Mapping the network of Shiʿi clerical relations in the Middle East: an analytical approach

ABSTRACT

While Shiʿi clerics are among the most influential political actors in the Middle East today, there remains insufficient understanding of the network of relationships that links them together and shapes their politics. This article argues that understanding the influence of Shiʿi clerical elites in the region requires a more systematic exploration of their internal dynamics, including how individuals within the clerical elite ascend through the hierarchy and the legacy they carry. Without this knowledge, comprehending where, when, and how new leaders (maraji’) will emerge, as well as the impact they are likely to have on the politics of their societies and beyond, remains challenging. To illuminate these dynamics and better understand the influence of this important group of actors, this article proposes an approach that aims to map the interconnected world of the Shiʿi clergy using quantitative network analysis. It argues that network analysis can complement existing qualitative studies by illuminating latent connections among the Shiʿi clergy; and, ultimately, can offer a more authoritative foundation for forecasting the identity and politics of future Shiʿi marjaʿiyya.

KEYWORDS:

• Shiʿism
• clerical authority
• network analysis
• Iraq
• Iran

Following the 1979 Iranian revolution, there came to power a group of clerics who established a Shiʿi theocracy under the leadership of Ayatollah Ruhollah Khomeini, a religious leader who was unknown to many observers just a year earlier (Amir Arjomand 1989, 134). About 25 years after the revolution – and as US officials were devising a new, post-Saddam political order in Iraq – another unfamiliar religious figure appeared in the region’s political mise en scène, Ayatollah Ali Sistani. Sistani managed to consolidate Shiʿi authority in the country by encouraging Iraqis to participate in national elections and by calling for a policy of ‘one man, one vote’ in 2005 (Nasr 2006, 231). As was the case with Khomeini in 1979, few people anticipated that Sistani would become the most influential political player in post-war Iraq (Sayej 2018, 8). Is there a more systematic approach to unveiling the Shiʿi clerical milieu and eliminating these blind spots in the contemporary Middle East? Khomeini, Sistani, and their clerical colleagues form one of the networks which have shaped the politics of the contemporary Middle East; a region consisting of myriad interconnected social elements, ranging from individuals to institutions.

Interconnectedness significantly influences the dynamics of the Middle East, an aspect that can be meticulously examined through the lens of network analysis. Although previous scholarship has delved into network structures within the region, these inquiries have primarily been confined to areas such as conflict dynamics, terrorist affiliations, and networks of privilege. Network analysis as a comprehensive methodology for understanding the multifaceted aspects of the Middle East has thus been largely underutilised (Barber 2015; Haddad 2011; Heydemann 2004; Maoz and San-Akca 2016; Wu, Carleton, and Davies 2014).

This article seeks to stimulate nuanced discussion and contribute to this relatively nascent field of study, employing network analytic approaches to studying the Middle East by illustrating the utility of the approach for studying Shiʿi clerical elites in the region.¹ This important group of political actors has been for centuries, and remains today, embedded in a transnational network of individuals, institutions, and ideologies. This network represents an evolving social world: a world informed by Islamic history, convictions, and more recently, political meanings. Identifying the clerical nodes (high-ranking Shiʿi clerics) and their ties could lay the foundation for a more in-depth analysis of the Shiʿi clerical network in the Middle East.

A cleric’s position within this network arguably contributes significantly to his trajectory to becoming a marjaʿ, the highest clerical position in Shiʿism.² This is predicated on the notion that his position within a network structure signals its status among his fellows, because it signals his rank amongst his colleagues; more central figures are more prominent. Likewise, as these individuals move up through the clerical hierarchy, their political viewpoints will shape and be shaped by the links they establish with their fellow clerics, the community, and other hubs of power such as the state. To this end, employing a systematic and quantitative network analysis can complement qualitative observations of the contemporary Shiʿi clergy and its behaviours. It brings to the surface latent connections, making them visible, demonstrating the importance of such ties in the collective system of Shiʿi clerical authority. It also provides a baseline that allows us to predict future trends, by excavating recurring patterns from the past. Network analysis therefore offers a resourceful toolkit capable of contributing to our understanding of why an individual ascends to the position of a marjaʿ, who could potentially attain this status in the future, and may elucidate the political and theological orientation of such an individual. This approach, when combined with available qualitative observations, significantly enhances our understanding of the internal dynamics of the clergy, and thereby provides insightful projections on future marjaʿ candidates and their potential political leanings.

The aim of this article, then, is to further illuminate the politics of the Shiʿi clergy and its influence in the region by introducing a structured approach of mapping, visualising, and analysing the network of Shiʿi clerical authority (marjaʿiyya) in the modern Middle East. The first section outlines the emergence of the clerical network in the region. It delves into the various properties of this structure, which has evolved around the Shiʿi marjaʿiyya since the position was institutionalised in the mid-nineteenth century. In the second section, I will elaborate the social network approach. The first step involves the definition of the clerical network’s boundaries through identifying its constituting nodes and links. The second step delineates how available social network measurements contribute to and enhance our understanding of this network. The third section of this article applies the network methodology to the case study of Ayatollah Khomeini. His clerical network is mapped out, identifying associated clerics, and establishing their interconnections. Subsequent application of selected network measures illustrates Khomeini’s commanding presence within this clerical network. The article concludes with a discussion on the potential implications of broadening this network-focused approach to include Shiʿi clerical networks throughout the Middle East and beyond. By advocating for the integration of a social network approach in studying Shiʿi clerics and their intricate internal dynamics, this article contends that our comprehension of their roles within the larger regional contexts can be further enriched and nuanced.

The emergence of the Shiʿi clerical network in the modern Middle East

Today, Shiʿi clerics are among the most influential political actors in the Middle East, as evidenced by the growing number of scholars across a range of disciplines who are studying Shiʿi clerical authority, its history, its role within the community of the faithful, and in the so-called Shiʿi revival in the region (Calder 1979; Eliash 1969; Kamrava 2018; Kohlberg 1988; Lambton 1964; Rizvi 2010). In contrast to mainstream Sunni Muslims, Shiʿis believe in the necessity of an intermediary between divine authority and the community of the faithful, the first and pre-eminent example of whom was the Prophet Muhammad himself. After the Prophet’s death, they believe that this role was taken up by a series of 12 divinely guided imams who acted not only as the religious authority for the community of the faithful but, according to many accounts, as righteous political leaders as well. Therefore, when – according to mainstream Shiʿi beliefs – the last imam went into occultation in the tenth century, the community was faced with a leadership vacuum. To fill this vacuum, a class of religious scholars emerged (Kalantari 2019).

As early as the eleventh century this elite class of religious scholars had appropriated some of the prerogatives of the imams that had come before them, functioning as his general deputies. Without the guidance of an imam, clerics took over the role of promoting the faith’s principles and doctrines, and as such the community of scholars became the point of reference for Shiʿi in their daily religious affairs. This led to the development of the relationship between ijtihād and taqlīd. A qualified Shiʿi cleric (a mujtahid) utilises juristic reasoning to extract specific laws from Islamic principles, thus conducting ijtihād. However, a lay Shiʿi (a muqallīd) who lacks religious expertise should follow the mujtahid’s verdicts and emulate them; he does taqlid (Clarke 2001). Thus, a dynamic relationship between mujtahids and muqallīds has formed within Shiʿism, meaning that the very existence of the community of mujtahids depends on the support (which can be financial) that it receives from lay followers.

At the apex of the community of mujtahids sits the supreme exemplar office, the marjaʿ al-taqlīd. For centuries, aspiring religious students travelled to Shiʿi scholarly hotbeds – in Rey, Baghdad, Hilla, Najaf, and Qum – to learn from marajiʿ, graduate and themselves become mujtahids, and then return to their hometown to propagate their knowledge of the faith. At the same time, lay Shiʿis went to them with their religious inquiries, respected and supported them financially. Since the tenth century, Shiʿi communities scattered across the Islamic world have formed nebulous clusters around marajiʿ and mujtahids, looking to them as the religious authorities in their locality.

The development of communication and transport means, in the early decades of the nineteenth century, transformed this socio-religious context. Against the backdrop of pro- ijtihād movement revivalism, a sole Shiʿi marjaʿ emerged – Muhammad Hasan al-Najafi (1785–1849), whose leadership in Najaf seminary institutionalised the position to the form that we still see today (Kazemi Mousavi 1996, 188). Al-Najafi succeeded in consolidating a network of mid-ranking teachers at the seminary, his students, and his ever-expanding community of followers (Litvak 1998, 56). Members of this network began to promote the authority of the marjaʿ, to collect religious taxes from lay followers (including the merchant class), and to remit these to Najaf (Kazemi Mousavi 1985; Louër 2012, 9). With the expansion of this network, al-Najafi was able to impose his authority over ever-larger parts of the Shiʿi world; and this expanding patronage and influence network became the most salient asset for the generation of Shiʿi marajiʿ who succeeded him. When Shaykh Murtadha Ansari (1781–1864) succeeded al-Najafi as the new marjaʿ around 1850, the position of marjaʿiyya and the very characteristics of its holder were well established among mujtahids and lay followers (Cole 1983).

Yet the question remains, how is a Shiʿi mujtahid recognised as a marjaʿ? Addressing this, academics have alluded to the clerical network, the position of a mujtahid in it, and his aptitude in exploiting that position. The clerical network, at its heart, is a representation of scholarly bonds, patronage, and allegiance of lower ranking mujtahids to their superiors. General convention implies that a prospective marjaʿ should possess exemplary knowledge of the jurisprudence and, by some accounts, be known as the most learned (aʿlam) among his colleagues at a given time (Walbridge 2001).

Beside scholarly acumen and ties, many also note the centrality of familial relations in the ascendancy of a marjaʿ. Elvire Corboz argues that a mujtahid’s family background forms a symbolic capital that goes hand in hand with his scholarly capital, and ultimately bolsters his chance of becoming a marjaʿ (Corboz 2019). The role of family ties in making a marjaʿ is so consequential, Sajjad Rizvi notes, that it impacts a marjaʿ’s ability to achieve wide renown. For instance, he points to Shaykh Ishaq al-Fayyad (b. 1930), who despite being widely recognised as the most learned figure in today’s Najaf, is little known outside because of his ‘inability to develop a social constituency for himself as he comes from a simple, peasant background in Afghanistan and does not have the family or class connections’ (Rizvi 2018).

Another factor in the rise of a marjaʿ is their ability to cultivate followers who will support them financially. Compared to Sunni religious authority, at least up to the 1979 Iranian Revolution, the Shiʿi marajiʿ were relatively independent of state sponsorship.³ Therefore, raising funds from followers is a crucial factor in their survival (Khalaji 2006). Whoever can secure more funds improves his chances of becoming a marjaʿ, and as these funds, at least in part, are distributed among one’s students and clients, it also further strengthens his personal network.

Little is known about the structural properties of the Shiʿi clerical network, despite its significance in making a marjaʿ and defining the boundaries of Shiʿi religious authority. While scholars unanimously acknowledge the existence of a transnational patronage and influence network which determines who becomes a marjaʿ, and which shapes both his political orientation and how he deploys his authority over the community, the structure and internal dynamics of this network have not been studied in a systematic way.⁴ Considering the pivotal role of the Shiʿi clergy in social mobilisations, sectarian conflicts, and political turmoil in the Middle East, we need to know not just about prominent individuals within this group, but about their internal and structural dynamics.

The Shiʿi clerical network: mapping connectivity

Social network theory posits that an array of social constructs – extending from individuals to groups, corporations, and states – are intricately interlaced within a comprehensive network of relations and interactions. Emphasising the significance of understanding these entities’ interconnections, mutual dependencies, and exchanges, the theory aims to explain not only individual behavioural patterns but also collective characteristics.⁵ Incorporating a systematic approach that encompasses the mapping and visualisation of relationships and the rigorous analysis of complex networks, social network analysis provides a robust framework for tracing various elements, including similarities in memberships or attributes, familial or kinship ties, and the flow of resources across the network.⁶ It tries to present a social phenomenon and offers nuanced understanding of the interplay between individual behaviours and larger structural dynamics.

Building on this premise, diverse methodologies involving social network analysis have been employed by scholars over decades to explore a variety of research questions. These include understanding the intrinsic bonds that hold a collective entity together within a network, discerning the nature of cohesion and group formation, determining the means by which individuals communicate within the network, analysing hierarchical structures, and investigating relational status.⁷ In order to fully leverage the methodological advantages of network analysis within the context of Shiʿi clerical milieu, two critical steps are needed: firstly, the clerical network needs to be defined and accurately mapped out, and secondly, network measurements need to be carefully applied. Once the network has been accurately mapped out, the use of quantitative measurements can unearth hidden attributes of the clerical structure at different levels, revealing its evolution and transformations over a given period.

Mapping the network

In every network, including the clerical network, links represent different types of ties between two nodes, sometimes with different weights. Nodes in our network are those religious figures who form the corpus of Shiʿi clergy in the twentieth century. The links among these nodes, however, are of at least three different types: family, social, and scholarly.

Given that the Shiʿi community identifies strongly with the charisma of the Prophet and 12 imams, family links and prestige are an important characteristic of any cleric who has grand ambitions to consolidate his clerical authority (Corboz 2014; Dakake 2007). Sayyids – clerics wearing the black turban, who purport to be descendants of the Prophet – are perceivably more popular within the community of believers (Morimoto 2012, 2). Moreover, the Shiʿi clergy are populated with family dynasties (e.g. Sadrs, Hakims, Bahr al-Uloums, Kāshif al-Ghiṭas, Khawnsaris), members of whom benefit from default social and religious capital for various reasons, from name recognition to pre-established access to resources (Mauriello 2011, 29). In the case of Shiʿi marajiʿ, it is in essence the job of their family members on the ground to drum up support for a marjaʿ and to coalesce his position among his peers with the community of lay followers. These family connections, and the aptitude of one’s household in promoting one’s leadership among people, is a determining factor in consolidating one’s marjaʿiyya (Walbridge 2001, 232). In our clerical network, these family links are established through diagramming father – son, father – son-in-law, and grandfather – grandson ties.⁸

Ultimately, family ties are just a tool to cultivate a large social following. This is where the marjaʿ’s authority really lies and provides the base for collecting religious taxes to distribute amongst his students. For growing their followings, the marajiʿ have historically relied on their exclusive patronage network of social links. This is where the marjaʿ’s representatives come to the fore (Jabbari 2003). Their responsibility is not only to improve the public image of their patron but also, as mentioned, to collect religious taxes on his behalf. In particular, it is the relationships that a representative forges with non-clerical sources of power (e.g. politicians, merchants, and wealthy businessmen) which has proven crucial in consolidating the marjaʿiyya of his patron (Khalaji 2006, 6).

Last but not least there are the scholarly links, which epitomise a mujtahid’s academic stature among his peers. Traditionally, it is a cleric’s scholarly credentials which facilitate his ascent through the ranks of the clergy, his recognition as an aʿlam, and ultimately his ability to assume the position of marjaʿiyya. Teachers who nurture exceptional students, and religious students who join the learning circles of the more revered teachers, are viewed as more learned amongst their peers. These teacher-student ties establish the scholarly links of our clerical network.

Besides taking these three distinct forms, clerical links can also have different weights. In the network analytical approach, every unique tie between two nodes establishes a link with a weight of 1. For example, in a case where the student of a marjaʿ is also his son-in-law and his representative, the link between the two has a weight of 3, because essentially it is composed of three separate links (familial, social, and scholarly) that reinforce one another, and thus the link ends up being more than the sum of its parts. The links in our network denote relationships between the nodes which are mutually inclusive; the existence of one type of link does not preclude the forming of another type. Therefore, the clerical structure displays an ‘undirected and weighted network’.⁹

Based on these characteristics of nodes and links, the process of mapping our clerical network involves the delineation of these links among the members of the clergy, the nodes. Once the scholarly, familial, and social links are mapped around Shiʿi marajiʿ, what emerges is the comprehensive clerical network, containing not only high-ranking marajiʿ but also all pertinent members of the clergy. The next stage, then, is to analyse various properties of this network using available measurements.

Analysing the network

Social network theory offers a range of measures that capture the structural properties of a network, both at node and network levels. These measures each illustrate a distinct facet of positional embeddedness for a given node within the network. A pivotal component of this is the concept of centrality, which over the last seven decades has been progressively refined, giving rise to a multitude of interpretations among scholars.¹⁰ Despite their differing conceptualisations, these approaches are unified by a common focus on point centrality, or the relative centrality of nodes within a network. Yet, perspectives significantly diverge beyond this shared focus. Some conceptualisations of centrality are structured to elucidate how certain nodes, due to their central positioning, might operate as intermediaries, thereby mediating the demands and influences within their network. This underscores the numerous ways an actor can be deemed ‘central’ within a network. Conversely, lower measures of centrality can serve to illuminate instances of exclusion or peripheral positioning within the network. At the node level, social network theory discerns several such centrality measurements, including degree, closeness, betweenness, and eigenvector (Borgatti 2005).

Degree centrality measures the centrality of a node in terms of the number of ties it has to other nodes in the network. The assumption here is that a hub-node, the most active node in the network, is the one with the most links with other nodes in the network. The higher the degree centrality of a clerical figure, the more connected and active he is within the clerical network. Degree centrality takes into account a node’s immediate links. A cleric might be linked to a large number of other clerics who are not necessarily well-connected to the rest of the network. This cleric could be important within his immediate context, but not outside of it. To this end, the degree centrality of a cleric signifies his level of access within the broader network, in comparison to his fellow clerics. Closeness centrality of a node, however, measures average distance to all other nodes; it can be thought of as the degree of separation between one node and another. A node with high closeness centrality – so long as it can reach other nodes in fewer steps – is closer to other nodes in the network, and can often influence the network more quickly and efficiently. Therefore, it is more central than a node with lower closeness centrality. A cleric with higher closeness degree is closer to other nodes in the network, and those with lower closeness, even if they possess many links to others, are more peripheral.

Betweenness centrality of a node denotes its ‘brokerage’ role between two other non-adjacent nodes. It is a score which ascertains a node’s importance in the network, based on measuring the extent to which it falls on the shortest path between other pairs of nodes in the network.¹¹ For example, according to the rules of clerical hierarchy, if a religious student wants to send a message to a marjaʿ, or request permission to attend his teachings, he must forward his request via an intermediary cleric who handles such requests. This intermediary cleric occupies a favoured position within the network, and the more people depend on him to make connections with the marjaʿ, the more authority he has. This cleric’s position within the network is indicated by his betweenness centrality score. A cleric with higher betweenness centrality is relatively more influential, as other clerics are more dependent on him to establish connections with additional nodes within the network.

Finally, eigenvector centrality measures the significance of a node for the connectivity of the whole network. In contrast to the above measurements, the eigenvector score emphasises the proportional centrality of a given node; it signifies the influence a node has in a network and is based on the assumption that connections to other high-scoring eigenvector centrality nodes contribute more to the score of the node than equal connections to low-scoring nodes. In particular, connections to clerics who are themselves influential will lend a cleric more influence. Thus, clerics with the highest eigenvector scores are those who should be watched within the clerical network; they will be those who have the best chance to hold the highest positions within the community of Shiʿi clergy.

In addition to node-level measurements, network theory also provides comprehensive network-level measures. One such key measurement that aids in assessing the collective behaviour of clerics is network density. This refers to the ratio of existing connections to possible connections within a single, bounded network, thereby gauging the extent of collaboration among all nodes. In essence, network density quantifies how many connections exist between different nodes relative to the maximum possible connections (Scott 2017, 84). The denser a network, the more nodes are interconnected and the greater the volume of interactions among individual nodes. It provides insight into the degree of interaction, interdependence, and integration among various clerics within a specific timeframe and locale. Changes in network density can illustrate collective shifts, such as whether the clergy becomes more interconnected (dense) in response to external threats. If we applied this density measure to the 1963–1979 period, for example, it might reveal a strengthening and consolidation of the clerical network, possibly serving as a precursor to a power shift. Another critical aspect of network theory is community structure, a concept that divides the network into areas of greater density (Scott 2017, 29). Numerous algorithms exist to find such structure, and it has been primarily employed in large-scale social networks to simplify network description. Moreover, the presence of smaller network patterns can shed light on underlying structural processes within the network, such as reciprocation levels or cooperation within small groups (like triangles) (Robins 2015, 178). An interesting application of this measurement would be to compare the properties of two communities within the network: how well connected are the clerics in Najaf in contrast to their colleagues in Qum?

Above, I set the boundaries of the modern clerical network by pointing out its constituting nodes and links, its characteristics, and how a network approach can further develop our understandings of the Shiʿi clergy and their internal dynamics. Beside what we already know about these major agents of modern politics of the Middle East through qualitative observations, quantitative network measures of the interconnectedness of Shiʿi clerical authority would further advance our knowledge. They reveal those often-latent meanings which every single link carries within this network by ranking, for example, the significance of a cleric at a given time, or which links are more or less central or peripheral to the relations that constitute the clerical authority of a figure individually, or within a network of clerics collectively. These measurements, hence, not only further clarify the genealogy of the Shiʿi clergy, but also render competent tools to forecast its future developments.

In the next section, the theories and concepts outlined previously will be further elucidated through an in-depth examination of Ayatollah Khomeini’s network. Despite his status as a high-ranking cleric, Khomeini remained relatively unknown to western observers prior to the 1979 Iranian Revolution. This case has been chosen to illustrate the potential contribution of social network theory approaches to the existing literature concerning the status and politics of Shiʿi clerical elites. Consequently, his case provides a compelling example of how the application of network analysis could have substantially enhanced our understanding of his position within the clerical network and presaged his ascent to the role of Supreme Leader of the Islamic Republic of Iran in 1979.

Khomeini’s network: a case study

Ruhollah Khomeini was among the first students of the Qum seminary and its founder, Shaykh Abd al-Karim Haʿiri (d. 1937). After Haʿiri’s death, the newborn seminary was in a dire financial situation, and not on good terms with the then more important Najaf seminary. This is when Khomeini, along with a group of younger teachers including Haʿiri’s eldest son, Morteza, invited Sayyid Hussein Broujerdi to Qum (Ansarian 2019, 228). Broujerdi was among the most learned mujtahids of the time, and had been teaching at the local seminary of Broujerd in western Iran for more than three decades (Ansarian 2019, 438).

Broujerdi finally accepted the invitation, moved to Qum in 1945, and remained there until his death. During this time, his marjaʿiyya became the most influential in the contemporary era. Broujerdi, who had avoided the spotlight for more than three decades, happened to be at the right time and place to become the last ‘sole Shiʿi marjaʿ’ (Walbridge 2001, 5). When he passed away in 1961, at least five mujtahids in Qum and Tehran were considered likely to be Broujerdi’s successor, and one of them was Ayatollah Khomeini.

Mapping Khomeini’s network

To map Khomeini’s network, the first step is to consider his family ties. We can begin with his eldest brother, Sayyid Morteza Pasandideh (d. 1996), who at the time was a respected figure in clerical circles who supported the Iran National Front and Prime Minister Muhammad Mossaddegh (Pasandideh 2013, 199). Mustafa Khomeini, the Ayatollah’s eldest son, was 30 years old when Broujerdi died, and was also the son-in-law of Morteza Haʿiri, son of the founder of the Qum seminary (Sharif Razi 1973, Vol II: 128). Ayatollah Khomeini’s second son, Ahmad, was only 15 at the time. Khomeini’s own son-in-law, Shahab al-Din Eshraghi (d. 1981), was from a wealthy and established family in Qum.¹² Establishing brother, son, and son-in-law ties results in the map of Khomeini’s family links (see Figure 1).

Figure 1. Khomeini’s family links.

A review of Khomeini’s family relationships reveals that he is linked to at least four previous marajiʿ before him, through marriage ties. Mustafa was son-in-law of Haʿiri, who himself was son-in-law of Ayatollah Muhammad Hujjat Kuhkamarei, one of Abd al-Karim Haʿiri’s disciples and successors. Later, Ahmad Khomeini married the daughter of Ayatollah Muhammad Baqir Sultani Tabatabai, who was cousin of Ayatollah Broujerdi and son-in-law of Ayatollah Sayyid Sadr al-Din Sadr (Tabatabai 2010, Vol I: 3–15).

Many narratives suggest that Khomeini was not keen to be recognised as a marjaʿ (Moin 2009, 75). It was only in 1964, three years after the death of Broujerdi, when he published Taḥrīr al-Wasīla, his collection of fatwas, that he officially announced his marjaʿiyya.¹³ By then, Khomeini was already in exile in Iraq, had started teaching at Najaf seminary, and had appointed a handful of representatives in Iraq and Iran (Khomeini 2010, Vol. I: 448–512). In Qum, his brother, his son-in-law, and Shaykh Ali Akbar Eslami Turbati were his full-fledged representatives (Sharif Razi 1973, Vol III: 286). Khomeini also named his lifelong friend, Muhammad Sadiq Lavasani, as his representative in Tehran (Moin 2009, 31). In additions, three of his close disciples and students – Muhammad Hashemian, Ali Akbar Hashemi Rafsanjani, and Muhammad Mahdi Rabbani Amlashi – orchestrated Khomeini’s social network throughout the rest of Iran (Hashemi Rafsanjani 1997, 127). They promoted their patron’s marjaʿiyya among lay followers, collected his religious taxes, and forwarded them to his office in Najaf from time to time. Shaykh Nasrallah Khalkhali, who was in charge of Khomeini’s office finances in Iraq, dramatically enhanced his social prestige in Iraq. Khalkhali was known as a major financer of religious schools in Najaf, had served as Broujerdi’s representative in Iraq, and was, at the same time, in charge of Ayatollah Sayyid Ahmad Khawnsari’s (d.1985) office in Najaf (Nakash 2003, 258). His social links to two marajiʿ, Broujerdi and Khawnsari, add to his degree centrality within Khomeini’s social network (see Figure 2 below).

Figure 2. Khomeini’s social links.

The third and final type of link is scholarly, which denotes teacher – student ties. As mentioned above, Khomeini was among the followers of Shaykh Abd al-Karim Haʿiri in the city of Arak, and he migrated with his teacher to Qum, which ultimately led to the pre-eminence of the town’s seminary. For his middle-level seminary studies, Khomeini attended the teaching circles of Muhammad Taqi Khawnsari (d.1952) and Sayyid Ali Yathrebi Kashani (d.1960). At the age of 25 Khomeini began study in earnest at the higher-level studies, al-bahth al-kharij, of Haʿiri. He also studied philosophy with Shaykh Muhammad Reza Masjid Shahi (d.1943). The one teacher Khomeini later referred to as his ideal mentor, however, was Mirza Mohammad Ali Shahabadi (d. 1950), with whom he studied gnosis (Moin 2009, 42).

Khomeini started teaching at the seminary in Qum following the death of Haʿiri. By the time Broujerdi assumed the marjaʿiyya, Khomeini was regarded as one of its most famous scholars. He taught in Qum until 1964 and resumed his teaching gradually in Najaf between 1966 and 1978. He taught scores of religious students in Qum and Najaf, some who became marajiʿ themselves in subsequent decades (see Figure 3).¹⁴

Figure 3. Khomeini’s scholarly links.

Combining all of Khomeini’s scholarly, familial, and social links results in his clerical network. In this case, a total of 109 nodes are interrelated with 268 links (see Figure 4).

Figure 4. Khomeini’s network.

This network, which can be mapped for every Shiʿi marjaʿ, essentially represents a section of the more comprehensive Shiʿi clerical network. Thus, an analysis of its nodes and links offers insights not only about Khomeini but also about the broader clergy network of which he was an integral part. This approach also illustrates the potential of social network analysis in further informing our understanding of Shiʿi clerics in modern times.

Analysing Khomeini’s network

As I have noted, every mujtahid’s network represents a section of the broader Shiʿi clerical network. Analysing Khomeini’s network, I focus here particularly on centrality measures (i.e. degree, closeness, betweenness, and eigenvector), which signal the importance of individual nodes in the network. The calculated centrality measures of all nodes in Khomeini’s network are presented in Table 1.¹⁵ Following Khomeini’s position in this network, looking at a couple of other figures and their position within this network unveils interesting information about his contemporaries and colleagues.

Table 1. Khomeini’s network centrality measures.

	Degree		Closeness	Betweenness	Eigenvector
1	R. Khomeini (49)	R. Khomeini (1)		R. Khomeini (1)	R. Khomeini (1)
2	Mo. Mohaghegh (25)	A. Hairi (0.907)		A. Khoei (0.273)	Mo. Mohaghegh (0.599)
3	H. Borujerdi (19)	M.T. Khawnsari (0.849)		Mo. Mohaghegh (0.270)	A. Hairi (0.501)
4	A. Hairi (16)	M. B. Sultani (0.846)		A. Hairi (0.261)	H. Borujerdi (0.418)
5	A. Khoei (16)	H. Amoli (0.837)		M. Hujjat (0.210)	M. Hujjat (0.388)
6	Z. Araghi (15)	M. Hujjat (0.833)		Z. Araghi (0.194)	H. Montazeri (0.354)
7	M. Hujjat (15)	Mo. Mohaghegh (0.824)		M. B. Sadr (0.192)	Z. Araghi (0.346)
8	A. Isfahani (14)	M. H. Naini (0.817)		M. H. Naini (0.178)	M. B. Sultani (0.344)
9	M. B. Sultani (11)	M. Khomeini (0.815)		H. Amoli (0.174)	A. Isfahani (0.319)
10	M. H. Naini (11)	Z. Araghi (0.811)		H. Borujerdi (0.174)	H. Amoli (0.315)
11	M. Hairi (10)	H. Montazeri (0.811)		A. Isfahani (0.170)	M. Khomeini (0.312)
12	H. Amoli (10)	M. Lavasani (0.811)		M. Shahroudi (0.123)	M.T. Khawnsari (0.311)
13	M. Rouhani (10)	A. Isfahani (0.802)		M. B. Sultani (0.115)	M. Hairi (0.302)
14	H. Montazeri (8)	M. Pasandideh (0.802)		H. Rashti (0.111)	A. Meshkini (0.261)
15	M. B. Sadr (8)	M. Hairi (0.800)		M. K. Khorasani (0.107)	M. H. Naini (0.252)
16	H. Rashti (8)	N. Khalkhali (0.800)		H. Wahid (0.103)	M. Lavasani (0.249)
17	H. Bojnordi (8)	A. Meshkini (0.788)		M. H. Shirazi (0.102)	A. Javadi (0.237)
18	M.H. Tabatabai (8)	M. K. Khorasani (0.785)		M. Rouhani (0.095)	N. Khalkhali (0.229)
19	M.T. Khawnsari (7)	M. A. Shahabadi (0.785)		S. Sadr (0.087)	M. Gerami (0.225)
20	M. K. Khorasani (7)	R. Sadr (0.782)		M. Hairi (0.077)	Sh. Eshraghi (0.218)
21	H. Wahid (7)	A. Yathrebi (0.782)		H. Bojnordi (0.072)	A. Azari (0.218)
22	M. Hakim (7)	A. Javadi (0.779)		M.T. Khawnsari (0.069)	Mus. Sadr (0.214)
23	M. Khomeini (6)	Mu. Mohaghegh (0.779)		M. Hakim (0.067)	A. Yathrebi (0.211)
24	A. Khawnsari (6)	H. Borujerdi (0.773)		M. Ansari (0.050)	R. Sadr (0.208)
25	S. Sadr (6)	Mus. Sadr (0.773)		M. A. Shahabadi (0.045)	M. Pasandideh (0.201)
26	R. Sadr (6)	M. Gerami (0.771)		M.H. Tabatabai (0.043)	M. A. Shahabadi (0.201)
27	H. Hilli (6)	S. Sadr (0.768)		A. Javadi (0.042)	H. Bojnordi (0.200)
28	M. A. Shahabadi (5)	J. Sobhani (0.760)		H. Hilli (0.041)	M. K. Khorasani (0.194)
29	M. Lavasani (5)	A. KhatamYazdi (0.760)		H. Montazeri (0.039)	A. Ardabili (0.194)
30	A. Javadi (5)	M. Shahroudi (0.760)		M. Khomeini (0.035)	A. Khawnsari (0.193)
31	M. H. Shirazi (5)	H. Bojnordi (0.757)		R. Sadr (0.034)	M.H. Tabatabai (0.192)
32	M. Golpaygani (5)	A. Khomeini (0.749)		A. Milani (0.034)	A. Rafsanjani (0.191)
33	M. H. Milani (5)	A. Khamenei (0.749)		M. Golpaygani (0.033)	Mu. Mohaghegh (0.190)
34	N. Khalkhali (4)	H. Rasuli (0.749)		J. Imani (0.032)	Y. Sanei (0.187)
35	A. Meshkini (4)	M. MasjidShahi (0.744)		M. H. Gharawi (0.032)	M. Muqtadai (0.186)
36	M. Gerami (4)	M. Bojnordi (0.741)		A. Yathrebi (0.026)	M. Fazel (0.175)
37	M. A. Araki (4)	H. Wahid (0.731)		A. Khawnsari (0.022)	M. Motahhari (0.175)
38	Sh. Marashi (4)	A. Khawnsari (0.731)		M. MasjidShahi (0.022)	J. Sobhani (0.173)
39	Mus. Sadr (4)	A. Rafsanjani (0.729)		Mu. Mohaghegh (0.020)	M. Hashemian (0.172)
40	Mu. Mohaghegh (4)	M. Fazel (0.729)		Mus. Sadr (0.020)	H. Mazaheri (0.172)
41	M. H. Gharawi (4)	M. Motahhari (0.729)		M. Pasandideh (0.018)	H. Noori (0.172)
42	A. Milani (4)	M. B. Sadr (0.726)		M. H. Milani (0.018)	M. Emami (0.172)
43	J. Imani (4)	M. H. Shirazi (0.726)		A. Khamenei (0.017)	M. M.Rabbani (0.172)
44	F. Gharawi (4)	Sh. Eshraghi (0.726)		H. Qumi (0.017)	M. Momen (0.172)
45	H. Qumi (4)	A. Ardabili (0.726)		A. Meshkini (0.017)	N. Makarem (0.172)
46	M. Shubayri (4)	A. Azari (0.726)		M. Lavasani (0.015)	A. Khoei (0.169)
47	A. Khomeini (3)	M. Muqtadai (0.724)		A. KhatamYazdi (0.014)	H. Rasuli (0.167)
48	Sh. Eshraghi (3)	M. H. Milani (0.714)		J. Sobhani (0.012)	A. KhatamYazdi (0.164)
49	M. Pasandideh (3)	Sh. Marashi (0.705)		K. Heidari (0.011)	A. Khomeini (0.162)
50	A. Yathrebi (3)	A. Eslami (0.705)		J. Tabrizi (0.011)	A. Khamenei (0.162)
51	J. Sobhani (3)	I. Amini (0.705)		M. Bojnordi (0.011)	S. Sadr (0.159)
52	A. Rafsanjani (3)	H. Taghavi (0.705)		M. Gerami (0.011)	M. Hakim (0.157)
53	A. Khamenei (3)	F. Mahallati (0.705)		M. Muqtadai (0.011)	A. Eslami (0.154)
54	A. KhatamYazdi (3)	M. Mahdvai (0.705)		F. Gharawi (0.010)	I. Amini (0.154)
55	M. Fazel (3)	M.H. Tabatabai (0.703)		N. Khalkhali (0.008)	H. Taghavi (0.154)
56	M. Ansari (3)	Y. Sanei (0.703)		M. S. Hakim (0.008)	F. Mahallati (0.154)
57	M. Motahhari (3)	A. Safi (0.698)		M. Shubayri (0.007)	M. Mahdvai (0.154)
58	M. Muqtadai (3)	M. Hashemian (0.698)		H. Rasuli (0.007)	Sh. Marashi (0.152)
59	H. Rasuli (3)	H. Mazaheri (0.698)		Sh. Marashi (0.007)	M. Bojnordi (0.149)
60	Y. Sanei (3)	H. Noori (0.698)		Y. Sanei (0.006)	H. Wahid (0.148)
61	M. Bojnordi (3)	M. Emami (0.698)		M. H. Fadlallah (0.005)	H. Rashti (0.142)
62	A. Ardabili (3)	M. M.Rabbani (0.698)		M. Shamsuddin (0.005)	M. H. Milani (0.135)
63	A.S. Hakim (3)	M. Momen (0.698)		M. Fazel (0.005)	H. Hilli (0.134)
64	M. H. Fadlallah (3)	N. Makarem (0.698)		M. Motahhari (0.005)	M. A. Araki (0.132)
65	M. Shamsuddin (3)	A. Khoei (0.696)		A. Ardabili (0.005)	M. MasjidShahi (0.131)
66	A. Safi (3)	M. Hakim (0.691)		M. A. Araki (0.004)	A. Safi (0.122)
67	K. Heidari (3)	M. Golpaygani (0.689)		M. Bahjat (0.004)	M. Shahroudi (0.122)
68	M. S. Hakim (3)	J. Imani (0.680)		A. Khomeini (0.004)	M. Shubayri (0.117)
69	M. K. Yazdi (3)	H. Rashti (0.674)		A. Safi (0.004)	M. H. Shirazi (0.115)
70	A. Azari (3)	H. Taheri (0.667)		Sh. Eshraghi (0.004)	M. Golpaygani (0.113)
71	M. MasjidShahi (2)	H. Rasti (0.667)		A. Azari (0.004)	J. Imani (0.111)
72	M. Hashemian (2)	Mj. Tehrani (0.667)		L. Safi (0.003)	H. Taheri (0.108)
73	A. Eslami (2)	Mz. Tehrani (0.667)		M. K. Yazdi (0.003)	H. Rasti (0.108)
74	I. Amini (2)	M. Malakuti (0.667)		A. Shubayri (0.003)	Mj. Tehrani (0.108)
75	H. Taghavi (2)	S. Khalkhali (0.667)		J. Alavi (0.002)	Mz. Tehrani (0.108)
76	H. Mazaheri (2)	A. Nateq (0.667)		A. Sistani (0.002)	M. Malakuti (0.108)
77	H. Noori (2)	M. A. Araki (0.665)		A. Rafsanjani (0.001)	S. Khalkhali (0.108)
78	F. Mahallati (2)	K. Heidari (0.661)		A. Eslami (0.001)	A. Nateq (0.108)
79	M. Emami (2)	M. Shubayri (0.660)		I. Amini (0.001)	M. Rouhani (0.097)
80	M. Mahdvai (2)	H. Hilli (0.657)		H. Taghavi (0.001)	F. Gharawi (0.096)
81	M. M.Rabbani (2)	A. Milani (0.646)		F. Mahallati (0.001)	M. B. Sadr (0.096)
82	M. Momen (2)	J. Tabrizi (0.646)		M. Mahdvai (0.001)	A. Milani (0.080)
83	N. Makarem (2)	F. Gharawi (0.640)		M. T. Shirazi (0.001)	N. Mustanbet (0.079)
84	L. Safi (2)	N. Mustanbet (0.638)		M. Hashemian (0.001)	A. Shubayri (0.068)
85	J. Alavi (2)	M. H. Gharawi (0.636)		H. Mazaheri (0.001)	M. H. Gharawi (0.068)
86	B. Najafi (2)	A. Shubayri (0.634)		H. Noori (0.001)	J. Tabrizi (0.067)
87	N. Mustanbet (2)	M. Rouhani (0.628)		M. Emami (0.001)	M. K. Yazdi (0.064)
88	J. Tabrizi (2)	M. K. Yazdi (0.621)		M. M.Rabbani (0.001)	A. Mohaghegh (0.064)
89	A. Shubayri (2)	M. Bahjat (0.609)		M. Momen (0.001)	M. S. Hakim (0.063)
90	M. Bahjat (2)	M. S. Hakim (0.605)		N. Makarem (0.001)	L. Safi (0.059)
91	M. Shahroudi (2)	H. Qumi (0.600)		H. Taheri (0.000)	K. Heidari (0.058)
92	M.M. Sadr (2)	L. Safi (0.600)		H. Rasti (0.000)	M. H. Fadlallah (0.058)
93	M. T. Shirazi (2)	S. Larijani (0.591)		Mj. Tehrani (0.000)	M. Shamsuddin (0.058)
94	A. Sistani (2)	A.S. Hakim (0.589)		Mz. Tehrani (0.000)	A.S. Hakim (0.052)
95	H. Taheri (1)	A. Mohaghegh (0.584)		M. Malakuti (0.000)	H. Qumi (0.049)
96	H. Rasti (1)	M.M. Sadr (0.579)		S. Khalkhali (0.000)	M. Bahjat (0.046)
97	Mj. Tehrani (1)	M. H. Fadlallah (0.575)		A. Nateq (0.000)	A. Sistani (0.040)
98	Mz. Tehrani (1)	M. Shamsuddin (0.575)		N. Mustanbet (0.000)	B. Najafi (0.039)
99	M. Malakuti (1)	M. Ansari (0.564)		S. Larijani (0.000)	M.M. Sadr (0.038)
100	S. Khalkhali (1)	J. Alavi (0.561)		A.S. Hakim (0.000)	S. Larijani (0.035)
101	A. Nateq (1)	M. T. Shirazi (0.557)		A. Mohaghegh (0.000)	M. Ansari (0.034)
102	A. Mohaghegh (1)	A. Sistani (0.545)		M.M. Sadr (0.000)	J. Alavi (0.033)
103	S. Larijani (1)	K. Hairi (0.533)		K. Hairi (0.000)	S. Rouhani (0.024)
104	M. Sanad (1)	B. Najafi (0.526)		B. Najafi (0.000)	M. Fayadh (0.024)
105	S. Rouhani (1)	S. Rouhani (0.517)		S. Rouhani (0.000)	M. T. Shirazi (0.021)
106	M. Fayadh (1)	M. Fayadh (0.517)		M. Fayadh (0.000)	M. H. Najafi (0.018)
107	H. Kuhkamarai (1)	M. H. Najafi (0.505)		M. H. Najafi (0.000)	M. Sanad (0.015)
108	K. Hairi (1)	M. Sanad (0.479)		M. Sanad (0.000)	K. Hairi (0.014)
109	M. H. Najafi (1)	H. Kuhkamarai (0.441)		H. Kuhkamarai (0.000)	H. Kuhkamarai (0.006)

Take, for instance, Ayatollah Broujerdi, who despite his high degree score (number of connections) has a relatively low closeness score (interconnectedness with all the parts of the network) at 0.773. This indicates that although he is a well-connected figure, he is embedded in a cluster which is peripheral to the rest of the network. In other words, although Broujerdi’s scholarly, familial, and social links form a significant cluster within Khomeini’s network, he is not consequential for tying the network together. In contrast, Sayyid Muhammad Taqi Khawnsari is a node with low degree but relatively high closeness ( = 0.849). His scores denote that although he has limited connections with other figures, the nodes to which he is tied are themselves key players within the network. In other words, were resources to flow from Khawnsari, they would likely reach more parts of the network than if the same resources originated with Broujerdi.

Ayatollah Khoei’s centrality measures also deserve attention. In Khomeini’s network, Khoei exhibits a very rare combination of high betweenness ( = 0.273) and relatively low closeness ( = 0.696).

As mentioned earlier, betweenness centrality represents a node’s brokerage position within a given network. Khoei’s high betweenness score despite his low closeness indicates that he monopolises the ties between a small group of people in the network. It shows that to reach Khomeini, Khoei is the main node for his surrounding cluster of figures. Sayyid Mahmoud Hashemi Shahroudi’s case – with high betweenness and low degree – is also remarkable, suggesting that he has few ties, but that the nodes to which he is linked are central in the network.

Finally, a quick review of eigenvector centrality singles out four key figures, who despite high scores are not among the known Shiʿi marajiʿ: Sayyid Mohammad Mohaghegh Damad (d. 1969), Aqa Zia Araghi (d. 1942), Sayyid Muhammad Baqir Sultani Tabatabai (d. 1997), and Mirza Hashem Amoli Larijani (d. 1993). In general, their relatively high eigenvector scores indicate that their status within Khomeini’s network is based not only on the high number of ties they have, but also on the high centrality of the people to whom they are linked. The former two are among the most prestigious teachers of Qum and Najaf seminaries. Despite training many students who later became marajiʿ, neither had the opportunity to hold the position themselves.¹⁶ Their relatively high eigenvector centralities simply indicate how instrumental their students were in Khomeini’s network. The high scores of Sultani and Amoli, apart from their scholarly status, partly stem from their family relationships within the broader network.

Although a thorough analytical examination of the clerical ties – especially at network levels – demands mapping the exhaustive clerical network, the case of Khomeini provides a clear example of how a quantitative approach and network analytic methods can further inform our knowledge about clerical elites in the contemporary Middle East.

Analysing the Shiʿi clerical network: where to start?

Recent decades have witnessed a remarkable evolution of analysis of the social world through networks. This is partly due to the convergence of the data revolution, advances in network-science methodology, and the development of computational modelling and network software across disciplines (Barabási 2016, 26). This approach has the potential to add new perspectives to the qualitative data analysis already being done on this community.

For instance, analysing clerical authority through a network approach means arguing that position within the network matters more than personal attributes or resources. The highest-ranking and most influential clerics (i.e. marajiʿ) tend to be situated in positions that enable them to exercise control by virtue of their connections, while lower-ranking clerics reside in positions that constrain their ability to create connections within the community of followers. Indeed, every mujtahid possesses his own scholarly merits, family prestige, and social resources; but his ability to deploy these resources is determined by his position within the broader clerical network. A scholar on the periphery of the network, regardless of his personal virtues, does not have the ability to mobilise resources and support.

Another real boon for network analysis is the ready availability of immense amounts of data surrounding Shiʿi religious figures. Compiling biographies of clerics is an established tradition among Shiʿi, with a history going back to the early ages of Islam.¹⁷ Biographers have documented the scholarly lives – and sometimes sociopolitical activities – of all major clerical figures. Another source in tracing inter-clerical ties, especially scholarly and social links, is to refer to the ijāza system. Ijāza is a licence issued by an established marjaʿ that bestows authority on its holder (Kondo 2009). Two types of ijāza exist: (1) ijāza fi al-ijtihād (licence in ijtihād), and more recently (2) ijāza fi al-omour al-hesbiah (permission in managing non-litigious affairs). Tracing these lines of ijāza can be useful for understanding scholarly and social links between clergy. In the case of the former type, many mujtahids rely on licences they have received from their well-known teachers in order to issue juridical rulings, to perform ijtihād. The latter ijāza corresponds more with a social link between a marjaʿ and his representatives. Through issuing this ijāza, a marjaʿ gives power of attorney to the licensee to collect religious taxes from the community of followers on his office’s behalf; an evident indication of a social link. More recently, the personal websites of Shiʿi marajiʿ have become prolific sources of data. Almost every contemporary Shiʿi marjaʿ has a personal website that he uses as a means of interacting with his followers, and which also contains extensive materials, such as his biographies and writings. These provide a promising resource for mapping a marjaʿ’s scholarly, family, and social links.

Going forward, qualitative and quantitative methods should be employed to complement one another in the analysis of the Shiʿi clergy. In other words, the analytical network approach presented here should be treated as an auxiliary method to further understand the clergy, to excavate patterns in its genealogy, and to predict its future positioning. Patterns that are identified in the data can lead to new ways of looking at qualitative sources and vice versa.

Let us turn again to the limited case study of Khomeini’s network to show what I mean in practice. Analysis of this network clearly indicates that, despite what Shiʿi sources and many observers suggest, a marjaʿ is not necessarily the ‘most learned’ cleric in a given time. If this were the case, nodes with more scholarly links – such as Mohaghegh Damad in our network – would inherently be perceived as the most learned and, consequently, would have a higher likelihood of being recognised as a marjaʿ. This would suggest that other inter-clerical ties, including social and familial links, are as important as scholarly links. Is it possible to find qualitative evidence to support this hypothesis? In a letter to his son written in 1971, Ahmad, Ayatollah Khomeini testifies to the importance of his representatives (social links) in the legitimation of his religious authority. He writes:

It has been a while that I feel from your letter that you, unaware, have been instigated to limit or expel my representatives […] Those which I have chosen, whether in Najaf or Tehran [i.e., Khalkhali and Lavasani], occurred after long associations with them. If we ourselves put them aside, it injures me and not the individuals concerned […] Again by your harsh tone it is apparent that you have been agitated a great deal: know that my representative in Tehran does not act contrary to my wellbeing. (Khomeini 2010, Vol. II: 342–3)

Khomeini’s belief in the importance of social links in consolidating one’s marjaʿiyya has also been discussed by number of scholars and observers, albeit in a looser fashion than what the quantitative network approach is able to reveal (Moin 2009, 70). Thus, by combining qualitative and quantitative analysis, we can make the argument that for an established cleric to become a marjaʿ, he needs more than the scholarly acumen singled out by the Shiʿi primary jurisprudential text and a vast number of academic studies.

Another hypothesis that could be tested in this way is that nodes with the highest eigenvector centrality are the most likely to become marajiʿ. This is based on the assumption that a cleric’s ties with other prominent clerics within the network at a given time inevitably affect his chances to ascend the clerical hierarchy, and ultimately to assume the position of a marjaʿ. Besides these node-level observations, there are some measurements which could be applied at network-levels. This would help us understanding the clerical elites’ sociopolitical behaviours in a more collective way.

For instance, at the same time that Khomeini was rising as a marjaʿ in exile, teaching at Najaf seminary, a group of his students was forming the Society of Seminary Teachers of Qum, a clerical body which played a substantial role promoting his status amongst Iranian Shiʿi and consolidated his political ambitions at the time of 1979 revolution. This group of clerics, mostly tied to the Ayatollah through scholarly links, comprised a close-knit cluster. Later on, they held high-rankings positions in the Islamic Republic. Analysing this cluster and delving into their collective behaviour through network analysis would unearth patterns of clerical relationships, which helps us to understand future orientations in similar situations.

Ultimately, mapping the entire Shiʿi clergy would provide an unparalleled pool of information about the sociopolitical properties of the group, at a network level. We could see quantitatively whether changes in political context led the clerical elite to adopt different postures vis-à-vis the state and the community of believers: whether, for example, they support their colleagues when facing contentious circumstances by solidifying their internal connections.

It is evident that analytical network thinking can contribute to studying the Shiʿi clerical milieu. In order to promote the use of such methods, one should follow a three-stage network approach. First, inter-cleric connections should be mapped. This involves, amongst other things, a process of thorough qualitative examination of available databases (e.g. Shiʿi biographical encyclopaedias, records of ijāza, marajiʿ’s websites, etc.) to extract data on scholarly, social, and familial ties. Second, it is necessary to visually map the clerical network. This network, like all social networks, is a structure in process – one that evolves as the relations among members of the clergy evolve. Thus, to compare across time, multiple networks maps should be created. And third, data analysis is needed at both the micro and macro levels. The micro node-level analysis, as showcased above, mainly involves centrality measures. The network-level analysis consists of statistical procedures which generate, among others, measures of density, centralisation, and reciprocity (Jackson 2010, 20–51).

Conclusion

Networks form the backbone of Shiʿi clergy and their sociopolitical interactions within the modern Middle East. To effectively understand their role in the region’s rising political Shiʿism, it is crucial to delve into the nodes and linkages that comprise the clerical network. These elements shape the clergy’s strategic orientations, mould their political postures, and give rise to religious leaders. The clergy has been described by many observers as a loosely structured yet complex system, where the utility of network analysis becomes evident in deciphering the sometimes-latent ties between figures. To interpret the complexity, we need to unpack the nodes and links of the network and start focusing on the dynamics that pass through their interconnectivities.

Although the Shiʿi clergy emerged as a focus of western scholarship over six decades ago, a systematic approach to studying and explaining this community’s processes and practices remains elusive. While it is true that social network analysis has been a field of study for almost a century now (Wasserman and Faust 1994, 10), and scholars have long pondered the social aspects of religious networks (Everton 2018), an analytical application of this theory to comprehend the Shiʿi clergy is a frontier yet to be explored.

This article, through its case study of Ayatollah Khomeini, has sought to pioneer such exploration and to stimulate debate about how employing social network analysis can generate a more profound understanding of the Shiʿi clerical milieu. Even though one cannot claim that future trends within the clergy are predictable, modelling past dynamics with the quantitative tools of network analysis enhances our understanding of the underlying structures that form the backdrop from which Shiʿi marajiʿ emerge, relationships are built, and resources are diffused. Numerous policy briefs, academic articles, and scholarly books have sought to address significant queries such as who are the potential successors to Ayatollahs Sistani and Khamenei in Iraq and Iran, respectively. However, the arguments that underpin these works often fail to withstand rigorous analytical scrutiny. Frequently, these narratives elucidate the process through which mujtahids ascend to the status of marjaʿ by recounting historical incidents or invoking religious texts, but they lack a systematic methodology that predicts the emergent marjaʿ. Incorporating the toolkit provided by social network analysis – as exemplified by the case of Ayatollah Khomeini in this article – can shed substantial light on these pressing questions. By implementing meticulous analysis of data rooted in both historical and contemporary contexts, this approach facilitates a more authoritative comprehension of the Shiʿi clergy.

Furthermore, looking at the Shiʿi clergy through the prism of a network approach enables us to address broader questions. By mapping the network structure and focusing on its component parts we can capture their dynamic interplay, and the impact of these dynamics on surging sectarian conflicts and regional rivalries. We can understand how a decision made in Iraq will be received in Iran, and then how its broader consequences will impact the entire region. Shiʿi clerical elites are among the most influential political actors in the contemporary Middle East, and in order to fully understand their individual and collective behaviours, scholars need to start mapping them, like early cartographers. This task, to map the network node by node and to establish their interconnecting links, involves a painstaking process at the beginning, but the result will be worthwhile. When we have a map, it is much easier to see what we should be looking at and where we need to go.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research received no external funding.

Notes on contributors

Mohammad R. Kalantari

Mohammad R. Kalantari is a Lecturer in International Relations of the Middle East and the Director of the Centre for Islamic & West Asian Studies at Royal Holloway University of London. His research centres on the interplay of regional doctrines, elite ideologies, and political Islam. His latest book, The Clergy and the Modern Middle East: Shiʿi Political Activism in Iran, Iraq, and Lebanon (Bloomsbury, 2021), examines the role of Shiʿi clerics in contemporary politics. Through interviews with key figures, the book scrutinises pivotal events like Iran’s 1979 Islamic Revolution, the 2003 Iraq War, and the 2006 war in Lebanon.

Notes

1. The research in this article is based on historical figures, centring on Ruhollah Khomeini (1902–1982) and his contemporaries. Because this approach did not require current research engagement with human subjects, no institutional ethical approval was necessary.

2. In this article, (Shiʿi) clerics, ulama, and mujtahids are used interchangeably and all in reference to religious elites belonging to the mainstream Usuli School. For background on Usulis and their counterpart Akhbaris, see Cole (1985).

3. The prominent Shiʿi jurist and thinker Morteza Motahhari (d. 1979) notes this contrast by acknowledging marajiʿs independence from the state yet indicating that their dependence on the financial support of their followers may compromise their independence and lead to demagoguery among the clergy’s rank and file. For more details, see Motahhari (2003, 283–89).

4. For an overview of these transnational networks, see Louër (2008).

5. Two central tenets that underpin much of the research in social network analysis are that the structure of a network significantly influences collective outcomes; and second, that an actor’s position within the network impacts individual outcomes. For an in-depth discussion of these principles, see Robins (2015, chapter 2).

6. See, for example, Padgett and Ansell (1993), which investigates how the leveraging of central positions in a network of familial intermarriages, economic relationships, and political patronage could consolidate political and economic power.

7. Light and Moody (2020) propose a conceptual model wherein the significance of social networks can be framed into two principal categories: Connectionist approaches, which concentrate on the links among nodes and the role these connections play in facilitating diffusion, and Positional approaches, which prioritise the understanding of how relation patterns establish distinct roles within networks, over the emphasis on the structure of paths linking nodes across a population (see Light and Moody 2020, 19–22).

8. It is worth noting that this list is not exhaustive of the recognised relationships within this context. The general conviction is that these highlighted attributes encapsulate a majority of other familial ties. This includes, for instance, brother-in-law relationships, which can be traced back by linking the two individuals through their mutual father-in-law. Therefore, while the primary focus may lie on direct links of familial lineage, an intricate network of broader family ties is also acknowledged and mapped out in the clerical network as proposed here.

9. An undirected network is a network in which links indicate a two-way relationship. For more information, see Barabási (2016, 51).

10. For an overview of early formalisation and evaluation of the centrality concept within the context of social networks, see Freeman (1978).

11. The notions of ‘bonding’ and ‘bridging’, as outlined in the social network theory literature, underline the potential value derived from differing types of social connections. See Burt (2004) and Everett and Valente (2016).

12. Shahab al-Din Eshraghi was the grandson of Mirza Muhammad Arbab (d. 1922), a wealthy landowner in Qum with close ties to Shaykh Abd al-Karim Haʿiri (see Jawahir-Kalam 2003, Volume III: 1796–1803).

13. Compiled in two volumes, Taḥrīr al-Wasīla is Khomeini’s jurisprudential comments on the Wasīla al-Najāt of Sayyid Abu al-Hasan al-Isfahani (d. 1946). See Tehrani (1983, Vol. XXV: 85).

14. In mapping the clerical network, every distinct tie (i.e. scholarly, family, social) between two nodes is shown with a link with weight = 1. Therefore, in Figure 3 scholarly ties with students who are also linked to Khomeini familially (i.e. his son Mustafa Khomeini) or socially (e.g. Ali Akbar Hashemi Rafsanjani) are represented as doubly or triply weighted links.

15. The table presents normalised values of closeness, betweenness, and eigenvector centrality measures to facilitate comparisons across different networks. The data within the table are organised based on the scores obtained under each specific category of centrality measure.

16. For instance, Ayatollah Javadi Amoli, a student of Mohaghegh Damad, believes that his teacher was not recognised as a marjaʿ, despite being one of the most established figures among his colleagues, partly due to his lack of publications. Aqa Zia Araghi, similarly, has few publications. For the full account, see Kazemeini (2019, 60).

17. Two of the most famous contemporary Shiʿi biographers are Muhsin al-Amin (d. 1952) and Aqa Bozorg Tehrani (d. 1970) who together have compiled and documented biographies of more than 12,000 clerical figures. See Al-Amin (1960) and Tehrani (1991).