Network analysis in general
The basic premise of network analysis is that possible courses of action and, as a result, human actions are determined partly by the social relationships of the individual performing the action. Thus a further level of investigation is added between that of the individual perspective on the one hand and wider categories of analysis, such as class or gender, on the other. However, individuals are by no means regarded as being determined entirely by social demarcations. If we consider interpersonal networks, we are looking at an essential tool employed by individuals in order to overcome or manipulate social barriers.
The most important term, which is at the same time the most methodologically difficult in a practical context, is the quality of relationships. Besides kinship relations, relevant relationships in the field of Islamic networks of culture and learning are especially the relationship between teacher and pupil and, within mystical Islam or Sufism, the relationship between a shaykh and his disciple. Using the self-representations that are available in the sources, the intensity of such relationships is generally difficult to determine. Consequently we will have to look not only at multiple network relationships (e.g. when the shaykh–disciple relationship is complemented by a relationship by marriage), but also at diachronic aspects such as the frequency of visits, the duration and/or stability in times of crisis of the relationship. These considerations will enable us to reach an assessment of the qualitative aspects within the network analysis. The most important of the quantitative elements are the density and the extent of networks. "Density" refers to the frequency of direct relationships within a network.
The extent of the network is a useful analytical category, particularly in regard to "ego-networks", i.e. relationship networks which evolved around a certain person who frequently holds the position of the founder of the community. In these cases it is easiest to measure whether a person has a direct or an indirect relationship (via one or several intermediate links) with the central person of the "ego-network".
Networks are frequently formalised to only a small degree and consequently have fuzzy borders, i.e. their outer limits cannot be clearly determined. Thus the networks of one person or one quality often overlap with other networks. Within network analysis, the theoretical framework for this structural characteristic is weak tie concept among others. Granovetter pointed out that people who occupy a marginal position within a network have a potential advantage over central positions because they can be more easily integrated into more than one network. It is true that persons in this situation would only ever have a marginal position within different networks. However, from the point of view of network analysis they would still have more possible courses of action open to them. In times of crisis they would find it easier to move between different networks than persons who may have more numerous and more intensive contacts but only within a single network, or who are identified primarily with this one network.1
A separate category of analysis is the absence of network contacts. Particularly in informal contexts the absence of network contacts highlights the membership of, or dissociation from, groups. This aspect can be employed – in conjunction with, for example, discourse analysis – to better comprehend historical phenomena.
Network analysis in Islamic Studies
Network analysis has been introduced to Islamic Studies in Stefan Reichmuth (*1950) and Roman Loimeier (*1957) as well as other scholars, on whose studies this article is based. Reichmuth and Loimeier argue that, in view of the fact that Islam does not know any centralised religious institutions apart from the pilgrimage, relations between various Islamic groups and their religious authorities play a decisive part. They add that a network perspective seems to suggest itself when analysing the activities of religious scholars. The connections among religious scholars as well as the relations between them and their pupils can overcome not only the limitations of social stratification and borders between urban, rural and nomadic societies, but also spatial distances. These connections are ultimately the essential channels by means of which information will be exchanged even over great distances, and cultural capital may be acquired. Educational journeys and pilgrimages in particular play an important part in this exchange.2in the 1990s by
Stefan Reichmuth followed the ideas of the American Islamic historian John Voll (*1936) and, also taking into consideration World-Systems theory, attempted to understand network analysis not merely as a method for analysing sources but also as a theoretical model. This model would make it possible to define the subject matter "Islam" as a sphere of communication within which Muslims could refer to their tradition of religious culture and learning in order to communicate among themselves about the world and their position in it; in short: about their identity.3
The groundbreaking works by Loimeier and Reichmuth had already placed a strong emphasis on the analysis of "ego-networks". This tendency was continued in the subsequent writings of some other scholars who examined mainly ego-networks of the seventeenth to twentieth centuries.
Henning Sievert studied Rāghıb Meḥmed Paşa (1699–1763), one of the most influential political figures of the during the eighteenth century. Rāghıb grew up in a family of bureaucrats and from the days of his childhood his education prepared him for a career in the Imperial administration. By way of many posts – in , , , , and – he ultimately reached the position of Grand Vizier. Sievert shows that Rāghıb's career and final pre-eminent position were due to his networks and his cultural capital. Thus he lost his position as governor in Egypt in 1748 after a shift in the constellations of power in Istanbul.4 Similar circumstances led to his appointment to the position of Grand Vizier in Istanbul in 1757, at a time when he was governor in Aleppo.5 He remained Grand Vizier until his death in 1763, which was an unusually long term of office for an Ottoman Grand Vizier in the eighteenth century.
Rāghıb's career was characterised to a high degree by contacts in the Ottoman provinces as well as in the capital Istanbul. This balancing act is maybe most clearly visible in Rāghıb's marriage strategies, as he married one of his daughters to a civil servant in Istanbul, while the other daughter married a civil servant who was an influential figure in the regions of 6 Rāghıb's ability to forge lasting contacts in the provinces and in the capital was greatly supported by his excellent mastery of the standard literary-bureaucratic canon of the Ottoman Empire (in particular the three languages Persian, Arabic and Turkish, complete with their respective literary traditions) and his profound knowledge of religious studies. In this way Rāghıb achieved high prestige in both the capital and the province, perpetuating contacts which provided him with economic opportunities and political support and allowed him to participate in the Ottoman Empire's cultural life open to the provincial elite. With Rāghıb as his example, Sievert shows how widespread the networks were that, in a way, held the Ottoman Empire together. However, he firmly rejects the dichotomous view of "capital vs. province" and emphasises the importance of the relations between the provinces.7/ / .
Stefan Reichmuth analysed the works of the polymath Murtaḍā al-Zabīdī (1732–1791), who was born in and received his first education there. From 1749 to 1754 he spent time in and the cities and in the , where he forged ties with a number of other important scholars. He then moved to where he lived until his death in 1791, leaving the city frequently for short journeys. Zabīdī left several extensive works, among them an unfinished dictionary of biography (Mu'jam), within which he noted the biographies of scholars whom he had met personally. A number of different genres, such as diary notes, travel writing and autobiographical elements, make up the text of the Mu'jam. Due to the large number of entries (over 643) and Zabīdī's extremely detailed accounts, the Mu'jam is an exceptional source even within the genre of biographical dictionaries, a source that is suitable for the application of the tools of network analysis.8 Of the studies discussed here it is Reichmuth's work which applies the quantitative methods of network analysis in the most intensive way, achieving a number of statistical results. Thus Reichmuth shows how Zabīdī's relationships changed through different stages of his life, and how this correlates with the various types of cultural and intellectual subject matter conveyed within these relationships. It becomes clear, for instance, that Zabīdī maintained relationships relevant to Sufism (i.e. Islamic mysticism) mainly in the , while logic and speculative theology (kalām) were mainly covered by contacts with scholars from the heartland of the Ottoman Empire ( and the ) and the .9 Furthermore it is possible to statistically record relations between different cultural and intellectual topics as well as their relations to certain professions within Zabīdī's network of relationships.10 All told we are able to record statistically the process that made Zabīdī a famous Muslim scholar of the eighteenth century, with contacts reaching from India through the to and the Southern fringes of the desert. What is particularly remarkable about these contacts is the high percentage of scholars in regions beyond the so-called Islamic heartland (namely the Fertile Crescent, Egypt, ). This is proof that regions such as Yemen or were by no means on the outer fringes of Islamic history in the eighteenth century but rather constituted a vital component of it.11
Unlike Reichmuth, Michael Kemper's (*1966) studies of Dagestani scholars of the nineteenth century maintain a critical position towards the quantitative approaches of network analysis, since he regards the sources as insufficient and therefore prefers the term "network perspective".12 He applies this approach especially to the analysis of the resistance movements against the advancing in the nineteenth century. These movements centred around the three Imams Ghāzī Muḥammad (b. ca. 1785 or 1795, d. 1832, Imam 1828–1832), Ḥamzat Bek (1789–1834, Imam 1832–1834) and Shāmil (1798–1871, Imam 1834–1859). Earlier descriptions used to postulate a connection between these resistance movements (which led to the establishment of lasting rule under Shāmil) and Sufism, partly because Shāmil presented a Sufi Shaykh to support his claim of being the legitimate ruler. It was postulated that Sufism not only provided the ideological justification of jihād but also, through the structure of the Sufi brotherhood (ṭarīqa), a sort of logistical base for the fight against the Russian colonial power. Kemper, however, was able to show that Sufi connections were not immediately relevant to the topic of jihād. By correlating Sufi networks (especially by looking at chains of initiation) and analysing various pamphlets he was able to demonstrate the political dimensions of Sufi debates in nineteenth-century Dagestan. Thus it became clear, for instance, that there had for some time been a discussion among the followers of Sufism of whether Sufi ritual formulae ought to be recited loudly or quietly. In Dagestan this evolved into a criticism of the use of these formulae as battle cries, which was a dissociation from the jihād against the Russians.13 The approach of "network perspective" was also successful in explaining the durability of Shāmil's reign. He created a system of deputies whose power rested almost exclusively on the military. Shāmil's overarching power, on the other hand, was based on a combination of military and – in the person of his close ally the Sufi grand master al-Ghāzīgumūqī (1837–1901) – religious factors. Consequently none of Shāmil's deputies could challenge his position as leading figure of the Imamate.14
Thomas Eich (*1973) studied the rise and fall of Abū'l-Hudā al-Ṣayyādī (1850–1909), a shaykh of the Rifā'iyya ṭarīqa, a mystical brotherhood which traced its origin back to Aḥmad al-Rifā'ī (1119–1182). Abū'l-Hudā, who came from a simple background in Northern Syria, succeeded within thirty years to become a member of the court of the young Sultan Abdülhamid II (1842–1918, r. 1876–1908). The analysis of Abū'l-Hudā's personal relationships shows that before he established himself in Istanbul in the late 1870s, he had been able to overcome several setbacks in his career thanks to his having forged contacts with diverse groups of people who wielded political influence in the Ottoman Empire and not only with the reformers who had made the early steps of his career possible. Starting in the early 1880s, Abū'l-Hudā published numerous writings, an activity that would continue almost without interruption until his death in 1909. Eich demonstrated the correlation between the development of Abū'l-Hudā's personal networks and the subjects discussed in his writings. Until the mid-1890s, Abū'l-Hudā's writings served the purpose of getting him accepted as the leader of the Rifā'iyya in the Ottoman Empire. In particular in Iraq, the Rifā'iyya became an instrument of Ottoman policy of integration.15 During this phase the network of Abū'l-Hudā's relationships expanded rapidly. Success led to conflicts with other influential groups, which was reflected in Sufi publications and in polemic writings which were published, among others, in India and . As Abū'l-Hudā lost some of his good contacts to influential men in the Sultan's environment at the same time, he was increasingly marginalised in Istanbul. The highly political implications of Sufi debates on miracles and on genealogies become clear in the synopsis of the development of personal networks and the subjects of Abū'l-Hudā's writings.16 Eich also extended the scope of his research beyond Abū'l-Hudā and applied his findings of his earlier work and the network approach to analyse the Iraqi uprising against the occupation in 1920. This enabled him to explain the varying patterns of action of different political parties.17
As regards the quantitative methods of network analysis, Eich steers a middle course between the exceedingly diversified application of mathematical methods employed by Reichmuth, and all other studies presented here, whose authors are very critical of the application of the quantifying tools of network analysis. By analysing elements such as the frequency of visits or the nature and tendency of communications Eich, on the other hand, can make quantifying statements for the description of Abū'l-Hudā's network as a whole.18 The low density and intensity of connections within the network explains how Abū'l-Hudā's contacts were able to expand so quickly until the mid-1890s, and then, once lasting crises erupted around Abū'l-Hudā and precipitated the loss of his power, to crumble equally quickly.
Jan-Peter Hartung (*1969) devoted his study to the activities of the Indian scholar Sayyid Abū l-Ḥasan 'Alī al-Ḥasanī Nadwī (1914–1999). Over a long time Nadwī was seen as an important and unifying figure for the integration of educated Muslims in India and was an essential link between Indian Muslim groups and organisations in the . As early as the 1950s he was visiting professor in ; he was a founding member of the Muslim World League which has its headquarters in Mecca, and he was involved in establishing the Islamic University in Medina (both in the 1960s).19 There are two factors which meet the eye when looking at the development of his position as an important link between Muslim India and the Arab world. On the one hand, Nadwī went on several pilgrimages to Mecca during which he was able to form important personal contacts, which had the effect (probably rather surprising to him) of the Saudis asking him in the 1960s to cooperate in founding the Islamic University in Medina. In forming these contacts he was, secondly, aided by his excellent knowledge of Arabic owing to which he was soon invited to give public lectures and radio addresses in several Arab countries.20
Nadwī was of the opinion that the global Islamic community was in a long-term process of decline which had begun in the early days of Islam and manifested itself in the present-day Muslims' impotence in regard to realpolitik. He contrasted this with the Qur'anic idea that the Muslims are chosen to be the political and spiritual leaders of all humanity. The main requirement to achieve this ideal was, in his view, a Muslim community that was united by its faith. It was this point of view that would ultimately contribute substantially to Nadwī's failure on the international level, as he was not willing or able – in the various committees and different contexts – to allow regional particularities into his way of thinking. This became particularly clear in the context of his work with the Muslim World League, where he became increasingly marginalised after the assassination of the Saudi King Fayṣal (1906–1975). Not the least of the reasons behind this development was Nadwī's understanding of Islam, which was strongly influenced by his Indian socialisation, including a positive attitude towards Sufism (which is opposed to the puritanical thrust of the Wahhabi school of thought prevalent in ). Another reason was his demand for recognition of the important role of in modern Arab-Islamic history.21
Hartung's study combines network analysis with the concept of the "social field" coined by Pierre Bourdieu (1930–2002): a delimited and pre-structured area within which several agents compete for symbolic goods.22 Hartung analyses Nadwī's activities on the basis of three categories, the closed, half-open and open fields. These are distinguished from one another by gradations: The more a field is open to the outside, the more equal are the agents and the smaller is the likelihood that communicative codes are restricted permanently and are restrictive in their effects on social actions.23 According to this definition, Hartung analyses Nadwī's important Indian networks of contacts as a closed field. He discovers that three different kinds of relationship overlap systematically: kinship, teacher-pupil and Sufi master-adept relationships. It turns out that for several generations these different relationships had not been clearly distinguishable in Nadwī's environment and that connections of intellectual instruction were understood as similar to kinship ties.24 A half-open field, according to Hartung, is defined by a combination of "closed field" and "open discourse", an open field by a combination of "open field" and "open discourse". There are examples of both in India as well as in the Arab world. Nadwī's involvement in the Muslim World League, at first so successful, should be seen as an example of an open field.
Bekim Agai studied the activities of the Turkish thinker Fethullah Gülen (*1938), who has lived in the since the late 1990s. The central term for Agai is the cemaat, which is used in to refer to autonomous religious organisations, without, however, being clearly defined. Agai defines cemaat as
a network containing relations based on the recognition of a discourse and of expedient motives. Anyone who shares the discourse and subordinates himself to it may form contacts to the cemaat and enter into multiple relationships within it.25
Central to Gülen's ideas is his positive assessment of culture and learning even if it they do not immediately relate to religion: to him, science is a means of comprehending God in a rational way, and it is also a means to achieve prosperity and political independence. According to him the reason for the fall of the Ottoman Empire (and for national and international ills) is having neglected culture and learning. Thus Gülen sees even non-religious education as serving religious aims, and supporting it becomes a religious act. In this way he appeals to many people beyond Turkey's borders: his Islamic statements are in the main conventional and conservative but they do leave room for interpretation thanks to the new combination of various elements of discourse. Consequently they are interesting to many people, not least a secular audience.26 The cemaat in Agai's sense then works, and grows, to a high degree in educational institutions that differ from one country to another, e.g. in centres for extra tuition in Germany.27 The decisive factor in the formation of the network around Gülen was the transformation of the very dense network of his earlier activities, connected to a very specific discourse, into an open network that offers a choice of different relationships and positions itself in a less strictly regulated discourse. In this way it has become possible to include non-religious social areas. The resulting image is of a central figure, Gülen, surrounded by circles representing the social areas. They are all at different distances from him, which in some cases is expressed in the relationships themselves, but also in the fact that they employ the discourse shaped by Gülen in different ways. These non-religious social areas, in descending order, are: early followers, religious audience, national secular audience (always with reference to Turkey), and international audience.28
What all the studies summarised here have in common is that they make the connection, each with its own emphasis, between personal networks and Islamic cultural and educational content. In this way they enable us to determine not only the geographical spread and extent (in regard to space) of certain cultural and educational subject matters, which can probably be seen most clearly in Reichmuth's work on Zabīdī. They also allow us to point out the pragmatic-social or political dimension of particular discussions, as we see in Kempner's work on Dagestan or Eich's work on Abū'l-Hudā. Furthermore we are able to see how networks that are more defined by discourse can reach the limits of their growth – witness Hartung's work on Nadwī – or not, as shown in Agai's study of Gülen. This depends in every case on how flexibly cultural and educational content can be adapted to local circumstances. Finally, we can see clearly how closely the creation and maintenance of personal networks are linked to very specific social capital, exemplified by the respective language skills of Rāghıb in Sievert's study and Nadwī in Hartung's.
Critical reception of network analysis in Islamic Studies
The critical reception of works on Islamic networks, which were mostly carried out after the year 2000, was varied (disregarding some polemic reactions which were clearly not related to the subject).29 The majority of reviews examine the individual studies with regard to the results. There was, however, hardly any discussion of the application of the tools of network analysis.30 The great difficulty in reading the texts due to the numerous names which appear in them is a concern which was repeatedly expressed.31 It seems that the results achieved by the methods of network analysis are on the whole to be seen as substantial contributions Islamic Studies. The method that made these results possible, on the other hand, is hardly discussed at all.32 This is rather surprising, as German Islamic Studies in particular have been reflecting critically since the late 1990s on the fact that throughout the history of the discipline it has defined itself, and still does, primarily by philological expertise. This was no longer seen as sufficient; and consequently, after diagnosing the theory and methodology as deficient, there have been calls for a new approach that includes the theories and methods that have been lacking so far.33 It is in this context that we must understand the great interest with which network analysis was at first greeted by German Islamic Studies.34 However, after only a few years, the deliberations on the future of Islamic Studies appeared to head into an entirely different direction, namely "area studies": Islamic Studies should free themselves from their Arabo-centric approach, which was also reflected in the choice of mainly Arabic sources and which led to a centre-periphery model with the Arab world heavily influencing other regions with large Muslim populations. Rather, it should be borne in mind that regions such as , India and are independent and must be seen as equally worthy of study and should, therefore, become firmly anchored within Islamic Studies, after having long been relegated to the periphery. Influenced by Art History and Gender Studies other kinds of sources and areas of society have been pointed out as being apparently peripheral within Islamic Studies.35
It is indeed surprising that network analysis was not discussed in this context, as the studies by Stefan Reichmuth, which meant to contribute to establishing the theory of network analysis in German Islamic Studies, had aimed in the same direction as a critique of the outdated centre-periphery model.36 Furthermore, the concrete study projects employing the methods of network analysis had nearly all discussed those subjects which had been marginalised in Islamic Studies so far: India, Sub-Saharan Africa and the Caucasus may serve as geographical examples.37