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

[Patrick Waterson]

  One may recall that Persians were taught to do three things from boyhood: ride a horse, shoot a bow and tell the truth.[/i]

Alas that our history was written by a Greek, had a Persian written it we might have got the correct numbers 

As it happens, we are in luck.  We have a Greek writing down the numbers he was given by a Persian.  Had it been the other way around, the Persian would have recorded the numbers he was given by the Greek.

[Jim Webster]

It's a fair assumption that it can't generally march more than about 4 men abreast.

sorry but where did the four men abreast come from?

Width of a cart. Many tracks would probably be narrower than this, obliging the army (presuming it keeps to the track) to march 2 abreast.

Which is why it becomes more possible to get anywhere with an army of 25,000 and an army of several million. I seem to remember that was why Patrick came up with the swarm of locust movement technique

[Jim Webster]

On the logistical front, an interesting observation by Dr Brian Dobson on p.199 of Warfare in the Ancient World (ed. General Sir John Hackett):

"Polybius gives rations which work out to 1.6 kg per horse, and a papyrus from Oxyrhyncus attests to a similar weight in the 6th century. English cavalry horses, at an average height of 15 1/2 hands, and therefore much heavier than Roman, at about 14 1/2 - 15, received 5.4 to 6.3 kg of oats or equivalent (e.g. barley)a day, with about 5.4 kg of hay and some greenstuff when available. The protein content of grains in the ancient world was much higher than it is today."

This would transform logistical calculations involving mounted troops.  It also means that a diet of bread for the soldiery was more balanced than we would suspect.  These are good reasons to support a general principle of not retroimposing 20th century figures onto the 5th century BC.

But they're not a reason to support the discoveries of two thousand years.
A bread diet will not protect you against scurvey
Similarly please don't assume that others know as little about agriculture as you do. When I quoted figures of feeding horses and mules I did take the largest most modern mules, because they were the only ones that came near to being able to carry what Justin needed for his calculations
But for horse feeding I took a smaller horse. Also note I did my calculations on a man eating 3lb (not 1.5) and a horse eating twice that, so I was going on 2.7kg
I took 3lb for the man (as recommended by American civil war armies and also Rome) because 1.5lb is less that the bread ration in the gulags, and the bread ration in the gulags formed only half the prisoners daily calories (calculated at 1300)

So however much you tinker with horse rations, you've find you do not come a long way adrift of the huge figures needed for grain that I came up with
However if you accept that a state cannot put upwards of 40% of men of military age outside the state in an invading force and come down to something reasonable, you don't have to postulate superior ancient baking techniques, grain varieties, and a human ability to to metabolise their own vitamin C from sun light

[Jim Webster]

  One may recall that Persians were taught to do three things from boyhood: ride a horse, shoot a bow and tell the truth.[/i]

Alas that our history was written by a Greek, had a Persian written it we might have got the correct numbers 

As it happens, we are in luck.  We have a Greek writing down the numbers he was given by a Persian. 

so because he wasn't a persian and taught to tell the truth he doesn't actually have the write down the same numbers as he was told.

But actually we're not saying Herodotus lied (Although Plutarch put it pretty much that bluntly), we're saying he got figures he didn't understand and quoted them in the wrong context.

[Flaminpig0]

Actually there are several impediments, beginning with the fact that Herodotus is not alone.  If he had managed to get his figures wrong, we can be pretty sure that Thucydides, who delighted in proving Herodotus wrong on small points of detail, would have dropped some very heavy hints about the matter.

why would any Greek bother contradicting him? This was now their story.

But was it?  Herodotus recorded it ''in the hope of thereby preserving from decay the remembrance of what men have done, and of preventing the great and wonderful actions of the Greeks and Barbarians from losing their due meed of glory," which suggests it was not 'their story' but rather one whose details needed to be set down before memories faded or were lost.  And some Greeks were happy to contradict Herodotus in other matters purely to show what a bad historian he was (or rather, how good they were by comparison). They would not have missed an opportunity to trip him up over this.

We might also turn this argument on its head and say: our own culture produces writers who consistently under-estimate and under-state figures from the Biblical and classical periods. 

Can you give examples of this?

Practically every book or article written about every battle involving forces of more than 20,000 men per side.  Two egregious examples are Delbruck on Alexander's battles against the the Achaemenids and Robin Lane Fox following him.  Even Cannae is not immune, as can be seen in the Wikipedia entry.

"However, some have suggested that the destruction of an army of 90,000 troops would be impossible. They argue that Rome probably had 48,000 troops and 6,000 cavalry against Hannibal's 35,000 troops and 10,000 cavalry." - the reference cited is the Cambridge Ancient History volume VIII of 1965.

It is a pervasive disease, and not one which contributes to understanding.

So there are two schools of though within the 'modern' study of ancient history- Patrick and those who are part of a disease which hobbles human knowledge. There are good reasons to doubt  ancient/biblical sources;for example the Hebrew numbers in Exodus which you use as an argument  for the size of Xerxes forces is predicated on supernatural intervention, and is taken from a source that doesn't seem to be contemporary to the events described.

[Erpingham]

So, you see no impediment to a smaller but still very large force other than it involves assuming that Herodotus got his figures wrong?

Actually there are several impediments, beginning with the fact that Herodotus is not alone.  If he had managed to get his figures wrong, we can be pretty sure that Thucydides, who delighted in proving Herodotus wrong on small points of detail, would have dropped some very heavy hints about the matter.

So we have consistency (at least in terms of Xerxes' army being really large) among our original sources.  They would not all have been hypnotised by Herodotus' account because quite a few classical historians delighted in disparaging Herodotus for one reason or another.

You did read the argument further on, didn't you?  The one about a cultural artefact of the "barbarian horde" and how it affected the reporting of numbers?

Conventional history states that the figures for " the other" are consistently over estimated.  That Herodotus, having been told some tales about the size of the Persian army, believes them, because culturally he is conditioned to think of vast barbarian hordes.  I don't think he fabricated his figures - I think he accepted some figures he was fed, because they fitted his pre-conceptions, then worked out from there, filling in gaps in a consistent manner. In this latter position, I confess I diverge from a lot of historians who do think he just made the numbers up.

Although if he did this, we get the questions 1) why would he have any such preconceptions int he first place, 2) who fed him these numbers, and why would he believe them?* and 3) how is it that he still has minor inconsistencies in his arithmetic?  If he were 'filling gaps in a consistent manner', why the inconsistencies (notably with regard to the number of smaller ships and auxiliary vessels)?

1. Because he came from a culture with these preconceptions
2. I believe you said he spoke to Persian officials?  He's believe them because they fed his cultural expectation.
3. You yourself have excused Herodotus' arithmetic errors as not undermining his narrative.  He gets things in the right ball-park throughout but, when it comes to creating the details of lists, he makes arithmetic errors (like in the breakdown of ships - we can see easily how he creates an itemised list to back up the traditional figure of 1207 but slips in 50 ships he didn't need)

Later writers produce figures which are consistent because they are from the same cultural outlook.

But some writers from the same cultural outlook, notably Thucydides and Plutarch, get their knives into Herodotus at every opportunity. 

But do they produce smaller numbers for the size of barbarian armies?  You've said they don't earlier. 

Furthermore, a cultural outlook which compulsively over-estimates and/or overstates enemy numbers is not going to fare as well as did the Greeks in combat against the Persian Empire.

I actually don't get what you mean here.

We might also turn this argument on its head and say: our own culture produces writers who consistently under-estimate and under-state figures from the Biblical and classical periods.  I think we are on much firmer ground with this assertion.

If we had known figures, we could test them against modern estimates.  But all we have are the numbers in the Bible and classical sources, which is what we are trying to test against our knowledge of the wider world.  So I doubt you could prove your assertion.  Where I do think you have some point is something I said right at the beginning of this discussion - some historians automatically downsize ancient figures as a kind of reflex, without critical thought and that is a failure to engage with cultural preconceptions.

[Erpingham]

I was unable to find the source of his view about a "table of organisation" of the Persian army at 300,000 infantry and 50,000 cavalry with equivalent numbers of hangers on.

Rereading this, its because he takes the list of officers in Herodotus and the number of men that rank would normally command and says this is the normal "table of organisation". 

The inconsistency of the ranks and numbers does turn up in several rejections of Herodotus' numbers but I don't know if anyone else using them in the same way as Stecchini.

[Duncan Head]

The inconsistency of the ranks and numbers does turn up in several rejections of Herodotus' numbers but I don't know if anyone else using them in the same way as Stecchini.

Barkworth points out that the next rank down from the six archontes, commanding generals, of the Persian army appears to be the myriarchs, commanders of 10,000. He therefore thinks that each of the 29 national infantry contingents, ethnea as Herodotos calls them, was a myriad; there is no intervening rank between the myriarchs and the archons, but each contingent has a named commander, therefore the named contingent-commanders are the myriarchs. This would give us 290,000 infantry (plus the Immortals and other guards) rather than the 1,700,000 which Herodotos was apparently told. He's not the first or the only scholar to opt for an infantry strength of 300,000 or so. This would appear to be the same as Stecchini's "normal table of organization" which he thinks Xerxes doubled in 480.

[Erpingham]

On the logistical front, an interesting observation by Dr Brian Dobson on p.199 of Warfare in the Ancient World (ed. General Sir John Hackett):

"Polybius gives rations which work out to 1.6 kg per horse, and a papyrus from Oxyrhyncus attests to a similar weight in the 6th century. English cavalry horses, at an average height of 15 1/2 hands, and therefore much heavier than Roman, at about 14 1/2 - 15, received 5.4 to 6.3 kg of oats or equivalent (e.g. barley)a day, with about 5.4 kg of hay and some greenstuff when available. The protein content of grains in the ancient world was much higher than it is today."

This would transform logistical calculations involving mounted troops.  It also means that a diet of bread for the soldiery was more balanced than we would suspect.  These are good reasons to support a general principle of not retroimposing 20th century figures onto the 5th century BC.

I think you may have misunderstood this a bit.  The 1.6 kg is not the total feed of the horse, it's the grain feed in addition to the fodder and/or grazing.  If you look at the figures Haldon uses for the Byzantines, referred to earlier, he reckons a cavalry ration of 2.2 kg grain and 6.8 kg fodder per horse per day.  All the figures on the internet I've seen on horse feeding reckon a horse eats 1-3% of its body weight a day, towards the higher end if working.  We reckoned that the rule of thumb that Persian cavalry horses weighed 500kg was a bit high - say 400kg - so that means a minimum ration of 4kg up to 12 kg.  I think if you went for the low ration, the horses would break down before they get to Greece, because it is for sedentary animals.  Incidentally, for consistency, you can't use load weighting from large modern mules if you are breeding from a cavalry horse the size of a large pony.

Addendum : I did a bit more research on the figures. 
Richard Duncan Jones Structure and Scale in the Roman Economy p107 gives these figures in detail.  The cavalry were issued 1/10 arbata (4 choinices) of barley a day for their horses.  That's 3.2 litres - about 2 kg (weight conversions for barley volume vary a bit).  He confirms the papyrus states there is a hay ration too.  Ptolemaic cavalry rations are given at 3 choinices a day.  Polybian cavalry received 8.6 litres if allied, 12.1 litres if Roman.  The figures do not match at all.  Southern and Dixon The Roman Cavalry p.210 explain this by each cavalry man having three horses and an attendant, so each horse 1.5 kg a day, which roughly matches the Byzantine figure (presumably, allied cavalry had one less horse or no attendant).  I guess the discrepancy is caused by the conversion factors for the weights and measures - Polybios is working in medimnoi a month, the Byzantines in arbata per day. 

[Justin Swanton]

When I quoted figures of feeding horses and mules I did take the largest most modern mules, because they were the only ones that came near to being able to carry what Justin needed for his calculations

Actually I based my calculations on the legal limit for a late Roman military pack mule which was well below the theoretical limit the animal of that period could stand.

[Jim Webster]

When I quoted figures of feeding horses and mules I did take the largest most modern mules, because they were the only ones that came near to being able to carry what Justin needed for his calculations

Actually I based my calculations on the legal limit for a late Roman military pack mule which was well below the theoretical limit the animal of that period could stand.

This means that the late Roman military pack mule was a better animal than the modern Indian and American army mules, even through it was bred from smaller horses. I honestly doubt it

[Flaminpig0]

Cannot help but feel that the only evidence that would persuade Patrick and possibly Justin is if a revised version of  Herodotus work was found giving a different account as to the size of the Persian army.

[Justin Swanton]

When I quoted figures of feeding horses and mules I did take the largest most modern mules, because they were the only ones that came near to being able to carry what Justin needed for his calculations

Actually I based my calculations on the legal limit for a late Roman military pack mule which was well below the theoretical limit the animal of that period could stand.

This means that the late Roman military pack mule was a better animal than the modern Indian and American army mules, even through it was bred from smaller horses. I honestly doubt it

See post #768. I took Anthony's figure of 65.5kg stipulated for a Roman army pack mule and went with that.

[Jim Webster]

When I quoted figures of feeding horses and mules I did take the largest most modern mules, because they were the only ones that came near to being able to carry what Justin needed for his calculations

Actually I based my calculations on the legal limit for a late Roman military pack mule which was well below the theoretical limit the animal of that period could stand.

This means that the late Roman military pack mule was a better animal than the modern Indian and American army mules, even through it was bred from smaller horses. I honestly doubt it

See post #768. I took Anthony's figure of 65.5kg stipulated for a Roman army pack mule and went with that.

i was remembering back to about 711/13, I perhaps hadn't noticed the bit with Anthony sorry

[Erpingham]

Just for fun, here is the US special forces mule packing guide (really) on loads

  good  load  for  most  horses  or  mules  (1,100-  to  1,200-pound  packhorses  and  800-  to  1,000-pound mules) is 160 to 170 pounds. However, if there are some small animals in  the  pack  string,  this  amount  could  be  too  much  for  them.  In  extreme circumstances,  the  packer  can  load  an  animal  with  up  to  250  pounds;  however,  this  amount  would  limit  his  speed  and  endurance. 

So, about 75 kg for the US army as normal, without major impact on speed and endurance.  The pakistani army, according to wikipedia, load up to 72 kg.  All not that far above Roman standards, with probably bigger animals.