The Empire is dead, long live the army

[Justin Swanton]

This takes up Patrick's suggestion of a separate thread to discuss the persistence of Roman military units in the West after the fall of the Western Empire. The fact that there were post-imperial military formations that contemporary observers considered as being 'Roman' is not in question. Procopius makes this clear in his well-known quote from his History of the Wars, XII:

Now other Roman soldiers, also, had been stationed at the frontiers of Gaul to serve as guards. And these soldiers, having no means of returning to Rome, and at the same time being unwilling to yield to their enemy who were Arians, gave themselves, together with their military standards and the land which they had long been guarding for the Romans, to the Arborychi and Germans; and they handed down to their offspring all the customs of their fathers, which were thus preserved, and this people has held them in sufficient reverence to guard them even up to my time. For even at the present day they are clearly recognized as belonging to the legions to which they were assigned when they served in ancient times, and they always carry their own standards when they enter battle, and always follow the customs of their fathers. And they preserve the dress of the Romans in every particular, even as regards their shoes.

The questions are these:

1. To what extent were these military formations actually Roman? Were they recognisably Roman in their equipment, training, dress, lifestyle, etc. or were they just barbarian mercenaries who carted a Roman standard around with them?

2. How were they maintained and who maintained them?

3. What was their function and how did they integrate into the barbarian kingdoms they became part of?

4. How many and how large were they?

5. How long did they survive the demise of the Empire?

[Justin Swanton]

Taking up Roy's query in the Slingshot thread I'd like to focus on point 2.

Justin, it would be interesting to see your evidence for Roman units being maintained as units by Post Roman officials taking on the diversion of taxes to pay them and maintain them as units. There is evidence for landowners recruiting their own men as Jim says, but that is rather different from ieeping the state apparatus going.

The late imperial tax system and state apparatus collapsed in Gaul (and elsewhere) in the second half of the fifth century. My point is that local landowners could and did maintain military units on their territories out of the rents they collected from their tenants. Not money as such, but kind: food, clothing, equipment. The landowners in effect became these units' commanding officers, and there was a shift from a civil aristocracy leading a life of otium on splendid country estates (as remained the case for a while in southern Gaul) to a militarised aristocracy who lived a much simpler lifestyle in fortified towns.

Chris Wickham (Framing the Early Middle Ages, Part II) argues for the persistence of a landowning aristocracy in northern Gaul. The will of Remigius (+533) shows him to be a landowner in the typical Roman manner, and his status can hardly have been unique in that period. Later wills towards the end of the sixth century onwards reveal the Paris basin to be under almost exclusive landowner control, with little sign of independent peasant communities, though many of these landowners were from the newly-created Frankish landed aristocracy.

This is the square hole. Into it I fit the square peg: the Roman units that persisted after the Empire that bankrolled them disappeared. Someone had to be paying their wages. Note that Procopius mentions that these units  "gave themselves, together with their military standards and the land which they had long been guarding for the Romans" to the Armoricans and Franks. Military garrisons did not control land in this period; the Gallo-roman aristocracy did. The implication here is that the commanders of these units were the owners of the land.

This is deductive. I can't quote any text that says 'the Romans soldiers after the fall of the Empire fell under the authority of the local senators.' It just seems the most logical conclusion.

[aligern]

It is plausible and I think I said in the parent thread that Liebschutz leans that way. I'd suggest again that some units just collapsed, some hired out to local landowners, and the commanders of some strong armed their way to local power via say a marriage alliance. All are logical, as is hiring a barbarian with his war band to garrison your city or defended site. As most troops are going to be limitanei and farmer/soldiering  I would not see them as having terrific military qualities.  I would also doubt that fancy cavalry units survive, so the Equites Illyricani  and Sagittarii  would default to being ordinary cavalry, much reduced in numbers and serving  a lord who could provide for them. Why, after all would a landowner need to have a large specialist mounted unit when most of the work is chasing down raiders, strong-arming peasants and escort duty.
Of course , if you gather together enough bodyguard units from an area  then you have a force of cavalry and I imagine that is the composition of the Arvernian noble forces that were hurrying to help Alaric II at Vouille. I am not aware that the Arvernians are given a named leader which would support the idea of many smallish contingents
Do we find fortified rural villa sites in Gaul?  I am not sure that we do. That suggests that the soldiers of whatever provenance are related to the fortified towns.

We are all in the land of conjecture here, of course.
Roy

[Justin Swanton]

As most troops are going to be limitanei and farmer/soldiering  I would not see them as having terrific military qualities.

Presuming that there was a state of war between the Franks and Syagrius's realm from 486 until Clovis's baptism in 496/7, these are the troops that "proved their valour and loyalty to the Romans and shewed themselves brave men in this war." They were able to stop Clovis dead in his tracks for a period of ten years - the only one of his enemies who succeeded in doing so. After the peace the Franks and Gallo-romans held them "in sufficient reverence to guard them even up to my time." This argues quality, which would suggest (looking at point 1) that they were a good deal more Roman than barbarian in their military makeup.

In Visigothic Gaul the landed gentry would have used any ex-soldiers they had in the manner you describe, and these would in consequence most likely have been low-calibre troops.

[Erpingham]

This argues quality, which would suggest (looking at point 1) that they were a good deal more Roman than barbarian in their military makeup.

I'm reading this exchange with interest in the hope of learning more about this period.  Can you explain why quality is directly related to Roman-ness?  I've not yet seen anything that suggests we aren't dealing with a faction, perhaps a particularly coherent one based around a Roman heritage, similar to the forces ranged against them?  The idea that these are regular Roman units with Roman discipline and tactics would need a deal more evidence.  You may, however, have such evidence up your sleeve 

[Justin Swanton]

If there was hard and detailed evidence that the Syagrian military were substantially Roman in discipline and tactics then this discussion would probably not be taking place. The trouble with history is that the theories grow but the evidence does not (or hardly does)  .

I look at it this way: Procopius's account that the 'Arborychi began to fight for the Romans' some time in the 470's makes sense only as a recruiting drive by Syagrius among his own subjects, the Gallo-romans of Armorica, replacing his reliance on the Franks with a surer dependence on a home-grown army. It was this army that fought Clovis up to his baptism in 496/7 and persisted as units under the overlordship of the Franks after that.

Question is, what was the nature of this army's training, equipment and structure? There are two opposing theories, with plenty of gradations in between.

Theory A: the Roman military tradition in northern Gaul had been dead for decades by the 470s, i.e. there were no longer any disciplined units with experienced officers capable of training new recruits in the Roman manner. The only model Syagrius had when he recreated his army was the barbarian one. His troops were equipped and trained in the barbarian manner, abeit with Roman battle standards and in Gallo-roman clothing, and that is how they fought.

Theory B: the Roman military tradition did survive in northern Gaul. Old formations were intact, albeit reduced in size, and they formed the nucleus of Syagrius's new army which was trained and organised by them. This force, substantially Roman in character, kept its identity even after the fall of Syagrius, only gradually losing it in the course of the 6th century.

Well, you can take your pick, but let me offer a few arguments in favour of theory B.

Nothing proves that the Roman army entirely disappeared in Gaul in the mid 5th century. It is more likely that it was much reduced in size and no longer capable, by itself, of fighting large battles. Hence Aetius's need for barbarian allies to face the Huns. Procopius's affirmation that 'even at the present day they are clearly recognized as belonging to the legions to which they were assigned when they served in ancient times', implies that the old legionary formations persisted right to the end of the 5th century and beyond. The fact that the soldiers belonged to 'legions' seems to rule out the notion that they were barbarian foederati, as these kept their own tribal structures and did not organised themselves into legions.

Of course, one could cast doubt on Procopius's use of the word 'legion' just as one could cast doubt on his whole account, but I prefer to take it at face value unless there is solid evidence for not doing so.

The economic and social network in northern Gaul remained intact throughout this period, as, for example, the pottery record shows, hence the means existed to maintain the traditional Roman formations. If these were privatised by the landed gentry then their upkeep was assured.

One can question whether these formations preserved all the nuances of the old Imperial army, but what would have made for differences between them? The Roman Army had always been an autonomous entity - the troops owed loyalty to their generals, not to the Empire as a whole. This attitude would have remained intact even when there was no longer an emperor and the 'Empire' had shrunk to northern Gaul. What motivated Roman troops to keep their standards up would have remained in place even after 476. The one thing that would have degraded the Roman character of the army is if it had ceased to be professional: paid troops living apart from the general populace. Nothing suggests that this was the case in northern Gaul.

There's a parallel to this in modern-day Britain. With the British Empire gone the British army has remained as professional as it ever was - even more so. It just became smaller. If you don't have numbers but you do have money then the natural inclination seems to be to go for quality.

[Owen]

Might the "arborychi" to whom Procopius refers be the remnants of Riothamus' British army - the timing is right?
Justin's analogy with the modern British army raises a possible comparison with post-colonial Africa, where there are essentially two effective military models: first, the traditional European system inherited from the colonial powers, second the "revolutionary" model acquired either fighting the colonialists or - more often - overthrowing an African government.  The Malawi Army, for example, has very strong British traditions - one of its battalions is still sub-titled King's African Rifles.  The Rwandan Defence Force on the other hand comes from the guerrilla movement which came to power in 1994.  Both models can be effective in creating competent forces and, coincidentally, there can be a strong "warrior" tradition too (like the Tuareg rebels in Mali).  The point is that such models create an ethos which can be quite durable, even in straitened financial circumstances.  The Malawi army is not well-resourced but, like Procopius "Romans" perhaps, it's still fiercely attached to its Regimental Colours.
Owen

[Jim Webster]

Just to put things into context with regards landowners supporting units. If we're wanting people to be 'full time' soldiers then they really have to have enough land to support servants as well.
So the figures I've come across I used in a slingshot article some time back.
The figure is the number of acres needed to support them.
   
Nicephorus II cataphract cavalry    460.8
Norman Miles in England      180   
Cavalry of Themes   144
Cavalry of Tagmata   144
Cibyrrhaeot marines   144
Infantry of Themes   30

To get things in perspective, the landholding of a Thematic infantryman should produce enough grain to support 3.25 peasant families. So one family can actually farm the land, and the surplus is enough to support him.
We now have to look at how large some of these estates were,  has anyone got any figures?

Jim

[Justin Swanton]

Might the "arborychi" to whom Procopius refers be the remnants of Riothamus' British army - the timing is right?

Jordanes affirms that Riothamus's men fled to the Burgundians after his defeat by the Visigoths:

(XLV.237) "Now Euric, king of the Visigoths, perceived the frequent change of Roman Emperors and strove to hold Gaul by his own right. The Emperor Anthemius heard of it and asked the Brittones for aid. Their King Riotimus came with twelve thousand men into the state of the Bituriges by the way of Ocean, and was received as he disembarked from his ships. (238) Euric, king of the Visigoths, came against them with an innumerable army, and after a long fight he routed Riotimus, King of the Britons, before the Romans could join him. So when he had lost a great part of his army, he fled with all the men he could gather together, and came to the Burgundians, a neighboring tribe then allied to the Romans. But Euric, king of the Visigoths, seized the Gallic city of Arvernum; for the Emperor Anthemius was now dead."

At this time the 'Romans' under Aegidius or Syagrius had a force of their own, though to what extent it was a mix of Frankish federates and Roman troops is unclear. Anthemus ruled from 467 to 472.

On the subject of Syagrian legions, the account of the 'British legion' north of the Loire in the Vita Sancti Dalmatii is interesting:

'Naturally, after the realm of the Franks [who were] pious and illustrious and devotees of the Christian religion, had subjugated the city of Rodez (the people themselves conspiring in their [the Franks'] favour), the priest [Dalmas], filled with desire, strove to look upon the presence of the Christian king Theudebert. As the devout one [Dalmas] was tirelessly hurrying to him [Theudebert] in the region beyond-Loire [or: beyond-Loir], it is said he enjoyed an evening's hospitality in a certain place where some sort of Breton [or: Brittonic] legion (so to speak) nearby was stationed [or: was waiting].'

This event took place after 534 when Theudebert became king. The best sense of the term 'ultralegeretannis' - 'beyond legeretannis' is that legeretannis is a scribal corruption of ligeranis 'pertaining to the Loire', which places it north of the Loire.

This mention of a legion is couched in prosaic terms in an account full of the marvellous, which if anything confirms its authenticity. What is interesting is that the legion is in exactly the place one would expect to find those 'other Roman soldiers' that defended the Loire frontier against the Visigoths. Its name is also interesting: 'legio bretonum'. The only legion with a name like this is the II Britannica,  formerly posted in Britain before being withdrawn to Gaul at the end of the 4th century. Three of its vexillations - stationed in Gaul - are mentioned in the Notitia Dignitatum.

Speculation of course, but adding up to a coherent picture.

[Erpingham]

Question is, what what the nature of this army's training, equipment and structure?

I would essentially agree.  While the facts are admittedly scattered across time and locations, there is enough to say that forces identifying with former Roman army units persisted for some time after the disappearance of the Western Empire.  I think where we differ is how militarily, rather than culturally or traditionally, Roman they were.  We have evidence that some, at least, kept their standards but does this mean a traditional command or simply an affirmation of tradition and identity.  They probably had a legacy stock of arms and armour - perhaps more than a passing warband would own.  But other than that?  I don't know.

[Justin Swanton]

Just to put things into context with regards landowners supporting units. If we're wanting people to be 'full time' soldiers then they really have to have enough land to support servants as well.

We now have to look at how large some of these estates were,  has anyone got any figures?

Jim

I'm no expert on this. Just a couple of ideas:

Roman farms were measured in iugera (1 iugera = 0.65 acres). Kehoe Economics of Agriculture on Roman Imperial Estates in North Africa, affirms that you needed 2 jugera to support a man.

The Gallo-roman notables tended to own many large farms (latifundia, each 500 iugera or more), though usually within a single region. Collectively, these could amount to thousands or even tens of thousands of iugera. Given that the great majority of the population lived in the countryside and most of these, in Syagrius's realm at least, resided on these huge estates, one can assume that sustaining a unit of several hundred men would not have been an undue burden for a landowner.

I can't however locate any precise information on the size of a typical landowner's holdings in Gaul. Can anyone help?

[aligern]

On the Italian frontier we have two examples. The first is the Life of St Severinus which speaks of the Roman soldiers at Passau on the Danube who send off a detatchment to get the pay, then see their bodies come floating back downstream and so disperse as no more pay would appear to be forthcoming. This is around 470 Military units that are not paid and especially not fed, can collapse very quickly.

Contra this, in the early sixth century Cassiodorus pens a letter of Theoderic the Great ( IIRC) writes to a commander of limitanei who are probably in NE Italy. In Procopius 'Buildings' there is a reference to Justinian replacing a garrison of farmer soldiers with a garrison of regulars which is unpopular, presumably because that means the regulars need a tax contribution.

However, town garrisons need not be regulars, or even limitanei. In the Vth century Caesarius of Arles refers to sections of the walls being maintained by sections of the community, in this case the Jews and in the 530s The Jews defend their section of the walls of Naples against the Goths. Hence it would be a fair presumption that the quarters of a town were allocated wall sections to maintain and to defend. In the mid Vth century Majorian specifically removed the penalties for civilians carrying arms so that they could defend themselves against Vandal raids.
As to Procopius' accuracy, he is working in an Herodotean tradition and thus salts his account with stories which may have a basis in fact and may be embroidery. The story of surviving legions is in this category because we just do not know enough to be sure. Justin can see them as militarily effective and  others can choose to show them as decayed remnants that provide part of a town defence, but are soldier farmers whose radius of action is the farmland of their city. Sadly we do not seem to have an example of them being involved in military activity, but then many of these multi contingent armies are not listed as conveniently as Majorian's is listed.
One pice of comparative continuity is the case of Goths in Septimania. these go on into the ninth century, after the fall of the Visigoth kingdom and the Frankish eviction of the Moors from the region. I remember reading that the longevity of the appellation 'Goth' might well be due to their being financial privileges attached to being a Goth which would have meant hereditary military service. Given there was a benefit to bearing the label it continued on for 200 years after the  political entity to which it related had gone under. If Justin's Romans had a reason to muster and dress up at the drill hall once a month and receive a donative, or held land by reason of their membership of, in effect,  a guild then they would keep it going.
My doubts about these Romans in Gaul are that we do not know of any political organisation that keeps them going for 60 years. being attached to cities is fine, but that will be small scale and with no coherent means of assembly as the cities do not have expeditionary armies, well not large ones.

Roy

[Jim Webster]

I included Nicephorus II's cataphract cavalry with their  460.8 acre estate because they would to an extent be prominent local landowners
It's also not all that far from 500 iugera

From memory here, wasn't the iugera relative rather than absolute? Like the hide it was a measurement  based on yield rather than mere area.

Mind you, if we take a landlord with 1500 acres.
He'd need 500 to maintain himself and his family (minimum)
Another 500 would support three or four full time cavalry
Another 500 would support 16 infantry.

To support a cavalry unit you'd need over 40,000 acres.

Jim Webster

[Justin Swanton]

To support a cavalry unit you'd need over 40,000 acres.

Jim Webster

That's about 160 square km, or 100 square miles. Syagrius's realm would have been in the region of 80 000 square km, most of it arable. Devote a twentieth of that to military upkeep and you have either 3500 cavalry or 40 000 infantry. Go fifty/fifty and you have 1750 cavalry and 20 000 infantry - far more than one can theorize as actually having been subsidized.

[Patrick Waterson]

As a slight detour from the mainstream discussion we might argue for the persistence of sagittarii on the basis that a number of Alans seem to have settled in northern Gaul during the 5th century AD and their horse archery skilld are likely to have lasted for at least one generation.  We might also consider the possibility of Illyricani being supplemented and eventually replaced by the light scout/raider types which Celts in general and Bretons in particular were fond of fielding.

The essential quality of 'Romanishness' that I see mattering is discipline (and for that matter fighting technique), with organisation as a secondary but useful feature.  If you have an army that does what you want it to then you have an advantage over an opponent whose army does more or less what he wants it to. 

I see a general pattern in that the quality of Roman units seems to have been in fairly direct proportion to their proximity to a seat of rule.  Troops far from a centre of rule, e.g. the units in Spain, seem to have been lacklustre, but the Illyrian units under the Nepos family were crack troops.  The armies of Britain and Gaul were traditionally among the best troops the late Empire possessed, and had a tradition of making (usually ultimately unsuccessful) emperors.  Aetius had built up a power base in Gaul (and presumably done his non-Hunnish recruiting there), with Aegidius and Syagrius apparently adopting this as a more or less going concern. 

The question that seems less easy to address is the extent to which an effective taxation system still prevailed.  If the Roman civil service persisted, then it may have drawn useful revenue from a basically prosperous province, as the usual contribution to the Imperial capital would no longer be made, allowing the entire revenue (such as it was) to be used by the ruler of what remained of Roman Gaul.  In practice, one would envisage a degree of cutting out the middleman, with local nobility maintaining their own units (a revised limitanei system) and the ruler keeping a core of good quality palatini, both to give him credibility with barbarian tribes and to ensure that he remained the ruler.

One might assume that overall the towns provided the money, the countryside the manpower and the fabricae the equipment.  Most importantly, having a ruler on the spot provided the means to keep everything coordinated and viable.