“Miss, can you speak English?”: raciolinguistic ideologies and language oppression in initial teacher education

Abstract Racism is pervasive within the lives of racially minoritised pre-service teachers in England, but little work has explored how perceptions about language feature here. Based on interviews and workshops with 26 racially minoritised pre-service teachers, I describe their experiences of language oppression whilst on school experience placements, where they were instructed by mentors to modify, flatten, and completely abandon their ways of talking if they were to be perceived as legitimate. I show how language oppression gets justified by mentors in reference to national policy, and how perceptions about the quality of speech are ideologically anchored to perceptions about the quality of teaching. I show how language oppression often materialises under seemingly benevolent and humanitarian guises, but inevitably maintains the raciolinguistic status quo because it instructs racialised teachers to adapt their speech so that it appropriates whiteness. I argue that language oppression is a key reason why England continues to fail to retain racially marginalised teachers.


Introduction
Racist and classist ideologies about language are pervasive in England's schools, universities and broader society (e.g. Cushing and Snell 2022;Drummond 2018;McKenzie 2015;Sharma, Levon, and Ye 2022). This article explores the affective dimensions of language ideologies in the lives of racially minoritised teachers enrolled on university-based initial teacher education (ITE) courses. I conceptualise this as language oppression -a form of violence, hostility and stigma based on how people are perceived to violate linguistic expectations defined by normative whiteness. Reproducing European colonial logics in which colonised communities were deemed to display linguistic deficiencies, language oppression can intimidate people into modifying their use of language or abandoning it completely, often under the guise that is in the interests of the oppressed to do so (Heller, and McElhinny 2017;Taff et al. 2018).
In this article I argue that language oppression is a significant yet overlooked reason that England continues in its failures to retain racially minoritised teachers. This is symptomatic of institutional racism within schools (e.g. Bradbury, Tereshchenko, and Mills 2022) and attempts by the state to actively derail and water down anti-racist efforts in university-based ITE (Smith and Lander 2023). As Smith and Lander (2023) show, these attempts are taking place amidst a context where people of colour represent the vast majority of applicants to ITE programmes but are in the minority when these programmes end, suggesting that institutional racism is a major reason that white teachers continue to dominate classrooms. This context provides the backdrop for this study, where I argue that racially minoritised pre-service teachers are seen and heard by the state as inadequate -and ultimately disposable -because of how they are perceived to display linguistic deficiencies. I focus on experiences of language oppression in terms of racially marginalised pre-service teachers being instructed by senior colleagues to modify, flatten, and in some cases, completely abandon their inherently heterogenous language practices. This analysis stems from a raciolinguistic perspective (Rosa and Flores 2017), a way of thinking about the co-construction of language and race which centres intersectional oppression, coloniality and white supremacy in the production of inequality. Raciolinguistic ideologies are key to this, denoting deeply-embedded beliefs that racialised speakers display linguistic defects, even when their language practices resemble those produced by powerful white speakers. It is my hope that this research contributes to ongoing anti-racist efforts in teacher education, by placing a focus on a dimension of oppression which has been previously overlooked: language.

Race, teacher education and language oppression
The majority of postgraduate university-based ITE programmes in the UK are 10 months long, with students typically completing two short periods of teaching at university and two longer placements in schools, where their progress is assessed by school-based mentors and university-based academics. UK-focused research has long demonstrated that racism is pervasive in the lives of many minoritised pre-service teachers (e.g. Bradbury, Tereshchenko, and Mills 2022;Bhopal 2015;Bhopal and Rhamie 2014;Gabi et al. 2023;Lander 2014;Poku 2022;Wilkins and Lall 2011). Some of these studies discuss language, particularly accent, as one of the intersection points where racism materialises. For example, Bhopal and Rhamie (2014), Bradbury, Tereshchenko, and Mills (2022) and Poku (2022) describe racialised pre-service teachers' reported feelings of being discriminated against due to comments from white colleagues about their purportedly unsuitable accent. Such studies take an intersectional approach to understanding experiences of racism, with social class also bearing influence on how racially minoritised teachers are perceived to be inadequate.
Alex Baratta's research (2018) describes the experiences that many pre-service teachers face in relation to accent policing, particularly those from working-class backgrounds. Baratta's work pays very little attention to race, failing to draw connections between individual acts of accent discrimination and broader structures of institutional racism and white supremacy within ITE. Of equal concerns are Baratta's recommendations that 'accent standards' for teachers should be developed, that marginalised teachers should engage in code-switching as a 'conscious strategy' , and that 'the need to monitor one's own speech may be regarded as a mere nuisance at times, but worth the effort' (2018: 91). A raciolinguistic perspective outrightly rejects these recommendations, instead arguing that this simply places responsibility on racialised speakers to modify their voice towards homogenous white standards and reproduces the message that their own, natural language practices should be left at the classroom door (e.g. Baker-Bell 2020).
Other, more productive initiatives in the UK have also brought public attention to the intersections of race, class and accent discrimination (such as Accent Bias Britain and The Accentism Project). However, these projects have not discussed the experiences of pre-service teachers. My aim here then, is to adopt a raciolinguistic perspective which challenges the dominant narrative that marginalised pre-service teachers must modify their ways of talking if they are to be perceived as legitimate professionals. This is done by interrogating what Rosa and Flores (2017) conceptualise as the white listening subject. The prototypical white listening subject is not simply a malicous individual, but refers to an ideological position and racialised mode of language perception which can be inhabited and enacted by anyone, irrespective of their racialised identity, and can exist in policies, assessments, and other instruments of linguistic surveillance. An analytical focus on the white listening subject shifts attention away from the purportedly deficient language practices of racialised speakers, and instead interrogates how they are perceived, categorised and policed.

Raciolinguistic ideologies and the teacher education policy architecture
Given my focus on structures rather than individual acts of language policing, I now describe the broader policy architecture of ITE in England and how this is designed to maintain and reproduce raciolinguistic ideologies. In previous work (Cushing 2023) I showed how ITE policy has long positioned pre-service teachers who use non-standardised language practices as deviant, impure, and in violation of professional requirements. From deficit discourses about language in cross-partisan policy texts such as the 1921 Newbolt Report and the 1981 Scarman Report, through to the current version of the 2011 Teachers' Standards, racialised and classed language oppression is written into teacher education policy as a design feature. These long histories of language oppression must be read within a context of anti-racist and decolonial efforts in ITE, led by community activists, unions, and grassroots movements (see Gabi et al. 2023).
A dense web of national policies introduced since the standards-based reforms of 2010 have further embedded raciolinguistic ideologies into schools and ITE (see Cushing 2023). This policy assemblage includes the Teachers' Standards, which instructs pre-service teachers that they must speak in 'correct ' , 'articulate' standard English (DfE 2011: 11), and the Core Content Framework, which instructs pre-service teachers that they must use 'high-quality oral language' and 'high-utility vocabulary' (DfE 2019: 15). A raciolinguistic perspective rejects the premise that these are empirically definable linguistic categories and instead construes them as racialised modes of perception that are dichotomised with the home language practices of racially minoritised speakers (see Flores 2020). Ofsted sit at the centre of the ITE policy assemblage, as institutional language police who inspect ITE providers and monitor their alignment with documents such as the Teachers' Standards and the Core Content Framework. For nearly 200 years, the inspectorate has acted as an institutional agent of language oppression and an archetypal example of the white listening subject (Cushing and Snell 2022). In Cushing and Snell (2022), we showed how the inspectorate has consistently produced deficit judgements about the non-standardised speech of students and teachers, with the most hostile comments reserved for institutions serving working-class and racialised communities. These acts of language oppression are carried out by an institution marked by its whiteness, with Ofsted's latest statistics showing that 92% of inspectors identify as white (Ofsted 2020a). Ofsted continue to engage in language oppression today, subscribing to a pathological, colonial and missionary narrative which claims that racialised communities require remediation if they are to achieve social mobility. The state grants Ofsted power to punish ITE providers if they are deemed to resist this narrative, categorising them as non-compliant if they challenge the raciolinguistic status quo.
In the discussion that follows then, I argue that language oppression materialises in the lives of pre-service teachers not as individual acts of racism, but is a design feature of an ITE policy architecture underpinned by white supremacy (see also Gillborn 2005). Put simply, my argument is that language oppression is legitimised by the state and one way in which the state maintains the whiteness of teacher education (see also Sriprakash, Rudolph, and Gerrard 2022). As Smith (2023) shows, there are zero references to race and racism in England's contemporary ITE policy assemblage, an occlusion and denial which delegitimises and curtails discussions of race from taking place in ITE. Finally, this policy architecture is one in which complex power networks are at play: pre-service teachers are under pressure to show that they are complicit with these policies; school-based mentors are required to evaluate this conformity, and university-based teacher educators must demonstrate curricula and pedagogical alignment with Ofsted's regime of racialised linguistic surveillance. It is within this assemblage that language oppression is endorsed by the state.

This research
The primary objective of this study was to document how language oppression materialised in the lives of racially minoritised pre-service teachers, focusing on their experiences whilst on school experience placements. This was fulfilled through 26 semi-structured interviews which took place between 2021-2022. Ethical approval was granted by my institution (Project ID 30517), and all participants provided informed consent. Institutional ethical guidelines were followed in all aspects of the work.
Critical Race Theory (CRT) informed the design of the methodology, in steering the interview questions and the analysis of data. CRT was chosen because of how it centres racialisation and structural racism in its methodologies, including how people of colour are perceived to display linguistic shortcomings (e.g. Baker-Bell 2020; Kroskrity 2021). It insists on an intersectional approach which exposes the intricate fabrics of marginalisation and the active crafting of inequality (Gillborn 2005;Rollock 2012). As in previous work on language discrimination which draws on combined insights from CRT and a raciolinguistic perspective (e.g. Kubota et al. 2021), my interest is in exposing how colonially constructed linguistic standards reinforce raciolinguistic ideologies concerned with 'speaking properly' and maintain normative whiteness.
Participants were mainly recruited through my own experiences of working with pre-service teachers on ITE courses. The criteria for inclusion were that participants identified as being a person of colour, that they were currently or previously enrolled on a ITE programme, and that they had received comments about their spoken language during as part of this. All participants self-reported as being confident speakers of English, and Table 1 includes details about each participant deemed to be relevant for this study.
Interviews followed a loosely organised guide, where participants were invited to recount their experiences of language oppression whilst on their ITE course. Most of these recounts centred on experiences at school placements. Discussions touched on participants' feelings and responses to language oppression, the perceived relations between language, race and class, the school-based mentors' reasons for justifying language oppression, and the responses they received when they challenged it.
Interviews lasted between 45 and 60 minutes. These were audio recorded and professionally transcribed, with transcriptions checked back for accuracy whilst I designed an initial set of themes. All participants were invited to attend an online discussion where I shared the emerging themes and my initial interpretations of these. 21 participants attended this, bringing a level of nuance which simply would not have been possible without their additional input. For example, multiple participants talked about the challenges and denials from mentors they experienced when raising their concerns, especially when participants named race, class and racism in these conversations. As such, the ways that racially marginalised teachers had their concerns about linguistic dismissed by schools is a key part of the discussion that follows. All participants were sent a draft version of this article and invited to give feedback. I took my duty of care seriously here, regularly checking in on participants and asking whether they continued to provide their consent. None of the participants withdrew their consent, and all spoke of their desire to see the work be published and shared. Indeed, participants talked of their commitments to social and racial justice, and we spoke about how language ideologies played a crucial role here (see also Gandolfi and Mills 2022). I hope that this work contributes to these efforts.

Language at the intersections, accent modification and guises of benevolence
My analysis begins by drawing immediate attention to language oppression as part of an intersectional architecture of racism, especially where participants were told by mentors to modify their voice under guises of benevolence, care and pastoral support. This begins with Neha, whose experiences provide a stark example of intersectional discrimination concerned with race, Islamophobia, class, gender and language. As she told me right at the beginning of our interview: My language was a big issue in the school, it felt like a daily thing where other teachers and kids would make a comment on my voice, how I sounded foreign or whatever, even though I'm from Birmingham. […] Yeah, I mean, it felt absolutely like it was about race, like what people said about my language was just another way that racism was just so normal in that school.
Neha was raised in a working-class, predominantly Muslim area of Birmingham. She was a practicing Muslim, wore a headscarf, and spoke with a distinctive Birmingham accent. She was placed in a rural school which served a predominantly white community. Racism in the school was pervasive and legitimised by management. For example, on Neha's first day of placement, the headteacher suggested she remove her headscarf if she wanted to avoid racist comments, with students later calling her derogatory words denoting her Pakistani heritage. Many of these incidents were heard and ignored by white staff. Neha reported her experiences of racism to the headteacher and to her university, but both failed to act and she was forced to continue on the placement in order to complete her course.
As in the interview extract above, racist perceptions about Neha's language played a significant part in these experiences. She recounted how students would interrupt her teaching to shout out things such as 'we can't tell what you're saying' , 'Miss, can you speak English?' , 'you don't sound like you're from round here' and 'you sound like a terrorist' (see Khan 2020 for a discussion on the relationship between raciolinguistic ideologies and Islamophobia). At one point, a student wrapped a jumper around their head to mimic her headscarf and exaggeratingly imitated her Birmingham accent, an incident which went largely dismissed by her mentor. After reporting these things to a senior white teacher, Neha described how he told her to 'just deal with it' , that he was 'also marginalised' , and that he had also received comments about his Irish accent, but had 'learnt to ignore it' . Neha's mentor, however, appeared to be deploying the same tactics used by education policy makers which pit marginalised groups against each other and use the underachievement of white, working-class communities as a tool to derail anti-racist efforts (see Gillborn 2010).
As it was for Neha, social class was named by many other participants as one of the intersection points that language oppression materialised. Rebecca, an Iranian teacher who identified as working-class, described how the member of staff responsible for mentoring pre-service teachers in her school repeatedly questioned whether her accent was something she 'wanted to think about addressing' to improve her teaching, under a guise of benevolence which to Rebecca, felt 'horrible, like I was being singled out and like I didn't belong' . The same mentor suggested she seek out elocution lessons, subscribing to the colonial logics that racially marginalised speakers' language practices require pathological and corrective remediation. Rebecca's working-class identity, coupled with her being the first in her family to attend university, meant that she was positioned on the periphery on her ITE course and made to feel like her accent marked her out as unwelcome. She described how this consciousness led to her modifying her own accent to 'sound like the white, rich people on the course' in the hope that this would make her mentor perceive her as more employable (see also Bathmaker, Ingram, and Waller 2013). Similar feelings to 'play the game' were expressed by Derek, a Black pre-service teacher who had grown up in poverty and had a desire to work with children from similar backgrounds. He was placed in what he called a 'leafy, middle-class and properly white' school and described how because he was made to feel like a visual and audible imposter, he went to great efforts to modify his accent if he was to survive: I remember first walking into the school and just being like oh my god, I mean, I've never been somewhere like this before, like proper, Hogwarts it felt like. And I was just like, I just knew that if I was going to have any chance of surviving here then I' d need to get rid of my accent, you know, this east London boy is going to be caught out if I'm not careful, right? […] and I remember the effort that took, like serious you know, mental and physical effort, just tiring, knackering, and yeah, that was just hard work, tiring, on top of everything else.
Quite aware of the fact that the overwhelmingly white school was likely to be a space where he would experience racism, Derek talked about having to prepare himself for this, undergoing modifications to his voice which required physical and cognitive demands on his body. He felt an obvious discomfort here, given his own background, dedication to his community, and the way he had been brought up: I was brought up in east London, you know, on an estate, and my parents were always like, properly politicised you know? I was always being told by them to retain my identity, my community, my language, and I felt that, and did that, but as soon as I started teacher training it was like, ok you need to get rid of that now […] and I was fearful, fearful for that loss and what my parents would think. Puwar (2004) writes about how racially marginalised people are made to feel they are space invaders when navigating spaces dominated by white people, forced to modify aspects of their behaviour, appearance and language in their attempts to mitigate against racism. For Neha, Rebecca, Derek and others who faced pressure to whiten their voice, they were not just constructed as space invaders but sound invaders, perceived by their white colleagues as audibly deviant and cacophonous. As Derek put it: The message was like if you choose to come here and talk like that then you're not welcome, you know, like you don't belong and I mean, thinking back on that now, I'm just like fuck that you know, but at the time, I mean, what choice did I have, right?
Derek recounted how the very same linguistic features that were deemed to be unacceptable when he produced them would go unchecked when produced by other white colleagues, including his own mentor. Here then, language practices were perceived as deficient purely because they were produced by a racialised speaker (Rosa 2018). These raciolinguistic double standards are a marked characteristic of racism in ITE, where racially marginalised teachers face pressures to modify their language, conduct and appearance in ways that white teachers do not.
Various participants recalled how their experiences of language oppression reminded them of being at school as children, and felt dismay -but not necessarily surprise -that they were still experiencing it as adults. Both Abdul and Bashira spoke of how their accents were one of the reasons they were bullied in school and were pursuing a career in teaching to work with other children who faced intersectional discrimination, as part of their commitment to racial justice (see Gandolfi and Mills 2022). But this bullying, which they now named as racism, continued to permeate their lives because other teachers refused to hear their voice as appropriate for the classroom. Bashira described it was follows: You know for years, ever since being a kid in school people have made fun of my accent, it's something that marks me out as different, foreign sounding, you know. I didn't think I would still have that when I trained to be a teacher, I thought maybe things had moved on […] it's just like, endless, so draining […] all these years of having to deal with racist shit and feeling conscious about my accent on top of that. Just, ugh, I, it's just another thing.
Schools have long been a space where minoritised teachers are infantilised and deemed to be displaying inferior linguistic and cognitive capabilities (e.g. Mirza 2018), and thus in need of remedial guidance from their white superiors in the form of accent policing and 'advice' . This 'advice' however, simply transmits the message that the abandonment of language is in the interests of the marginalised to protect them from racism. These guises of benevolence and humanitarianism simply maintain responsibility on those that are already minoritised to change what they are doing and appropriate whiteness, whilst deflecting attention away from and leaving intact the oppressive structures which created that marginalisation in the first place.

Language oppression is sanctioned by the state
This section discusses how mentors' evaluative comments about language were often made in reference to external policy mechanisms, such as the Teachers' Standards and Ofsted. I elaborate on the argument that language oppression is not simply about individual attitudes, but an institutionalised phenomenon which is sanctioned by the white listening subject of the state.
The first policy mechanism discussed is the Teachers' Standards (DfE 2011), a set of state-designed professional guidelines which all pre-service teachers must conform to in order to gain qualification. The first version of these was introduced under Margaret Thatcher's government in 1983, and under successive Labour and Conservative administrations, have consistently reproduced raciolinguistic ideologies which frame idealised linguistic whiteness as the normative mode of using language (see also Smith 2023). The version of the standards currently in use instructs teachers that they must 'demonstrate an understanding of and take responsibility for promoting high standards of literacy, articulacy and the correct use of standard English' (DfE 2011: 10-11).
This clause was named multiple times by participants, refering to its use by mentors in their attempts to justify the policing of pre-service teachers' language if they were to be perceived as employable. For example, Felicia described her feelings of heightened self-consciousness after her mentor instructed her to speak 'more like a teacher' , and that her voice was 'incorrect' because it was perceived to deviate from homogenous ideals as laid out in the Teachers' Standards. Similarly, Mariatu was told to 'speak much clearer' and 'tone down her style' because her mentor was concerned that students might not be receiving the most 'appropriate' modelling in terms of 'the correct way to speak as a teacher' and that this risked breaching professional benchmarks. Hodan described how the person responsible for overseeing pre-service teachers in her placement school would regularly remind her cohort about the need to 'talk properly' , using the Teachers' Standards as a reference point. She felt this was 'definitely about race' because her mentor would often use Hodan's voice as an example of unsatisfactory practice, making public judgements on her language in ways which didn't happen for her white peers. Hodan recounted how she had been publicly shamed in a whole-cohort meeting because of her perceived linguistic deficiencies, in an enactment of language oppression under guises of respectability, inclusion and support: In one of those meetings with the other trainees […] it was quite early on in the placement, he said something like 'Hodan, you've got a very strong accent and we want to respect that, but it is something, that's something which would be a problem when coming to grade you against the Teachers' Standards' and I was like, oh my god, just remember feeling so embarrassed in front of all my coursemates, the other trainees, like singled out for my accent you know?
Hodan's experience demonstrates the racialised dimensions of mentor-teacher relationships (see also Wilkins and Lall 2011), and how mentors operate as gatekeepers with the power to pass or fail pre-service teachers based on how they are deemed to meet white linguistic benchmarks. A mentor's position as the white listening subject is both legitimised and encouraged by the Teachers' Standards, which further places responsibility on racialised speakers to modify their language and assimilate towards normative whiteness if they are to be granted professional status.
Participants also described how mentors referred to Ofsted in their attempts to justify language oppression. As two examples: I was warned that if Ofsted ever came to visit then my accent would be something I' d have to work on and change, because yeah, they said that this was something that the school got pulled up on the last time they were inspected. (Shanika) My accent was definitely something that the school were worried about I guess, because we were all given these kind of sheets about things to do to get ready for Ofsted […] one of those things was about voice and speaking properly […] in my next mentor meeting we were looking at this and she was like, Aditi this definitely applies to you. (Aditi) Recent work has exposed that language oppression is normalised within Ofsted school inspections (Cushing and Snell 2022), with the inspectorate operating as an institutional realisation of the white listening subject who have long subscribed to raciolinguistic ideologies in their perceptions of marginalised speakers. In Autumn 2022, an online leak of confidential Ofsted training materials exposed how inspectors are being actively trained to uphold these ideologies. For example, one of these documents states that 'teachers should also appropriately correct pupils' grammar when they are speaking formally' (Ofsted 2020b: 25), whilst another prescribes exactly how inspectors should respond to a hypothetical statement from teachers about language variation and social justice: Teaching pupils to speak Standard English is not our priority because it is elitist and pupils should not be forced to adopt middle class ways of speaking.
• This and similar attitudes are not in keeping with the National Curriculum which specifies that pupils are taught standard English, a term for accurate grammar use which has nothing to do with accents. If pupils cannot speak grammatically correct language, they are often excluded from certain jobs and universities. Teaching standard English is a matter of social justice. (Ofsted 2020c: 12;original emphasis) Ofsted's training materials instruct inspectors to take up listening positions which perceive non-standardised English as deficient and a barrier to social justice. The ideologies espoused in these training materials dovetail with findings from Cushing and Snell (2022), where we showed how the inspectorate reserve the most hostile comments on spoken language for schools serving low-income and racialised children. Evidence from interviews demonstrated the affective dimensions of Ofsted's espoused language ideologies, with the inspectorate named by mentors as a justification for policing the language of pre-service teachers under their supervision. Tamara, for example, reported how she had been warned by senior management that all staff must be using standardised spoken English in preparation for an imminent Ofsted inspection, and that teachers should not feel afraid to correct each other. Indeed, the recently leaked Ofsted training materials as described above include how inspectors are being trained to associate 'stronger practice' when teachers 'speak standard English with pupils' (Ofsted 2020c: 34).
Eric, too, described how pre-service teachers' accents in his placement school had been criticised by Ofsted in its previous inspection. School management's response to this was the introduction of compulsory training sessions branded as 'High Aspirations for All: Standards in Speech' . He recalled how he felt that because he spoke with a regional accent, he was perceived by his mentor as having low aspirations. Eric described how these training sessions reminded him of accent modification programmes, a global industry built on white supremacist logics (Ramjattan 2019). Ofsted's stance on accent modification operates under a guise of social justice, as part of a colonial narrative that the solution to structural inequalities lies in marginalised speakers modifying their accent to resemble whiteness. A raciolinguistic perspective reminds us, however, that no matter what minoritised speakers do with their language, they will always be perceived as deficient by the white listening subject -because ideologies about language are never just about language, and often not even about language at all (Rosa 2018). This chimes with Derek's experiences of language oppression, who continued to receive negative comments about his speech despite his efforts to sound like white colleagues. Whilst language is of course a central concern in struggles for social justice, including those in teacher education (see Picower and Kohli 2017), critical educational linguists have long challenged narratives which claim that the path to social justice is one where marginalised communities must adapt the way they talk (e.g. Lewis 2018).

Im/proper speaking and im/proper teaching
In this section I show how ideologies of speaking im/properly are connected to ideologies of teaching im/properly, and how this surfaced as a further way in which language oppression was justified in schools.
Many of the pre-service teachers I interviewed reported how the policing of their speech led to them becoming increasingly conscious of their voice, negatively affecting their confidence in the classroom, having a damaging effect on their teaching, and ultimately, leading them to question whether teaching was the right career choice for them. This is especially ironic given that justifications for language policing were often given under the guise that this would improve the quality of teaching. For example, Shanika, whose speech might be categorised as so-called multicultural London English, a variety that is systematically stigmatised in schools, described how perceptions of her voice had been the main reason that her teaching was deemed to be unsatisfactory and impeding her progress. Her mentor expressed disdain for Shanika's speech style, routinely criminalising it by telling Shanika she should 'stop using slang' and 'stop talking like you're a gangster' . According to Shanika's mentor, this was about 'self-respect' , but also about providing children with a 'good role model' . As a Black woman who was from the same, low-income community as many of the Black students she was teaching, Shanika felt that she provided exactly the right kind of role model for a group of children who were also marginalised in terms of race, class and language. She was one of the only Black teachers in the school, teaching a group of children whose language practices closely resembled those that Shanika's mentor was attempting to eradicate from her teaching. Indeed, Shanika felt that she enjoyed a strong relationship with her students, due to their linguistic alliances and community solidarity. Shanika's mentor, however, felt that her language style served as an impediment to good teaching and required modification if she was to make professional progress.
Other pre-service teachers reported similar narratives from their mentors and other colleagues, such as James, who had worked hard to build relationships with his mostly Black students who were categorised as 'difficult' and 'lost causes' by other members of staff. James described an incident where he had deviated from the departmental prescribed lesson plan and asked students about their own relationship with language, prompting enthusiastic discussion and what the 'sudden use of like, slang, Black language, like you know, all those proper Brixton accents came out, and I was like yes, cracked it!' . Post-lesson, James recalled how the teacher observing his lesson evaluated his teaching largely in terms of his perceived violation of normative and homogenous linguistic codes, as being 'too informal' , using 'inappropriate language' and 'not being as articulate as we would expect of a teacher' . James reported this instance of language oppression to his university tutor, who raised it with the school-based mentor, but was told that it was school and national policy that all teachers use standardised English, and James was deemed to be in breach of that.
Others, too, raised their concerns with tutors and mentors but received inadequate responses. For example, Amir, Femi and Zayan spoke with accents that they described as something which marked them out as, in Femi's words, 'foreign' . Femi recalled a time that her perceived inability to produce standardised spoken English had been the reason that her mentor had called a meeting with Femi's university tutor to discuss 'ways forward' if she continued to 'struggle with her accent' . Femi told me how this incident, along with a feeling that she was being discriminated against because of her race, led her to pause her studies. As I write this, she is yet to decide whether to return to teaching. Amir also doubted whether teaching was the right career for him after being placed in a school where largely white students had mocked his accent, adopting cartoonish voices and telling Amir that he sounded like the character of Apu from The Simpsons, an Indian character who is based on racist stereotypes (and until recently, was voiced by a white man producing a mock Indian-English accent). Although Amir's white mentor had witnessed this, Amir described how it was 'just brushed aside, like the students were just joking, and that I was making too big of an issue out of it' . For Amir however, who lived in a part of West London which had a long history of racial violence, the comments he received about his voice were simply another point on which he was made to feel like he was failing, unwelcome and had made the wrong career choice. Similar feelings were reported by Zayan, who reported how his mentor had questioned whether he was a 'native' speaker of English -a question that Zayan felt was motivated by the racist belief that his brown skin made him look like he lacked the ability to speak English, and ultimately was a body and a voice out of place in the inherently white system of teacher education (see Cho 2017;Gerald 2022;Rosa 2018).
Finally, Leon described how his placement school had commissioned external training about language, resulting in a staff development session which instructed teachers to use standardised accents and modify their voices to sound 'neutral' and 'professional' . Leon and I discussed how both these words were proxies for whiteness, and he recounted how colleagues had then later referenced the training programme to justify language oppression. This policing was something Leon himself experienced, being told by senior teachers that his London accent required modifying if he was to improve his practice. A particular target was Leon's use of th-fronting, where prototypical 'th' sounds are realised as either 'f ' or 'v' sounds (see Drummond 2018). Leon recalled a time where his mentor was meant to be providing constructive evaluations of his teaching but focused on policing his use of th-fronting instead: Like there wasn't much he said about my teaching at all, and if there was, I don't remember it, because all I remember was just him saying about the way I say the 'th' sound you know, like I say it like, 'fing' for 'thing' or whatever. But all he did say was I wasn't speaking properly, and started oh god, I remember, just saying fing fing fing over and over again and said repeat after me THING THING THING and I just, oh man, I think at that point I was like, mate, I'm out.
Leon's mentor referenced the external training they had received here, using this to deflect attention away from his own enactment of language oppression and doubling down on the argument that high-quality teaching must take place in standardised English. As this section has shown, racialised and evaluative perceptions about language are ideologically connected to evaluative perceptions about pedagogy, a connection which has gatekeeping potential in terms of who is heard not just as a legitimate voice, but as a legitimate teacher.

Raciolinguistic borders and resisting language oppression
England's schools are over-represented by white teachers, symptomatic of a systemic problem in attracting and retaining racially minoritised teachers (Bradbury, Tereshchenko, and Mills 2022;Smith and Lander 2023). This article has shown how raciolinguistic ideologies and language oppression add to this over-representation of whiteness, playing a central part in keeping out and pushing out racially minoritised teachers. Although I have centered discussions of language here, I want to re-emphasise that ideologies and perceptions about language are always about much more than this -and that language oppression is intricately connected to hierarchies of race and class. The stories that interview participants told brought explicit attention to this, in their analyses that comments they received about their language worked as proxies for race and class, and that instructions to flatten their accent were taken as instructions to assimilate towards whiteness.
My analysis has sought to draw attention to the broader sociopolitical conditions which allow structural racism to thrive in teacher education. Whilst I am critical of the mentors who chose to make or ignore racist comments, it is crucial to not see these simply as malicious individuals, but see them as symptomatic of a broader, hostile education policy architecture which is shaped by the logics of white supremacy (see also Kohli 2018). This raciolinguistic perspective urges us to move away from an individualised critique of discriminatory language attitudes and towards a critique of teacher education as a space in which racialised speakers are perceived as disposable because of institutional racism. In addition, this perspective relocates critical attention away from the purportedly deficient voices of racially minoritised teachers, and towards the white listening subject (Rosa and Flores 2017). As this article has shown, the white listening subject materialises in teacher education policy as a design feature, articulated and enacted in assessments, professional standards, Ofsted, and other such state sanctioned initiatives.
Raciolinguistic policing has deeply affective consequences, including the gatekeeping of borders in and out of the teaching profession. Geographical, institutional, and linguistic borders have a colonial and racialised dimension to them (Khan 2022), and the data I have discussed here demonstrates how perceptions of language run central to the construction and maintenance of borders in deeming who comes to be perceived as an il/legitimate teacher. For many of the teachers who participated in this research, the institutional policing of their language was one of the central reasons they came to question whether teaching was the right career for them, with some choosing not to pursue teaching after all. Raciolinguistic borders keep out -and push out -those teachers that are deemed by the white listening subject to be in breach of idealised linguistic whiteness.
The logics of white supremacy propose that if racially and economically minoritised individuals are to achieve social mobility, then it is their responsibility to modify their own ways of using language so that it appropriates whiteness. These logics are espoused in the ITE policy assemblage (Cushing 2023) and were at work in the reported experiences of the pre-service teachers I spoke to, where they faced institutional pressure to adapt their language if they were to be granted legitimate status. Prevalent advocates of such logics argue that racialised speakers must be adept in code-switching to produce the so-called codes of power and 'collegiate codes' (for instance, in Doug Lemov's Teach Like a Champion industry and its popular teacher education manuals). A raciolinguistic perspective, however, argues that this is simply a linguistic solution for a structural problem, and a 'mechanism for producing governable subjects that support the raciolinguistic status quo' (Rosa and Flores 2017: 642). Writing about speakers of African American English, Charity Hudley, Mallinson, and Bucholtz (2020: 217) describe how, under the code-switching argument, the message that racially marginalised speakers receive is that it is for their own betterment to leave their own language practices at the classroom door. This was the very same message that many of the pre-service teachers I spoke to had received from their mentors -that their own language was worthless, and that they would be best served if they abandoned the way they spoke all together. A raciolinguistic perspective reminds us that even if they had succumbed to the pressures of the white listening subject, they would still be encoded as deficient, because perceptions about language im/purities are never just about language, but about who constitutes a legitimate and im/pure body.
Whilst some scholars have suggested that simply training teachers in language awareness is the solution to eradicating language oppression (e.g. Baratta 2018), this is unlikely to produce any meaningful change because it assumes that racism can be eradicated if people just changed their attitudes. This theory of change is flawed because it reduces racism to a set of individual beliefs, rather than its reality as a state-designed structure (Lewis 2018).
Simply asking individuals to change their attitudes about language frames language as an isolated modality of stigma, masking and leaving intact the broader realities of intersectional and institutional racism. The risk here is that linguistic solutions continue to be posed as the solution to addressing racial inequalities (see Rosa 2018). Instead, a raciolinguistic perspective pushes us to connect issues of language struggle with broader structures of coloniality and white supremacy. A raciolinguistic perspective centres how racial violence has long been legitimised by colonial logics in which racialised populations are deemed to be biologically inferior because of their perceived linguistic shortcomings. These histories are important, too, in how we imagine anti-racist futures and show solidarity with those that are marginalised -for grassroots teacher organisations, unions and community activists in the UK have long sought to resist systemic racism in schools (Shafi and Nagdee 2022). In offering a structural critique of language oppression, it is only then that we might begin to bring about material change in the lives of racially minoritised pre-service teachers and continue to push back against state-sanctioned racism and the violence of the white listening subject. Through this, we can seek to transform education institutions as spaces free from racial injustice, rather than expecting racialised teachers to transform themselves.