Who Is Immersion for?: A Critical Analysis of French Immersion Policies

Elitism has been an issue in Canadian French immersion since its inception. This study examines how two racially diverse Ontarian school boards and Ontario French immersion policy, curricula, and other related documents construct and support an elite student within the immersion program. The elite student who emerged from immersion documents is a White, middle-class, English, established resident, mirroring the current demographics of the program. A middle-class bias emerged within the documents due to an assumed wealth, and lack of financial assistance, transportation and promotional materials. The program locations themselves favoured the middle-class. The curricula demonstrated a Eurocentric focus and colonial lens. In the documents of this study, it

far above Core French and Extended French 2 . Students enrolled in immersion programs are perceived to be more engaged in school, have less behavioural issues and are generally regarded as smarter and/or higher achieving than their non-immersion peers (Yoon & Gulson, 2010;Makropoulos, 1998;Wise, 2011). However, these benefits are contested within the field given that social selection can explain the success of the students in the program (Roy, 2015). Middle-class families are more likely to have supports for students in immersion, making them more likely to do well in the program than their lower-class peers. Students who do well in French often have parents who push them, value language learning, and provide them with language learning resources (Makropoulos, 1998). Socalled 'weak' students and English language leaners (ELLs) are discouraged from entering the program or pushed out when they encounter any issues (Mady, 2016). FI programs tend to reproduce social hierarchies around intellectual ability, language abilities, and socioeconomic status. It is not an accident that immersion offers elite public educationparents have a vested interest to keep it that way to give their child an 'edge' (Yoon & Gulson, 2010).
Despite these debated benefits, FI students do have cultural and linguistic capital (Bourdieu & Passeron, 1977;Lazaruk, 2007;Olson, 1983), meaning an advantage that one has over others due to possession of specific cultural elements or belonging to a certain cultural group. Thus, whether or not being a FI graduate gives FI students higher cognitive ability, it does give them an advantage in the workforce because of the cultural capital associated with FI and the linguistic capital of speaking both official languages of Canada. In fact, the reason that most students are enrolled in FI is economic; parents expect that enrolling their child in the program will give them a leg up in the global market (Duchêne, 2016). While school boards have trouble meeting the high demand for FI, attrition remains high (Brown & Bennett, 2017;Durham District School Board, 2021;Sinay et al., 2018). Although, some boards are closing the attrition gap (e.g., Upper Grand District School Board, 2016;. High attrition might in turn contribute to FI's general elite status since it is viewed to be difficult to remain in or graduate from FI.

Decolonizing French Programs
I would be remiss to not address the fact that French programs in Ontario and Canada are essentially colonial. They celebrate the narrative that Canada is an English-French bilingual country (Ontario Ministry of Education, 2013a) completely obscuring and de-valorizing Indigenous languages, peoples, and ways of life. Learning French has been distanced from its colonial past and is now simply seen as a way for people to enter the global market (Duchêne & Heller, 2012). The French viewed imposing their language on others as a way of 'civilizing' them since their language was viewed as 'civil' in comparison to the people who they colonized (Kasuya, 2001). The Canadian educational system, via English and French, has played a major role in Indigenous linguicide; "Language was the vehicle to replace Aboriginal culture with core European concepts and values" (Fontaine, 2017, p. 187). Both languages continue to be held in much higher regard than any other language in Canada. Understanding and reconciling the past and present colonial violence associated with French is a necessity within French programming across the country, one that has not been adequately addressed up to this point. The elite status of FI further alienates and de-valorizes Indigenous languages for which there is no widespread equivalent. Why are we not required to learn a local Indigenous language? What does the continued value we place on colonial languages say about us as a nation?
There are very few studies about Indigenous peoples in French as a second language (FSL) programs (Masson et al., 2021); Lavoie (2015) and Crépeau and Fleuret (2018) are notable exceptions. Côté (2021) also explored the integration of Indigenous perspectives into FI programs in British Columbia. That being said, Indigenous people and issues have not been well researched in FSL.

Students with Special Education Needs 3
Students with special education needs have been excluded from FI programming, even though it has been known for decades these students can succeed in FI (Bruck, 1978). Despite this, only in the past decade did the government begin to support this inclusion (Ontario Ministry of Education, 2013a, 2013b, 2014. Students with special education needs are often discouraged from enrolling and encouraged to leave the program (Arnett & Mady, 2010) under the guise of looking out for their best interests (Wise, 2011). However, it is clear that it is not the language itself that is the issue, but the method of instruction or lack of extra supports, as students continue to struggle when they move to an English stream program (Cobb, 2015;Davis et al., 2019). Support is often only offered within the English program or by those who do not speak French, so students are torn between struggling or getting support in another program. Since FI is an elite and optional program, people see it as enrichment (Mady & Arnett, 2009); if you are struggling in FI, that means you cannot handle it and should attend a 'normal' program. Unjustified exemptions from FSL programs based on special education status were common in the past (Arnett, 2013;Wise, 2011), and it is only the revised curriculum that advocates for the ending of these exemptions (Ontario Ministry of Education, 2013a, 2014. However, students with special education needs are still under-represented in immersion (Sinay et al., 2018), despite the revised curriculum being published nearly a decade ago.

Culturally and Linguistically Diverse Learners 4
Culturally and linguistically diverse learners (CLD) (Prasad, 2012) or English Language Learners (ELL) are likewise excluded from immersion programs across the country. Despite evidence that learning French can support the development of English and one's home language due to the interdependence of languages (first developed in 1979; Cummins, 2014) it took until the last decade for the province to support the inclusion of CLD learners in French programs (Ontario Ministry of Education, 2013a, 2013b, 2014. English is thought to be the more important language for CLD people to learn, so they are excluded from FI for their 'own benefit' (Mady & Masson, 2018;Mady & Turnbull, 2009). Many teachers and administrators believe that students must have a strong command of English to succeed in FI, even though this has been disproven (Cummins, 2014;Mady, 2013Mady, , 2015Mady, , 2016Mady & Arnett, 2015;Roy, 2020). Unfortunately, ELLs are regarded as a problem for teachers to fix (Mady, 2015;Cobb, 2015) and are thought to be unsuited for such a demanding program (Dagenais, 2003;Davis et al., 2019;Mady, 2013). Despite this, CLD families want their children to attend FI because they see it as a way to integrate into Canadian society and because they value language learning (Moore, 2010;Mady, 2013Mady, , 2015Dagenais & Berron, 2001). Moreover, ELLs and immigrants do as well or better than their Canadian-born Anglophone peers in FI (Mady, 2013(Mady, , 2015(Mady, , 2017Knouzi & Mady, 2017). Evidently, there needs to be more emphasis on supporting plurilingual strategies and multilingual learning (e.g., Cummins, 2014).

Socioeconomic Status
Taking into consideration that immersion began because of middle-class parents hoping to maintain (or raise) their children's social class (Heller, 1990), it is unsurprising that middle-and upper-class children dominate the program (Sinay et al., 2018;Olson & Burns, 1983;Makropoulos, 1998Makropoulos, , 2009Makropoulos, , 2010Olson, 1983;Parekh et al., 2011). What is surprising, is that in this age of equity and inclusion, this middle-class phenomenon remains firmly in place. This could be due to program locations in middle-class neighbourhoods (Gossling, 2006;Heller, 2002;Collins & Coleman, 2008), lack of transportation (Makropoulos, 2009;Olson & Burns, 1983) and childcare (Gossling, 2006), and word-ofmouth promotion through middle-class social networks (Mady, 2013;Olson & Burns, 1983). Moreover, people from low socio-economic backgrounds and communities are often associated with negative stereotypes when it comes to intelligence and schooling (Gorski, 2012) and as such, may be guided away from the program, much like CLD learners and students with special education needs are.

Gender
It is not uncommon for more girls than boys to be in my FI classroom, especially in the upper years of the program. The Toronto, Peel, and Durham school boards confirm that there are almost always slightly more girls than boys in FI programs (Brown & Benett, 2017;Durham District School Board, 2021;Sinay et al., 2018). Kissau's (2007) study reveals that girls are encouraged to pursue French programs over boys and that boys consider French to be a feminine subject. This gendering of French could impact girls' and boys' desires to remain in the program, particularly during a time when boys may be trying to assert or affirm their gender.
Gender diversity was not a focus of my 2019 study and is an emerging area of research in the field of FSL. The inherently gendered language and strong stance that many French language teachers take on gender can be a source of exclusion and can be identity non-confirming for transgender and non-binary learners (Spiegelman, 2022). Spiegelman (2022) questions: "To what extent are students with marginalized gender identities granted space for discursive existence in foreign language learning contexts, and what actions are they able to take in order to be and express who they are?" (p. 6). This line of research needs further investigation. Hakeem (2022) advocates for the "queering" of the French language classroom through queer resources and directly engaging with social justice questions and work with students.

Race and Intersectionality
Much of the literature on exclusions within FI centres around special education needs and ELLs. Interestingly, while it is known that White students comprise most of the program (Heller, 1990;Makropoulos, 2010;Olson & Burns, 1983;Sinay, 2010;Sinay et al., 2018), there has yet to be a study that centres race in immersion. The studies that consider race do so as a demographic consideration (Sinay et al., 2018) or simply state White prevalence as common knowledge (Makropoulos, 2010). Additionally, most studies are not intersectional (Crenshaw, 1991;Davis, 2008) in nature and consider only one element of a student's identity (i.e. special education status, home language). Mady (2013Mady ( , 2015 often combines her study of ELLs with immigrants as there is overlap between these two groups of people. There is at least one study looking at ELLs with special education needs, thus combining the two most documented exclusions (Mady & Arnett, 2015). There is a need to understand how different identities intersect with one another in FI programs, and of particular interest to me is how race intersects with other exclusions. Is an upperclass South Asian student more likely to be in the program than a lower-class Canadianborn White one?

Methodology Theoretical Framework
This study situates itself within the framework of critical antiracist theory (CART) (Dei, 2000(Dei, , 2013 and as such it is important to consider how my body moves through the world and how my positionality affects how I question, interpret and come to know (Dei, 2016). CART does not shy away from the personal, as you will have noticed throughout this article and encourages the centring of voices that have been traditionally silenced. Locating the self also reflects feminist standpoint theory (Simpson, 2003). I am a Black woman FI teacher from a low-income family. I am an Anglophone who learned French through Core French in public school. Race was the entry-point of this study, given the importance of speaking to and centring race in CART (Dei, 2016). That being said, it was only an entry-point and the study expanded from race to look at different identities. This study is also framed by critical sociolinguistics which considers how language and power interact with one another (Heller et al., 2018). With a critical sociolinguistics backdrop, this study considers how language is tied to power and how language reveals or obscures different power relations.

Research Questions and Significance
Considering the exclusions outlined in the literature review, along with my own experiences as a FI teacher, I was interested in understanding why these exclusions exist. So, this study considered documents related to and governing FI to see if and how they impacted exclusions in the program. The research questions for this study were: 1. Is there a dominant social construction of the typical French Immersion student in the Ontario French as a Second Language curriculum, Ontario school boards' policies and procedures, and French Immersion promotional materials? a. Who is implicitly and explicitly included or excluded in this construction? b. How are they included or excluded? 2. What other related provincial and regional policies or documents impact this construction?
This study is significant in the field because it combines both the regional and provincial level of policy documents and considers how provincial policies impact regional ones. The findings for this study show how policies impact exclusion and inclusion, pointing to possible ways to revolutionize FI programs. Understanding who is and is not supported is essential in making this elite program more equitable. Lastly, even though it is not the sole focus of this project, this study considers race and racial identity explicitly in FI for the first time.

Data Collection
This study took place during my master's thesis work (Kunnas, 2019). Data was collected at two levels, provincial (Ontario) and regional (two racially diverse Ontario school boards, one secular (OSB) and one Catholic (OCSB) 5 ), using the snowball method (Jorgensen, 1989). The two school boards of this study were chosen for their racial diversity, since race was of particular interest to me. All documents collected were publicly available online and were either website pages, pdfs, or videos. Documents were policies, operational procedures, promotional materials, program descriptions, guidelines, or curricula. The online format was chosen for ease of access, ethics, and to understand what is publicly presented about FI. Given that everything was publicly available through an easy Google search, the findings of the study represent what these institutions wish to present to the world about FI programs. Multiple types of documents were selected because I wanted to have a broad understanding of the context of FI documents (Braun et al., 2011), how different documents influenced others, and what dominant image arose from FI documents.
First, four primary documents were collected and analysed: the FSL curricula (provincial), then the OSB and OSCB FSL/FI policies. Next, provincial documents were collected directly relating to FI or FSL programming. Documents that were referenced within the main curricula were then considered for analysis, and so forth. This was expanded to include equity documents because they were referenced frequently and enriched the ambiguities around the word 'inclusion'. Lastly, regional collection began with the OSB then the OSCB. The regional process mirrored the provincial one, first starting with policies directly related to FSL/FI, then expanding to other referenced policies like transportation and optional attendance, and ending with a collection of equity documents. The collection and analysis processes were iterative with the analysis phase informing the collection phase (Tashakkori et al., 2020). While the provincial collection started before the regional, they were not two clear cut phases.

Data Analysis
This study was a combination critical sociolinguistic analysis (Heller et al., 2018), interpretive policy analysis (Yanow, 2000), and critical discourse analysis (CDA) (Wodak & Meyer, 2016). Policy in this study is understood as a "narrative" (Lejano, 2006, p. 12) demonstrating societal beliefs and power structures (Wodak & Meyer, 2016). Policies are thus capable of telling us who the FI program is conceived for. Critical sociolinguistic analysis considers "social categories" (Heller et al., 2018, p. 107) which were relied upon to see who was and was not represented in FI documents. Interpretive policy analysis was used to understand how groups were represented and interpreted by those who did not write the documents. Lastly, CDA "aims to investigate critically social inequality as it is expressed, constituted, legitimized, and so on, by language use (or in discourse)" (Wodak & Meyer, 2016, p. 10). CDA was used within this study to understand inequities within the documents. While 50 texts were collected, only 28 of the most relevant texts were analysed. The analysis process looked at three main elements: 1) implicit and explicit identity markers, 2) issues of inequity, and 3) critical inclusion. I used a combination of initial coding and descriptive coding to see which topics arose and used in vivo coding for identity markers (Saldaña, 2009). First, I looked specifically for any mention or implication of any identities, especially race, special education, ELLs, gender, and social class. For example, I would note specific and implied identity markers about race (i.e., Black, Indigenous; Mexican, Syrian), home language (i.e., ELLs, English speakers, Farsi, etc.), and gender (i.e., male, female, woman, girl, boy, man; "he or she"). I noted when generic terms were used to refer to identities (i.e., all students, race, sexual orientation, gender, etc.). With this list of identities, I examined to what extent they were included (frequency) and what was being said about them. I considered what identities had not been included in this list and what identities were implied but never stated explicitly implied (i.e., wealth of time and resources implying a certain level of socioeconomic status). The analysis process aimed to see who arose as the dominant student within FI, if anyone. Was it someone specific, or a generic student with no identity markers?
Second, when analysing for equity issues, I subjectively interpreted where there were structural inequities, again focussing on race, special education, gender, and social class. For instance, oftentimes the documents referred to an implied wealth through providing students with extracurricular resources. This raised an equity issue for me since not all families can afford buying the suggested resources. Lastly, critical inclusion became a focus of the analysis process because of the generic focus on 'diversity' and 'inclusion' in the FI documents of this study. If inclusion was just a "non-performative speech act" (Dei, 2016, p. 42) it was considered uncritical. I deemed inclusion critical if the document included considerations to dismantling power structures and was transformative and social action driven. This is based on Dei's (2006) assertion that "inclusion is not bringing people into what already exists; it is making a new space, a better space for everyone".

Limitations
This study is limited because it looks very specifically at two school boards and only one province. Given this, the specificities of the study may not be far-reaching. However, the findings show where inequitable practices arise and were reinforced or maintained within this context, which researchers and stakeholders can look for in their own context. Unfortunately, this study does not consider stakeholder voices. However, it can be argued that stakeholders' voices are present within the text of the documents analysed. Table 1 summarizes the findings in relation to the construction of the typical FI student. The leftmost column shows the identity constructed, the second column offers some precision as to how I came to that identity construction. The last three columns show in which documents this identity construction arose, whether it was present (x), not present (--), or somewhat present (~). A detailed discussion follows this table.

Middle-Class
The documents of this study supported the construction of a middle-class student through an emphasis on parental involvement, an assumed wealth of time and resources, locations of the program, optional attendance policies, lack of transportation, and promotional materials.

Parental Involvement
The provincial documents emphasized that ensuring success for their child meant that parents should attend meetings, parent-teacher interviews, and/or join school council. These activities require a lot of time that not every parent would be able to commit to, especially lower-income parents who may have more rigid schedules, often work long hours, and may not have easy transportation to these events or childcare (Bower & Griffin, 2011;Luet, 2017). Further, it is often the middle-class who get involved in their child's school (Bower & Griffin, 2011). I argue this assumed wealth of time favours middle-class families. It is worth noting that lower-class families do support their child's education in unique and powerful ways (Bower & Griffin, 2011;Luet, 2017).

Assumed Wealth
The province and the OSB assume a certain wealth of families by encouraging them to set up homework areas with various supplies. While lower income families could likely afford supplies like dictionaries and pencils, having a dedicated space to complete homework is a lot harder in smaller homes and apartments. There is an assumption within the documents that parents can afford enriching their child's learning through educational resources, which can be costly. Parents are encouraged by the province to "Arrange travel and exchange programs between communities.; Add a French component to camps and after-school programs.; Offer French films at local theatres" (Ontario Ministry of Education, 2013a, p. 24). These are all time and money costs; arranging travel takes time, effort and money, attending camps and after school programs are expensive, and, organizing film nights takes a lot of effort, time, and knowledge of appropriate French content. The OSB and OSCB barely consider how to best support students in FI, favouring those who have the means to seek out support (i.e., a tutor) or those who are part of (middle-class) networks who share resources within their network (Horvat et al., 2003).

Location of Programs
Both school boards analyzed had their FI programs located almost exclusively in middle-class neighbourhoods, especially at the secondary level. This bars many students from attending FI, unless they live in that school's neighbourhood. Families may also be unaware of FI programs located at different schools since it is not their home school.

Optional Attendance and Transportation
While it is possible to attend FI programs not located in your catchment zone through optional attendance, the spots are limited, and the local community is prioritized. Additionally, transportation is not provided for those who attend through optional attendance (there are some exceptions to this), meaning that children have to commute long distances to get to their school. I argue that transportation to FI programs may be easier for the middle-class who have more money than the lower-class. The middle class may be able to provide rides to their children to and from school, pay for taxis, ride shares, or bus fare, or even buy them a car.

Limited Promotional Materials
Despite its widespread popularity, there was a shocking lack of promotional materials for the FI program, which in and of itself supports the fact that FI is promoted through middle-class networks (Heller, 1990;Mady, 2013). In the few promotional materials present from the OSB, families were presented as middle-class, with time, space, and resources readily available.

Racelessness
The majority of the documents did not show a racial bias, instead defaulting to a raceless identity of 'all' students regardless of race. All students are welcome in FI/FSL programs is repeated throughout all the documents studied. Yet, this in and of itself favours White people since not stating race reinforces the dominant image and normalizes Whiteness (Apple, 1999). There was no consideration to how to critically include racially minoritized students.

Curricular Content
The curricula were the most inclusive and presented several racial and ethnocultural groups, however, the dominant race was still White. Throughout, the most referred to racial identity is White, followed by Indigenous, Black and Latinx. Each year the curricula focus on one cultural/geographical group. Grades 1 -4 focus on local francophone groups; grade 5, Quebec; grade 6 is Canada; grade 8 is France; grade 9 is half French-Canadian, half French-European; grades 11 and 12 consider all francophone groups worldwide. Grades 7 and 10 are the only grades that explicitly focus on racially minoritized areas of la francophonie: the Americas and Caribbean in grade 7 and Africa and Asia in grade 10. While there are mentions to racial minorities during other grades of the curricula, they are uncommon. Two years of dedicated study of racial minority groups seems excellent, however, the fact that three massive regions are crammed into two years would only allow a surface level exploration. Moreover, in grade 10, the literature focus is not even African nor Asian literature, but European French literature. This is the same for grades 11 and 12 which have a global focus for culture, but a European one for literature. It is also worth considering that many students do not continue in FI after grade 8 (Sinay et al., 2018), so they would not even be exposed to the African and Asian curricular focus. Instead of inclusion, the curricula present tokenism.
Lastly, the curricula do include Indigenous peoples at all levels, as the Truth and Reconciliation Commission Calls to Action (2015) demand. However it is done so through a colonial lens and almost entirely in relation to colonization. For example, "role-play a meeting between an individual from a First Nation community and an early European settler" (Ontario Ministry of Education, 2013a, p. 276), or "write a historical report about France's first settlers in Canada and their relationships and alliances with First Nations people" (p. 311). Students approach Indigeneity from a colonial perspective, and it is assumed that colonization was harmonious and not violent (otherwise why would children be role-playing this meeting?). Further, Indigenous relations are placed within the past. Additionally, assimilation camps 6 are only mentioned once, in the grade 12 year. As mentioned within the literature review, there is a pressing need to decolonize the FSL curricula.

White Women Teachers
In its promotional videos, the OSB presents a variety of racial minority students and even parents, challenging a White norm. Yet, all the teachers in the video were White women. Also, the featured family in this video was White. This shows how FI programs may be diverse, but the authorities within the program are not. Seeing ourselves reflected within the program is important (Cherng & Halpin, 2016), whether it be as students, or as teachers. Granted, many teachers are White women (Hoffman, 2017;Picower, 2009), so it is not known if White women are more dominant in FI than in other programs.

Race and Income
Given that the middle-class identity emerged clearly within all documents, a White identity emerged as well. While I cannot say that all White people are middle-or upperclass, they are more likely to earn more than racial minorities (Herring & Henderson, 2016;Traub et al., 2017). Because of this, it is safe to assume that the middle-class construction intersects with the White racial make-up of FI. To what extent, however, is harder to know.

English Speaking
The documents in the study all showed an assumption that families had at least a functional knowledge of English. The curricula made comparisons between English and French multiple times, despite saying that students' home language should be valued and encouraged. Notably, the OSCB had a feature on their website that allowed all website pages to be translated via Google Translate, making it the most accessible of the three groups studied. The OSB had one document about supporting ELLs that was professionally translated into several common languages within their school board. However, the regional FI and the provincial support guide for parents and families were only available in English.

Established Resident
At the regional level, the application process favoured established residents over newcomers. This is because of the strict deadlines to enter the program. Anyone is able to apply to FI. The OSB even guarantees placement in a FI school, as long as the application is received on time. However, this firm deadline means that recent immigrants, or even those who did not know about FI until after the application deadline, may not be able to access the program. The OSCB deadline is even earlier since FI starts in junior kindergarten, not senior kindergarten like the OSB. Mady (2013) highlights how many immigrants would have attended the program if they knew of it in time.

Basic Knowledge of French
Parents were encouraged to buy French resources for their children, but at least a basic knowledge of French is necessary to know what is appropriate. They are likewise encouraged to help children with their homework, which is in French. Despite the provincial documents claiming that parents do not need to speak French, it is evident that some knowledge of it would be helpful. Moreover, most documents were available in English and French, but not most other languages (as discussed above), showing that if you speak either language, you are more likely to be able to navigate the program. Table 2 presents a summary of the contradictions in the construction of a typical FI student. It follows the same format as above: the dominant construction is listed on the left, the second column offers some specificity, and the last three columns represent the documents where this construction was present (x), not present (--) or somewhat present (⁓). -x = present ; --= not present ; ⁓ = somewhat present

Gender
While girls are often over-represented in FI, especially at the senior levels, the documents of this study did not reveal a bias toward girls, but instead a generic genderlessness and sometimes a prioritization of boys. Activities that are more associated with boys are highlighted, like reading comic books (Andera, 2007;Moeller, 2011). The people that the curricula reference are overwhelmingly men; elementary curricula referenced 10 women and 30 men, and secondary curricula reference 9 women and 27 men. As Kissau (2007) concludes, girls are more encouraged in French and boys find French to be feminine, which may be why men are emphasized within FSL curricula, to better engage this under-represented demographic. The documents of this study did not reflect the overrepresentation of girls in FI. Further, the documents all reinforced the gender binary of "he" and "she", with only one document of those studied defaulting to the gender-neutral "they".

Special Education Needs
The literature shows that students with special education needs are uncommon in the program, however, the documents do not support this. The provincial documents explicitly called for inclusion of students with special education needs within the curricula and its companion document. Applied FI courses (versus academic) 7 have also been included in the FI curricula for the first time, and considering that many students with special education needs take applied courses (Wise, 2011), this is an inclusive move 8 . The OSB dissuades exemptions based on special education needs in its FI policy. That being said, the OSB and OSCB do have ways to exempt students based on special education needs, and the only time FI is mentioned in the OSB special education document is in relation to its suitability for 'gifted' students. While these facts contradict inclusion, the majority of documents called for inclusion of all, particularly those with special education needs.

English Language Learners
Here, there was a stronger contradiction because the documents advocate for inclusion of ELLs while also reinforcing the inclusion of English speakers above all others (as presented above). The provincial documents highly advocated for the inclusion of CLD students throughout FSL programs and pointed to the fact that they can thrive in them, however, very few ways to support ELLs were communicated. At the regional level, both school boards had far less support for CLD learners. The OSB claimed that ELLs can access FI in the same way as all others, ignoring possible barriers to them, such as not knowing about the program (Mady, 2013) or not having a strong understanding of English (Bower & Griffin, 2011). The OSB includes people with non-Canadian accents in their FI promotional videos, signalling that CLD learners are welcome and enrolled in the program. The OSCB does not mention ELLs at all in the FSL/FI policies. Both boards do have supporting documents specifically for ELLs, but they do not consider ELLs in FI specifically. At the time of this study, the provincial call for inclusion had not yet been taken up at the regional level.

Findings: Equity Documents
Given the lack of explicit information and generic inclusion in the provincial and regional FSL/FI documents, equity documents were considered. Equity documents mandated inclusion across all schools in the province and the school boards investigated. The provincial equity documents and the OSB's equity documents included intersectionality and named explicit identities, whereas the OSCB did neither. While these documents do include race, special education needs, and ELLs, they do not consider how to address economic inequalities, which has been a major concern in FI for decades (Olson & Burns, 1983). Further, these documents called for inclusion, but did not offer many practical ways of achieving it. As mentioned above, to be considered critical inclusion, we must move beyond simply including others into broken and inequitable systems and instead consider how to challenge and transform systems to make a place where everyone is equitably supported. The OSB had the most critical inclusion in their equity documents and advocated explicitly for changing power structures and reallocating funding to better serve poor communities. Overall, these documents revealed that inclusion was mostly generic and not critical, meaning it focused on 'all people' without considering how power structures needed to be transformed nor how to transform them.

Discussion
In this section, I respond directly to the research questions and discuss the results of this study. First: is there a dominant social construction of the typical French Immersion student in the Ontario French as a Second Language curriculum, Ontario school boards' policies and procedures, and French Immersion promotional materials? Who is implicitly and explicitly included in this construction? How are they included? The answer to this question is nuanced and has some contradictions. Nonetheless, a dominant construction of the typical FI student does emerge: a middle-class, White, established resident with knowledge of English and to a lesser extent, French. Considering directly my entry-point of race, the dominant identity of a raceless student emerged, reinforcing the White norm. The documents revealed above all, a bias toward the middle-class. Some less dominant identities were present within the documents of the study with contradictions, including: students with special education needs, ELLs, and girls. I do not include these in the dominant construction as they occurred almost exclusively at the provincial level.
The documents exclude many different identities, particularly racial and ethnic groups, religions, physical ability, social class, sexuality, and gender. When these groups are included, it is done through listing identities without specificity; for example, the document says "race", "gender", "faith" instead of "Black", "gender non-conforming", "Muslim", etc. There is little consideration of how to support marginalized groups within FI programs. Instead, the documents frequently recite the need to 'include everyone'; empty words that do nothing. Barriers to traditionally marginalized communities need to be addressed within FI/FSL documents with practical steps of how to transform the program to include minoritized and traditionally excluded groups. As presented above, there were little to no references to nor supports for marginalized and minoritized groups, reinforcing the dominant identity. For there to be inclusion, it must be critical, and it must not be simply adding minoritized people to the program, but instead creating a program that better supports everyone.
The second research question was: What other related provincial and regional policies or documents impact this construction? This is when equity documents were analyzed. Considering that equity documents did not mandate critical equity or inclusion and instead called for a generic inclusion, it is unsurprising that FI documents did the same. As equity documents govern inclusion within the province and the board, there is a need for them to be critical and transformative, otherwise FI programs will do the bare minimum required by these documents, as was the case with the OSCB equity documents which were anything but inclusive. These documents instead serve as an auditable trail to show that institutions are addressing diversity and inclusion instead of actually making institutions address diversity and inclusion (Ahmed, 2006). The mere presence of equity documents does not mean anything, particularly when they do not promote transformative action, as CART calls for (Dei, 2013). Transformation is required not only on an individual level with people's attitudes and actions, but on an institutional level. Transformative action in education involves targeting and changing inequitable practices, empowering oppressed communities and allies to tackle oppression, and equity-based pedagogy (Potvin & Carr, 2008). Moreover, equity documents need to be more practical and offer steps to a more equitable future, otherwise teachers will parrot ideas of inclusion and continue to reproduce oppressive hegemonies. While equity documents do contribute to the construction of FI students, since they are referenced within FI/FSL documents, they also are exterior documents that do not apply specifically to the immersion program. Requiring stakeholders to consult another document instead of embedding equity directly into FI shows how it is not a direct concern of FI programs. The status quo is maintained through omitting equity concerns from FI documents.
These results were not surprising given the current demographics of the immersion program, discussed in the literature review. It is a program for middle-class people, populated by middle-class people. CLD learners and immigrants are excluded along with students with special education needs, although this is changing. It was interesting to note how racelessness and curricular content implicitly contributed to the construction of a White student within immersion programs. Whiteness was celebrated through explicit study and focus on White and Eurocentric people and ideologies. Whiteness was normalized through the omission of race throughout the documents, reinforcing the White norm. Policies and documents of this study have an evident and purposeful focus on Whiteness and Eurocentrism. The White over-representation in immersion is less surprising based on this finding.
Considering the findings from this study, I conclude that these documents do contribute to the inequitable student population in FI programs in Ontario schools. While policies and related documents are expanding to include more and more people in immersion, there are still very clear systemic barriers such as lack of supports and transportation, program locations in wealthy neighbourhoods, and an uncritical understanding of equity. Equity documents are helping inclusion, but their direct impact on the inequitable policies in school boards are unknown. These documents are not the sole barrier to equitable inclusion in FI, but they do limit inclusion in many ways.

Conclusion: The Future of Immersion
This study has revealed how regional and provincial documents contribute to exclusion and maintenance of the elite student in FI programs in Ontario. The student who emerged from these documents was White, middle-class, English speaking, established resident with parents who could understand some French. FI is a program that offers significant capital to those who attend and graduate from it. As a publicly funded program, immersion should not only benefit a specific elite few, but instead benefit as many diverse students as possible. Policies must be adapted to make immersion more accessible to a broader student population, however, there are many systemic and financial barriers to doing so. Funding must be allocated toward making the program more inclusive and equitable; simply creating equity documents and hoping is not enough. Particularly when considering the over-presence of middle-class students in FI, this is a class issue. Moreover, for FSL and FI programs to continue in Ontario, they must be more inclusive of CLD learners who represent more and more of the student population. Policies need to be more reflective of these considerations, particularly the regional policies of this study.
Policymakers, program coordinators, and administrators should consider the following items to make FI more equitable and less homogenous: • More entry-points into the program • More critical and less Eurocentric / colonial curricula • Expanding and / or moving FI school locations • Provide transportation to all students • Create an equitable enrolment model where more diverse students are prioritized • Offer free resources and tutoring to students • Provide supports for students with special education needs in FI and make them aware of these supports • Provide more resources and training for teachers to support diverse learners in FI • Promote the program in poorer neighbourhoods and to newcomers, ELLs, and those with special education needs While policies and documents have an important role in determining what happens in classrooms and school boards across the province, they are not the be all and end all. How policies are enacted is key in understanding how they function in reality (Braun et al., 2011), and in this case, how inequities persist and are subverted. This is not to undermine the importance of this study, but to instead point to potential future research to see how principals, teachers, and families interpret, experience, confront, and enact policies. As this study revealed, there is still much to be investigated about FI programs, including questions of race, gender, sexuality, social class, and more. It is past time to diversify and decolonize immersion education so that all students can reap its benefits.
Correspondence should be directed to Marika Kunnas. Email: rmkunnas@yorku.ca Notes 1 Here, Francophone means a person whose first or dominant language is French.
2 Core French is the main French language learning program in Ontario wherein students spend approximately 600 hours of their elementary career learning French and must take one high school credit of French.
Extended French is often referred to among teachers as 'immersion lite' wherein students spend less than ½ of their day in French from grade 4 or 7 onward (instead of ½ of their day or more, like in FI).
3 Special education needs are very diverse and come range from mental to physical disabilities. Students with special education needs may have an Individual Education Plan (IEP), but it is not a requirement. Students with special education needs are also often referred to as students with learning difficulties or disabilities. 4 ELL is the most common term for this group of people, appearing in most studies and within official government documents. Considering notions of plurilingualism and the transfer between languages, ELL may not be the most appropriate term. For this reason, I alternate between ELL and CLD (culturally linguistic and diverse) learners. I avoid 'allophone' as a term as it implies 'Other' and technically would be anyone who does not speak an Indigenous language (see Prasad, 2012). 5 The school boards chosen have been obscured to maintain anonymity. 6 Assimilation camps are also referred to as residential schools. 7 Ontario is in the process of phasing out academic and applied courses. Academic courses are more theoretical and populated with higher achieving students. Applied courses are more practical and populated with lower achieving students, and often racialized students and students with special education needs (James & Turner, 2017).