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

[Justin Swanton]

Justin (not me - the martyr) and others give similar figures of these magnitudes.

Justin the epitomator of Trogus, presumably? Not generally thought to be the same bloke as the martyr.

Oh, right.

[Justin Swanton]

Justin the epitomator of Trogus, presumably?

Wow, you could build an entire Dr Who episode around that name !

Epitomate! Epitomate!

Justin always struck me as the sort of person who'd describe an episode of Doctor Who for you.
But he'd have missed the point, have seen only half the episode, but had also seen a trailer for the next season

And kept getting the companion's name wrong

He'd have to epitomate the Doctor and then find some other way of dealing with K9 - you can't epitomate inorganic entities (but I don't have to describe this in detail, do I?)

[Erpingham]

Justin the epitomator of Trogus, presumably?

Wow, you could build an entire Dr Who episode around that name !

Epitomate! Epitomate!

Justin always struck me as the sort of person who'd describe an episode of Doctor Who for you.
But he'd have missed the point, have seen only half the episode, but had also seen a trailer for the next season

And kept getting the companion's name wrong

Jim, have you ever thought such an unreliable epitomator might make an interesting character in a fantasy novel?

[Jim Webster]

Jim, have you ever thought such an unreliable epitomator might make an interesting character in a fantasy novel?

Now there is an idea!

[Flaminpig0]

Incidentally (to wind up my overlong and unintended foray into this thread), has anyone tried wargaming an army of 3,000,000, or even 800,000? At a figure scale of 1:20, 800,000 men would only need 40,000 figures. Or alternatively, take something like DBA. If we take a standard 12 element army to represent about 40,000 men, then 800,000 would be represented by 240 elements, 3M by 900. Sounds doable - I can't think of any practical objections to that.

I was going to suggest for the next conference a committee  game on planning the 3 million man plus assorted supporting creatures march to Thermopylae.

[Justin Swanton]

To paraphrase Richard, and repeat something said earlier, neither the position "If you can't show it's impossible, it must be true" nor "If you can't show it's possible, it must be false" take us far.  Instead we need to look at a balance of probabilities.

Though Justin fails to see any "killer" arguments, the conventional position does pile up a lot of obstacles.  So far, the "sources first" advocates have struggled to answer these and the balance of probability is strongly with the orthodox position.  It may be that is in part due to the fact that we only see bits of argument at a time and Justin would do better if he could provide a summary of his case. 

As I understand it, there was a 3 year preparation period where stores and armies were assembled and a large construction battalion was sent to Greece.  We know the construction battalion built a canal - it is archaeologically detectable.  Apparently they also surveyed the route, marking out daily camp sites and weekly depot sites. Sometime in this three years they built a 300 m wide road from the Hellespont to Thermopylae.  Then, with the approach of D-day, bridges were built and rebuilt and an advanced force set off to establish the first weekly depots.  They advanced carefully to avoid trampling any pasturage.  Then, a few days later, the engineering vanguard crossed the Hellespont and proceeded a day in advance of the main army, harvesting fodder, building irrigation systems and laying out camps.  These were then followed in a very orderly and disciplined fashion by the remained of the army (I don't think we've discussed march order) who are spread over two days.  The back of the army camps where the previous one did the day before (I'm unclear here whether the first half of the army has left fodder for the second half, whether the second half use transported fodder or whether they go further abroad to forage).  The naval supply chain is delivering supplies across beaches using lighters to stock the seven day depots.  When the depot is stocked, they move to the next and the fleet of lighters sails down the coast.  Warships are presumably covering this operation, based at a beach near the depot, close enough to draw supplies from it and with an independent water supply (we haven't discussed the naval side).  I hope this does justice to the theory Justin is developing.

I presume part of Justin's geographic research is to identify potential sites for his daily camps and his depots, plus operating beaches for 1207 warships?

A good summing up. I would modify it slightly thus:

The Persians marked out suitable campsites well in advance with the help of local knowledge. They also surveyed the route the army would take. They didn't so much build a 600m wide road as oblige the locals to spend four years clearing away obstructive vegetation and other obstacles - not every obstacle as most could simply be walked around. Keep in mind that the human body is designed to walk on ground - not road but ground - which means that if in good shape it can handle terrain that is not too difficult without exhausting itself. Wild animals do it all the time.

The locals are also required to prepare the campsites just before the arrival of the army, with perhaps an advance guard arriving to oversee operations. This preparation consists of gathering up the brush and grass which clears the ground and supplies fodder for the animals. If possible, channels would be dug to facilitate watering the animals. The army marches along the cleared 'road', in a loose and not particularly orderly fashion but with reasonable spacing between individuals and animals to allow for some bunching up when flowing around obstacles.

The entire army marches in one day from campsite to campsite. This is easier if the army splits into several columns which happens whenever possible but can be done even when the army is together. With more than 15 hours visibility each day, if the head of the column starts out at the crack of dawn and marches with one half-hour break round the three-hour mark, the tail of the column will reach the next campsite while it is still light and the entire army will have advanced about 20km. Men and animals will actually spend only 7 hours on the march at a moderate pace of 3km/h, so the journey will not be too fatiguing for them. In fact, I get the impression that Xerxes is not really hard on his army. Bar the whipping at the Hellespont - necessary to get the men across the narrow bridges as quickly as possible - he does not seem to have unduly driven his troops.

I envisage the food transports as ships small enough to beach (50 tonners) and unload grain in sacks by short chain gangs to waiting mules just a few yards away on the sand - or even right next to the ships' sides, why not? (unless the sacks would get wet) Unloading a ship must be done with maximum speed to permit enough ships to arrive at and leave the shore in a day which actually isn't too many ships. Most of the Aegean shore on the northern and western coast is beach - I checked - so the ships won't lack for space. Rocky shores are precisely where the army won't be marching. It'll be inland for those stretches.

The depots are already stocked when the army and ships arrive and serve to ease the logistics burden of the navy.

[Patrick Waterson]

Could I just check how the army is carrying its non-food supplies e.g. the large numbers of arrows it needs?  Are we assuming that these are being carried by the camp followers (disregarding those on special duties)?

Given the occasional reference to 'harmata' (which could be chariots or wagons) there may have been some wheeled transport in the Persian lineup.  Whether these rather than pack animals would be carrying ammunition, farriers' tools etc. is anyone's guess, but I would incline to iit being the likelier option.

[Patrick Waterson]

While Jim has raised it in something of an exasperated tone, I do think the "sources first" advocates perhaps need to sort out the apparent contradictions the model of the Persian army is generating.  As a force, the current model requires it to be extremely efficient, with discipline among its camp followers that many later regular armies would envy for their soldiers.  Its march and camp discipline are greatly in advance of the Romans.  Is this high degree of regularity and discipline consistent with other sources or even with Herodotus himself?  Why would such a force need to be driven forward with whips?  Does their battlefield performance indicate great discipline and drill?

'Discipline' may be the wrong word here; I would suggest 'practice'.  They are doing something they are intimately familiar with from generations of tradition and, for many but by no means all, actual campaigning.  The difference is principally scale, and we see this applied in planning: cutting the Athos canal, spending four years in preparation for the war, ordering the gathering of food and reasources ahead of the march and applying what looks like maritime overkill to supply requirements.

An analogy might be India in almost any age: Indian armies were large, often astonishingly so compared to their opponents, but were they 'disciplined' or effective?

[Jim Webster]

Could I just check how the army is carrying its non-food supplies e.g. the large numbers of arrows it needs?  Are we assuming that these are being carried by the camp followers (disregarding those on special duties)?

Given the occasional reference to 'harmata' (which could be chariots or wagons) there may have been some wheeled transport in the Persian lineup.  Whether these rather than pack animals would be carrying ammunition, farriers' tools etc. is anyone's guess, but I would incline to iit being the likelier option.

did Persians shoe their horses? I ask purely out of interest because the Romans used hipposandals which were 'tied on'
But even without horseshoes there'd be plenty for metal workers to do so I have no doubt there would be metal and leather workers with their tools. Whether they were men attached to units making stuff in traditional patterns I don't know.

You'd certainly need bronze smiths as well as blacksmiths because stuff would break through fair wear and tear

[Patrick Waterson]

did Persians shoe their horses? I ask purely out of interest because the Romans used hipposandals which were 'tied on'
But even without horseshoes there'd be plenty for metal workers to do so I have no doubt there would be metal and leather workers with their tools. Whether they were men attached to units making stuff in traditional patterns I don't know.

As far as I know they did not, nor did the Macedonians a century and a half later; Alexander's men in India were complaining that their horses' hooves were worn down almost to the frogs.

Given the number of people in Xerxes' army carrying cane spears and reed arrows, there may have been less metalwork than we think, but there must have been some on account of the better-armoured contingents, e.g. Medes and Persians.  Whether they needed to have smiths etc. with the army all the time or whether they could get away with carrying many of them and their materials on ships and just meeting up on occasion in coastal camps to fix things I do not know.

The item most likely to be in need of ongoing manintenance would, to my mind, be footwear.  I would guess this would fall to camp followers who were handy leatherworkers; leather, or at least uncured leather, would probably not be too hard to come by during the campaign.

[Patrick Waterson]

I presume part of Justin's geographic research is to identify potential sites for his daily camps and his depots, plus operating beaches for 1207 warships?

That is for Justin to decide.

On a related matter, one wonders how many of Xerxes' troops moved by sea, and in what circumstances.

"Xerxes marched past these Greek cities of the coast, keeping them on his left. The Thracian tribes through whose lands he journeyed were the Paeti, Cicones, Bistones, Sapaei, Dersaei, Edoni, and Satrae. Of these, the ones who dwelt by the sea followed his army on shipboard; the ones living inland, whose names I have recorded, were forced to join with his land army, all of them except the Satrae." - Herodotus VII.110

The 'sightseeing' at Thermopylae I have already mentioned.  The point here is that sealift is a capability the Achaemenids could easily have used to get at least part of their army past any constriction on their route.  This in itself complicates any choke-point calculations and increases the likely size of the Achaemenid army.

Getting back to warships and beaches, we have one example in Herodotus VII.118:

"The Persian fleet put to sea and reached the beach of the Magnesian land, between the city of Casthanaea and the headland of Sepia. The first ships to arrive moored close to land, with the others after them at anchor; since the beach was not large, they lay at anchor in rows eight ships deep out into the sea."

If I read this correctly, one eighth of the ships 'moored close to land' without actually beaching.  The rest dropped the hook (or stone) in open water.

[Patrick Waterson]

Well I think you go wrong there at the first step. Our understanding of the Persian Empire is absolutely fine if we reject these figures - indeed because most people reject these figures, our understanding of the Persina Empire is already based on the assumption that these figures are incorrect.

For circular arguments, this one has to win a prize.

This means we have to respect the sources as much as possible. If we cover them with a blanket of doubt right from the word go, affirming that the only certitude we have is from scientific experimentation and you can't perform scientific experiments on history, then history as a serious subject of study is dead.

Wrong again! Without getting all meta, there are lots of sources of information a historian can use, and naively accepting every statement of an ancient author is not the only option.

But we are not naively accepting statements; we are taking them as stated and working out whether there is any reason we should not do so.

I suggest that instead of scientific certitude (which is impossible) we go for moral certitude, i.e. examining the writer of the period in the same way a lawyer cross-examines a witness in court. A witness may be hostile - not inclined to spill the necessary beans - but it is still possible to glean the facts from what he/she says with enough certitude to arrive at a judgment.

Well I hope I never have to have you as my lawyer! I'm all for the method (moral certitude is fine by me). But in most legal cases, you wouldn't rely on a single witness statement without, at the very least, examining the expertise, knowledge and motivations of the witness, looking for corroborating statements and (above all) considering forensic evidence. You would reject such methods and rely entirely on the unvarnished witness statement.

This seems to me to be a misrepresentation of what Justin is saying.  Herodotus is a careful 'witness' and not credulous about what he reports, as anyone who reads him extensively will see.  Just what forensics we can get from 480 BC is an open question, but internal consistency in Herodotus - the extensive preparation time, the shipping required to support such a huge army, the crossing time to get it across the Hellespont on a bridge constructed from warships whose dimensions we can be reasonably sure of - works out rather too well to be dismissed.

[Justin Swanton]

You could also include Ctesias for example (800,000 for Scythia and Xerxes in Greece), or Simonides (3,000,000 at Thermopylae). You might also consider the fact that the numbers in each source are different from each other - but it's OK for the purposes of this to talk about orders of magnitude.

Since these are pretty much all the sources we have for Persian history and to reject them to this extent means putting our understanding of the Persian Empire in question,.

Well I think you go wrong there at the first step. Our understanding of the Persian Empire is absolutely fine if we reject these figures - indeed because most people reject these figures, our understanding of the Persina Empire is already based on the assumption that these figures are incorrect.

If a writer like Herodotus makes a Persian army ten times larger than it actually was then he can't be trusted in anything else he affirms. Keep in mind he goes into some detail about where he got the number from: Xerxe's counting his men in batches of 10 000's. Add to this the 7 days to get the army across the Hellespont, a campsite stretching 20 miles along the shore south of Therme, and so on. If other writers get Persian numbers as wildly wrong as Herodotus then they can't be trusted either. Size is significant - a general would have to have an accurate idea of how large his army was so as to ascertain if it was capable of beating his opponent. It's a fundamental piece of information. Saying Herodotus et al. can't be trusted to get it right completely undermines their reliability as historians. Saying they can get it right and deliberately exaggerate it by a factor of ten is worse.

... I propose that one examine if these numbers are feasible. Apart from the affirmations of the sources, I can't scientifically prove that Persian armies were this big. Nobody can. But neither can anyone scientifically prove that they were considerably smaller. There is no hard evidence either way other than the sources themselves.

That depends on what you mean by 'hard evidence'. Plenty of fairly hard evidence has been presented in this thread. And most people would argue that a statement of Herodotus is not hard evidence (this was already the view in antiquity of course - many 'sources' - writers and historians in antiquity - regarded Herodotus as wholly unreliable - whether justly or not is another matter - but it's not some modern fashion).

By hard evidence I mean something concrete - traces of the passage of the Persian army to show how wide it marched, of the layout of the camps to show how big they were, and so on. Something archaeological in other words. Herodotus himself has a good contemporary reputation as a reliable source, which is saying something.

This means we have to respect the sources as much as possible. If we cover them with a blanket of doubt right from the word go, affirming that the only certitude we have is from scientific experimentation and you can't perform scientific experiments on history, then history as a serious subject of study is dead.

Wrong again! Without getting all meta, there are lots of sources of information a historian can use, and naively accepting every statement of an ancient author is not the only option.

Actually there isn't. About 90% of what we know about human events in the past comes from what people during or near that past wrote about them. 'Human events' as opposed to architecture, burial practices, arms and armour, and the like. Knowing what a Persian arrowhead looks like tells us nothing about the route the Persian army took during the Greek campaign. And it's not about naively accepting. Check my courtroom response below.

I suggest that instead of scientific certitude (which is impossible) we go for moral certitude, i.e. examining the writer of the period in the same way a lawyer cross-examines a witness in court. A witness may be hostile - not inclined to spill the necessary beans - but it is still possible to glean the facts from what he/she says with enough certitude to arrive at a judgment.

Well I hope I never have to have you as my lawyer! I'm all for the method (moral certitude is fine by me). But in most legal cases, you wouldn't rely on a single witness statement without, at the very least, examining the expertise, knowledge and motivations of the witness, looking for corroborating statements and (above all) considering forensic evidence. You would reject such methods and rely entirely on the unvarnished witness statement.

Obviously the background, motivations and general reliability of the witness is examined, corroborating evidence brought forward and forensic evidence submitted if it exists. The point is that witness testimony is acceptable in a court of law, even the testimony of hostile witnesses (cross-examination reveals consistencies or lack of them). In this case we have the testimony of a number of witnesses all of whom give Persian armies huge numbers. We have little or no forensic evidence. Expert testimony thus far in this thread does not refute the affirmations of the witnesses.

Ditto for Herodotus. It just needs a good dose of common sense. 

Sure. Funnily enough, common sense is what I suspect most people in this thread feel you lack... It's a slippery concept, common sense.

Remarks like this I just prefer to let go.

The only thing we can establish with some sort of scientific accuracy is the practicality of moving 3 million men from Asia Minor to Greece. If it is possible then I'm quite happy to take Herodotus at his word. To establish if it is possible from my armchair (actually a rather uncomfortable plastic chair) simply means establishing whether there are any insuperable obstacles, as opposed to obstacles that can be overcome with some preparation, ingenuity and organisation.

OK I'm fine with the method - but you apply exceptionally high standards of proof to arguments against the practicality of it, while being remarkably easy to convince as to its practicality. For example, drawing a diagram of a 1 in 10 slope, placing some figures on it, and presenting this as proof that several million men can walk en masse across the hills of northern Greece, is not, I feel, an argument that would convince many juries. But maybe the proof of the pudding is in the eating - we are all unbiassed jurors here, full of common sense - if you can convince us then that's a big step forward for your argument. If you can't, then maybe you should consider the possibility that it is your argument, rather than our understanding, or common sense, or historical method, that might be flawed after all.

Showing that the terrain is traversible if cleared of obstacles like difficult vegetation is a fair argument IMHO. And one can hardly compare posters in this thread to a jury. When a jury walks into a courtroom it knows a crime has been committed but it has not seen the evidence proving the innocence or guilt of the accused. The jurors are, as far as is humanly possible, mentally neutral. They acquire their understanding of the case from the evidence and arguments put forward by Prosecution and Defence.

Here however everyone pretty much had settled opinions on the topic before the thread was posted. I'm willing to change my mind if the big army hypothesis can be proven to be untenable. I'm even willing to put forward arguments against it - I'll put forward another one in my next post. But the point is that nobody is going to change their minds on the basis of probabilities and circumstantial evidence, I accept that.

Incidentally (to wind up my overlong and unintended foray into this thread), has anyone tried wargaming an army of 3,000,000, or even 800,000? At a figure scale of 1:20, 800,000 men would only need 40,000 figures. Or alternatively, take something like DBA. If we take a standard 12 element army to represent about 40,000 men, then 800,000 would be represented by 240 elements, 3M by 900. Sounds doable - I can't think of any practical objections to that.

It would work in Optio terms. Each element has a frontage of 50 yards and occupies a battlefield square 2 elements wide or 100x100 yards. So fill a square with 8 bases and you get 10 000 men if you assume depth now corresponds to width in scale. Create a battlefield 40 squares wide and 20 squares deep - that's about 4 x 2m. The Persians deploy their 1 000 000 or so infantry 20 squares wide and 5 squares deep and put the cavalry on the 10 squares on either side of the infantry. The infantry will number about 800 bases. Have fun painting that!

[Flaminpig0]

The Persians deploy their 1 000 000 or so infantry

Just a thought but is it possible to work out how long it would take to deploy 1,000,000 or so infantry?

[Jim Webster]

Jim, have you ever thought such an unreliable epitomator might make an interesting character in a fantasy novel?

tis done Anthony
I'll stick it on the Tallis Steelyard Blog and put a link in to this thread for you