Code-switching Patterns in Bilingual Children When Speaking with Monolingual Peers


Wahiba Tamazi Mongiello

The City University of New York: Hunter College



Code-switching is a linguistic phenomenon that is evident from as early as two years of age. When one party cannot understand the other, there is often a switch in language in order to fill lexical gaps. The proposed study will include 60 French-English bilingual children from 3;6 to 4;6 years of age. Students will be paired with a monolingual peer and video recorded for a 30 minute free play session. Ten minutes of the recording will be transcribed and encoded for frequencies of code changes. Based on previous literature providing evidence of code-switching in the presence of an adult population, it is hypothesized here that children will exhibit less code-switching with same-age peers


Code-switching Patterns in Bilingual Children When Speaking with Monolingual Peers

Code-switching is a mechanism utilized by bilingual children that allows them to fill gaps in either their own lexical knowledge or that of their addressee. Subcategories of this phenomenon include code-mixing (replacing single words within speech) and code-changing (alternating use of both languages) (Vu, Bailey, & Howes, 2010). Some of the current literature demonstrates that bilingual children are able to appropriately differentiate which language to use based upon who they are speaking with (Tare & Gelman, 2010). Other studies have shown that when speaking with their parents, bilingual children will speak the mother’s native language more with the mother and conversely with the father (Comeau, Genesee, & Mendelson, 2007).  The latter had found that French and English speaking two and three year olds are not only sensitive to their interlocutor’s lingual capacities, but that they employ several methods in attempt to have their interlocutor understand them. These studies are further explained.

Comeau, Genesee, and Mendelson (2007) led a study in the Montreal area with English-French bilingual children from ages 2;3 to 3;7. Their research aimed to assess breakdowns in communication and how bilingual children attempted to resolve them. Five clarification requests were used by the experimenter (i.e. What is this? I don’t understand.) when the child spoke the language their interlocutor did not understand. The ways in which these children responded to requests were classified as: Appropriate Language Change (ALC), Inappropriate Language Change (ILC), Reformulation (REF), Repetition (REP), and No Change (NC). The results showed that all three-year-olds made ALC while five of the eight two-year-olds made ALC. Their study supported previous evidence that young bilingual children have the ability to recognize a language barrier as reason for miscommunication and repair accordingly.

A similar study conducted by Vu, Bailey, and Howes (2010) looked at 92 bilingual Mexican-heritage children and their code-switching patterns. They utilized the MacArthur Story Stem Battery (MSSB) which obtained children’s portrayals of child-teacher and child-parent relationships in the child’s language of choice. Stories were enacted by the experimenter using dolls to represent a child-mother story and a child-teacher story. After each presentation, the examiner asked the child to show what would happen next. The language the child predominantly narrated in was designated as their primary language. Code-switches to the other language were  coded separately for child-mother stories and child-teacher stories. Fifty-four of the 92 children engaged in any code-switching behavior. With regard to sociopragmatic reasoning, the researchers found that while code-mixing was more frequent amongst all groups, code-changing was more prevalent in Spanish-dominant children’s child-teacher narratives. This finding supported one of the authors’ main hypotheses that code-switching patterns would be affected by situational context. In this case, switches to English were made based upon narrating a story taking place in an institutional setting (i.e. school).

In addition to focusing on sociopragmatic differentiation in children’s code-switching patterns, some research has tried to relate this to theory of mind. Tare and Gelman  (2010) examined bilingual children’s code-switching patterns in object naming and free play while relating it to theory of mind. Twenty-eight English-Marathi bilingual children ages 2;7 to 4;11 were assessed for such patterns. The authors utilized an object naming task where children would be asked to name six objects in English and six in Marathi. If the child answered in a language other than what was presented, they were asked to clarify, much like in Comeau et al.’s study (2007). A free play task then followed utilizing toy sets where the experimenter played with the child in a three minute recorded session in the target language. Theory of mind was measured using Wellman and Liu’s task. The results of this study demonstrated that children are able to code-switch more sophisticatedly than previous studies done with a slightly younger age group.

Children’s sociopragmatic competencies can be assessed by observing their dyadic speech patterns. It seems that research in the field has focused on child-parent or child-experimenter sociolinguistic relationships, but there seems to be a lack of research regarding children’s same age peer relationships. What happens when monolingual children ask bilingual children questions in the same way that an experimenter would? Does the same theoretical framework apply? One may question why there would be a chance of difference at all, but the answer may lie in interlocutor relationship. When a child speaks with an adult, there may be a pattern of authority that is adhered to. Vu et al. (2010) addressed the fact that when contextual situations such as physical setting and personnel change, so does a child’s code-switching behavior. Investigation of code-switching tendencies among preschool age children is particularly important because it gives us a clue as to how children cross linguistic and cultural barriers outside of the home- namely, in schools. With an ever growing population of immigrant families in New York, the need for social understanding becomes more evident and starts with understanding a language.

The purpose of the current study will be to determine the level of code-switching that bilingual children exhibit with their monolingual peers. For example, if an English bilingual, but primarily French speaking three-year-old meets an English monolingual peer, will that child recognize and try to fix the fact that their new acquaintance cannot understand them? Will they revert to speaking in the monolingual’s language immediately or will they try to make them understand their dominant language? Are language repairs quicker or similar to those of demonstrated results in child-adult studies.



Participants will include 60 English-French bilingual children from ages 3;6 to 4;6. Monolingual participants used as peer interlocutors will include 60 French speaking children in the same age group.  Children will be from all socioeconomic standings and not yet enrolled in any preschool/kindergarten classes. The previous literature has looked at children in this age group. It may also be useful as a constant for future replications involving theory of mind, which generally evolves around four years old. Children will be recruited from Francophone populated areas in New York City. Recruiting will include giving pediatricians flyers, scouting in nearby parks, and attending events related to infants and parents.


            Bilingual children will be given the Peabody Picture Vocabulary Test (PPVT) to measure their lexical knowledge as other studies previously have. This test “measures an individual’s receptive vocabulary for Standard American English and provides, at the same time, a quick estimate of verbal ability or scholastic aptitude.” (Dunn and Dunn, 1981) This test was chosen over the MCDI since it was designed for ages two and up, while the MCDI measures from eight to 30 months of age. The French adaptation, Échelle de vocabulaire en images Peabody (EVIP), will be administered in a separate session as to not tire participants. The exam will last roughly half an hour. French monolinguals will also be given the EVIP to test for basic comprehension.

Bilingual children will come to the laboratory over the course of three sessions: the first for the PPVT, the second for the EVIP, and the third for a free play observation task. This will be done to prevent exhausting the child, as the test may take around 30 minutes. The PPVT will be administered in English as the EVIP will be administered in French. Monolingual children will come for two sessions: the EVIP and the free play observation task.

For the free play task, children will come to the laboratory where they will be placed into a room with appropriate setup. Children will bring one of their own toys and also will have objects available for play on a play table where they will be seated. The objects will include some children’s books and a house set with figurines of family members. Two chairs will also be placed approximately seven feet away for the parents of the children. Parents will stay in the room for five minutes until children are acquainted with each other. If the parent wishes to stay in the room or the child is not cooperative, the parent must stay silent while the children are interacting. There will be a hidden video camera that will be recording the children that the parents will be aware of. The free play session will last 30 minutes in total.

Transcription will be done by graduate students who are bilingual in French and English. As in other studies, Cohen’s kappa coefficient for interrater reliability will be calculated to ensure proper encoding. The first five minutes of the recording will not be transcribed, as children are still getting acquainted with each other during this period; the following 10 minutes will be recorded. The operational definition for code-switching will include the following: (a) anytime a bilingual child corrects him/herself by switching to the interlocutor’s language when asked a question directly, and (b) anytime a child speaks to the peer and switches to the appropriate language for two or more words.


            This study could potentially yield some enlightening results about children’s bilingual speech patterns. As previous studies have shown, children are able to recognize breakdowns in communication and the reasons for them at a very young age. (Comeau et al., 2007) Those studies, however, have been with adult interlocutors. Now that children’s peers are being introduced, there could be implications for bilingual curricula as a form of early social intervention. Dual language programs already exist in New York City and have recently been expanded. If evidence from this study shows that bilingual children make the same appropriate language changes with peers their age as they do with adults, programs for education may continue to see expansion and retention.

The current proposal is not without its limitations. For one, testing between only two children versus within a group may not serve as an accurate representation of real life scenarios, namely the classroom. Child cooperation is also an issue of concern. If children are not cooperative, recordings and transcriptions may not be obtained and sample size affected. The current study also does not assess monolinguals, which would further emphasize the bilingual advantage if studies showed a monolingual disadvantage in comprehension attempts.

Future studies may benefit from several alterations to this proposal. It may be worth looking at bilingual sensitivity as a function of theory of mind, since other researchers have purported connection between the two. (Goetz, 2003; Rubio-Fernández & Glucksberg, 2012; Rubio-Fernández, 2016) Are children sensitive to communicative breakdowns even before their theory of mind has evolved? The study would also benefit if another qualitative dimension was added, particularly concerning children’s motives for code-switching. Do they realize when they do it? Do they have a reason for why they do it? They are valid questions that, hopefully, this proposal will set the framework for.



Comeau, L., Genesee, F., & Mendelson, M. (2007). Bilingual children’s repairs of breakdowns in communication. Journal Of Child Language, 34(1), 159-174. doi:10.1017/S0305000906007690

Dunn, M., & Dunn, L. M. (1981). Peabody Picture Vocabulary Test—Revised. Circle Pines, MN: AGS.

Goetz, P. J. (2003). The effects of bilingualism on theory of mind development. Bilingualism: Language and Cognition, 6(01), 1-15.

Rubio-Fernández, P. (2016). Why are bilinguals better than monolinguals at false-belief tasks?. Psychonomic Bulletin & Review, 1-12.

Rubio-Fernández, P., & Glucksberg, S. (2012). Reasoning about other people’s beliefs: bilinguals have an advantage. Journal of Experimental Psychology: Learning, Memory, and Cognition, 38(1), 211.

Tare, M., & Gelman, S. A. (2010). Can you say it another way? Cognitive factors in bilingual children’s pragmatic language skills. Journal of Cognition and Development, 11(2), 137-158.

Vu, Jennifer A., Bailey, Alison L., & Howes, Carollee. (2010). Early Cases of CodeSwitching in Mexican-Heritage Children: Linguistic and Sociopragmatic Considerations, Bilingual Research Journal. The Journal of the National Association for Bilingual Education, 33:2, 200-219, DOI: 10.1080/15235882.2010.502798