Could the Persian Empire logistically support an army several million strong?

[Erpingham]

My own take is that after his father's first campaign failed Xerxes raised the stakes. Estimates from the primary sources put the army of Darius I at between 200 000 and 600 000 men. Working on the notion that Persians used huge numbers to intimidate enemies into submission and that Darius' army hadn't been big enough - in Persian eyes - for the job, Xerxes' solution was to increase its size fivefold or more.

Though I'll agree it is internally consistent, you can't really "prove" Xerxes numbers by reference to those of Darius.  If there is a systematic tendency to overstate the size of the barbarian horde, it will apply to both figures.  If we take modern estimates of Darius' army of 25,000-50,000 and multiply by five to get Xerxes, we get 125-250,000, remarkably similar to modern estimates.  But it doesn't prove anything.

But it's still crazy to put millions of men at the end of a fragile supply network that depended on good weather and the Greek fleet being overcome.

Napoleon in Russia in 1812 is probably a better comparator than Hitler.  A flawed supply strategy meant that even taking his main objective meant he was doomed and he had no chance of extracating his army, most of which was lost in the retreat. And he had the advantage of land supply and a wide front.

[Duncan Head]

To support Patrick's point about Greek poleis maintaining stockpiles of surplus grain, here's an anecdote from (pseudo-)Aristotle's Economics II:

The people of Selymbria had a law, passed in time of famine, which forbade the export of grain. On one occasion, however, they were in need of funds; and as they possessed large stores of grain, they passed a resolution that citizens should deliver up their corn to the polis at the regular fixed price, each retaining for himself a year's supply. They then granted right of export to any who desired it, fixing what they deemed a suitable price.

Not dated, unfortunately. Selymbria is on the Thracian coast, west of Byzantium.

the year's supply is what will carry them through to the next harvest, the extra would be whatever surplus the farmer would sell anyway.
A lot of Greek cities had laws forbidding the export of grain. Athens had at various times laws saying that any ship that entered the harbour with grain had to sell it. With cities perpetually on the edge of famine, these rules were important and policed.

So in the case of Selymbria, like a lot of cities they don't allow the export. They'd store grain, buying the surplus from their citizens every year, and every year they'd sell out of store to their citizens who didn't have their own grain. This allows the rotation of stocks and means that you rarely have 'old grain' lying about. If there was a poor harvest then the city could run the stock down.
(A bit like the intervention buying policy of the EU)
Finally after a number of good years, the city has got more grain that it needs so it'll sell the grain to anybody from abroad who wanted to buy it.

The point is:
- This is not quite a picture of a subsistence economy. Not that far off the subsistence level by modern standards, maybe, but the community overall expects to generate a surplus in most years.
- The polis decides what to do with that surplus, by compulsorily purchasing it at a price fixed by the state. So these are not completely "independent" farmers, in the phrase someone used earlier. This level of local state control is going to make it easier to meet the requirements of the Imperial superstate than if each individual farmer could decide what to do with his surplus.

[Justin Swanton]

My own take is that after his father's first campaign failed Xerxes raised the stakes. Estimates from the primary sources put the army of Darius I at between 200 000 and 600 000 men. Working on the notion that Persians used huge numbers to intimidate enemies into submission and that Darius' army hadn't been big enough - in Persian eyes - for the job, Xerxes' solution was to increase its size fivefold or more.

Though I'll agree it is internally consistent, you can't really "prove" Xerxes numbers by reference to those of Darius.  If there is a systematic tendency to overstate the size of the barbarian horde, it will apply to both figures.  If we take modern estimates of Darius' army of 25,000-50,000 and multiply by five to get Xerxes, we get 125-250,000, remarkably similar to modern estimates.  But it doesn't prove anything.

Sure. The idea was that if we accept Herodotus' figures for the Persian army, why would Xerxes go for such a huge number, since Persian armies before that time habitually numbered in the hundreds of thousands and not the millions? (if we accept the primary sources) It makes sense in the context that for the Persians quantity was the decisive battle-winner and thus if the Greeks were able to stop half a million men, then one and a half million would be needed to put them down.

But it's still crazy to put millions of men at the end of a fragile supply network that depended on good weather and the Greek fleet being overcome.

Napoleon in Russia in 1812 is probably a better comparator than Hitler.  A flawed supply strategy meant that even taking his main objective meant he was doomed and he had no chance of extracating his army, most of which was lost in the retreat. And he had the advantage of land supply and a wide front.

Fair enough. One could argue that Xerxes came closer to achieving his objective than Napoleon. If he could just have beaten the Greek fleet or at least kept it at bay he might have achieved something on land. But that's doubtful as well. Most of the Persian army were not actually fighting men - just a cheering chorus that looked big and frightening - and the part that were proper fighters fought the Greeks at Plataea and lost.

[Justin Swanton]

  It may be worth noting that when, later in the dynasty, rebellions did occur, the task of suppressing them was customarily given to another satrap and did not necessarily involve the bulk of the core Medo-Persian forces.  Of course by then mercenary Greeks had replaced the Medes and Persians as the cutting edge of Achaemenid forces, and Achaemenid control seems to have centred as much on manipulating the Greek cities as anything else.

I have made a point of ignoring the later Persian monarchs, because they cannot have had an impact on the thinking of Xerxes or why he took the army he did

But his successor Artaxerxes I was probably a success, in that he seized power in a coup, held in with only a few rebellions and probably died in his bed

His successor Xerxes II was assassinated after 45 days
His successor Sogdianus reigned for six months before being overthrown by his brother who executed him
Darius II took power in a coup and seems to have died of natural causes after a reign of 19 years
Artaxerxes II inherited the throne and immediately faced revolt from his brother Cyrus. He then lost Egypt but did defeat the Satraps revolt.
Artaxerxes III waded to the throne in blood, but the killing might have been done by his father to ensure the succession, finally put down the Egyptian revolt and may have died of natural causes or may have been murdered by  Bagoas, the  Vizier
Artaxerxes IV, raised to the throne by a coup plotted by Bagoas, and ended up being poisoned by Bagoas
Darius III  Killed Bagoas and ended up being murdered by Bessus, his cousin.

So of the first 5, one died in his bed and three were murdered, and one died in battle.
Of the next 8, 3 or 4 died of natural causes and 4 or 5 may have been murdered.

If you were Great King, King of Kings the one group of people you didn't trust was your family and the Persian aristocracy!

The point appears to be that the subordinates of the king either did what they were told or tried to assassinate their master. There was no middle ground. If they didn't do what they were told they were executed themselves. Knife or poison (or whatever was used) was the only means of objecting to the Great King's commands. It's not like, say, a Mediaeval monarchy in which the dukes and barons had to be persuaded to do what the king wanted or at least not oppose him.

[Andreas Johansson]

- This is not quite a picture of a subsistence economy. Not that far off the subsistence level by modern standards, maybe, but the community overall expects to generate a surplus in most years.

Well, a subsistence economy needs to. If the average year doesn't give some surplus you're doomed when a bad year comes around. If the average surplus is zero, expected reserves when a bad year happens is also zero.

(Mathematically, you could have a deficit or breakeven most years compensated for occasional bumper years that keep the average positive, but apart from the problems of storage Jim has mentioned that's not particularly realistic ecologically. You're more likely to have occasional disaster years when the crop is eaten by locusts (Persian or otherwise).)

[Flaminpig0]

  It may be worth noting that when, later in the dynasty, rebellions did occur, the task of suppressing them was customarily given to another satrap and did not necessarily involve the bulk of the core Medo-Persian forces.  Of course by then mercenary Greeks had replaced the Medes and Persians as the cutting edge of Achaemenid forces, and Achaemenid control seems to have centred as much on manipulating the Greek cities as anything else.

I have made a point of ignoring the later Persian monarchs, because they cannot have had an impact on the thinking of Xerxes or why he took the army he did

But his successor Artaxerxes I was probably a success, in that he seized power in a coup, held in with only a few rebellions and probably died in his bed

His successor Xerxes II was assassinated after 45 days
His successor Sogdianus reigned for six months before being overthrown by his brother who executed him
Darius II took power in a coup and seems to have died of natural causes after a reign of 19 years
Artaxerxes II inherited the throne and immediately faced revolt from his brother Cyrus. He then lost Egypt but did defeat the Satraps revolt.
Artaxerxes III waded to the throne in blood, but the killing might have been done by his father to ensure the succession, finally put down the Egyptian revolt and may have died of natural causes or may have been murdered by  Bagoas, the  Vizier
Artaxerxes IV, raised to the throne by a coup plotted by Bagoas, and ended up being poisoned by Bagoas
Darius III  Killed Bagoas and ended up being murdered by Bessus, his cousin.

So of the first 5, one died in his bed and three were murdered, and one died in battle.
Of the next 8, 3 or 4 died of natural causes and 4 or 5 may have been murdered.

If you were Great King, King of Kings the one group of people you didn't trust was your family and the Persian aristocracy!

The point appears to be that the subordinates of the king either did what they were told or tried to assassinate their master. There was no middle ground. If they didn't do what they were told they were executed themselves. Knife or poison (or whatever was used) was the only means of objecting to the Great King's commands. It's not like, say, a Mediaeval monarchy in which the dukes and barons had to be persuaded to do what the king wanted or at least not oppose him.

I can't see how we can know that for certain - the information provided could just as easily indicate the Persian King having to constantly manage his faction through negotiation and bribes etc.

[Jim Webster]

To support Patrick's point about Greek poleis maintaining stockpiles of surplus grain, here's an anecdote from (pseudo-)Aristotle's Economics II:

The people of Selymbria had a law, passed in time of famine, which forbade the export of grain. On one occasion, however, they were in need of funds; and as they possessed large stores of grain, they passed a resolution that citizens should deliver up their corn to the polis at the regular fixed price, each retaining for himself a year's supply. They then granted right of export to any who desired it, fixing what they deemed a suitable price.

Not dated, unfortunately. Selymbria is on the Thracian coast, west of Byzantium.

the year's supply is what will carry them through to the next harvest, the extra would be whatever surplus the farmer would sell anyway.
A lot of Greek cities had laws forbidding the export of grain. Athens had at various times laws saying that any ship that entered the harbour with grain had to sell it. With cities perpetually on the edge of famine, these rules were important and policed.

So in the case of Selymbria, like a lot of cities they don't allow the export. They'd store grain, buying the surplus from their citizens every year, and every year they'd sell out of store to their citizens who didn't have their own grain. This allows the rotation of stocks and means that you rarely have 'old grain' lying about. If there was a poor harvest then the city could run the stock down.
(A bit like the intervention buying policy of the EU)
Finally after a number of good years, the city has got more grain that it needs so it'll sell the grain to anybody from abroad who wanted to buy it.

The point is:
- This is not quite a picture of a subsistence economy. Not that far off the subsistence level by modern standards, maybe, but the community overall expects to generate a surplus in most years.
- The polis decides what to do with that surplus, by compulsorily purchasing it at a price fixed by the state. So these are not completely "independent" farmers, in the phrase someone used earlier. This level of local state control is going to make it easier to meet the requirements of the Imperial superstate than if each individual farmer could decide what to do with his surplus.

well it depends on the nature of the polis, but those farmers who were the major grain producers would be in the hoplite class and in most cases they would be the ones who voted or whose opinions would have counted. I'm not sure of the population size of Ancient Selymbria but in our terms it would have been small.
In Greek cities one of the causes of tension between the classes was that the Hoplite class as landowners and grain sellers wanted exports to keep the price up. The 'navy rabble' hoi polloi or whatever wanted grain prices kept down and where they had the vote, they would tend to vote to ban exports, except where there was enough grain in the market to ensure prices stayed low.

The other thing to remember is the structure of agriculture. In most cities it's probable that most farmers weren't members of the hoplite class. They'd be men with some land. There would be a drive to produce enough grain to guarantee their family's supply. But if you haven't got a lot of land, once you've sown your grain, you're better growing 'unpolitical' higher value, easy to sell crops than just growing more grain. So once you've got your family grain, then you've got your olive trees, figs, perhaps even a few vines, but more importantly your vegetables which can be a cash crop to sell in town.

The city cannot command the production of grain, it can merely attempt to control the sale. If it keeps prices too low then it'll merely discourage production and you'll get a long term drift to olives and other crops. If it has the prices too high without some other policy like a food dole, it will run the risk of discontent among the landless

[Jim Webster]

The point appears to be that the subordinates of the king either did what they were told or tried to assassinate their master. There was no middle ground. If they didn't do what they were told they were executed themselves. Knife or poison (or whatever was used) was the only means of objecting to the Great King's commands. It's not like, say, a Mediaeval monarchy in which the dukes and barons had to be persuaded to do what the king wanted or at least not oppose him.

the problem is, we don't know that. We haven't a clue what discussions went on within the higher echelons. Herodotus does mention members of the Persian aristocracy having problems with the whole invasion plan, then there is the story about Xerxes, Artabanus  and dreams. Herodotus portrays Xerxes in a particular light, but that need to have been how the Persian aristocracy saw him, or how he saw himself.
The Accession of Darius and the legend of the six or seven families show that some of the Persian aristocracy regarded the king as little more than first among equals

[Erpingham]

I seem to detect a drift to the periphery here, with much on the nature of Persian kingship and governance and the nature of the city-state economy.

Way back, the challenge was made to those who take an orthodox position as to why they felt Herodotus' figures couldn't be accepted at face value.  This led to a list of maybe fifteen reasons. 

Justin has determinedly tried to tackle some of them head on.  I think he has drawn out a lot of quantification of the problem and introduced some notable images, like the giant mobile squatter camp.  He has also tried to create a model of how the advance across northern Greece may have been done (though I think we still await the grand scheme of how the Asiatic magazine cities, Greek depots, supply conveyor, "feed for a day" cities, baggage train, fodder cutters, hydraulic engineers and camp builders all fitted together).

At the end of the day, I don't know if any minds will be changed.  I know that all the figure work has confirmed to me that the orthodox view is closer to a practical reality at the same time pointing up the flaws of some of the more minimal estimates, like Delbruck or Young.

Incidentally, I managed to find this estimate by Percy Molesworth Sykes in his History of Persia. Percy Sykes was a British general who fought the Turks in WWI.  I do not know whether that makes him a contemptable source, like Maurice, or a quotable one, like Lawrence, but here is what he says.  Having enumerated the army and navy following Herodotus' figures, he goes on

Reinforcements in Europe and servants bring up the figures
to over five millions, a total which it is impossible to accept.
In view of the reliance the Persians placed on numbers and
the size of the empire, we are perhaps justified in assuming
that the land and sea forces combine inclusive of followers,
aggregated about one million. After deducting the crews,
this total would give at the most 200,000 fighting men, so
numerous are camp followers in an Eastern army, an allowing
for strong detachments posted on the lines of communication,
for sickness, and other causes, the actual numbers that met the
Greeks at sea, and finally on land, were not overwhelming.

Surprisingly similar to his contemporary Maurice.  Perhaps they corresponded, or maybe the same staff training led to the same estimate of scale.

Note Sykes takes the fleet at face value, which I suspect is an error.  The trireme count can be said to come from Aeschylus (though the Greek is said to be ambiguous) but the 3000 lesser ships (triaconters, penteconters and horse transports) have no independent authority and may suffer nautical "horde" mode.

[Justin Swanton]

To support Patrick's point about Greek poleis maintaining stockpiles of surplus grain, here's an anecdote from (pseudo-)Aristotle's Economics II:

The people of Selymbria had a law, passed in time of famine, which forbade the export of grain. On one occasion, however, they were in need of funds; and as they possessed large stores of grain, they passed a resolution that citizens should deliver up their corn to the polis at the regular fixed price, each retaining for himself a year's supply. They then granted right of export to any who desired it, fixing what they deemed a suitable price.

Not dated, unfortunately. Selymbria is on the Thracian coast, west of Byzantium.

the year's supply is what will carry them through to the next harvest, the extra would be whatever surplus the farmer would sell anyway.
A lot of Greek cities had laws forbidding the export of grain. Athens had at various times laws saying that any ship that entered the harbour with grain had to sell it. With cities perpetually on the edge of famine, these rules were important and policed.

So in the case of Selymbria, like a lot of cities they don't allow the export. They'd store grain, buying the surplus from their citizens every year, and every year they'd sell out of store to their citizens who didn't have their own grain. This allows the rotation of stocks and means that you rarely have 'old grain' lying about. If there was a poor harvest then the city could run the stock down.
(A bit like the intervention buying policy of the EU)
Finally after a number of good years, the city has got more grain that it needs so it'll sell the grain to anybody from abroad who wanted to buy it.

Duncan's point though is that Selymbria was capable of falling short of its requirements only if it exported its grain. The law implied that if the city kept all its harvest it would manage in a poor year. The problem then becomes what to do with the excess grain in a normal or good year. This all suggests that a Greek polis like Selymbria on average produced more than it needed (their law had left them with a large surplus they now had to dispose of). Certainly enough to supply a large passing Persian army with a meal.

[Erpingham]

Duncan's point though is that Selymbria was capable of falling short of its requirments only if it exported its grain. The law implied that if the city kept all its harvest it would manage in a poor year. The problem then becomes what to do with the excess grain in a normal or good year. This all suggests that a Greek polis like Selymbria on average produced more than it needed.

I didn't want to get into this but I think the conclusion here goes to far.  In order not to be left with a danger of famine/massive inflation, Selymbria restricted exports to maintain a constant supply.   Occassionally, surpluses were sufficient to operate the buffer and have an exportable surplus but we don't know how often this happened.  We also don't know how often Selymbria needed to buy grain to overcome a shortfall and restock the granaries.  So assuming this points to a regular surplus is overstating things.

[Duncan Head]

This all suggests that a Greek polis like Selymbria on average produced more than it needed (their law had left them with a large surplus they now had to dispose of). Certainly enough to supply a large passing Persian army with a meal.

I don't actually think that the first sentence necessarily implies the second.

[Justin Swanton]

Duncan's point though is that Selymbria was capable of falling short of its requirments only if it exported its grain. The law implied that if the city kept all its harvest it would manage in a poor year. The problem then becomes what to do with the excess grain in a normal or good year. This all suggests that a Greek polis like Selymbria on average produced more than it needed.

I didn't want to get into this but I think the conclusion here goes to far.  In order not to be left with a danger of famine/massive inflation, Selymbria restricted exports to maintain a constant supply.   Occassionally, surpluses were sufficient to operate the buffer and have an exportable surplus but we don't know how often this happened.  We also don't know how often Selymbria needed to buy grain to overcome a shortfall and restock the granaries.  So assuming this points to a regular surplus is overstating things.

OK. What we can draw out from the passage is that Selymbria passed the law when they did have a famine and had exported their grain (hence the law). It was the combination of imprudent export plus poor harvest that got them in the dwang. The city then manages in bad years by not exporting and possibly by importing, which means that in a bad year they can get by (their economy is robust enough to pay for the imports) and in an average and good year they will produce a surplus. If they had a lot of bad years - in other words, if they grew only enough to feed themselves provided the harvest was decent - then I suspect they wouldn't have passed a law, just had a general practice of not exporting their grain.

[Jim Webster]

Duncan's point though is that Selymbria was capable of falling short of its requirements only if it exported its grain. The law implied that if the city kept all its harvest it would manage in a poor year. The problem then becomes what to do with the excess grain in a normal or good year. This all suggests that a Greek polis like Selymbria on average produced more than it needed (their law had left them with a large surplus they now had to dispose of). Certainly enough to supply a large passing Persian army with a meal.

The purpose of storing grain in a polis is that most years you're OK. But a bad year might destroy you. Even a 5% shortfall can lead to price gouging, riots and violent political change.
The fact that a city banned exports shows how important this is, and how close to the edge they habitually sailed.
If you look at the example Duncan gave, "On one occasion, however, they were in need of funds; and as they possessed large stores of grain....."
So it was a one off.
It wasn't a large surplus they now had to dispose of, they were in need of funds and the only way the city had of raising those funds was to sell some of their grain. It was a gamble the city felt prepared to take.
With the funds raised they appear to have gone back to squirrelling away grain again, breathing a sigh of relief that they'd 'got away with it' this once.

[Erpingham]

If they had a lot of bad years - in other words, if they grew only enough to feed themselves provided the harvest was decent - then I suspect they wouldn't have passed a law, just had a general practice of not exporting their grain.

The grain belongs to private individuals who could, without a law, sell to anyone they liked.  A bad harvest was unlikely to be highly localised, so demand would be high for any surplus.  If you don't constrain the market forces, local grain will be sold to a wealthy city down the coast for a tidy profit and  the polis will end up with paying top import prices or letting the poor starve.  So you constrain the market.