Researching race, racialisation, and racism in critical terrorism studies: clarifying conceptual ambiguities

ABSTRACT

This article seeks to ameliorate the conceptual ambiguities surrounding the concepts of race, racialisation, and racism within Critical Terrorism Studies (CTS), as well as encourage the use of these concepts. Race is a key signifier in counter-terrorism discourse, yet its meaning is often taken for granted or left unexplored in CTS. Hence, this article proposes definitions of race, racialisation, and racism that make the concepts into an analytical lens. In developing these definitions, it employs a marginally adapted version of Gerring’s criterial framework for concept formation. The article argues that the core attributes of the concepts race, racialisation, and racism are: (A) their socially constructed nature; (B) categorisation of a group as the “other”; (C) naturalisation; and (D) being concerned with community-like groups. The proposed definitions build on similarities and navigate differences in how the concepts are currently given meaning in CTS research. Considering that CTS is embedded in the context of 9/11, the concepts of race, racialisation, and racism may help to place counter-terrorism discourses and practices in a broader historical and political context and analyse the justifications supporting these discourses.

KEYWORDS:

• Race
• racialisation
• racism
• critical Terrorism Studies
• anti-Muslim discrimination
• concept formation

Introduction

The proliferation of counter-radicalisation and counter-terrorism research, policies, and practices has seen an equally growing interest in their critical investigation. One prevalent critique in these studies is that counter-radicalisation and counter-terrorism discourses and practices severely discriminate against Muslims (see Kundnani 2015). Several scholars have started to connect this discrimination of Muslims to racism. This link has been examined and argued for significantly more by authors publishing in other fields and journals, such as Patterns of Prejudice and Ethnic and Racial Studies (e.g. Anidjar 2007; Balibar 2004; Essed and Goldberg 2002; Goldberg 2006; Lauwers 2019; Meer 2013; Meer and Modood 2009; Topolski 2018; Modood 1992). These authors applaud the use of the concept of race as an analytical tool.¹

However, Critical Terrorism Studies (CTS) tends to avoid the concept, including in research on closely related topics, such as visibility (Martin 2018), notions of risk (Heath-Kelly 2012a, 2012b; Amoore and De Goede 2008; De Goede 2008), and the suspect community thesis (Awan 2012; Breen-Smyth 2013; Mythen 2012; Pantazis and Pemberton 2009; Taylor 2018). Adding the concept of race to these analyses can make visible the wider context in which counter-terrorism discourses and mechanisms operate and help to uncover the justificatory logics that form their foundation. Moreover, the publications in CTS that do employ the concept of race and elaborate on the definition they use, vary in their interpretations, mirroring debates regarding the concept in other fields (Murji and Solomos 2015). For example, while some believe race refers only to discrimination on the basis of phenotypical characteristics, most often skin colour (e.g. Hussain and Bagguley 2012), others argue it has to do with a process of essentialisation that could apply to any group, including those delineated by religion or culture (e.g. Crawford 2017).

The ambiguous use of the concept in CTS, though understandable, can hinder the recognition and analysis of racial logics in counter-terrorism and counter-radicalisation discourses. As such, this article aims to ameliorate the current conceptual ambiguity in CTS by proposing definitions of race, racialisation, and racism. In doing so, it aims to facilitate taking up race as a conceptual tool and thereby encourage its use in CTS research.

Accordingly, the first section discusses the current state of work in CTS on race, reflecting on how often the concept of race is used, potential reasons for the limited engagement with the concept, and what a more consistent and increased use can bring CTS. Section two then dives into CTS work where the concept of race is being used, providing an overview of the main similarities and ambiguities in definitions. It connects these findings to debates on the concept of race in related fields in order to clarify the background of the definitions and to contextualise the specific requirements for the concepts of race, racialisation, and racism within CTS. The third section engages with concept formation literature to discuss the factors that need to be taken into account when developing concepts for social scientific research, focusing on Gerring’s (Gerring and Christenson 2017) criterial framework.

Having laid the groundwork, the fourth section proposes definitions of race, racialisation, and racism that are useful for CTS scholars. It argues that the core attributes of the concepts race, racialisation, and racism are: (A) their socially constructed nature; (B) categorisation of a group as the “other”; (C) naturalisation; and (D) that they are concerned with community-like groups. Whereas races are the socially constructed groups, racialisation is the process by which these races are constructed and racism is negative discrimination on the basis of ideas regarding races. Consequentially, the fifth section discusses how these definitions follow the guidelines of Gerring’s (Gerring and Christenson 2017) criterial framework, explaining the trade-offs that were made and the main limitation and objection to the proposed definitions. Besides reiterating the main points of the article, the conclusion indicates further venues within CTS in which race, racialisation, and racism can play a promising role.

Using the concept of race in CTS

Before proceeding, let me clarify what I understand under CTS research. In line with the vision of the journal, Critical Studies on Terrorism, I understand CTS to entail research that investigates aspects of terrorism, counter-terrorism, and state terror. In this perspective, “critical” denotes a recognition of the inherently problematic nature of the terrorism label and a questioning of accepted orthodoxies (“Aims and Scope” n.d.). The main research agenda of CTS is to reveal and challenge “the ideological function served by much core terrorism research” and counter-terrorism discourses (Jackson, Breen-Smyth, and Gunning 2009, 64; see also Jackson et al. 2018). As one of the most prominent places to publish work in CTS, Critical Studies on Terrorism provides a good indication of how often the concept of race is employed in CTS. Searching for “race”, “racialisation”, and “racism” in the journal results in two publications marking “race” and two marking “racism” as a keyword, in addition to two texts using “race” in their title (respectively, Monaghan and Molnar 2016; Innes and Steele 2015; Breen-Smyth 2013; Clini 2015; Kowalski 2016; Zalewski 2013).² Some of this absence of the concept of race is a function of space, as the main focus of much CTS literature is actual counter-terrorism measures and their impacts. But other reasons may also play a role, of which three prominent ones will be discussed. Consequently, this section discusses how using the concept can benefit CTS scholarship.

Potential objections to using the concept of race

One potential argument for avoiding the concept of race is that it may reinforce the idea that races exist and that their associated stereotypes hold some truth, despite the evidence indicating that there is no biological ground for the idea of human races. However, as Goldberg (2015) has pointed out, this silencing of race-talk leads to the problematic view that racism is no longer a problem in contemporary Western societies. Without a concept that clearly connects to the processes behind racism, it becomes extremely difficult, if not impossible, to make such processes discernible.

One may also posit that the concept of race does not fully describe what is going on in counter-terrorism discourses. Counter-terrorism discourses are complex and multifaceted; the concept of race does indeed capture only part of what is going on within these discourses. But that is a crucial part of why the concept is useful, provided it is sufficiently clear. Bounded concepts help researchers disentangle complex relations and processes (Sartori 1970). Hence, one will need more than one concept to fully describe and analyse the different parts of a discourse. Together, such bounded and clear concepts can provide a theoretical lens that allows researchers to spot otherwise overlooked logics and the relations between them.

Furthermore, one can ask why race rather than other concepts should be used to understand racial othering. Concepts such as, and related to, colonialism, orientalism and Islamophobia are incredibly useful in our analyses as critical scholars. However, they only partially overlap with the concept of race, are less specific regarding the form of othering, or, inversely, focus on one specific manifestation of racial othering in comparison to the concept of race. As such, the concept of race can add to analyses using these concepts by providing more depth or a common ground for links to other manifestations of othering.

Additionally, the concept of race, and especially racism, indicates that the phenomena labelled as such are rightly seen as moral wrongs and should actively be resisted and challenged (Blum 2002). Race, racialisation, and racism are morally and politically loaded terms. They have significant potential to stir debates and start the breaking down of institutionalised hierarchies. This in not only obvious in real-life politics, demonstrated, for example, by the contemporary anti-racism protests spurred by the death of George Floyd, but also in academic debates, as is shown by the current dispute between Wæver and Buzan (2020) and Howell and Richter-Montpetit (2019). By using these concepts to analyse racial logics and structures, scholars thus make explicit their criticism of these phenomena. Alternative concepts, such as stereotyping, essentialisation, identify formation, and ethnicity do not carry the same moral and political weight. Naturally, this means that scholars ought to be careful when labelling phenomena racial/racist. But rather than being a reason to avoid using the concept, I maintain this should be seen as an opportunity to bring the appropriate weight to one’s analyses.

Reasons for using the concept of race

The concepts of race, racialisation, and racism can help to make visible, properly analyse, and contextualise racial logics, while furthering the critical research and political agenda of CTS. First, a racial lens helps to make discourses and practices visible as forces that potentially reproduce racist logics, even when the word race is not explicitly used (Goldberg 2015). For example, Dixit (2014) employs a racial lens to identify invisibilities in the official US narrative of the death of bin Laden and to direct attention to how a shift in standpoint result in a foregrounding of other cases, identities, and meanings of the event. The concept of race can thereby contribute both to analysing the reproduction of racist logics in counter-terrorism and counter-radicalisation discourses, and to avoiding the reproduction of these logics in our own work. After all, it is only when we are able to recognise racial logics that we can evade reproducing them. Considering that the field of International Relations, from which many CTS scholars hail, has been criticised for being predicated on Eurocentric premises (Hobson 2012),³ an explicit effort to critically reflect on these premises and their consequences seems appropriate. By being geared towards making visible processes that are perpetrated by a Eurocentric perspective, the concept of race is a useful tool for such endeavours.

Second, whereas CTS is set against the background of 9/11 and the War on Terror (Jackson et al. 2018) – a context in which the concept of Islamophobia has enjoyed extensive use (Garner and Selod 2015) – race can aid to place contemporary counter-terrorism practices in a broader historical and geographical context (e.g. Medovoi 2012; Bayoumi 2006). For example, Razack (2008) traces the creation and exploitation of a climate of fear, created by western states, used to limit the fundamental rights of those categorised as threats. Race is fundamental to these practices and the story of a clash of civilisation that justifies them. Similarly, Selod (2018) uncovers, by sharing the stories of Muslim Americans, how the “War on Terror created conditions where Muslim bodies are racialized”, made into bodies that threaten security, thereby “justifying the need for their surveillance” (p. 4). Other studies lay bare similarities of counter-terrorism measures with, for example, colonial counter-insurgency tactics (e.g. Dunlap 2016), the criminalisation of black activists (e.g. Speri 2019), and nineteenth-century categorising of Muslims as violent, barbaric, and fundamentalist (Delatolla and Yao 2018).

It may be useful to compare these studies to work based on the widely used suspect community thesis to clarify the value of the concept of race. The suspect community thesis draws on the seminal work of Hillyard (1993) regarding the construction of the Irish as a suspect community as a result of several aspects of the Prevention of Terrorism Act (PTA) 1974 in Britain. According to the suspect community thesis, individuals may be singled out for state attention, not because of suspected wrong doing, but rather because of their presumed membership of a sub-group that is considered/constructed as “problematic”. After 9/11, authors such as Pantazis and Pemberton (2009) argued that Muslims replaced the Irish as a suspect community. Research inspired by the suspect community thesis has contributed greatly to the analysis of securitised groups in society as a result of counter-terrorism measures and discourses.

Still, as the work of, amongst others, Razack (2008), Selod (2018), and Rana (2011) show, the concept of race adds value to CTS research by situating counter-terrorism discourses and their related exclusionary mechanisms in a broader (historical, governmental, discursive, and potentially geographic) context. In doing so, they are able to lay bare the (oftentimes implicit) justificatory logics behind processes of securitisation, thereby making visible why the securitisation of specific groups was not vigorously challenged by society at large. Using the concept of race in reflecting on Hillyard’s work, it becomes clear how the wider racialisation of the Irish supported the justification of the PTA (see, for example, Hickman 1998).

Additionally, the use of the concept of race, as well as the moral and political weight carried by it, is in accordance with the explicit aim of CTS to challenge accepted orthodoxies. It is also in accordance with the implicit aim to help those unjustly harmed and marginalised by (counter-)terrorism discourses and practices, apparent in the way many CTS scholars frame and use their research. By using the concept of race, CTS scholars explicitly position themselves as political actors working to counter racism in counter-terrorism discourses and practices. Nevertheless, in order to do so properly, it is important to work with a clearly defined concept of race. Hence, the next section clarifies the current definitions of race being used in CTS.

The meaning of race in CTS

CTS deploys approaches from a variety of disciplines, including sociology, criminology, political science, and philosophy. Given the relatively recent establishment of the field, it is hardly surprising that scholars diverge in the way they employ, explain, and frame the concepts of race, racialisation, and racism. Nonetheless, such use does result in ambiguity and potentially in under-interrogated norms being utilised within CTS literature. To clarify what race has been taken to mean for scholars in CTS, this section begins by discussing the similarities in the meaning of the concepts of race, racialisation, and racism in work in CTS where they play a central role.⁴ Consequently, it delves into the ambiguities in the definitions, ending by specifying what is special about race in CTS.

Similarities

CTS scholars working with the concept of race often build on theoretical literature on race. This includes CTS scholars who do not explain their understanding of race in their text – usually because the focus of the work is on a different, though related, topic (e.g. Alam and Husband 2013; Choudhury 2017; Kapoor 2013; Thomas 2016; Poynting and Mason 2006). The works of Said (2003), Goldberg (2001), and Hall et al. (1978) are repeatedly referenced. Other important authors are Sian (2015), Sayyid and Vakil (2008), Fanon (1986), and Fredrickson (2002). As such, there are several assumptions that CTS scholars seem to share, namely, an understanding of race as a social construct, a predominant focus on what race does and how it is enacted rather than what it is, and a treatment of racism as a form of negative discrimination.

The first of these similarities is scholars’ understanding of race as a social construct. Races do not refer to actual categories of human beings, as was theorised from the period of colonialisation up until the period of the second World War (Fredrickson 2002; Garner 2010; Topolski 2018). Rather, they refer to the construction of such categories.⁵ This view of race has two important implications. To begin with, categorisation into races is not limited to a perception of biological differences. Instead, races can be constructed on the basis of a variety of so-called markers. For example, religious and cultural markers – especially markers of Islam and Judaism – have played a leading role in racial categories within Europe and the Americas, as is argued by both scholars working with/on theories of race (e.g. Anidjar 2007; Balibar and Wallerstein 1991; Goldberg 2006; Meer 2013; Topolski 2018) and CTS authors (e.g. Bayoumi 2006; Fekete 2004; Kundnani 2014; Rana 2017; Smith 2018, 2016). Hence, Muslims can be constructed as a race on the basis of markers commonly associated with Islam (Morsi 2017; De Koning 2016), such as wearing a headscarf or having a full beard.

Next, the way the socially constructed nature of races is interpreted in CTS means that race is ascribed, i.e., one does not have to be or self-identify as part of a racial group in order to be classified as such (Kaufman 2018). The objection that discrimination against Muslims cannot properly be called racism is in this view unsustainable. Whereas Smith (2018) points to the ways in which race has been employed historically, Crawford defends the same argument by defining anti-Muslim racism as “the discrimination faced by anyone believed to be Muslim, usually determined by ideas about physical appearance and religious custom” (Crawford 2017, 200, emphasis original). This is reflective of theoretical scholarship on race. Although there are discussions regarding the role those regarded as belonging to a specific race (can) play in the construction of this category – a discussion seminally started by the works of Du Bois (1969) and Fanon (1986) – most scholars agree that the construction of a race does not depend on the self-identification of people with such a category (Essed and Goldberg 2002).

The second perspective that is shared within CTS Studies, which ties into the previous point, is the treatment of race as a relationship rather than an identity and a resulting focus on what race does and how it is enacted. For CTS scholars, race performs specific functions within the discourse and practices of counter-terrorism. For example, in her analysis of the Australian domestic anti-terrorism programme, C. Smith (2018) focuses on how race shapes the nexus between surveillance, disciplinary measures, and carceral interventions. Similarly, Sharma and Nijjar (2018) employ race and racialisation to explain how the category of “Muslim-(as)-terrorist” comes about in contemporary surveillance practices in the UK; Rana (2007, 2017) demonstrates how race constructs Muslims as permanently threatening within the discourse of the War on Terror; and Fekete (2004, 2009) analyses the way in which contemporary security concepts are shot through with racism directed at groups perceived as foreign. In order to fully understand how the different aspects of counter-terrorism work in tandem, it is thus essential to understand the work that race performs.

Lastly, authors regard racism as a form of negative discrimination, i.e., as an unjust treatment of a certain (constructed) category of people on the basis of their (perceived) belonging to this group. The perception of race as socially constructed does not mean that it is thought to have no “real” effects. On the contrary, the construction and function of race in counter-terrorism is examined precisely because it has extensive, intrusive, and unjust consequences. This is in line with the aspiration of much theoretical scholarship on race to consider issues of power, privilege and oppression in its social, historical, and political context and, through this consideration, empower marginalised voices and perspectives (Essed and Goldberg 2002). What is important to recognise in all these similarities is the adaptability of race, racialisation, and racism. In other words, how race, racialisation, and racism manifest themselves is dependent on the context.

Ambiguities

There are a number of ambiguities in the way CTS scholars define race, racialisation, and racism as well, which can be generalised to the question of how broad or narrow the concept of race should be. Within this discussion, two ambiguities are most apparent: variations in specific characterisations entailed by racialisations (e.g. none, deviant, demonic, inferior, threatening), and disagreement about whether or not the concepts of race, racialisation, and racism should be limited to constructing characteristics as hereditary and permanent, or even to a negative construction of biological/phenotypical features – in other words, whether naturalisation is part of the concept. This section explains what is at stake in both discussions through an illustration of diverging works within CTS.

First, authors vary in whether or not they attach a specific characterisation to the process of racialisation, and for those who do, what this characterisation is. One of the authors who does not require such a characterisation is Crawford, who defines anti-Muslim racism as “the discrimination of anyone believed to be Muslim” (Crawford 2017, 200). However, this broad a definition is relatively unusual. More common is the referral to notions of superiority and inferiority. For example, Smith (2018) conceptualises race as the construction of the inferior “other” in essentialising ways. Similarly, Smith (2016) connects racism to the justification and reinforcement of perceived white superiority. According to Medovoi (2012), the codes of superiority and inferiority correspond to one of two “axes” of racism, namely “colour-line” racism. Colour-line racism has a corporeal ideological core, meaning that colour-based “categories of race proceed on an ideological basis as if the divides in the population that they produce either are or should be visibly marked as distinctions between the white and the black, brown, or Asian body” (Medovoi 2012, 46–7, emphasis original). Colour-line racism entails the view that the racial other, through their inferior body, threatens “our” whiteness, sex, and biological superiority.

The other form of racism Medovoi (2012) identifies is dogma-line racism. Dogma-line racism has a mental or psycho-political ideological root; it “proceeds as if its racial other either was or could be corporeally undetectable” (Medovoi 2012, 48, emphasis original). The dogma-line racialised other threatens through a covert war: “due to how they think and what they believe, they actively seek to infiltrate and destroy us” (Medovoi 2012, 68). Here, the codes are that of friend and enemy. This type of characterisation can be found in the articles of Sharma and Nijjar (2018) and Fekete (2004). By focusing on how the category of the “Muslim-(as)-terrorist” comes about in contemporary surveillance practices in the UK, Sharma and Nijjar demonstrate how these practices create and sustain the “the fear of an unknowable-yet-to-come terrorist threat” that is embodied by “the Muslim” (Sharma and Nijjar 2018, 82).

Similarly, Fekete (2004) argues that contemporary national security concepts have come to regard Muslim as strangers within the country, who are to be treated with continuous suspicion. In Fekete’s words, “[a]ll Muslims who practise their faith, that is, are guilty by association until proved innocent – except that the proof is never enough; their allegiance is, at best, irredeemably split and, at worst, cover for something far more sinister” (Fekete 2004, 23). Although both Fekete (2004) and Sharma and Nijjar (2018) thus include the code of friend-enemy in their concept of race, they differ slightly in their characterisation of the enemy; whereas Fekete refers to the demonisation of the racial other, Sharma and Nijjar argue Muslims are constructed as deviant.

As mentioned, scholars working on theories of race emphasise the importance of context for manifestations of racialisation and racism. As the CTS analyses discussed above focus on particular discourses, practices, and countries, it is not surprising that they use different characteristics or codes on which to map racialisation. Nevertheless, there is a difference between arguing that racialisation and racism manifest themselves in different ways depending on the context versus using specific characteristics to define these concepts. The former concerns the operationalisation of one’s concepts through the development of indicators relevant within the context of one’s study, whereas the latter concerns the conceptualisation itself (Collier and Gerring 2009). Unfortunately, many CTS scholars hitherto do not clearly stipulate whether they see the characteristics used as indicators or definitional attributes.

The second difference between CTS authors lies in whether or not they contend that the concept of race should be limited to a construction of characteristics as hereditary and permanent, or even to cultural constructions of biological features. Some authors, such as Crawford (2017), Smith (2018) and Smith (2016), employ a broader definition of race, referring to processes of essentialisation rather than biology. Within the theoretical framework upon which these scholars build, essentialisation refers to the reduction of some person, group, or category to their presumed fundamental characteristics (Goldberg 2006, 2009; Hall et al. 1978). Although manifestations of essentialisation often entail the construction of characteristics as more or less unalterable, this is not necessarily the case. In contrast, Bayoumi (2006), employing the work of Fredrickson (2002), asserts racism involves constructing characteristics as hereditary and unalterable, allowing for a distinction with religious discrimination. According to Bayoumi (2006), there is no escape from racial categorisation under racist regimes. By contrast, religious discrimination entails unjust treatment on the basis of perceptions about someone’s beliefs or practices.

As these are not perceived as hereditary or unalterable characteristics, one has – in theory at least – the opportunity to escape religious discrimination by conversion. In this framework, racialisation of religion takes place when people are perceived to have hereditary and unalterable characteristics based on markers associated with a specific religion. Hence, Bayoumi (2006) does not demand race to be the negative construction of presumed biological features; if religious characteristics are constructed as unalterable this may just as well be raced. In this respect, Hussain and Bagguley (2012) employ the narrowest definition of race, as they claim racism “entails the negative signification or cultural construction of biological or somatic characteristics” (Hussain and Bagguley 2012, 718). According to Hussain and Bagguley (2012), extending the concept of race to characteristics without biological or somatic features, such as religion, would improperly inflate it, resulting in the loss of its distinctiveness and specific empirical referents.

In general, authors in CTS scholars thus share a basic understanding of and approach to race, but differ in their conceptualisations. Both are reflective of dominant perspectives and prominent debates in theoretical scholarship on race. But how is the use of the concepts race, racialisation, and racism within CTS distinctive to the field? Naturally, due to the interdisciplinary nature of CTS as well as the intertwinement with related subjects, the definitions of race and racialisation employed in CTS are not unique. The answer to the question lies rather in the purpose for which the concepts are used: to critically investigate dominant, usually state-led, counter-terrorism discourses and practices. This means, firstly, that there is a focus on the structural features of these discourses and practices rather than individual dispositions of the persons developing and implementing them, and secondly, that the concepts are used in conjunction with concepts such as securitisation (Buzan, Wæver, and de Wilde 1998), risk (Beck 1992; see also Amoore and De Goede 2008; Aradau and Van Munster 2016) and suspect communities (Hillyard 1993). In other words, they are part of a specific type of theoretical framework to investigate a particular way of governing populations.

Before building on this overview to propose concepts that take the purposes of CTS as focal points, the next section delves into the method used to develop the concepts.

Methods of concept formation

CTS research often theorises from lived politics. This is without a doubt a valuable approach; not only does it add depth and context to analyses, it also ensures academic research is grounded in society and geared towards changing the world as we find it. Nonetheless, this approach runs the risk of under-interrogating the concepts being used, especially regarding their role in research. By reflecting on what is needed to develop an appropriate concept, this article aims to show the importance of taking into account the purposes of concepts in social scientific research projects. The methods in concept formation discussed here may thus provide CTS scholars with an extra tool in their theoretical exercises.

The section starts with a general introduction to the main aspects relevant to concept formation: a concept’s term, intension, and extension. Consequently, a number of approaches to concept formation are discussed in order to explain why this article employs the criterial approach based on Gerring’s (Gerring 1999; Gerring and Christenson 2017) classical formulation. The section further elaborates on Gerring’s work in concept formation and discusses the criteria of conceptual goodness.

Approaches to concept formation

The formation of a concept typically refers to three aspects: the term, extension, and intension (Toshkov 2016). The term, or word, is the label that covers the extension and intension. Extension refers to the phenomena to which it can appropriately be applied, whereas intension denotes the properties or attributes that define these phenomena (Gerring 1999; Sartori 1970). For example, the extension of the US legal definition of terrorism covers the flying of airplanes into the Twin Towers on 11 September 2001, the attacks in Mumbai on November 26–29, 2008, and the sinking of the Rainbow Warrior on 10 July 1985. All these cases share the attributes of being unlawful, using violence and intimidation, and being in pursuit of political aims, which together form the legal definition and hence the intension of the term. Good concepts are those that attain a proper alignment between extension, intension, and the term, thereby allowing scholars to mediate between theories and empirical observations (Toshkov 2016).

There are multiple ways to define a concept and its structure. The field of concept formation in the social sciences encompasses three main approaches: the classical, prototype, and theory approach (Toshkov 2016). In response to these approaches, Gerring (1999) developed a fourth, the criterial approach.

Under the classical view, a set of necessary and sufficient conditions are established that delimit which phenomena fall under the concept (Toshkov 2016; Goertz 2005). This tradition, seminally developed by Sartori (1970, 1975), thus leads to clear and strict definitions. However, Goertz (2005) argues that it tends to treat concept formation as an issue of definitions that are ultimately arbitrary. One could in principle define democracy as a two-legged animal with fur. This would of course be nonsense, but it does demonstrate that the approach does not require tying into the way people normally create, use, and process concepts.

In contrast, the prototype approach is based on Wittgenstein’s idea of family resemblance (Wittgenstein 1953; see also Collier and Gerring 2009). Like family members, concept members resemble each other not through a specific trait but through a variety of features that are shared by some, but not all, members of the family (Wittgenstein 1953). The prototype approach further developed this view by using the idea that people generally rate some phenomena falling under a concept as more typical than others (Toshkov 2016). For example, when asking people for an example of a fruit, they are more likely to point to apples than to figs (Toshkov 2016). As this effect is pervasive in many domains, it suggests that some examples are considered to somehow be more central to a concept than others, even though all examples fit the definition. Although this approach is closer to the way concepts and words are used in reality, it is also more ambiguous. Under this view, it is impossible and unproductive to specify which and how many aspects of a phenomenon need to resemble the prototype to quality for falling under a concept (Toshkov 2016).

The third main approach argues that concepts can only be defined in reference to the theories in which they are embedded (Gerring 1999; Toshkov 2016). According to this perspective, concept formation is subservient to theory formation (Gerring 1999). It is based on the work of Hempel, who argued that concept and theory formation are so closely related that they “constitute virtually two different aspects of the same procedure” (Hempel 1952, 2). As such, concepts are only as good as the theories of which they are part (Gerring 1999). While acknowledging this line of reasoning, Gerring (1999) argues that concept formation cannot be reduced to theory formation: a great theory with poorly crafted concepts will at best be a great idea that is poorly implemented.

Hence, Gerring (1999); Gerring and Christenson (2017)) proposes a fourth approach: a criterial framework. Responding to the varying concerns in the literature of concept formation, he argues that forming good concepts in the social sciences entails mediating among several criteria, including clarity and familiarity. The implication, made explicit in his work, is that concept formation involves a set of trade-offs; there is likely no single “best” solution, but rather a range of more-or-less acceptable alternatives. The job is then to aim for the conceptualisation that is the best relative to alternative formulations for the purpose of the project.

As mentioned, how the concepts of race, racialisation, and racism are currently used is relatively ambiguous. Moreover, like power, race can be considered an essentially contested concept (on the idea of essentially contested concepts, see Gallie 1956): while there is widespread agreement on the term, people disagree on its intension or extension. As such, dogmatism, scepticism, and eclecticism – respectively, the idea that there is only one right answer; that all answers are equally true and everyone has a right to their own truth; and that each meaning gives a partial view so the more meanings the better – are all considered to be inappropriate attitudes to the ambiguity surrounding the concept (Garver 1978).

Taking this into consideration, the classical approach seems too dogmatic: when one defines necessary and sufficient conditions, the implication is that this is the right definition. In contrast, the prototype approach is unlikely to resolve the ambiguities, as it implies that each definition may cover part of what race means. And the theory approach is too eclectic, since it maintains the definition of the concept depends on the theory of which it is part, resulting in potentially more ambiguity regarding the concept. Instead, a criterial framework that takes into account the different functions that concepts perform as well as the trade-offs that likely have to be made in their formation is a more promising avenue. Additionally, the acknowledgement that one should aim for the best alternative for the purpose of one’s project is in accordance with the aim of this article to propose definitions of race, racialisation, and racism for the field of CTS. Hence, the next part summarises the criterial framework, predominantly based on Gerring’s (1999) work.

A criterial framework for concept formation

Having made the model increasingly concise over the years, Gerring identifies five criteria of conceptual goodness in his third and last version: resonance, internal coherence, external differentiation, theoretical utility, and consistency (Gerring and Christenson 2017). Each criterion should be understood as a matter of degrees. For example, with regard to the first criterion, resonance, one has to establish the degree to which a new definition is faithful to established definitions and usages. If there is variety in how a concept is used, such as is the case for race, there is no way to perfectly satisfy this criterion. Rather, a definition may conform more or less to established uses. Every concept will need to mediate between the criteria, making trade-offs depending on the purpose(s) it is to serve.

As mentioned, resonance refers to the extent to which a concept is faithful to extant usages and definitions. As such, in includes the extent to which a concept is familiar to the audience and could be said to include the criterion parsimony, i.e., how short the term and its list of defining attributes is. A term resonates more when it is more parsimonious and easier to remember.

Internal coherence is concerned with a concept’s intension. In a coherent concept the defining attributes “belong” to the characteristics of the phenomena in question, rather than being grouped together due to a coincidence in time and space. Additionally, the more attributes the referents of a concept share, the more internally coherent the concept is. Internal coherence thus encompasses two criteria: coherence and depth. Deeper concepts are more meaning-full and convincing: the more characteristics the phenomena in question share, the more convincing the claim that they are a class of common entities. For the purposes of this article, and specifically to clarify the difference between religious discrimination and racism, the two criteria of coherence and depth will be kept apart.

External differentiation describes the external boundaries of a concept. Most important in this respect are the boundaries with concepts that are most similar. For example, in studies investigating discrimination against Muslims, racism and Islamophobia are often used to describe or analyse similar phenomena (Lauwers 2019). In order to have useful concepts and promote cumulative analyses, it is important to distinguish which instances and attributes are relevant for each of the most-similar concepts.

The fourth criterion suggested by Gerring is theoretical utility: the utility of a concept within a larger theory and research design. Concepts are the building blocks of theories, indicating that one should think about how useful a concept is within the broader array of concepts. Moreover, when (re)defining a concept, one should aim to enhance – or do as little damage as possible to – the utility of neighbouring concepts in order to increase the general utility of the theory and research design.

The last criterion, consistency, maintains that a concept “should mean the same thing in all contexts that are relevant to the study”, in order to avoid unwarranted ambiguity (Gerring and Christenson 2017, 61). Violating consistency can lead to conceptual stretching. A typical way to solve problems of conceptual stretching is to broaden one’s concept (i.e., to have less attributes defining the concept’s intension). Because this article aims to propose definitions in order to ameliorate some of the ambiguity in the field, having a consistent concept is inherent to its aim. As such, rather than taking it up as a criterion within the framework that can be given more or less priority, it is here treated as a primary goal.

There is one aspect of concept formation and concept use that Gerring (1999); Gerring and Christenson (2017)) does not refer to in his work, but that has rightly been taken into account by CTS scholars, which I will call a concept’s political consequence. Where the criterion of theoretical utility is relevant to the epistemic goals of a project, the criterion of political consequence refers to the (potential) political effects, both positive and negative, of using a specific concept. The political aspect of CTS scholars’ work may express and manifest itself in various ways. For example, the academic investigations themselves can be political in nature, such as is the case when aiming to lay bare the “inherently problematic nature of the terrorism label” (“Aims and Scope” n.d.). But CTS scholars may also be politically active in non-academic organisations and use their research as input for these activities. In such cases, the political consequence of using a specific concept – e.g., racism rather than Islamophobia – will depend on the experiences and strategies of these organisations and their members.

All in all, this article employs the criteria of resonance, coherence, depth, differentiation, theoretical utility and political consequence to develop and evaluate the concepts race, racialisation, and racism for CTS (see Table 1). As such, the time has come to propose definitions of these concepts. The next section aims to propose definitions that make race an analytical lens, allowing CTS scholars to see aspects of discourses and practices that have hitherto remained under-researched in a systemic manner in the field.

Table 1. Criteria of conceptual goodness (based on Gerring 1999; Gerring and Christenson 2017)

1. Resonance	To what extent is the concept in line with extant definitions and usage? How familiar and parsimonious is it?
2. Coherence	To what extent are the attributes of the concept logically related?
3. Depth	How many attributes are shared by the referents of the concept?
4. Differentiation	How differentiated is the concept from other (most-similar) concepts?
5. Theoretical utility	How useful is the concept within the broader theory and research design?
6. Political consequence	What are the political risks and benefits (the potential consequences) of using this concept?

Defining race, racialisation, and racism: A proposal

This section begins by reiterating the main similarities and ambiguities in the meaning of the concepts race, racialisation, and racism in CTS. On the basis of this information, the article proposes definitions of the concepts race, racialisation, and racism that revolve around four core attributes.

Section one clarified that there are three main similarities in how CTS scholars employ the concepts of race, racialisation, and racism. First, they see races as socially constructed, implying that races are not biologically determined and that they are ascribed. Second, they approach race and racialisation as a relationship rather than a fixed identity, which translates itself into a focus on what race does and how it is enacted rather than what it is. Third, they view racism as a form of unjust negative discrimination. There are also two main ambiguities in the use of the concepts. Specifically, there is ambiguity as to whether racialisation entails a specific characterisation, such as deviant, dangerous, demonic, or inferior, and if it should be limited to a construction of features as insurmountable or permanent (or even biologically determined). Before explaining how these ambiguities were navigated in section five, this section describes the proposed definitions.

As the concepts of race, racialisation, and racism are naturally connected, they should logically share a core set of attributes. I propose this core set consists of: (A) a socially constructed nature; (B) categorisation of a group as “other”; (C) naturalisation of characteristics thought to belong to a group; and (D) being concerned with presumed community-like groups. Whereas race denotes the presumed group, racialisation refers to the process or logic underlying the construction of a race, and racism signifies unjust and negative discrimination on the basis of ideas regarding a race.

Attribute A, a “socially constructed nature” reflects dominant consensus in how the concepts are interpreted. The implications of viewing race as socially constructed, mentioned above, are equally applicable to this definitional attribute. Attribute B, categorisation of a group as “other”, entails that people are categorised as a race on the basis of possession of certain markers, such as skin colour or religious dress, which then function as signifiers of the “otherness” of the person – as someone who is fundamentally different from “us” (Hall 1997; Lauwers 2019). Attribute C, naturalisation, refers to the process whereby characteristics considered to belong to a race are seen, framed, and/or treated as somehow always insurmountable. In some cases, this may be a very explicit naturalisation; for example, when blacks are deemed to be more aggressive than whites due to differences in biological make-up.

However, in contemporary counter-radicalisation discourses it is often more subtle. Here, naturalisation may show itself in the framing of target groups as always suspicious, regardless of the extent to which a racialised person may look, act, or seem like “us”, i.e., as those regarded as the norm. The suspicion remains that people thought to belong to these groups are somehow more prone to radicalisation, which entails a naturalisation of vulnerability to radicalisation. With regard to disadvantaged groups that are racialised, the characteristics are generally considered negative, whether it be abnormal, inferior, threatening, or even demonic. In contrast, when advantaged (often majority) groups in western countries are racialised, the characteristics that are naturalised are predominantly considered positive, such as rational, civil, and superior. Attribute D, being concerned with presumedly community-like groups, in combination with the other attributes, serves to distinguish races, and especially racism, from other categorisations of people and discrimination, especially gender and sexism. Whereas genders are often considered (at least in part) socially constructed and sexism may entail categorisation of people as “other” and a similar naturalisation of characteristics, these people are generally not seen as belonging to a community-like group. To put it simply: a sexist view of “women” does not presume they resemble a community, whereas a racist view of “blacks” or “Muslims” does.

Whereas attribute D is crucial in distinguishing between racism and sexism, it is attribute C – naturalisation – that is central in differentiating racism from pure religious discrimination. Naturalisation means that one cannot escape racism by choosing to leave a racialised group, for the choice does not exist. In contrast, one could evade pure religious discrimination by conversion (as is also argued by Fredrickson 2002). I use the word “pure” here, as racial and religious discrimination are often entwined (see for example, Goldberg 2017; Topolski 2018). In line with the socially constructed and context-dependent nature of race, racialisation, and racism, racisms manifest in various ways. In order to distinguish between different forms of these manifestations, I propose the use of adjectives. For example, one can refer to religion-based racism (see Topolski 2018), culture-based racism (see Balibar and Wallerstein 1991), and colour-based (or colour-line) racism (see Fanon 1986). Each of these racisms is based on markers that are associated with different types of community-like groups, but all share the fact that they are discriminatory in nature and base themselves on a logic of racialisation. I want to stress that the idea here is not to argue that these manifestations are separate and independent. In contrast, it is to ensure that each can be recognised as a form of racism and to help to see their connections, interdependence, or even co-constitution.

Discussion

This section discusses how the criteria guided the choices made in the development of the concepts. Specifically, it describes how the criteria of coherence, differentiation, and depth were used to navigate the ambiguities in the use of the race, racialisation, and racism in CTS (see Table 2 for an overview indicating the criteria that are most relevant per attribute).

Table 2. Primary criteria per attribute

Attributes	Criteria
A) Socially constructed	1, 2
B) Categorisation as “other”	1, 2, 6
C) Naturalisation	3, 4, 5, 6
D) Community-like group	1, 4, 5

Additionally, this section discusses the main limitation of the proposed definitions, as well as a potential objection to the project of the article as a whole.

Navigating conceptual ambiguities

Naturalisation is the key attribute when it comes to navigating the ambiguities. Whereas it is narrower than the concept of essentialisation, it is broader than specific characteristics or Hussain and Bagguley’s (Hussain and Bagguley 2012) definition of racialisation as a negative or cultural construction of biological characteristics. As has been argued, racialisation may employ different types of characterisations; people may be racialised and characterised as inferior, but also as the enemy – both of which are visible in counter-terrorism discourse (Medovoi 2012). Despite the fact that such characterisations are often empirically related, they are analytically distinct, i.e., not logically related or coherent. Nevertheless, both instances can be labelled as racial.

Hence, there must be a different attribute that is logically consistent with the other attributes of race, and clarifies why such different characterisations may still both be labelled racial, a task that is performed by the attribute “naturalisation”. Naturalisation is logically coherent with the other attributes, and, as section four clarified, helps to differentiate race, racialisation, and racism from similar concepts, such as religion and religious discrimination. By choosing a more coherent and differentiated concept, the definitions sacrifice some depth. After all, making the definition specific to, for example, colour-based or biological racism would increase the accompanying properties shared by the instances under the definition. Nonetheless, such a choice would have decreased the utility of the concepts for CTS scholars, who have recently focused on the framing of Muslims in counter-terrorism. Moreover, this article proposes to remedy the sacrifice in depth to some extent by using adjectives, when useful, to demarcate specific forms of racial constructions.

This brings us to the last criterion of Gerring (1999); Gerring and Christenson (2017)), theoretical utility, i.e., the added utility of concepts within a theory and in relation to other concepts. The relatively high degree of differentiation means that race, racialisation, and racism are easy to use besides other concepts, such as religious discrimination. Furthermore, as has been shown by several CTS investigations, the concepts of race, racialisation, and race, when properly demarcated, add to theoretical frameworks focused on notions of security and prevention (Bayoumi 2006; Fekete 2004; Hussain and Bagguley 2012; Razack 2008; Dixit 2014; Selod 2018).

Nevertheless, the concept of Islamophobia poses a challenge to the theoretical utility of race, racialisation, and racism in CTS. Based on a review of what is generally meant with the term Islamophobia, Lauwers argues the concept generally “does not differentiate between religion-based racism and religious or cultural bigotry” (Lauwers 2019, 330). As such, she argues for a clearer conceptualisation of Islamophobia as covering two “types”, namely, anti-Muslim racism and anti-Islam bigotry. In this conceptualisation, anti-Muslim racism would fall both under the concept of Islamophobia as well as under the concept of racism.

It is here that the added criterion, political consequence, becomes important. Part of the goal of much CTS research is to use analyses to help those unjustifiably and/or unjustly harmed by counter-terrorism discourse and practices. These people and communities often have experience regarding which terms are most effective in challenging harmful aspects of specific counter-terrorism discourses. As CTS scholars, it is important to take these experiences into consideration. If Islamophobia is used to investigate anti-Muslim racism – and is conceptualised in a similar manner as proposed in this article – it provides the same analytical lens and hence epistemological benefits as racism. In such contexts, scholars should reflect on which term (rather than which concept, as the attributes and referents are in this case the same) is most beneficial to further their political goals. As mentioned, whereas the term race is more readily associated with broader historical and political contexts (Topolski 2018), Islamophobia is closely linked to contemporary discussions regarding discrimination of Muslims, especially after 9/11 (Jackson et al. 2018).

Potential objections

This section discusses the main limitation of the proposed definitions and a potential objection to the project of this article. The main limitation of the definitions is the inclusion of naturalisation as a core attribute. This means that processes of othering without naturalisation are not captured by the concept of race. This choice was made on the basis of the belief that it is better to have multiple concepts to clearly delineate different types of logics and processes of othering, allowing one to establish the connections between them, over a more general concept that potentially foregoes these important details and interrelations.

A potential objection to the larger project of this article is that the theoretical exercise conducted here foregoes the messy reality of conducting research, especially research on race. In these studies, it is important to take into account how powerful actors instrumentalise the concepts by using them according to their need at particular times. Hence, the perceived lack of conceptual clarity may be due to the rapid mutability of the concept of race in practice. I fully support the idea that manifestations of race are dependent on their context and consequently are highly changeable. However, in order to see the patterns in these different manifestations and instrumentalisations of race, a concept is needed to capture what is common, without foregoing how each manifestation is specific. This article has aimed to provide such a concept by defining core attributes and encouraging the use of adjectives to specify the form of race researched by the study in question.

Nonetheless, as with any concept, these definitions are and will remain work in progress. This article aimed to offer what I consider to currently be the best option for the definition of race for CTS. The underlying goals of the article are to encourage CTS scholars to, first, study the role of race in counter-terrorism discourses and, second, be explicit about what they mean with the concept of race. Clarifying the state of research on race within CTS and consequently proposing definitions seemed to me the most useful way to work towards those goals, especially considering that not all CTS scholars are embedded in scholarship on race.

Conclusion

As is clear from the discussion above, developing a concept is a complex exercise, and one where perfect solutions do not exist (Gerring 1999; Sartori 1970). Nevertheless, one can distinguish better and worse formulations, and choose on the basis of several criteria a proposal as the best alternative for the purposes of one’s project. The purpose of this article was to provide proposals for definitions of race, racialisation, and racism that are useful for CTS scholars by making the concepts into an analytical lens. On the basis of this aim, the article reviewed similarities and ambiguities in the way the concepts of race, racialisation, and racism are employed and defined by CTS scholars. Using a criterial framework for concept formation based on Gerring’s classical formulation (Gerring 1999; Gerring and Christenson 2017), it proposed definitions of these concepts that build on the similarities and navigate the ambiguities.

Following this approach, this article proposes as core attributes of race, racialisation, and racism that they: (A) are socially constructed; (B) categorise people as “other”; (C) naturalise features thought to belong to the group of people regarded as race; and (D) are concerned with presumed community-like groups. Whereas race denotes the presumed group, racialisation refers to the process or logic underlying the construction of a race, and racism signifies unjust and negative discrimination on the basis of ideas regarding a race. Attribute A, the socially constructed nature of the concepts, indicates that manifestations of race, racialisation, and racism are context-dependent. As such, races are not limited to a cultural construction of biological features, but can also be fabricated on the basis of religious, cultural, or other markers. Moreover, one does not need to identify as part of a specific race in order to be classified as such.

Attribute B, categorisation as “other”, indicates the classification into supposedly different types. As this is also the case for the logic underlying, for example, religious discrimination, attribute C, naturalisation, helps to distinguish different ways of othering. Naturalisation – the construction of features as somehow insurmountable – is specific to the concepts of race, racialisation, and racism. Whereas one could, in theory, escape religious discrimination through conversion, no such escape is possible under a racist regime. However, as races are socially constructed, religious markers can function as a basis on which groups are racialised, potentially resulting in religion-based racism. Similarly, one can distinguish culture-based racism and colour-based racism. Each of these racisms is based on markers that are associated with different types of community-like groups, but all share that they are discriminatory in nature and base themselves on a logic of racialisation. Although analytically distinct, empirical manifestations are often connected, interdependent, or even co-constitutive of each other. The analytical distinction can help to recognise each as a form of racism and to clarify the links between them.

The last attribute, D, being about a community-like group, serves to distinguish racism from forms of discrimination such as sexism, while keeping in line with common ideas about what race refers to. Whereas gender and sexism may share attribute A-C, people categorised in a specific gender are not commonly thought of as a community-like group.

The underlying aim of these definitions is to clarify, facilitate, and thereby encourage the use of the concept of race in CTS. The definitions reflect how manifestations of race, racialisation, and racism constantly adapt to new contexts. This will allow scholars to explain the variations within these manifestations without losing sight of the fundamental process common to all. Moreover, they may provide a starting point to further analyses of CTS scholars who mention but do not focus on race (Bonino 2016; Coppock and Mark 2014; Mythen, Walklate, and Peatfield 2017; Ragazzi 2016; McDonald 2011; Skoczylis and Andrews 2019; Taylor 2018; Ali 2015). The proposed definitions of race, racialisation, and racism may also provide an interesting addition for studies working with closely related topics, such as visibility (Martin 2018), notions of risk (Heath-Kelly 2012a, 2012b; Amoore and De Goede 2008; De Goede 2008), and the suspect community thesis (Awan 2012; Breen-Smyth 2013; Mythen 2012; Pantazis and Pemberton 2009; Taylor 2018). Considering the continuing dominance of counter-terrorism and counter-radicalisation discourses, the work of CTS scholars remains critically important. By focusing on clarifying the concepts of race, racialisation, and racism, this article hopes to have sharpened one of the tools that can be employed in this enterprise.

Acknowledgments

An earlier version of part of this article was presented at the Reading Group “Conflict, Security and Governance: Critical Readings”, Vrije Universiteit Amsterdam, and at the Racism and Religion Conference 2019, Centre for Multidisciplinary Studies on Racism, Uppsala University. I am grateful for the comments received after these presentations, the observations of the anonymous reviewers, the feedback of Wolfgang Wagner and Govert Buijs, and the additional comments of Sophie Lauwers, Claes Täng Wrangel, and Sinan Çankaya on earlier drafts of this article. Any errors are mine.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Notes on contributors

Sanne Groothuis

Sanne Groothuis is a PhD candidate at the Department of Political Science and Public Administration of Vrije Universiteit Amsterdam. Her research falls within the disciplines of Political Science and Philosophy.

Notes

1. Excluding the part of section four that discusses the proposed definitions of the concepts of race, racialisation, and racism, “the concept of race” is meant to also imply its associated concepts, racialisation and racism.

2. When broadening the search to the use of any of these terms anywhere in the journal from the day it was first published, 129 results appear (specifically, 129 for “race”, five for “racialisation”, and 81 for “racism”). These results include publications that focus on other topics and mention race in passing.

3. For more insight on debates regarding (the struggle to avoid) Eurocentrism in scholarly work in general, see for example, Ashcroft, Griffiths, and Tiffin (1995).

4. I.e., this article concentrates on work within CTS that employs the concept of race, racialisation, and/or racism as an analytical or explanatory tool. Due to the aim of this article, it does not review CTS literature mentioning these terms without focusing on their role in counter-terrorism (e.g. Ali 2015; Sabir 2017), or those limiting their conceptual framework to terms as Islamophobia and/or anti-Muslim bigotry without mentioning race (e.g. Githens-Mazer and Lambert 2010).

5. While quotation marks around the term race may be used to denote its socially constructed nature, I refrain from doing so for two reasons. First, it eases reading this article, where the term race is used repeatedly. Second, and more importantly, one may interpret the use of quotation marks to indicate that there are some races that are socially constructed and others which have a basis in physical reality. In order to avoid any such inkling and foster the idea of all races as socially constructed, I think we need to use the concept as having – with regard to its socially constructed nature – no other possible interpretation.