Semper Victor, this is all really impressive and useful stuff. Can you recommend me some of the written sources you're drawing from? I'm working on a little project of my own and these kinds of materials would be really helpful to me.
Semper Victor, this is all really impressive and useful stuff. Can you recommend me some of the written sources you're drawing from? I'm working on a little project of my own and these kinds of materials would be really helpful to me.
Well, now you do, don't youI just realised that I know absolutely nothing about the inner politics of either the Parthian or Sasanian Empire.
5.3. THE ARSACID MILITARY. STRATEGIC COMPARATIVE WITH ROME.
At its peak, the Roman empire had a population between 50 and 75 million people, and fielded a well trained professional army which increased in numbers steadily from Augustus’ reign onwards. Apart from available manpower and trained soldiers, Rome also had at its disposal fiscal and economic resources much higher than those available to the Arsacid kings.
According to Yann Le Bohec and Adrian Goldsworthy, the total numbers of the Roman army (including the forces stationed in Rome itself) grew steadily in numbers from Augustus’ death to the age of the Severans:
In parallel to this overall increase in armed strength, the Romans kept increasing the garrisoning forces of their eastern provinces during the Principate. Under Augustus, the eastern provinces were garrisoned by the following legions (6 in total), with their corresponding auxiliary units (that according to Tacitus amounted usually to the same numbers as those of the legionaries they accompanied):
- Under Tiberius: 300,000 men.
- Under Trajan: 360,000 to 380,000 men.
- Under Septimius Severus: more than 500,000 men.
In parallel to the increase in numbers, the tendency was to move the legions closer to the borders as time went by; under the Flavians they were definitively posted at or very near to the borders. Under Titus (79-80 CE), there were in the East 8 legions:
- Legio VI Ferrata, based at Ancyra (Cappadocia).
- Legio IV Scythica, based at Cyrrhus (Syria).
- Legio III Gallica, based at Antiochia (Syria).
- Legio XII Fulminata, based at Raphanaea (Syria).
- Legio XXII Deiotariana, based at Nicopolis (Egypt).
- Legio III Cyrenaica, based at Coptos (Egypt).
By the time of Marcus Aurelius’ death (180 CE), this number grew to 9 legions, and Egypt had lost one of its legions in benefit of the Levant:
- Legio XVI Flavia Firma, based at Satala (Cappadocia).
- Legio XII Fulminata, based at Melitene (Cappadocia).
- Legio VI Ferrata, based at Samosata (Syria).
- Legio IV Scythica, based at Zeugma (Syria).
- Legio III Gallica, based at Raphanaea (Syria).
- Legio X Fretensis, based at Hyerosolima (Iudaea)
- Legio XXII Deiotariana, based at Nicopolis (Egypt).
- Legio III Cyrenaica, based at Coptos (Egypt).
And by the time of the death of Severus Alexander (235 CE), the grand total had grown to 11 legions:
- Legio XV Apollinaris, based at Satala (Cappadocia).
- Legio XII Fulminata, based at Melitene (Cappadocia).
- Legio IV Scythica, based at Zeugma (Syria).
- Legio XVI Flavia Firma, based at Samosata (Syria).
- Legio III Gallica, based at Raphanaea (Syria).
- Legio X Fretensis, based at Aelia Capitolina (Syria Palaestina).
- Legio VI Ferrata, based at Caparcotna (Syria Palaestina).
- Legio III Cyrenaica, based at Bostra (Arabia Petrea).
- Legio II Traiana Fortis, based at Nicopolis (Egypt).
- Legio XV Apollinaris, based at Satala (Cappadocia).
- Legio XII Fulminata, based at Melitene (Cappadocia).
- Legio III Parthica, based at Nisibis (Mesopotamia).
- Legio I Parthica, based at Singara (Mesopotamia).
- Legio IV Scythica, based at Zeugma (Syria Coele).
- Legio XVI Flavia Firma, based at Sura (Syria Coele).
- Legio III Gallica, based at Darsaba (Syria Phoenicia).
- Legio X Fretensis, based at Aelia Capitolina (Syria Palaestina).
- Legio VI Ferrata, based at Caparcotna (Syria Palaestina).
- Legio III Cyrenaica, based at Bostra (Arabia Petrea).
- Legio II Traiana Fortis, based at Nicopolis (Egypt).
Map showing the full deployment of all Roman legions in 125 CE (under Hadrian).
Considering the burden that the maintenance of a permanent army imposed on pre-industrial societies, it’s clear that this increase reflected a growing concern towards the eastern Roman border.
The Arsacid empire, with its much smaller human, economic and human resources, was able to compete with Rome (punching well above its weight class) because in the Iranian plateau it had a huge reserve of mobilizable experienced fighters which were not on the king’s payroll. Compared to Roman society, the society of the Iranian plateau was hyper militarized. Coupled with geography, this was the reason that turned the Iranian-based empires of the Arsacids and Sasanians into such dangerous enemies for the Romans.
But this key strength was also the Arsacids’ main weakness. Usually, a hyper militarized society is also a highly volatile and unstable one, and ancient Iran was no exception to the rule. The warrior clans that ruled the Iranian plateau behaved according to an ethos that judged that only war, hunting and drinking were suitable pastimes for them; coupled with the agnatic hereditary system, the result was a very unruly society. While the Arsacid empire was expanding in the III, II and early I centuries BCE, there were no troubles because the grandees and the Arsacid kings were united in the pursue of expansion, which meant more riches and lands to everybody. But when expansion stopped around the same time that the Arsacids first clashed with Rome, this warlike energy was turned inwards and led to continuous civil wars from Ūrūd II’s reign until the fall of the House of Aršak.
By comparison, in this same period of time the Roman empire enjoyed a golden age of (relative) political stability (especially under the Antonines) and great demographic and economic growth. This blatant difference between the internal situation in both empires incited Roman aggression, and in the II century CE led to three great scale Roman invasions that ended all of them in the capture and looting of the Arsacid main capital at Ctesiphon.
The “feudalizing” organization of Arsacid armies also imposed quite sharp restrictions on what the Arsacid kings could achieve with them.
First of all, as the Arsacid king would be waging war not with “his” own army, but with the armies of the great clans, he needed to convince them that it would be worth it. With hindsight, it’s clear that if they supported the Arsacids and Sasanians for so long, they had an interest in preserving the statu quo (with tradition and custom also playing a part in it), but Classical authors were very aware of this key weakness of the “Parthians”. The Iranian clans showed themselves usually keen to defend their homeland (although probably what they considered as “homeland” stopped at the Zagros), but they disliked long campaigns far from their lands, especially if they involved lots of sieges, hard fighting and little loot. It’s unclear if there were “feudal obligations” like in the European Middle Age, but ancient accounts seem to suggest that the nobility did not feel itself compelled to stay in the field supporting their king for long periods of time. Also, there were no centralized logistics: every clan’s army brought its own supplies, and when they were finished (or they thought their duty was done, whatever happened first) they just quit and went home. A part of those armies would also be formed by herders, peasants and other workers who could not afford to stay away from their livelihoods for extended periods of time.
This meant that, if pressed hardly by the Arsacid army and wanting to avoid a field battle, it was a sure bet for the Romans to just fortify themselves (in a field encampment or in a walled city) and wait the “Parthians” out. After his near defeat at Nisibis in 217 CE, emperor Macrianus just opened peace talks with Ardawān V and kept stretching them until Ardawān’s army melted away as time went by.
The nature of Iranian society meant that, if we count the years between the battles of Carrhae (54 BCE) and Nisibis (217 CE), which were the first and last battles between both powers, the Arsacids actually spent many more years fighting among themselves or quelling nobiliary uprisings than fighting the Romans or any other foreign enemy (even if we add all of them together). And in the case of civil wars, neither members of the House of Aršak nor members of the nobility showed any kind or reluctance in calling for Roman help.
Rome could support a war for a long time, provided it happened in only a single theater (the great weakness of the Roman military was its incapability of waging war at the same time in two or more fronts). But the Arsacids, due to all the handicaps already mentioned, could only handle short wars.
Arsacid cavalry armies could be used both in defensive or offensive roles, and could also be employed to fight conventional, pitched battles, or to wage “guerrilla” campaigns to wear the enemy down. Their great weakness was their uselessness against fortifications (which is quite puzzling, because during their expansion the Arsacids had to take many walled cities, some of them truly formidable ones).
Californian archer Mike Loades performing one of the shot modalities practised by ancient Iranians; as seen in some ancient reliefs, he holds two more arrows with the hand that holds the bow for rapid fire. This technique allowed for rapid "shower fire" at the cost of lower accuracy.
And again Mike Loades performing the same shot backwards; this is the famous "Parthian shot".
The situation was exactly the opposite for the Romans. Thanks to its large professional army, Rome could sustain long wars with plenty of losses, sieges, and lack of opportunities for looting. But these armies were mainly infantry armies, very cumbersome when compared with the Arsacid cavalry ones. They were useless in indirect, “guerrilla warfare” and to win campaigns, they needed “hard” objectives to hit: either enemy cities or fortresses, or to fight pitched battles against the enemy main army. As the Arsacid armies could (and would) always elude the Roman infantry armies if they wanted to, the Romans never managed to really inflict a crushing defeat against the Arsacids. The only targets they could hit without them running away were the great Mesopotamian cities, which they took several times without managing to bring down the Arsacid empire down.
But these attacks took their own toll. First, the annexation of northern Mesopotamia (Osrhoene and western Adiabene) temporarily under Lucius Verus and for good under Septimius Severus was a hard blow against the prestige of the dynasty, as it brought the Roman border dangerously close to Media, inserted a strategic wedge between Armenia (ruled by an Arsacid) and Mesopotamia and allowed the Romans to reach Ctesiphon in a much quicker and easier way. Also Mesopotamia, which became the main marching field for Rome’s legions, was the main power base of the House of Aršak itself (together possibly with royal estates in Parthia proper), and so, if Roman attacks did not bring the Iranian empire down, they did weaken the position of the Arsacid royal house in front of its dangerous nobility, both by sapping directly the fiscal and human resources available to them, and most seriously of all, by damaging the prestige of the dynasty.
Despite all the deep differences between Roman and Iranian societies, both the Roman emperor and the Iranian šahanšah (Arsacid or Sasanian) had something in common: they were war leaders first and foremost, and their fundamental role was to lead their armies to victory. If they failed in their main raison d’être, both could and would be replaced; the only difference was that as in Iran the dynastic principle was sacred, his successor would be chosen amongst his surviving relatives. The most extreme, almost comical, case is that of the Sasanian šahanšah Šabuhr II, who was said to have been crowned while still inside his mother’s womb.
Both cultures expressed this reality through elusive abstract concepts: Romans would talk about the fortuna, genius, felicitas or virtus of their emperor, while Iranians expressed it most succinctly through a concept rooted in Zoroastrian tradition, that of fārr, which translates as “kingly glory”. Fārr was an intangible quality with which Ohrmazd himself invested the king. In Iranian art, especially in Sasanian rock reliefs (although it has precedents in Arsacid times) this was expressed in a scene in which Ohrmazd or one of the yazdān (usually Mihr or Anahid) handed over a “ring of power” to the šahanšah.
Investiture relief of the Sasanian šahanšah Ardaxšir II. The king stands in the center, and is receiving the ring of power from the supreme god Ohrmazd (on the right), while the god Mihr (known to the Romans as Mithra) stands to the left.
Well, now you do, don't you
Thanks for this yet again informative and well written chapter on the Parthians!
BTW is it just me or does Ohrmazd have the exact same dress and appearance as the king to which he is giving his ring?? How comes.
"One Ring to Rule Them All"? xD
Diadem-with-scarf is how it looks to meScholars call them so because nobody has the slightest idea of what they were called in reality, or what they were exactly. Actually, they could also be a torques, or a diadem.
Don't you mean Ionic?How ironic.
I agree with you that Maddison's demographic information may not be that great outside of Europe (he seems to always pick the lower-level possibilities, such as his frankly ridiculously low populations for the New World, prior to European contact). However, I have a hard time seeing how Iran could support a larger population in the 6th century than in the 19th.In his monumental database, Angus Maddison gives a total of 4 million inhabitants for Iran by year 1 CE. To me, this seems like too low a number, due to several reasons:
- The comparison with population levels in later times. Basically, the demographic structure of the Iranian plateau remained the same under the Arsacids as in the Safavid and Qajar eras; perhaps with a little more area under of cultivation due to the extension of the qanat systems under the Sasanians and early Muslim dynasties. For example, a British traveler gave in the 1820s an estimate of 6 million inhabitants for Iran, and this taking into account that the XVIII century was a disastrous century for the country, which by then was just starting the recovery towards the higher demographic levels of the Safavid era. By the late XIX and early XX century, Iran had a population of a little more than 9 million inhabitants.
- The comparison with the eastern provinces of the Roman empire. For example, the most frequently quoted estimates (or rather guesstimates) give between 4 and 5.5 million inhabitants for Egypt, around 4,5 million for Syria (including Palestine), between 8 and 10 million for Anatolia (excluding Armenia, then an Arsacid vassal) and between 2 and 3 million for continental Greece. All of these territories are smaller when added together (for Egypt, counting only the inhabited areas of the Nile valley, not the surrounding desert) than the Iranian plateau, and although said plateau includes large wastelands, so do Anatolia or Syria. I think that population density must have been higher (or even quite higher) for the Roman eastern provinces than for Iran, due to the very different social and economic systems in both polities, but not stratospherically so.
- Then there’s the fact that ancient sources (Plutarch, Tacitus, etc.) state without reserves once and again that the Arsacid empire was able to mobilize very large armies, enough to match the ones fielded by Rome. Juxtaposing the tales of Justin and Plutarch, the Arsacid king Farhād IV opposed Mark Antony’s invasion force with an army of 90,000 men in three corps, while according to Porphirios Farhād I was able to muster an army of 120,000 against the invading Seleucid army of Antiochos VII Sidetes. The Chronicle of Arbela (a not too reliable document though) states that Walaxš IV (Latin Vologaeses) mustered an army of 120,000 men against the rebellious Median and Persian noblemen. And finally there’s the telling evidence of the battle of Nisibis, which we’ll deal with in detail later. Both Cassius Dio and Heraclianus, contemporaries of the events, stated unequivocally that Ardawān V’s army was superior in numbers to the Roman army led by emperor Macrinus, and it’s possible to reach a reliable estimate for Macrinus’ army (between 75,000 and 100,000 men), so Ardawān V must’ve fielded an army substantially superior in numbers.
With all these thoughts in mind, especially the numerous ancient testimonies about the abundant manpower reserves of the Arsacid empire, I find impossible to reconcile them with Maddison’s figure. If I had to take a chance, I’d make a guess and give to the Iranian plateau a population between 6 and 9 million (double than that of Roman Syria at its highest), which seems quite reasonable.
Well there are some problems with this. Firstly we need to define what we mean by "Mesopotamia". Are you talking about the whole of modern Iraq? Or just the flat, river-fed plains of the Tigris and Euphrates? Do you include the Iraqi part of the Zagros mountains in this (i.e. Kurdistan)?Mesopotamia was by far the agriculturally richest and most urbanized and populated province of the Arsacid empire, although its population was heavily concentrated in its central and southern parts. Angus Maddison’s database gives an estimate of 1 million inhabitants at 1 CE, which to me again looks like a too low number. If Roman Egypt was capable of supporting 4 – 5.5 million inhabitants, and Roman Syria 4,5 million, I’d take a number similar to these for Arsacid Mesopotamia: between 4 and 5,5 million inhabitants.
Sounds reasonable.In 197 CE, after being defeated by Septimius Severus, the Arsacid king Walaxš V had to cede northern Mesopotamia to the Romans, basically the area between Armenia, the Tigris river and the Khabur river. This area was not the most populated part of Mesopotamia, but still it included a series of important cities, like Nisibis, Edessa or Carrhae. So, we could think that the Arsacid empire lost between 0,5 and 1 million inhabitants with this territorial loss. With the annexation of Osrhoene by Caracalla in 216 CE (by then it was merely an island surrounded by Roman territory, all the autonomous northern Mesopotamian kingdoms disappeared, except for Adiabene (located east of the Tigris) and Hatra.
Hmm, I can't accept such a high figure for this area. Assuming by "Elymais" you mean roughly modern Khuzestan province, then I don't see how 1.5 to 2 million people could live there in the Early Middle Ages. If you are talking about a larger area than modern Khuzestan (such as including parts of Southern Iraq or something), then what I'm about to say may not apply, but here goes anyway.Elymais was again another agriculturally rich area, and densely populated since remote times. An alluvial plain, it was well watered by several rivers that descended from the Zagros, allowing for irrigated agriculture in most of its territory. Still, it was geographically small. I’d guess a total population around 1,5 - 2 million people for this land.
What area exactly are you talking about here for "Arsacid Armenia"? How big is it? and can you show a map of the area you mean (I assume it's significantly bigger than modern-day Armenia?). Modern Armenia (then a Soviet republic) only had a population of 1.35 million in 1950 (according to UN data), so I can only assume it had a much smaller population in the Middle Ages (often, counties increased their population 5 or 10-fold from the Medieval era to the Modern era). For 2 to 2.5 million you must be talking about a large section of (modern-day) Turkey as well.Armenia was a populous kingdom, which although mountainous had many valleys and plains scattered between them, and like Greater Media, it had excellent pasture land. Its social system was practically the same as that of Iran proper, and ancient authors state time and again that the Armenian king could levy among his nobles armies up to 60,000 men, with many armored knights amongst them. It was also a land without water scarcity like many other Middle Eastern countries. I’d take a guess of 2 – 2,5 million inhabitants for Arsacid Armenia.
Hmm, Azerbaijan (the then Soviet republic) had a population of 2.9 million by 1950 (and that doesn't include any of Iranian Azerbaijan, which I don't have any historic data for). And it is a much less arid region than most of Iran or Iraq, so you may be underestimating its population (especially if you include all of Iranian Azerbaijan in that figure).As for Media Atropatene, it was like a smaller version of Armenia. It corresponded in broad strokes to the territory of historical Azerbaijan, currently divided between the Republic of Azerbaijan and the Islamic Republic of Iran. It had not the abundance of water and pastures that Armenia had either, and included several semi-deserted steppe areas. Still, the proximity of the Armenian and Alborz mountains, as well as the heavy rainfall of the southern Caspian coast meant that irrigated agriculture was possible in parts of it. It was populated by an Iranian-speaking population, with religious traditions and a social system identical to that of the Iranian plateau. Although ancient authors give it a mobilization capability much lower than that of Armenia or Greater Media. Taking a guess, I’d give it a population around 0,5 – 1 million people.
I would perhaps conclude a figure closer to the low end of that estimate (though it depends, as I'm not fully sure of the exact areas you're talking about). Maybe 7 million at a push.Adding these numbers up, we get a total population in the western Arsacid territories (excluding northern Mesopotamia) between 7,5 and 10 million inhabitants.
Great pictures, but Mike Loades is English, for your information.
Californian archer Mike Loades performing one of the shot modalities practised by ancient Iranians; as seen in some ancient reliefs, he holds two more arrows with the hand that holds the bow for rapid fire. This technique allowed for rapid "shower fire" at the cost of lower accuracy.
And again Mike Loades performing the same shot backwards; this is the famous "Parthian shot".
Well that's quite a feat given that I don't see much of pre-modern empires being as militarized as Rome, but given the feudal/tribal nature of Iran I can go with itCompared to Roman society, the society of the Iranian plateau was hyper militarized
Hey, just thinking about this topic made me look up some old maps and diagrams, so some of you may also be interested in them:
Here's a map of the Satrapies of the Achaemenid Persian Empire, though I don't know if they are in any way similar to the Parthian and Sassanian periods.
And here's a useful diagram of the Iranian/Iranic languages (past and present):
Great pictures, but Mike Loades is English, for your information.
However, I have a hard time seeing how Iran could support a larger population in the 6th century than in the 19th.
Even though Iran was behind the West technologically by the 18th and 19th centuries, there was still surely significant new improvements in agriculture happen by that time. Iran must have received New World crops by that time, for instance (and indeed crops and varieties from Europe and elsewhere, due to the global trade network). So surely farming would be better, more productive and more sophisticated 200 or 300 years ago than it was 1,500 years ago, or 2,000 years ago. That would tend to suggest to me that the ancient and medieval populations couldn't have been as big as the 19th century population of the country.
Also, we should be a bit clearer with out definitions here. When you say the "Iranian plateau", do you mean basically the area of modern-day Iran? Or something larger, or smaller?
Well there are some problems with this. Firstly we need to define what we mean by "Mesopotamia". Are you talking about the whole of modern Iraq? Or just the flat, river-fed plains of the Tigris and Euphrates? Do you include the Iraqi part of the Zagros mountains in this (i.e. Kurdistan)?
In any case, I'm not sure if it is directly comparable to Egypt. I think that Egypt is a really special case of a pre-modern society that had a very high population density, because of the almost unique fertility of the Nile valley. Just go look at some satellite images (Google Earth) of the farmland in Egypt and the farmland in Iraq or Iran, and you'll see a massive difference in how green and productive it is (Mesopotamia and Iran look quite arid by comparison).
Frankly I would expect both Mesopotamia and even the good parts of Iran to have less population density than Egypt.
Modern Iraq has 437,000 square kilometres (or km2) and I would roughly divide that area up into: ~240,000 km2 of desert (probably with only a small nomad population), about 100,000 km2 is "northern Iraq" or "Zagros Iraq", or just "Iraqi Kurdistan", which is mostly hills and mountains and smaller river valleys (and some flat desert). Which leaves about 97,000 km2 of farmland in and around the Tigris and Euphrates rivers. Though I'm not sure how extensive the irrigation systems were in the Middle Ages and whether the interior area between the main river-courses was irrigated and farmable, or whether it was arid (this will mater a lot as most of the current arable area is not close to the rivers).
Actually 97,000 km2 may be far too generous an area, I've done some more measurements and come to perhaps only 39,000 km2 for the irrigated, arable area of Mesopotamia. That's if you exclude some of the main region between the rivers (some of which is dry even today) and the southern area, which used to be a giant swamp until very recently. Frankly I don't think it's possible to have 4 to 5.5 million people on an area of 39,000 km2 (that's a population density of 103 to 141 p/km2, which is ridiculously high) with pre-modern technology.
From my studies it seems that generally Medieval farming methods would allow an agricultural society to have densities from anywhere between about 10 persons per square kilometre (or p/km2), to perhaps as high as 50 p/km2. There may be some exceptional cases, like Egypt where the numbers would go as high as 70 p/km2, but no more than that is really possible with pre-modern technology.
If the 39,000 km2 of Mesopotamia had a density of say 30 p/km2, then it would have 1.17 million people (30 p/km2 is a perfectly reasonable figure for a highly developed Mediaeval society, it's about what France had in the 1300s). If it had 50 p/km2, then we're talking 1.95 million and if it had 70 p/km2 (which is about what the Nile valley and delta had) then it could be as high as 2.73 million. I can't imagine it could go any higher than that.
There might be an additional 0.5 to 1 million people in northern Iraq (the Kurdish area), which would give Iraq a total population of maybe 2 to 3.5 million, at most.
Sounds reasonable.
Hmm, I can't accept such a high figure for this area. Assuming by "Elymais" you mean roughly modern Khuzestan province, then I don't see how 1.5 to 2 million people could live there in the Early Middle Ages. If you are talking about a larger area than modern Khuzestan (such as including parts of Southern Iraq or something), then what I'm about to say may not apply, but here goes anyway.
I measured the area of farmland within Khuzestan, which is all located along a few river valleys/plains. And I came to a figure of about 16,500 square kilometres (or km2), which doesn't seem like enough to support such a large population. Although the total area of the province is 64,000 km2, most of that is very arid mountains and hills, which seem unlikely to support a major population (I'm sure there are scattered villages and nomads all over the place, but not dense agricultural settlements or cities). If I'm generous and increase the amount of farmland to say 18,000 km2, to account for all the little bits of arable and semi-arable land up in the mountains, that still isn't much.
Now if we calculate a population density for this area we can get somewhere.
If those ~18,000 km2 of arable land have a density of 10 p/km2, then it has a population of 180,000 people. If 20 p/km2 then 360,000, if 30 p/km2 then 540,000. If they had very high densities of 50 p/km2, then they would have 900,000 people. I don't see how it could be possible to support 1.5 to 2 million people though (that would suggest densities of 83 p/km2 and 111 p/km2, respectively, which is just too high). Now of course there are still going to be some people living in the remaining ~46,000 km2 area of the province, but since the climate is so harsh and the terrain so difficult I can only imagine that small numbers of nomads could live there, at a low population density.
Again, from my research, I determined that generally the population density of nomadic societies cannot go above 1 p/km2 and in many cases (like desert or arctic nomads) their density may be much less than that (I came to some general conclusions of about 0.5 p/km2 for steppe nomads, 0.2 to 0.1 p/km2 for desert nomads and 0.05 to 0.01 for arctic nomads). If the remaining 46,000 km2 of Khuzestan had say 1 p/km2, then it would only account for less than 50,000 people, which is a rounding error when considering the size of the settled agricultural population (almost an irrelevance when talking about the demography).
(you can read this post of mine, from a few months ago [and the following posts] to get an idea of the framework I'm using to calculate populations and some of my basic assumptions about pre-modern population density)
Here's an image from Google Earth where I've measured the areas of farmland in red:
(note that the large dark green area between the two main red areas is a swamp of some sort. You can see the Iraqi border in yellow and the Khuzestan province borders in light purple. I assumed that the last few kilometres of the river deltas are alluvial deposits which would not have existed yet by 2 millennia ago, as they get bigger gradually and the sea would have been further in-land back then. The red areas make up only about 25% of the province)
My guess is that Khuzestan in the Middle Ages would have only around 0.5 to 1 million people, but it would struggle to get higher than that.
There is one thing which you really need to take into account with all of this: both Iraq and Iran saw sharp deteriorations of their amount of arable land available during the Middle Ages. Modern maps are not an accurate reflection of the pre-Islamic period because of the extensive neglect/damage done to the qanat networks during the Middle Ages, as well as the extensive salination and desertification of medieval Iraq. This has often been placed solely on the backs of the Mongols, but (IIRC) it's been recognised more recently that a general neglect in the post-Islamic period also helped. The modern land potential in no way reflects the premodern land available, which explains the general stagnation in Iran's population (although particularly vicious periods of violence, such as the Mongol conquests and the whole of the 18th century, did not help, especially given the general damage done to agricultural infrastructure by such extended periods of anarchy).
Are you talking about Iran here? And do you mean "Iran" in its modern borders, or a larger Iranian empire?Incidentally, most estimates I've seen for the Safavid period (Floor certainly does, I think Matthee and others do so too) put the population at 6-10 million. They're all a little vague with their numbers, however, so take them with a pinch of salt; the destruction of Esfahan's archives in the 18th century mean there's very little quantitative data to work with. We do know that Esfahan's population was probably about half a million, however (IIRC that comes from a VOC report drawing from the Safavid censuses, although their method was a little primitive IIRC, taking the median population of a household and applying that to the number of occupied homes in the city. Still, half a million seems reasonable).
excellent posts! Very interesting and informative.
I have a few questions and a few points to raise though.
Firstly about the division of Iranian society into settled and nomadic populations. Where there long-standing ethnic or linguistic differences between the two groups (or three groups if we're talking about city-dwellers, hills people and desert nomads)? Or could nomads and agriculturalists speak the same language within each region? How separate were they? Was it a case of the two groups trying to dominate one another? Or a more cooperative relationship?
What was the nature of the nomad communities? Were they similar to the equestrian steppe nomads of the north (i.e. Scythians, Huns, Turks, Mongols, etc), i.e. a fiercely independent tribal society with a very warlike nature? Or are we talking more about simple sheep/goat herding groups?
What was their military power? What was their political power?
Did the cavalry armies of the Parthians and Sassanians come mostly from the equestrian nomads? Or from the estates of the landed-aristocracy/nobility? I'm trying to get an idea of whether we're talking about something like Turko-Mongol steppe tribe cavalry armies, or something more like the knights of medieval Europe (which required a large, settled farming economy to support them).
Did the nomads have their own nobility, or were the nobility entirely based around control of wealth and farming land? (were the nobles all descended from nomads who had conquered?)
Demographics will be my next issue. I've researched the subject somewhat, but Iran is indeed a difficult place to understand. I've seen the figure of 4 to 5 million quoted for the pre-modern population, but I don't know how good that number is.
I agree with you that Maddison's demographic information may not be that great outside of Europe (he seems to always pick the lower-level possibilities, such as his frankly ridiculously low populations for the New World, prior to European contact). However, I have a hard time seeing how Iran could support a larger population in the 6th century than in the 19th.
Even though Iran was behind the West technologically by the 18th and 19th centuries, there was still surely significant new improvements in agriculture happen by that time. Iran must have received New World crops by that time, for instance (and indeed crops and varieties from Europe and elsewhere, due to the global trade network). So surely farming would be better, more productive and more sophisticated 200 or 300 years ago than it was 1,500 years ago, or 2,000 years ago. That would tend to suggest to me that the ancient and medieval populations couldn't have been as big as the 19th century population of the country.
Also, we should be a bit clearer with out definitions here. When you say the "Iranian plateau", do you mean basically the area of modern-day Iran? Or something larger, or smaller?
Well there are some problems with this. Firstly we need to define what we mean by "Mesopotamia". Are you talking about the whole of modern Iraq? Or just the flat, river-fed plains of the Tigris and Euphrates? Do you include the Iraqi part of the Zagros mountains in this (i.e. Kurdistan)?
In any case, I'm not sure if it is directly comparable to Egypt. I think that Egypt is a really special case of a pre-modern society that had a very high population density, because of the almost unique fertility of the Nile valley. Just go look at some satellite images (Google Earth) of the farmland in Egypt and the farmland in Iraq or Iran, and you'll see a massive difference in how green and productive it is (Mesopotamia and Iran look quite arid by comparison).
Frankly I would expect both Mesopotamia and even the good parts of Iran to have less population density than Egypt.
Modern Iraq has 437,000 square kilometres (or km2) and I would roughly divide that area up into: ~240,000 km2 of desert (probably with only a small nomad population), about 100,000 km2 is "northern Iraq" or "Zagros Iraq", or just "Iraqi Kurdistan", which is mostly hills and mountains and smaller river valleys (and some flat desert). Which leaves about 97,000 km2 of farmland in and around the Tigris and Euphrates rivers. Though I'm not sure how extensive the irrigation systems were in the Middle Ages and whether the interior area between the main river-courses was irrigated and farmable, or whether it was arid (this will mater a lot as most of the current arable area is not close to the rivers).
Actually 97,000 km2 may be far too generous an area, I've done some more measurements and come to perhaps only 39,000 km2 for the irrigated, arable area of Mesopotamia. That's if you exclude some of the main region between the rivers (some of which is dry even today) and the southern area, which used to be a giant swamp until very recently. Frankly I don't think it's possible to have 4 to 5.5 million people on an area of 39,000 km2 (that's a population density of 103 to 141 p/km2, which is ridiculously high) with pre-modern technology.
From my studies it seems that generally Medieval farming methods would allow an agricultural society to have densities from anywhere between about 10 persons per square kilometre (or p/km2), to perhaps as high as 50 p/km2. There may be some exceptional cases, like Egypt where the numbers would go as high as 70 p/km2, but no more than that is really possible with pre-modern technology.
If the 39,000 km2 of Mesopotamia had a density of say 30 p/km2, then it would have 1.17 million people (30 p/km2 is a perfectly reasonable figure for a highly developed Mediaeval society, it's about what France had in the 1300s). If it had 50 p/km2, then we're talking 1.95 million and if it had 70 p/km2 (which is about what the Nile valley and delta had) then it could be as high as 2.73 million. I can't imagine it could go any higher than that.
There might be an additional 0.5 to 1 million people in northern Iraq (the Kurdish area), which would give Iraq a total population of maybe 2 to 3.5 million, at most.
Hmm, I can't accept such a high figure for this area. Assuming by "Elymais" you mean roughly modern Khuzestan province, then I don't see how 1.5 to 2 million people could live there in the Early Middle Ages. If you are talking about a larger area than modern Khuzestan (such as including parts of Southern Iraq or something), then what I'm about to say may not apply, but here goes anyway.
I measured the area of farmland within Khuzestan, which is all located along a few river valleys/plains. And I came to a figure of about 16,500 square kilometres (or km2), which doesn't seem like enough to support such a large population. Although the total area of the province is 64,000 km2, most of that is very arid mountains and hills, which seem unlikely to support a major population (I'm sure there are scattered villages and nomads all over the place, but not dense agricultural settlements or cities). If I'm generous and increase the amount of farmland to say 18,000 km2, to account for all the little bits of arable and semi-arable land up in the mountains, that still isn't much.
Now if we calculate a population density for this area we can get somewhere.
If those ~18,000 km2 of arable land have a density of 10 p/km2, then it has a population of 180,000 people. If 20 p/km2 then 360,000, if 30 p/km2 then 540,000. If they had very high densities of 50 p/km2, then they would have 900,000 people. I don't see how it could be possible to support 1.5 to 2 million people though (that would suggest densities of 83 p/km2 and 111 p/km2, respectively, which is just too high). Now of course there are still going to be some people living in the remaining ~46,000 km2 area of the province, but since the climate is so harsh and the terrain so difficult I can only imagine that small numbers of nomads could live there, at a low population density.
Again, from my research, I determined that generally the population density of nomadic societies cannot go above 1 p/km2 and in many cases (like desert or arctic nomads) their density may be much less than that (I came to some general conclusions of about 0.5 p/km2 for steppe nomads, 0.2 to 0.1 p/km2 for desert nomads and 0.05 to 0.01 for arctic nomads). If the remaining 46,000 km2 of Khuzestan had say 1 p/km2, then it would only account for less than 50,000 people, which is a rounding error when considering the size of the settled agricultural population (almost an irrelevance when talking about the demography).
(you can read this post of mine, from a few months ago [and the following posts] to get an idea of the framework I'm using to calculate populations and some of my basic assumptions about pre-modern population density)
Here's an image from Google Earth where I've measured the areas of farmland in red:
(note that the large dark green area between the two main red areas is a swamp of some sort. You can see the Iraqi border in yellow and the Khuzestan province borders in light purple. I assumed that the last few kilometres of the river deltas are alluvial deposits which would not have existed yet by 2 millennia ago, as they get bigger gradually and the sea would have been further in-land back then. The red areas make up only about 25% of the province)
My guess is that Khuzestan in the Middle Ages would have only around 0.5 to 1 million people, but it would struggle to get higher than that.
What area exactly are you talking about here for "Arsacid Armenia"? How big is it? and can you show a map of the area you mean (I assume it's significantly bigger than modern-day Armenia?). Modern Armenia (then a Soviet republic) only had a population of 1.35 million in 1950 (according to UN data), so I can only assume it had a much smaller population in the Middle Ages (often, counties increased their population 5 or 10-fold from the Medieval era to the Modern era). For 2 to 2.5 million you must be talking about a large section of (modern-day) Turkey as well.
Hmm, Azerbaijan (the then Soviet republic) had a population of 2.9 million by 1950 (and that doesn't include any of Iranian Azerbaijan, which I don't have any historic data for). And it is a much less arid region than most of Iran or Iraq, so you may be underestimating its population (especially if you include all of Iranian Azerbaijan in that figure).
I would perhaps conclude a figure closer to the low end of that estimate (though it depends, as I'm not fully sure of the exact areas you're talking about). Maybe 7 million at a push.
Great pictures, but Mike Loades is English, for your information.
Hey, just thinking about this topic made me look up some old maps and diagrams, so some of you may also be interested in them:
Here's a map of the Satrapies of the Achaemenid Persian Empire, though I don't know if they are in any way similar to the Parthian and Sassanian periods.
And here's a useful diagram of the Iranian/Iranic languages (past and present):