Global Religious History as a Rhizome: Colonial Panics and Political Islam in German East Africa

A Global History of Religion aims to trace connections, controversies, and contingen-cies in the emergence of “religion” as a global category. Its main intention is to de-center European epistemologies of religion by drawing out a more intricate global and plural genealogy. This is a very complex endeavour, however, especially when one leaves the realm of academic debate and considers the quotidian understandings of “religion” emerging in colonial encounters. Here one is often confronted by vast entanglements of practices, perceptions and politics, which need a historical methodology that foregrounds the plurality, complexity and historicity of all religious epistemes. Drawing on Deleuze’ and Guattari’s philosophical figure of the rhizome, this article sketches such an approach in a conversation between theory and historiographical practice, as it maps out a particular episode in the construction of “political Islam” in German East Africa.

guardian of the Prophet's tomb, had been found in several places. The letter urgently warned all Muslims to return to a more faithful religious practice. The holy duties of Islam shall be observed again with diligence, one must refrain from fellowship with unbelievers, and the reader was informed that this was the last warning before the imminent day of judgment, when the "sun would rise in the West." Although the missive, commonly referred to as "Mecca letter" (erroneously so, given that the Prophet's tomb is in Medina), lacked any stated political intentions, its millennial undertone was understood in German colonial discourse as a deliberate attempt to incite a religiously motivated rebellion. This fear of a politically disruptive potential of the letter was exacerbated by mosques filling up, religious practice intensifying, women attending prayer, and concubines ceasing their services to German colonial officials and African soldiers.
Until then, Muslims had largely been framed in German colonial discourse as politically inconspicuous and docile. Only missionaries had warned repeatedly of an impending Islamic threat, typically in combination with criticisms of Germany's "Islam-friendly" policy and calls for an official promotion of missionary work. Such interjections were roundly rejected by the government and heavily attacked by the settler press. The latter even framed Christian missionaries as a danger to colonial rule themselves, because of their attacks on Islam and their promotion of "free-thinking" individuals, which would only lead to "Ethiopianism," while the "this-worldliness" of Islamic religion and ethics was much closer to "Negro mentality" and more suitable for building a "work ethic" for colonial exploitation (Haustein 2018).
It seems that with the "Mecca letter" these perceptions began to change. The colonial government quickly found a culprit in Rumaliza, a former slave trader who now resided in Zanzibar, and began to crack down on suspected "Islamic agitators." Though these assertions remained unproven and the feared Muslim uprising never materialized, the "Mecca letter affair" of 1908 developed a life of its own. It now became a reference point for a certain "seditious" potentiality of Islam and the need to observe and regulate Muslim practice. Even the settler press began to concede this point to missionaries, and the specter of "Islamic danger" was now quite regularly invoked in political discourse. All of this was backed by the political opportunism of scholars like Carl Heinrich Becker, who repositioned Islamic Studies as a practice-oriented and politically useful scholarly discipline. While Becker rejected missionary generalizations about Islam, the genius of his approach lay in his proposition of a ubiquitously seditious potential of the idea of a Muslim umma, which necessitated a careful divide-et-impera Islam politics to dilute its appeal. This included proposals like the establishment of "Christian islands" in German East Africa to hinder the uniform spread of Islam (Becker 1909;Becker 1910). Much of this was in keeping with the time. The catchphrase of "Panislam" had arisen only a few years earlier, Muslim reformers were beginning to draw political attention, the Turco-Italian War and the Balkan Wars led to tensions in the German Islam policy, and the guerrilla-style resistance of the Sanūsīya to French and Italian colonizers in Libya were considered proof of the military potential of Islam. All this would eventually culminate in the German attempts to mobilize Islam in the Ottoman alliance in the First World War, with the two scholars consulted by the government on the "Mecca letter" -Max von Oppenheim and Becker -going on to become the architect and staunch defender of the German idea to mobilise the formerly contained "Islamic danger" to the Reich's advantage (Lüdke 2005).
It is therefore tempting to turn the "Mecca letter affair" into a key event for understanding the political dynamics behind Islam in the German East African colony. This is what Becker (e.g. 1909;1911) already did when he used the letter as evidence for his contention that Muslims regularly harbored Panislamic utopias that could always be activated for political purposes. More recently, Africa historians such as B. G. Martin (1969;1976: 153-176), John Iliffe (1979: 211-212), and Michael Pesek (2000;2002;2003) have used the letter as evidence for the contemporary spread of the Qādirīya and the socio-structural changes this engendered, based on the observation that some specimens of the letter were distributed by sheikhs from this Sufi order. Others have sought to lend the "Mecca letter" more contextual weight, like Felicitas Becker (2010), who saw it as a belated expression of the same millennialist hopes that had been invested in the Maji Maji rising, or Jennifer Kopf (2007), who framed the noticeable participation of women as an implicit female challenge to the masculine order of German East African colonial rule. My own inclination would be to frame the letter as a turning point in the German debate about Islam, which, for reasons that will become apparent in this article, had little to do with events in the colony (see also Haustein 2020a).
The problem with all of these interpretations, including my own, is that they overdetermine the event in order to endow it with a specific meaning, whereas the sequence of actors and interpretations involved will always exceed such monolithic frames of reference. Political utopias, religious revival, Sufi practices, female participation and colonial debates formed a complex and plural entanglement of historical genealogies, contemporary motivations and subsequent interpretations, which offer a rich quarry for historians to anchor their theories and historical trajectories. Yet this extractive mode of historical interpretation will always tend to sideline or eclipse aspects of the event that do not fit the story and thereby unavoidably arrest the multiplicity and ambivalence of a historical situation in the interest of a grand narrative. This prompts the question whether it is possible to develop a mode of historical analysis that is able to retain the multidimensionality and ambivalence of historical events like the "Mecca letter affair" in the interest of continually mapping out the complex interaction of multiple factors, intentions, and interpretations at work. This does not preclude historical narratives from tracing out causes, actions, and consequences, but none of these will be uniform, linear, or indeed final.
The present article seeks to sketch out such an approach by drawing on the figure of the rhizome as developed by Gilles Deleuze and Felix Guattari. The analysis will begin with some theoretical observations about the challenges that pose themselves when seeking to understand the construction of "political Islam" from the colonial source archive and show why the philosophy of the rhizome might offer a way to overcome these challenges. The remainder of the article will then follow the principles of the rhizome as laid out in Deleuze' and Guattari's introductory essay to A Thousand Plateaus (1987: 3-25) and apply them to the historical analysis of the "Mecca letter affair." The material and insights necessary for analysing this particular episode of German colonialism have been drawn from a recently completed, though not yet published monograph by the author (Haustein 2020b).

2
Overcoming the Duplicity of Postcolonial Historiography: "Religion 2" and "Pluralicity" When studying Islam in the colonial age, one easily engages in a postcolonial double play. On the one hand, the ideological transparency of the colonial archive readily prompts historians to deconstruct the articulated concepts of Islam as expressions of the colonial order of knowledge and its corollary strategies of domination. Concepts of Islam are thus historicized in the colonial encounter and possibly interrogated with regard to their influence on the (mis) understanding of Islam today. On the other hand, the primary interest for such a study is typically one of contrasting ideology with reality, and it is here that implicit and subtly normative concepts of Islam creep back in. How did colonial knowledge systems correspond to the "self-understanding" of contemporary Muslims? How did administrators, settlers, or missionaries misinterpret "lived" Islam, and how did their misunderstandings, categorizations, and political regiments influence and change this already constituted religion? In other words, studies of colonial Islam are subject to an epistemological oscillation: the knowledge category of Islam is being deconstructed and imposed at the same time. A recent example of this problem is Sebastian Gottschalk's (2017) otherwise insightful study of German and British colonial approaches to Islam in West Africa. In his introduction, Gottschalk notes that the colonial differentiation between Muslims and non-Muslims was perturbed by an "interference and entanglement between the categories of 'race' and religion," which made it difficult to sort the population according to a "European matrix" (Gottschalk 2017: 15). Nonetheless, he speaks of "Islamic actors" throughout his study, even though his analyses rely almost entirely on colonial texts which presumable operate precisely within this misplaced "European matrix." Thus, the colonial archive is used to deconstruct and reify Islam at once: it is framed as an epistemological problem and as a collective category, all in the same source analysis. A global history of religion seeks to avoid this double-play by foregrounding the genesis and deployment of religious concepts and identifiers in its analysis. As Michael Bergunder (2014) has laid out in his seminal article "What is Religion?" in this journal, the main aim of religious studies should be a genealogical reconstruction of the "unexplained religion" which lies underneath all attempts to define and conceptualize a religious entity or the category of religion as such. This "unexplained religion" ("Religion 2" in Bergunder's terminology) is the "contemporary, everyday understanding," which forms the "undisclosed subject matter of Religious Studies" because it provides the condition of plausibility for all definitions and categories of "explained religion" ("Religion 1") in scholarly discourse. Academic study, therefore, should not be occupied primarily with "explained religion" (such as discussing definitions or seeking to abolish the category of "religion"), but should reconstruct the genealogies of these quotidian, "Religion 2" notions in all their fluidity and contextuality. Drawing on Foucault, Laclau, and Butler, Bergunder provides a theoretical foundation for this "historicising of 'religion' ," but also admits that this is the "most complex and difficult" of the questions he poses.
For the study of colonial Islam, this call to pay heed to the quotidian understandings of religion means, first and foremost, that scholars should no longer try to look through the tainted glass of colonial constructions of Islam, but look at it as the very surface on which "Islam" appears. It is these colonial, quotidian constructions of Islam upon which all contemporary academic characterizations of Islam rest and which still form the background and source archive for any historian's investigation into the subject matter of colonialism and Islam. Therefore, instead of following the track laid out by the colonial archive in identifying certain actors or motivations primarily as Muslim, the inquiry looks at how this specific demarcation of Muslim identity emerged in everyday practices and perceptions and attained its plausibility for interpreting political events.
Focusing on the deconstruction of these colonial constructs of Islam does not mean to discount "native agency" or alternative concepts of Islam outside of the colonial gaze. (In fact, it is the relevance of the colonial perception of Islam for contemporary life in the colonies that stands in question!) Yet it does mean to take seriously the limitations of the colonial archive. In many cases, we are only able to reconstruct from the colonial sources those contemporary quotidian understandings ("Religion 2") of Islam that have informed the political, missionary, and academic attempts to describe, regulate, and influence the presumed phenomenon ("Religion 1") of Islam. Yet, as Dipesh Chakrabarty (2000) has shown, this does not mean to ignore concepts, practices and agencies outside of the colonial archive, it only means that one needs to resist the temptation to capture their essence from the mere traces they leave in the archive. Instead, it suffices to study how the epistemic violence of the colonial logic and teleology ("History 1" in Chakrabarty's terminology) is constantly interpunctuated by identities and agencies ("History 2s") that the colonial imposition needs to subject in order to instantiate its program but that it nonetheless fails to capture because they are not forward-looking to its logic. Therefore, in order to overcome the overbearing weight of the colonial archive and follow Chakrabarty's program of Provincializing Europe, it is important to read them as disruptions and not seek to reconstruct them as transparent reactions to or affirmations of a particular identity the archive proposes.
In other words, if the German archive speaks of an unexpected attempt to stir Muslim resentment against colonial rule with forged letters from the Prophet's tomb, which upended the perception of "docile" East African Islam, then one is confronted first and foremost with the breakdown of idiosyncratic colonial categories and interpretative routines rather than the sudden emergence of political Islam from outside of German discourse. For the notion of "political Islam" is itself a "Religion 1" category arising from the interpretative strategies of colonial debates and record-keeping, and the neat distinctions it draws between political and apolitical Muslims are precisely what breaks down in the interpretation of the "Mecca letter affair." Conversely, one must take seriously the absence of the voices of those implicated in the "affair." The failure of the colonial archive to come to terms with their unexpected agency is a "text-inscribed blankness" that needs to be rendered as the "quite-other" without assimilating it (Spivak 1988: 294). Where direct representation is absent, "political Islam" emerges first and foremost as an enigma of colonial interpretation.
So far, the attempt to avoid the criticized double play of simultaneously deconstructing and presuming "colonial Islam" has led to a simple and equally unsatisfactory juxtaposition of the accessible "colonial archive" and the inaccessible "other". And here post-colonial analysis quickly finds itself in a strange double bind. The effort to deconstruct colonial categories nonetheless continues to foreground them, the critique of colonial agencies continues to center the historical narrative around them, and the sheer weight of the colonial archive keeps asserting itself despite (and even because of) sources and situations that are read against its perceptions. We must push the analysis further and this means to work on dissolving the integrity and authority of the colonial gaze, to decenter its agency in historical events and the historical inquiry. Along this path, current ontological certainties -like the analytical object of Islam, of colonizers, or of politics -will recede back into the epistemology from which they emerged.
Seeking to overcome the epistemological limitations of the area studies approach in an age of globalization, Eva Spies and Rüdiger Seesemann (2016) recently coined the term "pluralicity" as a mode of analysis that foregrounds plurality as a theoretical choice. This means that more than simply acknowledging the empirical fact of diversity, one must account for "the existence of complex structures that do not necessarily refer to any given, prior entity" (Spies et al. 2016: 135). As a result, the analysis should avoid "master categories" that "deny difference and incommensurability by subordinating the multiple to the one," such as religion, Islam, Muslim actors, politics, or colonizers in the interest of mapping out the "entangled structure of the world and the plurality of lifeworlds" (Spies et al. 2016: 135). Dichotomies like center/periphery, colonizer/colonized, or global/local are thus dissolved into a multiplicity of relations which do not hearken to one particular dynamic and in fact escape any attempt to contain them in a bounded form.
Spies' and Seesemann's approach draws its inspiration from Gilles Deleuze' and Felix Guattari's figure of the rhizome (Deleuze et al. 1987: 3-25), which is also at the base of the analyses offered in this article. It is important to acknowledge at the outset that Deleuze' and Guttari's project is fundamentally a metaphysical one, but one that seeks to do away with metaphysical ontology, i.e. that avoids to be anchored in a "transcendent principle and instead seeks purely immanent principles" (Adkins 2015: 3). As such, Deleuze and Guattari position their model explicitly against Chomsky's tree, against the generative method of filiation, against a philosophy that is anchored in the unit, in oneness, in metaphysical identity. Instead, Deleuze and Guattari affirm radical continuity: in this flat ontology everything is embedded and relational, there is no outside, no "other." In the context of the analysis pursued here, this means to deny any object identity to Islam, native religion, colonialism, politics, and so on. These are labels, at best, that emerge in the context of a variety of interconnected practices, which in turn by far exceed the application of these label and disperse far beyond specific debates. This, then, is the surface of colonial Islam that emerges in a Deleuzian perspective: not the study of an object of Islam, but a flat ontology of interconnected practices and institutions that continuously produces and reabsorbs identity postulates, categorical specifications, and epistemological quests through which the object of Islam appears.
In what follows, the article will probe the usefulness of the figure of the rhizome for developing further the idea of "pluralicity" in the study of colonial Islam. We will follow the principles of the rhizome that Deleuze and Guattari posit and map them onto the complex entanglement of the "Mecca letter affair" and the corollary notions of Islam. These applications will allow us, firstly, to make important observations about the heterogeneity of the German colonial debates about Islam; secondly, to reinscribe subjectivity and agency into this complex entanglement; thirdly, to query the notion of an event as a significant rupture; and, finally, to reframe historiography as a mapping and tracing that is oriented toward the present.

Connection and Heterogeneity: The Rhizomatic Emergence of the "Mecca Letter Affair"
Deleuze and Guttari establish six principles of the rhizome, some of which are discussed in pairs and others individually. The first and second principle are connection and heterogeneity. This means that each point in a rhizome derives its essence from its differential relationality to other points; identity becomes a topographical effect. In a flat ontology, this of course applies to the world of articulated objects as well as to articulating subjects, both are essentially nodal points in rhizomatic structures. Each statement, therefore, is part of an "abstract machine that connects a language to the semantic and pragmatic contents of statements, to collective assemblages of enunciation, to a whole micropolitics of the social field" (Deleuze et al. 1987: 7). At its core, this is a rejection of all structures and hierarchies, of Chomsky's trees with their neat stemmatics. In the rhizome all elements are linked, but none is superordinate. All categories and codes are just elements within "collective assemblages of enunciation" (Deleuze et al. 1987: 7), and even "language in itself" is nothing but a "throng of dialects, patois, slangs, and specialized languages" (Deleuze et al. 1987: 8). In short, the rhizomatic method does not form hierarchies or categories, but rather decenters each element by reinscribing it into its relationships with any number of other elements. For the analysis of the "Mecca letter affair," this rejection of hierarchies and trees means that any invocation of root causes, core dynamics or primary social entities is to be rejected. Contrasts between colonizers and colonized, Sufi and orthodox Islam, Muslims and Christians, or Africans, Arabs and Europeans lose their analytical value outside of specific entanglements and interactions. Accordingly, the analysis endeavors to trace the network of relationships that enables the distribution, discovery, and interpretation of the "Mecca letter." Each role and person in this network derives its value from its relation to others, no authority is absolute or self-sufficient.
This can easily be shown when mapping out how the letter became a political object. The first mention of the "Mecca letter" in the archive originated with the district officer of Lindi (Tanzania), Walter Wendt, who immediately interpreted them as "missives of pan-Islamitic tendency [grossislamitische Tendenz]" and sought permission to "annex" the post of the local qāḍī (Wendt to Dar es Salaam, July 26, 1908, TNA G 9/46, f.1). Having just arrived in the district earlier that year and at the apex of a fairly steep colonial career, Wendt had enough reason to sound the alarm. He had served in the German colony in subordinate positions since 1898, which meant that his elevation to provisional district officer of Lindi was a significant step in his career. The post was a difficult one, though. While the district was south of the main arenas of the Maji Maji War, it had been affected severely by the War and its aftermath. Tax income had dropped by about twenty percent, and still in 1907 the district reported around 1,500 deaths of starvation per month (Bald 1976: 29, 44). Wendt's predecessor had also been a provisional appointment, and after lasting merely a year in office had been discharged dishonorably (Franz et al. 1973: 103, 180). Therefore, Wendt could hardly afford to overlook any risks, but had much to gain if he appeared energetic in securing the district and avoiding further conflict. He took draconian measures against those implicated in the "affair" and set out to prove -with wild conjectures and ultimately unsuccessfully -that the instigators behind the "Mecca letter" were also behind the Maji Maji rising. Governor Rechenberg, in turn, took a more cautious approach and at first sought to dampen Wendt's enthusiasm and rebuked his tendency to overreach. Yet he was quickly driven into a further investigation by newspaper leaks and pressure from his superiors in the Colonial Office, who were not amused to learn first of the matter from the press. Despite his scepticism toward Wendt's conjectures, Rechenberg adopted his conclusion that Rumaliza was the culprit behind the letter and disseminated this "insight" to all district offices with the instruction to keep an eye on similar movements (Runderlaß Rechenberg, August 12, 1908, BArch R 1001/701, f. 64-72). Invariably, this unearthed further letter specimens and culprits, now collected in a newly formed file, titled "religious movements" (TNA G 9/46). At the same time settler press increased the pressure as it began to use the "Mecca letter" as an indictment of Rechenberg's supposedly permissive stance on Indian and Arab immigration (e.g. Perrot 1908). This was part of a long-standing campaign against Rechenberg's economic policies, which were met with strong settler opposition as they aimed to boost growth and tax revenue through petty trade and native farming rather than cash-crop plantations and coercive wage-labor regimes (Iliffe 1969). For a while, Rechenberg's instructions and the alarmism of the press created a climate of paranoia, leading to a strict observation of Muslim "deviations", with minor incidents now prompting harsh reactions and reports to Dar es Salaam.1 Soon the Muslim establishment began to utilize the German nervousness to rid themselves of unwanted elements, like the Sufi preacher and miracle worker Zahur bin Muḥammad al-Ǧabrī, exiled by the colonial government in late 1908. This, in turn, put Sufism on the map for the German government, and once again, Rechenberg followed Wendt's conjectures, who now associated the "Mecca letter" with a "neo-Islamic movement" from Barawa. Arguing that this new movement used ḏikr to arouse irrational behaviour, Rechenberg banned Sufi practices from the public sphere (Runderlaß Rechenberg, February 15, 1909, TNA G 9/46, f. 164-180). At the same time, Rechenberg was concerned about losing the support of "moderate Muslims" in this increasingly confrontational climate and requested the assistance of the Colonial Institute in Hamburg for acquiring a reference library on Islam to educate his district officers. This in turn activated Carl Heinrich Becker, newly appointed Professor of Islamic Studies at the Institute, who not only sent (unsolicited) advice to Rechenberg, but produced a transcription, translation, and interpretation of one of the letter's specimens in his writings. Becker's text became the reference point for all subsequent research into the "affair" (Becker 1911: 43-47; for a critical reappraisal, see Haustein 2020a). This vaguely sketched entanglement of different interests and interpretations is what produced the "Mecca letter affair", and one could trace out in much greater depth this rhizomatic ensemble that stretched from Lindi, Tabora, Bagamoyo, and Dar es Salaam to Hamburg, Berlin, and even Istanbul (Haustein 2020b: 270-340). Yet it should already be clear from this initial overview that the discovery and interpretation of the "Mecca letters" were not driven by a single interpretative authority or agency, but an interconnected web of heterogeneous interests and pressures that turned them into a fullblown "affair". And even this complex colonial arrangement is still a rather incomplete picture, since the views of those spreading the letter are only available in very few interrogation minutes and reports, with all methodological caveats this genre implies for their interpretation. The motivations behind copying and distributing the letter were likely as diverse as those of the colonial interest groups reacting to it, and it is clear from textual evidence, that the German authorities only confiscated a fraction of the letters that were spread in the colony (Haustein 2020a).
This web of interconnected interests, pressures, and perspectives, therefore, does not amount to a superordinate authority or singular dynamic of "political Islam" producing the "Mecca letter affair" -neither in native agency or nor colonial perception. Instead, we are faced with a colonial entanglement, a "machinic assemblage" in Deleuze' and Guattari's terminology. It is this "machine" that a rhizomatic analysis seeks to map out by tracing its interconnected heterogeneity. Therefore, as Spies and Seesemann (2016) have already highlighted, taking plurality as a point of departure is more than simply emphasizing diversity and complexity, but it is the refusal of singular points of origin, of simple dichotomies, and of unidirectional trajectories. Rechenberg and his district officers are just as representative of the colonial machine as the settler press, missionaries, and local ʿulamā using the German panic to oust Sufi preachers. Certainly, power was not evenly distributed within this machine, but its flows are only legible when the various pressures acting upon the different nodal points of the rhizome are mapped out.

Multiplicity: The Productive Non-Identity of a Chain Letter
The above observations about the colonial assemblage significantly erode the notion of subjectivity, which instead becomes an effect of the heterogeneity of rhizomatic connections. Deleuze and Guattari underscore this observation in their third principle of the rhizome, which they call multiplicity. A multiplicity, they note, "has neither subject nor object, only determinations, magnitudes, and dimensions that cannot increase in number without the multiplicity changing in nature" (Deleuze et al. 1987: 8). An example of this is the puppet on a string, which is not connected to the will of the puppeteer, but to a "multiplicity of nerve fibers, which form another puppet in another dimension connected to the first." (Deleuze et al. 1987: 8). And one could continue to trace the connections of this nerve rhizome to the dimension of nutrients and the ecology of their procurement, or to hormones and the equilibria of biochemistry. With every step along the rhizome its dimensions increase and yet there is no outside, no unity that embraces and encloses the totality of the rhizome. That is why rhizomes are flat, they "fill or occupy all of their dimensions," (Deleuze et al. 1987: 9) any supposed outside, any "line of flight" is always reabsorbed: "There are no points or positions in a rhizome, such as those found in a structure, tree, or root. There are only lines" (Deleuze et al. 1987: 8). These observations are reminiscent of the "material turn," which similarly undercuts the anthropocentric object-subject distinction and traces connections and effects that are generated by material constellations that absorb and exceed human bodies, perceptions, and practices. For our example of the "Mecca letter," the principle of multiplicity means to spell out the consequences of understanding this missive no longer in terms of subjects and objects, but as a human-text machine, in which scribes, text, and reader produce one another. Original objects and subjects disappear in this machinic assemblage, which instead produces subjects and objects along its sprawl. This becomes very clear when studying the putative "object" of the "Mecca letter." There is no single authoritative version of the letter, but rather a rhizomatically proliferating text. Becker may have canonized a certain version of the letter but when returning to the hitherto neglected fourteen Arabic copies still extant in the colonial archive, a rather different picture emerges. In a historical-critical edition of these letters I have recently documented the variety of textual variants that ranges from orthographic variants and copyist addenda to theological corrections and copying errors with serious consequences (Haustein 2020a). The most consequential of these alterations was a longer omission owing to a haplography, which separates the extant letters into two main versions with further sub-groups. Yet it is impossible to draw up a neat stemmatic diagram of the letter versions from the available archive, because the haplography appears to precede the East African circulation and there are unexplainable cross-effects between both main versions. Chomsky's trees fail to reassert themselves.
Moreover, both main variants were sometimes found in one town without any attempts of reconciling them in a unified text. This indicates that contrary to the German endeavor to identify an original "culprit" behind the text, the letter's dissemination did not follow a simple model from original to dependent copies, but was disseminated in several versions by different groups at the same time. This eliminates subsequent attempts to tie the letter to one specific political intention or group dynamics. The versions spread by Rumaliza, whom the German colonial government had implicated, form a textually distinct corpus of only four of the fourteen letters in the archive. And contrary to Martin's and Pesek's above-mentioned contention that the Qādirīya was the vehicle of the letter's spread, the specimens connected to sheikhs from this ṭarīqa (Sufi order) are only two, both of which contain a theologically significant correction that is not echoed by the other copies (Haustein 2020a).
Even more important for understanding the "Mecca letter's" rhizomatic sprawl is the fact that missives with very similar content have been found in a variety of temporal and geographical contexts: North Africa and Iraq around 1860, Cairo in 1877, Indonesia in the early 1880s, among Indian soldiers in World War I, and even as an E-mail forward mentioned in a chat forum from 2005 (Hurgronje 1923;Katz 1994;Cole 1999;Singh 2014;Islamic Board 2005). The textual deviations are considerable in some cases, but the structure and content of the message are the same, as is the name and role of its fictitious author in most cases. This suggests that the East African "Mecca letter" was nothing more than a local circulation loop of a global chain letter that long preceded and outlasted the events of 1908.
A chain letter is of course a rhizome par-excellence, a "machinic assemblage" of letters, readers, scribes (and again letters), that is extremely productive in its autopoiesis. This rhizome, firstly, continuously produces the "Mecca letter" as a material and ideal object. Fortified with triple blessings and curses it instructs the reader to produce further copies, resulting in an exponential reproduction with multiple textual variants, including translations and vague copies penned from memory. Yet at the same time, all copies invoke an original, and each specimen accompanies this invocation by a solemn oath to not have deviated in the slightest from the received copy, an oath which, ironically, is traditioned in different textual variants. Beyond these text-immanent invocations, the rhizome of the Mecca letter gives rise to the postulate of an original text, a first letter, which was pursued in colonial investigations and subsequent scholarship. Even when this Urtext postulate is rejected or disregarded as in my analysis, one still needs to come to terms with the frayed ideal object of the "Mecca letter": Which variants are still part of this chain letter rhizome, and how far does one trace the connections to other millennial prophecies and warnings in Islam or beyond? Thus, the object of the "Mecca letter" arises within the entanglement of its copies. Its identity is not superordinate to the rhizome, but inscribed and read out.
Secondly, the "Mecca letter" rhizome also produces subjects. This is, first of all, its fictional author, guardian of the prophet's tomb with the suspiciously generic name "šayḫ Aḥmad." Without this putative authorial subject, all declarations of authenticity would be ineffective. Furthermore, the Mecca letter produces subjects in the signatures of the copyists who may appear anonymously ("I have reproduced everything exactly as I saw it") or as full names fortified with titles, place and date in lending their weight to the letter's authority. These copyists, as well as those accused of the letter's dissemination, also appear as subjects in the colonial investigation when hitherto invisible inhabitants of German East Africa and Zanzibar become corporeal subjects of the colonial state apparatus through interrogation minutes, incarceration notes and "most wanted" posters. These all are subject effects of the "Mecca letter" rhizome.
Finally, the "Mecca letter" rhizome is productive by continuously sprawling into new contexts, like Derrida's (1982) marque. It became a short-hand for "Islamic danger" in the subsequent colonial debate, and an anchor point for discussing religious change in later scholarship. Here, it is an object of methodological reflection, but all of these are an expression of the "Mecca letter's" rhizomatic multiplicity that knows no unity, superordinate object or original subject. Instead the "Mecca letter" rhizome produces analytical units, forges textual objects and disseminates subjects in its sprawl. Its implications for "political Islam" thus reside within an entanglement that proliferates between 1860s and 2005 (or 2021), between global and local circulation processes, between copy upon copy, between scribe and text, between finding letters and their analysis.
The main consequence of such a foregrounding of multiplicity for the analysis of "Religion 2" is thus apparent. Historical scholarship no longer centers its analysis around putative actors, objects, or dynamics, but on the emergence of said actors, objects, and dynamics in rhizomatic assemblages that are productive in manifold dimensions and contexts.

5
Asignifying Rupture: On Novelty and Development "The rhizome is productive," is a problematic contention, however, because the rhizome is not a subject either. The "Mecca letter" rhizome does not aim to produce "Mecca letters," copyists, suspects, rebels, or analytical effects in politics and scholarship. It just proliferates. This proliferation is total and without direction, as Deleuze and Guattari explain with their fourth principle, the principle of the asignificant rupture. An interruption of the rhizome at any one point only leads to a continuation at another; breaks get reabsorbed. Like an ant colony disturbed by a fallen branch, the rhizome reacts to deterritorialization with reterritorialization. Both processes, deterritorialization and reterritorialization, belong to the life process of the rhizome. The nectar-less orchid imitates the image of a wasp to attract drones to pollinate. The orchid is thus deterritorialized, so that it can reterritorialize the wasp on this adapted image. The wasp, in turn, is deterritorialized as a "piece of the orchid's reproductive apparatus" (Deleuze et al. 1987: 10), but it also reterritorializes the orchid pollination in its patterns of flight. Their heterogeneity forms a rhizome, which absorbs disruptions on both sides. Deleuze and Guattari call such entanglements an aparallel evolution: not imitation, reproduction, mimicry or other metaphors of original and copy (or independent and dependent variable), but entanglements that necessarily belong to the sprawling of the rhizome and constitute its elements. This insight has two main consequences for discussing events in historiography. Firstly, instead of studying how interest groups, factors, or dynamics are acting upon an event, one should focus on how they are constituted or reconstituted in the entanglements of said event. Secondly, the notion of events as significant ruptures, bringing fundamental transformation or novel developments, needs to be transformed via the recognition of previous entanglements that are merely rearranged through an asignificant break. The temptation to locate something completely new or a fundamental break in a historical event must be resisted. Rhizomes are diachronic. They change and proliferate, but never cease to exist, break apart or appear out of nowhere.
It therefore becomes imperative to locate the "Mecca letter affair" of 1908 within the wider history of the colonial engagement of Islam in German East Africa. This is a history of vicissitudes. The German occupation of Tanganyika mobilized an enemy construct of "Arabs", variously characterized as religious fanatics, an erosive Semitic influence, or cruel slave traders, with the last of these tropes being particularly useful for creating a public platform across different German interest groups (Haustein 2017). Yet after the suppression of the "Arab revolt" that had marked the beginning of the German conquest, the rhetoric of the government and settler associations had turned by 180 degrees: "Arabs" were now characterized as a useful cultural mediator between "Negro primitivity" and "high-standing" European culture, and Islam was seen as a particularly useful religion for colonizing due to its "fatalistic" and "this-worldly" orientation (Haustein 2017;Haustein 2018). A "religiously neutral" system of government schools was developed for the education of Muslim children, which led to most subaltern government positions being filled with Muslims, and Swahili, the language of the former "Arab" trade economy, became the lingua franca of the colony. All this was much to the chagrin of missionaries, whose call for a "Christian civilizing mission" was rejected by other stakeholders of colonial rule, leading to considerable political isolation in their reliance on vernacular languages and their educational efforts that offered no clear access routes to colonial service. When missionaries began to attack the German colonial system as "Muslim-friendly" and underestimating the "danger of Islam," this was either rebuffed with attacks on missionary failings to properly "civilize" Africans or with general treatises about the value of religious neutrality to German colonial politics (Haustein 2018). By characterizing East African Islam as "docile" and politically neutral, German colonial politicians were able to profile their program as religious neutral -forming a colonial-political rhizome of deterritorialization and reterritorialization of Islam and colonialism.
The "Mecca letter affair" acted as an asignificant rupture on this rhizome, a rupture, inasmuch as the rhizome needed to rearrange, but asignificant because the aparallel evolution of colonial politics and Islam concepts continued along the previous path of deterritorialization and reterritorialization. By reading the "Mecca letter" as a political affair, Islam was deterritorialized from the apolitical, "purely religious" corner and now required reterritorialization through a less-than-secular politics. In his first reports to Berlin about the letters, Rechenberg mused that it might be useful to have German East Africa declared as dar al-islām by the "supreme religious authority" for all Sunni Muslims in Istanbul (Rechenberg to Colonial Office, August 12, 1908, BArch R 1001/701 f. 64-71). The thought behind this was that such a recognition of German East Africa as belonging to the "House of Islam" would make it impossible for devout Muslims to rebel against the colonial government. The Colonial Office was skeptical toward this plan, but nonetheless asked the ambassador in Istanbul to investigate the matter. This led to a "strictly confidential" conversation with the Šayḫ al-Islām, who seemed open to the idea as long as Muslim practice in the colony was uninhibited by the colonial authorities and the ḫuṭba (the Friday sermon, typically including a supplication for the caliph) would be held every week as an expression of "dependence on the spiritual leader of Islam" (Bieberstein to Bülow, December 10, 1908, BArch R 1001. Inasmuch as this move deterritorialized politics from the realm of religious neutrality, it now made the government dependent on the opinion of experts of Islam, notably Becker and Oppenheim. Both advised against pursuing the dar al-islām declaration any further but formulated an Islam policy for the German colonies. Becker basically drafted his above-mentioned divide-et-impera policy, which amounted to recognizing Islam as the main religion of the East African protectorate, but only if combined with careful surveillance, a selective appropriation of Muslim diversity and measures to boost Christian missions in isolated areas in order to prevent a uniform spread of Islam (Becker to Stuhlmann, 24 March 1909, TNA G 9/46, f. 225-228). Oppenheim, on the other hand, drew a larger geopolitical picture of "Panislam" being weaponized in the competition of colonial powers. He conjectured that the "Mecca letter" was likely the result of French propaganda efforts in the aftermath of the Algeciras conference and advised a series of counter-measures: a public embrace of Muslims in the colony, even if necessarily combined with "far-sighted and smart surveillance," a stifling of further public discussion of the "Mecca letter affair," and an intensification of the alliance with Istanbul while avoiding a public show of unity through measures like the dar al-islām declaration, which would only giving rise to further French or English agitation of Muslims against Germany (Oppenheim to Bülow, July 12, 1909, BArch R 1001. It is not difficult to draw connections from here to Oppenheim's strategy paper in the early days of World War I, which now sought to wield the weapon of Panislam in German hands through embracing Ottoman religious propaganda (Epkenhans 2000). Becker was also won over to this reasoning by 1915, defending in particular the German embrace of the Ottoman ǧihād fatwā (declaring that all Muslims must join the War against the Entente) as good "realpolitik" against Snouck Hurgronje's attack of this "holy war, made in Germany" (see esp . Hurgronje 1915;Becker 1915a;Becker 1915b).
The "Mecca letter affair" of 1908 therefore was not some kind of singular or transformative event in German Islam politics. Rather, the notion of asignificant rupture allows us to see how it is embedded in a longer genealogical entanglement in which notions of Islam and colonial politics are engaged in a mutual deterritorialization and reterritorialization in the aparallel evolution of Islam and politics. This does not mean that religious history always has to be written in broad strokes of historical longevity. Rather, events like the "Mecca letter affair" can form useful microcosms to trace out such an aparallel evolution as long as they are not conceptualized as singularly transformative event, put placed within the larger chronology of the rhizome's sprawl.

6
Mapping and Tracing: What Is the Aim of a Global History of Religion?
This article began with a critique of historiographies of religion that neglect to historicize their terminology and thereby clandestinely introduce suprahistorical entities in their analysis, such as Islam and colonialism. When following Bergunder's program of studying the quotidian understanding of "Religion 2," such an approach no longer holds, because we are invariably thrown into a plurality of competing or even incommensurable usages of a term that are connected with specific practices. The "Mecca letter affair" of 1908 is a perfect example of such a plurality of practices which could all be understood as a form of "Islam politics": copying the letter to call the community of Muslims back to more faithful observance, or spreading it with seditious intent; seeing the letter as evidence of an "Islamic danger," or using it to attack Rechenberg's Islam policies; employing the colonial panic to expel Sufi preachers from within the local community, or seizing it as an opportunity to profile Islamwissenschaft (Islamic Studies) as a politically useful academic discipline. All of these practices could be construed as contributing to the "Religion 2" of "political Islam," but in different and often contradictory ways, and the "Religion 1" of describing this academic discourse must be careful to not re-introduce "political Islam" as a supercategory that collapses this plurality. In our case, this could be, for example, a notion of "political Islam" that draws mainly on a combination of colonial reports and Becker's Islamwissenschaft. In such a view, the "Mecca letters" would be a historical occurrence or dynamic to be explained from within the East African Muslim community in reaction to a colonial outside, rather than from an intricate colonial entanglement that crucially included a variety of perceptions and power struggles in defining "legitimate Islam." As antidote to such a misconception of "Religion 2," we have explored Deleuze' and Guattari's metaphysical figure of the rhizome for opening up a different historiographical methodology, which, firstly, rejects fixed categories, such as colonizer and colonized, and instead follows the principle of connection and heterogeneity to trace out a specific web of interconnected practices and pressures in order to understand the specific historical dynamic at work in the "Mecca letter affair." Secondly, this led us to give up on the explanatory category of historical subjects and objects, replacing them instead with the principle of multiplicity, in which the machinic assemblage of the chain letter constantly produces objects and subjects, not in a neat filiation from original to copy, or from an ideal to its instantiation, but in a rhizomatic sprawl. Thirdly, the principle of asignificant rupture led us to reject the initial hypothesis of the "Mecca letter" as a singular or key event and understand it instead as a microcosm for mapping out an aparallel evolution of Islam and politics in German East Africa.
Yet what is all of this good for? Who needs a global religious history that no longer draws even the roughest contours, that cannot take the simplest dichotomies as given? How can one come to terms with complex global processes such as the emergence of Islamic reform movements, when even such a small episode as the "Mecca letter affair" dissolves into a gigantic rhizome with no boundaries and dizzying depths of complexity?
These questions are understandable, and they may form a major obstacle to applying Deleuze' and Guattari's to historical inquiries, especially when the primary goal is indeed "coming to terms" with events -that is, delivering a factual account and an authoritative analysis of the past. Factuality and analytical incisiveness are of course the primary currency of historiography, and rightly so. But if Deleuze and Guattari are correct in pointing us to the infinite and irreducible complexity of historical and epistemological entanglements, then historiography is marked by a considerable analytical surplus as each account confines "the past" by demarcating events, constructing historical subjects and forging causal paths. History thus emerges as a function of the present, proliferating constructs of the past that are driven by and aimed at contemporary institutions and practices (Certeau 1988;Bergunder 2020).
For those interested in the study of this historiographical surplus, Deleuze' and Guattari's figure of the rhizome is of particular interest, especially in its final two principles: mapping and tracing (or cartography and decalcomania). Here the two philosophers show most clearly that their final aim is not metaphysics or ontology but a reorientation of epistemology. The rhizome, they note, does not itself exist as a unit, as an unconscious structure to be discovered by the tracing of the researcher. Instead, the rhizome is a map, "oriented toward experimentation in contact with the real" (Deleuze et al. 1987: 12). As such, it does not "reproduce an unconscious closed upon itself; it constructs the unconscious" (Deleuze et al. 1987: 12). The map is productive, it fosters connections, it is drawn upon all kinds of surfaces, and it can be reworked with different intentions. Indeed, as a productive practice, mapping the rhizome is part of the rhizome, the ontology remains flat even in this epistemological operation. But this does not mean that mapping a rhizome is an empty exercise, because once the map is drawn, Deleuze and Guattari tell us, "the tracing should always be put back on the map" (Deleuze et al. 1987: 13). For tracing is not the inverse operation of map making, instead, tracing a map is always oriented toward a practice of traversing the rhizome from an entry point to an exit point. Thus, tracing enables a variety of practices from historical discovery to critique, depending on the respective entry and exit points of the historian. Tracing is what makes the map useful, its depictions ground a multitude of historiographical operations that are aimed at the present.
Mapping the "Religion 2" of Islam in German East Africa via the "Mecca letter affair" therefore is the attempt to represent the rich tapestry of the historical archive in a web of lines or connections that enables a tracing of debates about Islam in German East Africa in their historical complexity. One useful entry point may be current discussions about the place of Islam in the contemporary constitutional arrangement of Tanzania. In the present political climate of Tanzania, some Muslims feel that they have been left behind or willfully marginalized in the postcolonial constitutional settlement. One prominent proponent of this view is Mohammed Said, a graduate of the University of Dar es Salaam in political science, who has written a number of historical accounts and is a well-connected campaigner for a better recognition of the contributions of Muslims in the struggle for Tanzanian independence. His work is driven by a sense of systematic erasure and marginalization of Islam in the history of Tanzanian nationalism, instigated by the Christian wing of the Tanganyika African National Union around Julius Nyerere (Said 1998: xxi-xxv;Said 2012;Said 2011).
As part of this revisionist historiography, Said claims that Muslims resisted German colonial rule while Christians collaborated with the colonial government. In particular, during the Maji Maji War of 1905, he contends that "some Christians fought alongside the German army against the people to safeguard Christianity," while "some Muslims were hanged particularly for killing missionaries and for waging a war against German rule" (Said 2012). This contention is not warranted by historical sources and established scholarship about the Maji Maji War, but it is not without historical grounding when one looks to the German debates of Islam around the "Mecca letter affair." Said cited August Nimtz for his claim, who briefly noted in his study about the Sufi ṭuruq in Tanzania that "the fact that Muslim teachers played a role in spreading the [Maji Maji] ideology is, no doubt, important" (Nimtz 1980: 13, emph. mine). Nimtz, in turn, drew on John Iliffe who had noted that from among the waalimu (teachers of Islam and often spiritual leaders in local communities) "some had probably been involved in Maji Maji" (Iliffe 1969: 194, emph. mine). Iliffe cited as evidence for this claim Rechenberg's decree of 1909, which sought to explain the dynamic behind the spread of the "Mecca letters." Yet as was already mentioned above, Rechenberg drew for his initial analysis on the claims of Walter Wendt, the over-zealous district officer of Lindi who without success had sought to prove a connection between Maji Maji and the Mecca letters. So once again, we see the "Mecca letter" rhizome sprawl, as Wendt's conjectures hardened into fact in academic writings and are then employed in the present struggle of Muslims against a perceived Christian dominance in the current political system.
It is not the primary intent of these observations to critique Said's historiography. A critical discussion of his claims appears necessary (Glassman 2001) but is better left to the Tanzanians involved in the identity debate from which Said draws his impetus. The point I wish to make here, is a more subtle, methodological one: Said may be wrong in his claims about Muslims in the Maji Maji War, but he stands in a long tradition of claims about "political Islam" in German East Africa that has influenced discussions about and political approaches toward Islam in the area since 1908. Merely correcting facts about the past does not suffice to address this longer history of misconceptions and misconstruals of Islam, nor the effects it has produced. Drawing the map of the "Mecca letter" rhizome in all its political and epistemological entanglements, by contrast, may make it difficult to come to conclusions about "political Islam" in the past, but it does enable a tracing of political and historical connections that is much better suited for a critical reappraisal of contemporary debates of Islam and politics. And herein lies the point: A rhizomatic history might not help us determine the past, but it certainly enables a fuller present.
A rhizome has no beginning or end; it is always in the middle, between things, interbeing, intermezzo.
The tree is filiation, but the rhizome is alliance, uniquely alliance.
The tree imposes the verb "to be," but the fabric of the rhizome is the conjunction, "and … and … and …" This conjunction carries enough force to shake and uproot the verb "to be." Deleuze et al. 1987: 25