The Ptolemaic army in Upper Egypt (2nd-1st centuries BC)

[davidb]

Thought people might be interested in this article:

https://www.researchgate.net/publication/265964759_The_Ptolemaic_army_in_Upper_Egypt_2nd-1st_centuries_BC

David

[Patrick Waterson]

Thanks, David: the organisation of the semaia/syntagma is particularly interesting.

Scheme of the semeion/syntagma (256 men)
— Commander: hegemon
— Officers, among them
– Kerux or Stratokerux or herald (of the army)
– Semeiophoros or standard bearer
— Unit of 100 men, commanded by a hundredman (hekatontarch)
Unit is called in 3rd-cent. B.C. = taxis, 2nd-cent. B.C. = tagma?
< two units of 50 men, each commanded by a pentekontarch
Officers: ? Kerux of the tagma and ? Hyperetes of the tagma
— Unit of 100 men, commanded by a hundredman (hekatontarch)
Unit is called in 3rd-cent. B.C. = taxis, 2nd-cent. B.C. = tagma?
< two units of 50 men, each commanded by a pentekontarch
Officers: ? Kerux of the tagma and ? Hyperetes of the tagma
— Rear guard of 50 men, commanded by the ouragos

Assuming this has been correctly understood and stated, we have five subunits each of 50 men and six additional officers.  Four of these are grouped into two 100-man units each under a 'hekatontarch' or commander of 100 men.  In reserve is a half-strength unit.

Remind you of anything?  Translate hekatontarch into Latin and you get 'centurion'.  Interestingly, we have the potential for two full-strength and one half-strength infantry lines, reminiscent of the Roman hastati-principes-triarii system.  The subunit strength is right for fighting ten deep (or five deep for the rear line, like triarii) but looks wrong for a pike-armed Hellenistic unit.

Could this be indirect evidence for the Romanisation of the Ptolemaic army?

[willb]

The timing would be about the correct period.   The defeat of the Seleucid army at Magnesia occurred at the beginning of the second century.  Rome would likely have created diplomatic ties with the Ptolemies to counter the Seleucids.   The Daphne parade with Romanized infantry was in the first half of the second centsury.

[RichT]

That is Nick Sekunda's theory - Seleucid and Ptolemaic Reformed Armies - short review here: http://discovery.ucl.ac.uk/85714/1/download15.pdf

[Patrick Waterson]

Thanks, Richard: I hope Hans van Wees gets over his sense of disappointment about the missing Gallic equipment!

One can see various significant clues amongst the generally ambiguous information, and what stands out more than the likely adoption of an 'imitation legion' is the unsuitability of the arrangement for traditional pike units and tactics, which pretty much tells us that a different system had been adopted.  The trick is to work out exactly what that system was.

The fact that the syntagma remains at 256 men without integral light infantry suggests an adaptation of the heavy infantry to a Roman-esque model but not a carbon copy of the legion.  It looks like a compromise between acquiring the important factors of the Roman system and fitting them into an existing Hellenistic organisation and, to an extent, command structure.

The lack of light infantry would not be a problem, as the Ptolemies would continue using their traditional light infantry types for skirmishing and would pull them back before the main battle lines met.  The apparent lack of cohorts may be more apparent than real, as a syntagma could act in much the same way as a small cohort and pairing two together would give an effective large cohort for tactical battlefield use.

The question this spawns is: was this approach a one-off, or would it represent the usual pattern for imitation legions such as those of the Selucids and Armenians, or those of Pontus?  Galatia seems to have followed a different course, with its own legions being directly Roman-trained and its surviving legion being transferred unchanged to the Roman army when Galatia was incorporated as a province.

And this brings us to a final speculation, which if accurate would be wonderful in its irony.  Did the Hellenistic imitation legion, which seems to lack distinction between hastati, principes and triarii, become the inspiration for the structure of the so-called Marian legion, which abolished these classifications within the Roman army?

[Swampster]

The Pontic imitation legionaries may well have been effectively Roman trained, e.g by Sertorian advisers sent. Unfortunately the details are lost - Dio seems to have referred in more detail to the rearming in the Roman manner in a now lost section.   

Sekunda goes into more detail in https://www.amazon.com/Hellenistic-Infantry-Studies-History-Medieval/dp/8389786834/ref=asap_bc?ie=UTF8 though many of the issues raised in the van Wees review still stand.

[Patrick Waterson]

One of the reviewers of Sekunda's book is definitely a confirmed thureophorophile, perhaps too much so considering how the type gave way to the phalangite as time progressed, notably among the Achaeans.  True, there is paucity of unequivocal evidence for Ptolemaic imitation legions, but trying to extrapolate a return to thureophoroi is illogical (why adopt a troop type inferior to both legion and phalanx?) and I anyway cannot see how this would square with the reorganised syntagma with its two-and-a-half 'centuries'.

Good point about Sertorius' officers being the likely trainers of Pontic imitation legionaries; I missed that.

[Andreas Johansson]

True, there is paucity of unequivocal evidence for Ptolemaic imitation legions, but trying to extrapolate a return to thureophoroi is illogical (why adopt a troop type inferior to both legion and phalanx?) and I anyway cannot see how this would square with the reorganised syntagma with its two-and-a-half 'centuries'.

Contemporaries must've seen some advantage to the thureophoroi - else the troop type wouldn't've been adopted in the 3rd century, well after Philip and Alexander had shown what phalangites could do. Presumably said advantage wasn't in pitched battle - more suitable to city fights perhaps, or easier to train due to less need for drill. One of the Amazon reviewers suggest they're well suited as city garrisons, which seems sensible.

But identifying the depictions in question as being of thureophoroi hardly commits one to thinking thureophoroi replaced phalangites. They might easily have been a constant element of Ptolemaic armies since well before the Romans showed up, we've hardly got the evidence to know how their numerical importance varied over time, and wargamer received wisdom would have it they originally replaced peltasts.

[Patrick Waterson]

Indeed, though a syntagma subdivided into fifties rather then sixty-fours would presumably be as odd for thureophoroi as it would for phalangites.

[Swampster]

I doubt the 'fifty' and 'hundred' are any more accurate than in the Roman usage.
Sekunda works on the 50 probably being an ideal of 32 and he gives some figures for very understrength units - one is only 7 men.

FWIW, he thinks there are 6 supposedly equal sized sub-units, based on the largest definite number given to a unit is the 6th. I think finding out Van 't Dack's theory would be necessary to weigh up the relative merits of their arguments.

One of Sekunda's prime arguments is that something has to have changed because of the unit becoming a semeion and that previous Hellenistic units didn't have standards. A doubt is raised if the Pergamene belt buckle does show a phalanx with banner.

[Jim Webster]

Just to go off on another track I was intrigued by the use of pseudo-ethnicity with native Egyptians being recruited into units of Persians and Macedonians.
Also it seems that men could progress through Persians, join the Macedonians and perhaps even advance to becoming cavalry.
Given that the Persians could be 'laid off' or put on the 'reserve' where they weren't paid it looks as if the ethnicity was about status and not equipment with Persians merely being second line troops as opposed to having different training and equipment

So at this point it might be worth considering the use of ethnicity elsewhere in the Ptolemaic army as well as potentially other armies? 

[Patrick Waterson]

I doubt the 'fifty' and 'hundred' are any more accurate than in the Roman usage.

Roman usage may be more accurate than is generally thought.  Using Polybius' assertion that when a legion increased above the 4,200 mark the hastati and principes were strengthened but other contingents remained the same, we get (omitting cavalry):

4,200-man legion
1,200 hastati + 400 velites (maniple strength 160; century strength 80)
1,200 principes + 400 velites (ditto)
600 triarii + 400 velites (maniple strength 100; century strength 50)

5,000-man legion
(adds 800 men, divided equally between hastati and principes)
1,600 hastati + 400 velites (maniple strength 200, century strength 100 - voila!)
1,600 principes + 400 velites (ditto)
600 triarii + 400 velites (maniple strength 100; century strength 50)

Following the Punic Wars, the 5,000-man legion seems to have become the norm, although we get occasional legions of 6,000 or so.  This would make the standard or at least modal 'century' the 100-man version, demonstrating that even Romans can count.

Curiously, if one halves a cohort from such a legion one gets exactly the 100/100/50 lineup of the misthophoroi semaia.

Regarding this revised syntagma (or semaia), I would read 50 men as 50 men, because the syntagma (or semaia) is still 256 strong of whom six are officers, so if it has five subunits then, by simple arithmetic, if of identical size they each have to be 50 strong (whatever Sekunda may think).  It is the departure from base 8 in favour of base 10 in the revised organisation which is so striking and to my mind the strongest indication of imitation legionarity.

Just to go off on another track I was intrigued by the use of pseudo-ethnicity with native Egyptians being recruited into units of Persians and Macedonians.

Probably quite off track, one of the degrees in Mithraic initiation was the Persian.

[Swampster]

While there were times when a century may well have been 100 men (and even then, this is the exceptional sized legion in your example), later practice was for a century to be perhaps 80 men.
We also have Asclepiodotus saying that a taxiarchos (in theoretical command of 128 men) has become known as a hekatontarch, though while saying the name has changed, he does not say that the theoretical unit strength has. See also the pentekontarchia below.
As for the 5 x 50 plus 6 officers, this model doesn't give 6 officers. 1 x hegemon, 2 x hekatontarchs, 4 x pentakontarchs, 1 x ouragos. This ignores the herald, standard bearer etc. To get 6, you could drop the ouragos as not being a unit commander but then you have 4 units, not 5. There is a good chance that these, and the herald etc. were supernumary as Asclepiodotus says is the case when the formation isn't a square (though that gets decidedly confused).

It does seem that the hekatontarch disappears very late in the period (being absent on the Jouget-Roeder stelai) which would give 6 officers but then diverges from the comparison with the Romans with the hekatontarch being the centurion. The change may be an acceptance that the units were so understrength that so many officers were unnecessary - the number of heralds is also droppe.


Looking back at Sekunda, I was mistaken about the 6 - he says six of the semeia make up a larger unit.
He divides the semeia into 4 - 2 pairs of '50's. He does not assign a unit to the ouragos. Why Van 't Dack does, I don't know. It is possible that the stele printed in Sekunda has the ouragos with a unit which is drastically understrength and only has a couple of members, but then some of the '50s' have only 5 men including the officer. Certainly there is no attempt to keep the 50s equal in size - if anything there is one large one and one or more small ones, perhaps acting as cadres.
His arguments for the size of 32 are rather involved but are partly based on the largest parade strength of any of the '50s' being 32. He then mentions the 'pentekontarchia' in the Taktika which has 64 men. There is some discussion as to why he thinks 256 would be unlikely. He follows this with an assumption that things are based on a tent party - he works off a 6 man tent, mentioning this as Roman practice though I've only seen the 8 man contuburnium.

I'm not seeking to support Sekunda as being a better interpretation, just that the clear cut unit structure shown in the other paper really isn't that clear cut. Even the unit names aren't secure.

[Patrick Waterson]

I'm not seeking to support Sekunda as being a better interpretation, just that the clear cut unit structure shown in the other paper really isn't that clear cut. Even the unit names aren't secure.

I shall accept that point as being well made.

Less sure about officers: if they were imitating Roman practice, the hekatontarchs would double as pentakontarchs, the herald would just be a herald and like the semeiophorus would not be an officer, leaving 1  hegemon, 2  hekatontarchs doubling as pentakontarchs, 2 'pure' pentakontarchs and the ouragos.

And I still wonder if the 2 x 100-man and 1 x 50-man pattern is indicative.  As it seems to depend upon Van t'Dack's interpretation more than anything else, I am happy to leave it hanging for now.

While there were times when a century may well have been 100 men (and even then, this is the exceptional sized legion in your example), later practice was for a century to be perhaps 80 men.

5,000 is hardly an exceptional size for a legion in the 3rd-2nd century BC, surely?  It appears if anything to have been the norm.  The later 80-man century incidentally follows the Marian reforms and would not be representative of the period in question.

[Swampster]

I'd wondered whether the hekatontarchs had their own 50 but it seems not. The stele he translates gives the name of a hekatontarch followed by a penta. which then has the names of the men in that 50. The hekatontarch doesn't have any unranked men directly below him.
Sekunda's model is to have two 50s, each with a pentakontarch and above them a hekatontarch. Two of these 100s combine under a hegemon (and sometimes a joint command).

This model certainly isn't secure though - the stele has some damage, there are men listed who seem to be supernummary e.g. those who are listed as 'sons of officers', the units seems to be very understrength so when an officer is absent from the above model is that because he was actually absent/ill/dead or is it the model that is wrong?

Vandorpe's paper follows Van 't Dack's but I haven't read this. She seems to make some assumptions which I would question - "Among  the  ofcers  (ἔκτακτοι)  of  the  semeion, are the kerux-stra-tokerux  (στρατοκῆρυξ) 70 or herald and the semeiophoros or standard bearer (σημειοφόρος, Demotic f3y-stn) 71. In the Hermopolite inscription SB I 599, they are mentioned immediately after the commanders-hegemones of the semeion and before the hundredmen, and are thus the second most important officers of the unit. In a letter from Pathyris, the standard-bearer is also addressed shortly after the commanders of the unit. In a text from Koptos, the standard-bearer appears to be the son of the commander of the semeion. "
I think their position in the list doesn't help with their seniority - they may well be there simply because they are the HQ section. Their mention in the letters may be a stronger point though.