Challenging anti-Black linguistic racism in schools amidst the ‘what works’ agenda

ABSTRACT Education policy in England’s schools is driven by the ‘what works’ agenda, characterised by interventions claiming to be scientifically objective and evidence-led. In this article I show how what works interventions reproduce anti-Black linguistic racism because to be perceived as someone who is ‘working’, racialised children must assimilate their language practices towards idealised whiteness. I present case studies of two teachers working in low-income, majority Black schools who rejected what works interventions concerning a commercially produced curriculum package and the so-called word gap, both of which framed racialised children as displaying linguistic deficiencies in need of correcting. I describe various institutional oppositions the teachers faced, including having their own language, expertise and evidence questioned by white management. I argue that the what works agenda is crafted by the state to delegitimise anti-racist efforts, and that for the state, what counts as ‘working’ is simply the reproduction of idealised linguistic whiteness.

(2011) critique of the what works agenda exposes how it operates as a mode of European coloniality which produces knowledge hierarchies built on the logics of white supremacy, anti-Blackness, and cognitive imperialism. He shows how the agenda is rooted in a discourse of superiority predicated on Western colonial logics in which non-European knowledge is constructed as inferior and irrational, and in which anti-racist efforts in education are subjugated, silenced, and discredited (see also Sriprakash, Rudolph, and Gerrard 2022;Rudolph, Sriprakash, and Gerrard 2018). In this way, the what works agenda can be seen as an effective means by the state to preserve the interests of the white status quo whilst continuing a long history of Black communities having their humanity called into question, especially in schools, and where 'racial equality' simply means becoming palatable to whiteness (Dumas 2014;Rollock et al. 2011).

Article overview
I contribute to the anti-colonial and anti-racist perspectives on the what works agenda by focusing on issues of anti-Black linguistic racism in schools. Previous critiques of the agenda have paid little attention to language, despite racist perceptions and ideologies about language playing a central role in coloniality, anti-Blackness, and white supremacy (e.g. Baker-Bell 2020;Bonfiglio 2002;Mignolo 1995). My discussion centres on case studies from two teachers in two different London secondary schools serving primarily Black children from low-income families. Both schools had subscribed to the what works ideology by implementing language-based interventions which were justified by management under a seemingly benevolent guise of social and racial justice. Both teachers felt that this agenda was not actually working at all but was causing harm to their students in how they exacerbated racial and class injustices and reproduced ideologies of anti-Blackness which blamed racialised children for their perceived failures in language use. I immediately frame this with the argument that for the state, the preservation of anti-Black language ideologies does constitute 'what works' because it keeps white people in power and preserves hierarchically arranged language ideologies.In short, what works interventions are implicated in the reproduction of racial and class linguistic normativity because to pass as someone who is 'working', children must model their language practices on idealised whiteness.
The first case study discusses Thea, who challenged anti-Black linguistic racism by designing a new curriculum which brought critical attention to histories of language, colonialism, and racial injustice. This curriculum was designed in response to her school's use of a commercial curriculum package, English Mastery, which Thea felt was prohibitive in how it policed her students' natural language practices and categorised them as incorrect. The second case study discusses Mowahib, who sought to challenge anti-Black linguistic racism at an institutional level in the form of a new whole-school language policy. This policy was designed in response to her school's use of interventions around the so-called 'word gap', an ideology that reproduces anti-Blackness in how it frames the size and quality of racialised children's vocabulary as lacking and limited when compared against the white status quo (see Cushing 2022). As commercial packages, both English Mastery and word gap interventions turn ideologies of linguistic deficit into economic profit -and like what works interventions in general, offer reductive linguistic solutions to what are structural and sociopolitical problems (see Rosa 2016).
Both case studies represent forms of teacher-led resistance to school initiatives which were justified by senior management on the grounds that they were 'evidence-led' and represented institutional commitments to social and racial justice. My criticism however, is not simply of two individual interventions rooted in anti-Black linguistic racism, but of a broader education policy architecture where the what works narrative is working to delegitimise and derail anti-racist endeavours. Whilst the complete eradication of anti-Black linguistic racism requires a fundamental rebuilding of society and a radical redistribution of power, I offer Thea and Mowahib's work as two attempts to enact anti-racist practice amidst challenging circumstances, institutional obstacles, and the seemingly benevolent narrative of 'what works'.

Language policing and anti-Black violence
Perceptions of language play an integral part in the production of racial and class hierarchies. These perceptions are consequences of raciolinguistic ideologies, in which racialised speakers are deemed to display linguistic deficiencies that require fixing, even when engaging in the same language practices that go unmarked when produced by middle-class white speakers (Flores and Rosa 2015). For example, Flores and Rosa discuss how nonstandardised forms such as 'she was' are heard as standardised English when spoken by privileged white students, but heard as defective and requiring correction when spoken by African American students. Raciolinguistic ideologies are never exclusively about language (and often not even about language at all), but coalesce with beliefs about biological im/purity and who counts as an ill/legitimate person. Beliefs around what constitutes un/acceptable ways of using language are not based on empirical linguistic practices but based on racialised perceptions that have long encoded lowincome, Black communities as exhibiting linguistic shortcomings (see also Dumas 2014). These raciolinguistic ideologies are an actively crafted design feature of education policy in England, such as in the racialised listening practices of Ofsted, the school inspectorate, who have long sought to stigmatise and eradicate language practices perceived to deviate from idealised whiteness (Cushing and Snell 2022).
Schools have long been a space where raciolinguistic ideologies and anti-Blackness circulate and transform into practices through various means of corrective procedures and language policing. Whilst these procedures certainly do include explicit instances of verbal chastisements and public shaming, this is not the default mode for language policing, which typically operates in much more covert ways. This includes curricula rooted in epistemologies of white supremacy, education policies built on the logics of linguistic deficit and verbal deprivation, and pedagogies that systematically erase the cultural practices of Black communities (e.g. Annamma 2018;Austin 2022;Johnson 2022;Sriprakash, Rudolph, and Gerrard 2022). Scholars have long demonstrated that language policing in schools is particularly marked for people racialised as Black (e.g. Coard 1971;Smitherman 1977), and contemporary research has (re)highlighted the linguistic violence that Black children and teachers continue to endure in schools (e.g. Cushing and Snell 2022;Baker-Bell 2020;Cabral 2022;Johnson 2022). Baker-Bell's (2020) work shows us how this materialises in terms of anti-Black linguistic racism, defined as follows: the linguistic violence, persecution, dehumanization, and marginalization that Black Language (BL) speakers endure when using their language in schools and in everyday life. It includes teachers' silencing, correcting, and policing students when they communicate in BL. It is the belief that there is something inherently wrong with BL; therefore, it should be eradicated. It is denying Black students the right to use their native language as a linguistic resource during their language and literacy learning. It is requiring that Black students reject their language and culture to acquire White Mainstream English, and it is also insisting that Black students code-switch to avoid discrimination. (Baker-Bell2020, 9) Anti-Black linguistic racism is part of a broader, interconnected typology of anti-Black violence, which systematically erases the cultural, linguistic, and cognitive knowledge that Black children bring to school with them. Johnson, Bryan, and Boutte (2019) describe such a typology, reproduced in Table 1 and where I also show some ways in which this materialises in schools in England. It should be clear from this table that the default mode for language policing in schools is not in the form of overt corrections and shaming but in more covert ways that permeate all aspects of Black children's educational experiences, as something that is actively woven into the fabric and structures of school life. In the following section, I argue that ongoing efforts against anti-Black linguistic racism and language policing in schools must focus on broader oppressive structures of white supremacy and provide a brief history of activist work that has sought to do this.

Resisting anti-Black linguistic racism in England
Educational linguists have long challenged oppressive language ideologies in the hope that this will modify teachers' attitudes about the language of Black children and that this shift in attitude will result in the modification of pedagogies and more racially just classrooms. Whilst this work has repeatedly shown the validity of Black children's language, deficit attitudes in schools remain durable and persistent. Combining insights from language ideology scholarship with Critical Race Theory, Lewis (2018) and Flores, Lewis, and Phuong (2018) argue that the failure of educational linguists to bring about any tangible impact is rooted in the fact that most work has concentred on modifying individual attitudes as opposed to connecting language-based struggles to broader histories of colonialism, anti-Blackness and white supremacy. Flores et al. argue that only a joint focus on language oppression with broader systems of oppression will allow teachers to be positioned as activists, begin to dismantle raciolinguistic ideologies in schools and bring about structural change. They conclude that While continued work with teachers to change their individual attitudes is important, more concerted efforts should be placed on the institutional mechanisms that constrain teachers from taking on alternative institutional listening subject positions in their classrooms. [. . .] if we frame the problem as one of individual teacher language attitudes, then our solutions will focus on changing individual teacher attitudes, leaving intact the broader institutional listening subject position teachers inhabit. If we frame the problem as one of institutional racism, then our solutions can more readily focus on dismantling institutional racism. (Flores, Lewis, and Phuong 2018, 24) It was this framing that was a starting point for Thea and Mowahib's work, where they approached their work not as a linguistic solution to a linguistic problem, but as a socially activist project that sought to connect issues of language with wider structures of racial injustice. They refused to allow attention to be directed away from the social, economic, and political dimensions of racism and refused to be complicit in the white supremacist logics which supposed that racially marginalised children would perform better in school if they would just modify their use of language. Their critical knowledge of language was central to this, for whilst language can be used to maintain existing power relations, it can also be used to resist them (Alim 2005). Whilst I argue that all teachers should see themselves as language activists, it is especially important for teachers of English given that they work so closely with language and thus are even more likely to reproduce raciolinguistic ideologies in their practice. As Seltzer and de Los Ríos (2018 write, English teaching is an inescapably political and ideological endeavour that involves colonial histories, ideals, and powers [. . .]. Thus, to build a teaching force that can destabilize the oppressive ideologies that result from these histories and unequal power relations, English teachers must commit to developing and extending their critical stance and their raciolinguistic literacies by reorienting their listening practices in ways that resist oppressive language ideologies. (Seltzer and de Los Ríos 2018, 72) Teachers experiencing and enacting raciolinguistic struggle have a long history in England (see, e.g. Gerrard 2013;Jones 1986;John 2006;Sivanandan 1981), and I now offer a brief overview of some of this work to provide a historical backdrop to which Thea and Mowahib's work sought to build on. As discussed above and by Flores, Lewis, and Phuong (2018), history reminds us that some of the most transformative changes have been brought about where activists have located anti-Black language ideologies as part of a broader structure of white supremacy and institutional racism. This work was particularly prominent in the 1960-80s, where Black educators and families led movements in rejecting educational systems which included the punishment of their children for how they used language. Organisations such as the East London Black Women's Organisation, the National Convention of Black Teachers, the Black Supplementary School Movement, the Caribbean Teachers Association, the Black Parents Association, and the Black Education Movement held regular meetings to imagine and design new ways of educating their children which sustained their natural language practices and called out racist policies at the same time. This resistance work took place amidst a prevalent national discourse on verbal deprivation, where lowincome, Black children were routinely perceived by teachers and policymakers as incapable of producing any adequate language. For instance, a 1968 seminar in Walsall called The Language of Failure brought (mainly white) teachers together who worked with lowincome and Black children and designed interventions that sought to fix purported linguistic and cultural defects (see Flower 1970). A year later, a racist document written by the white headmaster of Highgate School in Haringey, which came to be known as the 'Doulton Report', was leaked to a group of Black Caribbean parents. The report was underpinned by anti-Black eugenics, in its claim that the IQs of their children were 'below their English contemporaries' and consequently, 'academic standards will be lower in schools where they form a large group' (Haringey Council 1969). Much like narratives around the 'word gap' in today's schools (see Cushing 2022), Doulton's claim was largely based on perceptions that Black children exhibited linguistic deficiencies that were symptomatic of cognitive and behavioural problems. Black activists continued to resist such discourses of 'sub-normality', exposing how anti-Black policies were actively and systematically designed by the state, with racist perceptions of Black language being used as a key justification for segregationist schooling and the use of corrective-based language interventions (e.g. Coard 1971; La Rose 2014).
Newer histories of anti-racist work in English education remind us that these struggles still continue and that deficit perceptions of language continue to be a central part of this work (e.g. Richardson 2005; Wallace and Joseph-Salisbury 2022). However, contemporary activist work in education must be contextualised in relation to what is an increasingly hostile policy architecture, under which anti-racist work is routinely discredited by the state. For instance, in 2021 the chief inspector of Ofsted, Amanda Spielman, derided what she called 'militant activists' in response to school community protests about institutional anti-Black racism and Islamophobia. In 2020, the Conservative MP Kemi Badenoch criminalised teachers by declaring that they were breaking the law if they taught 'contested political ideas as if they are accepted fact', such as Critical Race Theory (The Independent 2020). More broadly, the state-crafted 2021 Report of the Commission on Racial and Ethnic Disparities (HMG 2021), known as the Sewell Report, attempted to deny the existence of structural racism in Britain whilst (re)attributing blame to Black families for their apparent failures to prepare their children for school (see Tikly 2022). This hostile and actively anti-Black policy architecture is the context in which I now discuss the two case studies of raciolinguistic struggle and resistance.

Case studies of resistance
Over the course of one academic year, I worked closely with two teachers, Thea and Mowahib, in co-designing policies and curricula that resisted anti-Black linguistic racism. They saw themselves as anti-racist teachers who were committed to challenging structural racism in their work (see Arneback and Jämte 2022). We talked about the long history of teacher activism as discussed in the previous section and shared readings which we felt resonated with the aims of the project. We discussed academic work on anti-Black linguistic racism in schools and used this to inform the development of pedagogy, curricula, and policy. I visited their schools, observed lessons, spoke with their colleagues and students, and formally interviewed them. We visited the Black Cultural Archives in Brixton together to explore the long history of anti-Black linguistic racism in schools. I collected various artefacts that their projects led to, including student work, curricula, and new policies.
As argued for by Lewis (2018) and Flores, Lewis, and Phuong (2018), both Thea and Mowahib committed to highlighting the broader sociohistorical and political processes that continue to produce anti-Black linguistic racism, and how teachers can inhabit listening positions that perpetuate this. Thea and Mowahib's schools had many similarities -both were in areas of economic disadvantage serving large proportions of Black children, with a majority white teaching staff. Both were under pressure to help students perform highly on standardised assessments, as well as pressure from Ofsted and from parents to prepare their children for further education and work. Management from both schools had responded to these pressures, at least in part, by implementing new policies around language with the justification that these were underpinned by what works research -in Thea's school, the English Mastery commercial curriculum package, and in Mowahib's school, policies subscribing to the so-called word gap. Thea and Mowahib expressed concern to management about these new initiatives, in terms of how they felt they were coercing them into being complicit in reproducing intersecting forms of anti-Black violence as described in Table 1. Their case studies then need to be read in light of this context, where their own activist work pushed back against a state-crafted narrative and institutional beliefs around what constitutes what works and what counts as evidence. This carried material risks for both teachers -for Thea, as a relatively new member of staff who expressed caution about being perceived by management as a 'troublemaker', and for Mowahib, as a Black, Muslim woman who occupied a position of racial disadvantage within a majority white workforce and a broader national environment of Islamophobia. Ethical approval was granted through my own institution, and Thea and Mowahib's names are pseudonyms. Management at both schools gave approval for the work to take place. Due to reasons around confidentiality, neither Thea nor Mowahib were able to be named as co-authors on this articleespecially Mowahib, who expressed worry that her school would act punitively in the face of our criticisms. Both, however, read and commented on the manuscript. This work would simply not have been possible without their courage, intellect, and innovation.

Case study 1: challenging anti-Black curricular and pedagogical violence
Thea identified as white and working-class. She was born in London but spent her preadolescent years between England and Europe. She spoke openly about her experiences of linguistic discrimination in both these settings, with other children telling her to 'go back to where she came from' due to her accent that marked her out as different and unwelcome. As a beneficiary of white supremacy, Thea felt strongly that her whiteness put her in a position of responsibility to challenge anti-Blackness in her work (see also Tanner 2019). She had long been engaged in anti-racist efforts in her teaching, through centring politicised discussions of language, race, and prejudice in her classrooms which made connections to the coloniality of national curricula, political populism, and police brutality.
Thea's school was located in an economically deprived area of southeast London and served a racially and linguistically heterogenous student population, with a large proportion of Black African children. Most of the teaching staff were white. The most recent Ofsted report graded the school as 'Requires Improvement', with criticisms aimed at the teaching of English and for the perceived lack of progress that racially marginalised and low-income children were making. It was against this context that Thea's management had subscribed to an intervention subscribing to the what works narrative: English Mastery.

English Mastery and anti-Blackness
Prior to Thea joining her school, management had bought into a commercially produced curriculum and resource package, English Mastery. Whilst flat pack, commercially produced curriculum packages have long been a feature of modern schooling, they have increased in popularity in recent years and are marketed to teachers under the promise of saving time, being produced by 'subject experts', and leading to increased test scores. English Mastery is sold by the education charity and large academy chain Absolute Return for Kids for between £2,100 -£7,350. Its branding and marketing are heavily geared around the what works narrative, focusing on standardised assessments, explicit grammar and vocabulary instruction, and the 'core knowledge' ideology of curriculum building which has its roots in the white supremacist and colonial logics of the North American academic E.D. Hirsch (see Hodgson and Harris 2022;Rudolph, Sriprakash, and Gerrard 2018). English Mastery claims to be a programme supported by scientific robustness, as well as a claim to social justice in a way that reproduces racist discourses of deficit -for example, it seeks to 'address the cultural capital deficit that our children experience' (English Mastery 2018, 3). In 2021, English Mastery published a report on equality and inclusion, where its designers outlined modifications in response to criticisms that materials were perpetuating racial injustices (see English Mastery 2021). However, this report focused on simply inserting writers of colour onto their curriculum, as well as making small tweaks to specific tasks (such as exploring issues of colonialism in The Tempest) rather than addressing broader structures of white supremacy and anti-Blackness. The report makes no reference to language and equality, and its designers continue to double down on the white supremacist logics that using standardised English is the key to unlocking social and racial justice.
The architects of English Mastery ground their claims around what works and research robustness largely in reference to a 2018 randomised control trial by The Brilliant Club, which found that reading and writing scores were higher in the group receiving the English Mastery curriculum, and a 2022 evaluation by the Education Endowment Foundation (Education Endowment Foundation 2022) 1 . The EEF report, based on teacher surveys, reported positive responses in terms of teacher workload (not particularly surprising given that the programme provides teachers with ready-made lesson materials) and in terms of increased confidence in relation to standardised English grammar. At the time of writing, English Mastery is the subject of a Department for Education's Accelerator Fund, illustrating the level of state-support that is lent to what works programmes. Management at Thea's school were convinced by this what works narrative and that English Mastery represented the kind of curriculum package that would improve disadvantaged students' test scores, especially in terms of their perceived lack of competence in standardised English. In an interview, the head of the English department talked about how she felt that the programme would 'address persistent errors' in children's grammar, curtail the use of 'colloquial' speech, and allow them to 'function properly in the world'. Whilst she suggested that she wanted to 'celebrate' the home language practices of students, this view suggested that she drew an ideological dichotomy between 'home' and 'school' in which nonstandardised language was a problem to be fixed. In this way, English Mastery was framed as a panacea for supposed defects in language in a way that continued a long tradition of pinning the blame on marginalised children for their apparent failures (see Labov 1971).
Thea, however, felt differently. The linguistic content of English Mastery sat uncomfortably with Thea's values as a teacher committed to challenging anti-Black linguistic racism in her work. For example, English Mastery lessons ask teachers to surveil their students' competence in standardised English grammar through exercises such as correcting 'errors' in sentences, producing choral responses of standardised English verb tenses, and engaging in artificial, decontextualised interactions in which only standardised English carries legitimacy. Thea repeatedly spoke of how these exercises simply did not allow her students to produce their natural language -with most children saying as little as possible due to fear of getting it wrong and a complete linguistic and cultural disassociation with what they were being asked to do. It is against this narrative and acts of anti-Black linguistic violence that Thea's activist work took place.

Resisting anti-Black language policing through curriculum change
Thea's act of resistance was in the form of a new unit of work that she designed, replacing English Mastery. The unit was built around the 2020 novel Windrush Child by the Black writer, Benjamin Zephaniah. This was more than simply inserting a writer of colour onto the curriculum however, for Thea also drew on anti-racist language pedagogies (e.g. Paris 2012) and her wider knowledge of critical linguistics which sought to affirm and centre the language practices of all children. Windrush Child tells the story of a young boy, Leonard, and his arrival as an immigrant to 1960s Britain from Jamaica. Leonard and his family experience various forms of anti-Black racism, with deficit perceptions of their language central to this -a narrative that is grounded in the reality of anti-Black 1960s education policy in England as discussed above. Zephaniah explores anti-Black linguistic racism in relation to enslavement and British colonialism, which Thea also chose to do in her own curriculum. For example, when talking to his son about the intersections of language, power, and race, Leonard's father says: There are some white people who think that white is de best, de standard, and everyone else is coloured. And because they think they are the best, they think they have de right to rule over us. You know 'bout slavery?' (Zephaniah 2020, 55) This is in stark contrast to the way that standardised English is conceptualised within the English Mastery unit that Thea's students were so used to. In English Mastery, standardised English is framed as a politically, ideologically, and racially neutral set of empirically identifiable grammatical features to which all should aspire if they are to achieve success in the world. The logic here is that children should assimilate toward linguistic whiteness or else be complicit in their own failures and struggles. Funding allowed us to interview Benjamin Zephaniah, where he explored the relationship between standardised English, power and racism which has long permeated education policy in England. He said, 'you know growing up, I've heard people in authority many times saying to me, you know, don't talk to me like that, speak properly when you speak to me'. Thea played a video clip of this interview to her class, generating a discussion of how standardised English is a product of white supremacy and a colonial construct used to justify covert linguistic racism (Kroskrity 2021). Thea's students explored critical questions concerning the origins of a standard, bounded language, and how it was designed in 1700s London by white people occupying positions of institutional power. This then led onto conversations around the policing of language deemed to be non-standardised when produced by Black speakers in contemporary London schools, and how this functioned as an act of anti-Black violence. Here, Thea drew on her knowledge of contemporary research which has highlighted the links between standardised English and anti-Blackness in neoliberal schooling (e.g. Austin 2022; Baker-Bell 2020), as well as key literature on white supremacy in education policy (e.g. Gillborn 2005). In interviews that took place throughout the year, Thea repeatedly expressed how the English Mastery unit had never allowed for some of these conversations due to the kind of limitations it places on teachers. I also argue that the what works agenda actively delegitimises these conversations as valid classroom activities because they offer a challenge to the White status quo. Additionally, these sanitised and ahistorical stances on language are entirely in keeping with the broader conservative argument that classrooms should be politically impartial, as discussed above. Thea and her student's discussions about language and anti-Blackness took place in a classroom where management had pinned posters on walls that all talk should be in 'fluent speech', use 'academic expression' and in 'standard English' -much like the kind of raciolinguistic ideologies espoused in the English Mastery materials. The nature of Thea's pedagogical identity, coupled with the principles of her anti-racist curriculum, meant that she disobeyed these instructions and refused to police the language of her students which management had deemed to require remediation. In fact, Thea did much more than actively disobey these directives from management -she encouraged her students to challenge them. One way in which she did this was by showing a clip of Benjamin Zephaniah discussing schools who have introduced punitive policies on nonstandardised spoken English (see Cushing 2021; Snell 2018 for a wider discussion).
Zephaniah describes this as 'linguistic fascism', talking about how socially constructed dichotomies such as 'academic'/'non-academic' and 'home'/'school' endorse language policing under the ideology that school is a place where marginalised children must be corrected for their linguistic misdemeanours and assimilate to idealised whiteness.
Bringing critical attention to the coalescing of linguistic and biological hierarchies was key to the Windrush curriculum that Thea designed. As an example, Thea asked her students to complete a linguistic profile about themselves, where they examined assumptions and ideologies surrounding their own use of language and the role that race and racialisation played in the construction of raciolinguistic hierarchies. Joy, a student with Nigerian heritage who spoke English and Yoruba, created a drawing of herself which depicted her mouth clamped shut, representing how her Yoruba had been silenced in schools. For Joy, she was able to interrogate the kinds of anti-Black linguistic violence and suffering she had experienced in school, situating this oppression in terms of the broader structures of white supremacy, patriarchy, and linguistic imperialism which had pushed her identity to the periphery of classroom spaces (see Morris 2016).

Case study 2: challenging anti-Black systemic and linguistic violence
The second case study describes the anti-racist efforts of Mowahib, a teacher in her third year of practice who was working at a secondary school in an economically deprived area of northwest London. Mowahib identified as Black, working-class, and an immigrant, having moved from Somalia to England when she was eight. She held a Masters degree in English and Education and was interested in pursuing doctoral research. She spoke in what might be perceived as a standard southern English accent, yet because of the ways that raciolinguistic ideologies construed her as deficient, she still reported experiences of being instructed by senior colleagues to modify her accent so that it corresponded towards idealised whiteness in the form of standardised English. She was a practicing Muslim and wore a hijab and, to her knowledge, was the only Muslim teacher in the school.
Her school served a community of largely Black Caribbean children but had a workforce of majority white teachers. Mowahib had recently been appointed as a whole-school literacy lead, a role that included leading on institutional initiatives in relation to language and delivering professional development for staff. This role was part of the school's efforts to rebrand themselves as a 'research-informed' institution and partly justified in response to a 'requires improvement' grading from Ofsted. Mowahib received no additional pay for this role and was granted just a single hour of teaching relief a week to do the work. In her case study, I focus on her efforts to enact institutional and policy change in the wake of linguistic and systemic anti-Black violence.

Punishing words: the word gap and anti-Blackness
Prior to Mowahib taking up her new post, senior management had bought into a cluster of commercial interventions around the so-called word gap. The word gap is a raciolinguistic ideology underpinned by the myth that low-income and racialised families exhibit linguistic shortcomings due to their lack of adequate vocabulary, and that this is the root cause of struggles in schools they might experience -and so they require remedial interventions designed to increase and improve their use of words. The original research that attempted to empirically prove the existence of a word gap (Hart and Risley 1995) is rooted in anti-Blackness and academic knowledge production about the perceived inferiority of language practices racialised as Black (see Aggarwal 2016). In Cushing (2022), I showed how the word gap has become recently normalised in England's schools, tethered to colonial logics in which low-income, Black communities have long been pathologised and encoded as displaying linguistic absences which require urgent attention from teachers. A raciolinguistic perspective on the word gap argues that it is simply another way of finding faults in the language and cultural practices of Black communities whilst posing linguistic solutions to structural inequalities, which continue to position the most marginalised members of society as a liability and responsible for their own failures (Rosa & Flores 2017).
Mowahib's school subscribed to word gap ideologies under their belief that this was in line with new research and the 'evidence-informed' narrative of England's education policy assemblage. For example, the word gap forms a major part of Ofsted's so-called research review on the teaching of English (Ofsted 2022; see also Cushing and Snell 2022), is integral to the 'evidence-base' of the Core Content Framework 2 (DfE 2019), and underpins a succession of research reports by the Children's Language Team at Oxford University Press (e.g. Oxford University Press 2018). Influential, uncritical books about the word gap were available for loan in the 'Staff Research Library' at Mowahib's school (e.g. Quigley 2018). In an interview, Mowahib's head of department told us how word gap interventions had been brought in because 'a lot of the poorer students don't have that language, the right kind of language, they have things that are missing'. Mowahib, however, knew that the children were perfectly capable of engaging in 'high-quality' language because she identified with many of them culturally, was from a similar community, had spent time listening to them, talked to their parents, and drew on her knowledge of educational linguistics research to know that language racialised as Black has a long history of being stigmatised in schools. Mowahib expressed deep anxiety at how her school's word gap interventions were encouraging her to police the vocabulary of Black children in particular, and how it was Black children who were more readily segregated into so-called 'vocabulary booster' and 'catch up' sessions.

Resisting word gap narratives through policy change
Mowahib's approach in resisting and rejecting word gap ideologies was to target institutional change in the form of new, whole-school policies. In an interview I conducted with her early on in the project, she described why she had made this decision: I feel like the new [literacy lead] role has kind of given me the space and like, power to try and do something different across the whole school. You know, like maybe people might actually listen if they feel like literacy is actually my job and that, I dunno, I know what I'm talking about? [. . .] So yeah, changing the policy I think is pretty important because I know that will have to be approved by the headteacher and things, so to get their signature on it would then give me that proper, official support. (Interview, 12 September 2021) In the same interview, we discussed ways of designing a policy that pushed back against word gap narratives and located this within an anti-racist tradition. Mowahib felt that student and parental engagement were crucial, as she looked to centre the voices of marginalised speakers within the policymaking process. She described her frustrations when she proposed this to senior management, who told her that there was no funding in place to support this, that any new policies should be designed 'purely in house', and that there was an institutional template for new policies -which did not include student and parent involvement. Despite this, Mowahib explained how she held multiple conversations with her own students about language, many of whom felt that the school's new focus on vocabulary was 'pointless' and 'felt weird' because they were being asked to talk in ways that sat uncomfortably with their own use of language and made them feel that their own ways of using language were inadequate. These students provided a stark illustration of how word gap interventions place a burden onto those that are marginalised to abandon their own language practices and modify their speech if they are to be perceived as legitimate. We spoke at length about how raciolinguistic ideologies work in these ways, and so Mowahib decided to centre this in the design of her new policy by shifting responsibility away from the marginalised speaker and towards the privileged listener. Her first draft of the policy included the following: At <school name> we do not believe in the existence of a word gap and wholly reject such deficit and racist descriptions of language. If a gap does exist, it exists in the way that people perceive language, rather than how they use it. We seek to value and understand what all children do with their language, regardless of their race or social class. We strive to be antiracist teachers. We recognise that teachers have disproportionate levels of power in the classroom and take this responsibility extremely seriously, in recognising that this power can cause harm. We commit to hearing and seeing the beauty in all children's language, and commit to nurturing that in our classrooms. Our focus is not on looking for what children cannot do with language, but what they CAN do. That is our starting point.
Mowahib felt it important to frame the policy as a research-informed document, especially given the school's claim to this narrative and how word gap policies had been justified by management under the logics of 'what works'. Accordingly, she included references to research from critical educational linguists whose work has called out anti-Blackness in schools, such as Baker-Bell (2020), and was granted 15 minutes in a senior management meeting to present the draft of her new policy. Later that evening she sent an email which recounted how senior management were policing and censoring her antiracist work: Am emailing from my own account because it feels a bit risky to be voicing my criticisms on my work email! SMT [senior management team] didn't really like the draft text of the new policy and want me to rewrite it. It was all a bit uncomfortable tbh and I suddenly lost my confidence with it and hate that I did. Ugh. I did take some notes but was so nervous that they're a bit of a mess. But they want me to remove the part about power and harm, and want me to remove the bit where I say the word gap is a racist description of language. (Email, 18 November 2021) Mowahib's new draft was eventually accepted by management, but only after numerous track changes and requests to change things. Mowahib described this as a 'whitening of the word gap' and we discussed how this was illustrative of white people's unwillingness to name racism (Leonardo 2002) and of broader attempts by the state to delegitimise anti-racist efforts in schools. The final, whitened version was as follows: At <school name> we always value what children do with their language, whatever their personal circumstances or background. We recognise that some ideas in language are contentious (such as the 'word gap' and Standard English) but encourage staff to use our CPD library to make informed decisions and be research literate. We always strive to see and hear the beauty in all children's language, and commit to nurturing that in our classrooms.
Just as Mowahib was attempting to push back against anti-Black language policing then, she herself had her language policed by white management. She described how management had silenced her efforts to bring critical attention to the anti-Black logics underpinning word gap interventions and that they were deliberately attempting to derail her work under the argument that the language around race and power was 'too strong' and 'unsuitable' for a school policy. Management had also expressed fears that the stance and wording in her original policy carried a risk for future Ofsted inspections, especially given Ofsted's subscription to word gap ideologies (Cushing 2022). The clause referring to staff becoming 'research literate' by engaging with the staff library was written by a member of senior management, and frustrated Mowahib because she knew that highly problematic literature about the word gap was in this library. Her requests to add anti-racist work on language to the library had been initially approved, but she was told they had been indefinitely delayed by budgetary bureaucracy.
Mowahib was granted a 20 minute slot at a whole-staff development day to launch the new policy. We spoke about how to maximise her time here, and felt that a productive approach would be introducing Rosa and Flores' (2017) notion of a raciolinguistic perspective. Here, attention is shifted away from the language practices of stigmatised speakers and towards scrutinising the perceiving practices of those in power, whilst connecting beliefs about language to broader sociopolitical structures such as institutional racism. This involved framing the introduction of the new policy with the request that her colleagues reflect on how their own modes of perception had the potential to reproduce harmful ideas about their students' language, and how these perceptions were shaped by the logics of white supremacy. It also involved asking her colleagues to consider how discourses of limits, absences, and gaps can be rejected in order to recognise, rather than repress, the natural language practices of marginalised children. Mowahib chose to script her presentation, and here I show a short extract of it: We need to stop thinking that the way children speak is the problem, and start thinking about the way that adults listen is the problem. Take, for example, the 'word gap'. This is the idea that poor, Black children use a limited vocabulary when contrasted against their white classmates.
It's what linguists call a deficit ideology, and has its origins in 1990s research from the USA which claimed that poor, Black children use and hear less words than wealthy, white children. But linguists have shown that all children are perfectly capable of using language in 'rich' ways, but it's the most disadvantaged children who get labelled as deficient. These labels can be racist and damaging. And it's us as teachers that have been complicit in this labelling, because we mishear, misjudge and mislabel them. This is why our institution must begin to address its own biases in how we listen to and talk about our children.
Mowahib asks her colleagues and management to adopt a raciolinguistic perspective in shifting critical focus towards their own potentially harmful modes of perception. She connects language stigmatisation to broader structures of institutional racism, as well as grounding her policy in research as to play into the institutional narrative of being 'evidence-based'. Despite this, Mowahib's school continued to use word gap interventions on the grounds that they represented, as one member of management told me, 'the most efficient way of getting disadvantaged children to catch up and boost their learning'. Here then, management interpret 'structural inequality as a linguistic problem requiring linguistic solutions, rather than as a politico-economic problem requiring politico-economic solutions' (Rosa 2016, 165). This reductive logic is commonplace within word gap narratives and the what works project more broadly, where it is posed that if marginalised children simply accept that their language is inadequate and look to change it, this will allow them to enjoy upward social mobility and fully participate in society. Despite attempts from management to police her work, Mowahib's activist work lives on amidst this narrative, a small but ongoing challenge to a global industry of anti-Black linguistic racism in schools.

Raciolinguistic justice and what really counts as 'what works?'
Whilst the complete eradication of anti-Black linguistic racism requires a fundamental rebuilding of society, a complete undoing of language ideologies and a radical redistribution of power, this article has presented work from teachers who are doing what they can to challenge it in their own practice. Both teachers involved in this research went beyond simply advocating for the legitimacy of their own racialised students' language practices but imagined ways of disrupting the sociopolitical and educational conditions that have delegitimised them in the first place, whilst connecting issues of language struggle to social struggles crafted by white supremacy (see Flores 2017). This activist work took place against a state-sanctioned narrative of what works and evidence-based practice in schools, a narrative that runs counter to the aims of anti-racist research because it classifies idealised linguistic whiteness as the only thing that is 'working'. The point here is not to compare how successful Thea and Mowahib's work was but to present their initiatives as something that was pushing back against a policy ecology in which their anti-racist endeavours were framed as questionable, suspicious, and in breach of institutional norms defined by whiteness.
Management at both schools had been seduced by the what works ideology under a belief that this represented their purported commitment to social and racial justice, and that the modification of stigmatised language practices was done so in the best interests of racialised children. However, as Thea and Mowahib's work showed, the what works interventions chosen by management simply perpetuated social and racial injustices because they were rooted in colonial and white supremacist logics that sought to annihilate language practices deemed to deviate from idealised whiteness. Put simply, the only thing that works in what works interventions is the maintenance of white linguistic supremacy. Both Thea and Mowahib knew that simply teaching their lowincome, Black students to use more standardised English or more academic sounding words would not grant them the socioeconomic mobility that their relative institutions so allegedly desired. But they came up against hegemonic ideologies and interventions that both subscribed to this myth. Thea and Mowahib found and widened cracks in an oppressive system (Weber 1997), imagined alternative worlds, challenged the narrative that the what works project is in the best of interests of marginalised children, continued a long tradition of anti-racist work in England's schools, and showed different ways that anti-Black violence can be resisted. Some comparison, however, is warranted. Whilst Thea was left largely alone by management and able to proceed with her work unchecked, Mowahib reported feelings of being under surveillance by management and experienced institutional attempts to edit, police, discredit, and censor her efforts. One simple reason for this was because Thea was enacting relatively small-scale change confined to her own classroom and department, whilst Mowahib was enacting structural change which involved trying to undo institutionalised raciolinguistic ideologies maintained by senior management. Curricular and pedagogical anti-Black violence were more immediately resistible because challenges to them generally involve changes to practicewhich are facilitated by the individual autonomy that teachers typically enjoy. But systemic and linguistic violence are much less resistible because challenges to them involve abolishing entire systems that are harder to penetrate and given greater protection by whiteness.
One key difference, of course, was Thea and Mowahib's different racialised positions in their schools. As a Black, working class, Muslim woman in a school with a predominantly white workforce, Mowahib reported an ever-present sense of surveillance and scrutiny which has long been central to white supremacist and anti-Black logics (Browne 2015). Mowahib described herself as someone who sought to actively reject the Islamophobic stereotype of oppressed, headscarf-wearing Muslim women who pose a threat to European society (e.g. Khan 2020; Mirza and Meetoo 2018). She was highly attuned to the fact that her body and language were Othered when with her white colleagues (Puwar 2004) and felt the most comfortable when working with Black students in her own classroom. Her reported discomfort, for example, at presenting the first version of her anti-word gap policy to management is illustrative of how perceptions of the policy were not just about the policy, but about the person who designed it. Here, Mowahib experienced what Fallas-Escobar and Herrera (2022) articulate as a raciolinguistic struggle, where racialised individuals' expertise about language is questioned and policed by institutional structures of white supremacy and anti-Blackness. Mowahib was also someone who experienced anti-Black linguistic racism in school, enduring regular instructions from her white mentor to modify and flatten her accent. A raciolinguistic perspective tells us, however, that no matter what Mowahib did with her voice, she would still be perceived as displaying linguistic inadequacies by powerful white listeners, because raciolinguistic ideologies are never just about language. Despite attempts to police and censor her work, Mowahib continued to see herself as a language activist and drew on her expert knowledge of critical linguistics to push back against anti-Black linguistic violence that permeated her school.
The point here is not to prove the 'efficacy' of Thea and Mowahib's initiatives, for that risks using the same crude logics that underpin the what works agenda. Instead, the what works agenda is, in Johnson, Bryan and Boutte's (2019) words, a dangerous and insidious backdrop against which anti-Black violence can be produced and maintained. The point here is to critically interrogate the colonial histories and structures which continue to shape what counts as normative and legitimate 'evidence' in language education and policy and to show different ways that educators might begin to apply that critical interrogation to their own anti-racist practice. Somewhat ironically given the 'researchinformed' agenda of contemporary education policy making, both Thea and Mowahib were highly qualified, expert teachers of their subject who actively read new research in educational linguistics, yet repeatedly had their expertise called into question when they started to challenge the white status quo. Their initiatives sought to denaturalise the durable discourses of linguistic deficit that permeate current policy and, in doing so, shifted the focus of attention away from their students' stigmatised language practices and towards political struggles against oppressive language ideologies. In the ongoing struggle against anti-Black linguistic racism, it is only this type of practice that can be counted as what works.

Notes
1. The EEF are one of the main propagators of the what works agenda in schools, and since 2013, have been one of fourteen members of the UK Government's What Works Network. Funded by the state, the network includes members from a range of policy areas such as social work, policing, public health, and youth offending. 2. The Core Content Framework is the state-designed 'minimum entitlement' for pre-service teachers, heavily policed by Ofsted in its inspection of initial teacher education providers.
Others (e.g. Smith and Lander 2022) have shown how the framework is an actively deracialised document in its notable absence of content around race and racism.