Rethinking Identity Fusion: A Critical Examination 3031469828, 9783031469824

Rethinking Identity Fusion presents a critique of Identity Fusion Theory, an identity-based social psychological approac

127 22 2MB

English Pages 135 [130] Year 2024

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
About the Book
Contents
About the Authors
List of Figures
Chapter 1: Introduction
Aims of This Book
Structure and Approach
References
Chapter 2: Conceptual Shortcomings of Identity Fusion Theory
Identity in the Social Identity Tradition
Personal and Social: A Subtle Redefinition
Functional Antagonism and the Novelty of the Concept of Fusion
What Would an Identity Fusion Consistent with SIT Look Like?
On Social Comparison and Dynamic Self-Definition
Conclusion: Hiding the Redefinition
References
Chapter 3: The Limitations of Extreme Cognitivism
Identities or Mental Entities?
Agency and Depersonalisation
In Search of the Lost Agency
Does Identity Fusion Account for Extreme Pro-group Actions?
The Problematic Concept of Relational Ties
Conclusion
References
Chapter 4: Methodological Shortcomings of Identity Fusion Research
The Problem of Measurement
The Discursive Alternative
Demarcation: Is Fusion Categorical or Continuous?
Conclusion
References
Chapter 5: Piling Con(fusion): Identity Fusion Theory Today
The Flattening of Identity Fusion
The Fading Out of Identification
The Roots of the Problem
References
Chapter 6: Conclusion
References
References
Author Index
Subject Index
Recommend Papers

Rethinking Identity Fusion: A Critical Examination
 3031469828, 9783031469824

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Rethinking Identity Fusion A Critical Examination

Metodi Siromahov Annie Hata

Rethinking Identity Fusion

Metodi Siromahov • Annie Hata

Rethinking Identity Fusion A Critical Examination

Metodi Siromahov Department of Experimental Psychology University College London London, UK

Annie Hata University College London London, UK

ISBN 978-3-031-46982-4    ISBN 978-3-031-46983-1 (eBook) https://doi.org/10.1007/978-3-031-46983-1 © The Editor(s) (if applicable) and The Author(s), under exclusive licence to Springer Nature Switzerland AG 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Palgrave Macmillan imprint is published by the registered company Springer Nature Switzerland AG. The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland Paper in this product is recyclable.

About the Book

Rethinking Identity Fusion presents a critique of Identity Fusion Theory, an identity-­based social psychological approach to understanding progroup extremism. It scrutinises the theory’s main theoretical claims and research methods, exposing serious inconsistencies and gaps in how the theory handles the concept of identity and in its research programme. The book demonstrates the flattening of the theory’s main concept, “identity fusion”, and the general state of confusion in the recent literature as to the theory’s claims and predictions. The book offers a reinterpretation of Identity Fusion Theory through a discursive perspective, critiquing its cognitivist assumptions about the nature of human relationships and identity. In this way, its scope extends to wider critiques of experimental and quantitative methods in contemporary social psychology. It argues that such theoretical and methodological shortcomings, rather than hindering a flawed approach, can accelerate its adoption in social psychology by creating an image of theoretical unity and consistency on top of a field characterised by confusion and contradiction.

v

Contents

1 Introduction  1 Aims of This Book   5 Structure and Approach   6 References   8 2 Conceptual  shortcomings of Identity Fusion Theory 11 Identity in the Social Identity Tradition  13 Personal and Social: A Subtle Redefinition  18 Functional Antagonism and the Novelty of the Concept of Fusion  20 What Would an Identity Fusion Consistent with SIT Look Like?  23 On Social Comparison and Dynamic Self-Definition  27 Conclusion: Hiding the Redefinition  28 References  29 3 The  Limitations of Extreme Cognitivism 31 Identities or Mental Entities?  32 Agency and Depersonalisation  37 In Search of the Lost Agency  40 Does Identity Fusion Account for Extreme Pro-group Actions?  42 The Problematic Concept of Relational Ties  46 Conclusion  49 References  52

vii

viii 

Contents

4 Methodological  Shortcomings of Identity Fusion Research 55 The Problem of Measurement  56 The Discursive Alternative  58 Demarcation: Is Fusion Categorical or Continuous?  62 Conclusion  69 References  71 5 Piling  Con(fusion): Identity Fusion Theory Today 73 The Flattening of Identity Fusion  75 The Fading Out of Identification  90 The Roots of the Problem  92 References  99 6 Conclusion103 References 111 References113 Author Index121 Subject Index123

About the Authors

Metodi  Siromahov is a social psychologist and a lecturer at the Department of Experimental Psychology, University College London (UCL). His academic interests are in the areas of ideology, group identity, and nationalism. Annie Hata  holds a BSc in Psychology from UCL and is undertaking an MSc in Developmental Psychology and Clinical Practice at the Anna Freud Centre in collaboration with UCL.  Her interests lie in community psychology and the intersections between social and clinical psychology in the context of race and identity. She was responsible for conducting the literature review on Identity Fusion Theory in her role as a research assistant.

ix

List of Figures

Fig. 4.1 Fig. 4.2 Fig. 5.1 Fig. 5.2

The pictorial measure of identity fusion Distributions of fusion scores obtained with the pictorial and verbal measures Categories of publications on Identity Fusion Theory in our literature search New research increasingly fails to distinguish between fusion and identification

56 64 74 92

xi

CHAPTER 1

Introduction

Abstract  The chapter provides a brief outline of Identity Fusion Theory (IFT) and describes the aims and approach of the present critique. Keywords  Social identity • Self-categorisation • Identity fusion • Cognitive psychology • Terrorism • Extremism Identity Fusion Theory (IFT) was first proposed by Swann et al. (2009) 15 years ago as an attempt to explain extreme acts, such as terrorist attacks or self-sacrifice, committed in the name of a group. It is a cognitive theory of identity which, although based on and borrowing a conceptual vocabulary from earlier theories of social identity, has become established as an elaborate theory and a vibrant research field in its own right. The hypothetical phenomenon of identity fusion has been studied using both quantitative and qualitative methods and has been used to interpret passionate forms of collective identity in contexts as diverse as Libyan rebel groups, Viking war bands, football fans, nationalism, brand loyalty, gaming subcultures, and wildlife conservation movements. It was also the subject of a special 2018 issue of the journal Behavioural and Brain Sciences, where it generated a lively critical discussion. Emerging in the post-9/11 world of the late 2000s, the theory is perhaps best understood in the context of contemporary debates about the hypothetical roots—cultural, ideological, personal, institutional, etc.—of © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 M. Siromahov, A. Hata, Rethinking Identity Fusion, https://doi.org/10.1007/978-3-031-46983-1_1

1

2 

M. SIROMAHOV AND A. HATA

acts of terrorism involving violence against civilians and self-sacrifice on behalf of the perpetrators. The first publication outlining the theory (Swann et  al., 2009) even opens with an explicit reference to the 2001 World Trade Center attacks, the 2004 metro bombing in Madrid, and the 2005 bombings in London. Rather than blaming the perpetrators’ ideology, as was common at the time, identity fusion theorists turned to more universal psychological processes to explain such extreme acts. Whitehouse, one of the main researchers working on IFT since its inception, has explicitly pointed to his earlier anthropological work on collective rituals and group cohesion as influencing his approach to the subject (Whitehouse, 2017). The theory views such extreme acts of violence as essentially pro-­ group acts in support of a tightly knit collective (such as a terrorist cell or a combat unit) carried out by particularly devoted members, and explains this extreme commitment as the product of psychological processes involved in the formation and maintenance of a collective identity. People who are willing to make extreme sacrifices in the name of a group are said to do so because they relate to that group in a specific way—seeing other group-members as an extension of themselves, like a close-knit family. This state of “visceral sense of oneness” with a group is called identity fusion. IFT starts with a long-established conceptual distinction between one’s personal identity (reflecting one’s individual characteristics) and the social or collective identity (reflecting one’s membership in a group, e.g., being an American or a Democrat). The theory holds that, for most people who belong to a group, the two identities are mutually exclusive—for example, when our group-memberships are important to us, our individual characteristics fade into the background, so that a strong social identity often comes at the expense of a weakened personal identity; such people are said to have little sense of who they are beyond their group-membership, and end up becoming conformist, obedient, and lacking in personal agency. But, crucially, in a minority of individuals the two identities can become ‘fused’ in a way that enables both to be active at the same time: … we propose that the personal self remains potent and influential among fused persons. In fact, for fused persons, group membership is intensely personal, for they feel that they care as much about the outcomes of the group as their own outcomes. (Swann et al., 2009)

1 INTRODUCTION 

3

In such a state of identity fusion, having a strong attachment to a group does not diminish one’s sense of personhood or agency. Instead, strongly ‘fused’ people perceive the wellbeing of the group as indispensable for their own wellbeing and view threats to the group as threats to their own life. It is precisely the fusion and simultaneous salience of the two identities that is said to enable such individuals to go to extremes in defence of the groups they belong to: [this disposition to extreme sacrifice] would be motivated by a highly salient personal and group identity between individuals, with a visceral feeling of deep union between the personal self and the social self, so that the delimitation between both identities becomes indistinguishable (Henríquez et al., 2020)

We can see examples of this deep enmeshment with others in close-knit families, where a person can go beyond merely identifying with the group category in the abstract, and can also form close and meaningful connections with group-members as individuals. Identity Fusion theorists go a step further by claiming that such a dynamic can exist not only in small family-like groups (like a football team or a terrorist cell), but also in large, impersonal groups, where the ‘fused’ individual cannot possibly know most of the other group-members personally—up to the level of whole nationalities. When a person’s individual and social identities become merged, they are said to ‘project familial ties’ onto the group, in effect perceiving it as an extended family, with all the deep emotional attachments and feelings of protectiveness that one has for one’s own family: … self-reported feelings of familial connection to other group members statistically mediated links between fusion and pro-group outcomes […] Apparently, highly fused persons view their group members as fictive family members, and these perceptions motivate them to take extreme actions on the behalf of these individuals. (Swann & Buhrmester, 2015)

In this way, identity fusion is used to explain not only deep emotional connections with one’s family or comrades but also large-scale social phenomena such as extremist nationalism or religious fanaticism (e.g., Besta et al., 2014). This ‘extended fusion’ with nations and other large, impersonal collectives has been a major focus of IFT research since its inception, and subsequent publications have extended IFT to study ‘identity fusion’

4 

M. SIROMAHOV AND A. HATA

with out-groups, famous individuals, animal species, and abstract entities, such as brands and ideologies. IFT can make that jump from deep personal connections with known group-members to equally deep attachments to strangers or abstractions because its approach is essentially cognitive—it assumes that the ‘fusion’ between the personal and collective identities is a cognitive process in which two mental representations become merged in the person’s mind; and, since any two cognitive representation can ostensibly undergo the same process, the range of things that one can become ‘fused’ with is potentially limitless. On this theoretical foundation a large body of empirical, predominantly quantitative research has been accumulated. Fusion theorists have conducted studies to support the main theoretical claims of IFT, for example demonstrating that identification and fusion with a group are two distinct phenomena, and that fusion is a more reliable predictor of extreme pro-­ group behaviour than identification (Swann, Gómez, Huici et al., 2010; Swann, Gómez, Dovidio et al., 2010). They have explored the causes of identity fusion, demonstrating that people are more likely to become ‘fused’ with groups with which they share the same genes or culture (Vázquez et al., 2017), or if they have gone through traumatic or otherwise dysphoric experiences together (Jong et al., 2015). Finally, they have traced some of the consequences of identity fusion, such as long-lasting attachments (Newson et al., 2016; Talaifar et al., 2021) and an increased propensity towards political violence (Kunst et  al., 2019). More recent studies have extended the explanatory framework offered by IFT beyond the study of group memberships: a particularly fruitful research area has been identity fusion with individuals (such as political or religious leaders; Kunst et al., 2019; Nikolic, 2021), animals (Buhrmester et al., 2018), or even abstract concepts like brands (Lin & Sung, 2014; Hawkins, 2019; Krishna & Kim, 2021). We have attempted to produce a précis of IFT’s key claims and ideas, which have been developed over the span of a decade and a half. However, such a summary can only ever be partially successful. As we will argue in this book, IFT’s use of language is often slippery: sometimes the theory gives multiple non-overlapping definitions for the same concepts, making it difficult to fix its claims in place; and other times it leaves important concepts under-theorised, especially ones that are already common currency in everyday language but are used in IFT with a peculiar, counterintuitive meaning. Distilling its tenets requires us to cut out aspects of the theory that sit uneasily alongside each other, and simultaneously elaborate

1 INTRODUCTION 

5

on ideas which were only hinted at by its authors but never stated explicitly, as will become clear in the next two chapters. At its core, Identity Fusion Theory does point to an important real phenomenon in group dynamics—the sense of deep enmeshment with others that can occur in close-knit groups (such as army regiments, secret societies, fraternities, close friendship circles, or religious cults), which is indeed distinct from mere identification with, or membership in, a group. When group membership involves frequent close and emotionally intense interaction with others and a high degree of self-disclosure and mutual trust, it can acquire a deep emotional significance for the group-members; people sometimes do describe such groups as feeling “like a family”, with the other group-members being like brothers or sisters to them—the romanticist language of kinship reflects the emotions and solidarity felt within such communities. By drawing attention to this quality of group membership and distinguishing it from mere self-labelling as group-­ member, fusion theory makes an important conceptual distinction. However, while IFT does point to this really existing phenomenon of social life, we will argue that its explanation for it—the fusion of the personal and social identities—is theoretically weak and inconsistently invoked across much of the IFT literature. Moreover, the theory runs into problems when it tries to extrapolate the existence of this phenomenon of personal attachment to group-members from small-scale groups onto large-scale ones, in which such personal relationships cannot exist.

Aims of This Book As fusion theory becomes better established and diffused throughout different sub-fields of the social sciences, it is important to take stock of the state of the literature today. IFT’s achievements and strengths have been summarised in literature reviews and meta-analyses before (e.g., Gómez et  al., 2020; Henríquez et  al., 2020), however, the growing volume of publications on the topic allows us to cast a critical look at certain theoretical assumptions and research practices which have, by now, become entrenched in identity fusion research. This book is an attempt to initiate a debate to that end. The present book offers a theoretical and methodological critique of the Identity Fusion literature. To our knowledge, this is the first publication that offers such a systematic critique of IFT in its entirety, beyond isolated commentaries on individual research papers. We believe that IFT’s

6 

M. SIROMAHOV AND A. HATA

weaknesses, both as a social psychological theory and as a research programme, are so numerous that they fundamentally undermine its integrity. In the process, we will attempt to challenge all of IFT’s foundational claims that we have summarised up to this point—from the hypothetical interactions between the personal and social identities and the validity of the concept of ‘identity fusion’ to the possibility of ‘extended fusion’ with nation-sized groups, individuals, and abstractions. We arrived at this position in the process of writing the book. Its first author began this journey as an insider to the Identity Fusion approach using the theory in his doctoral studies (Siromahov, 2020) and his published research (Siromahov et al., 2020), and in 2021 set out to write a short commentary criticising IFT’s overreliance on hypothetical cognitive processes, the inconsistent use of measurements, and some theoretical problems with extending fusion onto nations and abstractions. That critique grew as more inconsistencies and gaps became apparent, particularly in the way IFT handles ideas borrowed from the Social Identity approach and in its lack of concern for distinguishing fusion from identification in practice. At that point a more systematic approach was required, which led us to conduct a review of the identity fusion literature up to 2021, the results of which we report here. In the process, our stance on IFT has shifted from one of critical support with suggestions for future theoretical clarifications and methodological improvements to a wholesale rejection of the theory.

Structure and Approach The following two chapters deal with the early publications on IFT (spanning the years from 2009 to 2014) and explore the theory’s foundational claims. In Chap. 2, we argue that IFT’s grounding in the earlier and much better-established Social Identity tradition is plagued by fundamental misinterpretations (specifically regarding the notions of personal and social identity and depersonalisation) that undermine its own original concept of ‘identity fusion’. Chapter 3 deals with IFT’s hard-cognitivist and reductionist interpretation of social behaviour. We demonstrate that the theory treats identities as entities in the mind that drive behaviour, and ‘relational ties’—as a mere perception that can be ‘projected’ onto anyone, including strangers. This approach allows fusion theorists to extend IFT to national and other large-scale groups, but is also revealed to be under-theorised,

1 INTRODUCTION 

7

unsupported with evidence, and pregnant with the possibility of future misunderstandings about the theory’s main claims. Chapter 4 offers a critique of IFT’s experimental research methods, arguing that, in addition to being poorly theorised, its main variables are also poorly operationalised—the established self-report measures of fusion are not designed to capture the kind of ‘fusion of the personal and social selves’ that the theory is meant to describe. The chapter demonstrates the inconsistent treatment of the concept of fusion across different studies and offers a brief reinterpretation of the meaning of self-report data obtained using such measures from a discursive perspective. Chapter 5 then traces the evolution of IFT’s research programme up until 2021. In it, we argue that the concept of fusion has been stripped of much of its technical specificity and is today often used in a shallow and non-specific way as meaning a generic sense of strong attachment to something. We also demonstrate that, after an initial string of publications demonstrating identity fusion’s conceptual distinctiveness from other forms of alignment with groups, subsequent papers have eschewed such careful differentiation, collapsing the meaning of fusion into a generic sense of belonging to a group, and thus undermining the interpretability of much of their findings. Much of the debates on IFT aim to reconcile incongruent experimental findings or to map out identity fusion’s relationships with other quantifiable variables—in effect, relying on empirical findings to drive future developments in the realm of theory. In our critique we have taken the opposite approach by prioritising a discussion of theoretical gaps and inconsistencies above empirical findings. As we will argue in Chap. 5, the more recent shortcomings in identity fusion research (such as the inconsistent and inappropriate research methods, the confusion about basic aspects of the theory, and the flattening and blurring of the concept of fusion itself) lie downstream from fundamental weaknesses in how the theory was formulated at its inception. A critical assessment of IFT’s empirical findings is therefore secondary to the critique of its poorly theorised and operationalised constructs on which the empirical research rests. In our approach we owe an intellectual debt to critics of the cognitive approach to social psychology, particularly coming from a discursive perspective (e.g., Harré, 2001; Potter & Wetherell, 1987; Billig, 1987). Michael Billig’s critique of experimental social psychology (Billig, 2011, 2013) in particular has had a profound influence on the present work, and some of his criticisms of contemporary academic research are echoed here.

8 

M. SIROMAHOV AND A. HATA

References Besta, T., Gómez, Á., & Vázquez, A. (2014). Readiness to deny group’s wrongdoing and willingness to fight for its members: The role of poles’ identity fusion with the country and religious group. Current Issues in Personality Psychology, 2(1), 49–55. Billig, M. (1987). Arguing and thinking: A rhetorical approach to social psychology. Cambridge University Press. Billig, M. (2011). Writing social psychology: Fictional things and unpopulated texts. British Journal of Social Psychology, 50(1), 4–20. Billig, M. (2013). Learn to write badly: How to succeed in the social sciences. Cambridge University Press. Buhrmester, M.  D., Burnham, D., Johnson, D.  D., Curry, O.  S., Macdonald, D. W., & Whitehouse, H. (2018). How moments become movements: Shared outrage, group cohesion, and the lion that went viral. Frontiers in Ecology and Evolution, 6, 54. Gómez, A., Chinchilla, J., Vázquez, A., López-Rodríguez, L., Paredes, B., & Martínez, M. (2020). Recent advances, misconceptions, untested assumptions, and future research agenda for identity fusion theory. Social and Personality Psychology Compass, 14(6), e12531. Harré, R. (2001). The discursive turn in social psychology. In D.  Schiffrin, D.  Tannen, & H.  E. Hamilton (Eds.), The handbook of discourse analysis (pp. 688–706). Hawkins, M.  A. (2019). The effect of activity identity fusion on negative consumer behavior. Psychology & Marketing, 36(4), 395–409. Henríquez, D., Urzúa, A., & López-López, W. (2020). Identity fusion: A systematic review. Acta Colombiana de Psicología, 23(2), 410–437. Jong, J., Whitehouse, H., Kavanagh, C., & Lane, J. (2015). Shared negative experiences lead to identity fusion via personal reflection. PLoS One, 10(12), e0145611. Krishna, A., & Kim, S. (2021). Exploring the dynamics between brand investment, customer investment, brand identification, and brand identity fusion. Journal of Business Research, 137, 267–277. Kunst, J. R., Dovidio, J. F., & Thomsen, L. (2019). Fusion with political leaders predicts willingness to persecute immigrants and political opponents. Nature Human Behaviour, 3(11), 1180–1189. Lin, J. S., & Sung, Y. (2014). Nothing can tear us apart: The effect of brand identity fusion in consumer–brand relationships. Psychology & Marketing, 31(1), 54–69. Newson, M., Buhrmester, M., & Whitehouse, H. (2016). Explaining lifelong loyalty: The role of self-shaping identity fusion and self-shaping group events. PLoS One, 11(8), e0160427.

1 INTRODUCTION 

9

Nikolic, M. (2021). Identity fusion with a political leader: The role of value congruence in predicting violent intentions and prosociality (Master’s thesis). Potter, J., & Wetherell, M. (1987). Discourse and social psychology: Beyond attitudes and behaviour. Sage. Siromahov, M. (2020). The cognitive underpinnings of essentialist nationalism: Exploring the role of identity fusion and heuristic thinking. Doctoral dissertation, Royal Holloway, University of London. Siromahov, M., Buhrmester, M., & McKay, R. (2020). Beliefs in national continuity are related to essentialist thinking and to perceptions of the nation as a family. Nations and Nationalism, 26(4), 845–863. Swann, W. B., Jr., & Buhrmester, M. D. (2015). Identity fusion. Current Directions in Psychological Science, 24(1), 52–57. Swann, W.  B., Jr., Gómez, A., Huici, C., Morales, J., & Hixon, J.  G. (2010). Identity fusion and self-sacrifice: Arousal as a catalyst of pro-group fighting, dying, and helping behavior. Journal of Personality and Social Psychology, 99(5), 824–841. Swann, W. B., Jr., Gómez, Á., Dovidio, J. F., Hart, S., & Jetten, J. (2010). Dying and killing for one’s group: Identity fusion moderates responses to intergroup versions of the trolley problem. Psychological Science, 21(8), 1176–1183. Swann, W. B., Jr., Gómez, Á., Seyle, D. C., Morales, J. F., & Huici, C. (2009). Identity fusion: The interplay of personal and social identities in extreme group behavior. Journal of Personality and Social Psychology, 96(5), 995–1011. Talaifar, S., Ashokkumar, A., Pennebaker, J. W., Medrano, F. N., Yeager, D. S., & Swann, W.  B., Jr. (2021). A new pathway to university retention? Identity fusion with university predicts retention independently of grades. Social Psychological and Personality Science, 12(1), 108–117. Vázquez, A., Gómez, Á., & Swann, W. B. (2017). Do historic threats to the group diminish identity fusion and its correlates? Self and Identity, 16(4), 480–503. Whitehouse, H. (2017, February 14). Quelling radicalization through new understanding of ritual, fusion and identity [Video file]. YouTube. https://www. youtube.com/watch?v=2ohEo8J7VwI

CHAPTER 2

Conceptual Shortcomings of Identity Fusion Theory

Abstract  Identity Fusion Theory ostensibly describes a previously unnoticed type of relationship between an individual’s personal and social identities, in which the two become ‘fused’. However, its claim to originality is based on an unacknowledged redefinition of the meaning of certain key terms. IFT seems to understand the ‘personal self’ as encompassing the whole person, so the presence of a strong social identity is redefined in IFT as a fusion of the group and the self. In this way, IFT gives a new label (‘identity fusion’) to a phenomenon that has already been studied extensively in social psychology—that of self-definition in terms of group-membership. Keywords  Social identity • Self-categorisation • Identity fusion • Cognitive psychology • Agency

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 M. Siromahov, A. Hata, Rethinking Identity Fusion, https://doi.org/10.1007/978-3-031-46983-1_2

11

12 

M. SIROMAHOV AND A. HATA

Identity Fusion Theory began its life in 2009 as an extension and revision of the Social Identity tradition (SIT1). The new model rests on three foundational ideas that fusion theorists attribute specifically to Social Identity and Self-Categorisation Theory: (1) that people have a personal and a social identity (or ‘selves’) which are separate and distinct, (2) that one ‘self’ always operates at the expense of the other (i.e., identifying as a group member leads to the suppression of the ‘personal self’ and a loss of individuality), and (3) that therefore identifying with a group makes people more conformist, less individualistic, and thus less capable of extreme pro-group actions. Fusion theorists (Swann et al., 2009, 2012) accepted the broad validity of these three claims while proposing that there is a special kind of psychological state, called ‘identity fusion’, in which these dynamics change: the two ‘selves’ or identities do not suppress each other, as SIT allegedly claims, and instead the personal self remains active alongside the social. This gives the ‘fused’ individual a sense of personal agency that enables them to pursue extreme pro-group actions in a way that ordinary group-members cannot. In this chapter, we will subject these claims to a critical analysis. First, we will explore the basic principles of the Social Identity tradition and compare them to how they are recounted in the Identity Fusion literature. The goal will be to show that Identity Fusion Theory takes an overly mechanistic approach to social cognition that was not necessarily present in the original SIT. In particular, IFT’s treatment of personal and social identity as if they are cognitive entities that operate like machines in the mind and produce behaviours through their hydraulic relationship tends to distort the original meaning of the same concepts in the Social Identity tradition. In this way, IFT is able to produce a set of ‘laws’ that it attributes to SIT, which the original theories did not really hold as laws. Fusion theorists can then claim to have found exceptions to those laws which require a novel phenomenon to explain (‘identity fusion’), but in reality would have been largely compatible with the Social Identity tradition. We will then take a critical look at two of the main claims that fusion theorists make: that ‘fused’ individuals retain their ‘personal agency’, which enables 1  The acronym SIT is typically used to refer to Tajfel and Turner’s (1979) Social Identity Theory, which was later extended into and superseded by Turner’s Self-Categorisation Theory (SCT). For brevity, we will use the acronym SIT to refer to the broader social identity tradition which encompasses both theories, rather than only the original 1979 model, as most of our discussion here applies to both.

2  CONCEPTUAL SHORTCOMINGS OF IDENTITY FUSION THEORY 

13

them to engage in extreme pro-group acts, and that a state of identity fusion can be extended onto whole nations via the ‘projection of familial ties’. We will argue that both ideas are hampered by an excessive and unwarranted reliance on hypothetical cognitive mechanisms that are under-theorised and under-researched. It is important to highlight, however, that SIT does not form a univocal tradition, and many competing and often incompatible interpretations of it exist—some more nuanced or grounded in empirical evidence than others. In this way, fusion theory could be working with a particular (highly mechanical, cognitivist) interpretation of SIT that, although departing from the original formulation of the approach, has nevertheless found broad usage in the literature. If that is the case, then it is not made clear by fusion theorists, who routinely cite the earliest publications on the Social Identity approach to defend their interpretation of the ‘personal’ and ‘social’ selves and the functional antagonism principle. Therefore, we have opted to restrict our discussion to the early works that fusion theorists themselves rely on as an example of SIT. In comparing the two approaches, we are also not attempting to construct a value-laden opposition between SIT as “original” and “pure” and Identity Fusion Theory as a “deviation”, or to imply that the latter is somehow deficient simply for  disagreeing with the former. Our critique is not dependant on an unconditional acceptance of SIT, whether in the formulation outlined here, or in any of its other guises. Different strands of the Social Identity approach can themselves be found guilty of unfounded theorising, excessive cognitivism, or an overreliance on experimental methods. SIT is relevant to this discussion only to the extent that fusion theorists claim it to be the foundation of their own approach, which invites us to explore critically how they handle SIT’s original concepts. Beyond that, the reader is free to remain agnostic or even critical of the Social Identity tradition in its many variants.

Identity in the Social Identity Tradition Prior to the rise of the Social Identity tradition in the late 1970s, it was customary for social psychologists to view groups as mere collections of individuals, and group dynamics—as a function of individual psychological motivations (Turner & Reynolds, 2010). The concept of identity did not appear to add much explanatory value—it was generally accepted that people associate with others because they are attracted to them or they

14 

M. SIROMAHOV AND A. HATA

stand to gain from it. However, a series of seminal studies by Sherif (1967) and Festinger (1954) demonstrated that people are also capable of functioning on a qualitatively different psychological level when the group ceases to be merely a physical collection of individuals and its members come to view themselves as a collective with a shared identity distinct from other people. The broader goal of SIT, according to Turner and Reynolds (2010) was to emphasise that groups are psychologically real, in that consciousness of group identity changes their members’ thinking and behaviour, and to illustrate the complex social dynamics enabled by having a collective identity. The group, in other words, is “a distinct psychological process, not merely a summation of personal relationships” (Turner & Oakes, 1997, p.306). Tajfel and Turner’s earlier work in this vein explored the intergroup dynamics of competition, in-group favouritism, ethnocentrism, and prejudice (Tajfel & Turner, 1979). Their key insight was that people’s actions can be placed on a continuum from individual (i.e., acting in terms of ‘I’ and ‘me’, in pursuit of one’s individual interests) to collective behaviours (acting in terms of ‘we’ and ‘us’, in pursuit of the group’s goals and interests). Turner (1982) later extended this framework in the form of Self-­ Categorisation Theory (SCT), a general cognitive theory of self-definition, according to which people can define themselves at different levels of abstraction—from an individual distinct and separate from everyone else to a member of various social groups differing in size and abstraction: People categorise themselves at different levels of inclusiveness, of which the personal or individual level is only one. […] [S]elf-definition as an individual in contrast to other individuals is only one level among many. (Turner & Onorato, 1999, p.324).

The Social Identity tradition also understands the self (the mental representation of who one is) as contextually variable. People’s self-­knowledge is rich and comprises a variety of personality traits, biographical facts, interests, beliefs, values, skills, hobbies, group memberships, etc. These cannot be equally salient all the time—depending on the social context, some characteristics can become more relevant to one’s self-definition than others. One of the major factors that influence this process is social comparison—people’s sense of self is given meaning by the relevant Others that they are comparing themselves against. For example, interpersonal comparisons (‘me’ vs. ‘the other people in my in-group’) will emphasise

2  CONCEPTUAL SHORTCOMINGS OF IDENTITY FUSION THEORY 

15

the individual’s idiosyncratic characteristics that make him distinct from the rest of the group, resulting in a more individual level of self-definition; likewise, intergroup comparisons (‘my in-group’ vs. ‘an important out-­ group’) will bring to the fore characteristics that he shares with the in-­ group, while the characteristics that would have marked him as distinct from it will fade into irrelevance: … where people define themselves in terms of a shared social category membership, there is a perceptual accentuation of intragroup similarities and intergroup differences on relevant correlated dimensions. (Turner & Haslam, 2012, p.32)

This makes social comparison and the search for distinctiveness (either individual distinctiveness vis-à-vis the group or collective distinctiveness vis-à-vis an out-group) essential for the process of self-definition. Where social identity becomes relatively more salient than personal identity, people see themselves less as differing individual persons and more as the similar, prototypical representatives of their ingroup category. There is a depersonalization of the self—a ‘cognitive redefinition of the self’—from unique attributes and individual differences to shared social category memberships and associated stereotypes. (Turner & Onorato, 1999, pp.321–22)

The primacy of social comparison implies that the contents of both personal and collective identities are dynamic and context dependent—my identity as an individual will depend on which in-group I am comparing myself to: People do not have social and personal identities in a fixed, static sense as part of their individual identity. Variation in the level of self-categorisation is seen as normal and ever present […] Personal differentiation takes place in the context of ingroup memberships and not in the abstract; personal identities vary in content with context as do social identities (Turner & Onorato, 1999, pp.324–5)

The same dynamic plays out when collective identity is constructed. How the group sees itself is influenced by the relevant out-group against which it is compared—for example, which traits define the American identity will differ depending on whether we are comparing Americans to

16 

M. SIROMAHOV AND A. HATA

Canadians or to Iraqis (Haslam et  al., 1992, cited in Turner et  al., 1992, p.291): Our research suggests that there is no one-to-one correspondence between the long-term knowledge one has about what different kinds of people are like and the actual social category that is constructed to represent them in a given setting. […] Categories are not defined by a fixed prototype (or a fixed set of exemplars); they vary in the relative prototypicality of their members as a function of context [...] What the typical psychologist is like will vary with the in-group and out-group members compared in any setting. Change the outgroup, and the most prototypical psychologist [...] will change. (Turner et al., 1992, pp.291–2)

Turner (1984) was careful to warn against attempts to reify the ‘personal’ and ‘social identity’ into cognitive entities operating like machinery in the human mind. As he pointed out, a clear-cut distinction between personal and collective interests is not always observed in practice, as the interests of groups often overlap with those of individual group-members. SIT’s argument was that social identity can motivate pro-group action even in the absence of such individual self-interest or even at some personal cost. Turner has explained that the personal/social distinction is essentially an approximation intended to illustrate precisely this key insight—that the interests of the individual and of the group are dissociable in principle, even if not always in practice: [The personal/social distinction] is advanced as an approximation. Its main merit is to recognize the simple fact that sometimes we seem to perceive ourselves primarily or solely in terms of our relevant group memberships rather than as differentiated, unique persons: social identity is sometimes able to function to the relative exclusion of personal identity. (Turner, 1984, p.527)

The founders of SIT also rejected the more mechanistic interpretations of their theory (which, as we will show here, Identity Fusion Theory also engages in). Although social identity research showed very early on that the mere act of categorising people into different groups can foster a collective identity and enable a range of social dynamics from cooperation to in-group favouritism and ethnocentrism,2 Turner (1975, 1984) cautioned 2

 See the early “minimal groups” studies by Tajfel et al. (1971).

2  CONCEPTUAL SHORTCOMINGS OF IDENTITY FUSION THEORY 

17

that such consequences of a collective identity do not arise automatically, as the theory has sometimes been misrepresented. People in real-world social situations have multiple strategies at their disposal in pursuit of a positive identity—they can aim for individual upward mobility or engage in collective competition against a rival group; if that path is blocked by a rigid social hierarchy, they can organise to push for reforms or overturn the hierarchy altogether; alternatively, they can be socially creative and choose to compare themselves against the dominant group on a different set of values that cast their group in a more positive light, or compare themselves to another group that is even lower in the hierarchy. Groups will also behave differently depending on their social status, the values of the culture they inhabit, and the stereotypes about their group identity. For example, in one study (Hopkins et al., 2007), Scottish nationals who were aware of being stereotyped as selfish attempted to elevate their collective image by helping members of an out-group, rather than discriminating against them, demonstrating that prejudice and discrimination are not an automatic outcome of social identity processes. The social identity tradition also emphasises the ideological nature of identities—our knowledge about a group (e.g., “men are competitive”), as well as the value judgements we make about its traits (e.g., “competitiveness is desirable”), are socially shared and therefore the product of social ideologies. In this way, a social identity analysis should always be grounded in the analysis of the specific social circumstances and ideologies that inform the group dynamics being observed. As we will argue here, this is something that the literature on Identity Fusion Theory has consistently struggled with. The Social Identity tradition has not stopped evolving over the past four decades and the various theories derived from it are too diverse to recount here. We have chosen to restrict our summary to the key ideas and concepts that Identity Fusion Theory is ostensibly based on. SIT makes several important claims in this regard: that a person’s sense of self is situationally variable and can be defined both as a distinct individual and as a group-member; that these versions of the self are constructed through social comparison, so that ‘personal identity’ is built on one’s distinctiveness from the in-group, and ‘social identity’—on the group’s distinctiveness vis-à-vis another group; and that, although a person can have similar goals to other group-members, these are not essential for pro-group action to occur—sometimes the collective interest can take priority over the narrowly personal ones. As we will show in the next section, these are all points that end up misrepresented or ignored in Identity Fusion Theory.

18 

M. SIROMAHOV AND A. HATA

Personal and Social: A Subtle Redefinition Now that we have seen how the founders of the Social Identity tradition understood personal and social identity, we can turn our attention to the way the same concepts have been presented in Identity Fusion Theory. This exploration will require some creative interpretation because, after 15 years of theoretical developments and research on the topic, fusion theorists are yet to provide a detailed account of the personal and social selves that supposedly produce the state of ‘identity fusion’. By defining identity fusion as a feeling of ‘visceral oneness with a group’ characterised by a blurring of the ‘boundaries’ between the ‘personal and social self’ (Swann et  al., 2009), fusion theorists give the impression that they are merely building upon these concepts as they are used in the Social Identity tradition, so further theoretical elaboration is apparently not necessary—to learn more about these concepts, the reader could presumably consult Tajfel and Turner’s classic works on social identity. However, a closer examination reveals that IFT rests on a redefinition of SIT’s dichotomy between the ‘personal’ and the ‘social’, which might seem trivial at first glance but could turn out to be one of the major theoretical shortcomings of the theory. The Social Identity tradition has always worked with some form of a distinction between, broadly speaking, ‘the personal’ and ‘the collective’. Dichotomies of this kind (“Self and others”, “The person and society”, “The individual and the crowd”, “My inner world vs. my public persona”, “Subjectivity vs. social structure”, etc.) seem to be intuitively meaningful for many people, as they tap into human beings’ dual nature as self-aware individuals and fundamentally social creatures. However, these dichotomies cannot all be conflated and reduced to the same binary, as, despite their conceptual similarity, they reflect subtly different aspects of the personal/social divide. Over the history of its development, the Social Identity tradition has actually articulated two such continua. The first comes from the original Social Identity Theory (Tajfel & Turner, 1979) and can be described as a spectrum of social behaviour ranging from ‘Self’ to ‘Group’— the theory sought to explain why people sometimes act as individuals and other times band together in groups and large social movements. Social Identity Theory was then superseded by Turner’s (1982) Self-­ Categorisation Theory, which aimed to provide a cognitive basis for these social behaviours by proposing that a person can define themselves on a spectrum of self-categorisation: from ‘me-as-an-individual’ to

2  CONCEPTUAL SHORTCOMINGS OF IDENTITY FUSION THEORY 

19

‘me-as-a-group-member’. It differs from the Self/Group spectrum in that it recognises group membership as part of the self—people can construct their sense of self through different group-memberships (‘me as an American’, ‘me as a worker’, etc.) alongside being a distinct individual. In other words, the first spectrum describes the actions of physical people (either a single person or a group), whereas the second spectrum describes two possible modes of being for a single individual, which are meant to explain these social behaviours. When Identity Fusion Theory talks about the fusion of ‘the personal’ and ‘the collective’, it routinely slips back and forth between these two continua without acknowledging it, which becomes fertile ground for miscommunication and confusion. The explicit definition of fusion is the ‘merger of the personal and social identities’, i.e., a fusion of two versions of the self (‘me-as-an-individual’ and ‘me-as-a-group-member’) in the individual’s mind. However, fusion theorists also routinely talk about ‘fusion with a group’, ‘immersion in a group’, and ‘visceral sense of oneness with a group’, implying a merger between the individual’s entire self-­ concept and the idea of the group, not a merger between the individual’s personal and social identities. Individuals who score high on measures of identity fusion are described as “fused with Spain” (or India, Britain, etc.). In some cases, this is spelled out even more clearly: Some individuals […] may feel one or fused with a group. For fused individuals, the self-other barrier is blurred and the group comes to be regarded as functionally equivalent with the personal self (Swann et al., 2009)

The phrase “the self-other barrier” in this quote is particularly telling. Fusion theorists often write about a boundary between the personal and social identities (two aspects of the self that are said to exist in the individual’s mind), but here the boundary is between the self (the entire self, not its constituent parts) and ‘others’, i.e., people who we do not view as part of us. The inconsistency might seem trivial, but it obscures the fact that the notion of ‘identity fusion’ actually describes two distinct phenomena. One is the merger of two identities in a person’s mind, which makes fusion a kind of cognitive phenomenon arising from a transformation of one’s self-knowledge. The other phenomenon—the ‘immersion’ of an individual in a group—is more sociological in nature, as it seems to describe some set of social relationships and dynamics (e.g., feeling emotionally enmeshed and attached to the group, empowered by your

20 

M. SIROMAHOV AND A. HATA

membership in it, sharing your resources and dedicating your time and efforts to it, making the group an important presence in your personal life), but these are not elaborated on in much depth. The description of the cognitive type of ‘fusion’ can be taken literally, as it seems to describe a really existing process in a person’s mind, whereas the sociological ‘fusion’ sounds more metaphorical and encompasses a much fuzzier set of emotions, attitudes, and relationships towards the group. Identity Fusion Theory is fundamentally defined by this constant shifting between the two versions of the personal/social spectrum—it can never decide whether it is describing the fusion of two identities existing in a person’s mind, or the fusion of the whole individual and the group that he belongs to. IFT works with both definitions, calling both phenomena ‘identity fusion’, but also points to the cognitive phenomenon as the causal mechanism that gives rise to the sociological.

Functional Antagonism and the Novelty of the Concept of Fusion This conflation of two very different versions of the personal/social binary becomes particularly problematic when fusion theorists attempt to draw upon and challenge some of the key parts of the Social Identity tradition. IFT attributes to Self-Categorisation Theory the claim that, when a person identifies with a group, the social identity takes precedence over the personal. This is referred to as the principle of functional antagonism— meaning that, for social identity to be salient in any given interaction, the personal identity must be relegated to the background, or vice versa. The relationship between the two is described as ‘hydraulic’, which implies that it arises out of some automatic and mechanical cognitive process: as one identity becomes more salient, it suppresses the other. IFT then departs from SIT by arguing that this hydraulic relationship does not always play out in the same manner—‘identity fusion’ is supposed to be a psychological state in which “the personal self is activated along with the social self” (Henríquez et al., 2020), rather than cancelling it out. The problem with this account is that SIT’s understanding of identity is a lot less mechanistic and ‘hydraulic’ than what IFT describes. As we have seen, ‘personal identity’ is defined in Turner’s Self-Categorisation Theory as self-definition at the level of a sole individual distinct from other group-members. Therefore, the individual and social levels of identity are

2  CONCEPTUAL SHORTCOMINGS OF IDENTITY FUSION THEORY 

21

incompatible not because of some hydraulic relationships between the two, but by definition, because functioning as an individual is defined as not functioning as a group-member: Personal identity refers to self-categories which define the induvial as a unique person in terms of their individual differences from other (ingroup) persons. (Turner & Onorato, 1999, p.322)

In contrast, Identity Fusion Theory seems to ignore the role of social comparison in the construction of personal and social identity. Swann et al. (2012) state so directly: … fusion theory is not hamstrung by social identity theory’s assumption that intergroup behavior is shaped and informed by intergroup comparisons […]. In contrast, because fusion theory focuses just as much on relationships within groups as it does on the social category as a whole, it can readily explain instances in which highly fused persons make pro-group sacrifices in the absence of intergroup comparisons… (Swann et al., 2012)

It bears repeating that that the ‘functional antagonism principle’, like other components of the Social Identity approach, has been subject to a variety of interpretations, not all of which focus on social comparison. As some supporters of the identity fusion approach might correctly observe, the principle is often treated in the literature with significantly less nuance than the way we have described it here, so fusion theory should not be singled out in its treatment of these concepts. Fusion theory could therefore reject SIT’s original formulation, in which the two identities were incompatible by definition, and opt for a more mechanistic one, where one identity actively suppresses the other some of the time, leaving an opening for a fusion between the two to occur in some special cases. The abandonment of social comparison is certainly presented as a liberation, as if IFT is now free to explore social situations in which no direct social comparison is present—but the authors fail to acknowledge that this also entails a radical redefinition of the ‘personal versus social’ dichotomy, as ‘personal’ no longer means ‘individually distinctive’ and now merely refers to any attribute of the individual person. Having redefined ‘personal’ to mean anything related to the person, fusion theorists then correctly observe that the personal identity (thus redefined) is not always incompatible with group-membership—in fact, quite often members of a group will

22 

M. SIROMAHOV AND A. HATA

share the same characteristics (e.g., most members of a national community will speak the dominant language, follow the majority religion, engage in nationally distinctive cultural practices, or be similarly affected by major historical events in their country). Fusion theorists can point to these shared traits, habits, and experiences, and argue that they are simultaneously constitutive of a collective identity and experienced at a personal level; therefore, they argue, the social level does not have to suppress the personal—the two are not functionally antagonistic but can ‘merge synergistically’. It is, they claim, a novel phenomenon—a fusion of the personal and social identities—that SIT had not accounted for: Individualistic goals are not always conflicting with community values or with being a member of a group, and in the extreme form, they can be fused and display a phenomenon called “identity fusion” by Swann and colleagues (2009). (Brygoła et al., 2021)

However, we can see that the novelty of this dynamic rests on a mere redefinition of the terms. If ‘the personal’ means the whole self-image (and not just what is individually distinctive), then to have this self-­concept ‘fused with a group’ is analogous to including the group in one’s self-­ concept. It would mean partially defining oneself through the group, seeing the group-membership as an important aspect of who one is—i.e., it would be synonymous with having a social identity, which is something that SIT has already described in detail. The fact that a person shares certain traits with the group is not some newly discovered ‘fusion’ between the personal and the social, but an essential feature of the classic Social Identity model. We saw this in Turner et al. (1992, p.288): “a woman may define herself as intelligent both as an individual compared with other women and as a woman compared with men”. In this example, ‘intelligent’ can be simultaneously a personal attribute (“I am intelligent”) and a social one (“Women are intelligent”), so SCT did not assume a hydraulic or antagonistic relationship between the two. Its claim was that functioning as a distinct individual is antagonistic to functioning as a group-member. This is why, when IFT defines fusion as a merger between ‘the self’ and ‘the group’, it is not describing a new phenomenon, but is merely proposing a new term for the act of defining oneself through one’s group-membership. It is certainly true that fusion theorists are under no obligation to accept SIT’s original formulation of the functional antagonism principle,

2  CONCEPTUAL SHORTCOMINGS OF IDENTITY FUSION THEORY 

23

especially given that their own ideas are framed as a revision of the old approach. IFT falters not because it “fails” to adhere to a particular reading of SIT which ought to be taken as a priori correct merely because it was endorsed by the founders of the approach. Rather, the problem is that IFT fails to trace the implications of its redefinition of the two identities and the relationship between them. Fusion theorists have made the functional antagonism of the two identities—or rather, the rejection of that principle—the focal point of their own theorising, such that the whole theory hinges on the question of what exactly is meant by this antagonism. IFT is thus trapped in a double bind: if it goes with the original interpretation of personal and social identity (as two mutually exclusive levels of self-definition), then the fusion between the two becomes meaningless (in the sense that one cannot be simultaneously a member of a group and not a member of the same group); conversely, if IFT departs from SIT’s original formulation and instead views the two identities as compatible, then a fusion between them becomes possible, but also trivial, as hardly anyone would deny that people can define themselves through their group memberships. It is a dilemma that IFT appears unable to extract itself from, so long as it fails to acknowledge it.

What Would an Identity Fusion Consistent with SIT Look Like? If we were to take SIT’s original conceptualisation of personal and social identity, a fusion between the two would consist of the merger between the ‘me-as-an-individual’ and ‘me-as-a-group-member’ identities. Since these were originally understood as mutually exclusive levels of self-­ definition, IFT would be describing a novel phenomenon that arguably cannot be accounted for under the older Social Identity paradigm. Under these terms, fusion would mean operating on a personal and a collective level of identification simultaneously. That would lead to the pursuit of incompatible goals, interests, and identities. It would require the person to think in terms of “I am different from the other group-members” and “We are alike” in the same breath, without either level of self-definition taking primacy. In terms of agency, it would mean pursuing individual goals that benefit only the individual alongside collective goals that benefit the group. That would be a contradictory and unstable situation, one in which the fragile balance cannot be maintained for long. We could, of course, point

24 

M. SIROMAHOV AND A. HATA

to rhetorical strategies that individuals in such a situation could use to mitigate the conflict between the personal and the social identities—e.g., redefining one of the identities in such a way that it is not incompatible with the other, minimising the centrality of whichever aspect of identity causes the conflict, or framing the conflict as somehow beneficial or desirable. Nevertheless, these would be merely creative ways of resolving or managing the identity conflict, whereas IFT predicts that the conflict itself can strengthen both identities in a ‘synergistic merger’. Fusion theorists have not provided such examples of a synergy between distinct personal and social identities as conceptualised in SCT. In fact, the early papers on IFT are mostly devoid of practical real-world examples of how the described cognitive dynamics play out in real life. Whenever examples are presented, they typically feature people whose characteristics and interests align with those of the group (which in SIT could be the basis of a collective identity) rather than diverging from them. For example, some of the earliest studies that attempted to induce a state of identity fusion did so by priming (exposing experimental participants to subtle subtextual cues) with the idea that individuals share certain “core traits” with their in-group, such as genes or ancestral culture (e.g., Swann et al., 2014). Since then, the idea that beliefs in shared ancestry and culture foster identity fusion has become accepted as a fact in the literature. In contrast, descriptions of situations where a person’s traits mark them as distinct from the group while simultaneously strengthening their group identity are altogether absent from the major publications on IFT. The article “Brothers in arms: Libyan revolutionaries bond like families” by Whitehouse et al. (2014) provides a clear illustration of this trend. In it, the authors report the results of a logistically challenging investigation into the social identities of revolutionaries fighting in the Libyan Civil War that brought down the government of Muammar Gaddafi. The study is notable for investigating identity fusion in a naturalistic setting where self-­ sacrifice in the name of the group is a realistic possibility for many of the combatants. The unusual sample and the historical events unfolding in Libya at the time could have produced deep insights into the complex interplay between the fighters’ personal identities and attachment to the battalion or to the common cause of the revolution. Unfortunately, this route was not taken—the paper primarily reports participants’ fusion scores with a variety of groups, such as their battalion and the Libyan population in general. The authors mention conducting focus groups to develop the questionnaires used in the survey, but this richer qualitative

2  CONCEPTUAL SHORTCOMINGS OF IDENTITY FUSION THEORY 

25

data itself is not reported. Fusion scores are derived from a pictorial measure with two overlapping circles representing the ‘Self’ and the ‘Group’ (more on this in Chap. 4) and apparently no attempt was made to distinguish identity fusion from mere identification with the group. It is unclear whether the researchers were interested in learning more about the contents of their participants’ identities—both personal and collective—and to explore whether they experienced salient individual selves alongside their group-membership, as IFT would predict. Therefore, there seems to be nothing in the reported data to suggest that the fighters were not simply strongly identifying with their units, with the collective identity of the unit taking precedence over their individual differences. The revolutionaries’ commitment to their “brothers in arms” appears to be largely in line with Tajfel and Turner’s observations on social movements in the 1970s—groups of people banding together around a common identity and acting collectively in the group’s interest (over divergent personal interests). When the self takes on the interests of the whole group as its own, that is an example of a self-definition at a collective level—not a synergy between a personal and a social self. A more apt example of a personal identity working alongside the social, as SCT defines them, would be individuals realising that their personal interests diverge from the group’s interests and growing more attached to the group as a result. Interestingly, the authors of the study did include a question attempting to parse the relative personal and collective interests at play: the participants were asked to what extent their personal interests were the same as the interests of the revolution. Those who reported that their personal interests aligned with those of the revolution also expressed more commitment to it. This would not be surprising from the perspective of Social Identity or Self-Categorisation theory: According to SCT, the psychological basis for group behavior is the categorization of self with others, and a depersonalization in perception where one’s unique characteristics fade from awareness and one defines oneself in terms of stereotypical group characteristics. Depersonalization leads to a perceived similarity of needs, goals, and motives, and a mutual and shared perception by in-group members of their interests as interchangeable. (Levine et al., 2002)

In other words, according to SCT, a strong alignment between personal and collective interests would indicate strong identification with the

26 

M. SIROMAHOV AND A. HATA

group—when the self is extended to the level of the group, the individual takes the group’s interests as his own and begins to view them as ‘our interests’. However, confusingly, Whitehouse and colleagues take those same findings as evidence of fusion rather than identification: … self-rated commitment to the goals of the revolution [“To what degree are your personal interests the same as the interests of the revolution? From 0 (extremely different) to 6 (exactly the same)”] was associated with role in the revolution, such that fighters were more personally committed to the goals of the revolution than nonfighters (Whitehouse et al., 2014)

It is unclear how the described dynamic would be different from mere identification with the group—presumably, strongly identifying (rather than fused) group-members would also act in the group’s interests. The phrase ‘personally committed’ in the last sentence could provide a key for interpreting the whole paragraph—the word ‘personally’ sounds redundant at first glance, since in everyday speech we normally describe people as either committed or not. But by qualifying fused people’s commitment as ‘personal’, the text could be interpreted as suggesting that this is a special kind of commitment that comes from the ‘personal self’, which is said to be active in fused people. IFT does not explain what the alternative to personal commitment is—it cannot be a lack of commitment, since the Social Identity tradition has always described people with a strong collective identity as committed to their groups regardless of whether they are ‘fused’ or not, but it could be a ‘non-personal’ or a ‘social commitment’, since non-fused people are supposed to only have an active social identity. The word ‘personal’ is often used in works on IFT to imply that fused people are somehow distinct from non-fused ones—there are frequent mentions of them having ‘personal relationships’ with other group-­ members, caring ‘personally’ about them, or being driven by a ‘personal agency’. In this way, works on IFT often bifurcate a whole range of psychological events and experiences into ‘personal’ and ‘social’, suggesting that fused individuals experience their group-membership and relate to others in a unique way derived from their active ‘personal self’. The distinction is most often implied rather than spelled out directly.

2  CONCEPTUAL SHORTCOMINGS OF IDENTITY FUSION THEORY 

27

On Social Comparison and Dynamic Self-Definition In order to achieve its redefinition of ‘personal’ and ‘social’, IFT must first ignore the degree to which self-construal depends on social comparison according to SIT, and with it—the situational variability that is inherent to people’s personal and social identities. To become ‘fused’, the two identities or ‘selves’ must be more or less stable and enduring mental structures that include all relevant self-knowledge (e.g., Swann et al., 2009 describe the personal self as containing all the “idiosyncratic properties of the individual”, such as ‘being intelligent’ or ‘sociable’). Fusion theorists recognise that the Social Identity tradition views identity as dynamic, but they seem to misinterpret the nature and implications of this dynamism: Highly identified persons should remain devoted to the group only insofar as the immediate contextual influences support such devotion; removal of contextual support for identification may therefore produce corresponding diminutions in identification. By contrast, once fused, people will tend to remain fused. (Swann et al., 2012, p.443)

Here, Swann et al. propose that whereas individuals’ degree of identification with a group is situationally variable, their identity fusion is permanent. But the Social Identity tradition has always claimed that it is the contents of group identity, not merely its strength, that varies between contexts. In SCT, there is no singular, monolithic ‘personal self’ that can fuse with a similarly monolithic ‘collective self’. Instead, both identities are constructed situationally based on social comparisons with other individuals and groups: The situational specificity of self-images (see Gergen 1971) implies that people have a capacity to regulate their behaviour in terms of different self-­ conceptions in different situations. Different situations seem to ‘switch on’ different self-schemata so that social stimuli can be perceived and social behaviour controlled in the appropriately adaptive manner. (Turner, 1984, p.527)

For example, ‘being intelligent’ and ‘being sociable’ should become parts of one’s personal identity only when compared to an in-group that lacks these attributes, otherwise they would not be personally distinctive. Conversely, if the group also has the reputation of being intelligent and sociable, then these traits in the individual group-member will come to

28 

M. SIROMAHOV AND A. HATA

exemplify their social (not personal) identity. It follows then that the ‘personal self’ does not have temporally stable contents that can be kept distinct from the ‘social self’ by an impermeable boundary, as fusion theorists claim—rather, aspects of identity can become either individual or collective, relevant or irrelevant, depending on the social comparative context. SCT’s idea that the self exists at different levels of categorisation renders the idea of fusion meaningless—there is only one self (albeit contextually functioning as a distinct individual or as a group-member), so there is nothing else for it to become fused with. Therefore, the problem for IFT remains—if the identity of the relevant ‘Others’ that one is being compared to determines which traits are seen as personally distinctive and which are shared with those Others, then there are no permanent and stable personal and social identities that can become permanently fused.

Conclusion: Hiding the Redefinition In this chapter, we have argued that Identity Fusion Theory actually describes at least two distinct forms of fusion. The first is the merger of an individual person and the group—a metaphorical description of some deep sense of belonging which seems to be more emotionally charged and integral for the fused person’s sense of self than mere identification with a group, but is also poorly defined and not always clearly differentiated from it. The basic process—the integration of the group-membership into one’s sense of self—seems to describe what SCT understood by self-definition at the level of group-membership, i.e., social identity tout court. The second form is the merger of the personal and social identities in the individual’s mind, which has the benefit of being a novel claim, but it also rests on an apparent misunderstanding of what SCT meant by ‘personal’ and ‘social’ identity, by framing ‘personal identity’ as including all self-knowledge that a person has. Defined in this way, the personal and social identities are indeed compatible, but in a way that reduces IFT to common sense: it says that some individuals have characteristics and goals that are also shared with the group and reinforce their commitment to the group. That much is true, but to interpret that as a fusion between ‘personal’ and ‘social’ selves is to stick new labels to a dynamic that was already acknowledged in the Social Identity tradition.

2  CONCEPTUAL SHORTCOMINGS OF IDENTITY FUSION THEORY 

29

References Brygoła, E., Zięba, M., & Kwapis, K. (2021). The Identity Dynamics Questionnaire (IDQ): Operationalization and verification of the identity dynamics model. Polish Psychological Bulletin, 52(3), 270–288. Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7, 117–140. Henríquez, D., Urzúa, A., & López-López, W. (2020). Identity fusion: A systematic review. Acta Colombiana de Psicología, 23(2), 410–437. Hopkins, N., Reicher, S., Harrison, K., Cassidy, C., Bull, R., & Levine, M. (2007). Helping to improve the group stereotype: On the strategic dimension of pro-­ social behavior. Personality and Social Psychology Bulletin, 33, 776–788. Levine, M., Cassidy, C., Brazier, G., & Reicher, S. (2002). Self-categorization and bystander non-intervention: Two experimental studies. Journal of Applied Social Psychology, 32(7), 1452–1463. Sherif, M. (1967). Group conflict and co-operation: Their social psychology. Routledge & Kegan Paul. Swann, W. B., Jr., Gómez, Á., Buhrmester, M. D., López-Rodríguez, L., Jiménez, J., & Vázquez, A. (2014). Contemplating the ultimate sacrifice: Identity fusion channels pro-group affect, cognition, and moral decision making. Journal of Personality and Social Psychology, 106(5), 713–727. Swann, W. B., Jr., Jetten, J., Gómez, Á., Whitehouse, H., & Bastian, B. (2012). When group membership gets personal: A theory of identity fusion. Psychological Review, 119(3), 441–456. Swann, W. B., Jr., Gómez, Á., Seyle, D. C., Morales, J. F., & Huici, C. (2009). Identity fusion: The interplay of personal and social identities in extreme group behavior. Journal of Personality and Social Psychology, 96(5), 995–1011. Tajfel, H., & Turner, J. (1979). An integrative theory of intergroup conflict. In S. Worchel & W. G. Austin (Eds.), The social psychology of intergroup relations (pp. 33–47). Brooks-Cole Publishing. Tajfel, H., Billig, M. G., Bundy, R. P., & Flament, C. (1971). Social categorization and intergroup behaviour. European Journal of Social Psychology, 1(2), 149–178. Turner, J. C., & Oakes, P. J. (1997). The socially structured mind. In T. Postmes & N. R. Branscombe (Eds.), Rediscovering social identity [2016]. Routledge. Turner, J. C., & Onorato, R. S. (1999). Social identity, personality, and the self-­ concept: A self-categorization perspective. In T. Postmes & N. R. Branscombe (Eds.), Rediscovering social identity [2016]. Routledge. Turner, J. C., & Reynolds, K. J. (2010). The story of social identity. In T. Postmes & N. R. Branscombe (Eds.), Rediscovering social identity [2016]. Routledge. Turner, J. C. (1975). Social comparison and social identity: Some prospects for intergroup behaviour. European Journal of Social Psychology, 5, 5–34.

30 

M. SIROMAHOV AND A. HATA

Turner, J.  C. (1982). Towards a cognitive redefinition of the social group. In T. Postmes & N. R. Branscombe (Eds.), Rediscovering social identity [2016]. Routledge. Turner, J.  C. (1984). Social identification and psychological group formation (Chapter 25). In H.  Tajfel (Ed.), The social dimension (Vol. 2). Cambridge University Press. Turner, J.  C., & Haslam, S.  A. (2012). Social identity, organizations, and leadership. In M.  E. Turner (Ed.), Groups at Work: Theory and Research (pp. 25–66). Routledge. Turner, J. C., Oakes, P. J., Haslam, S. A., & McGarthy, C. (1992). Self and collective: Cognition and social context. In T. Postmes & N. R. Branscombe (Eds.), Rediscovering social identity [2016]. Routledge. Whitehouse, H., McQuinn, B., Buhrmester, M., & Swann, W. B. (2014). Brothers in arms: Libyan revolutionaries bond like family. Proceedings of the National Academy of Sciences, 111(50), 17783–17785.

CHAPTER 3

The Limitations of Extreme Cognitivism

Abstract  This chapter critiques the ways in which Identity Fusion Theory’s hard cognitivist approach distorts the meaning of established concepts in the Social Identity tradition. IFT appears to treat the personal and social identities or ‘selves’ not as two modes of social being but as mental entities that produce extreme behaviours through their mechanical interactions. This allows it to claim that an individual can form ‘personal relationships’ with in-group strangers on a large scale by ‘projecting’ relational ties onto them, in a process that is poorly theorised and lacking in empirical support. Keywords  Social identity • Self-categorisation • Identity fusion • Cognitive psychology • Agency • Motivation • Familial ties • Relational ties In addition to the conceptual confusion around what exactly is fused in identity fusion, IFT is hampered by its hardline cognitivist approach to social psychology, which reduces what are essentially social phenomena to the operations of cognitive machinery in people’s minds. In this chapter, we will explore the consequences of this unchecked cognitivism for IFT’s theorising, by using three prominent examples from the fusion literature: the idea that the ‘fusion’ of multiple ‘selves’ in the mind can drive pro-­ group behaviour; the notion of motivational energy or agency as used in © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 M. Siromahov, A. Hata, Rethinking Identity Fusion, https://doi.org/10.1007/978-3-031-46983-1_3

31

32 

M. SIROMAHOV AND A. HATA

IFT; and the idea that relational ties can be ‘projected’ onto strangers, creating familial bonds with whole nations.

Identities or Mental Entities? Identity Fusion Theory’s hard cognitivism is most clearly seen when it deals with the personal and social identities or ‘selves’. As we saw in Chap. 2, the Social Identity tradition generally understands them as two alternative levels of self-definition or, in John Turner’s words, as “approximations” describing two distinct modes of being. In contrast, fusion theorists regularly describe these ‘personal and social selves’ as object-like components of the mind which are said to be separated by ‘boundaries’, much like single-cell organisms separated by membranes: At the heart of the social identity perspective lays the distinction between the personal and social self (e.g., James, 1890; Tajfel & Turner, 1979).1 Whereas the personal self refers to idiosyncratic properties of the individual (e.g., “intelligent,” “sociable”), the social self refers to those aspects of self associated with group membership (e.g., “Democrat,” “American”). (Swann et al., 2012, p. 442) … whereas most people experience clearly demarcated boundaries between their personal and social selves, those who are highly fused with a group experience this boundary as porous and permeable. […] In fact, these boundaries become so permeable that aspects of both the personal and social self can readily flow into the other. (Swann et al., 2012, p.442) The authors propose that this disposition to extreme sacrifice would be motivated by a highly salient personal and group identity between individuals, with a visceral feeling of deep union between the personal self and the 1  These two references look like they might provide more clarity on the ‘personal and social selves’ described by IFT, but upon closer inspection they do not. While William James (1890) did discuss people having ‘social selves’, he clearly meant it in the sense of the different social roles that a person might have (he gives the example of the roles of a soldier, a priest, and so on). Central to his conceptualisation was the way a person was being perceived by others, in that a person has as many selves as there are people who know him or her. This is substantially different from the ‘social self’ being a singular cognitive representation of a group-membership, as IFT is assuming. Tajfel and Turner’s (1979) publication also does not provide us with more clarity, as it describes a continuum of individual versus collective behaviour, not different images of the self—the idea of the continuum between individual and collective identity would appear later, with Turner’s (1982) Self-Categorisation Theory.

3  THE LIMITATIONS OF EXTREME COGNITIVISM 

33

social self, so that the delimitation between both identities becomes indistinguishable (Henríquez et al., 2020)

IFT’s notion of the two ‘selves’ as mental entities appears to reflect a cognitive-structuralist view of the mind, according to which all of one’s self-knowledge is organised in a cognitive structure of self-representations (schemata). Structuralists understand these schemata as containing different and even contradictory bits of information, some more central to one’s identity than others, and some of which can be more ‘activated’ or ‘accessible’ at a given time than others (e.g., one becomes much more aware of being “an English-speaker” when travelling in a non-anglophone country than in England). Identity Fusion Theory seems to understand the ‘personal’ and the ‘collective’ self in precisely this way—as two interlinked cognitive schemata stored in the mind that can be either distinct or ‘fused’. If Identity Fusion Theory does take a structuralist approach and understands personal and social identity as cognitive schemata, as it appears to be the case, that puts the theory at odds with the Social Identity tradition it ostensibly comes out of. Turner and Oakes have explicitly challenged the structuralist cognitivist model of the self: Recent work on the self, however, has tended to borrow theories and concepts from individual cognitive psychology in order to devise models of the self which reduce it to a cognitive structure […]. The self is seen as a relatively fixed, separate mental structure that is activated according to the situation […]. This idea of the self as a fixed cognitive structure or system that processes information raises some fundamental and difficult theoretical questions. […] Our research suggests a different approach […] The self is a varying, reflexive representation of the perceiver which is inherently fluid and flexible because it is a comparative, relational judgement. (Turner & Oakes, 1997, p. 308, emphasis in original)

Likewise, Turner et al. (1992) emphasised that the difference between the personal and social identities is not a matter of informational content: The difference between personal and social identity is not a matter of the attributes that define the categories or of the abstract level of inclusiveness of the categories used to define self. For example, a woman may define herself as intelligent both as an individual compared with other women and as a woman compared with men. […] What matters is how the self is actually

34 

M. SIROMAHOV AND A. HATA

being defined in a specific instance, the level of comparison and self-­ categorization that is actually taking place, and the subjective sense of self that results. Is the perceiver being defined as an individual person or as a social group […]? (Turner et al., 1992, p.288)

Turner and Onorato (1999) have also argued that, although SCT’s symbolic interactionist approach can be reconciled to some extent with the cognitivist structural view of the self as a “bounded cognitive structure” (p.  317), the self cannot be reduced entirely to self-knowledge stored in the mind. People’s self-knowledge might consist of innumerable and sometimes contradictory pieces of information, but it is the social comparative context that determines which pieces are picked up in a specific context to serve as the raw materials for self-definition. In describing the personal and social ‘selves’ (or identities) as mental entities, fusion theorists seem to be reifying what was originally meant to be a metaphor for social behaviour. The ‘self’ in SIT serves as a metaphor for people’s ability to act from different standpoints or subjectivities: for example, the same person can think and act as a social scientist, as a mother, as a German, etc.; each standpoint that a person can take brings into focus certain characteristics, behavioural norms, and goals that are relevant to that identity. The metaphor of having multiple potential identities serves to illustrate this pluripotentiality of self-definition and social behaviour. As we saw in Chap. 2, John Turner himself claimed that the distinction between personal and social identity is more of a metaphor (an “approximation”) for this feature of human psychology, rather than a structural fixture of the mind. But when fusion theorists describe the personal and social identities, they do not seem to use them as a metaphor to illustrate a deeper social reality—they present them as really existing things in the mind which, through their dynamic relationship, cause behaviours to happen. For example, the fusion of the two selves is consistently described as producing or causing specific feelings, attachments, and actions (e.g., it fosters strong ties to individual group-members; it leads to perceiving the group as a family; it makes people feel invulnerable)—so it cannot be treated merely as a figurative description of these things. Likewise, the two selves are described as having a ‘boundary’ which normally prevents ‘aspects’ of the two identities from mixing, but in fusion that boundary becomes porous and the ‘aspects’ of the identities start flowing freely across it:

3  THE LIMITATIONS OF EXTREME COGNITIVISM 

35

… these boundaries become so permeable that aspects of both the personal and social self can readily flow into the other. The flow of influence may move in both directions: Just as highly fused persons come to view themselves through their group membership (“My group membership is a crucial part of who I am”), they also perceive the group through their personal self (“I am an important part of the group”) […] the personal and social identities of highly fused persons may combine synergistically to motivate pro-­ group behavior (Swann et al., 2012, p.442-3)

In this example, it is unclear what these “aspects” of the two selves are, because the ‘selves’ are poorly defined. Up to this point, IFT has sounded like a structuralist theory, i.e., viewing the ‘selves’ as cognitive schemata containing self-knowledge (specifically, self-descriptors like “I am intelligent”). If that is really the case, then the collapse of the ‘boundary’ and the free flow of ‘aspects’ from one self to the other could be intended to describe a confusion about the informational contents of the self—the ‘fused’ person becomes confused as to which traits are their own and which are the group’s (e.g., if I am fused with America and I am an atheist, I will start to believe that Americans tend to be irreligious like me—conflating my own stance on religion with the group’s). But we can deduce that identity fusion cannot mean such a conflation of self- and group-­ relevant knowledge, since fusion theorists have explicitly said that fusion means the maintenance of a distinct personal self alongside the social—not a conflation of the two. As we will see later in this chapter, fused people are said to not conform to the group’s prototype, which makes them less group-like in their self-perception and behaviour. This is the main thing that is supposed to distinguish fusion from mere social identification—it is identification that supposedly causes group-members to become more group-like, while fusion preserves their ‘personal self’ and individuality: … identification refers to an alignment of people’s personal identities (i.e., aspects of self that make people unique) and social identities (i.e., aspects of self that align them with groups—e.g., being Jewish or Catholic). As identification increases, however, the personal self fades into the background and people come to see themselves as exemplifying qualities of the collective category. This shift encourages them to adopt prototypical behaviors, such as favoring the in-group over the out-group (e.g., Turner et al., 1987). In contrast, although strongly fused persons align themselves with the collective, they nevertheless retain an agentic personal self and cultivate

36 

M. SIROMAHOV AND A. HATA

close ties to fellow group members as well as to the collective category. (Swann & Buhrmester, 2015)

Therefore, whatever ‘flows’ across the boundary of the selves is not informational content. Swann et  al.’s (2012) example of fusion in the quote above concerns beliefs about the individual’s place in the group (“I am an important part of the group”), rather than fusion with the individual’s social identity (i.e., ‘me-as-a-group-member’). The definitional problem appears again: what exactly is fused in identity fusion—the individual and the group, or the individual’s two identities? What adds to the confusion is that the individual’s perceived position in the group—one of the exceedingly rare examples of identity fusion in this seminal paper on fusion—is rarely used as a criterion for detecting identity fusion in empirical research. If the ‘porous boundary’ between the ‘personal and social selves’ does not describe the two becoming conflated in the individual’s mind, it is also left unclear what it does describe. If the two ‘selves’ do not carry informational content, then what is it? Surprisingly, fusion theorists often describe them not so much as having contents, but as actively doing things—as we will see in the next two sections, the ‘personal self’ is said to provide ‘personal agency’, to care about individual group-members, or to have its own perceptions (e.g., perceiving others as relatives). The two ‘selves’ and the blurring of the boundary between them do not seem to be metaphorical descriptions of how one’s self-knowledge or attitudes change—they are described as if they are really-existing things which mechanically produce these consequences. For example, Swann et al. (2012) use the interaction of the two selves as a causal mechanism explaining why people come to perceive a group as family: The tendency for highly fused persons to maintain permeable borders between their personal and social selves will not only magnify their feelings of connectedness to the group category, it will also foster connections to other ingroup members. […] In short, the state of identity fusion refers to a powerful union of the personal and social self wherein the borders between the two become porous without diminishing the integrity of either construct. The result is a powerful feeling of connectedness.

Here, the fusion of the two selves is not treated as a figurative or metaphorical description of a person’s deep attachment to the group, but as the

3  THE LIMITATIONS OF EXTREME COGNITIVISM 

37

mechanical cause of that attachment. This is far removed from SIT’s original understanding of personal and social identity as different ways of viewing oneself, but it also sounds less and less like a standard cognitivestructuralist theory, since we are dealing not so much with cognitive structures as with mental entities interacting seemingly on their own in the individual’s mind. In this way, the two selves somewhat resemble the Freudian agencies of the Id and the Ego, so IFT’s metapsychology in places sounds almost psychodynamic.

Agency and Depersonalisation The redefinition of ‘personal identity’ that we explored in Chap. 2 is also at play when fusion theorists propose that ‘fused’ people have a ‘potent personal self’ which enables them to commit extreme actions in service of the group. Swann et al. (2009, 2012) correctly observe that, according to the Social Identity model, a person’s identity plays a role in motivation, but they once again interpret this in an overly mechanistic way. As we have shown, one’s identity brings into focus specific goals, needs, and interests, and relegates others to the background. This is what Turner (1984) means when he says that “people have a capacity to regulate their behaviour in terms of different self-conceptions”: a person’s identity (whether personal or collective) is one of the factors influencing which goals, interests, and needs the person will pursue (e.g., one’s narrowly individualistic interests, or the interests of one’s family, colleagues, or nation). In other words, a collective identity can frame certain goals as ‘mine’ or ‘ours’, or frame a slight against a social category as a slight against ‘us’, or allow behaviour in groups to be “regulated by reference to social standards” (Reicher et al., 1995). However, because fusion theorists understand identity as a mental entity, they conclude that personal and social identity are themselves generators of ‘motivation’ and thus drive behaviour. The identity fusion literature routinely describes fused people as having a “motivationally potent personal self” (Swann et al., 2009), “a strong sense of personal agency that they channel into pro-group behavior” (Swann et al., 2012), and an “agentic self” (Newson et al., 2021). The ‘personal self’ thus seems to be acting like a dynamo producing ‘motivation’ and ‘feelings of agency’ which compel people to act, rather than merely providing a particular self-image that directs action, as in SIT. And, since the two identities are said to function simultaneously in cases of identity fusion, ‘fused’ people must have two

38 

M. SIROMAHOV AND A. HATA

such dynamos producing a double dose of motivation, which pushes them towards extreme pro-group actions: The personal and social identities of highly fused persons may combine synergistically to motivate pro-group behavior, thereby producing additional motivational “oomph.” (Swann et al., 2012) When this strong autonomous self becomes merged with the group, it can provide the motivational machinery needed for taking radical action on behalf of the group. (Swann et al., 2009)

IFT seems to attribute to SIT a view of social behaviour that the original theory did not express, one in which identity is “machinery” which provides “motivational oomph” on its own and directly “fuels” (Swann et al., 2012) behaviour. And, since non-fused group-members are said to lack an active personal self, IFT strongly implies that identifying with a group entails some loss of volition and an abdication of personal interests in service of the group’s needs. Swann et al. (2012) summarise their interpretation of SIT thus: The social identity perspective holds that when a social identity is salient, the actions of highly identified persons are regulated by a “depersonalized” social self associated with the group; feelings of personal agency2 presumably play no role in pro-group activities. (Swann et al., 2012, p.443)

They then offer a revision to SIT by proposing that in ‘fused’ people the two sets of motivations do not suppress each other but instead “work together”: In contrast, when people fuse with a group, they do not temporarily abdicate their personal self. Rather, when highly fused persons enact pro-group activity, their actions reflect both their personal and social identities, working together by virtue of the porous borders that define them. (Swann et al., 2012, p.443) 2  We again come across the peculiar use of the word ‘personal’ implying a distinction between personal and social without stating so directly. When people act, we say that they have agency, full stop. When the authors qualify it as ‘personal agency’ specifically, they could be suggesting that there is an alternative kind of agency that non-fused people have—perhaps a social or collective agency? We saw a similar turn of phrase (“personally committed”) in Whitehouse et al. (2014) in Chap. 2.

3  THE LIMITATIONS OF EXTREME COGNITIVISM 

39

The problem of IFT redefining the ‘personal self’ presents itself again. If Self-Categorisation Theory understands ‘personal identity’ as built on traits that mark someone as individually distinct from the group,3 then ‘fusion’ between the two would have to describe a situation in which a person’s individual interests are distinct and incompatible with those of the group, while this incompatibility simultaneously strengthens the person’s social identity and leads to a synergy with the group, rather than a conflict. Papers on identity fusion have not, to our knowledge, provided any such examples. On the contrary, whenever fusion theorists discuss the ‘fusion’ of personal and collective motivations, these include instances where the two coincide. For example, ‘fused’ members of militant groups are said to perceive the other group-members as family (Whitehouse et al., 2014), so fusion theorists take this as an example of the fusion of motivations coming simultaneously from the personal identity (“I need to protect my comrades”) and the collective identity (“I need to protect my combat unit as a whole”). The problem with labelling such an overlap between personal and collective motivations as ‘fusion’ is that its existence was already acknowledged in SIT, and even in earlier social-psychological thought. Hardly anyone would dispute the claim that people become more committed to a group’s cause if they also have a personal stake in it. Even prior to the formulation of SIT, social psychologists largely agreed that people will form groups and collaborate in pursuit of a shared goal; Tajfel and Turner did not dispute this either. Their novel idea was that people can form groups in the absence of such instrumental motivations: the perception of the group as a collective ‘us’ and of the group’s interests as ‘our interests’ can enable cooperation even in the absence of shared self-interest. This is also why we should not attribute to SIT the view that having a strong collective identity means abandoning all of one’s interests and submitting entirely to the will of the group or its leader. Collective identification does not deprive people of volition, agency, or identity. It merely allows them to recognise the group’s goals and interests as their own. 3  See e.g. Turner and Onorato (1999), p.324: “Depersonalisation is a dynamic process. It’s the shifting of self-definition away from individual differences and towards social similarities. Social identity also produces socially unitary collective behaviour; it produces a mutual orientation of attraction, cooperation and influence as members define and react to each other in terms of their common social category membership rather than as differing individuals.” Also Reicher et al. (1995, p.177): “Turner … refers to this shift from relating to others on an individual level to relating on a higher level of inclusiveness as a process of depersonalization.”

40 

M. SIROMAHOV AND A. HATA

Turner explicitly rejected the idea that social identity makes people submit to the group or act like a mindless herd: Depersonalization, however, is not a loss of individual identity, nor a loss or submergence of the self in the group (as in the concept of ‘deindividuation’) [...] It is the change from a more personal to a more social identity, a change in the nature or content of the self-concept corresponding to the functioning of self-perception at a more inclusive level of abstraction. In many respects depersonalization may even be seen as a gain in identity because it represents a mechanism whereby individuals may act in terms of the social similarities and differences produced by the historical development of human society and culture (Turner, 1985, p. 258)

In Search of the Lost Agency So far, we have been arguing that Identity Fusion Theory seems to rest on a misinterpretation of how agency is understood in the classic works of the Social Identity tradition. However, it is still possible that fusion theorists are working with a more recent reinterpretation of SIT which does conceptualise personal and social identity as cognitive dynamos producing quantities of motivational energy independently of each other. Swann et al.’s (2009) first paper on the topic does not cite research that specifically supports this idea, however Swann et al. (2012) cite two sources: a paper by Levine and Crowther (2008) and another by Hopkins et al. (2007). Levine and Crowther’s (2008) article offers a critique and a reinterpretation of the classic literature on the bystander effect, which in its traditional formulation held that the presence of bystanders reduces a person’s sense of responsibility and makes them less likely to help another person in need. The authors argue that researchers had up to that point ignored the role of social identity by studying the behaviour of groups of bystanders who neither knew each other nor had a shared collective identity. The paper’s main claim is that the presence of a physical group is not enough to enable helping behaviours—the group must be experienced psychologically as real, i.e., there needs to be a sense of shared identity with the other bystanders and with the victim. Levine and Crowther’s experiments also demonstrated the interplay between social identity and ideology (in this case, the ideological norms of masculine behaviour): while women were more likely to intervene and help a stranger when surrounded by

3  THE LIMITATIONS OF EXTREME COGNITIVISM 

41

their in-group (a crowd of other women), male participants were more likely to intervene when in the presence of an out-group (also a crowd of women). The authors attribute this difference to the effect of male “heroic and chivalrous norms”, which seem to become relevant in the presence of women rather than other men. Altogether, we do not see a straightforward effect of in-group identity suppressing agency—there is a complex interplay between in-group identity, group size, and social norms in guiding people’s decision to intervene. Levine and Crowther (2008) seem to suggest that a collective identity is necessary to facilitate pro-group action in defiance of the  conformity  of the crowd and the bystander effect, whereas IFT would attribute the same outcome to the agency of ‘the personal self’ and fusion, not identification. The second paper cited by Swann et al. (2012) to support the notion that identification brings about a loss of personal agency is also critical of the classic bystander effect hypothesis. Hopkins et al. (2007) discuss the social functions of helping behaviours, reject older theories about the “dilution” of individual responsibility in crowds, and remark that the presence of a strong group identity and collective norms for helping often do motivate people to help other in-group members. The authors support their critique with references to experimental research by Simon, Stürmer, and Steffens (2000, cited in Hopkins et al., 2007) showing that a strong in-group identity promotes charitable giving, and archival research by Reicher and colleagues (2006, cited in Hopkins et al., 2007) demonstrating that public figures campaigning to save Bulgaria’s Jews during the Holocaust discursively framed Jews as part of the Bulgarian nation (i.e., as in-group members) and emphasised Bulgarian cultural norms of opposing oppression. The paper further extends these findings to argue that, in some cases, a strong social identity can motivate people to engage in proactive helping of out-groups, not just in-groups, especially when that would help to disprove a harmful stereotype about their in-group. Neither of the papers cited by Swann et al. (2012) suggests that we can distinguish whether a ‘personal self’ or a ‘social self’ is driving such pro-­ group actions, or that identification with a group implies a heightened state of conformity or loss of personal identity (in the IFT sense of ‘personal’). The studies reported therein demonstrate that collective identity is implicated in helping behaviours in complex ways and is dependent on factors as diverse as historical context, cultural norms, and social ideology. It is revealing, however, that the authors of these publications, themselves working within the Social Identity tradition prior to the invention of

42 

M. SIROMAHOV AND A. HATA

Identity Fusion Theory, did not see a problem with the notion that social identification enables personally costly prosocial behaviour. Participants in studies on the bystander effect would have risked personal injury by intervening to protect an in-group stranger, and the Bulgarians who campaigned to save their Jewish neighbours would have known that they were putting their lives on the line by defying the Nazis’ deportation orders. This is precisely the kind of costly pro-group behaviour that, according to fusion theorists, can only be derived from identity fusion rather than mere identification. And yet, researchers working within the Social Identity tradition prior to the invention of IFT have interpreted such behaviours as stemming from social identification—they seem unaware of the idea, attributed to them by fusion theorists, that identification should extinguish ‘personal agency’ and induce a state of passive conformity. Likewise, when Swann and Buhrmester’s (2015) claim that “fused persons care about individual members of the group as well as the abstract collective”, they are implying that strongly identifying persons only care about the abstract collective and not about individual group-members.4 This assumption is absent from the earlier SIT research, in which social identity is routinely shown to be accompanied by behaviours intended to help individual group-members. In fact, this distinction between ‘individual members’ and ‘the group in the abstract’ would have been difficult to maintain in practice—behaviours aimed at helping the abstract idea of a group and not concrete group-members would be so rare and ineffective as to be pointless. Does Identity Fusion Account for Extreme Pro-group Actions? Leaving aside the question of whether Identity Fusion Theory accurately summarises the Social Identity approach to motivation and pro-group actions, it is worth exploring whether it provides a robust account of these dynamics on its own. How exactly does the ‘active personal self’ fuel 4  Here we see another version of the ‘personal vs. social’ dichotomy discussed in Chap. 2. In this instance, the authors seem to be implying that each ‘self’ can form attachments to specific targets: attachment to the group as an abstraction comes from the ‘social self’, and caring about individual group-members—from the ‘personal self’. This distinction does not seem to feature anywhere in the Social Identity approach and seems to come instead from IFT’s insistence on splitting every psychological phenomenon into two along the lines of the ‘personal/social’ dichotomy. Also, note that the distinction between these two forms of caring is never used to distinguish between fused and non-fused people in practice.

3  THE LIMITATIONS OF EXTREME COGNITIVISM 

43

extreme behaviour? Fusion theorists have provided at least three distinct explanations in different places. The first—we will call it the quantitative explanation—we have already seen glimpses of: it is the proposition that each ‘self’ generates ‘agency’, a kind of psychological energy that motivates behaviour. The simultaneous activation of the personal and social selves gives the fused person “an extra motivational oomph” (Swann et al., 2012), a double dose of motivation, so to speak, which makes them go above and beyond others in their pursuit of the group’s cause. In other words, both fused and non-fused members have some degree of motivation to benefit the group, but fused ones simply have more of it: Fighting or dying for one’s group represents a profound statement regarding one’s allegiance to the group. As such, simply noting that one shares many qualities with other group members may not be enough to motivate such extreme behavior. Instead, only people who possess extraordinarily high levels of motivation should theoretically be willing to engage in such extreme actions on behalf of the group. […] The possession of a potent personal self is especially relevant here because it may help fuel extreme behavior. (Swann et al., 2009)

This explanation seems to assume that it is the quantity of motivation that determines whether a person chooses to engage in moderate pro-­ group actions (like donating a small amount of money) versus extreme ones (like a suicide bombing). The authors seem to be suggesting that the main barrier preventing non-fused members from taking extreme actions for the group is insufficient motivation—these members would in principle be open to committing atrocities, but because they only have the agency of the ‘social self’ to drive them, they cannot quite muster the strength to do so; in contrast, fused people are similarly driven by an active ‘social self’, but in addition to it they get the extra dose of motivation from the ‘personal self’. With the two selves combined, ‘fused people’ are bursting with agency, and the additional quantity of this ‘motivational energy’ is enough to push them towards extreme pro-group actions. This seems like an exceptionally narrow idea of how radicalisation occurs and what range of factors can drive people to commit atrocities. The authors are correct that awareness of within-group similarities, on its own, is not enough to explain even more mundane examples of ethnocentrism, let alone why a person would give up their life or commit terrorist acts for the group, but theorists working in the Social Identity tradition

44 

M. SIROMAHOV AND A. HATA

have never claimed that to be the only factor. Social scientists (Turner & Onorato, 1999; Turner, 1984) have insisted on the decisive role played by concrete intergroup dynamics, the historical and social context, the person’s individual life trajectory (e.g., the absence of alternative sources of positive identity besides the group), the socially shared ideologies that are used to dehumanise out-groups and to render violence as the natural and moral course of action, etc. Once all of these additional factors have been accounted for, they might go a long way to explain radicalisation and specific individuals’ journey towards extremism, without positing the existence of a special psychological state like ‘identity fusion’ or some novel source of surplus agency. A second, more interpersonal, explanation for why fusion provides motivation for pro-group actions appears throughout the identity fusion literature. Whereas the first hypothesis referred to the ‘agencies’ of a person’s individual and collective identities (i.e., two components of a person’s self), the sociological explanation deals with the agencies of the individual and the group. It holds that non-fused group-members believe that only the group-as-an-abstraction has agency, whereas fused members believe that the agency of other individual group-members is added on top of it. The fused individual thus comes to believe that, just like he is willing to die for the group, so are other group-members willing to die for him, creating a feeling of invulnerability: … highly fused persons will feel that they and other group members synergistically strengthen each other. This perception of reciprocal strength should foster the perception that together, the members of the group are uniquely invulnerable. (Swann et al., 2012) Insofar as fused persons believe that they and other group members synergistically strengthen each other, fused persons may conclude that they are invulnerable relative to actors who do not enjoy complete confidence that their fellow group members will act on their behalf. (Gómez et al., 2011)

Finally, Swann et al. (2009) offer a third, qualitative explanation of the ‘agentic personal self’ principle: non-fused people can only conform to the group’s stereotypes and behavioural norms (because they only have an active social self), so their pro-group behaviours will be stereotypical and conformist; meanwhile, fused people have an active ‘personal self’ which enables them to step outside the boundaries of conventional group behaviour and thus go to extremes:

3  THE LIMITATIONS OF EXTREME COGNITIVISM 

45

Depersonalized individuals may be well suited for falling in line and obeying orders issued by the group leader, but they lack the initiative to enact extraordinary actions for the group, as such activities are, by definition, nonprototypical. The opposite is true for fused persons. When people become fused with a group, they do not relinquish their sense of personal identity in favor of their group identity nor do they come to regard themselves as undifferentiated members of the group. (Swann et al., 2009)

This version of the ‘agentic personal self’ principle implies that the contents of group identity should be at the forefront of any study of identity fusion. If the argument is that non-fused members will act in accordance with the group prototype, whereas fused people will act in ways unusual for the group, then it becomes impossible to distinguish between fused and non-fused people without knowing what their in-group’s prototype is. This idea does not seem to have been seriously pursued by much subsequent research on identity fusion, as studies rarely engage in qualitative analyses of group identity—instead, most research on IFT only deals with quantitative measurements of the strength of fusion or identification, without concern as to which actions are seen as prototypical and which are not. This is in line with IFT’s tendency to treat both benign and terrorist pro-group actions as lying on the same spectrum, with the main difference between them being the ‘quantity’ of ‘motivational energy’ that fuels them. More importantly, the principle poses a problem for the study of fusion in groups where violence is a normative part of the group identity. Swann et al. seem to be arguing that mere identification can only generate conventional (stereotypical, group-normative) pro-group behaviours, whereas fusion can generate unconventional pro-group behaviours that defy the norms of the collective identity. In this context, ‘extreme’ does not mean actions that are unusual in the scale of their brutality or violence, such as suicide bombings, but non-normative acts that deviate from the group’s behavioural norms, which makes IFT not so much a theory of political extremism as one of unconventional pro-group behaviour. By this definition, when we are dealing with a group whose identity normalises violence and brutality, it should be the non-fused (conformist) members who commit extremist acts of violence, while fused (individualistic) members should be free to help the group in more benign ways, if they so choose. This would seem to be particularly relevant for the types of social groups studied by identity fusion researchers: terrorist cells (Atran, 2017; Ferguson & McAuley, 2020; Corner et al., 2021; Gómez et al., 2021), military units (Whitehouse et  al., 2014; Hart & Lancaster, 2019), and football ultras (Newson et  al., 2018; Newson, 2019, 2021; Knijnik &

46 

M. SIROMAHOV AND A. HATA

Newson, 2021), all of which would have a certain degree of violence and conflict as their reason for being, and it would be expected of any ordinary member to be willing to fight and (in the case of terrorists and soldiers) die for the group. In such cases, researchers would not be able to distinguish between fused and non-fused extremists, as some group-members could be committing extreme acts out of conformity (caused by identification) rather than out of ‘personal agency’ (associated with fusion).

The Problematic Concept of Relational Ties Identity Fusion Theory’s extreme cognitivism becomes particularly problematic when it attempts to extend the idea of micro-level solidarity (close and personal relationships with concrete individuals) onto larger, impersonal groups, where such solidarity could no longer realistically apply. This is evident in the claim, prevalent in the literature, that the same cognitive processes that constitute identity fusion at the scale of small, personal groups can operate at the level of large social movements or whole nations. Fusion theorists often write about how the merger of the ‘personal and social selves’ leads to ‘relational ties’ being ‘projected’ onto other group-­ members, causing an individual to perceive those people as family. The projection of familial ties is foundational to IFT—it is not a mere correlate of identity fusion, but the very mechanism through which identity fusion is said to exert its influence. Apparently, highly fused persons view their group members as fictive family members, and these perceptions motivate them to take extreme actions on the behalf of these individuals. […] But if the perception of family ties motivates willingness to self-sacrifice, why do people die for large groups and abstract causes? One possibility is that extended fusion entails the projection of relational ties onto genetically unrelated group members, thereby transforming them into fictive kin (Atran, 2010). This projection process could persuade highly fused persons to sacrifice themselves for members of a heterogeneous group. (Swann & Buhrmester, 2015) … identity fusion is associated with relational ties (i.e., feelings toward individual group members) and collective ties (i.e., feelings toward the group as a whole […]), which create the perception of connection and reciprocal strength between personal identity and group identity […]. Because of this strong unity with the group, the merged individuals represent the other group members as if they were their own relatives […], motivating them to perform extreme behaviors in favor of the group. (Henríquez et al., 2020)

3  THE LIMITATIONS OF EXTREME COGNITIVISM 

47

This is an extraordinary claim: IFT seems to say that we can observe the effects of personal connections (such as trust, commitment, and dedication to the other) in the absence of personal connections. Like the two selves, ‘relational ties’ are treated as a purely cognitive phenomenon, but in much of the social sciences, relationships are understood in a somewhat broader sense—as meaningful, committed connections with people, places, or events, characterised by complex and varied patterns of thoughts, feelings, interactions, social expectations, etc. Contrastingly, IFT treats these relations as something created entirely within the mind and disconnected from social reality, a matter primarily of perception. The mind of the ‘fused’ person is said to miscategorise in-group strangers as family, and hence treat them in the same way that it is predisposed to treating real family-members—as people with whom they share a deep personal connection: These feelings of connectedness, in turn, foster strong relational ties to other group members and the perception of reciprocal strength. […] In local fusion, group members form relational ties with others with whom they have direct personal contact and thus have the opportunity to share experiences. […] In extended fusion, people may project the relational ties normally associated with local fusion onto large groups despite having little or no direct contact or shared experiences with individual members. (Swann et al., 2012) Because of this strong unity with the group, the merged individuals represent the other group members as if they were their own relatives (Henríquez et al., 2020)

The theory implies that ‘family’ is a perceptional category into which our minds can sort people; the mind seems to come equipped with a set of cognitive templates for different types of relationships, which are then mapped onto specific individuals. Supposedly, miscategorising strangers as family leads us to treat them with the warmth and care that our minds reserve for close family members. Like the ‘mental entities model’ described earlier in this chapter, these relational ties can only be understood in cognitive terms, rather than as a metaphor for some type of social relationship. First, the quote from Swann et al. (2012) above makes it clear that the projection of familial ties onto large groups (like nations) occurs in the absence of deep, meaningful interactions with individual members, so the familial ties cannot be a metaphor for such deep interactions—they must be a purely cognitive phenomenon, a (mis)perception of the nation as if it were a family. Second,

48 

M. SIROMAHOV AND A. HATA

the fact that relational ties are usually described in IFT as being “projected” onto others also hints at the purely cognitive meaning of the concept: real-world relationships are typically formed or built over time through social interactions with others, but in IFT they are treated exclusively as a matter of perception. Third, these ties can apparently also be fostered through simple social priming, by exposing the fused individual to a snippet of text convincing them that they share “core characteristics” with other group-members (Swann et al., 2014)—beyond this subliminal suggestion, no actual interactions or relationships with the other group-­ members are required for the projection to occur. Finally, according to IFT, the projection of relational ties is the mechanistic cause of people’s deep attachment to a group—so, again, the former cannot be interpreted as merely a figurative description of the latter: How then, do people who are fused with large groups come to perceive that they have familial ties with all group members? […] Even though it is impossible to form actual familial ties to all other members of large groups, under the proper conditions fused persons may project familial ties onto them. This projection process may be set in motion by priming shared characteristics of the group members. Priming shared core characteristics may foster the perception of oneness within the group, which should, in turn, encourage the perception of familial ties toward other group members. […] this shift to seeing group members in familial rather than merely relational terms will embolden extreme pro-group behavior, as group members are now family members whom it is the duty of the fused person to support and protect. (Swann et al., 2014) Swann et al. showed that, for strongly fused persons, recognizing that other group members share core values makes extended groups as the country seem “family like” and worth dying for. (Besta et al., 2014)

This cognitivist interpretation of the ‘projection of familial ties’ is shared by other fusion theorists—for example, Szocik and van Eyghen (2021) write that “fusion mechanisms could be culturally evolved gadgets for generating identity fusion in all sorts of groups. In this way, cultural evolution exploits biological mechanisms”, and Kunst et al. (2018) summarise the view that fusion is the “evolved logic of altruism directed towards kin”, supported with a reference to W. D. Hamilton’s “The Genetical Evolution of Social Behaviour”—further suggesting that the term ‘projection’ here describes an innate (evolved) cognitive mechanism.

3  THE LIMITATIONS OF EXTREME COGNITIVISM 

49

Swann et al. (2012) do provide a brief justification for this evolutionary psychological approach based on research showing that members of multiple animal species are more likely to sacrifice for others who are more genetically similar to them, which could be taken as evidence for an evolved cognitive “mechanism designed to detect the genetic relatedness of other group members”. Swann et  al. then propose that the same evolved mechanism, which humans share with other social animals, could be the basis for fusion with large, impersonal groups. However, even if we accept the evolutionary psychological view that social behaviour is partially guided by subtle, inherited intuitions about who to trust and to protect, it does not follow that these intuitions can emulate the full richness and emotional depth of interpersonal relationships that take place in small cohesive groups, and replicate them in the context of large, impersonal groups like nations. Fusion theorists have not explained what it is about close familial relationships that makes people more willing to make sacrifices for their relatives than for strangers. It would be reasonable to suggest that it is something about the rich history of knowing your relatives, learning to trust them, growing to feel indebted to them over countless everyday interactions, and feeling the social expectation to protect those closest to you—rather than the simple cognitive act of labelling someone as ‘family’. Without this kind of commitment, trust, and emotional attachment, what IFT is left with is the hypothetical evolved intuition that someone who is physically and culturally similar to you is therefore probably slightly more trustworthy than others. IFT’s contention that this insipid intuition, rather than some other motivation, is behind the rare cases where a person decides to make a costly sacrifice for complete strangers seems unfounded.

Conclusion We have explored the ways in which the key explanatory concepts used in Identity Fusion Theory—the ‘personal self’ and ‘the social self’, the hydraulic relationship between them, ‘personal agency’, and ‘relational ties’—only superficially resemble the concepts bearing the same names in the Social Identity tradition and the wider social sciences. Fusion theorists consistently interpret these concepts in strictly cognitivist terms: the people described in IFT do not believe, feel, or need things; rather, there are invisible cognitive machines in their minds which, through their mechanistic relationships, produce beliefs, feelings, and needs. People do not

50 

M. SIROMAHOV AND A. HATA

pursue certain needs and goals actively and consciously; on the contrary, their behaviours are ‘fuelled’ and compelled by the motivational energy churned out by these invisible cognitive dynamos. Relationships are not built on familiarity, trust, and mutual obligations; instead, complete strangers can become brothers and sisters in a flash, provided their brains can be tricked into sorting others into a pre-existing ‘family’ category, and so on. Crucially, although IFT draws upon this ontology of the human mind, it does so only implicitly—papers on identity fusion do not spell out the rudiments of this cognitivist view of the mind or its implications, nor do they attempt to prove its adequacy as a metapsychology. Papers on identity fusion do not cite other research defending this cognitivist interpretation of human relationships or the existence of the mental entities and the ‘projection of ties’. On the contrary, fusion theorists use terms that are already common currency in social psychology (personal identity, social/ collective identity, relational ties), but give them an idiosyncratic, hard-­ cognitivist meaning that social scientists from adjacent disciplines might struggle to recognise. In this way, fusion theorists’ references to the Social Identity approach might serve to paper over a substantial drift away from the more fundamental psychological assumptions of those earlier theories: IFT dresses up its novel ideas about metapsychology in the deceptively familiar language of established theories, but the familiar concepts here carry very unusual meanings. Michael Billig, a critic of the cognitivist approach to social psychology, has drawn on the writings of Hans Vaihinger to explain how social scientists tend to reify their theoretical abstractions: … scientists often creatively use concepts, which they know to be made-up fictions. Some of the most fruitful concepts in the history of science have involved fictional, even impossible, ideas […] One might say that by turning processes rhetorically into things, scientists routinely are creating ‘fictional things’—namely entities, which cannot actually exist as things, but which are treated as if they were things. Vaihinger argued that problems arise when scientists stop realizing that their fictions are fictions, but treat them as literal truths or actual things […] There are problems when we confuse the imagined world of our fictional things with the world that we are using these fictions to understand. The problems are likely to be acute if we take our fictional things to be more real than the things that ordinary people recognize in the social world. (Billig, 2011)

3  THE LIMITATIONS OF EXTREME COGNITIVISM 

51

This is precisely the kind of reification that appears to have occurred with the translation of SIT’s personal and social identity (as levels of self-­ definition or modes of being) into IFT’s personal and social selves (as mental entities). People are largely absent from publications on identity fusion, which deal predominantly with the operations of the theory’s invisible mental entities rather than with real people’s actions and motivations. The contrast with the earlier Social Identity tradition could not be clearer—in the classic works, the theoretical abstractions could always be illustrated with reference to concrete actions taken by real people. Turner et al. (1992) emphasised that SCT’s abstract theoretical model was only the first step in the analysis of any social phenomenon—it would then have to be applied to people’s concrete circumstances, social structures, historical context, ideologies, etc. Turner went so far as to describe this ascent from the abstract model to the concreteness and fullness of real life as one of the three pillars of the Social Identity tradition as envisaged by Henri Tajfel. Sadly, this does not happen in the Identity Fusion literature—its authors routinely take their theoretical concepts at face value, rather than providing detailed, full-bodied examples of the phenomena they strive to explain. Because they view their theoretical concepts as cognitive entities that have agency over human behaviour, they can treat them as the ‘real’ origin of the social phenomena that they want to study—therefore, once these entities and their cognitive mechanisms have been discovered, the analysis tends to stop there. As a result, IFT’s concepts get reified—the existence of the two selves and the physical properties of their ‘boundaries’ are taken not as visual metaphors of how people perceive themselves and relate to others but as really existing things in the mind that mechanically cause social perceptions and attachments to occur. The flipside of this reification of abstractions into mental entities is that, if one was to challenge the existence or nature of the ‘two selves’ as we have attempted to do here, that would deprive the theory of its foundational explanatory mechanism. Fusion theorists are correct in pointing out that, whereas some people merely self-­ categorise as members of a group, others feel a deep emotional attachment and dedication which is much more intense and consequential than mere self-categorisation. But if we reject the notion that this feeling is driven by a ‘fusion’ between a personal and a social self, that extremism is fuelled by some psychic energy emanating from these selves, or that people come equipped with a cognitive mechanism that sorts the world into “family” and “not-family” and confers on anyone in the “family” group automatic

52 

M. SIROMAHOV AND A. HATA

feelings of kinship and commitment, then IFT is left unable to explain where this difference between the ‘merely self-categorising’ and the ‘deeply committed’ members comes from. Because Identity Fusion Theory prefers to dwell in the cognitive world of fused selves, porous boundaries, and mental projections, the reader never gets to visit the social world of real people and their groups—we are never given a sense of what a “simultaneous activation of the personal and social self-concepts” looks like in practice, or what it means for “aspects of both the personal and social self [to] flow into the other”. This makes it impossible to generate new metaphors to describe the same phenomena. All the reader is left with is the original highly abstract formulation, which is probably why different publications on identity fusion over the years tend to repeat the same cryptic phrases about “fused self-concepts” and “porous boundary between the personal and social selves” without further elaboration. Meanwhile, in the rare cases that researchers have attempted to map IFT’s theoretical constructs to real social interactions, they have  often done so by distorting or ignoring the theory’s main principles, carelessly handling its concepts and research methods, or simplifying it beyond recognition. We will explore how later papers have distorted and ‘flattened’ the theory in the next two chapters.

References Atran, S. (2017). The role of the devoted actor in war, revolution, and terrorism. In J.  R. Lewis (Ed.), The Cambridge companion to religion and terrorism. Cambridge University Press. Besta, T., Gómez, Á., & Vázquez, A. (2014). Readiness to deny group’s wrongdoing and willingness to fight for its members: The role of poles’ identity fusion with the country and religious group. Current Issues in Personality Psychology, 2(1), 49–55. Billig, M. (2011). Writing social psychology: Fictional things and unpopulated texts. British Journal of Social Psychology, 50(1), 4–20. Corner, E., Taylor, H., & Clemmow, C. (2021). Assessing the behavioural trajectories of terrorists: The role of psychological resilience. Dynamics of Asymmetric Conflict, 15(2), 96–122. Ferguson, N., & McAuley, J. W. (2020). Staying engaged in terrorism: Narrative accounts of sustaining participation in violent extremism. Frontiers in Psychology, 11, 1338.

3  THE LIMITATIONS OF EXTREME COGNITIVISM 

53

Gómez, A., Belanger, J. J., Chinchilla, J., Vázquez, A., Schumpe, B., Nisa, C., & Chiclana, S. (2021). Admiration for Islamist groups encourages self-sacrifice through identity fusion. Humanities & Social Sciences Communications, 8(1), 1–12. Gómez, Á., Brooks, M. L., Buhrmester, M. D., Vázquez, A., Jetten, J., & Swann, W. B., Jr. (2011). On the nature of identity fusion: Insights into the construct and a new measure. Journal of Personality and Social Psychology, 100(5), 918–933. Hart, R., & Lancaster, S.  L. (2019). Identity fusion in US military members. Armed Forces & Society, 45(1), 45–58. Henríquez, D., Urzúa, A., & López-López, W. (2020). Identity fusion: A systematic review. Acta Colombiana de Psicología, 23(2), 410–437. Hopkins, N., Reicher, S., Harrison, K., Cassidy, C., Bull, R., & Levine, M. (2007). Helping to improve the group stereotype: On the strategic dimension of pro-­ social behavior. Personality and Social Psychology Bulletin, 33, 776–788. James, W. (1890 [2001]). The principles of psychology (Vol. 1). Dover Publications. Knijnik, J., & Newson, M. (2021). ‘Tribalism’, identity fusion and football fandom in Australia: The case of Western Sydney. Soccer & Society, 22(3), 248–265. Kunst, J. R., Boos, B., Kimel, S. Y., Obaidi, M., Shani, M., & Thomsen, L. (2018). Engaging in extreme activism in support of others’ political struggles: The role of politically motivated fusion with out-groups. PLoS One, 13(1), e0190639. Levine, M., & Crowther, S. (2008). The responsive bystander: How social group membership and group size can encourage as well as inhibit bystander intervention. Journal of Personality and Social Psychology, 95, 1429–1439. Newson, M. (2019). Football, fan violence, and identity fusion. International Review for the Sociology of Sport, 54(4), 431–444. Newson, M. (2021). High and highly bonded: Fused football fans who use cocaine are most likely to be aggressive toward rivals. International Journal of Drug Policy, 93, 103263. Newson, M., Bortolini, T., Buhrmester, M., da Silva, S. R., da Aquino, J. N. Q., & Whitehouse, H. (2018). Brazil’s football warriors: Social bonding and inter-­ group violence. Evolution and Human Behavior, 39(6), 675–683. Newson, M., Buhrmester, M., & Whitehouse, H. (2021). United in defeat: Shared suffering and group bonding among football fans. Managing Sport and Leisure, 28(2), 164–181. Reicher, S. D., Spears, R., & Postmes, T. (1995). A social identity model of deindividuation phenomena. European Review of Social Psychology, 6(1), 161–198. Swann, W. B., Jr., & Buhrmester, M. D. (2015). Identity fusion. Current Directions in Psychological Science, 24(1), 52–57. Swann, W. B., Jr., Gómez, Á., Buhrmester, M. D., López-Rodríguez, L., Jiménez, J., & Vázquez, A. (2014). Contemplating the ultimate sacrifice: Identity fusion channels pro-group affect, cognition, and moral decision making. Journal of Personality and Social Psychology, 106(5), 713–727.

54 

M. SIROMAHOV AND A. HATA

Swann, W. B., Jr., Jetten, J., Gómez, Á., Whitehouse, H., & Bastian, B. (2012). When group membership gets personal: A theory of identity fusion. Psychological Review, 119(3), 441–456. Swann, W. B., Jr., Gómez, Á., Seyle, D. C., Morales, J. F., & Huici, C. (2009). Identity fusion: The interplay of personal and social identities in extreme group behavior. Journal of Personality and Social Psychology, 96(5), 995–1011. Szocik, K., & Van Eyghen, H. (2021). Religion as Adaptive 2: Adaptationist approaches. In Revising cognitive and evolutionary science of religion (New approaches to the scientific study of religion) (Vol. 8). Springer. Tajfel, H., & Turner, J. (1979). An integrative theory of intergroup conflict. In S. Worchel & W. G. Austin (Eds.), The social psychology of intergroup relations (pp. 33–47). Brooks-Cole Publishing. Turner, J. C., & Oakes, P. J. (1997). The socially structured mind. In T. Postmes & N. R. Branscombe (Eds.), Rediscovering social identity [2016]. Routledge. Turner, J. C., & Onorato, R. S. (1999). Social identity, personality, and the self-­ concept: A self-categorization perspective. In T. Postmes & N. R. Branscombe (Eds.), Rediscovering social identity [2016]. Routledge. Turner, J.  C. (1982). Towards a cognitive redefinition of the social group. In T. Postmes & N. R. Branscombe (Eds.), Rediscovering social identity [2016]. Routledge. Turner, J.  C. (1984). Social identification and psychological group formation (Chapter 25). In H.  Tajfel (Ed.), The social dimension (Vol. 2). Cambridge University Press. Turner, J. C. (1985). Social categorization and the self-concept: A social cognitive theory of group behaviour. In T.  Postmes & N.  R. Branscombe (Eds.), Rediscovering social identity [2016]. Routledge. Turner, J. C., Oakes, P. J., Haslam, S. A., & McGarthy, C. (1992). Self and collective: Cognition and social context. In T. Postmes & N. R. Branscombe (Eds.), Rediscovering social identity [2016]. Routledge. Whitehouse, H., McQuinn, B., Buhrmester, M., & Swann, W. B. (2014). Brothers in arms: Libyan revolutionaries bond like family. Proceedings of the National Academy of Sciences, 111(50), 17783–17785.

CHAPTER 4

Methodological Shortcomings of Identity Fusion Research

Abstract  The key measures that Identity Fusion research relies on do not reflect identity fusion as defined in the early theoretical papers and appear to have been originally developed with the aim of capturing related but distinct constructs. The pictorial measure consists of abstract diagrams, the meaning of which is opaque to the research participants and requires overly literal interpretation on the part of the researchers. Meanwhile, the verbal measure uses metaphorical language which is then interpreted literally, ignoring the poetic nature of expressions of social solidarity and group belonging. The construct is also treated inconsistently in the literature—as continuous, multicategorical, or binary—often seemingly without justification. This produces a general state of confusion in the Identity Fusion literature as to who counts as ‘fused’ or how prevalent ‘identity fusion’ is in a given population, and presents a challenge to integrating findings from multiple research studies. Keywords  Identity fusion • Social identity • Personal identity • Quantitative methods • Experimental psychology • Discursive psychology In addition to the conceptual issues outlined in Chaps. 2 and 3, the literature on identity fusion suffers from some methodological shortcomings which can throw its findings into question. Most important among these © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 M. Siromahov, A. Hata, Rethinking Identity Fusion, https://doi.org/10.1007/978-3-031-46983-1_4

55

56 

M. SIROMAHOV AND A. HATA

are the choice of measures of the theory’s central constructs, the overreliance on experimental and quantitative methods, and the inconsistent demarcation between fusion and non-fusion in quantitative research.

The Problem of Measurement The first major methodological obstacle in the identity fusion literature concerns measuring its central variable—the fusion of the personal and social identities or selves. In experimental and correlational research, which form the majority of the studies on identity fusion, it is typically measured using one of three self-report measures: the pictorial, the dynamic, and the verbal. The pictorial measure of identity fusion was first introduced in Swann et al.’s (2009) seminal paper on IFT and continues to be the most widely used one in the field of IFT research. It consists of a circle labelled “Me” and another circle labelled with the name of the group, placed at varying stages of overlap (Fig.  4.1). The full overlap between the two circles (option E) was taken by its authors to indicate a state of identity fusion. A digital version of this measure—the dynamic identity fusion index, or DIFI (Jiménez et  al., 2016)—was later developed, which uses a similar format with the same two overlapping circles, but rather than offering a fixed set of responses, it allows the respondent to click and drag the “Me” circle to the position that best reflects their relationship to the group. It thus generates a continuous, not categorical, score, expressed as a number from 1 to 125, making it easier not just to sort people into “fused” and “non-fused”, but to quantify their degree of identity fusion.

Self

Group

A

Self

Group

B

Self

Group

C

Self

Group

D

Self

Group

E

Note: Adapted from Swann et al. (2009). Participants choose one of the ive diagrams to represent how close they feel to their group. Those who select option E (full overlap between oneself and the group) are considered to be “fused”.

Fig. 4.1  The pictorial measure of identity fusion

4  METHODOLOGICAL SHORTCOMINGS OF IDENTITY FUSION RESEARCH 

57

The problem is that neither of these pictorial measures seem to have been designed to capture the kind of merger between the personal and social self-concepts that fusion is said to describe. The pictorial scale is a visual representation of the overlap between the person and the group and does not attempt to probe the respondent’s identities in line with the markers of fusion we have seen across the IFT literature: e.g., testing whether ‘aspects’ of the group identity have become integral to the ‘personal’; whether both identities are activated simultaneously and synergistically; whether the participant cares about distinct individuals, rather than only about the group in the abstract; and whether their behaviour goes against the group’s prototype. Moreover, the measure is almost identical to Aron et  al.’s (1991) measure of the inclusion of others in the self, a concept which is closer to what SIT terms identification (i.e., the creation of a shared ‘we’-category) than what Swann et al. defined as the fusion of the two identities. In effect, the pictorial measure of fusion is meant to capture a construct that reflects the simultaneous activation of personal and social selves, yet it is identical to another measure that does no such thing. As we will see in the next chapter, some researchers have incorrectly interpreted fusion as a construct analogous to Aron et al.’s (e.g., Rousseau et al., 2021) or have rewritten the history of IFT, connecting it with older theories of cognition but erasing its roots in SIT (e.g., Hatvany et  al., 2018)—a confusion which is perhaps partially rooted in the way fusion was operationalised from the start. The pictorial measure itself does not make it clear to the participant what the different categories of overlap between the circles indicate, so it invites a double misunderstanding: the research participant can misinterpret what the researcher is asking of them (e.g., the proximity of the circles could be taken to represent attachment, attraction, or belonging), and the researchers are free to project whatever meaning they choose onto the participant’s answer (e.g., they take the overlapping circles ‘Me’ and ‘My group’ to mean that the participant’s individual and collective identities are merged, indicating fusion, rather than the more common-sense interpretation that the participant feels part of the group, indicating mere social identification). The pictorial measure thus becomes something of a Rorschach test for both parties involved in the research. A self-report verbal measure, developed by Gómez et al. (2011), is also widely used in the literature. The questionnaire consists of seven statements which the participant can agree or disagree with on a 7-point scale,

58 

M. SIROMAHOV AND A. HATA

and agreement scores for all statements are averaged to produce a mean fusion score ranging from 1 to 7. Verbal measure of fusion 1. I am one with my country. 2. I feel immersed in my country. 3. I have a deep emotional bond with my country. 4. My country is me. 5. I’ll do for my country more than any of the other group members would do. 6. I am strong because of my country. 7. I make my country strong.

The verbal measure has the advantage of consisting of concrete statements that participants can agree or disagree with, which makes it somewhat less opaque than the pictorial, in that participants are aware of what it is that they are expressing agreement with.1 However, it still has a potential weakness in that it treats participants’ self-reports as literal and objective self-disclosures. The researchers take participants’ words as an accurate reflection of the contents of their mind or their mental state, which requires an overly literal interpretation of the questionnaire items. For example, abstract pronouncements like ‘Spain is me’ or ‘Russians are my brothers’ are taken as evidence that the participant is literally treating their personal identity and their group-membership as interchangeable, or that they misperceive their co-nationals as family.

The Discursive Alternative As Rom Harré (2001) has argued, questionnaires are a highly controlled and constrained form of interviewing, one in which the interviewee has not been given the freedom to choose their answer outside of a limited range of options, elaborate on the meaning of their answer, or push back 1  This does not apply to all items, however. The questionnaire includes items about ‘my country’ but does not clarify whether this is referring to the nation as a collection of individuals (e.g., ‘Russians’), the nation as an abstract entity (‘Mother Russia’) or the country as a geographical space. Presumably, it is only the first definition that should be relevant to IFT, as people are said to ‘project relational ties’ onto other individuals, not to the nation as a monolithic abstraction, but the questionnaire items refer to the country rather than one’s countrymen. This again leaves a gap between fusion as explicitly theorised and as measured with the verbal scale.

4  METHODOLOGICAL SHORTCOMINGS OF IDENTITY FUSION RESEARCH 

59

against the researcher’s formulation of the question. Crucially, this approach does not lead us to some non-discursive, non-subjective form of communication that conveys the contents of the participant’s mind objectively to the researcher, but merely pushes the discursive character and subjective meaning of the exchange out of sight so that they do not interfere with the researcher’s quantification of identity fusion. How did participants interpret the two overlapping circles in the pictorial measure of fusion, or items such as ‘Spain is me’ in the verbal measure? What did they intend to say by responding with ‘Strongly agree’ to that item in the questionnaire? What did they understand as ‘sacrificing for Spain’? All of these questions remain unasked and unanswered. In order to produce its quantifiable constructs, identity fusion research is designed to take proclamations of identity at face value and treat them literally, ignoring the communicative functions they serve. This approach clearly ignores the poetic and metaphorical functions of language. Statements of identity should not be taken literally before considering their rhetorical function within their discursive context. “I am…” statements are often used figuratively to express political or social positions—we need only to think of the phrase “Je suis Charlie” (“I am Charlie”) which became a rallying cry for freedom of speech around the globe after the 2015 terrorist attack against the French satirical magazine Charlie Hebdo. Some anti-racist activists have also explained their participation in the global movement against racist police violence sparked by the murder of George Floyd in 2020 through the rhetoric of identification. Speaking about Floyd’s killing, British dancer and choreographer Ashley Banjo drew a parallel with his own family’s experience of racist policing: “I looked at George Floyd and I saw my dad” (BBC News, 2020). It would be naive to take these poetic expressions of solidarity as literal self-disclosures—as evidence that a person who agrees with these two statements somehow really believes that they are Charlie and misperceive George Floyd as their father. However, because the measures used in IFT are blind to the rhetorical functions of language, nothing would prevent a hypothetical researcher from using the exact same methods employed by fusion theorists to ‘demonstrate’ that a certain percentage of the population have become ‘fused’ with Charlie Hebdo, or that they are ‘projecting familial ties’ onto George Floyd. Doing that would require only a penchant for overly literal reading of Likert scale responses and a tendency to miss the statements’ discursive function as an expression of empathy and solidarity.

60 

M. SIROMAHOV AND A. HATA

This approach is particularly problematic when interpreting people’s expressions of national identity, which are often highly formulaic and follow established social conventions of describing the nation as a parent or other co-nationals as brothers and sisters. Scholars of nationalism, for example, have recognised that nation-states and nation-building movements routinely use the language of micro-level (familial) solidarity to naturalise their own political projects: In order to disseminate nationalist narratives, social organisations have to tap into the grassroots: the well-established networks of close kinships, friendships, neighbourhoods, peers, lovers and other face-to-face groupings. Since human beings are first and foremost emotional and moral creatures who attain solace, security and fulfilment in very small groups, the social organisations have to emulate the language and practises of the small group bonds. Hence, the successful nationalist projects are premised on the organisational translation of the ideological grand narratives into the micro, family and friendship-based, stories. (p.14) … Many representatives of nation states and nationalist movements invoke these kinship and friendship-based images deliberately aiming to link their anonymous and instrumentally driven organisations with the inherent warmth and intimacy of face-to-face micro-group interactions. (Malešević, 2019, p.66)

In other words, the parallel between interpersonal relationships and solidarity with large-scale, abstract groups, lies in the similar rhetorical resources that they employ, not in their identical nature as psychological phenomena. Fusion theorists notice the similarity in the rhetoric, but because they assume that speech merely reports internal mental states, they posit the existence of a single mental state (identity fusion) responsible for both instances of that rhetoric. However, because the language of family and kinship and the profession of one’s willingness to die for the motherland are one of the dominant ways in which patriotism is expressed in national societies, it is much more parsimonious to assume that research participants will interpret most of the items on the verbal scale (especially the more abstract ones, such as “I am Spain” and “Spain is me”) as generic expressions of patriotism. In practice, the measure is designed in such a way that it detects willingness to perform the rhetorical scripts of nationalist discourses—perhaps itself driven by a commitment to nationalism—rather than the hypothetical state of identity fusion.

4  METHODOLOGICAL SHORTCOMINGS OF IDENTITY FUSION RESEARCH 

61

It would be difficult to reconcile a discursively and sociologically informed approach to identity with IFT’s preference for quantitative and experimental methods, as well as with the theory’s cognitivist foundation. The experimenters want to demonstrate the existence of certain statistically significant differences between ‘fused’ and ‘non-fused’ individuals, which requires them to treat identities as different quantities of the same cognitive phenomenon (e.g., stronger or weaker identification, higher or lower degree of fusion). This is in contrast to the way identity is understood in much of the social psychological literature. For example, even Social Identity Theory recognises that identities are sociologically dense ‘packages’ of ideas, behavioural norms, social practices, attitudes, etc. The simple act of proclaiming oneself to be, e.g., British, presupposes the existence of such a thing as a British nation and carries the assumption that such a thing as ‘nation’ exists as a legitimate and meaningful form of identity. These assumptions should not be taken for granted—Billig (1995) has written on the consistent ideological and institutional labour that must be performed every day in order to reproduce the idea of nationhood and of individual nations as common sense. Moreover, different national identities come with particular scripts for performing that identity, so they cannot be treated as mutually interchangeable. In some societies, passionate proclamations of national pride and willingness to die for the motherland are considered normal and even socially desirable; children as young as four or five can be taught to recite patriotic poems about their love of their country. Meanwhile, such expressions of fervent patriotism are often considered gauche and distinctly un-English in the UK, prompting Susan Condor (1996; Condor & Abell, 2005) to argue that a critical disavowal of national identity is one of the more common ways in which English identity is performed today. In the English context, a critical distancing from nationalist rhetoric thus serves to mark the supposedly cool-headed, rational, and enlightened in-group from the ‘parochial’ foreigners who so easily succumb to nationalist passion. Finally, proclamations of identity can serve different purposes depending on the social context—describing oneself as a ‘Proud American’ could contain an implicit repudiation of one’s political opponents by framing them as un-American or unpatriotic; in this way, respondents could use their seemingly straightforward responses to a questionnaire to paint themselves as virtuous or to position themselves against some imaginary Other. Because declarations of identity can serve a complex range of functions, sociologist Siniša Malešević (2006)

62 

M. SIROMAHOV AND A. HATA

has gone so far as to call for social researchers to dispense with the concept of ‘identity’ altogether and instead rehabilitate that of ‘ideology’. Because IFT sees identity as a mental object, rather than something that is produced socially and is dependent on socially shared ideologies, it is blind to the importance of the contents of group identity. The main thing the IFT researcher is interested in is the fact that the two selves become ‘fused’, regardless of the normative contents of those selves, their subjective meaning for the individual, or how they are produced socially. Turner (Turner et al., 1992; Turner & Onorato, 1999) advocated against such an overly mechanistic understanding of social identity, arguing that the original Social Identity Theory was meant to integrate general principles with the concrete contents of specific identities. According to Reicher and Hopkins (2001, p.34–36) forgetting that makes the theory an encumbrance. Fusion theory is also not unique in its treatment of social identity— research in contemporary social psychology often relies on self-report questionnaires to quantify the strength of ‘identification’ with particular groups, often neglecting the qualitative contents of those identities. In this way, IFT is perhaps a symptom of a larger issue than an outlier in the social sciences. However, these methodological choices are surprising given that the theory’s earlier publications explicitly pointed to the contents of group identity (i.e., the prototypical traits that non-fused members ostensibly adhere to and fused members deviate from) as a main determinant of pro-group behaviour—so the abandonment of this principle is particularly problematic for IFT.

Demarcation: Is Fusion Categorical or Continuous? One major inconsistency that has become entrenched in the IFT literature is whether to treat fusion as a continuous construct or a categorical one, and if the latter, where to draw the boundary between fused and non-­ fused individuals. This question goes to the heart of what identity fusion is, but surprisingly, how researchers answer it is dictated not so much by theoretical considerations but by the measurement tools they choose. When they first introduced the pictorial measure of fusion (Fig. 4.1), Swann et al. (2009) considered only those who chose option E (full overlap between Self and Group) to be ‘fused’—everyone else, including those choosing the partial overlap, were ‘non-fused’. They consistently treated fusion as a binary variable (fused vs. non-fused) in their other

4  METHODOLOGICAL SHORTCOMINGS OF IDENTITY FUSION RESEARCH 

63

comparisons: “the results of the preliminary studies supported our assumption that fusion is a unique state of oneness with a group, a state that is categorically distinct from the state of nonfusion”. This categorical distinction between ‘fused’ people and everyone else is reflected in Swann et  al.’s reported fusion scores, which were distributed bimodally—most of their participants chose either the middle option (50% overlap between Self and Group) or the most extreme one (100% overlap), as shown in the top panel of Fig. 4.2. This was the empirical evidence, in addition to the theoretical justifications presented in their paper, that led the authors to conclude that fusion must be not just a categorical construct, but a binary one—people are either fused or they are not. This approach is also taken in some more recent papers, e.g., Bello et al. (2021). However, Swann et  al.’s (2009) empirical findings provide less than conclusive evidence that fusion and non-fusion are categorically distinct psychological states. The theory was premised on the idea that people with a strong group identity (those who are ‘merely identifying’ but not fused) will conform to the group’s prototype, while fused people will show more individuality because of the active personal self and will deviate from the prototype. However, Swann et al.’s results showed that neither fusion nor identification were correlated with perceived prototypicality. According to the authors, this raises the possibility that the measure of prototypicality was invalid, but it could also mean that the prototypicality hypothesis does not describe reality. This is strange, as the presence of an active personal self leading to counter-prototypical behaviour is the main way in which fusion and identification are meant to differ. As we will see in Chap. 5, this pillar of IFT has been largely abandoned today. The authors also claimed that fused grep-members would be willing to fight and die for the group, whereas non-fused members would not. The study did show that fused participants score marginally higher on a willingness-­to-die questionnaire than non-fused ones, but the difference was not practically meaningful (about 0.4 points out of a 7-point scale). The willingness-to-die scale consists of a set of statements that participants can agree or disagree with; scores from −3 to −1 indicate disagreement, 0—neutrality, and 1 to 3—agreement. Both ‘fused’ and ‘non-fused’ participants scored on average within the negative range, meaning that neither group expressed a clear willingness to fight or die for Spain. ‘Fused’ participants were only slightly less reluctant to fight (M = −1.96, where 2 presumably indicates “Strongly disagree”) than ‘non-fused’ ones (M=  −2.38). Interestingly, both groups were more willing (or, more

64 

M. SIROMAHOV AND A. HATA

Note: Top: Distribution of fusion scores obtained with the original pictorial measure by Swann et al.’s (2009). Notice the bimodal distribution, which led the authors to conclude that fusion is a categorical construct with a clear distinction between those who are ‘fused’ (reporting the highest level of fusion, Category E) and everyone else who is ‘nonfused’. Figure recreated based on descriptive statistics reported in Swann et al. (2009). Bottom: Distributions of fusion scores from an American and an Indian sample obtained with Gómez et al.’s (2011) self-report questionnaire. The data is taken from Siromahov et al. (2020). Notice the overall higher levels of fusion in the Indian sample, as well as the lack of a bimodal distribution that would have indicated a categorical difference between “fused” participants and everyone else.

Fig. 4.2  Distributions of fusion scores obtained with the pictorial and verbal measures

4  METHODOLOGICAL SHORTCOMINGS OF IDENTITY FUSION RESEARCH 

65

accurately, slightly less unwilling) to die than to fight for Spain. Fused people’s average willingness to fight had a score of −1.96, but it climbed up to −0.67 (an increase of 1.29 points) when it came to dying for Spain; likewise, non-fused people were more reluctant to fight (−2.38) than to die (−1.26, or an increase of 1.12). This is surprising on its face, given that dying is much more of an extreme and unambiguous sacrifice than merely fighting. Altogether, the data did not seem to show the kind of binary division into fusion and non-fusion that the theory predicts with its claim that ‘fused’ people will endorse extreme self-sacrifice while others would not. In this way, fusion theorists were willing to ignore recalcitrant data while IFT was still in its infancy, until the same hypothesis was supported in later research using different methods (e.g., Swann et al., 2010). Other studies have used the same pictorial measure while treating its output as continuous. For example, Zumeta et al. (2020): “The five response options range from A to E, where A symbolizes a lesser perception of closeness or fusion (i.e., circles without overlapping) and E is a greater closeness or fusion (i.e., completely overlapping circles)”. In this study, participants who chose any response representing less than perfect fusion would still be categorised as partially fused, whereas in Swann et al.’s original paper they would be categorised as non-fused. The dynamic identity fusion index (Jiménez et al., 2016) uses a similar format with two overlapping circles but generates a continuous, not categorical, score. This makes it easier to plug the resulting score into correlation analyses or regression models which, rather than comparing ‘fused’ and ‘non-fused’ individuals as two distinct groups on some outcome measure, instead track the relationship between the degree of fusion and some other continuous variable. The verbal measure (Gómez et al., 2011) also produces a continuous score—it consists of seven statements that the respondent can agree or disagree with on a 7-point scale, out of which an average score is calculated. By design the questionnaire is prone to producing more normally distributed scores than the original pictorial measure, because in order to produce the maximum score (the equivalent to the pictorial measure’s ‘complete overlap’ category), people who take the verbal measure would have to select ‘Totally agree’ on each one of the seven items; since most people tend to qualify their responses on at least some of the items, most of them will fall somewhere below the maximum score. That is why, when Swann et al. (2014) used the verbal measure, their fusion scores were normally distributed with a mean of 3.86 out of 7, suggesting that very few

66 

M. SIROMAHOV AND A. HATA

participants got the maximum possible score (which, let us not forget, was the threshold for fusion advocated by Swann et al., 2009). The difference between the data produced by the pictorial and verbal measures is illustrated in Fig. 4.2, which compares the distribution of Swann et al.’s (2009) fusion scores with those obtained from a sample of American and Indian participants using the verbal measure (using data from Siromahov et al., 2020)—the latter does not show the clear bimodal distribution that we get from using the pictorial measure, which was the original justification for treating ‘fused’ and ‘non-fused’ people as categorically distinct. Hence, most studies that use the verbal measure are likely to treat fusion as a continuous construct, in contrast to how it was originally formulated by IFT’s authors. However, this is not the only discrepancy in how fusion is measured. Some researchers use the verbal (continuous) measure while still parsing the spectrum of scores into discrete categories, such as Mason et al. (2021), who studied identity fusion among supporters of Brexit: “Fusion ‘categories’ were determined through mean item scores of the identity fusion verbal measure: Not fused