research paper about code switching

Our systems are now restored following recent technical disruption, and we’re working hard to catch up on publishing. We apologise for the inconvenience caused. Find out more: https://www.cambridge.org/universitypress/about-us/news-and-blogs/cambridge-university-press-publishing-update-following-technical-disruption

We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings .

Login Alert

• > Journals
• > Bilingualism: Language and Cognition
• > Volume 21 Issue 5
• > Code-switching as a marker of linguistic competence...

Article contents

• Introduction
• General discussion

Supplementary material

Code-switching as a marker of linguistic competence in bilingual children.

Published online by Cambridge University Press:  04 September 2017

• Supplementary materials

Code-switching is a common phenomenon that bilinguals engage in, including bilingual children. While many researchers have analyzed code-switching behaviors to better understand more about the language processes in bilingual children, few have examined how code-switching behavior affects a child's linguistic competence. This study thus sought to examine the relationship between code-switching and linguistic competency in bilingual children. Fifty-five English–Mandarin bilingual children aged 5 to 6 years were observed during classroom activities over five days (three hours each day). A number of different word roots and mean length of utterance for both languages, and a number of code-switched utterances for each child, were computed. English receptive vocabulary scores were also obtained. Additionally, teachers rated children's English and Mandarin language competencies approximately six months later. Correlational and hierarchical regression analyses support the argument that code-switching does not indicate linguistic incompetence. Instead, bilingual children's code-switching strongly suggests that it is a marker of linguistic competence.

1. Introduction

Code-switching is a common phenomenon that bilingual speakers regularly engage in. When bilinguals code-switch, words from two languages are used within a single discourse. In some studies, code-switching has been distinguished from code-mixing – code-mixing is defined as a practice of mixing languages in a single sentence while code-switching can occur either within or across sentence boundaries within a single discourse or constituent (e.g., Brice & Anderson, Reference Brice and Anderson 1999 ; Meisel, Reference Meisel 1989 ; Muysken, Reference Muysken 2000 ; Nicoladis & Genesee, Reference Nicoladis and Genesee 1997 ). In other studies, as well as the present study, code-switching and code-mixing are synonymously regarded as an alternation of two languages within the same speech act (Bokamba, Reference Bokamba 1989 ; Clyne, Reference Clyne 1987 ; Genesee, Reference Genesee 1989 ; Genesee, Paradis & Crago, Reference Genesee, Paradis and Crago 2004 ; Poplack, Reference Poplack, Smelser and Baltes 2001 ).

Code-switching has been well studied in bilingual adults, particularly with regard to the grammatical and communicative functions of the behavior (e.g., Cantone, Reference Cantone 2007 ; Gumperz, Reference Gumperz and Dil 1971 ; MacSwan, Reference MacSwan 2014 ; MacSwan & McAlister, Reference MacSwan and McAlister 2010 ; McClure, Reference McClure 1977 ; Poplack, Reference Poplack 1980 ). The complexity of bilingual adults’ code-switching generally reveals a sophisticated knowledge of the grammars of both languages and reflects the adults’ competency in using them appropriately. However, there has been much debate with respect to what children's code-switching behavior suggests about their linguistic competency.

Early studies on children's language alternation behaviors postulated that bilingual children mix or switch languages because 1) they are confused or 2) they are linguistically incompetent. According to proponents of the position that bilingual children mix languages because they are confused and cannot differentiate between the two languages (e.g., the Unitary Language System Hypothesis in young children aged 3 years and below; Genesee, Reference Genesee 1989 ), the lexicons and grammars of both languages in young bilingual children first exist in one single system, and only gradually develop into two separate linguistic systems by a process of language differentiation. In this framework, young bilingual children's mixing of two different language elements within the same utterance was seen as evidence of the pre-separation stage and, thus, was argued to be a reflection of their inability to differentiate two language systems (Köppe & Meisel, Reference Köppe and Meisel 1995 ; Redlinger & Park, Reference Redlinger and Park 1980 ; Volterra & Taeschner, Reference Volterra and Taeschner 1978 ). For example, Redlinger and Park ( Reference Redlinger and Park 1980 ) studied four 2-year-old bilingual children over five to nine months and suggested that the children experienced various stages of language differentiation. The children would start off with high rates of language mixing as they do not separate their two language systems. The high language mixing rates would then gradually decline as these children move from an undifferentiated single language system to two distinct language systems. Proponents of this position argued that the decrease in the language-mixing rate is, therefore, positively related to language development, at least for children aged 3 years and below.

Other researchers claimed that bilingual children code-switch not because they cannot differentiate the two language systems, but because they lack the lexical, grammatical and/or pragmatic competence in one or both of the languages known. Several studies have found that bilingual children aged between 2 to 6 years code-switch in order to fill in their lexical gaps – they tend to insert words from one language into another language when they do not have the translation equivalents (e.g., Deuchar & Quay, Reference Deuchar and Quay 2000 ; Cantone, Reference Cantone 2007 ; Lindholm & Padilla, Reference Lindholm and Padilla 1978 ; see Nicoladis & Genesee, Reference Nicoladis and Genesee 1997 , for a review). Further, Bernardini and Schlyter ( Reference Bernardini and Schlyter 2004 ), who examined code-mixing patterns of five Swedish–French/Italian children aged 2 to 4 years, posited that children code-switch because they are not yet competent in structuring grammatical sentences in their “weaker” language. In addition, Vihman ( Reference Vihman 1985 ) suggested that a young bilingual child aged 3 years old or younger would not be focused on the situational context when developing a dual lexicon. As such, the child may code-switch inappropriately during this period, reflecting the absence of pragmatic competence.

Recent studies, however, have provided more complex and contradictory evidence. First, many studies have failed to confirm the Unitary Language System Hypothesis . Results suggest that young bilingual children are able to differentiate their two language systems from an early age and their code-switched utterances are systematic and conform to the grammatical constraints of each known language (e.g., MacSwan, Reference MacSwan 1999 ; Meisel, Reference Meisel 1994 ; Nicoladis & Genesee, Reference Nicoladis and Genesee 1997 ; Paradis, Nicoladis & Genesee, Reference Paradis, Nicoladis and Genesee 2000 ; van Gelderen & MacSwan, Reference van Gelderen and MacSwan 2008 ). Genesee ( Reference Genesee 1989 ) argued that, contrary to the Unitary Language System Hypothesis , young bilingual children are able to use their developing language systems differentially in contextually sensitive ways. Second, case studies have found that children's code-switching behavior illustrates a good understanding of the grammatical systems of both languages. For example, 2- to 4-year-old French–English bilingual children displayed code-switching patterns that were largely similar to that of their adult counterparts – grammatical constraints were adhered to in mixing patterns that involved sentential negation and pronominal subjects (Paradis et al., Reference Paradis, Nicoladis and Genesee 2000 ). Siri's data in Lanza ( Reference Lanza 1992 ) indicated that the two-year-old did not use the inflections of both languages interchangeably. English grammatical morphemes were only used with English lexical morphemes, such as “look s ”, but Norwegian grammatical morphemes were used with both Norwegian and English words, such as “look er ” and “ husk er ”. Results from Cantone ( Reference Cantone 2007 ) also revealed how all instances of Italian–German code-switching in 2- to 5-year-old children in her corpus were grammatical. Furthermore, an English–Spanish bilingual child, M, exhibited language-specific syntax and morphology in both her pure and mixed utterances before the age of 3 years (Deuchar & Quay, Reference Deuchar and Quay 1998 ). Therefore, the results of studies such as the aforementioned suggest that bilingual children's code-switching behavior does not indicate an inability to differentiate their two language systems or a lack of linguistic competency. Instead, they strongly suggest that children's code-switching behavior illustrates that they possess adequate grammatical knowledge of both languages.

Additionally, numerous studies have demonstrated that bilingual children are pragmatically competent and can code-switch according to the situation and interlocutor (e.g., Genesee, Boivin & Nicoladis, Reference Genesee, Boivin and Nicoladis 1996 ). Even though bilingual children below the age of 4 years have more single-noun insertions in their mixed utterances, these utterances reflect their awareness of social norms (De Houwer, Reference De Houwer, Kroll and de Groot 2005 ). For example, Siri code-switched in bilingual contexts but not in monolingual contexts, thereby reflecting her sensitivity to the social demands of the conversation (Lanza, Reference Lanza 1992 ). Two-year-old bilingual children were also able to adjust their rates of code-switching according to that of their interlocutors (Comeau, Genesee & Lapaquette, Reference Comeau, Genesee and Lapaquette 2003 ), suggesting that they are sensitive to the language choices of their interlocutors. Bilingual children's code-switched utterances were also found to be in accordance with the language socialization practices in their families. Chung ( Reference Chung 2006 ) found that a 4.5-year-old and an 11-year-old Korean-American (who were regularly exposed to both Korean and English before age 3) switched between the two languages when conversing with their family members who had different language preferences. This facilitated communication and comprehension between the family members despite these different preferences. Moreover, Vu, Bailey and Howes ( Reference Vu, Bailey and Howes 2010 ) found that 4.5- to 5.5-year-old Spanish–English bilingual children code-switched in their attempts to draw the interviewer's attention or to change speaking roles. These studies show that bilingual children have the pragmatic competence to adjust their code-switching behavior appropriately depending on the situational contexts.

Some researchers have attempted to examine the code-switching-linguistic-competency relationship through investigating whether code-switching in elicited narratives can be a marker of language impairment (LI). Iluz-Cohen and Walters ( Reference Iluz-Cohen and Walters 2012 ) found that 5- and 6-year-old Hebrew–English bilingual children with LI code-switched more than bilingual children with typical language development (TLD). Using the Bilingual English–Spanish Oral Screener (BESOS), Greene, Peña, and Badore ( Reference Greene, Peña and Bedore 2012 ) found that 5-year-old children's risk status for language impairment affected their code-mixing frequency. With the English screener, children who were identified as at-risk of LI code-switched more than the no-risk group. Interestingly, the no-risk group code-switched more on the Spanish screener than the at-risk group. The authors cited limited awareness of the social interaction and challenges in suppressing the irrelevant language as possible explanations for these findings. Contrary to these findings, Gutiérrez-Clellen, Simon-Cereijido and Erickson Leone ( Reference Gutiérrez-Clellen, Simon-Cereijido and Erickson Leone 2009 ) did not find any differences in the use of code-switching between 5- to 6-year-old bilingual children with LI and those with TLD. Clearly, the inconsistent findings across studies call for more research to be conducted in these areas.

Thus, despite earlier attempts to understand the nature of bilingual children's code-switching, the relationship between children's code-switching and linguistic competency remains not well understood, or at best, controversial (Baetens Beardsmore, Reference Baetens Beardsmore, Gopinathan, Kam, Pakir and Saravanan 1998 ; Kamwangamalu & Leng, Reference Kamwangamalu and Lee 1991 ; Ong & Zhang, Reference Ong and Zhang 2010 ). Furthermore, no study has investigated children's code-switching behavior in bilingual preschool settings, especially for children who spend a significant amount of their awake time at these centers, and at a time when their language skills are becoming more complex. Most research on children's code-switching behavior consists of case studies of parent-child interactions or is based on children's narrative samples in a laboratory setting. Information on groups of children's code-switching behavior in a larger natural language environment would be relevant and important to understand the effects of code-switching on bilingual children's language development. Many children spend approximately 10 hours a day or more away from home, such as in childcare or daycare centers. Their language development is, thus, largely influenced by their interactions in this larger and more complex environment (Chung, Reference Chung 2006 ; Comeau et al., Reference Comeau, Genesee and Lapaquette 2003 ; Nicoladis & Genesee, Reference Nicoladis and Genesee 1997 ). In short, code-switching in such sociolinguistic contexts has not yet been adequately examined. The present study, therefore, seeks to fill these gaps by investigating the relationship between children's code-switching and linguistic competency in preschool settings and adopting a quantitative approach toward the analysis of children's language behavior. Measures of 55 English–Mandarin children's spontaneous speech were obtained through five 3-hour observation sessions in two childcare centers. Information on the children's receptive vocabulary was also obtained. In addition, teachers’ assessments of the children's language competency in English and Mandarin were collected approximately six months after the observation sessions in order to ascertain whether any predictive relationship exists between code-switching and linguistic competency.

Participants

Fifty-five English–Mandarin bilingual children aged between 5;5 to 6;7 ( M = 6.06, SD = 0.34) from two private childcare centers in Singapore Footnote 1 (33 from Center 1 and 22 from Center 2; 25 females, 30 males) were observed during their classroom activities. Four additional participants were excluded either because they had very low attendance during the observation days that resulted in very little recording time (less than 5% of the total recording time in the center) or because he or she spoke fewer than ten utterances throughout the entire observation session. Both childcare centers conducted classroom activities in English and Mandarin Footnote 2 .

Parents completed a demographic and language background questionnaire prior to the observation session. The questionnaire asked about the age and gender of the child, the language first acquired by the child, and the amount of time (in percentage) their child hears or speaks a language in a typical week (see Yow & Markman, Reference Yow and Markman 2016 ). The average amount of English and Mandarin exposure children had at home as reported by the parents was 55.30% ( SD = 19.93%) and 41.80% ( SD = 20.15%) respectively. All children were reported as simultaneous bilinguals (i.e., acquiring two languages at age 3 or younger), except for one who was reported as a sequential bilingual. Preliminary analysis found that including the single sequential bilingual child did not change the results significantly, thus all children were included in the final analyses. In addition, parents reported their highest education level as a measure of socioeconomic status (SES), ranging from 0 ( no formal education ) to 5 ( postgraduate degree ). The average parental highest education level was 3.98 ( SD = 0.54).

Parents were informed about the study and were requested to complete the language background questionnaires that were distributed with the teacher-parent communication book. The observation was conducted for about three hours each day across five different days in each childcare center. We were thus able to record children's conversations in different settings throughout the week, such as during meal times, craft sessions, and free play. Teachers split children from their respective childcare center into two groups of 2 to 6 as part of their normal preschool routine. Two research assistants each followed and recorded one group of children with a video camera that also had an audio recorder attached to it throughout the recording duration. The research assistants held the camera and audio recorder as close to the children as possible without interrupting their activities. As the audio and video recording were meant to be as naturalistic as possible, there was no form of intervention from the researchers during the entire recording duration. The video recordings were transcribed and crosschecked with the audio recordings, especially when the conversations were unclear from the video recordings. After the observation sessions ended, children were tested individually on their receptive vocabulary using the Peabody Picture Vocabulary Test (4 th Edition; PPVT-IV; Dunn & Dunn, Reference Dunn and Dunn 2007 ). The respective teachers were asked to complete a short questionnaire on the language competency of those children who participated in the study approximately six months after the observation sessions ended (see section on Materials).

Measure of receptive language competency

The Peabody Picture Vocabulary Test (4 th Edition; PPVT-IV; Dunn & Dunn, Reference Dunn and Dunn 2007 ) was administered individually to assess children's receptive English vocabulary. Each child was instructed to point to one of four pictures that depicted the word spoken by the experimenter. Eleven children from Center 1 and five children from Center 2 did not complete the PPVT. Raw scores were converted to age-based standard scores according to the manual. The average standardized score was 100.69 ( SD = 12.33). As there is currently no equivalent approved version in Mandarin, the same task was not conducted in Mandarin.

Teachers’ report of language competency

Teachers were asked to rate the students’ expressive and receptive language competencies from 1 ( very poor ) to 5 ( very good ) approximately six months after the observation session ended. This questionnaire consisted of eight items ( Table A1 ), which we developed based on the Language and Literacy section of the curriculum framework for kindergartens in Singapore (Singapore Ministry of Education, 2012 ). Examples include “he or she talks about drawings and artworks he or she has created” (expressive) and “he or she understands a good variety of words” (receptive). English teachers assessed English language competency while Mandarin teachers assessed Mandarin language competency of the respective children in their charge. The average teacher's rating of English and Mandarin competency was 3.75 ( SD = 0.95) and 3.85 ( SD = 0.79) respectively.

Transcription

Children's utterances during the observation sessions were transcribed in accordance with CHAT and the transcriptions were analyzed using CLAN (MacWhinney, Reference MacWhinney 2000 ). Four additional research assistants, who were also native language speakers of English and Mandarin, were involved in the transcription and checking process. All videos were divided among the six research assistants. The research assistants independently transcribed the videos assigned to them. In accordance with the transcription and reliability checking methods detailed in Lust and Blume ( Reference Lust and Blume 2016 ), a different research assistant (i.e., second transcriber) checked through each transcription for errors or missing data. All transcriptions were checked sentence by sentence by crosschecking the video and audio recordings. When there were discrepancies, the second transcriber would discuss them with the first transcriber before making changes to the transcriptions. A third transcriber was involved if the first two transcribers could not come to an agreement.

In all transcriptions, onomatopoeia (imitation of sounds, e.g., “woof woof”) and ambiguous communicators that can be used in either English or Mandarin, such as “uh”/“哦”, “ah”/“啊”, “oh”/“噢”, Singlish Footnote 3 particles (e.g., “meh”, “la”, “na”, see Rubdy, Reference Rubdy 2007 ) were marked as non-words and thus automatically excluded from all analyses. Words that were not English or Mandarin were also marked as non-words (e.g., “chaota”, a Hokkien word which means burnt). All forms of routinized speech, such as standardized greetings before meal, text or nursery-rhyme reading, and games with standard lyrics (e.g., “scissors paper stone”) were excluded from the analyses as well. The basic unit of our analyses is an utterance, which is defined as “a word or group of words with a single intonation contour” (Lanza, Reference Lanza 1992 , p. 638). A pure utterance (either in English or Mandarin) consists of words only in one language, and excludes single proper nouns, intra-sentential switches, and utterances that contain translations and imitations of other languages.

Expressive language measures: Number of different word roots (NDWR) per minute

Lemma or word roots have often been used as a measure of children's lexical development (Hewitt, Hammer, Yont & Tomblin, Reference Hewitt, Hammer, Yont and Tomblin 2005 ; Thordardottir, Reference Thordardottir 2005 ; Watkins, Kelly, Harbers & Hollis, Reference Watkins, Kelly, Harbers and Hollis 1995 ). We computed this measure separately in English and Mandarin from the transcription data. For English, different words originating from the same word root (e.g., ‘eat-ate-eaten’) were considered as a single word root. NDWR was divided by the recording duration of each individual child because the recording duration varied from child to child (see Aukrust & Rydland, Reference Aukrust and Rydland 2011 ). Proper nouns (e.g., “Tangled”–the title of an English movie, “小兔跳楼” (xiao3tu4tiao4lou2)–the name of a local hand game) and unintelligible words were excluded from the computation of this measure.

Expressive language measures: Mean length of pure utterances (MLU)

Mean length of utterances (MLU) for English and Mandarin were calculated from the transcription data based on the guidelines provided in CHAT and CLAN (MacWhinney, Reference MacWhinney 2000 ). MLU, the ratio of morphemes over utterances (Brown, Reference Brown 1973 ), is frequently used as a measure of sentence complexity (Klee, Stokes, Wong, Fletcher & Gavin, Reference Klee, Stokes, Wong, Fletcher and Gavin 2004 ; Mishina-Mori, Reference Mishina-Mori 2011 ; Thordardottir, Reference Thordardottir 2005 ). Some researchers have noted that MLU is only meaningful until approximately 4 to 5 morphemes (Bernstein & Tiegerman-Farber, Reference Bernstein and Tiegerman-Farber 1997 ), while others have claimed that MLU is a valid measure even into the grade school years (Jones, Weismer & Schumacher, 2000; Miller, Frieberg, Rolland & Reves, Reference Miller, Frieberg, Rolland and Reves 1992 ). As there is currently no consensus on the age limit and morpheme-count limit of MLU (the mean MLUs in our study fall in the range of 4 to 5 morphemes: Mean English MLU = 5.07; Mean Mandarin MLU = 4.26), we proceeded to calculate both English and Mandarin MLU for our study and analyses.

Utterances included in the computation of MLU were those that only consisted of English or Mandarin words (i.e., pure utterances in English and Mandarin). Utterances with unintelligible words were included in the analysis, but the unintelligible words were excluded from the morpheme count. This approach was employed because noise from the environment decreased the intelligibility of many words (see Thordardottir, Reference Thordardottir 2005 ); both childcare centers had open classrooms and thus, voices of children in other groups or classrooms sometimes interfered with the recordings.

Code-switching measures

We coded two types of code-switching from the children's utterances: intra-sentential switches and inter-sentential switches (see Table 1 ). The total amount of code-switched utterances was the sum of both types of code-switching. The percentage of the total number of code-switched utterances made by each child was obtained by dividing the total number of code-switched utterances by the total number of utterances spoken by each child.

Table 1. Types and Examples of Code-switching.

A total duration of 21:16:43 hours and 30:09:48 hours of observation in Center 1 and Center 2, respectively, was transcribed and analyzed. An average of 648.78 utterances per child was recorded ( SD = 542.37; see Table 2 ). The number of observed switches (intra-sentential and inter-sentential switches) in our sample of children constituted a small percentage of their total utterances ( M = 8.95%, SD = 9.60, range = .23% to 33.83%.). There was no child who did not code-switch at all. Of the code-switched utterances, children engaged in similar amounts of intra-sentential switches and inter-sentential switches ( M = 4.46% and 4.48% respectively). Children also produced a greater number of pure English utterances than pure Mandarin utterances and code-switch utterances ( M = 77.01% vs. 16.57% and 8.95%, respectively), and a greater amount of English NDWR per minute and MLU than Mandarin NDWR per minute and MLU ( M = 1.87 and 5.07 vs. .88 and 4.26), reflecting the population's dominance in English language. Preliminary analysis showed that both types of code-switched utterances (i.e., intrasentential and intersentential switches) did not differ in their relationship with the other measures of language competency, hence they were combined as the total number of code-switched utterances in subsequent analyses. Children varied in how much they spoke during the observation period, so the total number of code-switched utterances is divided by the total number of utterances to obtain a percentage of code-switched utterances for each child. In addition, we observed that the teachers themselves did not code-switch when they interacted with the children. The teachers also did not explicitly encourage or discourage code-switching from the children, although they made efforts to speak in only one language to the children.

Table 2. Measures of Children's Spontaneous Speech.

Note: The total number of all utterances is the sum of intra-sentential switch utterances, pure English utterances, pure Mandarin utterances, and other utterances such as single proper nouns, translation, and imitation. Inter-sentential switch utterances comprise only pure utterances.

Correlational analyses

As some of the measures of interest were not normally distributed, Spearman correlations were used. Partial correlations, controlled for age, between the various measures of language competency and percentage of code-switched utterances were conducted ( Table 3 ). No significant correlations between measures of English competency (MLU of pure English utterances, English NDWR per minute, and English PPVT) and percentage of code-switched utterances were found. Thus, the amount of code-switched utterances was not significantly related to both the expressive and receptive measures of English competency. On the other hand, correlations between measures of Mandarin competency (MLU of pure Mandarin utterances, and Mandarin NDWR per minute) and percentage of code-switched utterances were positive and significant, r = .72 and r = .91, respectively, p s < .001, indicating that children who code-switched more also produced Mandarin sentences that are more complex and consist of a larger variety of words than children who code-switched less.

Table 3. Spearman Partial Correlations between Measures of Language Competency and Percentage of Code-Switched Utterances (controlled for age).

*Bonferroni corrected p value = .003

Do children code-switch because of poor language competency?

If children code-switch because they are weak in a language, then their language competency (i.e., NDWR per minute and PPVT) would negatively predict the amount of code-switching (i.e., percentage of code-switched utterances) they engaged in. To test this hypothesis, two multiple hierarchical regressions were conducted, one controlled for age and home English exposure, and the other controlled for age and home Mandarin exposure, since home English exposure and home Mandarin Exposure were highly negatively correlated with each other ( r = −.96, p < .001). The two control variables were entered in Step 1 and the language competency variables were entered in Step 2. We noted that the field has not reached a consensus about the usefulness of MLU in measuring language complexity in children aged 5 to 6 years old. Nevertheless, separate regression analyses were conducted with and without MLU as one of the language predictors and similar results were obtained. Given also that NDWR per minute was highly correlated with MLU ( r s = .55 to .81, p s < .01), we thus presented the regression analyses that included only NDWR per minute as a language predictor of code-switching. The change in R 2 was significant in Step 2 for both regression models (see Table 4a for regression analysis controlled for age and home English exposure, and Table 4b for regression analysis controlled for age and home Mandarin exposure). Both final regression models significantly predicted the percentage of code-switched utterances, F (5, 30) = 11.87, p < .001 (home English exposure), and F (5, 30) = 11.79, p < .001 (home Mandarin exposure), accounting for 66.4% and 66.3% of the variance, respectively. However, only Mandarin NDWR per minute significantly and positively predicted the percentage of code-switched utterances, over and beyond age and home language exposure: β = .79, t (30) = 5.09, p < .001 (home English exposure), and β = .76, t (30) = 5.34, p < .001 (home Mandarin exposure). English NDWR per minute and PPVT were not significant predictors of the percentage of code-switched utterances ( p s > .10).

Table 4a. Summary of Hierarchical Regression Analysis for Language Competency Variables Predicting Percentage of Code-Switched Utterances when Home English Exposure is controlled for.

* p < .05, ** p < .01, *** p < .001

Note: Similar results were obtained if Mandarin MLU and English MLU were included in Step 2: Mandarin NDWR per minute remained significant (β = .50, t (28) = 3.00, p =.006) while English NDWR per minute and PPVT standard scores remained non-significant (β = .17, t (28) = 1.34, p = .19 and β = −.04, t (28) = −.36, p = .72, respectively).

Table 4b. Summary of Hierarchical Regression Analysis for Language Competency Variables Predicting Percentage of Code-Switched Utterances when Home Mandarin Exposure is controlled for.

Note: Similar results were obtained if Mandarin MLU and English MLU were included in Step 2: Mandarin NDWR per minute remained significant (β = .46, t (28) = 2.85, p =.008) while English NDWR per minute and PPVT standard scores remained non-significant (β = −.20, t (28) = 1.30, p = .21 and β = −.02, t (28) = −.17, p = .87, respectively).

Does code-switching affect language competency?

Another important question that the current literature has not yet been able to address is whether code-switching would negatively affect language development in bilingual children. Here, we analyzed teachers’ ratings of the children's English and Mandarin competency approximately six months after the observation sessions. First, Spearman correlation analyses (controlled for age) between measures of language competency, percentage of code-switched utterances, and teachers’ ratings were conducted. Teachers’ ratings of both English and Mandarin competency were positively correlated with the various measures of language competency and code-switched utterances, except English PPVT (see Table 5 and 6 ). Next, separate three-step hierarchical regression analyses were used to examine whether the percentage of code-switched utterances and the other measures of language competency obtained during the observation sessions (i.e., time 1 - T1) predicted teachers’ ratings of language competency 6 months later (i.e., time 2 - T2). Age was entered in Step 1 as a control variable. Percentage of code-switched utterances at T1 was entered in Step 2, to examine whether this variable is significant in predicting teachers’ ratings of language competency at T2 on its own (controlled for age). Finally, language competency variables at T1 (i.e., NDWR per minute and PPVT for English) were added in Step 3 as predictors of teachers’ ratings of language competency at T2. As with earlier regression analyses, we obtained similar results with and without MLU as one of the language predictors, thus we included only NDWR per minute, and not MLU, as one of the predictors of teachers’ ratings of language competency in our final regression analyses.

Table 5. Spearman Partial Correlations between Measures of English Language Competency, Percentage of Code-Switched Utterances and Teachers’ Ratings of English Language Competency (controlled for age).

*Bonferroni corrected p value = .005

Table 6. Spearman Partial Correlations between Measures of Mandarin Language Competency and Percentage of Code-Switched Utterances and Teachers’ Ratings of Mandarin Language Competency (controlled for age).

*Bonferroni corrected p value = .008

For teachers’ ratings of English competency at T2, the change in R 2 was significant when the percentage of code-switched utterances at T1 was added in Step 2 and when English NDWR per minute and PPVT standard scores at T1 were added in Step 3 (see Table 7 ). The final regression model with four predictors at T1 (age, percentage of code-switched utterances, English NDWR per minute, and PPVT) significantly predicted teachers’ ratings of English competency at T2, F (4, 24) = 14.01, p < .001, and accounted for 70.0% of the variance. Both percentage of code-switched utterances and English NDWR per minute at T1 were significant predictors of teacher's ratings of English competency at T2, β = .36, t (24) = 2.81, p = .01, and β = .60, t (24) = 5.15, p < .001, respectively.

Table 7. Summary of Hierarchical Regression Analysis for Language Variables at Time 1 (T1) Predicting Teachers’ Ratings of English Competency at Time 2 (T2).

Note: Similar results were obtained if English MLU was included in Step 3: Percentage of code-switched utterances and English NDWR per minute remained significant (β = .34, t (23) = 2.67, p = .014, and β = .50, t (23) = 3.03, p = .006, respectively), while PPVT standard scores remained non-significant (β = .17, t (23) = 1.34, p = .19).

For teachers’ ratings of Mandarin competency at T2, the change in R 2 was also significant when the percentage of code-switched utterances at T1 was added in Step 2, and when Mandarin NDWR per minute at T1 was added in Step 3 (see Table 8 ). The final regression model with three predictors at T1 (age, percentage of code-switched utterances, and Mandarin NDWR per minute) significantly predicted teachers’ ratings of Mandarin competency at T2, F (3, 46) = 15.68, p < .001, and accounted for 50.6% of the variance. Mandarin NDWR per minute at T1 significantly predicted teachers’ ratings of Mandarin competency at T2, β = .80, t (46) = 4.73, p < .001. While percentage of code-switched utterances at T1 was a significant predictor of teachers’ rating of Mandarin competency at T2 in Step 2, it was no longer significant when Mandarin NDWR per minute was added to the model in Step 3.

Table 8. Summary of Hierarchical Regression Analysis for Language Variables at Time 1 (T1) Predicting Teachers’ Ratings of Mandarin Competency at Time 2 (T2).

* p < .05, ** p < .01, *** p < .001, + p = .051

Note: Similar results were obtained if Mandarin MLU was included in Step 3: Mandarin NDWR per minute remained significant (β = .68, t (45) = 3.45, p =.001), while percentage of code-switched utterances remained non-significant (β = −.18, t (45) = −1.06, p = .29).

As percentage of code-switched utterances was no longer significant after Mandarin NDWR per minute was added to the model, it is possible that Mandarin NDWR per minute mediated the relationship between teachers’ ratings of Mandarin competency and code-switched utterances. A mediation analysis was conducted using a four-step approach in hierarchical regressions (Baron & Kenny, Reference Baron and Kenny 1986 ). Controlled for age, the percentage of code-switched utterances and Mandarin NDWR per minute at T1 individually predicted teachers’ ratings of Mandarin competency at T2, β = .46, t (47) = 3.47, p = .001, and β = .69, t (47) = 6.27, p < .001, respectively (see Figure 1 and Table A2 for the detailed regressions). The percentage of code-switched utterances also significantly predicted Mandarin NDWR per minute at T1, β = .76, t (47) = 7.98, p < .001. However, the percentage of code-switched utterances was no longer significant when Mandarin NDWR per minute was controlled for, β = −.15, t (45) = −.86, p = .39. Thus, Mandarin NDWR per minute at T1 fully mediated the relationship between percentage of code-switched utterances at T1 and teachers’ ratings of Mandarin competency at T2. The mediation effect of Mandarin NDWR per minute at T1 suggests that the amount of code-switched utterances indirectly leads to higher levels of Mandarin competency six months later through higher Mandarin NDWR per minute during the observation period.

Figure 1. Mediation effect of Mandarin NDWR per minute.

4. General discussion

The present study investigated the relationship between children's code-switching behaviour and their language competency in preschool settings. Five- to six-year-old English–Mandarin bilinguals were observed across five days in their childcare centers. The types and amount of code-switched utterances produced by children in their daily conversations, along with their NDWR per minute and MLU for English and Mandarin were measured and analyzed. In addition to the observation sessions, children were administered the English PPVT, and teachers were asked to rate the children's English and Mandarin language competency six months after the observation sessions.

Despite lower levels of proficiency in expressive Mandarin as compared to expressive English, results indicated that the number of code-switched utterances was positively related to Mandarin expressive language competency (NDWR per minute), over and above home language exposure. In other words, children who code-switched more tended to produce a larger variety of Mandarin words, even though this bilingual population is less dominant in Mandarin compared to English. In addition, English language competency (both expressive and receptive) was not significantly related to the amount of code-switched utterances. This is consistent with recent studies that showed that code-switching is not a result of language incompetency (e.g., Cantone, Reference Cantone 2007 ). Most importantly, analyses conducted with teachers’ ratings of children's language competences for Mandarin and English showed that the amount of code-switched utterances (over and beyond the current levels of expressive English) positively predicted children's English and Mandarin competency six months later, but the latter relationship is mediated by their current levels of expressive Mandarin.

These findings illustrate that, contrary to popular belief, code-switching in bilingual children does not signal linguistic incompetency . Rather, code-switching is positively associated with language competency. These findings put forward the possibility that children may be using code-switching as a platform to aid the development of their languages, especially the weaker one. Young bilingual children may not be able to express themselves fully and accurately in both of their languages yet. Code-switching thus allows them to explore and use both languages (the weaker language with the stronger one) while keeping the intended meaning intact. This is in line with the Ivy Hypothesis, which argues that children code-switch to improve their weaker language by using the grammatical structure they have acquired in their stronger language (e.g., Bernardini & Schlyter, Reference Bernardini and Schlyter 2004 ; Gawlitzek-Maiwald & Tracy, Reference Gawlitzek-Maiwald and Tracy 1996 ). Bernardini and Schlyter ( Reference Bernardini and Schlyter 2004 ) found that the majority of the mixed utterances produced by Swedish–Italian and Swedish–French bilingual children consisted of single words or simple phrases in their weaker language combined with more complex phrases from their stronger language. Additional support for this hypothesis comes from a study illustrating syntactic transfers of wh in-situ interrogatives and prenominal relative clauses from Cantonese to English in a Cantonese–English bilingual child during the period when his Cantonese syntactic development was significantly ahead of his English syntactic development (Yip & Matthews, Reference Yip and Matthews 2000 ). Thus, code-switching can be used as a scaffold for the weaker language, where more complex syntactic structures from the stronger language are used in combination with lexical items and simpler syntactic structures from the weaker language.

This argument that code-switching may be helpful to young bilingual learners is not dissimilar to that proposed by researchers on translanguaging. According to Garcia and Wei (2013), translanguaging refers to the idea that bilingual speakers have one linguistic repertoire that holds concepts socially constructed from both languages. Research on translanguaging, which focused mainly on the use of translanguaging within the classroom context, have found that translanguaging facilitates deeper thinking in bilingual students, and can be used by teachers to aid bilingual speakers in subjects taught in their weaker language (e.g., Creese & Blackledge, Reference Creese and Blackledge 2010 ; Hornberger & Link, Reference Hornberger and Link 2012 ). Teachers can leverage on the bilingual students’ stronger language to provide the students with a platform to participate, elaborate on their thought processes and raise questions. This is similar to our proposal that code-switching allows young bilingual children to leverage on their stronger language in daily communications and dual language learning contexts. It provides young bilingual children with an alternative tool to express their thoughts, feelings, and ideas. For these young learners of two languages, code-switching can be used as a form of communicative support and as a way to expand these emergent bilinguals’ understanding and linguistic competency.

This follows that, despite mounting concerns about the potential negative impacts of parental code-switching on an infant's language development (Byers-Heinlein, Reference Byers-Heinlein 2013 ), children's code-switching behaviour itself does not necessarily indicate linguistic incompetency, nor will it negatively affect children's language development. Research has shown that language input from parents and teachers are critical to children's language development, in terms of vocabulary size (Bowers & Vasilyeva, Reference Bowers and Vasilyeva 2011 ; Hammer, Davison, Lawrence & Miccio, Reference Hammer, Davison, Lawrence and Miccio 2009 ; Hoff, Reference Hoff 2006 ; Hoff, Core, Place, Rumiche, Señor & Parra, Reference Hoff, Core, Place, Rumiche, Señor and Parra 2012 ; Hurtado, Marchman & Fernald, Reference Hurtado, Marchman and Fernald 2008 ), grammatical development (Blom, Reference Blom 2010 ; Bohman, Bedore, Peña, Mendez-Perez & Gillam, Reference Bohman, Bedore, Peña, Mendez-Perez and Gillam 2010 ) and comprehension skills (Dickinson & Porche, Reference Dickinson and Porche 2011 ; Huttenlocher, Vasilyeva, Cymerman & Levine, Reference Huttenlocher, Vasilyeva, Cymerman and Levine 2002 ). However, children's usage of the language(s) is also an important factor in language development. Studies, for example, found that increasing the use of a second language is associated with improved proficiency in that language (Freed, Segalowitz & Dewey, Reference Freed, Segalowitz and Dewey 2004 ; Martinsen, Baker, Bown & Johnson, Reference Martinsen, Baker, Bown and Johnson 2011 ). Thus, while it remains important that the language input children receive should accurately reflect the linguistic characteristics of the target languages, bilingual children's regular use of both languages should be highly encouraged too, even if it involves switching between the two languages. Our results elucidate that the act of code-switching by children may have provided them with a way to engage both their languages more frequently, particularly the weaker language. To put it simply, code-switching in a multilingual environment may present bilingual children with opportunities to use both their languages in ways that a pure language environment alone would not be able to provide them with. This, in turn, has a positive outcome on language development with improved proficiency.

Language and literacy development guidelines from the local curriculum framework were used in this study to create a teacher's rating scale to measure children's language competencies. This scale is relevant to the local context and it captures the major receptive and expressive language proficiency requirements as expected of a 6-year-old. Similar uses of teachers’ assessment of children's language skills are often employed in both research and educational settings (e.g., August, Shanahan & Escamilla, Reference August, Shanahan and Escamilla 2009 ; Sundberg & Partington, Reference Sundberg and Partington 1998 ). Nevertheless, this scale has some limitations. First, the rating scale is not a standardized measurement of language competency. Second, teachers’ ratings could be subjective and were based on retrospective reporting. Thus, measures of children's language competency based on teachers’ ratings may not reflect children's full extent of their language ability. However, finding an assessment tool that has been validated for use in multiple languages, in this case English and Mandarin, is a challenge. Standardized measures of language ability such as the Peabody Picture Vocabulary Test (Dunn & Dunn, Reference Dunn and Dunn 2007 ), subtests of Woodcock-Johnson Tests of Achievement (Woodcock, McGrew & Mather, Reference Woodcock, McGrew and Mather 2001 ), and Clinical Evaluation of Language Fundamental Preschool (CELF; Wiig, Secord & Semel, Reference Wiig, Secord and Semel 2004 ) are available in English but not in Mandarin. Even if they are available in Mandarin, there remains a possibility that the materials of these language assessment tools may not be appropriate for the local children and thus may not accurately reflect their true language ability (Brebner, Rickard-Liow & McCormack, Reference Brebner, Rickard-Liow, McCormack and Lind 2000 ; Carter, Lees, Murira, Gona, Neville & Newton, Reference Carter, Lees, Murira, Gona, Neville and Newton 2005 ). Furthermore, researchers have not agreed on the assessment tools that are best for assessing the language ability of bilingual children (e.g., Bedore & Pena, Reference Bedore and Pena 2008 ; Gutiérrez-Clellen, Reference Gutiérrez-Clellen 2002 ; Saenz & Huer, Reference Saenz and Huer 2003 ), and published self-report tools are not appropriate for use by children (e.g., LEAP-Q; Marian, Blumenfeld & Kaushanskaya, Reference Marian, Blumenfeld and Kaushanskaya 2007 ). Taking these challenges into consideration, we believe that our teacher's rating scale of language competency is a relatively appropriate language measure in this context, albeit with limitations. Future studies should consider replicating the present study with a different population of children where standardized language assessment tools in both languages are available and well tested.

It is worth noting that a bilingual child's exposure to his or her two languages may vary a lot during six months due to external factors, such as changed input at home or at school, or internal factors, such as when the child identifies himself or herself more with one language's culture or when the child refuses to speak in one language, Such factors may affect a child's code-switching behavior as well. While we agree that bilingualism could vary substantially in an individual within six months, we believe that the bilingual status of the children in our study is relatively stable throughout the study. The children in our study had been in the same preschool for the past 6 months with no known significant changes to their family or parental background, or to their preschool routines and teachers. Thus, it is unlikely that these children experienced a significant change in their language environments. Nevertheless, future studies should consider collecting information about possible external and internal factors affecting children's language balance during the period of study, such as through additional parental and teachers’ surveys.

We have earlier raised the possibility that the use of MLU to assess children's language complexity in this study may be limited due to a ceiling in age and morpheme count. Compounding this issue is the challenge of computing MLU as an indicator of language competence across two typologically different languages like English and Mandarin. The difficulty in using MLU to assess children's English language complexity is particularly noted in the case of Singapore Colloquial English (SCE). Two common features of SCE are the absence of subject, for example, (e.g., “ (That car) very expensive, you know”) and the deletion of the copula ‘be’ (e.g., “that boat ø very short one”) (Leimgruber, Reference Leimgruber 2011 ). Thus, the use of English MLU to assess children's English language complexity in this context may be limited. Future investigations examining whether code-switching affects children's syntactic development of their two languages could employ more specific measures other than MLU, e.g., an elicited imitation task to assess sentence formation (Lust, Chien & Flynn, Reference Lust, Chien, Flynn and Lust 1987 ; Lust, Flynn & Foley, Reference Lust, Flynn, Foley, McDaniel, McKee and Cairns 1996 ), the truth-value judgment task (Lust & Blume, Reference Lust and Blume 2016 ), or an adaptation of adult syntactic complexity tasks that use not just length measures but also ratio measures such as sub-clauses/sentence (e.g., Bulté & Housen, Reference Bulté and Housen 2012 ; Norris & Ortega, Reference Norris and Ortega 2009 ).

Children code-switch for various reasons. Future studies can tap on multiple age groups or conduct a longitudinal study to further investigate the developmental shifts in the use of code-switched utterances. For example, studies that have found that children engaged in code-switching behavior in order to fill their lexical gaps considered younger children typically around or before the age of 3 years (Cantone, Reference Cantone 2007 ). Studies that have found that children make use of code-switches for sociocultural or pragmatic purposes have focused mainly on older children (Chung, Reference Chung 2006 ; Reyes, Reference Reyes 2004 ; Vu et al., Reference Vu, Bailey and Howes 2010 ). It is logical that very young bilingual children first start off with a limited lexicon, and would, therefore, code-switch when they do not have the translation equivalent of a particular concept (De Houwer, Reference De Houwer, Kroll and de Groot 2005 ; Cantone, Reference Cantone 2007 ). After acquiring a sizeable lexicon in both languages and learning their sociolinguistic rules and cultural practices, older bilingual children would begin to code-switch depending on the social demands of the conversation (Nicoladis & Genesee, Reference Nicoladis and Genesee 1997 ). The children in our study were 5 to 6 years of age and were likely aware, to some degree, of the sociolinguistic rules used in their community. Including multiple age groups in future studies would shed light on the developmental trends in bilingual children's code-switching behavior.

In conclusion, the goal of the present study was to further elucidate the relationship between bilingual children's code-switching behavior in their larger language environment and their linguistic competency. Debate about whether bilingual children's code-switching behavior reflects their linguistic incompetency is ongoing. The present study is the first attempt to investigate this relationship using a quantitative approach. Findings from the present study provide counter-evidence against the linguistic incompetency hypothesis – there was no indication that bilingual children's code-switching behavior was a result of their linguistic incompetency. Instead, bilingual children's code-switching behavior suggests greater language competency. The findings from the present study provide an alternative perspective on the linguistic incompetency hypothesis – that, far from being debilitating, code-switching plays an important and positive role in language development of bilingual children, especially in the context of the larger language environment.

To view supplementary material for this article, please visit https://doi.org/10.1017/S1366728917000335

Appendix 1. Questionnaire for Teacher's Ratings of Language Competency.

Appendix 2. Summary of Mediation Analysis of Mandarin NDWR Per Minute.

* We are grateful to the children and parents who participated in the study and to the teachers and staff of Creative O Preschoolers’ Bay, and Red SchoolHouse. We thank Ferninda Patrycia, Xiaoqian Li, Wanyu Hung, Yvonne Yong, Wei Xing Toh, Lu Xing, Qi Xuan Yap, Tony Zhao Ming Lim, and Yuxin Lou for their help in this study. The corpus from this study is published in CHILDES/Biling/Singapore. Portions of this work were previously presented at the BUCLD 39 (2014) and in Yow, Patrycia, and Flynn (2016). This research was supported by the SUTD SRG grant (SRG HASS 2011 011) and the SUTD-MIT IDC grant (IDG31100106 and IDD41100104) awarded to the first author.

Supplementary material can be found online at https://doi.org/10.1017/S1366728917000335

1 Singapore is made up of 74.3% ethnic Chinese, 13.3% ethnic Malay, 9.1% ethnic Indian and 3.2% Others (Singapore Department of Statistics, 2015). English is the official working language while Mandarin, Bahasa Melayu and Tamil are the official mother tongue languages. Singapore adopts a bilingualism policy that obliges all children enrolled in Singapore schools to learn English and a mother tongue according to their ethnicity, but most subjects are taught in English (Gopinathan, 1999). Hence, all Singaporean children speak English as a shared language, but they are also expected to be fluent in their mother tongue language.

2 Many childcare centers conduct lessons in English and one of the mother tongue languages to prepare the children before entering elementary school.

3 Singlish, also known as Singapore Colloquial English, is a creolized form of English spoken in Singapore (Platt, 1975).

Yow et al supplementary material

This article has been cited by the following publications. This list is generated based on data provided by Crossref .

• Google Scholar

View all Google Scholar citations for this article.

Save article to Kindle

To save this article to your Kindle, first ensure [email protected] is added to your Approved Personal Document E-mail List under your Personal Document Settings on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part of your Kindle email address below. Find out more about saving to your Kindle .

Note you can select to save to either the @free.kindle.com or @kindle.com variations. ‘@free.kindle.com’ emails are free but can only be saved to your device when it is connected to wi-fi. ‘@kindle.com’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.

Find out more about the Kindle Personal Document Service.

• Volume 21, Issue 5
• W. QUIN YOW (a1) , JESSICA S. H. TAN (a1) and SUZANNE FLYNN (a2)
• DOI: https://doi.org/10.1017/S1366728917000335

Save article to Dropbox

To save this article to your Dropbox account, please select one or more formats and confirm that you agree to abide by our usage policies. If this is the first time you used this feature, you will be asked to authorise Cambridge Core to connect with your Dropbox account. Find out more about saving content to Dropbox .

Save article to Google Drive

To save this article to your Google Drive account, please select one or more formats and confirm that you agree to abide by our usage policies. If this is the first time you used this feature, you will be asked to authorise Cambridge Core to connect with your Google Drive account. Find out more about saving content to Google Drive .

Reply to: Submit a response

- No HTML tags allowed - Web page URLs will display as text only - Lines and paragraphs break automatically - Attachments, images or tables are not permitted

Your details

Your email address will be used in order to notify you when your comment has been reviewed by the moderator and in case the author(s) of the article or the moderator need to contact you directly.

You have entered the maximum number of contributors

Conflicting interests.

Please list any fees and grants from, employment by, consultancy for, shared ownership in or any close relationship with, at any time over the preceding 36 months, any organisation whose interests may be affected by the publication of the response. Please also list any non-financial associations or interests (personal, professional, political, institutional, religious or other) that a reasonable reader would want to know about in relation to the submitted work. This pertains to all the authors of the piece, their spouses or partners.

• DOI: 10.15282/ijleal.v13i2.9585
• Corpus ID: 265267461

Issues and Functions of Code-switching in Studies on Popular Culture: A Systematic Literature Review

• Siti Nur Atikah Nazri , Asiah Kassim
• Published in International Journal of… 15 October 2023
• Linguistics
• International Journal of Language Education and Applied Linguistics

Figures from this paper

One Citation

Code switching in moroccan nursing education: exploring pedagogical functions and professors’ attitudes, 48 references, systematic literature review of conversational code-switching in multilingual society from a sociolinguistic perspective, the role of code-switching in the creation of an outsider identity in the bilingual film.

• Highly Influential

Functional and sociocultural attitudes of code-switching and its relation to the meaning-making process: The case of dubbing Kim Possible into Arabic

English–greek code-switching in greek cypriot magazines and newspapers – an analysis of its textual forms and functions, translation and code switching in the lyrics of bilingual popular songs, language and advertising effectiveness: code-switching in the korean marketplace, code-switching in italo-brazilian literature from rio grande do sul and são paulo: a sociolinguistic analysis of the forms and functions of literary code-switching, the the impact of code switching in conversation of “nebeng boy” youtube vlogs towards communication in english among the participants, code-switching in gol & gincu, bilingualism and identity construction in the digital discourse, related papers.

Showing 1 through 3 of 0 Related Papers

2003 Articles

Issues in Code-Switching: Competing Theories and Models

Boztepe, Erman

This paper provides a critical overview of the theoretical, analytical, and practical questions most prevalent in the study of the structural and the sociolinguistic dimensions of code-switching (CS). In doing so, it reviews a range of empirical studies from around the world. The paper first looks at the linguistic research on the structural features of CS focusing in particular on the code-switching versus borrowing distinction, and the syntactic constraints governing its operation. It then critically reviews sociological, anthropological, and linguistic perspectives dominating the sociolinguistic research on CS over the past three decades. Major empirical studies on the discourse functions of CS are discussed, noting the similarities and differences between socially motivated CS and style-shifting. Finally, directions for future research on CS are discussed, giving particular emphasis to the methodological issue of its applicability to the analysis of bilingual classroom interaction.

• Linguistics--Research
• Code switching (Linguistics)
• English language--Study and teaching--Foreign speakers
• Linguistics

Also Published In

More about this work.

• DOI Copy DOI to clipboard

The Decades Progress on Code-Switching Research in NLP : A Systematic Survey on Trends and Challenges

Genta Winata , Alham Fikri Aji , Zheng Xin Yong , Thamar Solorio

Export citation

• Preformatted

Markdown (Informal)

[The Decades Progress on Code-Switching Research in NLP: A Systematic Survey on Trends and Challenges](https://aclanthology.org/2023.findings-acl.185) (Winata et al., Findings 2023)

• The Decades Progress on Code-Switching Research in NLP: A Systematic Survey on Trends and Challenges (Winata et al., Findings 2023)
• Genta Winata, Alham Fikri Aji, Zheng Xin Yong, and Thamar Solorio. 2023. The Decades Progress on Code-Switching Research in NLP: A Systematic Survey on Trends and Challenges . In Findings of the Association for Computational Linguistics: ACL 2023 , pages 2936–2978, Toronto, Canada. Association for Computational Linguistics.

Information

• Author Services

Initiatives

You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.

All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess .

Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.

Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.

Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.

Original Submission Date Received: .

• Active Journals
• Find a Journal
• Proceedings Series
• For Authors
• For Reviewers
• For Editors
• For Librarians
• For Publishers
• For Societies
• For Conference Organizers
• Open Access Policy
• Institutional Open Access Program
• Special Issues Guidelines
• Editorial Process
• Research and Publication Ethics
• Article Processing Charges
• Testimonials
• Preprints.org
• SciProfiles
• Encyclopedia

Article Menu

• Subscribe SciFeed
• Recommended Articles
• Google Scholar
• on Google Scholar
• Table of Contents

Find support for a specific problem in the support section of our website.

Please let us know what you think of our products and services.

Visit our dedicated information section to learn more about MDPI.

JSmol Viewer

Building educational technologies for code-switching: current practices, difficulties and future directions.

1. Introduction

2. CSW in an ESL Context

2.1. terminologies, 2.1.1. code-switching vs. code-mixing, 2.1.2. code-switching vs. borrowing, 2.1.3. code-switching vs. translanguaging, 2.2. why do learners code-switch, 2.3. why do teachers code-switch, 2.4. current policies and practices, 2.5. code-switching in digital learning environments, 3. difficulties of current nlp technologies for esl teaching and learning, 3.2. difficulties of current nlp technologies for csw, 3.3. educational nlp case study, 3.3.1. case 1: feedback technologies.

3.3.2. Case 2: Assessment Technologies

4. future directions.

4.2. Improving NLP Models for Educational CSW

5. final thoughts, author contributions, institutional review board statement, informed consent statement, data availability statement, conflicts of interest.

• Aaron, Jessi Elana. 2015. Lone English-origin nouns in Spanish: The precedence of community norms. International Journal of Bilingualism 19: 459–80. [ Google Scholar ] [ CrossRef ]
• Adinolfi, Lina, and Lluïsa Astruc. 2017. An exploratory study of translanguaging practices in an online beginners’ foreign language classroom. Language Learning in Higher Education. Journal of the European Confederation of Language Centres in Higher Education (CercleS) 7: 185–204. [ Google Scholar ] [ CrossRef ]
• Aguilar, Gustavo, Sudipta Kar, and Thamar Solorio. 2020. LinCE: A centralized benchmark for linguistic code-switching evaluation. Paper presented at 12th Language Resources and Evaluation Conference, Marseille, France, May 11–16; Paris: European Language Resources Association, pp. 1803–13. [ Google Scholar ]
• Ahmad, Badrul, and Kamaruzaman Jusoff. 2009. Teachers’ code-switching in classroom instructions for low english proficient learners. English Language Teaching 2: 49–55. [ Google Scholar ] [ CrossRef ]
• Alikaniotis, Dimitrios, Helen Yannakoudakis, and Marek Rei. 2016. Automatic text scoring using neural networks. Paper presented at 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Berlin, Germany, August 7–12; Stroudsburg: Association for Computational Linguistics, pp. 715–25. [ Google Scholar ] [ CrossRef ]
• Andersen, Øistein E., Rebecca Watson, Zheng Yuan, and Kevin Yet Fong Cheung. 2021. Benefits of alternative evaluation methods for automated essay scoring. Paper presented at 14th International Conference on Educational Data Mining (EDM 2021), Virtual. June 29–July 2. [ Google Scholar ]
• Anderson, Jason, and Amy Lightfoot. 2021. Translingual practices in english classrooms in india: Current perceptions and future possibilities. International Journal of Bilingual Education and Bilingualism 24: 1210–31. [ Google Scholar ] [ CrossRef ]
• Backus, Ad, and Margreet Dorleijn. 2009. Loan Translations versus Code-Switching . Cambridge Handbooks in Language and Linguistics. Cambridge: Cambridge University Press, pp. 75–94. [ Google Scholar ] [ CrossRef ]
• Bell, Samuel, Helen Yannakoudakis, and Marek Rei. 2019. Context is key: Grammatical error detection with contextual word representations. Paper presented at Fourteenth Workshop on Innovative Use of NLP for Building Educational Applications, Florence, Italy, August 2; Stroudsburg: Association for Computational Linguistics, pp. 103–15. [ Google Scholar ] [ CrossRef ]
• Brevik, Lisbeth M., and Ulrikke Rindal. 2020. Language use in the classroom: Balancing target language exposure with the need for other languages. TESOL Quarterly 54: 925–53. [ Google Scholar ] [ CrossRef ]
• Bullock, Barbara E., and Almeida Jacqueline Ed Toribio. 2009. The Cambridge Handbook of Linguistic Code-Switching . Cambridge Handbooks in Language and Linguistics. Cambridge: Cambridge University Press. [ Google Scholar ] [ CrossRef ]
• Caines, Andrew, Christian Bentz, Calbert Graham, Tim Polzehl, and Paula Buttery. 2016. Crowdsourcing a Multi-lingual Speech Corpus: Recording, Transcription and Annotation of the C rowd ED corpus . Paper presented at Tenth International Conference on Language Resources and Evaluation (LREC 2016), Paris, France, May 23–28; Paris: European Language Resources Association (ELRA). [ Google Scholar ]
• Caines, Andrew, Helen Yannakoudakis, Helena Edmondson, Helen Allen, Pascual Pérez-Paredes, Bill Byrne, and Paula Buttery. 2020. The teacher-student chatroom corpus. Paper presented at 9th Workshop on NLP for Computer Assisted Language Learning, Gothenburg, Sweden, November 20; Linköping: LiU Electronic Press, pp. 10–20. [ Google Scholar ]
• Carstens, Adelia. 2016. Translanguaging as a vehicle for l2 acquisition and l1 development: Students’ perceptions. Language Matters 47: 203–22. [ Google Scholar ] [ CrossRef ]
• Chen, Jing, James H. Fife, Isaac I. Bejar, and Andre A. Rupp. 2016. Building e-rater ® Scoring Models Using Machine Learning Methods. ETS Research Report Series 2016: 1–12. [ Google Scholar ] [ CrossRef ]
• Chodorow, Martin, and Jill Burstein. 2004. Beyond essay length: Evaluating e-rater ® ’s performance on toefl ® essays. ETS Research Report Series 2004: 1–38. [ Google Scholar ] [ CrossRef ]
• Clark, Kevin, Minh-Thang Luong, Quoc V. Le, and Christopher D. Manning. 2020. ELECTRA: Pre-training Text Encoders as Discriminators Rather than Generators. Paper presented at International Conference on Learning Representations, Addis Ababa, Ethiopia, April 26–30. [ Google Scholar ]
• Clyne, Michael. 2003. Dynamics of Language Contact: English and Immigrant Languages . Cambridge: Cambridge University Press. [ Google Scholar ] [ CrossRef ]
• Conneau, Alexis, Kartikay Khandelwal, Naman Goyal, Vishrav Chaudhary, Guillaume Wenzek, Francisco Guzmán, Edouard Grave, Myle Ott, Luke Zettlemoyer, and Veselin Stoyanov. 2020. Unsupervised cross-lingual representation learning at scale. Paper presented at 58th Annual Meeting of the Association for Computational Linguistics, Online. July 5–10; Stroudsburg: Association for Computational Linguistics, pp. 8440–51. [ Google Scholar ] [ CrossRef ]
• Council of Europe. 2001. European Framework of Reference for Languages: Learning, Teaching, Assessment . Cambridge: Cambridge University Press. [ Google Scholar ]
• Council of Europe. 2020. Common European Framework of Reference for Languages: Learning, Teaching, Assessment. Companion Volume with New Descriptors . London: Council of Europe. [ Google Scholar ]
• Creese, Angela, and Adrian Blackledge. 2010. Translanguaging in the bilingual classroom: A pedagogy for learning and teaching? The Modern Language Journal 94: 103–15. [ Google Scholar ] [ CrossRef ]
• Daniel, Shannon M., Robert T. Jiménez, Lisa Pray, and Mark B. Pacheco. 2019. Scaffolding to make translanguaging a classroom norm. TESOL Journal 10: e00361. [ Google Scholar ] [ CrossRef ]
• De Felice, Rachele, and Stephen G. Pulman. 2008. A classifier-based approach to preposition and determiner error correction in L2 English. Paper presented at 22nd International Conference on Computational Linguistics (Coling 2008), Manchester, UK, August 18–22; Manchester: Coling 2008 Organizing Committee, pp. 169–76. [ Google Scholar ]
• Department of Education, Language Policy, Culture of the Basque Government, and The Sociolinguistics Cluster. 2013. Talking Pupils: The Arrue Proyect 2011: Research, Results and Contributions of Experts ; Vitoria-Gasteiz: Basque Government.
• Devlin, Jacob, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. BERT: Pre-training of deep bidirectional transformers for language understanding. Paper presented at 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers), Minneapolis, MN, USA, June 2–7; Stroudsburg: Association for Computational Linguistics, pp. 4171–86. [ Google Scholar ] [ CrossRef ]
• Doğruöz, A. Seza, Sunayana Sitaram, Barbara E. Bullock, and Almeida Jacqueline Toribio. 2021. A survey of code-switching: Linguistic and social perspectives for language technologies. Paper presented at 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), Online. August 1–6; Stroudsburg: Association for Computational Linguistics, pp. 1654–66. [ Google Scholar ] [ CrossRef ]
• Duff, Patricia A., and Heidi Byrnes. 2019. Sla across disciplinary borders: Introduction to the special issue. The Modern Language Journal 103: 3–5. [ Google Scholar ] [ CrossRef ]
• European Commission. 2019. Proposal for a Council Recommendation on a Comprehensive Approach to the Teaching and Learning of Languages . Brussels: European Commission. [ Google Scholar ]
• Faltis, Christian. 2019. Pedagogical Codeswitching and Translanguaging in Bilingual Schooling Contexts . New York: Routledge, pp. 39–62. [ Google Scholar ] [ CrossRef ]
• Faltis, Christian, and Valdés. 2016. Preparing teachers for teaching in and advocating for linguistically diverse classrooms: A vade mecum for teacher educators. In The Handbook of Research on Teaching , 5th ed. Washington, DC: American Educational Research Association, pp. 549–92. [ Google Scholar ]
• Fareed, Muhammad, Samreen Humayun, and Huma Akhtar. 2016. English language teachers’ code-switching in class: Esl learners’ perceptions. Journal of Education & Social Sciences 4: 3–13. [ Google Scholar ] [ CrossRef ]
• Felice, Mariano, and Zheng Yuan. 2014. Generating artificial errors for grammatical error correction. Paper presented at Student Research Workshop at the 14th Conference of the European Chapter of the Association for Computational Linguistics, Gothenburg, Sweden, April 26–30; Stroudsburg: Association for Computational Linguistics, pp. 116–26. [ Google Scholar ] [ CrossRef ]
• Felice, Mariano, Zheng Yuan, Øistein E. Andersen, Helen Yannakoudakis, and Ekaterina Kochmar. 2014. Grammatical error correction using hybrid systems and type filtering. Paper presented at Eighteenth Conference on Computational Natural Language Learning: Shared Task, Baltimore, MD, USA, June 26–27; Stroudsburg: Association for Computational Linguistics, pp. 15–24. [ Google Scholar ] [ CrossRef ]
• Félix-Brasdefer, J. César, and Rachel L. Shively, eds. 2021. New Directions in Second Language Pragmatics . Berlin: De Gruyter Mouton. [ Google Scholar ] [ CrossRef ]
• Gamon, Michael. 2011. High-order sequence modeling for language learner error detection. Paper presented at Sixth Workshop on Innovative Use of NLP for Building Educational Applications, Portland, OR, USA, June 24; Stroudsburg: Association for Computational Linguistics, pp. 180–89. [ Google Scholar ]
• García, Ofelia. 2011. Bilingual Education in the 21st Century: A Global Perspective . West Sussex: Wiley-Blackwell. [ Google Scholar ]
• García, Ofelia, and Tatyana Kleyn. 2016. Translanguaging Multilingual Students: Learning from Classroom Moments . New York: Routledge. [ Google Scholar ]
• García, Ofelia, Susana Ibarra Johnson, and Kate Seltzer. 2017. The Translanguaging Classroom: Leveraging Student Bilingualism for Learning . Philadelphia: Carlson. [ Google Scholar ]
• Gardner-Chloros, Penelope. 2009. Code-Switching . Cambridge: Cambridge University Press. [ Google Scholar ]
• Gardner-Chloros, Penelope. 2020. Contact and Code-Switching . West Sussex: John Wiley & Sons, Ltd., chp. 9. pp. 181–99. [ Google Scholar ] [ CrossRef ]
• Gretter, Roberto, Marco Matassoni, Stefano Bannò, and Falavigna Daniele. 2020. TLT-school: A corpus of non native children speech. Paper presented at 12th Language Resources and Evaluation Conference, Marseille, France, May 11–16; Paris: European Language Resources Association, pp. 378–85. [ Google Scholar ]
• Gumperz, John J. 1982. Discourse Strategies. Studies in Interactional Sociolinguistics . Cambridge: Cambridge University Press. [ Google Scholar ] [ CrossRef ]
• Han, Na-Rae, Martin Chodorow, and Claudia Leacock. 2004. Detecting errors in English article usage with a maximum entropy classifier trained on a large, diverse corpus. Paper presented at Fourth International Conference on Language Resources and Evaluation (LREC’04), Lisbon, Portugal, May 26–28; Paris: European Language Resources Association (ELRA). [ Google Scholar ]
• Ho, Wing Yee Jenifer. 2018. Translanguaging in Online Language Learning: Case Studies of Self-Directed Chinese Learning of Multilingual Adults. Ph.D thesis, University College London, London, UK. [ Google Scholar ]
• Ho, Wing Yee Jenifer, and Kevin W. H. Tai. 2021. Translanguaging in digital learning: The making of translanguaging spaces in online english teaching videos. International Journal of Bilingual Education and Bilingualism , 1–22. [ Google Scholar ] [ CrossRef ]
• Johansson, Sara. 2014. Code-Switching in the English Classroom: What Teachers Do and What Their Students Wish They Did. Available online: http://www.diva-portal.org/smash/record.jsf?pid=diva2%3A691914&dswid=5065 (accessed on 30 April 2022).
• Kaneko, Masahiro, and Mamoru Komachi. 2019. Multi-Head Multi-Layer Attention to Deep Language Representations for Grammatical Error Detection. Computing Research Repository 23: 883–91. [ Google Scholar ] [ CrossRef ]
• Khanuja, Simran, Sandipan Dandapat, Anirudh Srinivasan, Sunayana Sitaram, and Monojit Choudhury. 2020. GLUECoS: An evaluation benchmark for code-switched NLP. Paper presented at 58th Annual Meeting of the Association for Computational Linguistics, Online. July 5–10; Stroudsburg: Association for Computational Linguistics, pp. 3575–85. [ Google Scholar ] [ CrossRef ]
• Kim, Jeongyeon, and Bradley Tatar. 2017. Nonnative english-speaking professors’ experiences of english-medium instruction and their perceived roles of the local language. Journal of Language, Identity & Education 16: 157–71. [ Google Scholar ] [ CrossRef ]
• Lapierre, Cynthia. 2018. The Role of First Language Influence in the Learning of Second Language Grammar: The Case of His/Her in English. Master’s thesis, Concordia University, Montreal, QC, Canada. unpublished. [ Google Scholar ]
• Le Pichon-Vorstman, Emmanuelle, Hanna Siarova, and Eszter Szönyi. 2020. The Future of Language Education in Europe: Case Studies of Innovative Practices: Analytical Report . Luxembourg: European Commission and Directorate-General for Education, Youth, Sport and Culture, Publications Office. [ Google Scholar ] [ CrossRef ]
• Lichtarge, Jared, Chris Alberti, and Shankar Kumar. 2020. Data weighted training strategies for grammatical error correction. Transactions of the Association for Computational Linguistics 8: 634–46. [ Google Scholar ] [ CrossRef ]
• Lin, Angel. 2013. Classroom code-switching: Three decades of research. Applied Linguistics Review 4: 195–218. [ Google Scholar ] [ CrossRef ]
• Little, David, and Déirdre Kirwan. 2019. Engaging with Linguistic Diversity: A Study of Educational Inclusion in an Irish Primary School . Multilingualisms and Diversities in Education. London: Bloomsbury Publishing. [ Google Scholar ]
• Liu, Jingxia. 2010. Teachers’ code-switching to the l1 in efl classroom. The Open Applied Linguistics Journal 3: 10–23. [ Google Scholar ] [ CrossRef ]
• Liu, Yang, and Fan Fang. 2020. Translanguaging theory and practice: How stakeholders perceive translanguaging as a practical theory of language. RELC Journal , 1–9. [ Google Scholar ] [ CrossRef ]
• Macaro, Ernesto, Lili Tian, and Lingmin Chu. 2020. First and second language use in english medium instruction contexts. Language Teaching Research 24: 382–402. [ Google Scholar ] [ CrossRef ]
• Mahboob, Ahmar, and Angel M. Y. Lin. 2016. Using Local Languages in English Language Classrooms . Cham: Springer International Publishing, pp. 25–40. [ Google Scholar ] [ CrossRef ]
• Maluleke, Mzamani. 2019. Using code-switching as an empowerment strategy in teaching mathematics to learners with limited proficiency in english in south african schools. South African Journal of Education 39: 1–9. [ Google Scholar ] [ CrossRef ]
• Masna, Yuliar. 2020. Efl learners’ code-switching: Why do they switch the language? Englisia: Journal of Language, Education, and Humanities 8: 93–101. [ Google Scholar ] [ CrossRef ]
• Mayfield, Elijah, and Alan W. Black. 2020. Should you fine-tune BERT for automated essay scoring? Paper presented at Fifteenth Workshop on Innovative Use of NLP for Building Educational Applications, Seattle, WA, USA, July 9–10; Stroudsburg: Association for Computational Linguistics, pp. 151–62. [ Google Scholar ] [ CrossRef ]
• Muysken, Pieter. 2000. Bilingual Speech: A Typology of Code-Mixing . Cambridge: Cambridge University Press. [ Google Scholar ]
• Myers-Scotton, Carol. 1993. Duelling Languages: Grammatical Structure in Codeswitching . Oxford: Clarendon. [ Google Scholar ]
• Ngo, Ngoc Bich, and Yen Hoang Phuong. 2018. The frequency and functions of teachers’ use of mother tongue in EFL classrooms. European Journal of English Language Teaching 3. [ Google Scholar ] [ CrossRef ]
• Nguyen, Li. 2018. Borrowing or code-switching? Traces of community norms in Vietnamese-English speech. The Australian Journal of Linguistics 38: 443–66. [ Google Scholar ] [ CrossRef ]
• North, Brian. 2022. Plurilingual Mediation in the Classroom Examples from Practice . London: Routledge, chp. 17. pp. 319–36. [ Google Scholar ]
• Orellana, Marjorie Faulstich, and Ofelia García. 2014. Conversation currents: Language brokering and translanguaging in school. Language Arts 91: 386–92. [ Google Scholar ]
• Otheguy, Ricardo, Ofelia García, and Wallis Reid. 2015. Clarifying translanguaging and deconstructing named languages: A perspective from linguistics. Applied Linguistics Review 6: 281–307. [ Google Scholar ] [ CrossRef ]
• Otto, Ana, and Estrada José Luis. 2019. Towards an understanding of clil assessment practices in a european context: Main assessment tools and the role of language in content subjects. CLIL. Journal of Innovation and Research in Plurilingual and Pluricultural Education 2: 31–42. [ Google Scholar ] [ CrossRef ]
• Pennycook, Alastair. 2010. Language as a Local Practice . London and New York: Routledge. [ Google Scholar ]
• Piccardo, Enrica, Aline Germain-Rutherford, and Geoff Lawrence. 2021. The Routledge Handbook of Plurilingual Language Education , 1st ed. London: Routledge. [ Google Scholar ] [ CrossRef ]
• Piribauer, Gerda, Ute Atzlesberger, Irmgard Greinix, Thomas Ladstatter, Franz Mittendorfer, Helmut Renner, and Belinda Steinhuber. 2015. Designing and Implementing Plurilingual Oral Exams: Framework for the Austrian Upper Secondary Level Oral Leaving Examination at Colleges for Higher Vocational Education . Vienna: Center for Vocational Languages CEBS. [ Google Scholar ]
• Poeste, Meike, and Natascha Müller. 2020. Code-mixing in trilingual children: Domains and dimensions of language dominance. In Contact, Variation and Change in Romance and Beyond . Studies in Honor of Trudel Meisenburg. Berlin: Erich Schmidt Verlag, pp. 301–20. [ Google Scholar ]
• Poeste, Meike, Natascha Müller, and Laia Arnaus Gil. 2019. Code-mixing and language dominance: Bilingual, trilingual and multilingual children compared. International Journal of Multilingualism 16: 459–91. [ Google Scholar ] [ CrossRef ]
• Poplack, Shana. 1980. Sometimes i’ll start a sentence in Spanish y termino en español: Toward a typology of codeswitching. Linguistics 18: 581–618. [ Google Scholar ] [ CrossRef ]
• Poplack, Shana, David Sankoff, and Christopher Miller. 1988. The social correlates and linguistic processes of lexical borrowing and assimilation. Linguistics 26: 47–104. [ Google Scholar ] [ CrossRef ]
• Qasem, Fawaz. 2020. Crosslinguistic influence of the first language: Interlingual errors in the writing of esl saudi learners. Macrolinguistics 8: 105–20. [ Google Scholar ] [ CrossRef ]
• Rei, Marek, and Helen Yannakoudakis. 2016. Compositional sequence labeling models for error detection in learner writing. Paper presented at 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Berlin, Germany, August 7–12; Stroudsburg: Association for Computational Linguistics, pp. 1181–91. [ Google Scholar ] [ CrossRef ]
• Rei, Marek, Mariano Felice, Zheng Yuan, and Ted Briscoe. 2017. Artificial error generation with machine translation and syntactic patterns. Paper presented at 12th Workshop on Innovative Use of NLP for Building Educational Applications, Copenhagen, Denmark, September 8; Stroudsburg: Association for Computational Linguistics, pp. 287–92. [ Google Scholar ] [ CrossRef ]
• Rozovskaya, Alla, and Dan Roth. 2011. Algorithm selection and model adaptation for ESL correction tasks. Paper presented at 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies, Portland, OR, USA, June 19–24; Stroudsburg: Association for Computational Linguistics, pp. 924–33. [ Google Scholar ]
• Sah, Pramod K., and Guofang Li. 2020. Translanguaging or unequal languaging? unfolding the plurilingual discourse of english medium instruction policy in nepal’s public schools. International Journal of Bilingual Education and Bilingualism , 1–20. [ Google Scholar ] [ CrossRef ]
• Saville, Nick, and Graham Seed. 2022. Language assessment in the context of plurilingualism. In The Routledge Handbook of Plurilingual Education . Edited by Geoff Lawrence, Enrica Piccardo and Aline Germain-Rutherford. London: Routledge, chp. 19. pp. 360–76. [ Google Scholar ]
• Seed, Graham. 2020. What Is Plurilingualism and What Does It Mean for Language Assessment? Cambridge: Cambridge Assessment English, Research Notes. [ Google Scholar ]
• Seed, Graham, and Martine Holland. 2020. Taking Account of Plurilingualism in Cambridge Assessment English Products and Services . Cambridge: Cambridge Assessment English, Research Notes. [ Google Scholar ]
• Séror, Jérémie. 2022. Plurilingualism in digital spaces. In The Routledge Handbook of Plurilingual Education . Edited by Geoff Lawrence, Enrica Piccardo and Aline Germain-Rutherford. London: Routledge, chp. 22. pp. 449–64. [ Google Scholar ]
• Shinga, Sibongile, and Ansurie Pillay. 2021. Why do teachers code-switch when teaching english as a second language? South African Journal of Education 41: 3–13. [ Google Scholar ] [ CrossRef ]
• Sun, Yachao. 2022. Implementation of translingual pedagogies in eal writing: A systematic review. Language Teaching Research , 1–25. [ Google Scholar ] [ CrossRef ]
• Taghipour, Kaveh, and Hwee Tou Ng. 2016. A neural approach to automated essay scoring. Paper presented at 2016 Conference on Empirical Methods in Natural Language Processing, Austin, TX, USA, November 1–5; Stroudsburg: Association for Computational Linguistics, pp. 1882–91. [ Google Scholar ] [ CrossRef ]
• Tetreault, Joel R., and Martin Chodorow. 2008. The ups and downs of preposition error detection in ESL writing. Paper presented at 22nd International Conference on Computational Linguistics (Coling 2008), Manchester, UK, August 18–22; Manchester: Coling 2008 Organizing Committee, pp. 865–72. [ Google Scholar ]
• Torres Cacoullos, Rena, and Catherine E. Travis. 2018. Bilingualism in the Community: Code-Switching and Grammars in Contact . Cambridge: Cambridge University Press. [ Google Scholar ] [ CrossRef ]
• Treffers-Daller, Jeanine. 2005. Evidence for insertional codemixing: Mixed compounds and French nominal groups in Brussels Dutch. International Journal of Bilingualism 9: 477–508. [ Google Scholar ] [ CrossRef ]
• Treffers-Daller, Jeanine. 2009. Code-Switching and Transfer: An Exploration of Similarities and Differences . Cambridge Handbooks in Language and Linguistics. Cambridge: Cambridge University Press, pp. 58–74. [ Google Scholar ] [ CrossRef ]
• Tsimpli, Ianthi Maria, Margreet Vogelzang, Anusha Balasubramanian, Theodoros Marinis, Suvarna Alladi, Abhigna Reddy, and Minati Panda. 2020. Linguistic diversity, multilingualism, and cognitive skills: A study of disadvantaged children in india. Languages 5: 10. [ Google Scholar ] [ CrossRef ]
• Vaswani, Ashish, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. In Advances in Neural Information Processing Systems 30 . Edited by Isabelle Guyon, Ulrike von Luxburg, Samy Bengio, Hanna Wallach, Rob Fergus, S. V. N. Vishwanathan and Roman Garnett. Red Hook: Curran Associates, Inc., pp. 5998–6008. [ Google Scholar ]
• Walker, Ute. 2021. From target language to translingual capabilities. harnessing plurilingual repertoires for language learning and teaching. Language Education and Multilingualism—The Langscape Journal , 117–34. [ Google Scholar ] [ CrossRef ]
• Wang, Danping. 2019. Multilingualism and Translanguaging in Chinese Language Classrooms . Basingstoke: Palgrave Macmillan. [ Google Scholar ]
• Wilson, Joshua, Dandan Chen, Micheal P. Sandbank, and Michael Hebert. 2019. Generalizability of automated scores of writing quality in grades 3–5. Journal of Educational Psychology 111: 619–40. [ Google Scholar ] [ CrossRef ]
• Wilson, Sonia. 2021. To mix or not to mix: Parental attitudes towards translanguaging and language management choices. International Journal of Bilingualism 25: 58–76. [ Google Scholar ] [ CrossRef ]
• Winford, Donald. 2003. An Introduction to Contact Linguistics . Malden and Oxford: Blackwell. [ Google Scholar ]
• Winford, Donald. 2009. On the unity of contact phenomena and their underlying mechanisms. In Multidisciplinary Approaches to Code-Switching . Edited by Ludmila Isurin, Donald Winford and Kees de Bot. Philadelphia: John Benjamins Publishing Company, pp. 279–306. [ Google Scholar ]
• Yamashita, Junko, and Nan Jiang. 2010. L1 influence on the acquisition of l2 collocations: Japanese esl users and efl learners acquiring english collocations. TESOL Quarterly 44: 647–68. [ Google Scholar ] [ CrossRef ]
• Yannakoudakis, Helen, Øistein E. Andersen, Ardeshir Geranpayeh, Ted Briscoe, and Diane Nicholls. 2018. Developing an automated writing placement system for esl learners. Applied Measurement in Education 31: 251–67. [ Google Scholar ] [ CrossRef ]
• Yao, Mingfa. 2011. On attitudes to teachers’ code-switching in efl classes. World Journal of English Language 1: 19. [ Google Scholar ] [ CrossRef ]
• Yuan, Zheng, and Christopher Bryant. 2021. Document-level grammatical error correction. Paper presented at 16th Workshop on Innovative Use of NLP for Building Educational Applications, Online. April 20; Stroudsburg: Association for Computational Linguistics, pp. 75–84. [ Google Scholar ]
• Yuan, Zheng, and Ted Briscoe. 2016. Grammatical error correction using neural machine translation. Paper presented at 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, San Diego, CA, USA, June 12–17; Stroudsburg: Association for Computational Linguistics, pp. 380–86. [ Google Scholar ] [ CrossRef ]
• Yuan, Zheng, Felix Stahlberg, Marek Rei, Bill Byrne, and Helen Yannakoudakis. 2019. Neural and FST-based approaches to grammatical error correction. Paper presented at Fourteenth Workshop on Innovative Use of NLP for Building Educational Applications, Florence, Italy, August 2; Stroudsburg: Association for Computational Linguistics, pp. 228–39. [ Google Scholar ] [ CrossRef ]
• Yuan, Zheng, Shiva Taslimipoor, Christopher Davis, and Christopher Bryant. 2021. Multi-class grammatical error detection for correction: A tale of two systems. Paper presented at 2021 Conference on Empirical Methods in Natural Language Processing, Punta Cana, Dominican Republic, May 10; Stroudsburg: Association for Computational Linguistics, pp. 8722–36. [ Google Scholar ] [ CrossRef ]

Share and Cite

Nguyen, L.; Yuan, Z.; Seed, G. Building Educational Technologies for Code-Switching: Current Practices, Difficulties and Future Directions. Languages 2022 , 7 , 220. https://doi.org/10.3390/languages7030220

Nguyen L, Yuan Z, Seed G. Building Educational Technologies for Code-Switching: Current Practices, Difficulties and Future Directions. Languages . 2022; 7(3):220. https://doi.org/10.3390/languages7030220

Nguyen, Li, Zheng Yuan, and Graham Seed. 2022. "Building Educational Technologies for Code-Switching: Current Practices, Difficulties and Future Directions" Languages 7, no. 3: 220. https://doi.org/10.3390/languages7030220

Article Metrics

Article access statistics, further information, mdpi initiatives, follow mdpi.

Subscribe to receive issue release notifications and newsletters from MDPI journals

Help | Advanced Search

Computer Science > Computation and Language

Title: the decades progress on code-switching research in nlp: a systematic survey on trends and challenges.

Abstract: Code-Switching, a common phenomenon in written text and conversation, has been studied over decades by the natural language processing (NLP) research community. Initially, code-switching is intensively explored by leveraging linguistic theories and, currently, more machine-learning oriented approaches to develop models. We introduce a comprehensive systematic survey on code-switching research in natural language processing to understand the progress of the past decades and conceptualize the challenges and tasks on the code-switching topic. Finally, we summarize the trends and findings and conclude with a discussion for future direction and open questions for further investigation.

Submission history

Access paper:.

• Other Formats

References & Citations

• Google Scholar
• Semantic Scholar

BibTeX formatted citation

Bibliographic and Citation Tools

Code, data and media associated with this article, recommenders and search tools.

• Institution

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs .

EDITORIAL article

Editorial: behavioral and neurophysiological approaches to code-switching and language switching.

• 1 Department of English Language and Applied Linguistics, University of Reading, Reading, United Kingdom
• 2 Department of Dutch, University of Oldenburg, Oldenburg, Germany
• 3 Institute of Education, University College London, London, United Kingdom

Editorial on the Research Topic Behavioral and Neurophysiological Approaches to Code-Switching and Language Switching

One of the unique characteristics of bilinguals is that they can freely switch between languages, both between and within utterances, a phenomenon that is generally described as code-switching (CS). Since the seminal papers of Pfaff (1979) and Poplack (1980) many linguists working on CS have focused on where switching can take place in a sentence and attempted to formulate (universal) linguistic constraints on this behavior. This branch of research into the linguistic characteristics of CS has led to in-depth insights into the variability in CS patterns found in speech communities across the world, to the development of new CS typologies as well as a renewed understanding of the ways in which sociolinguistic factors interact with these typologies ( Poplack, 1988 ; Muysken, 2013 ).

Although the term code-switching is used in both sociolinguistic and experimental studies, in their overview of research techniques used in code-switching research, Gullberg et al. (2009) make a distinction between internally generated CS, for which data are collected using corpus linguistic and sociolinguistic techniques, and language switching (LS), which is externally induced in a laboratory situation, where respondents switch languages, e.g., in response to an external cue. Researchers interested in LS generally aim to arrive at a better understanding of the ways in which switches are processed rather than the end product of this process. In this branch of research, experimental methods are used for which the stimulus materials as well as the situation under which respondents respond to stimuli are carefully controlled. We believe that sociolinguistic and experimental approaches are complementary in that each brings vital evidence to our understanding of the ways in which bilinguals switch between languages and the cognitive processes supporting this behavior. A better understanding of CS could therefore be achieved if researchers drew cross-disciplinary conclusions, integrating insights based on both linguistic studies of naturalistic CS and on experimental studies of LS, as in Pablos et al. (2019) , who test theory-driven linguistic hypotheses on spontaneous data as well as with EEG methodology. We hope that the current Article Collection will help to further this integration, by bringing together interdisciplinary evidence from different research strands in the field.

In recent years, novel psycholinguistic, as well as neuroscientific methods, such as brain imaging and electrophysiological approaches, have allowed researchers to obtain insights into online processing that cannot be obtained using more traditional offline or behavioral methods which rely on the measurement of the end product of processing or measure reaction times (RTs) needed to complete tasks. Methods from psychology and neuroscience have the potential to revolutionize CS research because they provide a more direct insights into the working of bilingual mind than other methods. They make it possible to observe potential relationships between cognitive processes and language use as well as the neurophysiological correlates of these processes much more directly than had been possible so far, which has led to new insights in these fields [see e.g., Christoffels et al. (2007) ].

The development of new models of bilingual speech processing and bilingual visual word recognition ( Green and Abutalebi, 2013 ; Green and Li, 2014 ; Dijkstra et al., 2019 ) also led to a renewed interest in CS, for example among researchers interested in Cognitive Control and Executive Functions. Work in this field of research focuses on the attentional control mechanisms that are needed to enable bilinguals to switch between languages. Some studies on the relationship between CS and attentional mechanisms have found that CS practices modulate performance on inhibitory control tasks ( Hofweber et al., 2016 , 2020 ), while others have failed to reveal a relationship between CS and attentional processes ( Kang and Lust, 2019 ). Further evidence is therefore needed to study the causes of these inconsistencies.

A new line of enquiry focuses on the neurophysiological correlates of CS with the aim of analyzing brain reactions to CS in real time ( Moreno et al., 2002 ; Ruigendijk et al., 2016 ; Zeller et al., 2016 ; Van Hell et al., 2018 ; Pablos et al., 2019 ). These studies have the potential to shed more light on the psychological reality of different types of CS, on the magnitude of the processing cost involved in CS, and on the role of variables that may modulate the processing cost of CS, such as speakers' relative proficiency in the two languages, the direction of the switch (i.e., from L1 into L2 or vice versa), and the typological difference between the languages (processing CS in closely related languages vs. in structurally different languages). Specifically, ERP studies can be used to gain insights into the cognitive processes underlying CS.

In this volume four broad topics are addressed: (1) the relationship between CS or LS and cognitive control; (2) linguistic processing of CS and LS; (3) neural and electrophysiological correlates of switching; and (4) linguistic and orthographic analyses of CS and LS. In the remainder of this Editorial we will present each Part in turn.

The focus of the first Part of this Article Collection is on the relationship between CS or LS and cognitive control. In their study among proficient bilingual adults, Barbu et al. found clear evidence for a positive effect of the frequency of reported LS on cognitive flexibility, but not on alertness or response inhibition. In a similar vein, in a study investigating the Adaptive Control Hypothesis (ACH, Green and Abutalebi, 2013 ), Lai and O'Brien found positive relationships between the frequency of CS and cognitive control performance. Crucially, the Lai and O'Brien study offers partial support for the ACH, but suggests that the three interactional contexts (single, dual, and dense) distinguished by the model should not be seen as a categorical distinction but placed along a continuum. Interestingly from a methodological perspective, the observed effects were stronger when CS was measured using naturalistic conversational data, than when the CS measure was based on self-reports. Hofweber et al. (2016) investigated the effects of experimentally induced language modes and bilinguals' regular CS habits on proactive and reactive control. They also found support for the ACH in that inhibitory performance in the L2-single-language condition was enhanced, possibly because suppressing the L1 requires heightened levels of inhibition. In a highly innovative study taking into account bilinguals' socio-cultural identities, Treffers-Daller et al. explored the relative contribution of informants' CS habits and their multicultural identity styles, that is the strategies individuals use to manage multiple identities, and found that the latter explained most variance in inhibitory control.

For the last two papers in this Part, attention shifts toward the analysis of cognitive control in bilingual children. In the first of these two, Gross and Kaushanskaya tease apart the interaction between cognitive control, language dominance, and language ability. They found an increase in cross-language intrusions among children with lower cognitive control, particularly in the dual-language context, irrespective of children's levels of language ability. The second paper, by Timmermeister et al. focuses on LS and task switching in bilingual children. While the authors found that response times in the LS and nonverbal switching tasks were related, bilingual children did not outperform monolinguals in cognitive control in this study.

In the second Part of the Article Collection, we turn to linguistic processing of CS and LS, for which a range of experimental techniques and behavioral measures are used. In the first contribution, Beatty-Martínez et al. use Green and Abutalebi's (2013) notion of opportunistic planning and suggest that CS can serve as an opportunistic strategy for optimizing task performance, for which they provide evidence on the basis of data from an innovative CS map task. In the next paper, Suurmeijer et al. use another novel technique, namely auditory sentence matching, to study how switch site and switch directionality affect the processing of CS sentences. Contrary to expectations, only effects of the direction of switching but no effects of the switch site were found. The third paper, Kootstra et al. studies the combined effects of interactive alignment (that is alignment between CS behavior of dialogue partners) and lexical triggering ( Clyne, 1980 ) on bilinguals' CS behavior. On the basis of an experimental task which had not yet been used to study these phenomena, they show that lexical triggering is driven by interactive alignment. In the final paper in this Part, Zhang et al. focus on the differences between the cognitive processes underlying language switches and concept switches using a bilingual picture naming task. They found that trials, which involved semantically unrelated items as well as switching between languages led to the longest naming RTs.

In Part three, the focus is on the neural and electrophysiological correlates of switching. These four studies all follow-up on the already mentioned earlier ERP studies that examined the processing of CS ( Moreno et al., 2002 ; Ruigendijk et al., 2016 ) by zooming in on some relevant factors. Valdés Kroff et al. asked whether semantic and language unexpectancy result in similar processing effects. Their ERP results clearly differ for the effect of semantically unexpected vs. highly expected words, and for CS in Spanish to English switches, with a classical N400 effect for the semantic manipulation and a late positive component (LPC) for the CS, in line with earlier studies. Additionally, these data were related to self-reported experience with CS, which suggested that certain effects are linked to having less experience with CS.

Zeller compared the effect of switching at different positions in a sentence, a preposition or a noun, in German–Russian listeners. He found clear differences between the positions on the relevant ERP components, indicating that the underlying psycholinguistic processes for these two types of CS are indeed not the same. Vaughan-Evans et al. studied adjective-noun order in Welsh–English nominal constructions. They tested predictions of the Matrix Language Framework (MLF, Myers Scotton, 1993 ) and the Minimalist Program (MP, Cantone and MacSwan, 2009 ). The ERP data showed different patterns for MLF vs. MP violations. Furthermore, the data suggested that noun insertion is preferred over adjective insertion supporting MLF. Interestingly, the ERP was also modulated by the Matrix Language: when the ML was Welsh, effects were found that were absent when the ML was English. These two studies thus contribute to our theoretical understanding of the rules that governing intra-sentential CS and they do so by examining language combinations that have not received much attention in neurolinguistic approaches to CS so far. The final paper in this Part took a slightly different approach by examining the role of the social situation in which CS takes place by comparing processing in Spanish–English bilinguals in the presence of another bilingual or in the presence of a monolingual speaker of English. Kaan et al. found that relevant ERP effects were smaller in the presence of a bilingual. This indicates that listeners activate their languages in a bilingual social situation and thus CS lead to less processing cost. These results are important for our understanding of language control (see Green and Li, 2014 ).

The final Part of the Article Collection consists of two papers with in-depth linguistic analyses of CS and two papers which focus on the effects of language-specific letter sequences (i.e., letter sequences that are illegal in one of the two languages) on word recognition. The linguistic analyses start with a paper by Alexiadou , who offers a detailed study of mixed nominal compounds, showing that one of the two contact languages generally provides the underlying structure, i.e., is the matrix language of the compound. The results from a wide range of language pairs are discussed with a view to informing theory building in word formation. The second paper, by Cacoullos , shows how speakers deploy CS strategies, considering prosodic and syntactic variables at switch points of variable equivalence, as is the case, for example, for switches between main and complement clauses where languages have different requirements regarding the use of complementizers. In the third paper, Duñabeitia et al. investigate to what extent bilinguals from different ages use orthotactic cues to recognize to which language a word belongs, on the basis of an innovative language decision task. They found that bilinguals are very good at detecting orthotactic markedness in their L2 even for pseudowords and that this ability increased with age. While their study focused on languages which share the same alphabet but are orthotactically distinct, Chen and Liu focus on trilinguals who use languages that use different scripts. They found no switch costs in a bilingual lexical decision task, nor did they find evidence for effects of the non-task language on lexical processing. Both papers interpret their results in the light of recent models of bilingual visual word recognition ( Dijkstra and Van Heuven, 2002 ; Dijkstra et al., 2019 ).

The current Article Collection has brought together cutting edge research in the field of CS and LS. The papers illustrate the importance of ensuring experimental work in the field is informed by insights obtained in more naturalistic circumstances, for example by creating experimental stimuli for psycholinguistic and neuroscientific experiments that are representative for the kinds of switching that are found in the real world in a particular language pair. Conversely, as bilingual corpora are generally small and unlikely to provide the necessary evidence about all switches that are possible in a language pair, experimental methods can help drive forward research into constraints on CS ( Munarriz-Ibarrola et al., 2018 ; Treffers-Daller, 2021 ). As the current volume illustrates, making links between evidence from naturalistic and experimental approaches is not always straightforward, but the combination of insights from different disciplines can lead to the creation of innovative methods, which shed new light on the key problem of how bilinguals manage to keep their languages separate on some occasions while they can switch freely between languages when the situation allows it. We hope the current volume has also contributed to developing models of processing in bilinguals and multilinguals, an endeavor that is urgently needed in the face of the divergent findings in the field.

Author Contributions

All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication.

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Acknowledgments

The Editors would like to thank all authors and reviewers of the 18 papers in this volume for submitting their work to this Article Collection. We are also very grateful to Ilaria Prete and all members of staff from the Frontiers team for their support in producing this volume.

Cantone, K. F., and MacSwan, J. (2009). “Adjectives and word order,” in Multidisciplinary Approaches to Code Switching , eds. L. Isurin, D. Winford, and K. de Bot (Amsterdam: John Benjamins), 243–278.

Google Scholar

Christoffels, I. K., Firk, C., and Schiller, N. O. (2007). Bilingual language control: an event-related brain potential study. Brain Res. 1147, 192–208. doi: 10.1016/j.brainres.2007.01.137

PubMed Abstract | CrossRef Full Text | Google Scholar

Clyne, M. G. (1980). Triggering and language processing. Can. J. Psychol. 34, 400–406. doi: 10.1037/h0081102

CrossRef Full Text | Google Scholar

Dijkstra, A., and Van Heuven, W. J. (2002). The architecture of the bilingual word recognition system: from identification to decision. Bilingual. Lang. Cogn. 5, 175–197. doi: 10.1017/S1366728902003012

Dijkstra, A., Wahl, A., Buytenhuijs, F., Van Halem, N., Al-Jibouri, Z., De Korte, M., et al. (2019). Multilink: a computational model for bilingual word recognition and word translation. Bilingual. Lang. Cogn. 22, 657–679. doi: 10.1017/S1366728918000287

Green, D. W., and Abutalebi, J. (2013). Language control in bilinguals: the adaptive control hypothesis. J. Cogn. Psychol. 25, 515–530. doi: 10.1080/20445911.2013.796377

Green, D. W., and Li, W. (2014). A control process model of code-switching. Lang. Cogn. Neurosci. 29, 499–511. doi: 10.1080/23273798.2014.882515

Gullberg, M., Indefrey, P., and Muysken, P. (2009). “Research techniques for the study of code-switching,” in The Cambridge Handbook of Linguistic Code-Switching , eds B. Bullock and A. J. Toribio (Cambridge: Cambridge University Press), 1–17.

Hofweber, J., Marinis, T., and Treffers-Daller, J. (2016). Effects of dense code-switching on executive control. Linguist Approaches Bilingual. 6, 648–668. doi: 10.1075/lab.15052.hof

Hofweber, J., Marinis, T., and Treffers-Daller, J. (2020). How different types of code-switching modulate bilinguals' executive functions - a dual control mode perspective. Bilingual. Lang. Cogn. 23, 909–925. doi: 10.1017/S1366728919000804

Kang, C., and Lust, B. (2019). Code-switching does not predict Executive Function performance in proficient bilingual children: bilingualism does. Bilingual. Lang. Cogn. 22, 366–382. doi: 10.1017/S1366728918000299

Moreno, E. M., Federmeier, K. D., and Kutas, M. (2002). Switching languages, switching palabras (words): an electrophysiological study of code switching. Brain Lang. 80, 188–207. doi: 10.1006/brln.2001.2588

Munarriz-Ibarrola, A., Parafita Couto, M. d,. C. P, and Wyngaerd, E. V. (2018). Methodologies for intra-sentential code-switching research. Linguist. Approaches Bilingual. 8, 1–4. doi: 10.1075/lab.17082.mun

Muysken, P. (2013). Language contact outcomes as the result of bilingual optimization strategies. Bilingual. Lang. Cogn. 16, 709–730. doi: 10.1017/S1366728912000727

Myers Scotton, C. (1993). Duelling Languages: Grammatical Structure in Code Switching. Oxford: Clarendon Press.

Pablos, L., Couto, M. C. P., Boutonnet, B., de Jong, A., Perquin, M., de Haan, A., et al. (2019). Adjective-noun order in Papiamento-Dutch code-switching. Linguist. Approaches Bilingual. 9, 710–735. doi: 10.1075/lab.17036.pab

Pfaff, C. W. (1979). Constraints on language mixing: intrasentential code-switching and borrowing in Spanish/English. Language 55, 291–318. doi: 10.2307/412586

Poplack, S. (1980). Sometimes I'll start a sentence in Spanish y termino en español. Toward a typology of code-switching. Linguistics 18, 581–618. doi: 10.1515/ling.1980.18.7-8.581

Poplack, S. (1988). “Contrasting patterns of code-switching in two communities,” in Aspects of Multilingualism , eds E. Wande, J. Anward, B. Nordberg, L. Steensland, and M. Thelander (Uppsala: Borgström), 51–77.

Ruigendijk, E., Hentschel, G., and Zeller, J. P. (2016). How L2-learners' brains react to code-switches: an ERP study with Russian learners of German. Second Lang. Res. 32, 197–223. doi: 10.1177/0267658315614614

Treffers-Daller, J. (2021). “Turkish-German code-switching patterns revisited: what naturalistic data can(not) tell us,” in Aspects of Contact , eds. N. Smith, T. Veenstra, and E. Aboh (Amsterdam/Philadelphia: John Benjamins), 237–260.

Van Hell, J. G., Fernandez, C. B., Kootstra, G. J., Litcofsky, K. A., and Ting, C. Y. (2018). Electrophysiological and experimental-behavioral approaches to the study of intra-sentential code-switching. Linguist. Approaches Bilingual . 8, 134–161. doi: 10.1075/lab.16010.van

Zeller, J. P., Hentschel, G., and Ruigendijk, E. (2016). Psycholinguistic aspects of Belarusian-Russian language contact. An ERP study on code-switching between closely related languages. Slavic Lang. Psycholinguist . 554, 257.

Keywords: code-switching, language switching, cognitive control, executive functions, task switching, event related potentials

Citation: Treffers-Daller J, Ruigendijk E and Hofweber J (2021) Editorial: Behavioral and Neurophysiological Approaches to Code-Switching and Language Switching. Front. Psychol. 12:660695. doi: 10.3389/fpsyg.2021.660695

Received: 29 January 2021; Accepted: 09 February 2021; Published: 09 March 2021.

Edited and reviewed by: Niels O. Schiller , Leiden University, Netherlands

Copyright © 2021 Treffers-Daller, Ruigendijk and Hofweber. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Jeanine Treffers-Daller, j.c.treffers-daller@reading.ac.uk ; Esther Ruigendijk, esther.ruigendijk@uni-oldenburg.de ; Julia Hofweber, j.hofweber@ucl.ac.uk

Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.

College & Research Libraries ( C&RL ) is the official, bi-monthly, online-only scholarly research journal of the Association of College & Research Libraries, a division of the American Library Association.

C&RL is now on Instragram! Follow us today.

C&RL News

ALA JobLIST

Advertising Information

• Research is an Activity and a Subject of Study: A Proposed Metaconcept and Its Practical Application (79936 views)
• Information Code-Switching: A Study of Language Preferences in Academic Libraries (40964 views)
• Three Perspectives on Information Literacy in Academia: Talking to Librarians, Faculty, and Students (29782 views)

Information Code-Switching: A Study of Language Preferences in Academic Libraries

Frans Albarillo *

Initially coined by sociolinguists, the term code-switching refers to the alternation of languages by multilinguals. Code-switching is an active research area that has significant implications for academic libraries. Using data from focus groups and a survey tool, this paper examines language preferences of foreign-born students for particular information tasks. The main finding is that students’ culture and language represent an active influence on and important part of their identity, information consumption, and academic socialization. The author discusses the practical implications of these findings on academic library services in relation to ACRL’s 2012 Diversity Standards Cultural Competency with an emphasis on standard 6, linguistic diversity.

Introduction

This paper is about foreign-born students’ language preferences and information use in academic libraries. Based on a quantitative and qualitative data set collected by the author, this study examines a concept called code-switching, a linguistic phenomenon where speakers change between two or more languages or between varieties of a language within a speech act or discourse. 1 The author was concerned with how individuals switch languages for different information tasks. Since information is coded in language, the author will refer to the phenomenon as simply information code-switching, referring to switching languages for a particular information task. The author coined the term “information code-switching” to distinguish it as a kind of code-switching that refers specifically to the preferences of a person when it comes to choosing a language or dialect for a particular kind of information behavior.

The basic research question the author investigates is whether there are types of information activities and places where multilingual students code-switch. The answer is yes. The data analyzed here, from focus groups and a survey, show how code-switching works in relation to topics of concern to academic libraries. Librarians are interested in providing access to resources, creating a friendly and welcoming environment, and accommodating the behaviors and preferences of the library’s users. This paper provides qualitative and quantitative evidence that foreign-born students who use academic libraries do switch languages when communicating and use different languages for different kinds of information tasks.

This triangulated approach shows a nuanced picture of the language preferences and information use of foreign-born students as a student population. For students with another language besides English, these data show that language and information use and seeking are intertwined. Even though English is widely considered to be the language of the academy, non-English languages are spoken in academic libraries all the time and used for a variety of information seeking and media consumption. Code-switching has many implications for library service, especially relating to professional standard 6, Linguistic Diversity, of ACRL’s 2012 Diversity Standards Cultural Competency for Academic Libraries, as will be examined in the discussion section.

Literature Review

This literature review is a selective review of the vast literature on code-switching (CS), covering some key studies and works in CS. In the literature, code-switching can also be referred to as “codeswitching” or “code switching.” This author chooses to use the hyphenated version, as it appears in a seminal article by Jan-Peter Blom and John Gumperz describing code-switching in Norway between two language varieties, Ranamål and Bokmål. In their study, Blom and Gumperz look for “social meaning” in how people use their “linguistic repertoire.” 2 Ranamål is a “prestige” dialect and Bokmål is a standard language taught in schools in northern Norway. 3 Blom and Gumperz describe CS: “In their everyday interactions, they [people in the village of Hemnesberget] select among the two as the situation demands. Members view this alternation as a shift between two distinct entities, which are never mixed. A person speaks one or the other.” 4 Carol Myers-Scotton, a linguist writing in the 1990s about research in code-switching, summarizes Gumperz’s characterization of CS as the use of language codes as a type of “social strategy” and extends it to include not only changing languages for content, but speakers using it to frame discourse to show social meanings like solidarity or power relations. 5 Similarly, in the data for this paper, the author will discuss how languages are used strategically for different kinds of information including normal situational and demographic contexts like language choice at home as an expression of cultural identity.

Another concept that is important to this discussion is the notion of a language domain, or the context in which a certain language is chosen. Joshua Fishman traces the history of this concept in his article on the topic, and defines domains as created by “institutional contexts” and their “congruent behavioral co-occurrences.” 6 Fishman further elaborates that domains can have different settings, have a “sociopsychological” component, and, most important, the “domain is a sociocultural construct abstracted from topics of communications, relationships between communicators, and locales of communication, in accord with institutions of a society and the spheres of activity of a speech community, in such a way that individual behavior and social patterns can be distinguished from each other and yet be related to each other.” 7 Fishman’s domain model of language use is the construct used in this study. While Fishman did not initially consider libraries as a separate domain from school, in this study the author looks at information code-switching as an extension of the domain model to language use in libraries. An extended analysis of Fishman’s concept of linguistic domain is given by L.B. Breitborde: “A domain is not the actual interaction (the setting), but an abstract set of relationships between status, topic, and locale which gives meaning to the events that actually comprise social interaction.” 8 In the case of academic libraries in the United States, the status of non-English languages and their presence in library collections, signage, and user experiences are rarely discussed, and, in most cases, languages other than English are invisible.

Fishman includes several domains in his language census of 431 individuals in a Puerto Rican neighborhood in Jersey City, such as “home, church, school, work place, and beach.” 9 In his language census, Fishman asks his respondents about their language abilities in English and Spanish, including “reading, writing, understanding, and speaking, writing letters, language of instruction, conversational languages, etc.” 10 In this paper, the language questions were designed in a similar manner using Fishman’s work and the language questions in the Children of Immigrants Longitudinal Study (CILS). 11 Language domains allow bilinguals to use their linguistic repertoires for a variety of social functions, according to Breitborde: “The use of their linguistic repertoire by bilingual speakers has been linked to situation, setting, social relationships, identity, and topic.” 12 The relationships, identity, and topics in this paper involve the academic library and the relationships foreign-born students have to information topics, tasks, and people as mediated through language.

Another oft-cited code-switching study is by Shana Poplack, who examines Puerto Rican Spanish and English in New York City. 13 Poplack provides examples of “intra-sentential” CS, where English and Spanish are mixed in the same sentence. For example, “Me iban a layoff” is translated as “They were going to lay me off” 14 and includes both English and Spanish in the same sentence. Poplack also notes the influence of Spanish phonology on the perfectly formed grammatical English when “That’s what he said” is pronounced “[da ‘wari se].” 15 And Poplack reports that most of the CS discourses were grammatical: “Perhaps the most striking result of this study is that there were virtually no instances of ungrammatical combinations of L1 and L2 [first language and second language] in the 1,835 switches studied, regardless of the bilingual ability of the speaker.” 16 Another important finding by Poplack is the positive attitude of the speakers who code-switched more toward their language. 17 This landmark research underscores how important CS is for multilingual individuals in defining social relations when speakers have access to more than one language. CS has a situational context, in terms of location, prestige, and institutional support. CS is fluid and natural and is not something that is planned ahead of time. Furthermore, switches have grammatical patterns and should not be seen as “broken” utterances. A significant theme in CS research aims at dispelling these myths about ill-formed languages and dialects. The languages in an individual’s linguistic repertoire will be important to the individual’s or community’s identity, and it is important to think about how libraries recognize or do not recognize languages through their signage, public services, collection development, and implementation of technology.

CS as a research topic today has crossed from linguistics into fields including education, English composition studies, and cultural studies. A search using “codeswitching OR code switching OR code-switching” in the Scopus Social Science and Humanities Index (limiting the search to articles) shows that there were 67 articles published about code-switching in 2016. For a fascinating look at CS in hip-hop the author recommends the book Global Linguistic Flows: Hip Hop Cultures, Youth Identities, and the Politics of Language by Samy Alim, Awad Ibrahim, and Alastair Pennycook. Sociolinguist Jannis Androutsopoulos’s chapter “Language and the Three Spheres of Hip-Hop” has a fascinating discussion on how English is used in code-switching and code-mixing to exemplify style, social identity, and glocality [global and local practices] when English is code-switched or code-mixed with a non-English language. 18 While hip hop is one artistic genre where language and code-switching play a major role in global media communities, the localization of English into specific local varieties is a rich and thriving academic field of study called World Englishes. This field of study recognizes that “The unprecedented spread of English has not [led] to a uniform global language; English is indigenizing into new vernaculars and specializing into national and international varieties of the lingua franca. As Mufwene puts it, ‘rather than driving the world towards monolingualism, differential evolution of English appears to be substituting a new form [of] diversity for an older one’ (2013:50).’” 19 Although this paper does not focus on World Englishes or code-switching between varieties of a single language, an instance of this can be seen in the qualitative data when students dialect-shift to British Standard English—their preferred variety of English when looking for news media.

Language teaching is another field that actively researches code-switching. A recent example is the work of Marta Fairclough and Flavia Belpoliti, who examine code-switches in the writing of English/Spanish heritage language learners to look at the transfer of vocabulary from English to Spanish. 20 A heritage language learner (HLL) is “an individual who is raised in a home where a non-English language is spoken.” 21 Fairclough and Belpoliti investigate how code-switching between English and Spanish improves HLLs’ Spanish literacy by analyzing essays written in response to a prompt measuring a learner’s vocabulary in the language they are trying to learn. 22 As an occurrence closely associated with globalization and transnationalism, CS is a popular phenomenon to study as a topic that involves multilingualism, language instruction, and education. The topic of CS is very relevant today, as other disciplines incorporate this behavior into their research areas. The current study is not a linguistic study or education study; rather, it looks at how an understanding of CS can be applied to academic libraries as an important site to study the intersections of culture, language, information, and immigration.

The same search described above for “codeswitching OR code switching OR code-switching,” when done in Library, Information Science & Technology Abstracts and Library & Information Science Source, returned only two results. The first article, by Magdalena Malechová, discusses code-switching as an intercultural communication trend and a contact linguistic phenomenon. 23 Contact linguistics is a subdiscipline of linguistics that has an active history of research in CS. For data, Malechová’s article counts occurrences of “grammatical code-switching” between English and German in two online German newspapers, concluding that CS is a “strong communication trend.” 24 The second result is an article by Bettina Kümmerling-Meibauer, who analyzes the “visual codes” 25 in Korean and Iranian bilingual picture books for children. She concludes that, in addition to having two separate languages, multilingual picture books have “an elaborate visual code, that are both universal and cultural,” which needs to be accounted for in how children are taught to read. 26 Malechová’s work and Kümmerling-Meibauer’s study show how CS is being extended to analyze media and visual codes, respectively. At the time of writing, the author could not find articles in any library and information science journals on code-switching. This study aims to fill that gap by providing examples of CS by foreign-born students in an academic library context. This study also aims to persuade librarians that information code-switching, or the use of different languages for different information tasks, is a research area that librarians can pursue to understand the role that language plays in the information behavior of multilingual individuals. For librarians working with multilingual populations, it is important to be aware of CS and how populations use their languages to consume information and media in academic libraries. While there are no articles specifically on code-switching in library and information science journals, there are a range of works that look at language issues in libraries.

Language in Libraries

Research about language in libraries exists in the literature and can be found as a topic associated with populations like international students 27 and immigrants. 28 Considering the complexity of the relationship between language and information, very few studies exist that look directly at language preferences, language attitudes, bilingual outreach, and linguistic diversity. At the time of writing, a study by Ignacio Ferrer-Vinent on language preferences at the reference desk 29 is the only study that directly investigates language preferences in academic libraries with reference interactions. This is a rich area that needs further exploring. In relation to library instruction, work by scholars like Karen Bordonaro 30 on students of English as a second language is an important example of how language learning happens in the academic information-seeking context. Sara Luly and Holger Lenz apply the model for Language Oriented Library Instruction (LOLI) 31 to learners of German as a foreign language, who have varying pedagogical needs compared to international students and immigrant students. A critical gap that no one has looked at is how librarians need to vary their instruction according to the different library populations: international students, immigrant students, generation 1.5 students, and foreign language learners. The author has written a literature review about these various different kinds of English language learners 32 and how important it is to distinguish between these populations. In addition to this variation across the kinds of English language learners, it is also important that librarians learn about the varieties of Englishes that students can speak, with attention to World Englishes, as there may be more than one kind of English used in the classroom and the library. Sonia Smith’s article “Library Instruction for Romanized Hebrew” discusses her experiences at McGill University in Canada creating a library instruction session to help students in an advanced Hebrew class navigate the romanized Hebrew catalog records. 33 Smith emphasizes how important the role of library instruction 34 is to scholars who wish to access scholarly materials that are romanized in the library catalog. There is a lively discussion on romanization, language, and access of titles in non-Latin scripts in the journal Cataloging & Classification Quarterly ; these languages include Persian, 35 Korean, 36 and Japanese. 37 The author has also discussed issues of romanization and transliteration as barriers to accessing content in library databases in his article “Evaluating Language Functionality in Library Databases.” 38 For outreach to international students, academic librarians Xiang Li, Kevin McDowell, and Xiaotong Wang write about their experiences creating videos about the library in Arabic, Chinese, English, Japanese, and Korean to help international students “navigate new systems and to bridge the gap between past library experiences and US academic library settings.” 39

Language is an important aspect of cultural competency. In a survey by Misa Mi and Yingting Zhang, two health sciences librarians exploring their perceptions of culturally competent library services, they found that “…those who spoke another language in addition to English rated their own levels of cultural competency higher than those who only spoke English. Those with the ability to speak another language might have an advantage of better understanding a given culture, which could lead to higher levels of cultural awareness and sensitivity.” 40 Mi and Zhang go on to argue that “Culturally competent librarians should regard the ability to speak a second language as an asset that demonstrates greater cognitive ability [33], rather than a deficiency [5]. It would be worthwhile for librarians to develop awareness and knowledge of language differences (which does not require an ability to speak that language) that are reflected in verbal and nonverbal communication processes and norms for effective cross-cultural interactions with and service provision for users from different backgrounds.” 41 The author agrees with Mi and Zhang that, in general, second languages should be viewed as a form of cultural capital, and a cognitive advantage, and that multilingual individuals are more sensitive to cultural differences that they negotiate daily in their lives as speakers of minority languages. It is important to be conscious of essentializing culture into language. While language is an important facet of culture, there are other dimensions of culture like race, ethnicity, and religion that are also meaningful cultural factors. However, the focus of this article is language, and because language is a part of cultural competency, the author recommends that managers should hire multilingual individuals, or at least include multilingualism as a preferred qualification in job advertisements. Managers should also provide training for monolingual staff who work with multilingual populations.

Linguistic diversity training is often ignored in organizational contexts. By linguistic diversity training, the author means education on accents, creoles, code-switching, identifying relevant varieties of English (World Englishes) and other linguistic facets that characterize the patron population. Additionally, it’s important for librarians to understand that language groups like Russian, French, Spanish, or Mandarin are often lingua francas for other linguistic minorities. Ideally, multilingual colleagues could talk about their language use with their monolingual colleagues to increase their awareness of linguistic diversity. For example, the Spanish language is incredibly diverse; and, in areas where it is spoken, cross-cultural trainings could be led by the individuals in the library who speak it so that people who are not aware of these differences can be mindful and at the very least know that differences exist in spoken Spanish that could indicate race, ethnicity, identity, religion, and other potential categories that a speaker may self-identify with. Knowing about linguistic diversity and linguistic behaviors like code-switching is important in the goal of becoming a culturally competent organization. Medical librarians have created a roadmap for hospitals 42 to move their organizations closer, and the author believes that it is important for academic libraries to do the same by studying the demographic effects of factors like language, race, ethnicity, income, age, gender, and able-bodiedness, and how these factors affect or influence library use.

Methodology

With approval from the City University of New York Institutional Review Board and using a small grant awarded by the PSC-CUNY Research Award Program, the author conducted focus group interviews in the spring of 2014 and a survey in the fall of 2014.

Population and Data Collection Procedures

The author used SurveyMonkey, an online survey tool, to create a survey to screen candidates for focus groups. A link to the screening survey was posted on flyers around campus to recruit foreign-born students. The total number of responses for the screening survey was 66, with 33 complete replies and 4 respondents who did not meet the main selection criteria (at least 1 year of high school in their home country and a current student at Brooklyn College at the graduate or undergraduate level). There were a total of 29 qualified respondents and a 45 percent response rate (# interviewed / # eligible to participate). Thirteen students participated in the focus groups. Recruiting for the focus groups was difficult, and the focus groups were very small, ranging from 2 to 4 individuals per group. Participants received $20 each to take part in the focus group. The interviews were semistructured, took 45 minutes to 1 hour to complete, and were based on the following prompts: What languages do you use in your daily life? What does research mean to you? How do you do research? The author recorded the data using a Zoom H2 audio recorder and transcribed the recordings in NVivo.

The main survey was designed during and after the 2014 summer Institute of Research Design in Librarianship. Before launching the survey, the author did a pilot test of the survey with students who participated in the focus groups. The author hired a research assistant for the survey portion of the study, and together the researchers piloted the survey, used a screening survey to recruit survey participants, and collected data during the fall 2014 semester. Flyers were distributed on campus advertising the study with a link to the screening survey. The researchers also staffed an informational table about the survey in various spaces around the campus, including the library. Several offices agreed to promote the screening survey on their Facebook pages and their mailing lists (Women’s Center, the Office of Graduate Admissions, and the office of student activities). The screening survey collected demographic and educational background information, which allowed researchers to include participants based on the following criteria:

• Participants are foreign-born
• Participants are undergraduate and graduate students

Qualifying participants received an e-mail with a link to the full survey within two days of completing the screening survey. The full survey took 30 to 40 minutes to complete and was divided into the following parts: demographics, educational background, language use, library use, and cultural questions (including views on American-style research). For this paper the author examines the data from the language use and demographics sections.

Because the survey was linked to their e-mail, respondents could finish one part of the survey and return to the survey at another time to complete the other parts. The researchers sent out e-mails to remind participants that they needed to complete the full survey. Once SurveyMonkey indicated that a survey was complete, the researchers made an appointment with the student via e-mail to distribute the $10 incentive.

Participants also had the option of doing an in-person survey, where the researcher would help the participant complete the survey in a classroom setting, after which they would immediately receive their incentive, though none of the eligible participants chose this option.

There were 3,004 foreign-born students at Brooklyn College in the fall 2014 semester. 43 A total of 274 students were screened, and 123 eligible students were invited to participate in the full survey. Of these, 103 responded and participated. Ten surveys contained partial responses and were discarded, and one survey was discarded because the person self-reported low English reading and writing ability. For this paper, 92 complete surveys were analyzed. The survey response rate was 74 percent (# complete surveys / # eligible to participate).

Survey Data and Analysis

The author downloaded the survey data from SurveyMonkey and analyzed CS and language use data in Excel, and then in SPSS 21 using independent samples t-tests to look for associations between mean scores in language use variables (CS, language domains, information tasks, and language ability) and demographic grouping variables, which include student status (undergraduate or graduate), immigration status (permanent or temporary status), first-generation student status, gender, and race/ethnicity. T-tests were not conducted for grouping variables that had fewer than 10 respondents (for example, there were only 5 respondents who identified as Hispanic and 2 respondents who identified as Middle Eastern). The author used Somers’ d test to look for associations between language use variables and the following dependent variables: age, age arrived in the United States, years lived in the United States, and median income (estimated from the 2014 American Community Survey 44 using respondent zip codes).

The author has chosen to treat the Likert-type variables as continuous rather than ordinal. The survey data meet all the assumptions for an independent samples t-test 45 and the Somers’ d test for associations. 46 The author has carefully followed Arlene Fink’s survey methods book 47 in this analysis.

The author created a variable called code-switching based on the following survey question:

I switch between English and my non-English language(s)

b. Very frequently

c. Occasionally

e. Very rarely

The author coded this variable in SPSS 21 (1 = Never, 6 = Always).

The grouping variables for language domains and information tasks were created from a matrix as shown in figures 1 and 2, with Language 2 (L2) referring to the main language the student spoke in addition to English.

The responses were dummy-coded (0 = no attribute, 1 = presence of the attribute).

The variable for language ability was created from the questions below with responses coded in a five-point Likert-type scale (1 = not at all, 5 = very well).

How well do you speak English?

How well do you understand English?

How well do you read English?

How well do you write English?

How well do you speak L2?

How well do you understand L2?

How well do you read L2?

How well do you write L2?

The first of the demographic variables, student status, was determined by the following question:

Are you a high school, undergraduate, or graduate student?

a. I am not a student [disqualify]

b. I am a high school student [disqualify]

c. I am an undergraduate student

d. I am a graduate student

The author used two questions in the survey that asked about immigration status to construct two variables: permanent immigration status and temporary immigration status. The first question was asked in the screening survey: Are you an international student with one of the following visas: F, J, M, A, H1B, or K? (a. Yes b. No) The second question was asked in the main survey:

What is your immigration status?

a. U.S. citizen

b. U.S. citizen by naturalization

c. Permanent resident

d. Not a U.S. citizen

e. Dual citizenship or nationality

f. Deferred Action for Childhood Arrivals (“DACA”)

g. I do not wish to answer this question

h. I don’t know

Respondents who chose a, b, c, or e in the main survey were grouped as permanent, and dummy-coded in SPSS (0 = not permanent, 1 = permanent). Respondents who chose a in the screening survey question and d or f in the main survey question were grouped as temporary status.

Gender, age, age arrived in the United States, years in the United States, and zip code were constructed from the survey questions below.

What is your gender? a. Male b. Female

What year were you born? [numerical text field]

What year did you move to the United States? [numerical text field]

What is your current zip code? [numerical text field]

First-generation college students were identified using the following question: Ideally, what’s your intention for completing a degree? Check all that apply. Nineteen answers were available, and respondents could choose as many as applied; answer L, “I am the first in my family to get a college degree,” was used to identify first-generation college students. The author dummy-coded the variable (0 = not a first-generation college student, 1 = first-generation college student).

Race and ethnicity categories were taken from adapted from a White House document on reporting race. 48 The question appeared as follows in the survey:

Please indicate your race:

a. Hispanic or Latino (a person of Cuban, Mexican, Puerto Rican, South or Central American, or other Spanish culture or origin, regardless of race)

b. American Indian or Alaska Native (a person having origins in any of the original peoples of North and South America, including Central America, who maintains cultural identification through tribal affiliation or community attachment)

c. Asian (a person having origins in any of the original peoples of the Far East, Southeast Asia, or the Indian Subcontinent, including, for example, Cambodia, China, India, Japan, Korea, Malaysia, Pakistan, the Philippine Islands, Thailand, and Vietnam)

d. Black or African American (a person having origins in any of the black racial groups of Africa)

e. Native Hawaiian or Other Pacific Islander (a person having origins in any of the original peoples of Hawaii, Guam, Samoa, or other Pacific Islands)

f. White (a person having origins in any of the original peoples of Europe)

g. Middle Easterner (a person having origins from the Middle Eastern countries)

h. North African (a person having origins from the North African countries)

i. From multiple races

j. Other (please specify)

The author dummy-coded all the race and ethnicity variables.

Figure 3 shows the binary grouping variables.

Finally, reading preferences were created from the matrix in figure 4. The author created variables for language preference for academic and nonacademic reading to determine whether students who are more likely to code-switch prefer a particular language for academic and leisure reading.

The data illustrate that students who speak a language besides English are more likely to do so in certain language domains such as home, school, or with friends. The ability of students to speak in English and their second language is also associated with how likely these students are to switch between languages.

Independent samples t-tests (hereafter t-tests) were not conducted for race and ethnicity categories with fewer than 10 respondents, which included Hispanic, American Indian, Native Hawaiian, Middle Easterner, North African, and multiple races. The t-tests showed that the following variables did not have a statistical difference in the mean reported CS score for the multilingual respondents:

• Immigration status
• Student status
• First-generation status
• White or nonwhite (race and ethnicity variable)
• Other (race and ethnicity variable)

The author found no associations using Somers’ d test between reported CS and the following variables:

• Years in the United States
• Median income based on zip code
• Ability to speak English
• Ability to understand L2 [language spoken other than English]
• Ability to write L2

Table 2 shows the variables where the author found that foreign-born students are very likely to switch to L2 (the non-English language) for information tasks and in linguistic domains such as home, school, and with friends.

Table 3 shows associations between language ability and CS using Somers’ d test. According to Laerd Statistics, “A value of –1 indicates that all pairs of observations are discordant and a value of +1 indicates that all pairs of observations are concordant.” 49 Table 3 shows moderate to weak effects.

Reported CS is positively associated with higher L2 speaking and reading ability with a moderate effect strength. Additionally, reported CS has an inverse association with the students’ ability to understand, read, and write English. The strength of the inverse association is weak (–0.305, –0.299, –0.237).

Focus Group Data and Analysis

In the focus groups, all but one of the participants were bilingual (counting English and Jamaican Creole English as separate languages), and the single monolingual speaker of English spoke British English. There were 10 instances of students reporting code-switching in the focus group transcripts. The author also counted instances of switching from American to British English as a CS. Below are select quotes from the transcripts when participants were asked about how they used their languages. The following languages were discussed in the focus groups: English, Vietnamese, Hindi, French, French Creole (Kreyol), Cantonese, Mandarin, Hakka, Armenian, Bengali (Bangla), British English, Patwa (Jamaican Creole), Portuguese, Spanish, Russian, and Ukrainian. The data collected and described in this section illustrate the complexity of code-switching, the use of different languages for different types of information, the importance of community, and the presence of dialects both in English and non-English languages.

A quote that exemplifies the complexity of code-switching is from a Chinese student whose parents do not speak the same variety of Chinese. The student referred to the Hakka language as a “dialect,” describing her upbringing in a multi-Chinese dialect household, where neither of her parents spoke English very well. She explained that if she doesn’t want her parents to understand what she is saying she switches to English as her secret language. This quote from her is also illustrative of the linguistic diversity of Chinese languages:

“Yeah, when I am in school. Of course, most of the time I speak English in school. When I go home, I speak to my dad, my mom. I talk with my mom in dialect. Because she doesn’t know how to speak Mandarin. My dad speak to me in dialect. I respond back in Mandarin because my dad is not much good. Not much fluent in the dialect. But my dad I can hear and listen. You know, for my sisters, my brothers, I speak in like half English, half dialect, and maybe some has just like all everything mixed one or I don’t want to know what my parents to know what I am talking about, I speak English.”

A female Colombian undergraduate student described which languages she uses for different types of information:

‘‘I do like Spanish literature. When it comes to reading for pleasure, 50/50 for Spanish and English. I read a lot of religion in Spanish. Research, I don’t. And it’s a little limited in Spanish, because my entire college career has been in English. So even when I’ve tried, it’s difficult. I’ve done it, like when I travel back home. Or if I’m trying to read a text that’s in Spanish, and it’s academic, it takes a little bit of work. Because I have the information in my head in English.”

The same student also said:

“I pretty much read what I come across. As long as it’s interested, is interesting, and I’m interested in it. I don’t really have a preference with the news per se. I have always read the Bible in Spanish. Family and friends, and music. It’s like 80% in Spanish, 20% in English. Television is mostly English, just because lack of good stuff (laughter). Seriously. All those telenovelas, get out! So that’s out.”

Academic vocabulary in English and Spanish is learned. This student is obviously a fluent Spanish speaker and reader, but she knows her academic discipline in English only so she cannot readily code-switch to Spanish for academic tasks.

Community plays an important role in CS and, because this male Indian graduate student does not have a community of Bengali speakers, he has limited ability in Bengali. But his comprehension is maintained through preference for Bengali music:

“For research, basically I do everything in English because I have little family and friends here and back home so I can barely speak in Bengali. Other than that, I use English everywhere. For research, I follow the Wall Street Journal and the Economist. I don’t follow Bengali newspapers. Per se about music, I prefer Bengali, and Hindi, and English as well.”

This quote illustrates how a language may not manifest in one ability (like speaking) but can still be used and enjoyed for other kinds of information (music, movies, oral poetry, etc.). And when probed about mixing Hindi and English, the male Bengali student said, “Yeah. It’s Hinglish. Hindi and English. Yeah.”

English, like any language, has dialects. An undergraduate student from Grenada described how British English influences the kinds of information sources she seeks: “Yeah, in Grenada we’re more influenced by the British, so the spelling of our words are along the British standard.” And when probed about her preferences for British English sources, she said:

“Yes, maybe because I’m familiar with it. If I go on the Internet, I would look up bbc.uk. I’m only here three years, so I’ll go back and forth. So it’s not like I’m going to assimilate into the culture. I’m just like an outsider looking in. I feel most comfortable with the British way of news, expressing, so their writing. I don’t know how to explain. It’s so different here. When I did A levels, years ago. It’s just a whole different experience. I started at a community college, BMCC [Borough of Manhattan Community College]. It was like A level all over again, although it was an associate degree. I had to focus on the spelling of the words, how we spell color and you spell color. C-o-l-o-u-r, we spell it different. It was just, no other difficulty really.”

This is an example of how culture influences information seeking. This preference for British sources was verified by the two other Jamaican students who also spent some of their formative education in Jamaica.

This female Bengali undergraduate student uses Bengali exclusively with her family:

“So in school I speak in English. And I live with my aunt. And she works 14 hours a day, so I hardly see her. Whenever I see her for half an hour or an hour, we’re going to talk in Bangla [Bengali], but we don’t like mix stuff. Some of the words are in English. Like we don’t have a word for chair that’s in Bangla. So some of the words has to be in English. But um, most of the cases we don’t mix two languages. And even when I’m talking to my parents I’ll be talking Bangla.”

In response to a probe about making friends in school, this student noticed that the variety of Bengali spoken by the students she meets in Brooklyn contain an unusual accent. This is an example of the linguistic diversity that can exist in the same language between diaspora speakers and recently arrived speakers. Her CS shows that she enjoys and uses her languages for a variety of information consumption:

“I have met like a couple of Bangladeshi people in the school. But as she was saying, they don’t speak Bangla. What I feel they have really weird accent when they speak in Bangla. I feel weird when they speak in Bangla. So they like speak okay, if you speak comfortable speaking in English, I don’t have a problem. And in case of music, I feel like I’m more into Bangla and Hindi music. More than English music. English music is okay,…. If I listen to a song I’ll only listen to like, once or twice. But if I’m gonna listen to music, like the whole day, it has to be Bangla and Hindi. And even in the case of movies or soaps. I watch a lot of English soaps. Like I watched How I Met Your Mother, the whole 9th season. And I watch like Scrubs, and Big Bang Theory. And what’s the other one? I like watching CSI and Breaking Bad. …Yeah sorry, Breaking Bad.”

The following quote from a male Vietnamese undergraduate student shows his sophistication when looking for news about his home country: “I do read Vietnamese news and writings that’s about Vietnamese because I want to know what’s going on at the university. We still have government controlled media. So sometimes the news has bias. I like Vietnamese pop, for my personal life.” This male Chinese undergraduate student, after multiple probes, simply stated that both English and Chinese are important in his life: “I think the language is important, because I prefer the both language in my life for my enjoyment.”

The focus group data clearly show that multilinguals are using languages in different ways to consume and communicate information; they are switching readily in their daily lives before going to school, at school, and after school. These data highlight the complex and fluid ways language and culture influence how multilingual foreign-born students consume information. It’s important to note that within large language categories exist smaller subgroupings or varieties, whether they are varieties of English or Chinese dialects. These language groupings are not homogeneous populations, and care should be taken by academic librarians to acknowledge these differences in both accent and grammaticality. Interactions with foreign-born students should focus on communicating and not correcting.

The survey and data provide evidence of CS in foreign-born students in academic environments like the library at Brooklyn College (“I use L2 at school,” P < .001, MD = 5) and most likely with co-ethnic friends (“I speak L2 with my friends,” P < .001, MD = 1.81). The data show that students who report a second language are more likely to use it for a variety of information tasks like searching the Internet, consuming media, and communicating through social media or similar texting mediums, as shown in table 1. These data show that home is the place that L2 speakers will most likely code-switch ( P < .001, MD = 2.559). They also show that L2 is used for academic purpose like reading ( P < .05, MD = .644), especially for news and current events ( P < .001, MD = 1.194). The moderate association in table 2 between code-switching, L2 speaking, and L2 reading needs to be further investigated. Does it suggest that people who speak and read a language are more engaged in that language? It’s difficult to tell and is further complicated for languages that don’t have a strong print culture (Creoles, for example). The inverse association between English ability and CS is also interesting in that the higher the ability reported in English comprehension, reading, and writing, the lower the likelihood for CS. Brooklyn College library is a multilingual space, as the survey data show, and the language environment of foreign-born students is complex and fluid.

An important detail about the data collected in the survey is that 38 of the 92 respondents self-identified as Asian, and they are more likely than any of the other demographic category surveyed to code-switch ( P < .001, MD = 1.283). Additionally, there were respondents from the Caribbean who did not self-identify with the category of black/African American; they chose the “other” response to write in Caribbean. Similarly, there were Bengali students who did not choose to self-identify with Asian (the author considered, but did not include, a category for South Asian), and several of the respondents wrote in Bengali by selecting the “other” category. It was unfortunate that there were very few Hispanic foreign-born students, since Spanish is widely spoken in New York City.

Furthermore, it’s important to think about how Creole languages do not really appear in more formal academic language domains. For example, there are no calculus books written in Haitian Kreyol as there would be calculus books written in Vietnamese, Mandarin, or Spanish. Another question that these survey data cannot effectively comment on is CS for speakers of other varieties of English, like British English: do those speakers code-switch across varieties of English? For example, would the Trinidadian student from the focus group have reported in the survey or even be conscious of Americanizing her spelling and accent to function in an American English academic environments; is that a form of code-switching? These finite, qualitative distinctions are difficult to tease out in a survey. Fortunately, the focus group data show and identify some of these experiences and processes.

Synthesis of the Survey and Focus Group Data

This section discusses two instances when the quantitative findings are complemented by the qualitative findings. There is one instance when the quantitative findings are contradicted by the qualitative data.

Both data sets in this study capture the importance of CS at home. The survey data correlate that home is an important place for language use, and the narrative from the Chinese student who speaks various dialects of Chinese and English with her parents, brothers, and sisters reveals the complexity of her language use. Both data sets show that there is diversity within Chinese and that, as librarians who value linguistic diversity, it is important to be aware that Chinese is not a monolithic language or cultural category. The application to practice is clear: if a library has large groups of Chinese speakers, it will be important to know and identify which cultural and linguistic groups use the academic library.

The second pattern found in both data sets includes the influence of language over the preference for consuming information like academic reading and news. The findings suggest that it is likely that students will code-switch in academic reading contexts and when reading news. The focus group data give us more insight into this information behavior. For example, the data from the Colombian undergraduate student tell us that her code-switching is in very specific domains, like leisurely reading and religious reading, and that she has difficulty reading academic Spanish. This information is important to know for building collections and for supporting students who may want to take their academic experiences to their home country for an internship, for a job, or for pursuing graduate work. In this study we also see how dialect preferences in English influence how Caribbean speakers gravitate toward the British Standard English with consuming news and current events, which is also an example of dialect-switching. These data showing how English dialects can affect information behavior is important for helping librarians understand that, not only is there diversity across languages, there are dialectal differences within major language groups like English, Chinese, Spanish, and others.

The third pattern found in the survey, that students are more likely to use L2 with friends, is contradicted by the example of the female Bengali student who finds that Bengali accents spoken by Brooklyn students are unusual. Again, while students are likely to switch with friends, students who have just arrived may not have friends who speak the same dialect. This is important when distinguishing between immigrant students and international students when it comes to conducting further studies of co-ethnic language use in academic libraries.

Future studies involving code-switching should focus on capturing both quantitative and qualitative data. It’s just as important to understand individual experiences and processes around code-switching, because this helps in interpreting the statistical data. Qualitative data can also reveal missed or new variables that will give the researcher better insight into the relationships between language and information.

Limitations

This survey and analysis contain some limitations that should be kept in mind when considering the findings. The first limitation is the nature of self-reported usage data, though the author attempted to increase the internal reliability of the data by using focus groups to triangulate the results. The second limitation is related to sampling: the focus groups were very small, and the survey was a convenience sample. Third, income as a variable was collected as a rough estimate using ACS 2014 data by zip code. This is not the best way to capture income data, but it does provide some information. Finally, it should be noted that many of the terms used throughout this article, including race, gender, ethnicity, and immigration status, are social science terms often used in survey research; their use is not meant to give offense or intentionally exclude any groups. There is always some aspect of reductionism that occurs in survey-based research, and the author welcomes feedback on how to make the survey categories more inclusive.

Implications for Academic Library Services

CS is a well-established phenomenon outside of library and information science, and this paper aimed to introduce librarians to this concept and document this behavior in an academic library setting. There are many implications of CS for academic library services, especially in the area of linguistic diversity, standard 6 of ACRL’s Diversity Standards: Cultural Competency for Academic Libraries, which reads: “Librarians and library staff shall support the preservation and promotion of linguistic diversity, and work to foster a climate of inclusion aimed at eliminating discrimination and oppression based on linguistic or other diversities.” 50

The data from this study provide evidence that language influences the information behavior of students in the form of code-switching and dialect switching. More research could be done to investigate information code-switching, which the author has broadly defined as changing languages or dialects for particular information tasks. This kind of research would allow librarians to map language use, language choice, and language preferences of students to actual library collections, services, and resources. Furthermore, this research would be valuable for serving first-generation college students, generation 1.5 students, international students, and immigrant students.

Another area of critical importance is being inclusive of non-English languages in collection development: “collection managers should be attentive to represent the linguistic needs of library constituents, and assure that library resources in print or electronic formats are available, especially to support the academic curricula reflecting all diversity issues, including those of visually disabled constituents.” 51 Increasing the visibility of non-English scholarly sources can be as simple as creating library guides that show students how to access peer-reviewed journals and open access indexes in non-English languages. Engaging with the scholarly literature in non-English languages is particularly important in the social sciences. For example, librarians could create a guide that would allow Spanish-speaking students of urban sociology to engage with and synthesize sociological ideas in Spanish language journals with concepts from English language journals. The practice of incorporating non-English sources into English-language papers is a long scholarly tradition in the humanities. Language access is also an important concept in the diversity standards: “Provide and advocate for the provision of information, reference, referrals, instruction, collection management, and other services in the language appropriate to their constituencies, including the use of interpreters.” As the CS data suggest, our libraries are not monolingual spaces, so making sure that printers can print in different scripts (and in general having technology capable of supporting users’ linguistic preferences), as well as having welcome signage in other languages are steps that libraries can take to make non-English speakers feel more included when using the library as a space for studying or meeting with classmates. Ideally, supporting linguistic diversity in academic libraries would include multilingual staff who could create library instruction and other academic library services that cater to large linguistic populations served by the academic library.

The purpose of this study was to explore language use, language choice, and language preferences in academic libraries, and the author found evidence for code-switching patterns in both qualitative and quantitative data. In the analysis, the author maintains that code-switching patterns are correlated with information tasks and argues that more research could be done on information code-switching to give librarians data on language use and apply those data to library services. For the foreign-born students analyzed in this study, it is clear that their culture and their non-English language represent an active and important part of their identity, information consumption, and academic socialization. In their language choice for information, there is enough statistically significant evidence for information code-switching, when students switch languages for a particular information task. Yet, in most of the academic library literature, these active language communities, their patterns of use, and their preferences have not been the subject of research. These data show that Brooklyn College Library is a rich multilingual space, yet there are only a few studies that discuss multilingualism in academic libraries. What can academic librarians do with these kinds of data?

There are many practical recommendations relating to linguistic diversity, including creating a multilingual-friendly environment. Reference assistance could include offering specialized library instruction or orientations for immigrant students, first-generation college students, and international students. Public services staff could also receive linguistic diversity training that includes information on ESL and EFL populations. Linguistic diversity training might also focus content on creating sensitivity and awareness of patrons who are linguistic minorities (for example, Spanish speakers who speak other Central American languages such as Mayan languages), as well as information about Creoles and pidgins, how accents work, nonwritten languages, and varieties of spoken English that may be relevant to the patron population. Computer labs can be language-friendly, with a variety of keyboard formats and printers available for people who need to print e-mails, share notes, and look up concepts in languages other than English. In the focus group interview with the Colombian undergraduate student, she spoke about wanting to gain some experience working in Colombia; however, she lacked the Spanish academic vocabulary to be competitive. To help students like her, academic librarians could create LibGuides for non-English scholarly sources that include Latindex, for example ( www.latindex.unam.mx/latindex/inicio ). Libraries could focus on hiring multilingual librarians. There is clearly a need for more research in transnationalism, especially in academic libraries that have a high number of foreign-born students. Are these students trying to use their American college degrees and create transnational careers that take advantage of their cultural capital? Are academic libraries spaces currently treated as monolingual rather than multilingual spaces? How does this affect our practice, and how can libraries change to support these students in their information needs? Monica Jacobe, director of the Center for American Language & Culture at The College of New Jersey, speaks about immigration trends in student success:

First-generation college students will no longer be primarily American-born students from working class families. Instead, many more students in that category will be recent immigrants, born all over the world, who completed high school in the U.S. For many schools, they will “look” on paper like domestic applicants, but the support they need will be very different.

How can academic librarians imagine a shift from monocultural and monolingual approaches to multilingual approaches? And what services will need to be rethought? These are the kinds of questions additional studies on language use and libraries can answer. CS is just one conceptual tool from sociolinguistics that has very practical applications in our work with international and immigrant students.

1. Barbara E. Bullock and Almeida Jacqueline Toribio, “Themes in the Study of Code-Switching,” in The Cambridge Handbook of Linguistic Code-Switching , eds. Barbara E. Bullock and Almeida Jacqueline Toribio, Cambridge Handbooks in Linguistics (Cambridge; New York: Cambridge University Press, 2009), 1–19.

2. Jan-Peter Blom and John J. Gumperz, “Social Meaning in Linguistic Structure: Code-Switching in Norway,” in Directions in Sociolinguistics: The Ethnography of Communication , eds. John J. Gumperz and Dell Hymes (New York: Blackwell, 1986), 409.

3. Ibid., 411.

4. Ibid.

5. Carol Myers-Scotton, “The Rise of Codeswitching as a Research Topic,” in Social Motivations for Codeswitching: Evidence from Africa (Oxford: Clarendon Press, 1995), 57.

6. Joshua Fishman, “Domains and the Relationship between Micro- and Macro-Sociolinguistics,” in Directions in Sociolinguistics: The Ethnography of Communication , eds. John J. Gumperz and Dell Hymes (New York: Blackwell, 1986), 441.

7. Ibid., 442–43.

8. L.B. Breitborde, “Levels of Analysis in Sociolinguistic Explanation: Bilingual Code Switching, Social Relations, and Domain Theory,” International Journal of the Sociology of Language 1983, no. 39 (Jan. 1983): 19.

9. Fishman, “Domains and the Relationship between Micro- and Macro-Sociolinguistics,” 447.

10. Ibid., 448.

11. Alejandro Portes and Rubén Rumbaut, “Children of Immigrants Longitudinal Study (CILS), 1991–2006,” Children of Immigrants Longitudinal Study (CILS), 1991–2006 (ICPSR 20520) , 2009, available online at www.icpsr.umich.edu/icpsrweb/RCMD/studies/20520 [accessed 24 March 2015].

12. Breitborde, “Levels of Analysis in Sociolinguistic Explanation,” 5.

13. Shana Poplack, “Sometimes I’ll Start a Sentence in Spanish Y Termino En Español: Toward a Typology of Codeswitching,” Linguistics 18, no. 7/8 (1980): 581–618, doi:10.1515/ling.1980.18.7-8.581 .

14. Ibid., 583.

15. Ibid., 584.

16. Ibid., 600.

17. Ibid., 610.

18. Jannis Androutsopoulos, “Language and the Three Spheres of Hip Hop,” in Global Linguistic Flows: Hip Hop Cultures, Youth Identities, and the Politics of Language , eds. H. Sammy Alim, Awad Ibrahim, and Alastair Pennycook (New York: Routledge, 2009), 54–58.

19. Elena Seoane, “World Englishes Today,” in World Englishes: New Theoretical and Methodological Considerations , eds. Elena Seoane and Cristina Suárez Gómez, Varieties of English around the World G57 (Amsterdam; Philadelphia: John Benjamins Publishing Company, 2016), 1.

20. Marta Fairclough and Flavia Belpoliti, “Emerging Literacy in Spanish among Hispanic Heritage Language University Students in the USA: A Pilot Study,” International Journal of Bilingual Education and Bilingualism 19, no. 2 (Mar. 3, 2016): 185–201, doi:10.1080/13670050.2015.1037718 .

21. Ibid., 186.

22. Ibid., 189.

23. Magdalena Malechová, “Multilingualism as a Sociolinguistic Contact Phenomenon with Regard to Current Forms of Multilingual Communication Code-Switching as One of the Contemporary Communication Trends,” Višejezičnost Kao Sociolingvistički Fenomen Kontakta Imajući U Vidu Suvremene Oblike Višejezične Komunikacije Mijenjanje Kodova Kao Suvremeni Komunikacijski Trend 49, no. 1/2 (July 2016): 86–93.

24. Ibid., 91.

25. Bettina Kümmerling-Meibauer, “Code-Switching in Multilingual Picturebooks,” Bookbird: A Journal of International Children’s Literature (Johns Hopkins University Press) 51, no. 3 (July 2013): 12–21.

26. Ibid., 19.

27. Amanda B. Click, Claire Walker Wiley, and Meggan Houlihan, “The Internationalization of the Academic Library: A Systematic Review of 25 Years of Literature on International Students,” doi:10.5860/crl.v78i3.16591 .

28. Nadia Caidi, Danielle Allard, and Lisa Quirke, “Information Practices of Immigrants,” Annual Review of Information Science and Technology 44, no. 1 (Jan. 1, 2010): 515, doi:10.1002/aris.2010.1440440118 .

29. Ignacio J. Ferrer-Vinent, “For English, Press 1: International Students’ Language Preference at the Reference Desk,” Reference Librarian 51, no. 3 (Sept. 2010): 189–201, doi:10.1080/02763871003800429 .

30. Karen Bordonaro, “Is Library Database Searching a Language Learning Activity?” College & Research Libraries 71, no. 3 (May 2010): 273–84.

31. Sara Luly and Holger Lenz, “Language in Context: A Model of Language Oriented Library Instruction,” Journal of Academic Librarianship 41, no. 2 (Mar. 2015): 140–48, doi:10.1016/j.acalib.2015.01.001 .

32. Frans Albarillo, “Is the Library’s Online Orientation Program Effective with English Language Learners?” College & Research Libraries , 78, no. 5 (July 2017): 656–59, doi:10.5860/crl.78.5.652 .

33. Sonia Smith, “Library Instruction for Romanized Hebrew,” Journal of Academic Librarianship 41, no. 2 (Mar. 2015): 197–200, doi:10.1016/j.acalib.2014.08.003 .

34. Ibid., 199.

35. Molavi Fereshteh, “Main Issues in Cataloging Persian Language Materials in North America,” Cataloging & Classification Quarterly 43, no. 2 (Dec. 8, 2006): 77–82, doi:10.1300/J104v43n02_06 .

36. Kim SungKyung, “Romanization in Cataloging of Korean Materials,” Cataloging & Classification Quarterly 43, no. 2 (Dec. 8, 2006): 53–76, doi:10.1300/J104v43n02_05 .

37. Hikaru Nakano, “Non-Roman Language Cataloging in Bulk: A Case Study of Japanese Language Materials,” Cataloging & Classification Quarterly 55, no. 2 (Feb. 2017): 75–88, doi:10.1080/01639374.2016.1250853 .

38. Albarillo, “Is the Library’s Online Orientation Program Effective with English Language Learners?” 3.

39. Xiang Li, Kevin McDowell, and Xiaotong Wang, “Building Bridges: Outreach to International Students via Vernacular Language Videos,” Reference Services Review 44, no. 3 (July 2016): 325, doi:10.1108/RSR-10-2015-0044 .

40. Misa Mi and Yingting Zhang, “Culturally Competent Library Services and Related Factors among Health Sciences Librarians: An Exploratory Study,” Journal of the Medical Library Association 105, no. 2 (Apr. 2017): 135, doi:10.5195/jmla.2017.203 .

41. Ibid., 136.

42. Ibid., 133.

43. Institutional Research and Data Analysis at Brooklyn College, City University of New York, “Fall 2014 Final Enrollment Reports, Enrollment Table 21 Country of Birth,” available online at www.brooklyn.cuny.edu/bc/offices/avpbandp/ipra/enrollment/F14/Enrollment-Table21.pdf [accessed 20 November 2016].

44. United States Census Bureau, “American FactFinder,” available online at http://factfinder.census.gov/faces/nav/jsf/pages/index.xhtml [accessed 20 November 2016].

45. Laerd Statistics, “Independent-Samples T-Test Using SPSS Statistics,” Statistical Tutorials and Software Guides (2015), available online at https://statistics.laerd.com [accessed 21 November 2016].

46. Laerd Statistics, “Somers’ D Using SPSS Statistics,” Statistical Tutorials and Software Guides (2016), available online at https://statistics.laerd.com [accessed 21 November 2016].

47. Arlene Fink, How to Conduct Surveys: A Step-by-Step Guide (Los Angeles: SAGE, 2013), 45.

48. Executive Office of the President, Office of Management and Budget, “Revisions to the Standards for the Classification of Federal Data on Race and Ethnicity,” The White House, available online at https://www.whitehouse.gov/node/15626 [accessed 21 November 2016].

49. Laerd Statistics, “Somers’ D Using SPSS Statistics.”

50. Association of College and Research Libraries, “Diversity Standards: Cultural Competency for Academic Libraries” (2012), available online at www.ala.org/acrl/standards/diversity [accessed 30 August 2016].

51. Racial and Ethnic Diversity Committee Members, “Diversity Standards: Cultural Competency for Academic Libraries (2012),” Association of College and Research Libraries, available online at www.ala.org/acrl/standards/diversity [accessed 21 November 2016].

* Frans Albarillo is Assistant Professor/Reference & Instruction at Brooklyn College, City University of New York; e-mail: [email protected] . ©2018 Frans Albarillo, Attribution-NonCommercial ( http://creativecommons.org/licenses/by-nc/4.0/ ) CC BY-NC.

Article Views (Last 12 Months)

Contact ACRL for article usage statistics from 2010-April 2017.

Article Views (By Year/Month)

© 2024 Association of College and Research Libraries , a division of the American Library Association

Print ISSN: 0010-0870 | Online ISSN: 2150-6701

ALA Privacy Policy

ISSN: 2150-6701

• Personal Essay

Editor's Letter: We're Talking Code-Switching This Latine Heritage Month

"Pórtate bien; esta gente son blanca" is a line I'd often hear my mom say when I was growing up. It translates to: "Behave yourself; these people are white."

I was introduced to code-switching at a relatively young age — significantly younger than a lot of my inner-city Latine peers. I learned the skill before I even attended school because my parents were desperately trying to assimilate back into American culture after attending college in the Dominican Republic.

At the time, in the early '90s, my parents weren't even aware that there was an actual term to describe what they were learning as two college-educated Dominican immigrants trying to achieve the American dream. They didn't know that trying to mask their Latin American accents and attempting to reduce the volume at which they spoke was actually a form of code-switching. Back then, the only folks using that term were academics. But that's what it was. My parents were teaching my younger siblings and me how to speak, behave, dress, and carry ourselves in predominantly white spaces. They were teaching us an inherent survival tool.

What does it mean to code-switch, exactly? For many marginalized communities, particularly communities of color, it means assimilating to white-centric ideas around what it means to be "well-behaved" or "professional." It's shifting the language you use or the way you physically present or express yourself to make others more comfortable.

Don't I deserve to be treated fairly and respectfully regardless of my background or the color of my skin?

By the time I was in college, code-switching was so natural to me that I wasn't even always cognizant of when I was doing it. I learned how to do it everywhere, from school to jobs and even in friend groups where I was the only person representing both Latinidad and Blackness. Of course, I knew the spaces where I could authentically be myself — often those occupied by Black and Latine folks — but I had learned how to instantly tone myself down in spaces where the folks in the room were mostly white. I soon began to consider my code-switching to be a superpower. And in many ways, it can be. But I also learned that it could come at a serious psychological cost.

My perception of code-switching almost immediately changed when I started working for Latin media startups. My direct reports were all Latina women, and they often had either Latin American, Miami, or New York accents. Many of them were extremely assertive when they spoke — and unapologetically spoke loudly and in Spanglish. Not to mention everyone wore whatever they wanted: long acrylic nails, red lipstick, oversized hoop earrings, heels, and crop tops. There was no official dress code.

Going to work felt like getting together with family. To be clear, it wasn't perfect. Latine families can be toxic, too. But I always felt safe enough to be my most authentic self, and I've carried that into every role I've taken throughout my career. It made me realize that before that point, whenever I found myself in a white-dominated space, I was adjusting everything about myself, from my style of speech to how I express myself, in the hopes that I would be treated fairly and respectfully. Then it hit me: at what cost? At the cost of honoring myself? And don't I deserve to be treated fairly and respectfully regardless of my background or the color of my skin? It taught me that being myself and being able to embrace all the things that make me uniquely Johanna, including my Latina identity and Dominican heritage, has absolutely nothing to do with professionalism.

Coming to this realization early on in my career made me realize how much energy it takes for one to code-switch for eight hours a day, five days a week, and in some cases more. It became crystal clear to me how that can easily come in the way of one's creativity, confidence, and even productivity.

Research shows that code-switching often occurs in spaces where negative stereotypes of marginalized communities run high , and it is often the folks who subscribe to these stereotypes who lead and dictate what is considered "appropriate" or "professional behavior." I am proud to be living in a day and age when Latines not only have been unapologetically doing away with the code-switching — and the hiding, masking, and toning down of their Latinidad — but are also thriving as a result of it. Some of our biggest stars are perfect examples of that. Bad Bunny is the No. 1 artist in the world, singing Latin trap and reggaetón music while refusing to do it in English. Sofía Vergara proved she's an excellent actress in some ways because of her heavy Colombian accent with a recent Emmy nomination for her outstanding performance in "Griselda."

That's why there's no better time than this Latine Heritage Month to delve into how this generation of Latines are thriving and finding success from simply being their most authentic selves, regardless of the spaces they occupy. Through opinion pieces, profiles, and personal essays, we're exploring how Latines across industries drastically transformed their careers and their lives when they started to embrace themselves fully. You can start reading here .

For every Latine reader who comes across this package, I hope it reminds you that you should behave however you want!

Con amor, Johanna Ferreira, Juntos content director

Johanna Ferreira is the content director for PS Juntos. With more than 10 years of experience, Johanna focuses on how intersectional identities are a central part of Latine culture. Previously, she spent close to three years as the deputy editor at HipLatina, and she has freelanced for numerous outlets including Refinery29, Oprah magazine, Allure, InStyle, and Well+Good. She has also moderated and spoken on numerous panels on Latine identity.

• Conferences
• Last Updated: September 13, 2024
• In AI Mysteries

Top Machine Learning Research Papers

• by Dr. Nivash Jeevanandam

Advances in machine learning and deep learning research are reshaping our technology. Machine learning and deep learning have accomplished various astounding feats, and key research articles have resulted in technical advances used by billions of people. The research in this sector is advancing at a breakneck pace and assisting you to keep up. Here is a collection of the most important scientific study papers in machine learning.

Rebooting ACGAN: Auxiliary Classifier GANs with Stable Training

The authors of this work examined why ACGAN training becomes unstable as the number of classes in the dataset grows. The researchers revealed that the unstable training occurs due to a gradient explosion problem caused by the unboundedness of the input feature vectors and the classifier’s poor classification capabilities during the early training stage. The researchers presented the Data-to-Data Cross-Entropy loss (D2D-CE) and the Rebooted Auxiliary Classifier Generative Adversarial Network to alleviate the instability and reinforce ACGAN (ReACGAN). Additionally, extensive tests of ReACGAN demonstrate that it is resistant to hyperparameter selection and is compatible with a variety of architectures and differentiable augmentations.

This article is ranked #1 on CIFAR-10 for Conditional Image Generation.

For the research paper, read here .

For code, see here .

Dense Unsupervised Learning for Video Segmentation

The authors presented a straightforward and computationally fast unsupervised strategy for learning dense spacetime representations from unlabeled films in this study. The approach demonstrates rapid convergence of training and a high degree of data efficiency. Furthermore, the researchers obtain VOS accuracy superior to previous results despite employing a fraction of the previously necessary training data. The researchers acknowledge that the research findings may be utilised maliciously, such as for unlawful surveillance, and that they are excited to investigate how this skill might be used to better learn a broader spectrum of invariances by exploiting larger temporal windows in movies with complex (ego-)motion, which is more prone to disocclusions.

This study is ranked #1 on DAVIS 2017 for Unsupervised Video Object Segmentation (val).

Temporally-Consistent Surface Reconstruction using Metrically-Consistent Atlases

The authors offer an atlas-based technique for producing unsupervised temporally consistent surface reconstructions by requiring a point on the canonical shape representation to translate to metrically consistent 3D locations on the reconstructed surfaces. Finally, the researchers envisage a plethora of potential applications for the method. For example, by substituting an image-based loss for the Chamfer distance, one may apply the method to RGB video sequences, which the researchers feel will spur development in video-based 3D reconstruction.

This article is ranked #1 on ANIM in the category of Surface Reconstruction. 

EdgeFlow: Achieving Practical Interactive Segmentation with Edge-Guided Flow

The researchers propose a revolutionary interactive architecture called EdgeFlow that uses user interaction data without resorting to post-processing or iterative optimisation. The suggested technique achieves state-of-the-art performance on common benchmarks due to its coarse-to-fine network design. Additionally, the researchers create an effective interactive segmentation tool that enables the user to improve the segmentation result through flexible options incrementally.

This paper is ranked #1 on Interactive Segmentation on PASCAL VOC

Learning Transferable Visual Models From Natural Language Supervision

The authors of this work examined whether it is possible to transfer the success of task-agnostic web-scale pre-training in natural language processing to another domain. The findings indicate that adopting this formula resulted in the emergence of similar behaviours in the field of computer vision, and the authors examine the social ramifications of this line of research. CLIP models learn to accomplish a range of tasks during pre-training to optimise their training objective. Using natural language prompting, CLIP can then use this task learning to enable zero-shot transfer to many existing datasets. When applied at a large scale, this technique can compete with task-specific supervised models, while there is still much space for improvement.

This research is ranked #1 on Zero-Shot Transfer Image Classification on SUN

CoAtNet: Marrying Convolution and Attention for All Data Sizes

The researchers in this article conduct a thorough examination of the features of convolutions and transformers, resulting in a principled approach for combining them into a new family of models dubbed CoAtNet. Extensive experiments demonstrate that CoAtNet combines the advantages of ConvNets and Transformers, achieving state-of-the-art performance across a range of data sizes and compute budgets. Take note that this article is currently concentrating on ImageNet classification for model construction. However, the researchers believe their approach is relevant to a broader range of applications, such as object detection and semantic segmentation.

This paper is ranked #1 on Image Classification on ImageNet (using extra training data).

SwinIR: Image Restoration Using Swin Transformer

The authors of this article suggest the SwinIR image restoration model, which is based on the Swin Transformer . The model comprises three modules: shallow feature extraction, deep feature extraction, and human-recognition reconstruction. For deep feature extraction, the researchers employ a stack of residual Swin Transformer blocks (RSTB), each formed of Swin Transformer layers, a convolution layer, and a residual connection.

This research article is ranked #1 on Image Super-Resolution on Manga109 – 4x upscaling.

Artificial Replay: A Meta-Algorithm for Harnessing Historical Data in Bandits

Ways to incorporate historical data are still unclear: initialising reward estimates with historical samples can suffer from bogus and imbalanced data coverage, leading to computational and storage issues—particularly in continuous action spaces. The paper addresses the obstacles by proposing ‘Artificial Replay’, an algorithm to incorporate historical data into any arbitrary base bandit algorithm. 

Read the full paper here . 

Bootstrapped Meta-Learning

Author(s) – Sean R. Sinclair et al.

The paper proposes an algorithm in which the meta-learner teaches itself to overcome the meta-optimisation challenge. The algorithm focuses on meta-learning with gradients, which guarantees performance improvements. Furthermore, the paper also looks at how bootstrapping opens up possibilities. 

Read the full paper here .

LaMDA: Language Models for Dialog Applications

Author(s) – Sebastian Flennerhag et al.

The research describes the LaMDA system which caused chaos in AI this summer when a former Google engineer claimed that it had shown signs of sentience. LaMDA is a family of large language models for dialogue applications based on Transformer architecture. The interesting feature of the model is its fine-tuning with human-annotated data and the possibility of consulting external sources. This is a very interesting model family, which we might encounter in many applications we use daily. 

Competition-Level Code Generation with AlphaCode

Author(s) – Yujia Li et al.

Systems can help programmers become more productive. The following research addresses the problems with incorporating innovations in AI into these systems. AlphaCode is a system that creates solutions for problems that require deeper reasoning. 

Privacy for Free: How does Dataset Condensation Help Privacy?

Author(s) – Tian Dong et al.

The paper focuses on Privacy Preserving Machine Learning, specifically deducting the leakage of sensitive data in machine learning. It puts forth one of the first propositions of using dataset condensation techniques to preserve the data efficiency during model training and furnish membership privacy.

Why do tree-based models still outperform deep learning on tabular data?

Author(s) – Léo Grinsztajn, Edouard Oyallon and Gaël Varoquaux

The research answers why deep learning models still find it hard to compete on tabular data compared to tree-based models. It is shown that MLP-like architectures are more sensitive to uninformative features in data compared to their tree-based counterparts. 

Multi-Objective Bayesian Optimisation over High-Dimensional Search Spaces 

Author(s) – Samuel Daulton et al.

The paper proposes ‘MORBO’, a scalable method for multiple-objective BO as it performs better than that of high-dimensional search spaces. MORBO significantly improves the sample efficiency and, where existing BO algorithms fail, MORBO provides improved sample efficiencies over the current approach. 

A Path Towards Autonomous Machine Intelligence Version 0.9.2

Author(s) – Yann LeCun

The research offers a vision about how to progress towards general AI. The study combines several concepts: a configurable predictive world model, behaviour driven through intrinsic motivation, and hierarchical joint embedding architectures trained with self-supervised

learning. 

TranAD: Deep Transformer Networks for Anomaly Detection in Multivariate Time Series Data

Author(s) –  Shreshth Tuli, Giuliano Casale and Nicholas R. Jennings

This is a specialised paper applying transformer architecture to the problem of unsupervised anomaly detection in multivariate time series. Many architectures which were successful in other fields are, at some point, also being applied to time series. The research shows improved performance on some known data sets. 

Differentially Private Bias-Term only Fine-tuning of Foundation Models

Author(s) – Zhiqi Bu et al. 

In the paper, researchers study the problem of differentially private (DP) fine-tuning of large pre-trained models—a recent privacy-preserving approach suitable for solving downstream tasks with sensitive data. Existing work has demonstrated that high accuracy is possible under strong privacy constraints yet requires significant computational overhead or modifications to the network architecture.

ALBERT: A Lite BERT

Usually, increasing model size when pretraining natural language representations often result in improved performance on downstream tasks, but the training times become longer. To address these problems, the authors in their work presented two parameter-reduction techniques to lower memory consumption and increase the training speed of BERT. The authors also used a self-supervised loss that focuses on modelling inter-sentence coherence and consistently helped downstream tasks with multi-sentence inputs. According to results, this model established new state-of-the-art results on the GLUE, RACE, and squad benchmarks while having fewer parameters compared to BERT-large. 

Check the paper here .

Beyond Accuracy: Behavioral Testing of NLP Models with CheckList

Microsoft Research, along with the University of Washington and the University of California, in this paper, introduced a model-agnostic and task agnostic methodology for testing NLP models known as CheckList. This is also the winner of the best paper award at the ACL conference this year. It included a matrix of general linguistic capabilities and test types that facilitate comprehensive test ideation, as well as a software tool to generate a large and diverse number of test cases quickly. 

Linformer is a Transformer architecture for tackling the self-attention bottleneck in Transformers. It reduces self-attention to an O(n) operation in both space- and time complexity. It is a new self-attention mechanism which allows the researchers to compute the contextual mapping in linear time and memory complexity with respect to the sequence length. 

Read more about the paper here .

Plug and Play Language Models

Plug and Play Language Models ( PPLM ) are a combination of pre-trained language models with one or more simple attribute classifiers. This, in turn, assists in text generation without any further training. According to the authors, model samples demonstrated control over sentiment styles, and extensive automated and human-annotated evaluations showed attribute alignment and fluency. 

Reformer 

The researchers at Google, in this paper , introduced Reformer. This work showcased that the architecture of a Transformer can be executed efficiently on long sequences and with small memory. The authors believe that the ability to handle long sequences opens the way for the use of the Reformer on many generative tasks. In addition to generating very long coherent text, the Reformer can bring the power of Transformer models to other domains like time-series forecasting, music, image and video generation. 

An Image is Worth 16X16 Words

The irony here is that one of the popular language models, Transformers have been made to do computer vision tasks. In this paper , the authors claimed that the vision transformer could go toe-to-toe with the state-of-the-art models on image recognition benchmarks, reaching accuracies as high as 88.36% on ImageNet and 94.55% on CIFAR-100. For this, the vision transformer receives input as a one-dimensional sequence of token embeddings. The image is then reshaped into a sequence of flattened 2D patches. The transformers in this work use constant widths through all of its layers.

Unsupervised Learning of Probably Symmetric Deformable 3D Objects

Winner of the CVPR best paper award, in this work, the authors proposed a method to learn 3D deformable object categories from raw single-view images, without external supervision. This method uses an autoencoder that factored each input image into depth, albedo, viewpoint and illumination. The authors showcased that reasoning about illumination can be used to exploit the underlying object symmetry even if the appearance is not symmetric due to shading.

Generative Pretraining from Pixels

In this paper, OpenAI researchers examined whether similar models can learn useful representations for images. For this, the researchers trained a sequence Transformer to auto-regressively predict pixels, without incorporating knowledge of the 2D input structure. Despite training on low-resolution ImageNet without labels, the researchers found that a GPT-2 scale model learns strong image representations as measured by linear probing, fine-tuning, and low-data classification. On CIFAR-10, it achieved 96.3% accuracy with a linear probe, outperforming a supervised Wide ResNet, and 99.0% accuracy with full fine-tuning and matching the top supervised pre-trained models. An even larger model, trained on a mixture of ImageNet and web images, is competitive with self-supervised benchmarks on ImageNet, achieving 72.0% top-1 accuracy on a linear probe of their features.

Deep Reinforcement Learning and its Neuroscientific Implications

In this paper, the authors provided a high-level introduction to deep RL , discussed some of its initial applications to neuroscience, and surveyed its wider implications for research on brain and behaviour and concluded with a list of opportunities for next-stage research. Although DeepRL seems to be promising, the authors wrote that it is still a work in progress and its implications in neuroscience should be looked at as a great opportunity. For instance, deep RL provides an agent-based framework for studying the way that reward shapes representation, and how representation, in turn, shapes learning and decision making — two issues which together span a large swath of what is most central to neuroscience. 

Dopamine-based Reinforcement Learning

Why humans doing certain things are often linked to dopamine , a hormone that acts as the reward system (think: the likes on your Instagram page). So, keeping this fact in hindsight, DeepMind with the help of Harvard labs, analysed dopamine cells in mice and recorded how the mice received rewards while they learned a task. They then checked these recordings for consistency in the activity of the dopamine neurons with standard temporal difference algorithms. This paper proposed an account of dopamine-based reinforcement learning inspired by recent artificial intelligence research on distributional reinforcement learning. The authors hypothesised that the brain represents possible future rewards not as a single mean but as a probability distribution, effectively representing multiple future outcomes simultaneously and in parallel. 

Lottery Tickets In Reinforcement Learning & NLP

In this paper, the authors bridged natural language processing (NLP) and reinforcement learning (RL). They examined both recurrent LSTM models and large-scale Transformer models for NLP and discrete-action space tasks for RL. The results suggested that the lottery ticket hypothesis is not restricted to supervised learning of natural images, but rather represents a broader phenomenon in deep neural networks.

What Can Learned Intrinsic Rewards Capture

In this paper, the authors explored if the reward function itself can be a good locus of learned knowledge. They proposed a scalable framework for learning useful intrinsic reward functions across multiple lifetimes of experience and showed that it is feasible to learn and capture knowledge about long-term exploration and exploitation into a reward function. 

AutoML- Zero

The progress of AutoML has largely focused on the architecture of neural networks, where it has relied on sophisticated expert-designed layers as building blocks, or similarly restrictive search spaces. In this paper , the authors showed that AutoML could go further with AutoML Zero, that automatically discovers complete machine learning algorithms just using basic mathematical operations as building blocks. The researchers demonstrated this by introducing a novel framework that significantly reduced human bias through a generic search space.

Rethinking Batch Normalization for Meta-Learning

Batch normalization is an essential component of meta-learning pipelines. However, there are several challenges. So, in this paper, the authors evaluated a range of approaches to batch normalization for meta-learning scenarios and developed a novel approach — TaskNorm. Experiments demonstrated that the choice of batch normalization has a dramatic effect on both classification accuracy and training time for both gradient-based and gradient-free meta-learning approaches. The TaskNorm has been found to be consistently improving the performance.

Meta-Learning without Memorisation

Meta-learning algorithms need meta-training tasks to be mutually exclusive, such that no single model can solve all of the tasks at once. In this paper, the authors designed a meta-regularisation objective using information theory that successfully uses data from non-mutually-exclusive tasks to efficiently adapt to novel tasks.

Understanding the Effectiveness of MAML

Model Agnostic Meta-Learning (MAML) consists of optimisation loops, from which the inner loop can efficiently learn new tasks. In this paper, the authors demonstrated that feature reuse is the dominant factor and led to ANIL (Almost No Inner Loop) algorithm — a simplification of MAML where the inner loop is removed for all but the (task-specific) head of the underlying neural network. 

Your Classifier is Secretly an Energy-Based Model

This paper proposed attempts to reinterpret a standard discriminative classifier as an energy-based model. In this setting, wrote the authors, the standard class probabilities can be easily computed. They demonstrated that energy-based training of the joint distribution improves calibration, robustness, handout-of-distribution detection while also enabling the proposed model to generate samples rivalling the quality of recent GAN approaches. This work improves upon the recently proposed techniques for scaling up the training of energy-based models. It has also been the first to achieve performance rivalling the state-of-the-art in both generative and discriminative learning within one hybrid model.

Reverse-Engineering Deep ReLU Networks

This paper investigated the commonly assumed notion that neural networks cannot be recovered from its outputs, as they depend on its parameters in a highly nonlinear way. The authors claimed that by observing only its output, one could identify the architecture, weights, and biases of an unknown deep ReLU network. By dissecting the set of region boundaries into components associated with particular neurons, the researchers showed that it is possible to recover the weights of neurons and their arrangement within the network.

Cricket Analytics and Predictor

Authors: Suyash Mahajan,  Salma Shaikh, Jash Vora, Gunjan Kandhari,  Rutuja Pawar,

Abstract:   The paper embark on predicting the outcomes of Indian Premier League (IPL) cricket match using a supervised learning approach from a team composition perspective. The study suggests that the relative team strength between the competing teams forms a distinctive feature for predicting the winner. Modeling the team strength boils down to modeling individual player‘s batting and bowling performances, forming the basis of our approach.

Research Methodology: In this paper, two methodologies have been used. MySQL database is used for storing data whereas Java for the GUI. The algorithm used is Clustering Algorithm for prediction. The steps followed are as

• Begin with a decision on the value of k being the number of clusters.
• Put any initial partition that classifies the data into k clusters.
• Take every sample in the sequence; compute its distance from centroid of each of the clusters. If sample is not in the cluster with the closest centroid currently, switch this sample to that cluster and update the centroid of the cluster accepting the new sample and the cluster losing the sample.

For the research paper, read here

2.Real Time Sleep / Drowsiness Detection – Project Report

Author : Roshan Tavhare

Institute : University of Mumbai

Abstract : The main idea behind this project is to develop a nonintrusive system which can detect fatigue of any human and can issue a timely warning. Drivers who do not take regular breaks when driving long distances run a high risk of becoming drowsy a state which they often fail to recognize early enough.

Research Methodology : A training set of labeled facial landmarks on an image. These images are manually labeled, specifying specific (x, y) -coordinates of regions surrounding each facial structure.

• Priors, more specifically, the probability on distance between pairs of input pixels. The pre-trained facial landmark detector inside the dlib library is used to estimate the location of 68 (x, y)-coordinates that map to facial structures on the face.

A Study of Various Text Augmentation Techniques for Relation Classification in Free Text

Authors: Chinmaya Mishra Praveen Kumar and Reddy Kumar Moda,  Syed Saqib Bukhari and Andreas Dengel

Institute: German Research Center for Artificial Intelligence (DFKI), Kaiserslautern, Germany

Abstract: In this paper, the researchers explore various text data augmentation techniques in text space and word embedding space. They studied the effect of various augmented datasets on the efficiency of different deep learning models for relation classification in text.

Research Methodology: The researchers implemented five text data augmentation techniques (Similar word, synonyms, interpolation, extrapolation and random noise method)  and explored the ways in which we could preserve the grammatical and the contextual structures of the sentences while generating new sentences automatically using data augmentation techniques.

Smart Health Monitoring and Management Using Internet of Things, Artificial Intelligence with Cloud Based Processing

Author : Prateek Kaushik

Institute : G D Goenka University, Gurugram

Abstract : This research paper described a personalised smart health monitoring device using wireless sensors and the latest technology.

Research Methodology: Machine learning and Deep Learning techniques are discussed which works as a catalyst to improve the  performance of any health monitor system such supervised machine learning algorithms, unsupervised machine learning algorithms, auto-encoder, convolutional neural network and restricted boltzmann machine .

Internet of Things with BIG DATA Analytics -A Survey

Author : A.Pavithra,  C.Anandhakumar and V.Nithin Meenashisundharam

Institute : Sree Saraswathi Thyagaraja College,

Abstract : This article we discuss about Big data on IoT and how it is interrelated to each other along with the necessity of implementing Big data with IoT and its benefits, job market

Research Methodology : Machine learning, Deep Learning, and Artificial Intelligence are key technologies that are used to provide value-added applications along with IoT and big data in addition to being used in a stand-alone mod.

Single Headed Attention RNN: Stop Thinking With Your Head 

Author: Stephen Merity

In this work of art, the Harvard grad author, Stephen “Smerity” Merity, investigated the current state of NLP, the models being used and other alternate approaches. In this process, he tears down the conventional methods from top to bottom, including etymology.

The author also voices the need for a Moore’s Law for machine learning that encourages a minicomputer future while also announcing his plans on rebuilding the codebase from the ground up both as an educational tool for others and as a strong platform for future work in academia and industry.

EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks

Authors: Mingxing Tan and Quoc V. Le 

In this work, the authors propose a compound scaling method that tells when to increase or decrease depth, height and resolution of a certain network.

Convolutional Neural Networks(CNNs) are at the heart of many machine vision applications. 

EfficientNets are believed to superpass state-of-the-art accuracy with up to 10x better efficiency (smaller and faster).

Deep Double Descent By OpenAI

Authors: Mikhail Belkin, Daniel Hsu, Siyuan Ma, Soumik Mandal

In this paper , an attempt has been made to reconcile classical understanding and modern practice within a unified performance curve. 

The “double descent” curve overtakes the classic U-shaped bias-variance trade-off curve by showing how increasing model capacity beyond the point of interpolation results in improved performance. 

The Lottery Ticket Hypothesis

Authors: Jonathan Frankle, Michael Carbin

Neural network pruning techniques can reduce the parameter counts of trained networks by over 90%, decreasing storage requirements and improving computational performance of inference without compromising accuracy. 

The authors find that a standard pruning technique naturally uncovers subnetworks whose initializations made them capable of training effectively. Based on these results, they introduce the “lottery ticket hypothesis:”

On The Measure Of Intelligence 

Authors: Francois Chollet

This work summarizes and critically assesses the definitions of intelligence and evaluation approaches, while making apparent the historical conceptions of intelligence that have implicitly guided them.

The author, also the creator of keras, introduces a formal definition of intelligence based on Algorithmic Information Theory and using this definition, he also proposes a set of guidelines for what a general AI benchmark should look like. 

Zero-Shot Word Sense Disambiguation Using Sense Definition Embeddings via IISc Bangalore & CMU

Authors: Sawan Kumar, Sharmistha Jat, Karan Saxena and Partha Talukdar

Word Sense Disambiguation (WSD) is a longstanding  but open problem in Natural Language Processing (NLP).  Current supervised WSD methods treat senses as discrete labels  and also resort to predicting the Most-Frequent-Sense (MFS) for words unseen  during training.

The researchers from IISc Bangalore in collaboration with Carnegie Mellon University propose  Extended WSD Incorporating Sense Embeddings (EWISE), a supervised model to perform WSD  by predicting over a continuous sense embedding space as opposed to a discrete label space.

Deep Equilibrium Models 

Authors: Shaojie Bai, J. Zico Kolter and Vladlen Koltun 

Motivated by the observation that the hidden layers of many existing deep sequence models converge towards some fixed point, the researchers at Carnegie Mellon University present a new approach to modeling sequential data through deep equilibrium model (DEQ) models. 

Using this approach, training and prediction in these networks require only constant memory, regardless of the effective “depth” of the network.

IMAGENET-Trained CNNs are Biased Towards Texture

Authors: Robert G, Patricia R, Claudio M, Matthias Bethge, Felix A. W and Wieland B

Convolutional Neural Networks (CNNs) are commonly thought to recognise objects by learning increasingly complex representations of object shapes. The authors in this paper , evaluate CNNs and human observers on images with a texture-shape cue conflict. They show that ImageNet-trained CNNs are strongly biased towards recognising textures rather than shapes, which is in stark contrast to human behavioural evidence.

A Geometric Perspective on Optimal Representations for Reinforcement Learning 

Authors: Marc G. B , Will D , Robert D , Adrien A T , Pablo S C , Nicolas Le R , Dale S, Tor L, Clare L

The authors propose a new perspective on representation learning in reinforcement learning

based on geometric properties of the space of value functions. This work shows that adversarial value functions exhibit interesting structure, and are good auxiliary tasks when learning a representation of an environment. The authors believe this work to open up the possibility of automatically generating auxiliary tasks in deep reinforcement learning.

Weight Agnostic Neural Networks 

Authors: Adam Gaier & David Ha

In this work , the authors explore whether neural network architectures alone, without learning any weight parameters, can encode solutions for a given task. In this paper, they propose a search method for neural network architectures that can already perform a task without any explicit weight training. 

Stand-Alone Self-Attention in Vision Models 

Authors: Prajit Ramachandran, Niki P, Ashish Vaswani,Irwan Bello Anselm Levskaya, Jonathon S

In this work, the Google researchers verified that content-based interactions can serve the vision models . The proposed stand-alone local self-attention layer achieves competitive predictive performance on ImageNet classification and COCO object detection tasks while requiring fewer parameters and floating-point operations than the corresponding convolution baselines. Results show that attention is especially effective in the later parts of the network. 

High-Fidelity Image Generation With Fewer Labels 

Authors: Mario Lucic, Michael Tschannen, Marvin Ritter, Xiaohua Z, Olivier B and Sylvain Gelly 

Modern-day models can produce high quality, close to reality when fed with a vast quantity of labelled data. To solve this large data dependency, researchers from Google released this work , to demonstrate how one can benefit from recent work on self- and semi-supervised learning to outperform the state of the art on both unsupervised ImageNet synthesis, as well as in the conditional setting.

The proposed approach is able to match the sample quality of the current state-of-the-art conditional model BigGAN on ImageNet using only 10% of the labels and outperform it using 20% of the labels.

ALBERT: A Lite BERT for Self-Supervised Learning of Language Representations

Authors: Zhenzhong Lan, Mingda Chen, Sebastian Goodman, Kevin G, Piyush Sharma and Radu S

The authors present two parameter-reduction techniques to lower memory consumption and increase the training speed of BERT and to address the challenges posed by increasing model size and GPU/TPU memory limitations, longer training times, and unexpected model degradation

As a result, this proposed model establishes new state-of-the-art results on the GLUE, RACE, and SQuAD benchmarks while having fewer parameters compared to BERT-large.

GauGANs-Semantic Image Synthesis with Spatially-Adaptive Normalization 

Author: Taesung Park, Ming-Yu Liu, Ting-Chun Wang and Jun-Yan Zhu

Nvidia in collaboration with UC Berkeley and MIT proposed a model which has a spatially-adaptive normalization layer for synthesizing photorealistic images given an input semantic layout.

This model retained visual fidelity and alignment with challenging input layouts while allowing the user to control both semantic and style.

📣 Want to advertise in AIM? Book here

Subscribe to The Belamy: Our Weekly Newsletter

Biggest ai stories, delivered to your inbox every week..

Discover how Cypher 2024 expands to the USA, bridging AI innovation gaps and tackling the challenges of enterprise AI adoption

© Analytics India Magazine Pvt Ltd & AIM Media House LLC 2024

• Terms of use
• Privacy Policy

Subscribe to Our Newsletter

The Belamy, our weekly Newsletter is a rage. Just enter your email below.

Subscribe to Our Youtube channel

Code Switching and Students' Performance in English

• October 2018

• Department of Education of the Philippines

• Cagayan State University

Abstract and Figures

Discover the world's research

• 25+ million members
• 160+ million publication pages
• 2.3+ billion citations

• Samalesu Graham

• Abdeslam Jamai
• Judy M. Saligumba Maed

• Ushas Richel R. Rino

• John Frederick B. Tesoro
• Fernandette Timcang Gamotin

• Recruit researchers
• Join for free
• Login Email Tip: Most researchers use their institutional email address as their ResearchGate login Password Forgot password? Keep me logged in Log in or Continue with Google Welcome back! Please log in. Email · Hint Tip: Most researchers use their institutional email address as their ResearchGate login Password Forgot password? Keep me logged in Log in or Continue with Google No account? Sign up