Being ‘the lowest’: models of identity and deficit discourse in vocational education

ABSTRACT This paper critically analyses and aims to denaturalise models of identity that circulate in discourse about vocational education in the Netherlands. It is argued that discourse about the vocational track is characterised by a pervasive focus on deficits, framing vocational education as unprestigious, and its students as unintelligent and insubordinate. The analysis focuses on three levels at which this model of identity circulates and is reproduced: it is rooted in the historical emergence of tracks in the Netherlands, is re-enforced throughout the educational trajectories of students in the vocational track, and is reproduced on the event level in routine interactions among students and teachers. The paper contributes to existing scholarship on the sociocultural and personal dimensions of tracking, which predominantly comes from studies based on survey and interview-based data, by building on data from ethnographic fieldwork and participant observation.


Introduction
This paper critically analyses, and aims to denaturalise, models of identity that circulate in discourse around tracking in the Netherlands, generally, and vocational education specifically. In the Netherlands, children are sorted into different educational tracks according to their supposed academic ability upon entering secondary education at the age of twelve. This system of tracking (sometimes called 'streaming' or 'ability grouping') is designed to enable schools to cater to the needs, abilities, and interests of each child. Teaching groups with homogeneous learning styles and paces is often considered easier for teachers and is intended to enable schools to prepare children and adolescents for different educational and professional trajectories (Van Houtte, Demanet, and Stevens 2012). However, much research has demonstrated the adverse effects of tracking. Highly stratified educational systems increase inequality as not everyone has equal opportunities to thrive in the system (van de Werfhorst and Mijs 2010). Access to the different tracks does not depend on academic ability alone. In the Netherlands, children with disadvantaged socio-economic backgrounds and children with migration backgrounds are at a higher risk of being under-referred, i.e. placed in 'lower' tracks than expected as based on school performance, than children with more advantaged backgrounds (Inspectie van het Onderwijs 2018; van den Bulk 2011; van de Werfhorst and Mijs 2010). Highly stratified educational systems furthermore increase inequality since they cause learning results to diverge, whereas more integrated systems reduce differences (van den Bulk 2011; van de Werfhorst and Mijs 2010;Oakes 2005).
Learning is not only a cognitive, but also a sociocultural process, however. Dividing children and youth into different tracks on the basis of their (supposed) ability at age twelve creates a social reality in which tracks are associated with social types not only cognitively, but also in terms of behaviour and personality (cf. Tarabini, Curran, and Castejón 2022). A pervasive hierarchy of prestige, in which working with the 'head' is considered more valuable than working with the 'hands', results in vocational education having an image of being a lesser form of education than academic education (van Daalen 2010). Vocational education suffers from stigma in many societies (e.g. Chankseliani, Relly, and Laczik 2016;Spruyt, Van Droogenbroeck, and Kavadias 2015). Students internalise those value hierarchies: international studies show that academic track students consistently show a more positive self-image than vocational track students (Van Houtte and Stevens 2008). Academic track students in van den Bulk's (2011) study saw themselves 'at the centre of society' and expected to become part of an elite, whereas vocational track students more often emphasised the long road in front of them to reach a 'higher' position. Youth allocated to high status (high performing) groups in school understand themselves as able to make valuable contributions to class, whereas low status students do not (Verhoeven, Poorthuis, and Volman 2019). They consider themselves, and are considered by others, to be intelligent, diligent, compliant, and sociable; and vocational track students are seen as daring, rebellious, lazy, with defective language (Knigge and Hannover 2011;Jonsson and Beach 2013). Consciousness and internalisation of stigma has furthermore been suggested to have deleterious effects on school engagement: It is correlated with sense of futility, defined as a fatalistic image of the future and feeling that one's efforts in school will not pay off (Knigge and Hannover 2011;Spruyt, Van Droogenbroeck, and Kavadias 2015;van Daalen 2010). Sense of futility, in turn, has been found to strongly influence school involvement and motivation, and is associated with school misconduct and delinquency (Van Houtte and Stevens 2008; Demanet and Van Houtte 2011).
An educational system that categorises youth at a young age according to academic ability thus not only creates inequality, but also puts in place a social reality in which different tracks are organised into a hierarchy of prestige and are associated with particular kinds of people. The aim of this paper is to trace the origins of models of identity related to educational tracks in the Netherlands, and to show, on the basis of nine months of ethnographic fieldwork in a Dutch secondary school, how these circulated among a group of students in the vocational track that is frequently referred to as the lowest track of the Dutch educational system. The paper is organised as follows. First, I discuss Wortham's (2006) theory of social identification in the context of education, which forms the theoretical framework that guides the analyses. Following Wortham, three levels of analysis are used to examine the emergence and circulation of models of social identity: the sociohistorical, local meta-pragmatic, and the event level. After discussing the methods by which the data were gathered, those three levels organise the analysis sections. The section on the sociohistorical level builds on secondary literature, while the sections about the local metapragmatic and event level build on empirical data gathered through ethnographic fieldwork. In the discussion and conclusion, I review the most important findings and make some suggestions about how changing the discourse surrounding vocational education could lay a basis for an educational system in which 'all tracks are created equal'.

Models of identity on the sociohistorical, metapragmatic, and event level
The analyses presented in this paper are guided by Wortham's (2006) theories of social identification in the context of education. Although a common understanding is that academic activities in school (mainly) contribute to learning, and social identification happens (mainly) during non-academic activities, Wortham (2006) argues that both are inseparable parts of a process of 'making up people' and contribute to social identification. Assignments that students receive, the way in which material is explained, or task division in the classroom might be primarily intended to bring about learning, but also constitute messages that make up a model of identity. A model of identity is defined as 'either an explicit account of what some people are like, or a tacit account that analysts can infer based on people's systematic behaviour toward others' (Wortham 2006, 6). In identifying students as 'exemplary', 'disruptive', 'fast' or 'slow' in specific events, Wortham (2006) argues, teachers draw on more widely circulating models of identity. In this paper, models of identity are located within discourse, which is defined as 'an organisation of talk or text that does something, in the broad social world, or in the immediate interaction, or in both' (Antaki 2012, 431).
Models of identity develop historically, over decades or even centuries. For instance, Foucault (1966Foucault ( , 1977 shows how, in the last few centuries in Europe, powerful institutions such as the government, the church, and the school, have developed ways of classifying people as, e.g. normal or deviant, or (as is typical in education) as slow, fast, able, or unable. In the process of development and ongoing circulation, such modelswhich are historically embedded and culturally situatedcome to be seen as natural and inevitable ways of classifying people: some people just are able, and others are less so. Analysing their historical emergence, analysts such as Foucault denaturalise these models of classification, highlighting the relations of power underlying them. Wortham (2006) takes such analysis a step further by arguing that to denaturalise a model of identity, it is necessary to pay attention not only to the sociohistorical origins of naturalised systems of classification. The sociohistorical timescale provides the basis on which models of identity that circulate today build, but identification really exists in actual events and must thus also be analysed in such events. Distinguishing between a sociohistorical model of identification and an event-level instance of identificationwhich is common in many analyses that build on macro-micro analysesstill does not suffice, however, since 'any phenomenon is constrained and made possible by processes at several disparate timescales' (Wortham 2006, 9). We can look at timescalesdefined as 'the spatiotemporal envelope within which a process happens' (Wortham 2004, 166)ranging from minutes within a specific interaction, to decades of an individual's life. It depends on the empirical data which timescales are relevant to analyse social identification.
In the analysis sections of this paper, three timescales are used to critically analyse and denaturalise models of identity related to being a student in the vocational track. Firstly, I use secondary literature to trace the sociohistorical origins of the tracking system, and specifically, to critically analyse the background of the value hierarchies and models of identity related to the different tracks. Then, zooming in on the empirical data gathered among vocational school students, I analyse a middle-level, metapragmatic timescale by discussing the educational trajectories of the students. Here, I argue that socially and historically embedded value hierarchies manifested as a discourse of deficit around their enrolment in vocational education. Finally, analyses of interactions at the event-level, between students as well as teachers and students, highlight the enactment and continuing circulation of models of identity in specific events of identification in daily interaction. In this last analysis paragraph, interactional analysis inspired by membership categorisation analysis (Hester and Eglin 1997) is used to argue that, in talk-in-interaction, students and teachers routinely associated the category of vocational track student with the characteristics of being unintelligent and insubordinate.
Analysing the circulation and reproduction of the model of identity on those three timescales enables me to show that the tracking system, and the relative position of vocational education, is not inevitable or natural, arguing that it can have harmful consequences to the students, and that there are alternative ways to conceptualise educational tracks that could better serve all students.

Methods and data
The data on which two of the three analysis sections of this paper are based were collected during nine months of ethnographic fieldwork with the students of one school class ('class 3/4b') between January 2017 and March 2018 in a secondary school in the middle-sized Dutch city of Venlo. 'South High School' was one of the larger secondary schools in the city, catering to around 1500-2000 students. The school offered all tracks of secondary education, from vocational to academic. I joined class 3b (who became '4b' the next year), a class with students of the least prestigious ('lowest') track in the entire system: the vocational track basis. During their vocational instruction time, as is common in many Dutch schools, the basis students were in class together with students of another, slightly more prestigious vocational track (kader) who had chosen the same specialisation. I went to school with them for 2-4 days a week, for approximately five months during their third year, when they were around 15 years old, and for four months in their fourth (and final) year, amounting to a total of approximately 333 class hours.
In total, I met 37 students, spending most time with a subset of 23 students as I followed courses with them in their specialisation 'Care and Well-being'. There were 13 students of the vocational sub-track basis in this course, and 10 students of the sub-track kader, and there were 6 boys and 17 girls in this group. About half of the students in Care and Well-being had a migration background, with mostly their parents having migrated to the Netherlands. I once overheard a student remark on a relation between migration background and tracking, 1 but as this was not part of regular discourse about tracking among these students it is not elaborated upon in this paper. For extensive analyses of the role of ethnicity in these students' interactions, see van de Weerd ( , 2020. Around two-third of school hours consisted of vocational classes, which were characterised by an informal atmosphere, independent (group)work, and much occasion for informal interaction. Therefore, I got to know these students quite well. They came to orient to me, in their words, as 'a kind of friend, or classmate, but different'. Nevertheless, there were many differences between us: I was about ten years older than them, (audibly) came from a different part of the Netherlands, and my role in the school was that of a researcher rather than student. Students knew that I worked at the university, and sometimes made comments such as 'my mother would be so proud if I went to university'. They thus oriented to the difference in our educational background, but as it was one of many differences, it was difficult to gauge how much this specific one affected our interactions. Even though there were undeniable power imbalances between us, once they had experienced that I did not act as an authority in the school and would not reprimand them, they did not show much preoccupation with my differential status in the school.
The data produced consist of fieldnotes and about 140 h of audio-recordings of classroom interaction, recorded with a lapel microphone that I wore attached to my clothing. Initially, the focus of this study was not on discourse surrounding the vocational track. Rather, my initial focus was on the role of diversity and multilingualism in the classroom (van de Weerd 2020). I informed the students of those interests at the start of the fieldwork. Alongside those topics, my interest in tracking and educational hierarchies emerged as students' characterisations of themselves (and teachers' characterisations of students) often included an appraisal of their educational level. I thus started noting down these comments and tagging them in fieldnotes and recordings. Interactions about educational hierarchies or students' own educational level were described in fieldnotes and/or recorded and subsequently transcribed. The data were coded inductively in NVivo, and then analysed using discourse analysis (as in the section 'Metapragmatic timescales of identification') and interaction analysis (as in the section 'Models of identity on the event level').
The fieldnotes and excerpts analysed in this paper have been selected for being illustrative examples of discourse and interactions that I heard regularly during the fieldwork. They come mostly from informal, casual, non-plenary interaction among students, between students and me, or between students and teachers during class. Since my research focus was not on tracking at the time of the fieldwork, I did not explicitly ask for students' opinion on their track, nor were they aware that I developed this interest. Aside from rare moments where I asked about pupils' enrolment, I also did not regularly talk to teachers about tracking. The fact that the topic often came up spontaneously, however, shows that it was a central concern in this setting.
This study was reviewed by the Ethical Review Committee of Maastricht University and approved under case code ERCIC_001_04_01_2017. All participants have provided informed consent and their names have been pseudonymised.

Timescales of identification: tracked identities in the Netherlands
The sociohistorical model: the present and past of tracking in the Netherlands When analysing the current Dutch educational system and the models of identity related to tracking within it, it is necessary to first review shortly what this system looks like and how it originated. The Dutch tracking system is characterised by early selection and strict differentiation between tracks. Around the age of twelve, when children enter secondary school, they are referred to a track based on their primary school teacher's evaluation of their abilities and effort, and their score on a standardised test. The tracks into which children can be sorted are (in order of prestige, low to high): vmbo, which is subdivided into the four tracks basis, kader, gemengd, and theoretisch, which differ in time spent on theoretical versus practical training (vocational education, four years), havo (higher general education, five years), and vwo (pre-university education, six years). These tracks give access to different forms of higher education: the tracks in vmbo gives access to different levels of upper vocational education, a havo diploma to universities of applied sciences, and a vwo diploma gives access to university. Although in principle it is possible to switch between tracks during secondary school, in practice especially upward mobility (i.e. to tracks perceived as more challenging) is difficult because of curriculum differentiation, variation in quality of education, and furthermore, because schools have financial incentives to deter those at risk of failing final exams (Mijs and Paulle 2016;van de Werfhorst, Elffers, and Karsten 2015). Over half (53.8%) of all secondary school students obtains a vmbo diploma, 26.7% a havo diploma, and 19.5% a vwo diploma (Inspectie van het Onderwijs 2020). Generally, havo and vwo are considered to be for 'good' students and vmbo for 'slower' students (van den Bulk 2011).
The defining features of today's educational systemselection into highly differentiated tracks around the age of twelve, and the differentiated valuation of those tracks -find their roots in the developments of that system from the nineteenth century onwards. Until about the mid-nineteenth century, only primary education until the age of twelve was compulsory, which meant that only upper-class children attended secondary education, which focused on academia and morality (van Daalen 2010). With industrialisation in the second half of the nineteenth century, however, the demand for skilled workers grew and primary education did not suffice to prepare youth for many jobs. A new form of secondary education was thus created explicitly for the middle classes. Some well-off members of the working classes attended a form of vocational education organised by municipalities in collaboration with workplaces and companies. This was not compulsory, though, and meant delaying entrance to the labour market, and following such education thus constituted a sign of distinction among the working classes (van Daalen 2010). In the early twentieth century, compulsory education was expanded, making vocational education increasingly accessible and attended, though it was still aimed primarily at working class youth (van Daalen 2010). Vocational schools lost the element of distinction and came to be seen as education for the poor. The middle classes had their own form of secondary education (Middelbaar onderwijs) and the elites attended the gymnasium. Until well into the twentieth century, social origins thus largely defined the length and content of a child's educational path (Bakker, Noordman, and Rietveld-van Wingerden 2010).
After the second World War, the influence of social origins on educational careers was put into question. It came to be seen as a waste of potential talentspecifically cognitive talentthat it was primarily elite youth who entered the gymnasium and then university. In the 1960s, all-encompassing educational revisions were implemented, one of the goals of which was to make achievement more important than social class in determining a child's education (Dronkers 1993). Compulsory education was further expanded, and it was decided that all children should be together in the first year(s) of secondary school, which would grant them more time to decide which form of further education to pursue. After those first joint years, children would be divided into different forms of education, which were formulated as tracks. Today's tracks are largely the same as then (except for some changes to their names and minor aspects of structure): a vocational track (today's vmbo), higher general education (havo), and the pre-university track (vwo).
In the developments from the nineteenth century onwards, then, a line can be traced where today's pre-university education stems from the historical schools for the elites, which focused on academic knowledge and preparation for university. Similarly, today's vocational education finds its roots in historical education for the working classes, which prepared youth for the (manual) labour market. The tracks are no longer explicitly aimed at the working classes, middle class, or the elite, but the logic behind the system, much of the discourse around it, and the unequal distribution of students of different social classes in it, strongly evokes that classed system of the nineteenth century and earlier (van de Werfhorst 2019). The notion that working with the 'head' and working with the 'hands'i.e. theoretical versus practical, knowledge versus skills, or academic versus vocational educationare clearly distinguishable, which is reflected in today's stark division between what is taught in vocational versus academic tracks, originates with that classed division of labour (and hence, education). The greater prestige associated with working with the 'head' that causes those who follow an academic track to be seen as 'good' students, who are compliant, diligent and intelligent (Knigge and Hannover 2011;Jonsson and Beach 2013) can also be traced back to when theoretically-focused forms of education were reserved for the elites. Vocational forms of education on the other hand were, and still are, considered of less value: vmbo is often framed as the 'drain' of Dutch education, where rowdy, difficult, undisciplined and untalented pupils end up (van den Bulk 2011; van Daalen 2010). The characteristics associated with students in different tracks evokes not only a discourse of (in)ability related to different forms of education, but also one of historically emergent and culturally situated moral evaluations of different socioeconomic classes (Skeggs 2004).
This section's examination of the model of identity surrounding tracking and vocational education on a sociohistorical timescale, then, argues that the tracking system, though often considered an inevitable way of providing youth with diverse skills and abilities with a suitable education (Van Houtte, Demanet, and Stevens 2012), finds its roots in class-based thinking of the nineteenth and early twentieth centuries. This model of identity is naturalised to such an extent that it is seen as inevitable that students should be categorised in some way (e.g. Horn 2007; Puckett and Gravel 2020). Tracking and the subjects produced by it are part of a 'formidable discursive hegemony', as Francis et al. (2017) argue. The next sections illustrate how this discursive hegemony, and the resulting social identities and value hierarchies, find their way into students' lives through their experiences in the educational system.

The metapragmatic timescale of identification: the vocational track as 'the lowest' in educational trajectories
The model of identity that positions vocational education students as following the least prestigious and valuable kind of education, being unintelligent and, moreover, carrying associations of nonconformism and insubordination, circulates on a metapragmatic scale, that is, in discourses that these youth encounter throughout their lives in the educational system. The hierarchy of prestige that positions academic tracks as 'high' and vocational tracks as 'low' could be observed particularly well in the way students and their teachers of class 3/4b discussed the educational trajectories that had led them to their current enrolment. None of the students had experienced a selection into their track because of a particular talent that fit the vocational profile, say, for their cooking skills, knack for mechanics, or caring nature. Rather, for many students, enrolment had been the result of a rather explicit process of elimination, or what in the Dutch educational system is often referred to as 'demotion' or 'descending'. For instance, Nikki had started secondary school in havo (senior general education) and had been 'demoted' to vmbo theoretisch (the most theoretical of the vocational tracks) in her second year. She did not pass her third year in that track, after which she had continued to vmbo kader, where she was enrolled when I met her. Telling her friends about all those track changes, Nikki said: 'You know the thing? I started at havo, and this is going to sound really dumb, but I did this intelligence-test last summer and it turns out my school level is actually basis. So I'm now doing a high level considering my intelligence'. Tracking decisions were communicated to students as focusing on a limited set of skills, a conception of (un)intelligence, which for these students resulted in a deficit discourse that they themselves readily reproduced in talking about themselves.
Other students of class 3/4b had immediately been put into the basis or kader track. This was generally because, at the end of primary school, they had been judged as not having the necessary skills for any of the other tracks. Yildiz, for instance, who had been referred to vmbo basis at the end of her primary school, mentioned that she had not been allowed to go to the kader track even though she had 'even cried [to the teacher]'. Enrolment in the vmbo basis track felt to her like a failure. Like several students of class 3/4b, Yildiz argued that she could have done 'better', if only her teachers had believed in her or if she had worked harder. Teachers also frequently lamented that some students 'could have done better if they only tried'. Doing betterin the words of teachers, as well as studentsalmost without exception referred to being enrolled in other, 'higher' tracks. Being in the basis and kader tracks was thus often constructed as a sign of deficiency, without considering performance in that track. For those who had started in a different ('higher') track, the framing of each next track as 'lower' and each switch as a demotion, reinforced the idea of enrolment in vmbo basis and kader as a last resort.
The behavioural dimension discussed earlier, in which the vocational track is associated with insubordinate and undisciplined students, was also present in (discourses around) tracking decisions. Being a 'difficult' or 'lazy' student could also be a reason for being tracked into vocational education, as illustrated in Excerpt 1.
Excerpt 1: Too smart to be here Fieldnotes 16 June 2017 -I ask Ms. Jansen, who has known many of the students of class 3/ 4b for a few years, about the vmbo basis enrolment of a few students. She starts discussing individual students: 'Ahmed started in basis/kader, I think. And he just descended at some point, just because he just didn't try hard enough, you know? Handing in assignments, that kind of stuff, planning. So actually, he has the capacities, but they don't come out. And that's a shame, really. And that is of course also the case for a student like Khadija, she also started in kader, and she is much too smart for here. Jay as well. But he just doesn't do anything, you can't get him to … Stefan too, actually also a kader student'.
In this exchange, the teacher describes some students enrolled in the basis track that she considers do not fit there. The main associated characteristic of a vocational education student in the basis or kader track, in her explanation, is a lack of capacities. However, she also mentions that not putting in the effort is, for some students, the reason that their capacities 'don't come out'. Although she does not mention it explicitly, Khadija and Jay were by many teachers seen as (behaviourally) difficult, though capable, students.
The model of identity that emerges is one of vocational track studentsparticularly those in the vmbo basis track, the least prestigious of all tracksas lacking the skills that would be necessary for enrolment in other, more prestigious, tracks. Students' educational trajectories were filled with moments in which they were made clear that their track was the undesirable outcome of the sorting process. Decisions on track assignment, and changes between tracks, were made on the basis of a limited set of skills, namely (a perception of) ability or intelligence. Where a student stood on that single axis of abilitytogether with their behaviour in schoolwas framed as defining their place in the system. This focus on a limited set of skills led to a strong deficit discourse: tracking placement, especially when leading to vocational education, primarily focused on the skills or behaviour that students lacked. Aside from the harmful consequences this has been shown to have to students' image of self and motivation in school (e.g. Knigge and Hannover 2011;Spruyt, Van Droogenbroeck, and Kavadias 2015), this discourse also obscures the fact that track enrolment is often influenced by causes other than students' individual skills (such as socio-economic background) (Gamoran 2010;van de Werfhorst and Mijs 2010).
Models of identity on the event level: 'we're the lowest' Models of identity positioning students of the vocational tracks as unintelligent and insubordinate in the context of Dutch schooling, generally, and class 3/4b specifically, translated into this educational category being routinely equated with those characteristics in daily interactions (i.e. the event level) among students and teachers. Merely mentioning the category 'vocational school student' in an interaction could be enough to conjure up the image of someone unintelligent and insubordinate. The category could 'stand in' for those characteristics as well as vice versa: the characteristics would call up the category in talk-in-interaction.
For instance, one day in maths class, the teacher assigned study materials for an upcoming test. The students perceived the material to be too much and too difficult, and several of them called out: 'Remember that we're not havo, we're vmbo, we're the lowest!' In this interaction, reminding the teacher of their membership categoryof 'being vmbo'was designed to convince the teacher to adapt their study materials. There are many different reasons on which students could have drawn to achieve such results, like that it was simply too much, that they had other tests coming up, that there would not be enough time to study. However, drawing on their supposed inability as evidenced by their educational level was a reason that students frequently called upon. The association between educational categories and intelligence was also evident when Amira condemned the behaviour of an acquaintance at school, who was in a 'higher' track. She remarked: 'She's so dumb, and she is in [vmbo] theoretisch!' Being in vmbo theoretisch, the 'highest' sub-track of vmbo, was constructed by Amira as inconsistent with the 'dumb' behaviour of her peer.
Such reasoning was confirmed by teachers, who also made ready associations between ability and educational level in interactions with students. For instance, in the vocational training classes, which were shared between the students of the two vmbo sub-tracks basis and kader, all students had to take a weekly test of concept definitions. Whereas the kader students were expected to reproduce the exact phrasing of the definition as provided by their textbook, the students enrolled in basis were allowed to write a definition in their own words, which was said to be easier. Even in assignments that were the same for students of basis and kader, teachers regularly emphasised to the students that they had different expectations for students enrolled in the different tracks. For instance, when Leila (a student enrolled in kader) finished an assignment and asked her teacher what to do next, her teacher responded: 'You did a good job, but as you know there are people in this class who are basis and others who are kader, and those in kader will be a bit quicker in finishing assignments'. This scene happened in a group setting, where also basis students were present and thus overheard the teacher's differential expectations for them in comparison to the kader students.
The other dimension of the model of identity of the vocational school student mentioned in the previous sectionsthat of (im)morality and (non)conformismwas even more evident on the level of categorisation in talk-in-interaction, especially among students. During a class in which the teacher did not manage to settle down the group, Amira remarked to her classmate: 'He is used to havo students. Not us, vmbo'. The image of rowdy vmbo pupils was also expressed to me by Omer, shortly after I had intervened in a discussion between him and his classmate Amine, which I perceived to be on the verge of turning into a physical fight. Not long after the incident, I sat at a table with Omer and his friend Ben. The situation and conversation are described in the following excerpt.
Excerpt 2: We're not like those university boys 27 February 2018 -Omer has been talking much of the morning about Amine, who texted him last night about something that Omer considered unnecessary (i.e. to ask which sandwich he should buy Omer the next day, as part of an agreement where Omer gave Amine extra change for fetching him lunch). I ask Omer why he keeps talking about something that, to me, does not seem important enough to start a physical fight over. Omer says I just don't understand, to which I agree. In this exchange, Omer positioned himself as part of an unspecified group using 'we' and compares that to university boys. The category 'university boys' can be understood to be designed for my understanding, as I was known in the class as affiliated with the university. Going to university, furthermore, was not considered by these students to be a likely part of their future, as their (expected) secondary school diploma did not grant access to university education. By using this category, then, Omer makes level of education relevant to the conversation. By virtue of the comparison, 'we' is understood to refer to vocational track boys. Omer paints an image of 'university boys' as conflict avoiding, and by comparison, vocational track students as rowdy and, 'if necessary', physically aggressive. As such, he conjures up an image of vocational students (or boys, specifically) as insubordinate and nonconforming. Again, also teachers expressed associations between categories related to educational level and expected behaviour of members of those categories. When Dounia was quietly working on an assignment in a Dutch language class where many students were chatting, her teacher quietly told her: 'You're doing great. You are an exemplary student for basis, you can be proud of yourself'. Being an exemplary student 'for basis' thus was communicated to mean not engaging in disruptive behaviour. The teacher hereby expressed expectations of what a student in the basis track could normally be expected to be like.
Data in this section indicate the contextualised enactment and situated circulation of the same models of identity as discussed in the previous sections on the level of talk-ininteraction among the students of class 3/4b. Students associated the category of vocational school student with the characteristics of being unintelligent and insubordinate not only on a metapragmatic level, but also on the more fine-grained level of interactions or events. When students wanted to challenge a teacher in the assigned amount or level of study material, they built upon the model of identity of vocational education students being unintelligent. When students considered their own behaviour to be rowdy or insubordinate, they readily connected this to the category of vocational education student. They did not have to use those categories, as the topic of talk in many of the described instances was not educational level, but the fact that this was such a ready association for them shows the strength and taken-for-grantedness of the association between categories of educational level and ideas of (un)intelligence and (in)subordination.

Discussion and conclusion
This paper has analysed the origins and circulation of a deficit discourse that vocational school students in the Netherlands contend with in their experiences with the educational system generally, and with tracking specifically. Using three timescales of identification (Wortham 2006), I have analysed the origins of this deficit discourse at the sociohistorical timescale, as well as its continuing circulation in the daily experiences of vocational school students.
Analysis on the sociohistorical timescale indicate that models of identity related to tracking emerged in the history of the Dutch educational system, where, until the mid-twentieth century, educational paths were strongly linked to social class. Elites engaged in theoretically oriented education, and education focused on skill development was oriented to the working classes. A redesigning of the educational system in the mid twentieth century laid the bases for today's system of tracking in which achievement would be more important than class in defining a child's educational trajectory, but parents' socioeconomic background is, today, still highly influential on children's achievement (Dronkers and Korthals 2010). More importantly, reforms did not change the hierarchy of prestige attached to different forms of education: theoretically oriented education retained its higher status, and vocational education its inferior image. Schools still focus and reward primarily that prestigious, but limited, set of skills that allows one to engage in theoretically oriented education.
Based on ethnographic fieldwork with students of the two most vocationally oriented tracks of vocational education, I argued that the focus on that limited set of skills is particularly pervasive in the tracking system, where the skills that students are perceived not to possess determine their track placement. This leads to a pervasive presence of a discourse of deficit in these students' experiences with the educational system. For several students, the individual trajectories that led to their current enrolment were marked by changes between tracks characterised as descension, communicating to them that their enrolment was the undesirable outcome of the sorting process. Mere enrolment in the vocational track was framed as a last resort, and many students considered it a sign of failure without considering performance in their track. I furthermore discussed how students often brought up the associations of lack of intelligence and insubordination related to their educational level in daily interactions at school.
The focus on a limited set of valued skills clashes with an ideal image of tracking in which 'all tracks are created equal' underlying the often-expressed reasoning behind tracking, stating that all youth simply have different skills and interests, and should be prepared for different educational and professional trajectories. As I have argued in this paper, though, different tracks are, in practice and experience, hierarchically ordered. The perceived value and prestige of vocational education is thoroughly embedded in a historically developed cultural construct that distinguishes between working with the 'head' and working with the 'hands' (Nylund, Rosvall, and Ledman 2017;Tarabini, Curran, and Castejón 2022;van Daalen 2010) and values the former over the latter. It is also linked to the quality of this type of education and its prestige and reward in the labour market. The data presented in this paper have also pointed to the role of metapragmatic and event-level models of identity in reinforcing and reproducing sociohistorically embedded models of identity, however. Even if it is impossible to change the historical legacy of education aimed specifically at elites or working classes, which is at the basis of this hierarchy of prestige, a broad discourse change on the metapragmatic and event levels could reframe the relative value of tracks today, and particularly the position of vocational education. A broadening of the definition of educational success to encompass more than cognitive skills would be a possible step in the direction of modifying models of identity on the metapragmatic and local level (van Daalen 2010; Wheelahan 2015). If we are to stimulate children to pursue the education (and career) that suits them best, and to enable all students to thrive in the system in their own way, it is not only necessary to ensure equal opportunities for all but also to strive for true equality within difference. If tracking policies are to be maintained, it is important that track a is considered (and communicated as) suitable for students who are good at x, and track b for students who are good at y, instead of the current situation in which track a is presented as suitable for students who are good at x, and track b is for students who are considered bad at x. Furthermore, praising effort and improvement in school rather than ability, and rewarding curiosity rather than knowledge, could enable all students to find motivation in education (Covington and Teel 1996). In a system where the focus is on a broader set of skills rather than on deficit, no student would have to speak of themselves as 'the lowest'.

Notes
1. Talking about her primary school teacher's tracking advice, Yildiz commented: 'My school did not allow me [to go to a different track]. (…) My teacher was a bit of a racist. All the Dutch kids got higher advice than the foreigners. All the foreigners at this school are in basis and kader'.