SASSANIAN MILITARY SYSTEM

 

I herein divide the history of the Sassanian military system into three periods—the Early Sassanid Period (ca. 210-498 CE), including the military system of the Sassanid Kingdom of Pars (210-225 CE) and the early period of Sassanid imperial hegemony (226-498 CE), during both of which the Sassanids were heavily influenced by the Parthian military system that preceded it; the Middle Sassanid Period (499-591 CE), which saw a break from the military traditions inherited from the Parthians, in part due to the reforms of a series of Shahanshahs aimed at eschewing the quasi-feudal system of army recruitment in favor of a standing professional army, and in part due to the increasing influence of the Xionites and Hephthalites (see my Huns, Bulgars, Avars gallery); and the Late Sassanid Period (590-651 CE), during which there was a full return to the quasi-feudal system of recruitment, albeit with an expanded role for the lower nobility, although the shift in cavalry equipment and tactics that had taken hold during the Middle Period significantly changed the appearance and performance of Sassanian armies thereafter, and this was further reinforced under the influence of the Göktürks.

 

Unfortunately, the relative dearth of primary source materials; the unsystematic way in which the Persians themselves seem to have used titles and ranks; the limited understanding of non-Persian primary sources regarding the Persian language and Persian culture; the way Persian social, political, religious, and military ranks (rang) and their attendant responsibilities seem to have had significant crossovers; and the apparent evolution of terms and practices; all makes it very difficult to speak in absolute terms. However, I used five excellent primary source materials, in English translation, from the pre-Islamic and Islamic Persian periods that directly relate to the military matters of the Sassanid state—the Shah-nama/Shahnameh (“Book of Kings”), an epic poem by Abul-Qasem Ferdowsi Tusi (940-1020 CE) that gives a semi-mythical account of the history of the Persian Empire that is believed to have preserved earlier Parthian and Persian myths and legends; the fifth volume of the Tarikh al-Rusul wa al-Muluk (“History of the Prophets and Kings”), a history of the world from a Persian perspective, written by Muhammad ibn Jarir al-Tabari (838-923 CE); the Madigan-i Hazar Dadistan (“A Thousand Legal Decisions in Dadistan”), completed in 620 CE by Farrukhmard of Gur, which cites data from a census of available manpower resources in the Persian heartlands (district-by-district); the Ain-nama (“Book of Protocols”), a Sassanid tactical manual preserved in the Uyun al-Akhbar of Ibn Qutaybah (828-889 CE); and the Avesta (“Praise-song”), the Zoroastrian holy book, which includes both religious/philosophical concepts and some historical information about castes and their roles in society (including the warrior caste). I also used several Greco-Roman primary sources—the Rerum Gestarum Libri (“Book of Achievements”) of Ammianus Marcellinus (330-400 CE), which covers the conflicts between Rome and Persia during the fourth century CE; the Historia Nova (“New History”) of Zosimus (490-510 CE), which deals with Roman-Persian relations during the period 238-410 CE; the De Bello Persico (“On the Persian War”) of Procopius of Caesarea (500-570 CE), which details the conflicts between Rome and Persia during the early sixth century CE; and the Strategikon/Strategicon, a tactical manual written by the Byzantine emperor, Maurikios/Maricius/Maurice (582-602 CE). Modern works that I used to inform my research and modeling choices include Rome and Persia in Late Antiquity: Neighbors and Rivals by Beate Dignas and Engelbert Winter, Sasanian Persia: The Rise and Fall of an Empire by Touraj Daryaee, Decline and Fall of the Sasanian Empire: the Sasanian-Parthian Confederacy and the Arab Conquest of Iran by Parvaneh Pourshariati, A Glossary of the Construction, Decoration, and Use of Arms and Armor in All Countries and in All Times by George Cameron Stone, Sassanian Armies: The Iranian Empire by David Nicolle, Sassanian Elite Cavalry by Kaveh Farrokh, Rome's Enemies 3: Parthians and Persians by Peter Wilcox, and The Armies of Islam: 7th-11th Centuries by David Nicolle. I also found several articles from the Encyclopaedia Iranica website helpful—Sasanian Dynasty by Shapur Shabazi, Army of Pre-Islamic Iran by Shapur Shahbazi, Class System in the Parthian & Sasanian Periods by Mansour Shaki, and Elephants in the Sasanian Army by Michael B. Charles.

 

Strategic Command

At the apex of the Sassanian state, the commander-in-chief of the Persian Empire was the Shahanshah (Middle Persian, “King-of-kings”) or Padishah (“Master-king” or “Master-of-kings”), invariably a member of the aristocratic House of Sassan. The Persian monarchs were members of the warrior caste (arteshtaran), and they were expected to be fighters and battlefield commanders as well as administrators and the supreme representatives of the Mazdayan religion. The office was conceived as an autocratic monarchy, but the power and independence of the empire’s aristocracy (warrior and religious castes) tended in practice to limit the power of the monarchy in the social, political, military, and religious arenas. In the capacity of commander-in-chief of the military, the Shahanshahs also held the honorific titles of Arteshtaran-salar (“Leader of the Warrior Caste”) or Arteshtaran-sardar (“Head of the Warrior Caste”), although these titles were sometimes conferred as honorifics on members of the imperial court (in the Middle and Late periods, these titles were sometimes used in place of Eran-spahbad, see below). Ostensibly, the right to command any of the military forces of the empire flowed from the person of the Shahanshah down through designated subordinates, although the quasi-feudal structure of the landholding aristocracy (discussed below) tended to favor the development of semi-autonomous grandees with vassals and armed retainers of their own that they could call up with or without express imperial authority. Chief of the designated military subordinates of the Shahanshah was the Eran-spahbad (“Master of the Aryan Army”), a kind of vice-royal over-general that acted as the Shahanshah’s second-in-command of the imperial field army (Eran-spah), although the Eran-spahbad would take over as commander-in-chief if the Shahanshah was indisposed (e.g., seriously ill or too aged), or when there was a need for two separate imperial field armies on two separate fronts (the Shahanshah would command one, the Eran-spahbad the other), or when there was an imperial heir that needed command experience (and the heir would be named Eran-spahbad). In the first two instances, the Eran-spahbad was usually appointed from among the provincial governors (shahsaban or marzbanan), although prominent members of the imperial court (framadaran, singular, framadar) were also eligible. The imperial field army included imperial guard units, quasi-feudal provincial levies, allied contingents, and mercenary units. The imperial guard units were the standing central army of the empire kept in a perpetual state of war-readiness, mostly stationed in the capital city of Tisfon/Tesfon (Middle Persian, Tyspwn or Tysfwn), the imperial palaces and estates (ostan) spread across the empire, and key fortresses (arg) established at strategic locations within the empire and along its frontiers. The number of imperial guard units varied, although there were three primary groups throughout most of the history of the dynasty (these groups are attested between 224 and 642 CE)—the Zhayedan (“Immortals”), commanded by the Warahnigan-Khwaday (“Lord of Victory”), the Pushtighban (“Noble-bodyguards”) commanded by the Pushtighban-salar/Pushtighban-sardar (“Leader/Head of the Noble Bodyguards”), and the Darigan (“Palace-guards”) commanded by the Hazarbad (“Commander of 1,000”). The Zhayedan was a large division (gond/gund), numbering 10,000 men, while the Pushtighban and Darigan both seem to have been drafsht units (1,000 men). The Zhayedan was essentially the private army of the Shahanshah that would form the core around which an imperial field army could be built, although recruits for the Zhayedan were drawn primarily from the top two tiers of the warrior caste (wuzurgan/vaspuhran and azadan/azatan), which meant that the ranks of the Zhayedan were often riven with the aristocratic rivalries of the great houses. Although the azadan/azatan long dominated this guard division, the reforms of Kavad I and Khosrow I (ca. 488-579 CE) began to favor the recruitment of dehgan/dahigan, a move that seems to have been aimed at suppressing the power of the aristocracy and improving the loyalty of the corps to the Shahanshah. The Pushtighban was recruited exclusively from the younger sons of the great houses of Persia—the wuzurgan (hence the name, meaning bodyguards who were themselves nobles)—and they served as the command unit of the Shahanshah in the field (i.e., forming a screen around him, serving as dispatch riders, and serving as a crack reserve of shock cavalry). There was an inner circle of Pushtighban known as the Gyan-avspar (“Those that sacrifice themselves”), a vasht (100 men) of personal bodyguards to the Shahanshah who were sworn to lay down their lives before allowing harm to come to him. The Darigan were drawn from across the spectrum of the noble classes (wuzurgan/vaspuhran, azadan/azatan, and dehgan/dahigan). They were primarily palace guardsmen in charge of the security of whichever imperial palace the Shahanshah was in residence, although they sometimes took the field—some Shahanshahs were military men who preferred to fight alongside their Pushtighban shock cavalrymen, but sometimes the Shahanshah commanded the army from a litter or pavilion located on any convenient high point overlooking a given battlefield, and at such times the Darigan formed an outer perimeter and the dismounted Gyan-avspar formed an inner perimeter. Unlike the Zhayedan and Pushtighban, which were both shock cavalry units, the Darigan was composed of infantry archers (kamanderan). To these primary imperial guard units were sometimes added supplementary imperial guard units (bandagan, ayyaran, or janbazan), some of the better-known examples being the Khosrowgetae (“Khosrow’s Own”) and Peroozetae (“Victorious Ones”)—both probably created by Khosrow I (531-579 CE) as part of his military reforms, and both remaining in service at least until the end of the reign of Khosrow II (590-628 CE)—and the Gond-i Shahanshah (“Division of the King-of-kings”)—a unit of Daylamites (an Iranian people that inhabited the Alborz mountain region in northern Iran) that numbered 4,000 men, originally recruited by Khosrow II (probably around 600 CE), although the division only served until 636 CE (in the aftermath of the Battle of Kadisiya, the members of the division converted to Islam en masse, and thereafter served the Rashidun caliphs). The Khosrowgetae and Peroozetae were likely composite cavalrymen of a type that was replacing cataphracts (see below) in the Sassanian military system by the Middle Period (ca. 499-591 CE), while the Gond-i Shahanshah was a mixed unit of medium cavalrymen and infantrymen.

 

If an imperial field army was mustered, the available guard units were almost always supplemented with provincial levies, allied contingents, and mercenaries. The great houses of the empire (Persian wuzurgan and Parthian vaspuhran) sat atop a quasi-feudal or patron-client socio-political order in which they owed fealty/allegiance to the Shahanshah, although technically they held their estates in their own right, not as fiefs granted by the Shahanshah (the authority of the Shahanshah was derived from his khvarenah/khwarenah, an essence that made him the physical representation of the Mazdayan Wise Lord on Earth, and from his position as the titular head of the warrior caste). Thus, the provincial levies reflected the virtual independence of the great houses—the socio-political order of the Persian Empire has been compared both to the feudal system in Medieval Europe and the patron-client system in the Roman Empire, although it lay somewhere in-between. The underlying autocratic concepts that saw the Shahanshah as ruling by divine right also saw the grandees of the warrior and religious castes as having similarly divinely mandated authority, even if they were recognized as subordinates. There was a strong tendency for both political and social position to become inheritable, and dynastic lineages tended to dominate political-military posts. There were two types of province/satrapy—inner province (shahr) and border protectorate (marz). Both types of province were subdivided into imperial estates (ostan) and districts (tasug/tasuk), and districts were further subdivided into subdistricts (rustag/rustak). To use the analogy of the European feudal system, the provincial governorships (singular, shahsab/shahrab and marzban/marzpan) were basically treated as an appanage (an inheritable title and office granted by the monarch that gave the holder various perquisites and responsibilities, including the right to govern a given territorial area on behalf of the Shahanshah), while the actual landholdings of the great houses were treated like feudal demesnes (they had management rights over their tenants, although unlike European feudalism, the estates of the grandees of the Persian Empire were not granted from the sovereign, they were owned outright by families). Thus, the Shahanshah only had the right to appoint or remove governors (shahsaban or marzbanan) or court officials (framadaran) from official offices (or have them executed on various charges), but he did not have the right to seize their estates. If a grandee was removed from office and not executed, he was usually “exiled” to his own estates (essentially exiled from the court, but not the empire, and although no longer holding official office, still an influential landholder), and if he was executed, his heir automatically retained control of his estates, even if someone from a different house was given the governorship or court office thus vacated (although sometimes the heir was allowed to take the vacated position, and toward the end of the empire many of these offices had become semi-hereditary and the heir expected to take over the office as well). Indeed, the estates of the grandees were not specifically linked to the offices they held—their primary, ancestral landholdings were usually located in a province in which they tended to dominate the governorship, but most had estates throughout the empire, and branches of great houses sometimes came to control multiple provinces. Nevertheless, despite the disconnect between office-holding and landownership, the lines of demarcation between the authority exercised by the grandees as holders of office and as large-scale landowners became blurred, since governors of provinces in which they were the largest landholder tended to conflate their roles (in essence, treating the entire province like their own estate). Further, the districts (tasug) were originally dominated by families of the azadan rank, although following the reforms of Kavad I and Khosrow I, the dehgan, who had formerly served only as subdistrict (rustag) or village (deh) headmen, began to displace the azadan as district leaders, and this led to a confusion in terms (i.e., dehgan literally only means “village headman,” but oftentimes they were referred to as dehgan even when they held higher offices like the district and subdistrict leadership positions). As with the provincial governorships during the Early and Middle periods, by the Late Period (ca. 592-651 CE) the district governorships tended to be treated like a combination between a feudal appanage and a demesne, except that in this case the holders of office were generally not wealthy enough to have estates throughout the empire, and therefore their landholdings were usually entirely within the district that was their appanage. During the Early Period (ca. 210-498 CE), the azadan and dehgan generally did not hold their lands as tenants/clients of the estates of the grandees (nor did they have many tenants/clients themselves), and therefore they were not technically dependents of the grandees, but if they held office as a district or subdistrict leader/headman, they were the political and military subordinates of whichever great house controlled the governorship of the province in which they were landholders. And this put them in a tricky position vis-à-vis the grandees. This is where the Roman patron-client model becomes useful. There were more azadan and dehgan in each province than there were districts and subdistricts (each subdistrict was composed of a collection of rural villages called deh, from which the dehgan class took its name). The grandees generally dominated the imperial court and access to the Shahanshah (who ultimately determined who governed what), so despite being economically independent of the grandees by virtue of having their own landholdings (i.e., they were not held as fiefs of the larger estates of the grandees), they were often politically dependent on the grandees if they ever wanted to be appointed to a position in the bureaucracy or military. Competition between the azadan and dehgan for the favor of the grandees, often represented by informal but longstanding agreements, therefore created a degree of dependence that was not reflected in any formal patron-client system or direct vassalage, but that was basically what it was. And like the feudal magnates of Medieval Europe, the greater and lesser nobles of ancient Persia—most of whom were members of the warrior caste (arteshtaran)—were therefore simultaneously officers of the civil bureaucracy and the natural leaders and primary fighters of the provincial levies, and so their own network of social hierarchies determined what types of units made up a provincial levy and who commanded them, often making these levies something akin to the private armies of the grandees. Indeed, when the imperial field army (Eran-spah) was called together, the forces contributed by each province were considered an army (spah) and its commander—the governor (shahsab or marzban)—held the additional title of spahbad (“army master”). Persian spahbads were known to operate their own provincial armies (spah) independently of imperial guard units and the central command (e.g., when responding to a foreign incursion across the border in their region of the empire), or sometimes in coordination with one-another, but if they were mustered as part of an imperial field army, they were either subordinated to the commander-in-chief (i.e., the Shahanshah or Eran-spahbad) or the most prominent of the spahbads was given a special command as a kind of field-marshal (nakhvaegan). Confusingly, various commanders at different times preferred to use their rank (rang) in the socio-political hierarchy of the warrior caste (e.g., shahsab/shahrab or marzban/marzpan) in place of any specific field command title to which they had been assigned (e.g., Eran-spahbad, spahbad, or nakhvaegan), or they used their house name as a title (e.g., Kanarang, Surena, or Karena), or they even used honorific epithets in place of titles (e.g., Shahrbaraz, “Wild Boar of the Empire”). To add to the confusion, Greek and Roman writers often confused titles with names in their accounts of encounters with the Persians, and some of the names of Persian noble houses seem to have had their origins in hereditary titles of rank (e.g., Aspahbadh).

 

As one can well imagine, these quasi-feudal provincial levies/armies, commanded by powerful noble grandees to whom most of the fighters owed some sort of allegiance via legal dependence or informal social contract, tended to make said grandees overmighty in the eyes of the Shahanshahs. The political settlement made by Ardashir I (211-242 CE) when he became the first Sassanid Shahanshah (ca. 226 CE) may have attempted to address this, which was essentially a problem that had dogged the throne from the time of the Achaemenids and had become acute during the Seleukid and Arsakid dynasties. Some modern historians have suggested that Ardashir may have intended the governorships of the inner provinces to be divided between a civil administrator (shahsab/shahrab) and a military commander (paygosban, often translated as “district-guardian”), while the margraves of the border protectorates (marzban) combined both civil and military functions—the idea being that those governors closest to the throne would not have all the reins of power in their hands, while the heightened need for security on the frontiers meant that those governors needed more independence of action (although the border protectorates were usually much smaller than the inner provinces, with more limited resources). However, lack of evidence for the holders of the title of paygosban has made it impossible to determine for certain whether this was actually the situation. The translation, or mistranslation, of the word paygosban is also problematic. In Middle Persian, -ban means “nobleman” and paygos- seems to be derived from the same root as paygan, which meant “footman,” so it seems to me that paygosban should not be translated as “district-guardian” but as “noble-footman.” Districts were known as tasug, not paygos. Since we know that many shahsab fought as cavalrymen and commanded their own provincial troops, it seems to me likely that paygosban might have been a title used by a noble who was selected to command a province’s infantry (i.e., the commander of a gund/gond of infantrymen), while the governor was assumed to be the commander of the province’s cavalry. Further, the royal estates of the House of Sassan formed a kind of parallel administration to that of the political-military hierarchy in some provinces (the royal estates tended to be clustered more in southern Iran and Mesopotamia)—these estates were called ostan and the majordomo/administrator of them was called an ostandar. Ostan were analogous to tasug (i.e., districts), although rather than being subdivisions of a province whose administrator answered to the governor, the ostandaran answered directly to a member of the imperial court known as the Darigbad/Darigbed (“Palace-commander” or “Superintendent of Palaces”). Therefore, ostandaran may have served as representatives of imperial interests in the provinces, which sometimes included command of imperial guard units that had been posted there. During the Middle (499-591 CE) and Late (592-651 CE) periods, there is increasing confusion in the sources between the terms tasug, rustag, ostandar, and dehgan that seems to derive from the gradual increase in importance of the dehgan—by the Middle Period, they had largely displaced the azadan as headmen of tasug (district) and ostan (imperial estate district), and the class rank (rang) of the dehgan seems to have supplanted the other titles. Regardless, it is obvious from the historical record that most governors commanded their own provincial armies as personal, quasi-feudal forces, and most of the Shahanshahs accepted this as the status quo. In at least two cases we know that there were long-term special commands that gave a governor a kind of status as an overlord of the other provincial governors in an entire region of the empire (similar to a nakhvaegan, although nakhvaegan were appointed for short-term commands that generally did not extend beyond a single specific campaign). The first such special command was that of the governors of the Abarshahr (an inner province), traditionally from the House of Kanarangiyan, who combined both civil and military functions under the eponymous title of Kanarang (ca. 484-652 CE). The Kanarangs were recognized as the overlords of the neighboring border protectorates in northeastern Iran (Goyman, Khorasan, Marv, and Harev)(ca. 484-652 CE). The second was the Kushanshah (“King of the Kushans”), a title created after the Sassanian conquest of the Kushan Empire. The Kushanshah governed the Kushanshahr (“Kingdom of the Kushans”) as an inner province (southern Afghanistan, formerly known as Arachosia) and was the overlord of the other provinces carved out of the Kushan Empire—Fararud (formerly Sogdiana, spanning eastern Uzbekistan and western Tajikistan), Tokharistan (formerly Bactria, northern Afghanistan), Gandhara (northwestern Pakistan), and Turgestan/Turestan (formerly Hindush, southwestern Pakistan)(ca. 225-365 CE). Ironically, the Kanarangs and Kushanshahs may actually have influenced or inspired the reforming Shahanshahs of the Middle Period when they attempted to weaken the power of the grandees by reorganizing the Persian political-military hierarchy.