At the National Propaganda Work Meeting in August 2013, Xi Jinping 习近平 warned that the “collapse of a regime often begins in the ideological sphere.” While “ideological evolution is a long-term process,” he cautioned the attending cadres that once the “ideological front was breached, it becomes extremely difficult to defend other fronts.” To him, this was existential, as “political unrest and regime change may occur overnight.”Footnote 1 His stark language echoed the now-infamous Document No. 9 on the “State of the ideological sphere,” which was circulated later that year.Footnote 2 Yet, at the time, these words were invisible to the outside observer, as neither the official readout of the speechFootnote 3 nor the authoritative Xi Jinping Speeches Database included them. As of 2025, the full transcript remains unavailable. However, various sections of his speech have appeared in compilations on cultural construction and national security, and additional fragments have been published in study readers and official anthologies. This disjointed publishing makes it near impossible to reconstruct his speech in its entirety. This opacity might be taken as emblematic of the Party’s secrecy surrounding elite discourse. But more fundamentally, it reveals something deeper about how ideological meaning is produced, circulated and transformed in China’s political system.
The treatment of this speech illustrates more than the black box surrounding China’s elite politics. It reveals a complex system of textual mediation: Xi’s oral remarks, delivered behind closed doors to Party officials, are disassembled into fragments and reassembled into curated ideological formats. These fragments are removed from their original temporal and textual contexts and inserted into new configurations, often alongside excerpts from unrelated speeches given years or even decades apart. This editorial logic is not anomalous but systematic. Fragments circulate throughout the corpus of 115 books published up until the end of 2024 in a constant process of extraction and recomposition, appearing in “collected works” (zhuzuo ji 著作集), “compilations of discourses” (lunshu zhaibian 论述摘编), “work and life experiences” (gongzuo shenghuo jingli 工作生活经历) and “study readers” (xuexi duben 学习读本). Each booklet may contain hundreds of fragments drawn from dozens of speeches and restaged to support new argumentative trajectories that often bear little relation to the source materials from which they came.
The practice of ripping excerpts from contexts and remixing them across genres and timelines is one of the hallmark features of Xi Thought.Footnote 4 It turns each published volume into what Roland Barthes called a “tissue of quotations drawn from the innumerable centres.”Footnote 5 Here, more literally, it becomes a tissue of fragments drawn from Xi’s speeches. Each work is a composition, “eternally written here and now,” and its meaning emerges not from authorial intent or historical fixity but from the connections that are assembled in the present.Footnote 6
What kind of ideological formation emerges when meaning is produced not through coherent exposition but through the continual disassembly and recombination of textual fragments? This article posits that rather than functioning as a system of stable, doctrinal propositions, Xi Thought emerges as a dynamic assemblage shaped by the continual reconfiguration of speech fragments, textual excerpts and conceptual linkages. Moreover, the practice of fragmenting and re-territorializing Xi’s statements across disparate publications does not merely obscure original meanings, it creates new ones, enabling different elements to cohere in novel constellations.
This process exemplifies what Gilles Deleuze and Félix Guattari describe as a “rhizomatic” structure: a formation that expands not through a hierarchical logic or linear development but through lateral connections, recombinations and emergent linkages. Meaning is not imposed from the top down but produced through the dynamic arrangements of fragments. The result is an ideological system marked by heterogeneity, multiplicity and an open-ended logic. As a result, Xi Thought cannot be reduced to a coherent doctrine but rather must be understood as a flexible, dynamic and resilient ideological formation that is constantly being reassembled, repurposed and rendered intelligible through new textual architectures.
Fragments are reclassified under new topical headings, stripped of context and repurposed to support arguments often far removed from their original meaning. In this process, Xi is not the architect of a stable doctrine but a source of raw textual material to be reorganized by the Party’s ideological apparatus and other actors. Traditionally, authors guarantee a “certain unity of writing”;Footnote 7 however, Xi’s name serves less as a guarantor of coherence than as a generative placeholder under which an ever-expanding corpus is produced. The result is not one Xi Thought but rather a potentially infinite proliferation of “Xi Thoughts.” This contrasts with the prevailing literature, which tends to treat Xi Thought as a changing yet ultimately closed doctrinal system. What these approaches share is a commitment to asking what Xi Thought is. The analytical shift proposed in this article asks how Xi Thought works: how it mutates, recombines and circulates as a fragmented and dynamic ideological formation.
What is at stake here is not simply an alternative way of reading Xi Thought but a different model of how ideology operates in contemporary China. A rhizomatic approach moves analysis away from the presumption of a unifying source of ideological meaning. Instead, it highlights the existence of decentralized, dynamic and modular processes during which meaning is distributed across the ideological apparatus, shaped by editorial decisions, institutional needs and a multitude of minuscule shifts of degree enacted by local actors that accumulate and, over time, lead to shifts in kind. As ideology increasingly bleeds into policy, understanding this logic of production becomes essential not only for scholars but also for analysts, diplomats and policymakers seeking to interpret the signals emanating from China’s political system. The question of how Xi Thought works is, therefore, not merely academic; it is of acute political importance.
The article proceeds in four parts. First, I review the existing scholarship on Xi Thought and Chinese ideology, showing how it tends to reduce the phenomenon to a static and hierarchical essence. Second, I introduce the conceptual vocabulary of Deleuze and Guattari’s rhizome. Third, I demonstrate how the core features of the rhizomatic structure, fragmentation, multiplicity, malleability and cartography manifest in the production of Xi Thought. Finally, I reflect on the implications for understanding ideology in China.
Approaches to Ideology in China
Scholarship on ideology in China can be grouped, albeit imperfectly, into five broad approaches: functionalist, practice-based, slogan-centric, belief-oriented and historicist accounts. Despite their differences, these approaches share a tendency to analytically stabilize the fragmented and evolving nature of Xi Thought.
First, from a functional perspective, in the classic Marxist understanding, ideology is intimately linked to maintaining the political and economic interests of the ruling class. Perhaps the most widespread understanding of ideology in the Chinese context links it to legitimacy and cohesion. Heike Holbig argues that ideology signifies continuity with the past and thereby enhances regime legitimacy.Footnote 8 John Garrick and Yan Chang Bennett argue that under Xi Jinping, ideology has become a tool for power consolidation.Footnote 9 However, power consolidation is not always framed as Xi-centric. Gang Chen, for instance, contends that it is meant to provide “a renewed sense of cohesion and mission” for the Party as a whole.Footnote 10 Beyond the Party, Huang Haifeng points to the signalling function of ideology to show the strength of the party-state,Footnote 11 while Jean Christopher Mittelstaedt argues that ideological campaigns under Xi have established hegemonic narratives and crowded out competing narratives.Footnote 12
A second strand reduces ideology to practices. This scholarship builds on ideas of ideology as practice and its relatively autonomous existence, manifested in everyday rituals and practices. Mittelstaedt has examined the production and reproduction of ideology at the grassroots of the Chinese Communist Party (CCP) through its monthly rituals.Footnote 13 Clyde Yicheng Wang shows in his study of propaganda that through its online dissemination, ideology becomes part of the “inescapable everyday experience.”Footnote 14 Marina Svensson focuses on another practice: the university classes through which ideology is transmitted.Footnote 15 Likewise, Mittelstaedt and Patricia M. Thornton look at how ideology is produced through a mixture of cultural and volunteering practices at the New Era Civilization Centres.Footnote 16 Like the functionalist view, the practice-based literature also assumes a unifying logic.
This logic is also evident in a third category of literature that explores ideology through slogans and key concepts. For example, Kerry Brown and Una Aleksandra Bērziņa-Čerenkova identify 12 keywords that embody Xi’s overall ideology.Footnote 17 Other times, ideology under Xi is seen through the lens of the “new era.”Footnote 18 In the foreign policy realm, under the theme of “slogan politics,” Zeng Jinghan shows how key foreign policy slogans are subject to manipulation, leading to a host of different interpretations and “driving policy discussion.”Footnote 19 Eun A. Jo and Jessica Chen Weiss reduce ideology under Xi to the “China model,” as an alternative to the West and broader visions of socialist modernization.Footnote 20
Fourth, some scholars see ideology as emanating from individuals’ beliefs. In this argument, Xi Thought mirrors Xi’s personal convictions. François Bougon sees Xi’s texts as reflecting his beliefs.Footnote 21 Similarly, Willy Lam traces Xi’s obsession with ideology to the ambition of becoming the “equal of Deng Xiaoping and Mao Zedong.”Footnote 22 Focusing on Xi’s early life, Joseph Torigian identifies toughness, idealism, pragmatism and caution as the characteristics that influence his worldviews.Footnote 23 Linking Xi Thought back to his relationship with his family and his upbringing, Torigian finds these traits reflected in Xi’s speeches and articles.Footnote 24 This approach tends to essentialize Xi Thought by locating its coherence in Xi’s personal convictions.
Lastly, the official understanding of Xi Thought stresses its historicity. According to Chinese legal scholar Jiang Shigong, Xi Thought did not emerge “automatically” but through the “struggle of the entire Party, the entire country and the entire people.” It is both the “crystallization of the wisdom of the entire Party” and the product of the “integration of Marxism with the practice of contemporary China” and “Chinese traditional culture.”Footnote 25 The most authoritative and recent redefinition in the 2021 History Resolution defines Xi Thought teleologically as “the historical experience gained since the founding of the Party” and proceeding from “new realities.”Footnote 26
Steve Tsang and Olivia Cheung provide the most systematic analysis of Xi Thought. While they posit it as evolving and adapting, their treatment reproduces many of the same tendencies found across the literature. They frame Xi Thought as a system organized around “key tenets,” with the China Dream at its core and rooted in Xi’s personal ideas, and they interpret it as a teleological project, aimed at elevating it to the stature of Mao Zedong Thought.Footnote 27 In doing so, their analysis presumes Xi Thought’s unity and stabilizes it as a coherent ideological object. By contrast, this article’s starting point is to treat unity as an effect, not a premise.
Whether functional, performative, conceptual, personalistic or teleological, a shared epistemic structure exists: the assumption that an underlying unity ultimately renders Xi Thought intelligible. Existing scholarship gestures towards fragmentation, flexibility or contextual adaptation, particularly in descriptive analyses of policy variation and discursive change. What it does not do, however, is treat ideological unity itself as a problem requiring explanation. Functionalist accounts focus on ideological form while ignoring shifting content; practice-based approaches obscure the intertextual connections between ideological acts; and slogan-centric analyses fixate on keywords while missing their evolving and unstable meanings. Personalist interpretations reduce ideology to biography, while historicist readings crush agency under the weight of inevitability. By contrast, this article proposes a different epistemological orientation, one that sees Xi Thought as a rhizome. This focuses attention not on origin, order or coherence, but on movement, multiplicity and recombination as the primary analytical focus. Xi Thought can, in the words of Barthes, only be “followed, ‘run’ (like the thread of a stocking).”Footnote 28 Take the idea of Party leadership, a core political axiom. The proposed analytical shift does not deny its centrality but rather shows how its centrality is produced. The difference between the literature and understanding Xi Thought as rhizomatic is, therefore, not empirical but analytical: it is not whether Xi Thought changes or produces unities, but how they are established. Below, I develop this theoretical lens and trace its implications for understanding the structure and operation of Xi Thought.
The Rhizome
Gilles Deleuze and Félix Guattari introduced the concept of the rhizome in A Thousand Plateaus to challenge traditional hierarchical models of thought.Footnote 29 Unlike a tree, which has roots from which a trunk, branches and leaves emerge to form a static vertical structure, a rhizome resembles a wall of vines: sprawling, decentred and continuously branching out. It has no defined beginning or end, and parts of it may die off without disrupting the whole. Rhizomes are heterogeneous and non-hierarchical by nature, defined not by a fixed identity but by continual recombination, adaptation and relational motion. In rhizomatic space, hierarchical or tree-like patterns, such as dominant ideological configurations, may appear temporarily at points of convergence, but they remain contingent and reversible; they emerge at the intersection of lines that can always shift, branch out or break. While linked to power, such as an organization that wants to stabilize meaning, this is not sufficient.Footnote 30 Rather, it is the growth of the rhizome that determines the shifts in meaning and identity that adapt to new contexts and connections. This framework understands Xi Thought not as a doctrinal totality but as a dynamic assemblage of ideas and practices, which are continually and haphazardly adapting to institutional changes, policy shifts and local practices and interpretations.
As a rhizomatic structure, Xi Thought has several characteristics. First, it embodies heterogeneity and the principle of connection. The corpus of work related to Xi Thought, such as his speeches, writings and policy directives, forms a vast interconnected network. The disassembly of Xi’s speeches into fragments and their recombination in disparate works shows how past texts are recontextualized and linked to new ideas, creating new meanings and interpretations. Once separate and peripheral, fragments can become central nodes in a growing network and can connect to any other fragment, enabling new meanings and growth.
Further, Xi Thought adopts multiple identities. Each “type” of Xi Thought (military, economics, law, diplomacy, culture or ecology) is both autonomous and interconnected. For instance, “foreign-related law” embeds differently in Xi Legal Thought than it does in Xi Diplomatic Thought, yet both are linked through their connections to principles like Party leadership or national rejuvenation. This multiplicity means Xi Thought adapts to different contexts and can develop at varying speeds and intensities.
The third principle of rupture suggests that damage to one part of a rhizome does not impede the whole. Rather, the rhizome remains resilient and capable of forming new connections. In Xi Thought, certain themes have been de-emphasized or replaced over time without undermining the overall framework. For example, the initial prominence of the China Dream in the first volume of The Governance of China diminished significantly in later volumes, as new concepts like “Chinese-style modernization” took its place. New concepts and ideas emerge and are incorporated, while old ideas are discarded or repackaged and linked to the new ideas, thereby creating new meaning.
Lastly, the concept of mapping underscores how Xi Thought resists being confined to a linear narrative. Instead of tracing a direct path from past to present, mapping involves creating new linkages and shifts over time. The continual recombination of fragments generates novel connections between past and present. Through this process, small shifts of degree can produce shifts in kind. Xi Thought is thus not a stable doctrinal corpus but a flexible, evolving map.
In sum, rhizomatic thinking provides a framework for understanding how Xi Thought exists as a set of interconnected ideas and practices that are constantly being redefined and reorganized. The rhizomatic structure thus enables us to see Xi Thought not as a closed, coherent system but as an open-ended process of becoming, which can only be mapped rather than traced back to a fixed essence. Below, I examine these four characteristics in turn.
Heterogeneity in Xi Thought
Xi Thought is heterogeneous on two levels: at the corpus level and at the level of individual works. The corpus level comprises a diverse array of works that, although produced across different times and contexts, form a dynamic and evolving network.
At the end of 2024, 115 publications constituted Xi Thought, divided into four main categories.Footnote 31 The first, “collected works,” presents abridged versions of individual speeches, tailored to specific themes. These include the four volumes of The Governance of China and collections on youth work or scientific and technological self-reliance. By the end of 2024, 31 such collections had been published. In the second category, “compilations of discourses,” fragments from multiple speeches are brought together under a common thematic heading. By the end of 2024, 44 volumes had appeared in this format. “Work and life experiences” includes biographical volumes that narrate Xi’s time as a sent-down youth or his various Party appointments. In 2024, 20 books of this type had been published. The fourth category, “study readers,” which includes the 2018 and 2023 Xi Thought study outlines, as well as topical volumes on Xi’s thoughts on law and the economy, offers thematic syntheses. By 2024, 20 of these readers had been published. These categories frequently blur: some volumes, such as those on Party self-revolution, combine abridged speeches and thematic fragments.
The publication of leadership speeches has long been a routine practice within the CCP, and Xi Jinping is no exception. Prior to 2012, a 1992 collection from his tenure in Ningde 宁德市 and a 2006 volume on Zhejiang’s development circulated within Party channels. However, it was only after Xi’s rise to the Politburo Standing Committee in 2007 and later to the general secretaryship in 2012 that these texts were republished and reclassified. What had once been disparate local reflections were now repositioned within a broader ideological corpus and related to each other. His 1992 Ningde speeches were re-issued in summer 2014 and linked to the then-concluding mass line campaign. Similarly, his 2006 Zhejiang remarks were republished in October 2013, followed by other Zhejiang excerpts a month later. The first compilation focusing on national rejuvenation and the China Dream was published in December 2013, further weaving disparate fragments together to form a cohesive narrative. This process marked the emergence of new textual linkages, transforming previously isolated materials into ideological components of what would in 2017 become Xi Thought.
This ongoing process of republication and recontextualization shaped not only the corpus as a whole but also the internal structure of individual works. These volumes are heterogeneous assemblages comprising abridged speeches, excerpts and summaries: fragments drawn from multiple contexts and time periods. Speeches are broken down into segments, which are then re-territorialized within thematic collections. These fragments do not revolve around a single conceptual pivot; rather, their meaning derives from the networks in which they are embedded. Thus, in principle, every fragment has an unlimited range of potential meanings, depending on where and how it is deployed. Given this relational ontology, works are inherently provisional: most are not republished, and their relevance is often fleeting. Such works also lack overarching arguments. They do not require linear reading, and any chapter or fragment can serve as a point of entry. The rhizomatic structure ensures that no privileged order exists. Rather, coherence is always partial and emerges in the present.
The 2016 volume, Selected Excerpts from Xi Jinping’s Discourses on Comprehensive and Strict Governance of the Party (Xi Jinping guanyu quanmian congyan zhidang lunshu zhaibian 习近平关于全面从严治党论述摘编, hereafter Selected Excerpts on Strict Governance), which was republished in amended form in 2021, exemplifies this structure. It contains 371 fragments drawn from 87 speeches. Three speeches alone account for 94 excerpts (25 per cent), while 36 speeches are cited only once. There is no dominant speech, no unifying trajectory. The volume is both textually and temporally fragmented: the most frequently cited speech dates to January 2016, followed by others from the Sixth Plenum and the national organizational work meeting. The volume has no beginning or end and does not build a linear argument. It is a rhizome: it is open, fragmented and non-totalizing.
This also shapes temporal organization. More recent materials are favoured: speeches from 2016 are cited 129 times, compared to 75 citations from 2015 and only 13 from 2012. Yet these fragments are not arranged chronologically. Time does not function as a hierarchy. Instead, fragments from 2012 and 2016 coexist within the same conceptual space. As Henri Bergson’s concept of the “virtual” suggests, past fragments persist not as fixed historical artefacts but as resources. Every paragraph or line from any speech can become a productive unit, capable of de- and re-territorialization. All of Xi’s published and unpublished speeches and texts become an inexhaustible repository of “multiple potentials,” with the original context and publication date being secondary to the fragment’s current use and linkage.Footnote 32 Once delivered, Xi’s speeches are, therefore, not immediately forgotten. Dis-assembled into fragments, they can re-appear at any point in a different context. The ideological significance of a fragment is not what it once meant, but what it can now do within a new textual assemblage. As such, Xi Thought is best understood not in real time but in retrospect, as its meaning emerges through the ever-changing constellation of fragmentary linkages.Footnote 33
Take the example of a paragraph from a speech Xi delivered at the Central Economic Work Meeting in December 2014, which subsequently re-appeared across multiple authoritative compilations, each time recontextualized. In the original speech, Xi described a “law-based economy” as the essence of the socialist market economy, pairing this formulation with remarks on cadres’ capacity for economic work and criticism of the use of administrative orders that exceeded legal bounds. Situated within a discussion of the “new normal” and structural economic adjustment, the passage functioned as a subsidiary reflection within a broader argument about economic transition. When incorporated into a 2015 compilation on law-based governance, the same paragraph was reframed as a statement on the legal accountability of leading cadres. In a 2016 compilation, it was redeployed to support the strategic objective of building a moderately well-off society. By 2019, the unchanged passage appeared in a compilation on Party leadership, where law-based governance and cadre competence were subordinated to the imperative of strengthening Party leadership over economic affairs. Subsequent incorporations further narrowed its function: the 2022 volume on Xi Economic Thought retained the reference to a law-based economy while excising legal accountability, and a 2024 compilation on financial work reduced the passage to a technocratic emphasis on economic capacity alone. A paragraph initially tethered to structural economic change was initially mobilized as a constraint on cadre behaviour, then stripped of its normative content, leaving behind a depoliticized language of competence available for redeployment across domains. What began as a minor aside in a speech on economic restructuring was thus recontextualized several times.
Fragments remain perpetually in motion. They rupture existing unities and form new configurations. There is no master text and no central origin, only an ever-expanding mesh of texts, concepts and linkages. The ideological field is continually reconfigured, not by external imposition but through the ceaseless recombination of its parts. In this sense, Xi Thought is not a bounded doctrine but an ongoing process of recomposition in which heterogeneity is not an attribute but the very condition through which new meanings are continuously generated.
Multiplicities in Xi Thought
Building on the heterogeneity of Xi Thought, multiplicities arise as fragments link and re-link in dynamic configurations, enabling it to manifest in diverse forms and identities. Each paragraph or line of a speech provides a distinct perspective while simultaneously connecting to all other works. These fragments function as resources that both shape and constrain the potential identities Xi Thought can adopt, influencing its re-territorialization, its recontextualization and redefinition, in new thematic settings. Each volume assembled from these fragments constitutes a transient unity, relating to others through shifting linkages rather than stable thematic cores. As Deleuze and Guattari put it, a rhizome “operates by variation, expansion, conquest, capture, offshoots.”Footnote 34 Xi Thought proliferates through similar logics: any fragment can connect with others in novel configurations, giving rise to multiple ideological forms. Identity, in this framework, derives from its linkages across all elements of Xi Thought. However, not all connections are equal in density or speed: some areas develop more rapidly and become saturated, while others evolve more slowly. This uneven distribution of linkages creates a fluid and continuously shifting network of meanings and identities.
By 2024, six thematic types of Xi Thought had emerged: military, economic, ecological, diplomatic, legal and cultural. Although autonomous in their orientation, these types are sustained by the fragments that constitute their internal logics. Consider the concept of “foreign-related law.” In Xi Legal Thought, it emphasizes harmonizing domestic and international legal spheres through legal reform and institutional alignment. In Xi Diplomatic Thought, however, the same term is reframed to emphasize sovereignty, legal warfare and the protection of national interests. Similarly, “common prosperity” links to the Party’s founding mission and social policy in Xi Economic Thought, connecting poverty alleviation, education and upward mobility, but it appears only vaguely in Legal Thought as a broad aspiration associated with the China Dream. Thus, while concepts may recur across types, they are embedded in different linkages, producing different identities.
Individual types of Xi Thought are broadly self-sustaining and temporarily unified in the form of study guides. These include not only canonical study outlines and study Q&As but also materials like the “30 Lectures,” a widely disseminated set of PowerPoint presentations designed for cadre training. Xi Military Thought, established in October 2017, saw guides published in 2019, 2022 and 2023. Xi Economic Thought, formalized in December 2017, received a study guide in 2022. Ecological Civilization Thought, established in May 2018, was accompanied by a guide in 2022 and 2025, while Diplomatic Thought, established in June 2018, saw publications in 2021, 2022 and 2023, including volumes of collected diplomatic speeches. Legal Thought (November 2020) was supported by guides in 2021 and 2023, while Cultural Thought (October 2023) saw a guide published in late 2024. These materials periodically systematize and stabilize each type of Xi Thought, albeit only momentarily. They represent instances where doctrinal consolidation, hierarchical ordering and unified interpretation briefly reassert themselves within the rhizome. The ideological field is thus not purely horizontal or fluid: moments of vertical codification punctuate its movement, inscribing temporary boundaries and sanctioned meanings. Yet these instances never fully resolve the system into a fixed structure. Their stabilizing effect is always provisional and soon overtaken by the centrifugal momentum of new fragments.
Although updated regularly, study guides often fall out of sync and are rendered obsolete by newly emerging concepts. For instance, the 2022 guide to Xi Economic Thought does not incorporate the concept of “New quality productive forces” (xin zhi shengchanli 新质生产力), which emerged only in September 2023, thus requiring readers to understand its place within a more recent field of linkages. If updated, the guide will again impose a temporary coherence. Even this unity will eventually dissolve as the conceptual terrain continues to evolve. As new concepts and elements emerge, new “orientations of thought [come] into being” and new meanings are generated through reconfigured fragmentary associations.Footnote 35
Rhizomes do not resolve into a single point of control or identity and so each part of the system can function independently. As some areas of Xi Thought become outdated and fall into disuse, others continue to expand. This ensures the continued evolution of the ideological system and also means that new types of Xi Thought may always emerge. Wherever sufficient linkages are formed, fragments may be assembled into a new unity, which, however, is provisional, partial and always in motion.
Ruptures in Xi Thought
The third characteristic of the rhizome, and a key difference from a traditional hierarchical structure, is rupture. Whereas disruption can render a conventional structure inoperative, this is not the case for a rhizome. In rhizomatic systems, rupture does not entail collapse; rather, it opens new trajectories of growth.Footnote 36 As Deleuze and Guattari note with the example of ants, even if an ant colony is almost entirely destroyed, the remaining ants can re-establish old lines or form entirely new ones. In this sense, rupture is not terminal but generative, and existing concepts and elements allow Xi Thought to continue to evolve.
This logic is evident in the fading prominence of the China Dream. First invoked in 2012 and long treated as a “signature ideology,” it was central to early formulations of Xi Thought and appeared frequently in official discourse throughout the mid-2010s.Footnote 37 But by the early 2020s, it had receded from view, appearing less frequently in key publications and study guides. This retreat did not, however, signal an ideological crisis. On the contrary, Xi Thought continued to expand, absorbing and pivoting to new themes.
This dynamic is also visible in the evolution of Xi Thought’s study guides. Their shifting content illustrates how certain elements are discarded, others subjugated or reformulated, and still others brought back. Rupture and recomposition are not episodic but rather are constitutive features of the ideological system itself. The contrast between the 2019 and 2023 Xi Thought study guides offers a clear illustration of this adaptation. Both guides retain the same high-level structure of 19 chapters, but the 2023 edition expands to 113 subsections, compared to 95 in 2019. Approximately 55 per cent of the subsections remain unchanged, while others are restructured, dropped or newly added.
The 2019 version opens by emphasizing the historical continuity with Deng Xiaoping’s 邓小平 “socialism with Chinese characteristics,” whereas the 2023 edition foregrounds Party leadership and the “two upholds” (liangge weihu 两个维护).Footnote 38 This signals a shift from legitimating Xi Thought’s emergence to reinforcing its authority through the central role of the CCP. Similarly, the chapter on “underlying principles” shows both structural stability and conceptual change. While four of its six subsections remain unchanged, the 2019 subsection, “Socialism with Chinese characteristics is socialism, not any other ‘ism’,” is replaced in 2023 by a focus on the new historical juncture. Subsection five, originally about Marxism for the 21st century, is reformulated around the “two integrations” (liangge jiehe 两个结合), linking Xi Thought with both Marxism and traditional Chinese culture, in line with its redefinition in the 2021 History Resolution. Chapter three exhibits a more profound transformation. In the 2019 version, it centres on the “people-centred approach”; in 2023, this theme is moved to chapter four, with chapter three now devoted to “Chinese-style modernization” as the new “aim and task” of Xi Thought. This reframing positions modernization, and not the China Dream, as the ideological centrepiece, drawing on themes such as national rejuvenation, Party leadership and a critique of Western developmental models.
This reorganization extends across domains. While major themes such as law, Party-building, the economy, environment and foreign affairs remain, their ordering and internal emphases shift. Owing to its increasing importance, Party-building moves from chapter 18 in 2019 to chapter 8 in 2023. A new chapter on cadres is added, stressing themes like struggle, risk mitigation and political unity. On law, new content on constitution-based governance appears. In economics, the emphasis shifts from constructing a modern economic system to advancing a modern manufacturing system. In culture, soft power is replaced with the “dissemination and influence of Chinese civilization.” Environmental sections now include green and low-carbon development; national security has gained a new focus on political security.
Additionally, several new concepts introduced in 2023, including “whole-process people’s democracy,” “breaking the historical cycle” and the “new form of human civilization,” signal a broader recalibration through the introduction of newly coined terms. Rather than replacing earlier goals outright, they are layered onto the existing framework, gradually shifting the centre of gravity within Xi Thought. Change here is not a rupture in the sense of breakage, but rather recomposition.
This same logic of recomposition applies at the level of individual subfields and works. Some compilations are refolded into new lines of ideological work. For example, the 2017 compilation on socialist economic construction was retroactively subsumed into Xi Economic Thought. Others are left behind: the 2020 collection on COVID-19 and economic recovery has faded into irrelevance. Still others are neither wholly abandoned nor absorbed but updated and reframed. Comparing the previously mentioned 2016 version of Selected Excerpts on Strict Governance with its republished 2021 version makes this clear. The 2021 version expands the number of documents from 87 to 223 and the number of fragments from 371 to 788. While 26 per cent of the 2016 materials are retained, 65 per cent are newly added and 9 per cent discarded. This is not mere accumulation but rather a selective recomposition. Themes like the “two upholds” and new emphases on policy implementation and cadre responsibilities signal a widening of strict Party governance beyond internal CCP mechanisms, extending its reach into broader governance and state policy.
Taken together, these developments show that Xi Thought does not merely adapt or adjust. Rather, it mutates. Mutation here refers to an emergent, contingent process that can only be grasped retroactively. It is not merely additive but selectively generative. Fragments that once anchored major concepts, like the China Dream, can fade into marginality. Others gain new prominence through shifting linkages. Although every fragment theoretically holds the potential to connect with any other, these connections are uneven, contingent and shaped by institutional conditions and political demands. Not all potential is actualized, and not all paths are taken.
Xi Thought, in this view, is not a doctrine with a fixed kernel and orientation but a rhizomatic process of continual recomposition: reassembling, discarding and mutating. It is a field of ideological production that gains resilience not from stability but from the capacity to change its shape while remaining legible. Rupture is not exception but method, and mutation is not an anomaly but a norm, while incoherence is a feature, not a bug.
Mapping Xi Thought
These characteristics mean that Xi Thought can only be mapped, not traced. This distinction features prominently in Deleuze and Guattari’s work. They contrast mapping, a generative, open-ended process, with tracing, which assumes an origin and replicates existing structures. A rhizome cannot be traced in advance; rather, it must be mapped through encounters.
The party-state, however, engages in a paradoxical practice. In producing study guides, speech compilations and thematic volumes, it seeks to stabilize meaning through authorized interpretation. Yet once these texts are published, they circulate beyond the Party’s control. Released into the public sphere, they become open to new readings and unpredictable reconfigurations. As Paul Ricoeur reminds us, interpretation always generates a “surplus meaning,” resonances the author cannot foresee.Footnote 39 In the case of Xi Thought, these accretions are often small and local, but over time they accumulate, producing mutations within the ideological corpus. This surplus cannot be traced back to its origin; it must be mapped as it emerges.
The emergence of Xi Cultural Thought in October 2023 provides a paradigmatic example. It materialized not in response to a specific historical juncture but through the convergence of multiple pre-existing lines: Party leadership in propaganda work, socialist core values, news production, cultural industries and heritage, external propaganda, international cultural exchange and soft power. These lines rearranged familiar materials into an original ideological assemblage. Like the naming of Xi Thought in 2017, the formation of Cultural Thought created a provisional unity that invited actors to fill it with content. A subsequent Party symposium called on cadres to “interpret [it] to achieve deeper understanding,” a signal to begin decentralized elaboration.Footnote 40
At the first level, Xi Cultural Thought interlinks with other types of Xi Thought to produce internal recombinations. For instance, linking law and culture, the Supreme People’s Court declared that “legal interpretation, reasoning, and mediation in judgments are also a form of cultural dissemination and creation.”Footnote 41 Similarly, links to Xi Economic Thought emerge via the claim that the “cultivation and practice of socialist core values” foster an “innovative culture” that can “accelerate the formation of new quality productive forces.”Footnote 42 Integration with big data, AI and VR has been cited as central to enhancing cultural production under Cultural Thought.Footnote 43 In diplomatic registers, it intersects with the Global Civilization Initiative and discourse construction, reinforcing its entanglement with Xi Diplomatic Thought.Footnote 44 These recombinations show how Cultural Thought becomes constitutive of institutional practice through its transversality.
At the second level, Xi Cultural Thought is absorbed into state discourses through institutional codification. The 2024 Government Work Report, for instance, lists it under “people’s spiritual and cultural life” and associates it with a wide array of domains: philosophy and social sciences, news publishing, broadcasting, literature and art, cultural digitalization, online governance and even the Olympic Games.Footnote 45 On the surface, this appears to reduce Cultural Thought to an administratively bounded category. Yet this very act of formalization generates new lines of potential. By incorporating Cultural Thought into disparate policy arenas, the state unintentionally proliferates points of contact: new surfaces where fragments may be recontextualized and interpreted. Rather than closing down meaning, such institutional incorporations produce a surplus: they make the concept newly available, inviting appropriation and transformation.
At the third level, Xi Cultural Thought becomes an object of creative recomposition at the local level. In Baoting 保亭, a minority region in Hainan, a local court has invoked Cultural Thought to fuse modern jurisprudence with ethnic traditions, creating a culturally hybrid judicial framework.Footnote 46 Municipal bodies such as the Chengdu Audit Bureau have elaborated an “audit culture” in the name of Xi Cultural Thought.Footnote 47 These localized expressions underscore the rhizomatic principle that any actor, starting from any point, can generate new linkages. As Yayoi Kato notes, ideology possesses both legitimating and dividing functions: it unites actors through shared vocabularies while simultaneously enabling differentiation through interpretation.Footnote 48 The rhizomatic structure of Xi Thought permits both effects, depending on the lines actors choose to pursue.
Across these levels, a consistent logic is visible. Different actors begin from diverse positions, navigate different paths, and arrive at different destinations. As Deleuze and Guattari put it, the rhizome “establishes connections between semiotic chains, organizations of power.”Footnote 49 Each reading of Xi Thought, no matter how minor, constitutes an act of creation. Infinitesimal differences accumulate, generating new forms. Over time, some lines are intensified and others extinguished. This is what Deleuze and Guattari call “aparallel evolution”: the emergence of new constellations through distributed, immanent differentiation.Footnote 50
Rather than a static doctrine, Xi Cultural Thought is best understood as a generative map, that is, a structure continuously actualized through new connections. While core elements may be traced, their significance arises through mappings that actualize different connections. The formation of Xi Cultural Thought, like the formation of Xi Thought itself, was a creative act: a cartographic intervention that opened up new terrain for ideological elaboration. These openings do not simply reflect decentralized appropriation; they are moments of mutation, where fragments acquire new identities and functions through unexpected linkages. It is in these moments of differentiation that Xi Thought evolves, not despite attempts to fix meaning but through them.
Conclusion and Analytical Implications
The rhizomatic nature of Xi Thought has several implications. Most importantly, it shifts the leading question from “What does it mean?” to “What could this text, fragment or concept do?” What possibilities does it generate, foreclose or rearrange? This demands a rebalancing in epistemological and methodological orientation, from a hermeneutics of meaning towards a functional cartography of where recombinations acquire institutional force. As a governance technology, the rhizomatic nature of ideology suggests a methodological bridge between interpretive and relational analysis: not just reading texts for meaning but tracking how they link, mutate and reassemble in networked form. For scholars, this retools familiar skills: close reading remains indispensable, yet its purpose shifts from adjudicating doctrinal truth to charting the institutional life of these fragments: who cites them, in what venue and with what administrative bite.
This also means that historical tracing is of limited use. Concepts may have genealogies, but their function within Xi Thought is determined by their current linkages, not by their origins. Meaning does not simply flow from past to present; it assembles in the present through recontextualization. For the analyst, this demands reflexivity. As Ricoeur writes, the text “may be viewed from several sides, but never from all sides at once.”Footnote 51 Interpretation begins from a point, not a totality. Given that Xi Thought can only be mapped, the analyst’s task is to chart fragmentary itineraries through a shifting field of connections. The challenge, then, is to recognize that Xi Thought is always only momentarily fixed at the point at which the assemblage emerges. It is fluid in how its individual elements link, acquire new inflections and are redeployed by different actors. Analysis must, therefore, be attuned to moments of composition and not just continuity, focusing on relations and linkages, while staying attuned to the deeper, accumulated meaning of ideas.
This has implications for how to approach ideology in China. The sifting through dozens of documents to stay attuned to and understand shifts through interpretation remains important. However, Simon Leys’ famous dictum, which requires the analyst to command the Chinese language, absorb bureaucratic minutiae and decipher Party jargon, is no longer sufficient.Footnote 52 Rather, it has become a precondition for analysis. Hermeneutical analysis needs to be rebalanced towards network analysis. Xi Thought resembles a wall of vines, where “everything must be disentangled, nothing deciphered.”Footnote 53 The point is not to decode a stable truth but to map the evolving relations and effects that texts, fragments and concepts produce. This requires methodological openness and political sensitivity to what different actors do.
Second, approaching Xi Thought as a rhizome has implications for ideological governance. Its fragmented form enables continual recombination. Certain nodes gain prominence and then recede, only to be reactivated later. Some domains develop with greater density and intensity, while others remain sparse. This unevenness is not an error but a feature. It allows the ideology to expand, contract or reorient, as needed. Here, Deleuze’s concept of control societies is useful: authority no longer disciplines from above but modulates through continuous calibration. Xi Thought functions similarly. Its modularity means that different elements can be emphasized at different times, depending on circumstances, and operationalized through continuous recombinations. Texts do not simply convey meaning: they do things. This also opens a field for institutional analysis. Rather than assuming top-down coherence, one might examine how actors at different levels interpret available fragments, what lines they activate and what kinds of institutional effects these mappings generate.
Rather than Xi Thought becoming meaningless because it could mean anything, Xi Thought can be made to do almost anything, provided it is recombined within recognizable Party grammar. Understanding Xi Thought as a rhizome does not dissolve its power. Rather, power is redistributed across sites of implementation, where it acquires force at the point where fragments are embedded in institutional practices. At this point, it becomes clear that Xi Thought is participatory in a specific and non-voluntaristic sense. It does not circulate as a finished doctrine to be received and interpreted but as a set of fragments that demand continual elaboration, recombination and application. Participation here does not denote deliberation or consent but compulsory ideological labour, distributed unevenly across actors embedded in sharply hierarchical institutional settings. Cadres, judges, editors, SOE managers and scholars are required to work with Xi Thought: to cite it, embed it in justifications, align it with institutional tasks or translate it into evaluative criteria. This distributed participation is what sustains ideological coherence. At the same time, this mode of participation introduces indeterminacy. Because recombination occurs across heterogeneous sites under asymmetric constraints, the same fragments can be mobilized in divergent ways, generating effects that are not fully predictable. Participation thus stabilizes Xi Thought institutionally while simultaneously multiplying its possible trajectories. In this sense, unity is not imposed from above but is continuously produced through distributed acts of ideological production. It is, therefore, always provisional, reversible and contingent on ongoing participation.
Third, the rhizomatic nature of Xi Thought also imposes limits. Since 2022, the pace of new elaborations has slowed. At the local level, a sense of fatigue is coupled with increasing pressure to discipline thought. This raises two related concerns. First, what happens when ideological work becomes routinized and excessive? If every concept can be justified ad hoc and retroactively, does the system risk stagnation? As Mittelstaedt notes, this may result in ritualized governance, where ideological labour is performed less for conviction than for compliance.Footnote 54 Lisa Wedeen points out that belief is harder to trace than unbelief, and in the case at hand, actors could be performing “as if” they believe.Footnote 55 By contrast, Alexei Yurchak suggests that the binary between real belief and performative dissimulation is irrelevant and argues that the significance of ideological practice lies not in intention but in effects. His example is the courtroom oath: whether one intends to lie or tell the truth, the oath binds one to a set of institutional consequences.Footnote 56 Belief becomes irrelevant to function. So, too, with Xi Thought. Participation, citation and elaboration generate effects, regardless of belief. In this sense, what matters is not internal conviction but the binding force of ideological form. To invoke Yurchak’s logic, the repetition of Xi Thought, including its invocation in slogans, speeches or policy documents, performs its own reality. This is especially true given that ideology in China is increasingly organizationally enforced.
This brings us to Louis Althusser. In an interview, he claimed that “Marxist philosophy does not exist, and it cannot exist.”Footnote 57 Likewise, Xi Thought does not exist as a coherent, stable entity. What exists are fragments, the identity and meaning of which are produced only through the connections made in a particular moment. Once formed, unities dissipate. Xi Thought, then, exists only in the millisecond of its composition. In this narrow sense, Xi is “dead” to his Thought. He is responsible for it, but he is not in control of it. It escapes his intent. To impose a coherent authorial will onto the system is to “impose a limit on the text.”Footnote 58 The work of ideology proceeds not from a sovereign author but from the widely distributed labour of those who cite, interpret, implement and reconfigure. Xi Thought is not a book to be read; it is a map to be walked and redrawn.
Acknowledgements
The author would like to thank Olivia Cheung, Patricia M. Thornton, Chloé Froissart and the other participants of the 2025 “The fabric of CCP: ideology, circulation, impact, and resistance” workshop at INALCO in Paris
Competing interests
None.
Jean Christopher MITTELSTAEDT is professor of modern Chinese studies in the Institute of Asian and Oriental Studies at the University of Zurich, where his research explores the ideological and institutional foundations of the Chinese party-state’s rule.
Open access
