Noyaux sacrés-symboliques et périphéries mouvantes : une approche structurale du changement idéologique dans l'extrémisme

Introduction

A wide range of extreme ideologies has emerged in the internet era (Brace et al., 2024; Gartenstein-Ross et al., 2025), underscoring the need to identify core features shared across them and to explain their underlying mechanisms. The proliferation of typologies used to characterize ideology in extremism studies—exemplified by the ‘salad bar’ model, which treats ideological content as a composite—has increasingly drifted away from adequately addressing the challenges facing the field. As these typologies multiply and grow more varied, what emerges is cumulative confusion rather than clarity about precisely what is being studied. It is this inadequacy that draws attention to a structural framework of belief systems, which is considered more relevant than adding yet another content-based typology; in this respect, a structural approach offers a more coherent basis for explaining ideology’s role in extremism.

As Horgan and Shayler (2026) note, the mixing of ideologies in extremism is not a new phenomenon. We extend this observation with evidence from Indonesia, where we have documented a comparable pattern: extremists shifting between right-wing and left-wing extremist ideologies and vice versa (Muluk et al., 2014; discussed below). The phenomenon still has a limited explanation, exemplified by the fact that terminologies like “salad bar” proliferate.

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 →

Horgan and Shayler (2026) advanced a more fine-grained classification where so-called salad bar extremists are classified based on whether they mix or migrate between ideologies and whether the ideologies are convergent or divergent. Ultimately, we agree with Horgan and Shayler’s (2026, p. 16) point that it is more fruitful to utilize “language that incorporates, disentangles, and distinguishes respective processes and dynamics.” Consequently, we also agree with Rousseau and Veissière’s (2026) comment that the field must move beyond discussing which categories are correct and instead focus on the dynamic process, and Perliger (2026) calls for deeper engagement with the ideology itself.

At this point, we depart from Baele’s (2026) commentary, which is more sympathetic to the salad-bar conceptualization, and from McCauley’s (2026) argument, which privileges ideology less and views extremism primarily as intergroup conflict. Though we do not deny the possibility of non-ideological extremism as outlined by Cassam (2026), we argue that ideology still plays a key role in explaining extremism. Instead, we contend that integrating the dynamics of both ideological content and group processes is essential to addressing this issue. As we propose through the belief system framework below, such structural features include group process (identity), which is held to be central and symbolic, thereby integrating both the group process and ideological content. We argue that understanding extremism requires attention not only to the ideological content present but also to how that content is connected and structured.

Our claim that extremist ideology is not merely a collection of content but a structured system is supported by data from Islamists and jihadist groups in Indonesia, whose belief systems are anchored by a sacred value. This may limit the scope of our argument because sacred values are central and thus produce stable structures. Our findings may not generalize to extremist ideologies lacking such an anchor, limiting our challenge to the salad-bar concept more broadly construed. Nonetheless, the structural principle we observe in jihadist groups—centrality organizing an otherwise composite set of beliefs—may extend to extremist groups whose ideologies lack an apparent sacred anchor.

A structural approach of this kind makes it possible to trace connection strength and structural shifts in individuals undergoing extreme-to-moderate ideological moderation—in this regard, we can specify the psychological process that has so far remained underspecified in the literature. In what follows, we develop an alternative, structural account of ideological belief and use it to specify the psychological mechanism of change that the salad bar metaphor leaves unaddressed.

The Extremism Belief System Network

Drawing on a belief-system approach, ideologies can be understood as networks of interconnected nodes (belief elements) and edges (Brandt & Sleegers, 2021). We define ideology as a constrained set of interconnected beliefs about how society is and how it ought to be (Brandt, 2022; Kinder & Kalmoe, 2017). Belief systems vary in the degree of interconnection among elements: ideologies exhibit tightly connected networks, whereas lay beliefs remain loosely or unconnected. This framework also captures variation in centrality, as certain beliefs connect to many others and thus occupy a structurally central position.

In our research in the Indonesian context, we used self-report measures on radicalized individuals who have never been incarcerated and non-incarcerated. With a total sample of 42 participants, we assessed participants’ responses to the extent they are fused with religion and their willingness to fight for religion. We found that fusion to religion remained the same between the incarcerated (thus, deradicalized) and the non-incarcerated group, while willingness to fight for religion was lower for those incarcerated (Milla et al., 2026). We interpret this finding to indicate that the core feature of their belief system consists of religion. Furthermore, utilizing a conceptual similarity measure (see Brandt, 2022), we found that participants’ perceived connection between Islamic identities and Sharia predicted positively the willingness to fight for religion, controlling for fusion to religion itself. Conversely, participants’ perceived connection between Islamic identities and democracy was negatively linked to the willingness to fight for religion. Taken together, this preliminary finding suggests that extremist ideology indeed forms a structure and that a network structure framework is a fruitful way to explain extremism.

This framework also exemplifies its fruitfulness in explaining ideological change. Considering how central nodes tend to be symbolic (Brandt et al., 2019; Hasan et al., 2025; van Noord et al., 2024) and might be more “sacred” (Ginges & Atran, 2014; which we call a “sacred-symbolic”), ideological change can be posited as a reorganization around the new belief’s element (a shift), rather than an outright change. The system accommodates it peripherally while the sacred value retains its centrality.

Furthermore, drawing from the Goal System Theory (GST) framework, we conceptualize the sacredness of a belief as a motivated structure embedded within a broader goal system. GST posits that goals organize knowledge and beliefs into associative networks composed of means-goal links (Kruglanski et al., 2013, 2023). Within this structure, the constant-sum principle governs motivational processes: because the network maintains a fixed total associative strength, an increase in the number of connections reduces the strength of each individual link. Consequently, connecting more means to a single goal lowers the perceived instrumentality of each individual means (Bélanger et al., 2015; Schumpe et al., 2018). Applying this principle to belief-system networks, we argue that altering the connections among operational beliefs linked to a sacred symbolic goal will weaken their motivational potency. Moszczynska et al. (2025) demonstrate the same mechanism, showing that multiple interconnected commitments can facilitate goal attainment in political activism.

Consequently, a new belief can be adopted by adding loose connections to these sacred central nodes, which allows the new belief to attenuate prior, more extreme nodes and loosen the overall structure. The belief system may thus appear increasingly non-ideological on the surface while remaining tethered to the same central, sacred belief.

Applied to the “salad bar” phenomenon, what appears to be a composite mixing ideology might be a structured belief system with a central organizing node (Figure 1). Figure 1 illustrates this contrast, showing how belief elements are composed differently under a belief-system view than under a ‘salad bar’ view of ideology. This phenomenon is exemplified by a case of ideological shift from jihadism to socialism (Muluk et al., 2014), in which a militant Islamist activist became a left-wing environmental activist, advocating for extreme self-sacrificial behaviors, including sewing mouths and hunger strikes, while maintaining his Islamist status. As he stated himself, “For me, if I defend the people, their interests … it is jihad …” (Muluk et al., 2014, p. 31). What appears, on the surface, to be a mere salad-bar assortment of beliefs is in fact a structure—an idiosyncratically coherent belief system composed of central and peripheral nodes, with religion (Islam) occupying the central node throughout: “What is the most hated Islam, not fair, is it? Dzalim [oppression] right? Greedy, oppressive, overbearing, and slavery—the disease is not favored by Islam. Socialists fight for that right….” (Muluk et al., 2014, p. 31). His understanding of leftism is based on his anchoring in Islam and Jihad.

Figure 1
“Salad bar” view vs. Belief Structure View

Note. the apparent salad-bar composite ideology exhibited by a left-wing Islamist (left) and his actual belief structure (right; Muluk et al., 2014)

Three questions remain unresolved: which beliefs constitute the ideological core; how this structure comes to be formed; and—most critically—through what psychological mechanism it sustains the legitimacy of violence or gives way under pressure to change. Absent answers to these questions, extant accounts of ideology remain descriptive rather than explanatory. They may be useful for explaining what extremists believe, but not why that structure endures or what precipitates its transformation.

The Centrality of the Sacred-symbolic

As Cassam (2026) observes, although Horgan and Shayler (2026) caution against privileging ideological content over process, their taxonomy remains centred on content. We align with this observation but extend it further. Content anchored in sacred values is especially critical to understanding why some ideological structures resist change while others do not.

The composite ideology introduced by the salad bar metaphor may indeed describe contemporary extremism, in which individuals increasingly pick and mix from ideological narratives without the shared experiences with the groups that once anchored them—a pattern the internet era has made visible. Yet ideologies that legitimize violence typically revolve around sacred, change-resistant core attitudes. In jihadism, the focus of our argument is that violence is justified through the belief that jihad is divinely mandated—a sacred value not open to compromise (Milla et al., 2019). Our findings in jihadist groups in Indonesia, discussed earlier, for example, show that Islam occupies the ideological core that is relatively stable; however, a higher level of perceived connection between Muslim identities and democracy resulted in a weakened willingness to fight (Milla et al., 2026). Ideological content, in this sense, cannot be treated as an unstructured set of attitudes to be mixed and matched—individuals’ justification for violence stems from a sacred-symbolic core that tends to be non-negotiable. A shift in willingness to fight occurs once the more peripheral component (e.g., democracy) is more connected to such a core.

This finding bears directly on Horgan and Shayler’s (2026) alternative for a fine-grained analysis of the “salad bar” phenomenon, classifying whether the divergent or convergent beliefs are mixing or migrating. We agree that their classification scheme is an important analytical starting point, but our findings suggest a refinement: ideology is better conceptualized as an interrelated structure instead of merely as a mix of beliefs. Structure, in other words, is what makes ideology coherent as a subject of study. The salad bar framing, by contrast, offers no such structural account. It accommodates neither an explanation of content stability nor a specification of the relational processes that drive radicalization and moderation. What our findings point to instead is the role sacred-symbolic plays within the structure of a belief system. An approach that attends to ideological content in isolation, without this structural anchoring, therefore, leaves precisely the phenomenon Horgan and Shayler (2026) set out to explain analytically underspecified.

Taken together, these considerations suggest that the salad bar metaphor, however intuitively appealing, obscures more than it reveals about the mechanisms underlying ideological continuity and change in extremist movements. A belief-system framework—one that conceives of ideology as an organized structure of interlinked central and peripheral components, as Brandt (Brandt, 2022; Brandt & Sleegers, 2021) proposes—offers a more analytically useful lens for understanding extremism. Future research should prioritize empirically tracing this transformation process directly, rather than inferring ideological (in)coherence from content alone.

Funding

This research is funded by the Indonesian Endowment Fund for Education (LPDP) on behalf of the Indonesian Ministry of Higher Education, Science and Technology, and managed under the EQUITY Program (Contract No. 4302/B3/DT.03.08/2025 and 573/PKS/R/UI/2025).

Acknowledgement

The authors used ChatGPT (OpenAI) during the editing process to correct grammatical errors and improve the manuscript’s readability. Suggestions from the tool were then reviewed and edited by the authors before inclusion in the final manuscript. The authors take full responsibility for the final content of the manuscript.

Author Contributions

M.N.M. wrote the original draft and reviewed the final manuscript. W.Y. and U.H.S. contributed to the review and editing of the final manuscript. All authors provided critical revisions and approved the final version.

Data Availability Statement

No empirical data was analyzed for this manuscript.

Conflicts of Interest

The authors declare no competing interests.

Readers of this article also read

‹

Perspective

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

by John Horgan & Morgan Shayler Cited by 7

Commentary

The erosion of conceptual clarity in the study of political violence

by Arie Perliger Cited by 5

Commentary

Extremism without ideology

by Quassim Cassam Cited by 4

Commentary

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

by Stephane Baele Cited by 4

Commentary

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

by Cécile Rousseau & Samuel Veissière Cited by 2

Commentary

The end of ideological violence

by Clark McCauley Cited by 2

Research Article

The radicalization syndrome: Reciprocal dynamics of obsessive passion and deliberate engagement

by Jocelyn J. Bélanger, Daniel W. Snook, Muhammed Rayyan Ahmed, & Jais Adam-Troian

Review

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

by Teja Pammi et al.

›

References

Baele, S. (2026). Why the ‘salad bar’ might actually help extremism research – A reply to Horgan and Shayler. advances.in/psychology, 1, e359159. https://doi.org/10.56296/aip00060

Bélanger, J. J., Schori-Eyal, N., Pica, G., Kruglanski, A. W., & Lafrenière, M.-A. (2015). The “more is less” effect in equifinal structures: Alternative means reduce the intensity and quality of motivation. Journal of Experimental Social Psychology, 60, 93–102. https://doi.org/10.1016/j.jesp.2015.05.005

Brace, L., Baele, S. J., & Ging, D. (2024). Where do ‘mixed, unclear, and unstable’ ideologies come from? A data-driven answer centred on the incelosphere. Journal of Policing, Intelligence and Counter Terrorism, 19(2), 103–124. https://doi.org/10.1080/18335330.2023.2226667

Brandt, M. J. (2022). Measuring the belief system of a person. Journal of Personality and Social Psychology, 123(4), 830–853. https://doi.org/10.1037/pspp0000416

Brandt, M. J., Sibley, C. G., & Osborne, D. (2019). What Is Central to Political Belief System Networks? Personality & Social Psychology Bulletin, 45(9), 1352–1364. https://doi.org/10.1177/0146167218824354

Brandt, M. J., & Sleegers, W. W. A. (2021). Evaluating Belief System Networks as a Theory of Political Belief System Dynamics. Personality and Social Psychology Review, 25(2), 159–185. https://doi.org/10.1177/1088868321993751

Cassam, Q. (2026). Extremism without ideology. advances.in/psychology, 1, e945134. https://doi.org/10.56296/aip00062

Gartenstein-Ross, D., Zammit, A., Chace-Donahue, E., & Urban, M. (2025). Composite Violent Extremism: Conceptualizing Attackers Who Increasingly Challenge Traditional Categories of Terrorism. Studies in Conflict & Terrorism, 48(12), 1343–1369. https://doi.org/10.1080/1057610X.2023.2194133

Ginges, J., & Atran, S. (2014). Sacred values and cultural conflict. In Advances in culture and psychology, Vol. 4 (pp. 273–301). Oxford University Press.

Hasan, K. F. A., Yustisia, W., Milla, M. N., & Brandt, M. (2025). What is the most important issue in Indonesia’s political belief system? Network analysis of attitude and voting behavior in 2014 and 2019. Makara Human Behavior Studies in Asia, 29(2), 115–136.

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

Kinder, D. R., & Kalmoe, N. P. (2017). Neither Liberal nor Conservative: Ideological Innocence in the American Public. University of Chicago Press. https://press.uchicago.edu/ucp/books/book/chicago/N/bo25841664.html

Kruglanski, A. W., Fishbach, A., & Kopetz, C. (Eds.). (2023). Goal systems theory: Psychological processes and applications. Oxford University Press. https://doi.org/10.1093/oso/9780197687468.001.0001

Kruglanski, A. W., Köpetz, C., Bélanger, J. J., Chun, W. Y., Orehek, E., and Fishbach, A. (2013). Features of multifinality. Personality and Social Psychology Review, 17(1), 22-39. https://doi.org/10.1093/oso/9780197687468.001.0001

McCauley, C. (2026). The end of ideological violence. Advances.in/Psychology, 1, e145270. https://doi.org/10.56296/aip00066

Milla, M. N., Putra, I. E., & Umam, A. N. (2019). Stories from jihadists: Significance, identity, and radicalization through the call for jihad. Peace and Conflict: Journal of Peace Psychology, 25(2), 111–121. https://doi.org/10.1037/pac0000371

Milla, M. N., Sulaiman, U. H., Arifin, H. H., & Louis, W. H. (2026, July). The operational structure of a radical belief system of an Islamic Radical Group: A mixed-method analysis. Paper presented at the 49th Annual Scientific Meeting of the International Society of Political Psychology, Newcastle, UK.

Muluk, H., Milla, N.M., & Cahyono, W. (2014). An Introduction to Radicalism of Civil Society Movements in Indonesia: Intersection and Shifting Ideology between Socialists and Jihadists. SSRN. http://dx.doi.org/10.2139/ssrn.2479191

Perliger, A. (2026). The erosion of conceptual clarity in the study of political violence. Advances.in/Psychology, 1, e824472. https://doi.org/10.56296/aip00059

Rousseau, C., & Veissière, S. (2026). Shifts and continuity in contemporary violent extremism: Moving toward a dynamic and developmental understanding? advances.in/psychology, 1, e631837. https://doi.org/10.56296/aip00068

Schumpe, B. M., Bélanger, J. J., Dugas, M., Erb, H.-P., & Kruglanski, A. W. (2018). Counterfinality: On the increased perceived instrumentality of means to a goal. Frontiers in Psychology, 9. https://doi.org/10.3389/fpsyg.2018.01052

van Noord, J., Turner-Zwinkels, F. M., Kesberg, R., Brandt, M. J., Easterbrook, M. J., Kuppens, T., & Spruyt, B. (2024). The nature and structure of European belief systems: Exploring the varieties of belief systems across 23 European countries. European Sociological Review, jcae011. https://doi.org/10.1093/esr/jcae011