Late Roman Army Grade/Rank List under Anastasius

[Condottiero Magno]

Thanks for the heads up Longovicium. I hope to have an electronic copy in the next couple of days.

Granted my Turkish isn't up to par, but I didn't notice an electronic version, just a physical copy at 96 TL(~USD$28) and postage free, since it's above 80 TL.

[Steven James]

Mango wrote:
I didn't notice an electronic version, just a physical copy at 96 TL(~USD$28.
 
What I have been told by the university, is when a request is made for a paper that is not electronic, there is a company in the UK I believe, that scans the article and then adds it to its electronic data base. My contacts at the university have been getting me papers for years via this company.  

The problem at the moment is my friends are on holidays for 2 weeks, so I will have to wait.

[Nathan Ross]

the published article is now available in Gephyra 14, 2016

Thanks Francis!

Certainly a very thorough paper, and it's good to have the full texts and translations. I don't think it necessarily adds much to what we already know about Slab C (notitia) in particular - the numbers do not appear to have been revised since the previous publication.

It is interesting to see Professor Onur's notes on the text, however, and especially on the military ranks/grades. He does not attempt a reconstruction of the legion's full composition or structure, which is probably sensible, but a shame nonetheless.

There were a few points in the paper that I would query, or which seem unclear or perhaps suggest differing interpretations to the ones we've discussed so far on this thread. If anyone could offer further comment or clarification I'd be most grateful!


p.159: "The 10 signiferi given in the list in the Perge inscription were each in charge of 10 centuriae of the unit." - Can this be right? It would give the full legion 100 centuriae, which would be difficult with only 20 ordinarii to (presumably) lead them.

p.160 (about Veredarii): "The title is given to two groups in the inscription of Perge, the first being of 50 men, while the second was of 225 men, who were probably candidates for the first group." - Onur agrees that these men were cavalry, but offers no clues as to the odd numbering. If we accept that the 10 vexillarii suggest ten subunits of horsemen, then the only solution I can find is the one I suggested above somewhere: 9 units of 25 veredarii alii and one double unit of 50 veredarii. It's not a very comfortable solution, but I can't see any other way to make the numbers fit, unless we go with Marcel's idea of adding some of the Augustales and Flaviales to the cavalry complement.

p.163: "Unfortunately the numbers related to clerici et deputati are also lost. However, the number ?73 on a fragment... might have belonged to them" - Onur agrees that the clerici were military priests, which seems right. I notice that the old reading of the text has clerici ve deputati, whereas this one is clerici et deputati: 'and' rather than 'or' (I think). Even so, 73+ priests and 'assistants' (or whatever deputati means) seems a very high number for a relatively small legion!

Could it be possible that the "?73" fragment fits elsewhere, or belongs to a different inscription? The last lines of Slab C are very broken up - I'm not sure of the reading for the munifices number either. Can anyone make anything else out here?

   

p.186: "The unit in Perge edict is a legio, either comitatensis or perhaps palatina." - Onur discounts the possibiliy of this being a limitanei formation, although I don't see why this would not be plausible: the magister militum surely having overall control of all military units, not just the field armies?

p.187: "So these numbers actually represent a unit larger than 1200 men" - leaving out the clerici etc, it still appears that the minimum strength of the legion must be 1272 men (with 159 munifices); this could probably be rounded to 1280 with a nominal 8 clerici and/or deputati. This figure would step up at increments of 100 depending on how many munifices there were. However, the ?73 is troubling - if the proposed positioning of the fragment is correct, it could increase the overall size by an unknown margin!

[Steven James]

Well after reading the Perge paper, I am a little perplexed. Sometimes these payroll records produce more problems than solutions.

 
Page 157 “20 ordinarii.”
 
20 ordinarii matches my research ...for the smaller legion.
 
Page 159 “The 10 signiferi given in the list in the Perge inscription were each in charge of 10 centuriae of the unit.”
 
Well if this is true, in relation to my research, it is accurately giving the numbers for a full strength legion, and also giving the number for same legion with one third missing, which is a common practice of the Romans. Also the 1 signiferi and 10 centuries would relate to those under the command of a tribune.

[Nathan Ross]

Meanwhile, like some fiendish crossword, the list of numbers continues to obsess me...

It occurs to me that if the cornicines and bucinators might be infantry hornblowers, and the tubicines belong to the cavalry, their numbers could be added to the breakdown of ordines (I'm using 'ordo' in preference to 'century' or 'turma' here!).

We could do the same with the various other 'double pay' specialists: the praeco (herald) would be best mounted, for example, and gets the same annona as the veredarii alii. The others could be added to the infantry ranks. But we would have to decide whether the Augustales and Flaviales belonged in the infantry or cavalry ordines, or both (probably the latter, I think...)

So for the infantry we might get a breakdown like this:

The Armaturae Duplares and Semissales (20 men each), the Torquati Semissales (136 men) and Bracchiati Semissales (256 men) are all skilled infantry. Total = 432 men
Plus 8 cornicines and 2 bucinators = 442 men
Plus 5 librarii and mensores = 447 men
Plus 4 beneficiarii = 451 men
Plus 100 Augustales Alii (both grades) = 551 men
Plus 140 Flaviales Alii = 691 men
Plus 10 Optiones = 701 men
Plus 10 Signiferi = 711
Plus 10 Imaginifers = 721
Plus 60 Flaviales = 781
Plus 159 (minimum number) munifices = 940 infantry

Divided between 10 infantry ordines = 94 men + 1 Ordinarius = 95 men per infantry ordo.



For the cavalry, like this:

225 Veredarii Alii
50 Veredarii
20 Augustales
10 Vexillarii
4 Tubicines
1 Praeco
Total = 310 cavalry

Divided between 10 cavalry ordines = 31 men + 1 Ordinarius = 32 men per cavalry ordo.


With 10 infantry ordines of 95 men, 10 cavalry ordines of 32 men, and the 2 tribunes, we get a total full-strength legion size of 1272 men, all inclusive, which fits exactly with the numbers given as the minimum possible. To this we must add an unknown number of (supernumerary?) 'clerici et deputati'.

For a larger legion, the number of munifices could increased by the hundred, with 10 extra infantry being added to the ordines each time. So with 259 munifices the infantry ordo would number 105 men and the legion total 1372. And so on.

Plenty of speculation there, of course - and the relative strengths of infantry and cavalry ordines would change depending on where we put the Augustales and Flaviales - but it does appear to make sense of the jumble of numbers!

[Steven James]

Nathan wrote:

So with 259 munifices the infantry ordo would number 105 men and the legion total 1372. And so on.
 
Armenian Passion of St. Callistratus (BHO 185) 13: page 334, But when we saw the light which shot forth over the heads of the saints, and heard the blissful voice along with the earthquake and the breaking of the idols, we believed, - we, the soldiers, 105 of us.
 
It is unfortunate that you do not recognise the martyrdom stories as having any authenticity. This means the baby gets thrown out with the bathwater.
 
Nathan wrote:
Plenty of speculation there, of course - and the relative strengths of infantry and cavalry ordines would change depending on where we put the Augustales and Flaviales - but it does appear to make sense of the jumble of numbers!
 
That is debatable.

[Nathan Ross]

That is debatable.

It is intended to be!

[Steven James]

My impression still remains the same. It appears to be telling us of the legion at full strength, its strength with one third removed (10 signiferi command 10 centuries), and the numbers given seem to relate to the men of the one third of the legion that are removed or act as garrison troops. The two tribunes indicate this is what is going on. Then there is the cavalry aspect to be included.

 
The number of Augustales (120) and Flaviales (200) comes to 320 men, which from my own research has my attention. Having studied the Roman army from the Servian constitution to the sack of Rome, I have found some interesting patterns that have become tools for further investigation. No one has done an intensive study on how the legions make cavalry lanes, and the rules that apply. Basically, it is plain geometry. The frontage of a cavalry lanes created by the legion must be equal to the frontage of a cavalry squadron with squadron intervals, and knowing the frontage allocated to each horse, I can determine the size of a cavalry squadron by the frontage of a cavalry lane. As everything is formulised with the Roman military, it becomes predictable, and it is no different with the Late Roman army, and when the size of the squadron as determined by the cavalry lanes matches the primary source data, it is all systems go.
 
However, putting that aside, it still does not properly answer what is the size of a numerus or cuneus is. How do we know? If a person works out the size of an arithmoi as given by Soz., does that really prove an arithmoi is a numerus? No it doesn’t. And how large is an ala? How large is a cavalry vexillation? In the end I have to admit that I don’t really know, and have made the following based on organisational patterns of the past, which produce the following:
 

Ala                     24 squadrons

Numerus            16 squadrons

Equites              12 squadrons

Cuneus                8 squadrons

Vexillation           6 squadrons

 

The cavalry numbers are based on the Pythagorean cosmos, which uses the ratio 3/2 (the perfect fifth) and 4/3 (the perfect fourth). The cavalry in the Pythagorean system are part of the five elements that make up the cosmos, with the cavalry representing the element Heaven. The ala of 24 squadrons and the numerus of 16 squadrons represent the ratio 3/2. The numerus (16 squadrons) and the Equites produce the ratio 4/3, and the Equites and the Cuneus produce the ratio 3/2, while the cuneus and the vexillation produce the ration 4/3.

In the end, like all pay records, the Perge document does not produces answers to many a question. Pay roll records, I believe are incomplete or like the Perge document, has missing fragments, which make them deceptive and to be applied with caution. Working with pay records is akin to the five blind men trying to describe an elephant. Just look at how divergent are the interpretations of myself and Nathan concerning the Perge document. We are miles apart on nearly every aspect. So for those who believed the Perge document was to be the instrument of great understanding into the Late Roman legion, for me it has too many pitfalls and short comings.

[Nathan Ross]

Pay roll records, I believe are incomplete or like the Perge document, has missing fragments, which make them deceptive and to be applied with caution.

I don't agree. This document is not an 'effective strength' snapshot like the Dura papyri or Vindolanda tablet - it's an official list, carved into stone, telling us exactly how many men were supposed to be in the legion, and intended to stand for some time.

There are only two missing figures in Onur's reading. One is for the 'clerici et deputati', who were probably supernumerary to the legion strength, and the other is the 'hundreds' figure for the munifices, which can be reconstructed within a narrow field.

If we accept that the ten vexillarii are cavalry subunit standard-bearers (which I think we must), then it's interesting how few ways we can reconstruct the legion organisation. Short of coming up with an entirely different and quite radical organisational structure, we're dividing the available men into tens and twenties. And unless we envisage the infantry ordinarii commanding very large numbers of men, the total legion size cannot, I think, rise much above 1600.

My attempt above is one way, perhaps, of calculating things - there are a number of alternatives available, but they only make small differences to the overall picture. We could, for example, move the 140 Flaviales Alii onto the cavalry strength, giving a nicely traditional-looking 'Hyginian' 80-man infantry ordo - but that leaves us with a 45-man cavalry ordo!

Even at the smallest size for this legion, we appear to be looking at either an infantry subunit, or a cavalry one, or both, somewhat larger than we are used to from the Principiate.

Nevertheless, the similarities with earlier practice are quite striking, not only in the rank titles. As mentioned previously, this 'numerus legionum' does look surprisingly similar to a milliarian cohors equitata, like the XX Palmyrenorum from Dura.

There's also something quite 'Vegetian' about the unit breakdown: infantry ordines of c.95-105-115 (depending on number of munifices) and cavalry ordines of c.30-32 are reminiscent of V's comment (2.14) "As 110 infantrymen are controlled by one centurion under one ensign, so 32 cavalrymen are governed by one decurion under one ensign."

The numbers aren't exactly the same - both 'centurions' and 'decurions' are now called ordinarii, and the infantry units have two ensigns not one - but it is quite amazing that a military unit of this late a date still seems to bear the traces of the organisational structures of centuries past!


(EDIT - using the above structure, a figure of 259 munifices gives us an infantry ordo of exactly 100 infantrymen + 1 signifer, 1 imaginifer, 1 optio, 1 musician and 1 ordinarius. The infantry could be divided into 10 10-man contubernia, each led by one of the Augustales Alii as decanus.

The cavalry complement would remain at 30 men + 1 vexillarius and 1 ordinarius.

I realise this is almost certainly too neat to be true!)

[Nathan Ross]

Ala                     24 squadrons

Numerus            16 squadrons

Equites              12 squadrons

Cuneus                8 squadrons

Vexillation           6 squadrons

 

The cavalry numbers are based on the Pythagorean cosmos, which uses the ratio 3/2 (the perfect fifth) and 4/3 (the perfect fourth).

But in what way does this relate to the information in the Perge lists?

A unit of Equites was called a Numerus Equitum or a Vexillatio Equitum - the terms were synonymous. So these are not different-sized units.

A cuneus may be different again, but the only clear distinction is between the old-style ala (still known in the limitanei) and the new-style vexillatio (aka numerus equitum). Lydus, as you know, gives an ala 600 men and a vexillation 500 men (de mag 1.46) - which does not help us much!

[Julian de Vries]

Perhaps of some help

Constitution of Anastasius I on Libyan administration 501 AD (Seg 9, 356 and 414)

5. Present or subsequent accountants or chiefs of the office staff shall not be permitted to levy anything in the name of document fees from  the five detachments (numeri) of soldiers.

7. In regard to the inspection of assembled troops: the first names on the roll of each detachment or camp shall not be subject to review as ill or unfit of service, that is, the first five, if there are 100 names, or the first ten, if there are 200. The same proportion shall be maintained, whether there are more or less.

8. The honorable duke or those persons who must send messengers for such matters shall select by lot from all the detachments, and not from any one detachment, fifteen soldiers for carrying letters and dispatches, five for doorkeepers, and seven for guards at the public prison. They shall be provided from those persons in each detachment who are permitted to be of the rank of orderlies and not from those persons who are necessary for active service in battle, that is, from the group of camp patrols and those soldiers ranking next below them. All others shall continue in camp duties.

Source:

Ancient Roman Statutes: A Translation with Introduction, Commentary ... Allan Chester Johnson,Paul Robinson Coleman-Norton,Frank Card Bourne:

[Marcel Frederik Schwarze]

This document is not an 'effective strength' snapshot like the Dura papyri or Vindolanda tablet - it's an official list, carved into stone, telling us exactly how many men were supposed to be in the legion, and intended to stand for some time.

but the slab only tells us about the number of soldiers in this so called Legion. There is no reference at all to a widespread validity here. On the contrary, there seem to have been such local abuses that Anastasius felt compelled to set up the slabs or tables at this point and in this city. Whether such troop-lists have been set up at other locations as well, has to remain speculation so far. And if so, then perhaps with other figures.
Just few years later, when Justinian raised the throne, new regiments were just numbering round about 500 men.

I cannot deny the possibility that a normal legion had a small corps of cavalry in the late 4th, 5th or 6th centuries, let's say messengers, officials and some guards. But a mounted group of such a high quota compared to the infatry has nothing to so with a legion in the late antiquity and transition periode: as already described roughly 1000 : 300.
Also the gap of annona given to the higher posts compared to the lower posts is not big enough to speak about a classical legion. We know that the gap of allowances in auxilary units between soldiers and centurions was indeed much smaller compared to the legions.

The best essay to understand the very early numeri and its evaluation is still that of Marcus Reuter (also Nemeth is good, but Reuter is more extensive).
XXX see Studien zu den numeri des römischen Heeres in der mittleren Kaiserzeit. In: Berichte der Römisch-Germanischen Kommission 80, 1999, p. 359-569 

Most of the allegedly barbarian numeri in the 3rd century were auxiliarian vexillations (cohortes et alae). That means they were paired with other auxillaries and sent away from the parent unit. Just a small number of numeri was of real barbarian origin - the so called ethnic or national numeri. But even the latter group was according all studies quickly absorbed or integrated into the roman system and many of those numeri were indeed mixed with infantry and cavalry.
The cavalry of the numeri at Kapersburg and Nidensium were called "veredarii". The cavalry of legions and cohortes equitata had other names.

Then a last word to the deputatii.
a) they were most likely field- or combat medics, as described in the Srategicon (by Mauricius) and Tacticon (of Leo VI.). They were dressed in a light way, sometimes with horses and stirups for two people.
b) also factory specialists (e.g. smiths) were called like that in the same time frame. Specialists who were sent from the weapon factories to the troops - to explain how to use machines or to describe new developed technics when making new weapons etc. also called deputatii.
And indeed, by reading some tales of Procopius and others one get the impression that those factory-specialists were travelling on company business - and there was a high traffic of technical staff between all the roman troops and the factories.
But since our deputatii are grouped together with Military chaplains, and the list is made for the constantly present staff, it makes sense that they were deputies of the first group, and acting as combat medics as well.
In both cases, as fabricienses or medics - they were unarmed and for sure not integrated into the active compat groups of the centuriae or turmae. (as others as well, e.g. librari or the other handcraftsmen). Also the Armaturae Duplares and Semissales are probaly forming a schola of instructors. If they stood in a row with the other fighters is debatable. The Optiones are writers and belong to the administrative staff of the numerus like the actuarius or chartularius in other regiments.

[Nathan Ross]

this so called Legion.

Yes, I was aware of the difficulty of identifying this unnamed unit! Referring to it always in quotes as a "legion" seemed a little unecessary though. It was, more correctly, a numerus (or arithmos) - although, as you say, the text of Slab B on three occasions refers to a legion (λεγεoσιν / λεγεόνων / λεγεoναςas) as well. Quite possibly, as we discussed before, these terms could have become interchangeable by this late date.

However, I still think that the permanency and monumental nature of this inscription indicates that we are looking at a single military unit of a set establishment, both in structure and location, at a known historical moment; that makes it uniquely valuable, although no less challenging to interpret!

a mounted group of such a high quota compared to the infatry has nothing to so with a legion in the late antiquity and transition periode: as already described roughly 1000 : 300.

Quite so. It's the 'cavalry component' that causes the problems here, I think. Were it not for the veredarii we could quite feasibly reconstruct an infantry-only numerus legionum of around 1000-1200 men, and happily believe that we were looking at a surviving example of a 'new style' legion of the Constantinian era, perhaps.

But, as you say, we don't know how representative this particular unit might have been. It could be, perhaps, that military units in Pamphylia and Isauria during this period operated extensively in a counter-insurgency (or 'anti bandit') role, and therefore may have had additional mounted sections of light cavalry 'hunters' attached to aid them in this work.

the Armaturae Duplares and Semissales are probaly forming a schola of handcraftsmen. If they stood in a row with the other fighters is debatable. The Optiones are writers and belong to the administrative staff of the numerus like the actuarius or chartularius in other regiments.

I'm not sure about this. The Armaturae I would guess to be men skilled in the armatura fighting drill, rather than craftsmen - so certainly part of the ordines. The optiones and others (mensores, librarii, beneficiarii) would, on the model of the earlier legion, also be included on the strength of the century (or 'ordo'). Onur gives an example in his paper of a 6th-century Optio fighting in battle. So I suspect that all of these men were included in the fighting ranks, although they may also have had other tasks of an administrative nature.

[Marcel Frederik Schwarze]

as said before, members of the unit are named as legionaries at slabs A and B, but we have also examples from vexillations of the 3rd century which were called a "legion" - and in several cases it is clear that parts of the vexillation weren't true legionaries. Also several numeri-camps from the german limes showing that sometimes legionary detachments were attached to those camps, supplying officers or to add a detachment of specialists. This is important to understand why some camps were bigger and the number of barracks is not always representing or matching the number of the original regiment.
Esp. the limes in germany was very mixed by troops, also the Saalburg here in front of my house was not just a place for a cohors. Some fibulae findings are indicating the presence of a legionary detachment which was garrisoned here from time to time.

Interesting enough that at slab C - the so called notitia - refers to a "numerus" only. Here the unit is called by its official technical term and named as that what is was - a number, a regiment, a flag.
So
a) it could be that the soldiers were called legionaries in a kind of colloquial speech, since the slabs A and B just try to prevent future irregularities in promotion and allowances, from which the latter one (the annona) shows little similarities with the old legions
Or
b) it was indeed a numerus legionum, in this case the second term can just be considered as a remaining part of old troop-names, independently if the unit still kept a certain drill, better than other units (pure speculation).

It is also true that just 20 years later, when Justinan came to the throne, all new deployed numeri (at least we now that they were freshly recuited) come along with numbers between 400 and 500. Cavalry and infantry strictly divided as known from the 4th century. There is simply no dissent here, between what we have learned about the 4th, 5th and 6th centuries.

So our unit is indeed a remnant part of an old army (including cavalry like numeri and cohortes equitata) or it was a quite new unit, especially deployed to guard the difficult area of Pamphylia.
All evidence speaks about a unit which left a long way behind.

Regardless of how it actually was, the unit cannot - at least to my opinion and reseach - be a template for other units. For this, the regiment is too individual.
But I would like to believe that there were more of these old troops left. And this is also very likely as my troops list clearly shows.

Regarding the armaturae: this was my mistake and I have corrected this on my upper post. Of course I mean here drill instructors. That they were fighting in a combat group is possible, albeit difficult to say. What happens if such a man made mistakes during the fight? Who listens to his advice after the battle? But sure, it could be that they were fighting.
The armaturae were instructors acc. Gerhard Horsmann, Untersuchungen zur militärischen Ausbildung im republikanischen und kaiserzeitlichen Rom, page 93
and according my research they formed their own schola or collegium (AE 1908, 0009: ARMATVRA Q F ARMATVRAE).
Responsible officer for them within the Legion of Perge was the campiductor.

[Steven James]

Nathan wrote:

There's also something quite 'Vegetian' about the unit breakdown
 
Perge document list the Bracchiati semissales at 256 men, and the story of Saint Meletius an officer in the Roman army lists 252 of his soldiers put to death. So following your analogy, I could say that the Perge document is “something quite martyrdom about the unit breakdown.”
 
Nathan wrote:
But in what way does this relate to the information in the Perge lists?
 
Just showing how difficult the problem is and also information passed on for someone else (killing two birds with one stone).
 
Nathan wrote:
A unit of Equites was called a Numerus Equitum or a Vexillatio Equitum - the terms were synonymous. So these are not different-sized units.
 
I have been following the ND, which lists titles such as Equites, Numerus, Vexillation etc. Why would they have distinct titles when the terms are synonymous? It would be illogical.
 
Nathan wrote:
A cuneus may be different again, but the only clear distinction is between the old-style ala (still known in the limitanei) and the new-style vexillatio (aka numerus equitum).
 
You speak with authority that a vexillation is a numerus. What proof can you provide? Or is this just one of your theories?
 
Nathan wrote:
Lydus, as you know, gives an ala 600 men and a vexillation 500 men (de mag 1.46) - which does not help us much!
 
It would be helpful had Lydus provide the time frame for the unit size.
 
Marcel wrote:
Regardless of how it actually was, the unit cannot - at least to my opinion and research - be a template for other units.
 
Agreed. The information is too fragmented.