政治暴力研究中概念清晰性的侵蚀

Introduction

On May 6, 2023, Mauricio Garcia opened fire at an outlet mall in Allen, Texas, murdering eight people before he was killed by a responding police officer. Garcia, a Hispanic man in his early thirties, wore a patch bearing the initials RWDS (Right Wing Death Squad), and his extensive online footprint combined neo-Nazi iconography, incel grievances, and an assortment of conspiratorial narratives. Analysts described the case as “fuzzy” and “bizarre,” and within days it was absorbed into a now familiar interpretive template, first articulated by FBI Director Christopher Wray in his September 2020 Senate testimony, in which contemporary attackers are understood to assemble a “salad bar of ideologies,” a little of this and a little of that, while “what they are really about is the violence” (Threats to the Homeland, 2020).

But why has a field that spent five decades constructing increasingly refined models of ideological mobilization become so willing to conclude that ideology itself is receding? How exactly did we arrive at a conceptual landscape in which any perpetrator whose belief system does not mirror an organizational catechism is classified as “mixed, unclear, or unstable”? And what analytic work, if any, do labels such as “salad bar” or “nihilistic violent extremism” actually perform? Despite the growing interest of academics and practitioners in these ostensibly ambiguous cases, we are still unable to answer such questions authoritatively. Horgan and Shayler’s (2026) persuasive critique of the salad bar construct provides an ideal point of departure for doing so. Their article demonstrates that the term is imprecise, that it conflates several distinct processes, and that it overstates novelty. I would argue, however, that their diagnosis, while accurate, does not go far enough. The terminological proliferation they document is a symptom of a deeper erosion of conceptual clarity, one rooted in the field’s declining willingness to engage in deep and systematic study of violent ideologies themselves.

Part of a Special IssuePsychology of Violent ExtremismThis article belongs to a curated special issue. Discover more research on this topic from the full collection.Explore the special issue →

There is much in Horgan and Shayler’s account to endorse. Their distinction between mixing and migrating, and between convergent and divergent ideological combinations, restores a measure of order to a literature that has generated at least nineteen competing terms for phenomena that remain poorly specified. Their insistence that ideology should not be treated as a synonym for motivation is similarly well taken, and their observation that rising counts of “mixed, unclear, or unstable” cases may reflect how practitioners classify under conditions of limited training and heavy workload, rather than a change in the underlying population, deserves to reorient the empirical agenda. Their reminder that cross-ideological cooperation, side-switching, and idiosyncratic belief systems have deep historical roots is a needed corrective to a policy discourse enamored with novelty. Yet the article stops short of confronting the disciplinary conditions that allowed the salad bar metaphor to flourish in the first place. Four interrelated concerns follow.

The first relates to ideological literacy. The designation of a case as “mixed” or “unclear” is a measurement of the classifier as much as of the classified. An analyst who has never worked through the foundational texts of the American racist right, who cannot trace the genealogy connecting Christian Identity theology, the militia movement’s constitutional mythology, and the accelerationist milieus of the past decade, will perceive incoherence where a more immersed observer perceives a recognizable, if evolving, tradition. Horgan and Shayler (2026) gesture toward this possibility when they contrast the United Kingdom, where a majority of Channel referrals were designated mixed or unclear, with a Toronto intervention program in which fewer than five percent of clients received the equivalent label, and when they call for practitioner-engaged research into how such classifications are produced. I would push their point further. This variation is not merely administrative. It reflects a field in which sustained engagement with ideological content has become unfashionable, displaced by process models of radicalization that treat beliefs as interchangeable inputs and by a security-driven preference for behavioral indicators. Few doctoral programs in our field train students to read extremist intellectual production the way historians of political thought read canonical texts, and the consequences of that gap surface every time a perpetrator’s belief system is declared unclassifiable within days of an attack. Thus, it is not surprising that Garcia’s fusion of white supremacist and misogynist content struck practitioners as bizarre. Misogyny has been integral to far-right thought for a century, binding racial anxieties to the control of women’s bodies and of reproduction, and the incel vocabulary is in many respects a new dialect for that older grammar. Ideologies appear mixed when analysts decline to take white supremacy seriously as an organizing structure (Meier, 2023, as cited in Horgan & Shayler, 2026). Our reluctance to engage deeply with idea systems we find repugnant is understandable, but it produces a trained incapacity to recognize ideology when it stands in front of us, and it encourages us to set aside clear evidence that contemporary perpetrators remain motivated by specific ideological visions.

The second concern relates to motivational complexity. Implicit in the salad bar framing, and occasionally in critiques of it, is the assumption that earlier generations of militants were ideologically coherent in ways that contemporary attackers are not. The historical record does not support this baseline. Blee’s (1991) study of the 1920s Klan documented women who joined for community, status, and family ties as much as for doctrine, while the organization nonetheless pursued a coherent nativist and white supremacist program. Laqueur (1999) observed decades ago that terrorists are generally “not the people most deeply convinced of the righteousness of their cause” (p. 274). Interviews with foreign fighters, militia members, and former neo-Nazis have repeatedly surfaced the same braid of thrill-seeking, belonging, personal grievance, and conviction. Organizations from the Weather Underground to the Provisional IRA accommodated members whose command of doctrine was thin, without anyone concluding that those campaigns lacked ideological direction. Perpetrators of political violence were always motivated by a complex set of factors and ideas, and that complexity coexisted, in every era, with organized action in the service of specific ideological programs.

The analytic error, and here I part company with some readers of Horgan and Shayler’s argument rather than with the authors themselves, is the inference from motivational heterogeneity at the individual level to ideological absence at the level of the phenomenon. Ideology rarely operates as a private psychological engine. It operates as a collective framework that defines enemies, selects targets, legitimizes tactics, and supplies the narrative within which personal grievances acquire political meaning (Berger, 2018). A young man may arrive at violence through humiliation and isolation and still act, deliberately, in the service of a specific ideological vision. The target tells us so. Attackers who select a synagogue, a Black church, or a Walmart in a predominantly Hispanic border city are not distributing their violence randomly across a salad bar. The same logic applies over time. The fact that a perpetrator engaged with other ideas in the past does not mean he is not operating to promote specific ideas in the present. Migration between frameworks is evidence that people search, not evidence that no framework governs their conduct. Horgan and Shayler recognize this when they insist that ideology and motivation be separated analytically. Yet their suggestion that prevention might focus “less on the daunting task of understanding ideological commitment, which may be less important” (Horgan & Shayler, 2026) risks licensing precisely the dismissal their article otherwise resists.

Third, ideologies change, but in most cases they maintain their core principles, and much of what the field currently perceives as new ideological formations is better understood as new customs of old ones. The Great Replacement is routinely presented as a novel conspiracy theory. It is a re-customization of anxieties about “race suicide” that Madison Grant (1916) popularized, later filtered through the “Zionist Occupation Government” narratives of the 1980s. The conspiratorial architecture beneath it, a hidden elite orchestrating demographic destruction, descends in a direct line from the Protocols of the Elders of Zion, which is why attackers in Pittsburgh, Christchurch, and El Paso could deploy the same script against different proximate targets. Militant accelerationism is presented as an innovation of the Telegram era, yet its core scenario, the deliberate collapse of a corrupt system through exemplary violence, was elaborated in The Turner Diaries and in Mason’s Siege decades before the platforms existed (Macdonald, 1978). Incel ideology is treated as a freestanding category, though it grafts readily onto a much older hierarchy of gender that the far right has always policed. What presents as an ideological mixture is, in many cases, variation within a tradition whose stable core (anti-egalitarianism, conspiracism, and the dehumanization of out-groups) persists beneath rotating aesthetics, platforms, and vocabularies. Our own findings point in the same direction. Utilizing computational analysis of the online discourse of the major streams of the American violent far right, Hemmila and I found initial overlaps in designated adversaries alongside persistently distinct priorities, adversarial identities, and operational goals, a pattern consistent with adaptation and coalition-building rather than with the emergence of some new fused ideology (Hemmila & Perliger, 2025). This is also why the new category of nihilistic violent extremism should give the field pause, and here Horgan and Shayler’s (2026) caution is well placed. Labeling violence as nihilistic because we have not yet deciphered its ideological content converts an analytic failure of ours into a psychological attribute of the perpetrator. Residual categories have their uses in research. Institutionalizing one as a formal counterterrorism classification is how conceptual erosion hardens into policy.

Fourth, the evidentiary base on which these judgments rest has narrowed, to a troubling degree, to manifestos. Horgan and Shayler (2026) describe manifestos as “our most immediate and detailed glimpse into the kinds of ideas and content they want us to see.” The crucial clause is the final one. Manifestos are performative documents, frequently composed in the hours before an attack, addressed to an imagined audience of both followers and investigators, saturated with irony and in-group signaling, and heavily plagiarized from predecessors, as the chain of texts descending from Christchurch demonstrates (Ware, 2020). Inferring the depth or coherence of a perpetrator’s ideological commitment from such a document is akin to reconstructing a religion’s theology from a convert’s last letter. The sample is also badly skewed. Most perpetrators leave no manifesto at all, which means the cases that anchor our generalizations about “mixed” belief systems are precisely the ones curated for public consumption by their authors. When our primary instrument is this weak, the finding that perpetrators’ beliefs appear shallow, mixed, or incoherent is partly an artifact of the source. More specifically, ideological coherence is rarely located in any single text. It is located in years of participation, in forum posts, chat logs, memes, gaming servers, video playlists, and the organizational literature that circulates through them. The tools for analyzing such corpora at scale now exist, and the paradox Horgan and Shayler (2026) identify, that unprecedented access to extremists’ writings has coincided with diminishing consensus about their motivations, dissolves once we stop mistaking the manifesto for the discourse. When researchers utilize the full corpus of a perpetrator’s engagement, ostensibly incoherent cases frequently resolve into legible ideological trajectories.

Horgan and Shayler have performed a service in dismantling the salad bar metaphor, and their proposed schema offers a constructive foundation for empirical work. But abandoning a defective term will not by itself restore conceptual clarity. That will require the field to reinvest in what it has allowed to atrophy, namely the comparative, historical, and discourse-based study of violent ideologies, pursued by scholars willing to read the texts, trace the genealogies, and follow the conversations in which those ideologies actually live. It will also require us to hold two ideas simultaneously. Ideology is not a synonym for motivation, and ideology is not noise. The disciplines that study political violence did not lose their conceptual bearings because ideology vanished from the world. They lost them when they stopped studying it closely, and the remedy is available to us whenever we choose to take it up.

Conflicts of Interest

The author declares no conflicts of interest.

Readers of this article also read

‹

Commentary

The end of ideological violence

by Clark McCauley

Perspective

Why the ‘salad bar of ideologies’ does not help us understand contemporary violent extremism

by John Horgan & Morgan Shayler Cited by 5

Commentary

Extremism without ideology

by Quassim Cassam Cited by 2

Commentary

Why the ‘salad bar’ might actually help extremism research – A reply to Horgan and Shayler

by Stephane Baele Cited by 2

Review

Adverse childhood experiences and violent extremism: A scoping review of the literature

by Teja Pammi et al.

Research Article

Deadly persuasion: A multilevel analysis of leader influence tactics and lethality in violent extremist organizations

by Mackenzie Harms, Alex Bruens, Justin Jones, & Gina Ligon

Commentary

Shifts and continuity in contemporary violent extremism: Moving toward a dynamic and developmental understanding?

by Cécile Rousseau & Samuel Veissière

Research Article

System-disillusioned youth radicalize: The role of political alienation in the formation of political radicalism

by Marta Miklikowska & Tomasz Besta

›

References

Berger, J. M. (2018). Extremism. MIT Press.

Blee, K. M. (1991). Women of the Klan: Racism and gender in the 1920s. University of California Press.

Grant, M. (1916). The passing of the great race. Charles Scribner’s Sons.

Hemmila, T., & Perliger, A. (2025). Hybridization or salad bar ideology? Testing ideological convergence within the American violent far right. Crime & Delinquency, 71(3), 830–859. https://doi.org/10.1177/00111287241271288

Horgan, J., & Shayler, M. (2026). Why the ‘salad bar of ideologies’ does not help us understand contemporary violent extremism. advances.in/psychology, 1(1), e607682. https://doi.org/10.56296/aip00057

Laqueur, W. (1999). The new terrorism: Fanaticism and the arms of mass destruction. Oxford University Press.

Macdonald, A. (1978). The Turner diaries. National Vanguard Books. (Pseudonym of William Luther Pierce)

Threats to the homeland: Hearing before the Committee on Homeland Security and Governmental Affairs, United States Senate, 116th Cong. (2020) (testimony of Christopher A. Wray, Director, Federal Bureau of Investigation).

Ware, J. (2020). Testament to murder: The violent far-right’s increasing use of terrorist manifestos [Policy brief]. International Centre for Counter-Terrorism. https://icct.nl/publication/testament-murder-violent-far-rights-increasing-use-terrorist-manifestos