social media. In First Workshop on Computational Approaches to Code Switching (EMNLP 2014) (pp. 13-23). Association for Computational Linguistics.

Begum, R., Bali, K., Choudhury, M., Rudra, K., & Ganguly, N. (2016). Functions of code- switching in Tweets: An annotation scheme and some initial experiments. In Proceedings of the Tenth International Conference on Language Resources and Evaluation (LREC’16) (pp. 1644-1650). European Language Resources Association.

Bentahila, A., & Davies, E. (1983). The syntax of Arabic-French code switching. Lingua, 59(4), 301-330.

Bhatia, T. K. (1992). Discourse functions and pragmatics of mixing: Advertising across cultures. World Englishes, 11(2‐3), 195-215.

Bhatia, T. K., & Ritchie, W. C. (2006). Bilingualism in the global media and advertising. In The Handbook of bilingualism (pp. 512–46). Blackwell Publishing.

Bogaards, P. (2008) On ne parle pas franglais. Groupe De Boeck.

Bošnjak, M., Oliveira, E., Martins, J., Mendes Rodrigues, E., & Sarmento, L. (2012). Twitterecho: A distributed focused crawler to support open research with Twitter data. In Proceedings of the 21st International Conference on World Wide Web (pp. 1233-1240). ACM.

Chen, S. Y. (2014). From OMG to TMD – Internet and Pinyin acronyms in Mandarin Chinese. Language@ Internet, 11, article 3. Retrieved from: https://www.languageatinternet.org/articles/2014/chen

Chittaranjan, G., Vyas, Y., Bali, K., & Choudhury, M. (2014). Word-level language identification using crf: Code-switching shared task report of msr india system. In First Workshop on Computational Approaches to Code Switching (EMNLP 2014) (pp. 73-79). Association for Computational Linguistics.

Cohen, L. (2012). The French identity crisis: Fending off the franglais invasion. Honors Projects. Illinois Wesleyan University. https://digitalcommons.iwu.edu/french_honproj/5

Crystal, D. (1997). English as a global language. Cambridge University Press.

Crystal, D. (2001). Language and the Internet. Cambridge University Press.

Crystal, D. (2005). The scope of Internet linguistics. In Proceedings of American Association for the Advancement of Science Conference, Washington, DC (pp. 17-21).

Danet, B., & Herring, S. C. (Eds.). (2007). The multilingual Internet: Language, culture, and communication online. Oxford University Press.

Das, A., & Gambäck, B. (2015). Code-mixing in social media text: The last language identification frontier. TAL Journal, 54(3). Retrived from: https://ntnuopen.ntnu.no/ntnu-xmlui/bitstream/handle/11250/2391067/2.Das-TAL54-3.pdf?sequence=3&isAllowed=y

Dewaele, J. M. (2008). The emotional weight of I love you in multilinguals’ languages. Journal of Pragmatics, 40(10), 1753-1780.

Dewaele, J. M. (2010). Emotions in multiple languages. Palgrave Macmillan.

Dewaele, J. M., & Pavlenko, A. (2001-2003). Web questionnaire ‘Bilingualism and Emotions.’ University of London. https://doi.org/10.1177/0267658309337640

Dewaele, J. M. (2015). Bilingualism and multilingualism. In K. Tracy, C. Ilie, & T. Sandel (Eds.), The international encyclopedia of language and social interaction (pp. 1–11). John Wiley & Sons. https://doi.org/10.1002/9781118611463.wbielsi108

EF EPI (2018) – France. (n.d.). EF English Proficiency Index. https://www.ef.edu/epi/regions/europe/france/

Étiemble, R. (1964). Parlez-vous franglais? Gallimard.

Fishman, J. A. (1998). The new linguistic order. Foreign Policy, 113, 26-40.

Georgakopoulou, A., & Spilioti, T. (Eds.). (2015). The Routledge handbook of language and digital communication. Routledge.

Graciyal, D & Deepa, V. (2018). Freedom of expression in social media: A political perspective. Research Review International Journal of Multidisciplinary, 3(1), 110-113.

Grigg, P. (1997). Toubon or not Toubon: The influence of the English language in contemporary France. English Studies, 78(4), 368-384.

Gueldry, M., & Gott, M. (2009). The Americanization of France. Yale French Studies, (116/117), 37-51.

Gumperz, J. J. (1982). Discourse strategies (Vol. 1). Cambridge University Press.

Hadour, T. (2019). #LanguageMixing on Twitter. Unpublished doctoral dissertation, University of Missouri.

Hadour, T. (2021). Topic analysis of French-English tweets in France. Digital Scholarship in the Humanities, 37(1), 121-136.

Herring, S. C., Ed. (2010). Computer-mediated conversation, Part I. Special issue of Language@Internet, 7. http://www.languageatinternet.org/articles/2010

Hoffman, C. (1991). An introduction to bilingualism. Longman.

Isnards, A. D. (2014). Dictionnaire du nouveau français. Allary éditions.

Kearney, M. W. (2018). rtweet: Collecting Twitter data. R package version 0.6.8. https://cran.r-project.org/package=rtweet, retrieved October 20, 2020.

Kinzler, K. D. (2020). How you say it: Why you talk the way you do? And what it says about you. Houghton Mifflin.

Kraeber, S. (2021). English borrowed words in French gamer discourse: How do they function? In VI LexESP Seminar: The Language of Tourism, Sports and Outdoor Leisure Activities. https://hal.science/hal-03220935

Lee, C. (2016). Multilingual resources and practices in digital communication. In A. Georgakopoulou & T. Spilioti (Eds.), The Routledge handbook of language and digital communication (pp. 118–132). Routledge.

Le Page, R., & Tabouret-Keller, A. (1985) Acts of identity: Creole-based approaches to language and ethnicity. Cambridge University Press.

Lin, C. C., Ammar, W., Levin, L., & Dyer, C. (2014). The cmu submission for the shared task on language identification in code-switched data. In First Workshop on Computational Approaches to Code Switching (EMNLP 2014) (pp. 80-86). Association for Computational Linguistics.

Malinowski, B. (1923). The problem of meaning in primitive languages. In C. K. Ogden, & I. A. Richards (Eds.), The meaning of meaning (pp. 296-336). K. Paul, Trend, Trubner.

Marcoccia, M. (2012). The internet, intercultural communication and cultural variation. Language and Intercultural Communication, 12(4), 353-368.

Mesthrie, R. (Ed.). (2011). The Cambridge handbook of sociolinguistics. Cambridge University Press.

Myers-Scotton, C. (1993). Common and uncommon ground: Social and structural factors in codeswitching. Language in Society, 22(4), 475-503.

Nguyen, D., Trieschnigg, D., & Cornips, L. (2015). Audience and the use of minority languages on Twitter. Proceedings of the International AAAI Conference on Web and Social Media, 9(1), 666-669.

Page, R., Barton, D., Unger, J. W., & Zappavigna, M. (2014). Researching language and social media: A student guide. Routledge.

Paolillo, J. C. (1996). Language choice on soc.culture.punjab. Electronic Journal of Communication, 6(3). http://www.cios.org/EJCPUBLIC/006/3/006312.HTML

Papalexakis, E., Nguyen, D., & Doğruöz, A. S. (2014). Predicting code-switching in multilingual communication for immigrant communities. In First Workshop on Computational Approaches to Code Switching (EMNLP 2014) (pp. 42-50). Association for Computational Linguistics.

Pavlenko, A. (2006). Emotions and multilingualism. Cambridge University Press.

Pennel, F. (2016). Pourquoi les français adorent l’anglais mais le parlent mal. https://www.slate.fr/story/120169/francais-attrait-anglais-parle-mal, retrieved February 20, 2019.

Poplack, S. (1980). Sometimes I’ll start a sentence in Spanish Y TERMINO EN ESPAÑOL: Toward a typology of code-switching. Linguistics, 18(7-8), 581-618.

Poplack, S., & Sankoff, D. (1987). The Philadelphia story in the Spanish Caribbean. American Speech, 62(4), 291-314.

Portes, J., Fohlen, C., & Forster, E. (2006). Fascination and misgivings: The United States in French opinion, 1870-1914. Cambridge University Press.

R Core Team (2018). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.r-project.org.

Renwick, A. (2018). Les mots des réseaux sociaux. In C. Jacquet-Pfau, A. Napieralski, & J. F. Sablayrolles (Eds.), Emprunts néologiques et équivalents autochtones: études interlangues (pp. 73-87). Wydawnictwo Uniwersytetu Łódzkiego.

Rey-Debove, J., & Gagnon, G. (1980). Dictionnaire des anglicismes: Les mots anglais et américains en français. Les Usuels du Robert.

Roele, M., Ward, J., & van Duijn, M. (2020). Tweet with a smile: The selection and use of emoji on Twitter in the Netherlands and England. First Monday, 25(4). https://doi.org/10.5210/fm.v25i4.9373

Rijhwani, S., Sequiera, R., Choudhury, M., Bali, K., & Maddila, C. S. (2017). Estimating code-switching on Twitter with a novel generalized word-level language detection technique. In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Vol. 1 (pp. 1971-1982).

Saugera, V. (2017). Remade in France: Anglicisms in the lexicon and morphology of French. Oxford University Press.

Schiffrin, D. (1987). Discourse markers. Cambridge University Press.

Siserman, M. (2013). Sur quelques aspects du phenomène d’americanisation dans la France d’aujourd’hui. Buletin Stiintific, seria A, Fascicula Filologie, 22(1), 79-87.

Solorio, T., Blair, E., Maharjan, S., Bethard, S., Diab, M., Ghoneim, M., ... & Fung, P. (2014). Overview for the first shared task on language identification in code-switched data. In First Workshop on Computational Approaches to Code Switching (EMNLP 2014) (pp. 62-72). Association for Computational Linguistics.

Thody, P. (2000). Le franglais: Forbidden English, forbidden American: Law, politics and language in contemporary France: A study in loan words and national identity. A&C Black.

Volk, M., & Clematide, S. (2014). Detecting code-switching in a multilingual alpine heritage corpus. In First Workshop on Computational Approaches to Code Switching (EMNLP 2014) (pp. 24-33). Association for Computational Linguistics.

Taoues Hadour [taous.hadourmyers@ucf.edu] is an Assistant Professor of French Linguistics at the University of Central Florida. Her research interests include computer-mediated communications, sociolinguistics, and bilingualism. Her current research focuses on the influence of English on the French language in social media.

License

Any party may pass on this Work by electronic means and make it available for download under the terms and conditions of the Digital Peer Publishing License. The text of the license may be accessed and retrieved at http://www.dipp.nrw.de/lizenzen/dppl/dppl/DPPL_v2_en_06-2004.html.

Fulltext