THE NETWORK OF ISMAʿILI CASTLES IN THE ALA-MUT REGION: POWER AND GOVERNANCE
Seyedhamed Yeganehfarzand, University of Victoria
Abstract
In 1090, Hasan-i Sabbah (1050s –1124), the mysterious leader of the Nizari Ismaʿilis in Persia, obtained control of Alamut Castle, one of the major existing castles in the northern part of Iran, and reinforced it as the headquarters of his activities against the Seljuq government. The Nizari Ismaʿilis gradually became a very influential community within the political and intellectual history of the Islamic world until the fall of Alamut in 1256. The principal strategy that helped them to resist their enemies and to survive in a hostile milieu was the establishment of networks of castles within inaccessible regions.
This research concerns the network of Ismaʿili castles in the Alamut region. Ex-amining the distribution of the castles in the region, the paper questions the current schol-arly assumption that the network of the castles functioned as part of a defensive system, which separated the Ismaʿili territory from their enemies with a linear border. Proposing several examples as evidence, through a detailed examination of historical texts, this paper takes an alternative position, suggesting that the castles should be considered as centers of power, which formed important spheres of influence for their surrounding environment; they were also employed as effective units of governance in the expansion of the Ismaʿili dominions.1
In 1090 Hasan-i Sabbah (1050s – 1124), the charismatic leader of the Nizari Ismaʿilis in Persia, obtained control of Alamut Cas- tle (Fig.1), one of the major existing castles in the northern part of Iran, and reinforced it as the headquarters of his activities against the Seljuq government (1037-1194). During the so-called Alamut period (1090-1256), the Ismaʿilis, under the leadership of Hasan-i Sabbah and his succes-sors, succeeded in capturing, reinforcing and erecting about 200 large and small fortresses in different regions of Persia (including Rudbar, Alamut, Taliqan, Qumis, Quhistan, and Arrajan) and the region of Jabal al-Bahra in Syria (Fig. 2). They also stabilized their power as autonomous states in these regions. The principal strategy that helped them to resist their enemies and to survive in a hostile milieu was the establishment of networks of castles within inaccessible regions.
The Ismaʿili castles have been more or less studied as individual structures. However, the relationship between the castles as a network has been considerably overlooked. Focusing on this lesser known aspect of the Ismaʿili fortifications, this study makes an initial probe into the
network of the castles from the Ismaʿilis’ own perspective during the Alamut period. This purpose will be pursued through an examination of the distribution of the Ismaʿili castles in the Alamut region alongside a detailed study of principal historical texts that deal with Ismaʿili history. I will dispute the common notion of the network of the castles as part of a linear defensive system that separated the Ismaʿili territory from their enemies, as has been proposed by previous scholars. In contrast, I will suggest that the castles can be considered as the centers of power that formed the spheres of influence on their surrounding environment; of equal importance, these structures were applied in the expansion of the Ismaʿili dominions as offensive structures.
Figure 2: The approximate location of the main Ismaʿili regions in Iran and Syria (11th -13th Century) 1-Rudbar, 2-Alamut, 3-Quimis, 5-Quhistan, 6-Arrajan, 7-Jabal al-Bahra Google Earth, Google, Cnes/Spot2012 Data SIO, NOAA, U.S.Navy, NGA, GEBCO. Image U.S Geological Survey.
The Alamut region has been selected as a case study for three reasons: first, it was the first region that was captured by Ismaʿilis and was always considered to be a ‘template’ for the Ismaʿili movement in other regions; second, it was known as an inviolable Ismaʿili territory from the rise of Hasan-i Sabbah to the fall of Alamut (1090-1256); and third, the existing primary and secondary sources provide more material for the study of the castle in the Alamut region in comparison to other Ismaʿili centers.
Current scholarship on the network of Ismaʿili castles:
All the previous studies on the Ismaʿili castles have pointed to the relation ship between them as a network of defensive structures; however, they have never discussed this relationship in detail. The main focus of many of these studies is on the individual castles and their architectural features. Peter Willey, who carried out the most extensive research on the Ismaʿili castles for his book, Eagle’s Nest, Ismaʿili Castles in Iran and Syria, briefly discusses the strategic and tactical considerations of the Ismaʿilis in choosing a location for the establishment of a castle. Regarding the relationship between the castles as a network, he simply considers them a ‘chain’ or ‘line’ of formidable structures that defended the borders of the Ismaʿili state.
A relatively similar interpretation of the network of Ismaʿili castle is suggested by Wolfram Kleiss in Assassin Castles in Iran. Kleiss generally discusses the area of Ismaʿili influence through the distribution of castles in different parts of Iran. He suggests a schematic model of the relationship between the castles in different regions by displaying the location of only twenty castles on a map of Iran, and the arrows connecting the Ismaʿili regions together. According to this model, Alamut Castle was the political, military, and religious center of the Ismaʿili state; the other regions were linked to Alamut via the castles distributed in each region. For example, he considers the castles of Amamih and Firuzkuh as two links in the chain of castles towards the east, from Alamut to Khurasan, or the castle of Savih as a key connection towards the south, from Alamut to Isfahan and Fars. Furthermore, he mentions the possible relationship between Alamut and the Syrian castles through sites in the Azerbaijan region, such as the castles called Qalʿih Zahhak and Qalʿih Dukhtar. Although Kliess’ model is relatively vague and needs more detail to become acceptable, it clearly reveals his approach to the castles as a line of connected defensive structures. As a result of this analysis, at the end of his essay, he emphasizes a need for identification of more individual cases to fill the gap between the regions and complete this chain of castles.
A more recent model of the network of the Ismaʿili castles is suggested by Hamidih Chubak, exclusively for those castles located in the Alamut region (Fig. 3). Chubak considers the Alamut region to be an integrated and well protected area, with the centrality of Alamut Castle asthe capital (dāral-mulk) of the region. According to her model, the surrounding mountains functioned as a natural rampart around the region and defined its frontiers. Furthermore, the castles are considered the ‘gates’ which protected the Alamut valley against possible attack, through controlling the main access routes to this area.
Figure 3: The model of the interrelation between castles in the Alamut region suggested by Hamidih Chubak. After Chubak, “Alamut- Part II,” 120.
None of the above-mentioned studies have been able to suggest a clear picture of the network of the Ismaʿili castles, as a result of the lack of essential information regarding all the individual castles. However, they all share one common idea: the significant role of the castles in the protection of the Ismaʿili borders as a ‘chain’ or ‘line’ of defensive structures. In other words, these studies consider the castles to be bordering points, creating a linear boundary around the Ismaʿili regions and separating them from surrounding areas. This idea about the network of castles is not limited to the Ismaʿili fortifications. For instance, modern scholarship on the Crusader history suggests a similar idea about the distribution of the castles and their role in the protection of the Crusader boundaries.
Ronnie Ellenblum, in Crusader Castles and Modern Histories, criticizes the models suggested by previous scholars such as Rey, Prutz and Deschamps and their interpretation of the spatial distribution of the Crusader castles. Generally, these models considered all the castles to be part of a defensive system that separated the Crusader settlements from their enemies by means of a linear border. Ellenblum believes that the development ofwell demarcated border lines was a product of the establishment of modern states and influenced by the growth of the modern disciplines of geography and cartography. Examining the concepts of borders and frontiers in the medieval historical sources, he suggests that the notion of ‘linear border’ was meaningless to medieval people. Therefore, he refutes the idea of the existence of a linear connection between the Crusader castles in the protection of the frontiers of the Crusader states. Similarly, one might ask to what extent this would be true in the case of Ismaʿili castles.
Unfortunately, there is no explicit reference in the historical sources either to the criteria employed by the Ismaʿilis when selecting a location to establish their castles, or to their own understanding of the network of their castles. Some indirect hints do, however, exist. Regarding this issue Juwayni (d. 1283), an official at the Mongol’s court who accompanied Hulagu (r.1256-65) during his campaign against the Ismaʿilis, mentioned:
Hasan [-i Sabbah] was immoderate in releasing the places which were adjacent to Alamut or close to it, and wherever was possible did it by daʿwa guile and if not by murder and affront and assault and bloodshed and warfare, and captured those castles which were feasible and wherever he found a suitable rock for building, he built a castle upon it.
Juwayni does not explain the features of a ‘suitable rock (sangī mīyāft ki banārā mīshāyist)’ upon which to establish a castle. However, the account shows that capturing and building the castles was one of the main strategies of the Ismaʿilis for the extension of their territories from the beginning of the Alamut period. In addition to textual sources, the study of the archeological remains of the castles in their environmental context would provide valuable information about the network of Ismaʿili fortifications.
The distribution of the Ismaʿili fortifications in the Alamut region:
The Alamut region is an enclosed mountainous terrain with an eastwest extension in the Central Alburz Massif. To the north across the mountains lies the Caspian Sea coast. Southwest is the Qazwin Plain, the southeast is the Taliqan region, and to the west is the area of Southern Tarum. The surrounding mountains, with an average height of about 3,600 m on either side of the Alamut Valley, have turned the territory into an impenetrable refuge.
During the Alamut period, Ismaʿilis established several strongholds in the region. The topographical features of the mountains provided an ideal place for the erection of castles and contributed to their defensive qualities. The inaccessibility of the valley, particularly over the cold seasons when the passes fill with snow, was another factor which ensured safety for the castle dwellers during winter. According to Mustawfi (d.1349), the historian and geographer of the Ilkhanid period, there were more than fifty castles in the Alamut and Rudbar regions, of which Lamsar, Maymun Dizh, and Alamut Castle were the best (bihtarīn-i ān).
Recent research on the Ismaʿili fortifications has identified thirtythree large and small castles, forts, and watchtowers in the Alamut region (Fig. 4 and Table.1). However, the possibility of discovering more Ismaʿili sites through further survey work in the area should not be overlooked. The most common type of Ismaʿili structures in the region are the castles located on the summits of mountains. This positioning was, in fact, the principal defensive strategy of the Ismaili castles. These castles were established on the top of the rocks, which naturally are separated from their surrounding environment and are usually accessible from only one side. This specific location forms a passive defense for the castles and contributes to their defensive qualities. It is worth mentioning that most of the Ismaʿili castles show evidence of significant construction phases long before the Ismaʿilis took control of them. In these cases the Ismaʿilis modified and strengthened the castles, based on their requirements.
Figure 4: Schematic map of the distribution of fortifications in the Alamut region Google Earth, Google, Landsat 2014
In addition to these two types of fortifications, there are traces of very small individual structures related to the Ismaʿili period in some parts of the valley. These structures, which are located in relatively high positions in the middle of the valley, once dominated their surrounding environment. Their remnants suggest that they were small stone buildings with a simple plan. Although all of these structures have been largely destroyed over time, their small size and particular location raises the possibility that they were used as watchtowers and acted as communication elements within the network of the castles.
According to the archaeological surveys, all the sites were occupied during the Alamut period except for Qatran Shah castle in the extreme southwest of the region. This castle was probably active in a period between the ninth and eleventh centuries and was controlled by a local governor. The archaeological evidence indicates that it had been abandoned sometime before the Alamut period and was never occupied by the Ismaʿilis. In this regard it is considered to be a valuable case for the study of the evolution of building techniques in the region, before and after the presence of the Ismaʿilis. In this paper, I will briefly refer to it in the analysis of the relationship between the distribution of the castles and the defense of the Ismaʿili borders.
Challenging Current Theories on the Defense of the Ismaʿili borders:
Most of the secondary sources on the network of the Ismaʿili castles emphasize the linear connection between them. They suggest that these connections established defensible boundaries of the Ismaʿili regions through protection of the major routes. To test this claim and to understand what appears to be the Ismaʿilis’ own perception of the distribution of their castles and the geographical landscape in which they lived, one can examine the relationship between the location of the castles and the main routes leading to the Alamut region.
I have applied three sources to identify the possible communication routes during the Alamut period. The first are historical accounts that concern the history of the Ismaʿilis, as well as the campaigns of the Seljuqs and the Mongols to the Alamut region. The second sources are the accounts and maps provided by the scholars (including Freya Stark, Wladimir Ivanow, Manuchihr Sutudih, and Peter Willey ) who visited the valley22 23 24 before the construction of motor roads. Since they travelled through the access routes to the Alamut region which were used over the course of centuries, we can assume that, at least in some cases, they took the same historical routes that had been taken during the Alamut period.
In addition, I have examined satellite photographs as well as the geographical conditions of the terrain during my fieldwork in order to identify other possible locations through which the topography would allow access to the region. Using this information, I have mapped the potential historical routes to the region (Fig. 5). While one cannot assert that all of these routes existed or were taken for communication during the Alamut period, it can be claimed that the map displays all the possible
Figure 5: Schematic map of possible historical routes to the Alamut region and the distribution of the castles Google Earth, Google, Landsat2014
Ismaʿilis to stop building the castles, implies the offensive function of communication routes to the region, and thus provides a proper base upon which the relationship between the location of castles and the main access routes to the Alamut region can be fruitfully examined.
The distribution map of the fortifications, overlaid with this map, confirms that a great number of the castles were located at a considerable distance from these routes. Thus it would be difficult to think of any possible relationship between most of these castles and controlling the roads. In addition, the study of the historical texts suggests that almost all the attacks on the Alamut land have been made from the southern side of the valley. While one might expect the presence of a large number of the castles in the southern part of the valley (built with the purpose of defending the region’s border), the map shows fewer number of the castles in the southern part of
the valley in comparison with the northern part. In this regard, the case of Qatran Shah castle is even more notable. The castle is located very close to a route which enters the valley from the south. But according to the archeological surveys, there is no evidence of occupation of this site by Ismaʿilis, whereas if defending the bordering points was considered one of the main strategies for choosing the location of the castles, Qatran Shah castle, as an existing site, could certainly be modified and reused by the Ismaʿilis (incommon with many other castles in the region). Therefore, one can suggest that the protection of the borders was unlikely to be the main motivation for the establishment of the castles in the Alamut region.
However, there are a few cases in which the main function was to control transportation to the valley due to their proximity to the major routes. For example, the location of three castles, Shir Kuh, Burj-u Barak,and Dizh Kamar (numbers 17, 18, and 19), around the Du-āb area, which is one of the main entrances of the Alamut region, shows the important role of these castles in controlling the entry to the valley. Juwayni’s account of Girdkuh castle in the Qumis region, which was strategically an important site and controlled one section of the great Silk Road, confirms the above mentioned function of the castles. Juwayni says that:
.. Sultan (Sanjar, r.1097-1157) feared and became willing to make peace with them (the Ismaʿilis). He (the Sultan) accepted that travellers paid their tax to them (the Ismaʿilis) at the foot of Girdkuh castle and this covenant has remained true until now (i.e. the time of Juwayni’s writing)…
However, this fact does not mean that the castles were mere defensive structures established exclusively to protect the borders. If the Ismaʿili castles were not built to protect the borders of Ismaʿili territory, one may wonder what would be an alternative explanation for the distribution of these fortifications in the Alamut region. One possible answer might be obtained through examination of the relationship between the distribution of the castles and the human settlements in the region, and we turn now to examine the idea of the Isma’ili castle as a center of power and offensive strength.
The Ismaʿili Castles as centers of power:
Since water supply is a crucial factor in the establishment and development of permanent human settlements, all the villages in the region are located close to water sources (rivers, streams, and springs). Today, there are about 200 large and small villages in the Alamut region. The architectural patterns of the villages follow the features of the local architecture of the region. Therefore, it is difficult to attribute the formation of these villages to a specific period on the basis of the existing architectural qualities. However, the extant archaeological sites close to many of the villages are testimonies to the existence of human settlements in these areas during the Alamut period.
The overlay of the distribution maps of the settlements and the fortifications reveals that there are certain relationships between the location of the castles and the distribution of the villages (Fig. 6). There is at least one castle close to all of the north-south valleys in which a couple of villages were formed. Furthermore, in those cases where the castles were constructed close to a strategic location, such as Chinar-Galu and Qustinlar castles (numbers 16 and 20), a village was created in their vicinity perhaps in order to provide food and a workforce for them.
It is also worth mentioning that in the northern part of the region, where there are more villages in comparison to the southern parts, the number of the castles as well as their size and defensive qualities increases. Therefore, although the primary function of the castles was defense against possible attack, in many cases they were established by the Ismaʿilis to expand and stabilize their authority. In fact, during the Alamut period, the capture and establishment of a castle was considered to be a way of imposing the authority of the Ismaʿilis on their surrounding areas.
Fig 6: Schematic map of the distribution of villages and fortifications in the Alamut region Google Earth, Google, DigitalGlobe2014
An account in Jāmiʿ al-Tawārīkh about the peace agreement between the Ismaʿilis and the Seljuq ruler, Sultan Sanjar, confirms this claim. Rashid al-Din (d.1318) writes:
… the Sultan accepted the conciliation and said “I will conciliate them with three conditions: first, they should not build any new castle, second, they should not buy any more weapons, and third, they should not invite the people to their beliefs.”
It is noteworthy that this agreement was accepted after numerous attacks by Seljuq forces on the Ismaʿili regions, particularly Alamut, and their failure to seize them. In fact, one of the aims of the contract was to prevent Ismaʿilis from expanding their influence. Considering the fact that Sultan Sanjar asks the Ismaʿilis to stop building the castles, implies the offensive function of the castles as well as their active role in the extension of the Ismaʿili territories.
The historical texts show that this approach to the construction of the castles was a feature throughout the Alamut period. According to Rashid al-Din, during the time of Muhammad Ibn-i Buzurg Umid (third lord of Alamut, d.1162), when Ismaʿilis enjoyed the superiority of military power and expanded their borders, a large number of castles was built. For example, on the construction of the Qustinlar castle Rashid al-Din writes:
… and Rafīqān (Ismaʿilis) went to Lar and constructed a strong castle which is located close to Sarbashm of Qazwin. Qutluq Ubih who was one of the noblemen of Iraq requested assistance to prevent Rafīqān from building the castle. A wellarmed troop, consisting of soldiers from Khargam, Tarum, Abhar, Zangan, Kharaqan, Abih, Savih, Damavand, Damqan, Gurgan, and near Nayshabur was brought there in order to fight against Rafīqān. They (i.e. the troops of Qutluq Ubih) tried but could not defeat Rafīqān and just seized their sheep and returned. The castle was finished in that winter. Kiya Ali Ibn-i al-Kiya al-Kabir became the commander of the castle in Shaʿbān of that year. In 539 A.H. (1145) a group of Rafīqān attacked Qazwin but they returned without any act of aggression …
As we understand from the text, the castle was established in the immediate vicinity of the Seljuq territory. The Ismaʿilis were also strong enough to withstand the “well-armed troop” and finished the construction of the castle. Afterwards they threatened Seljuq’s lands by attacking Qazwin. A similar event was also reported about building (or capturing and modifying) Saʿadat-Kuh castle in the northern side of Alamut region:
.. and in Rajab of 536 A.H. (1142) Rafīqān of Alamut went to the plain of Diylaman, burnt Sijan, built (ʿimārat kardand) the castle and filled its storerooms and appointed Kiya Muhammad Ibn-i Ali Khusraw Firuzkuhi as the commander of SaʿadatKuh and Rafīqān attacked Gurjiyan occasionally and besieged there…
In this case also, the Ismaʿilis had possessed the military superiority when they succeeded in building the Saʿadat-Kuh castle. Interestingly, in both cases (i.e. Qustinlar and Saʿadat-Kuh castles), immediately after giving the report of the construction of the castles, Rashid al-Din mentions that Ismaʿilis attacked the territory of their enemies. Therefore, one can claim that building the castles did not necessarily mean that they strengthened the defensive power of the Ismaʿilis in order to protect their territory. The castles can beinterpreted as offensive structures rather than defensive ones which threatened the neighboring lands. As this study indicates, these buildings, as centers of power, assisted the Ismaʿilis in expanding their territory and establishing an insecure area for the surrounding regions instead of creating a safe and secure zone for themselves.
Conclusion
This paper challenged the existing scholarship on the network of Ismaʿili castles. A large body of current studies on this subject emphasizes a functional connection between the castles as a ‘line’ of defensive structures for the protection of the Ismaʿili borders. Analyzing the distribution map of the castles and the main routes leading to the Alamut region reveals that there is not a strong relationship between the location of most of the castles and the communication routes. Therefore, one can suggest that there is little connection between the castles as an integrated system for defending the borders of the Alamut region. On the other hand, examination of the relationship between the distribution of the castles and the villages demonstrates the important role of the castles in the control of their surrounding settlements as centers of Ismaʿili power. Based on these findings, it no longer seems reasonable to speak of the castles whose distribution created a linear border around the Ismaʿili regions. Instead, they should be considered as centers of power, which formed spheres of influence on their surrounding environment; they were in fact active locations employed in the expansion of the Ismaʿili territories. Consequently, although there undoubtedly have been different types of relationships between the castles as members of a ‘network’ (such as visual connections or mutual support through the provision of food and military forces), we cannot claim that there was a functional connection between the castles as a line of defensive elements for the protection of the Ismaʿili territories.
In the absence of a strong body of literature on the Ismaʿili castles in Iran to date, this study offers a new interpretation of the network of castles in the Alamut region. It is undeniable that other areas of Ismaʿili territory should also be studied in order to offer a better understanding of the Ismaʿili castles. There are large numbers of castles throughout other Ismaʿili regions in Iran and Syria. The study of their distribution and of the relationship between them will shed new light on the complex factors shaping the network of Ismaʿili castles during the Alamut period.