The 'Lost' Naval Commands of Late Rome

[Longovicium]

Duncan McAndrew's article entitled 'The Milites Hypothesis: The Lost Naval Commands of the Late Roman Army' makes for some intriguing reading.


The basic thesis is that the units described in the Notitia Dignitatum which bear the descriptor 'milites' are in fact naval units distinct from the classis and barcarii units. He marshals several pieces of statistical evidence to buttress the theory and it all stacks up quite nicely as far as it goes. The conclusion is that if the units are re-conceived as naval units then the 'map' as it were of infantry to cavalry to naval ratios changes quite dramatically and also plugs several gaps in riverine, lacustrine, and estuarial defenses.  His evidence rests primarily on analyses of geographic distribution, unit titles and their residual naval terms, and finally ethnic titles.

As far as hypotheses go, it makes for persuasive reading.

To my limited understanding of Late Roman history however, a number of questions arise which I wonder if anyone here can shed light on?

The first which springs to mind is the lack of epigraphical evidence in his article. In short, there is none. His hypothesis is bound to the Notitia Dignitatum and concurrent research around that document and the units it details. This led me to wonder what gravestones or inscriptions might survive which list a miles serving in any of the milites units listed in the ND? If any have survived, do any of them bear out any possible naval or riverine connection of any sort - or alternatively, dispel such a connection?

Secondly, following on from the above, can the rise of the term milites/militum as a unit descriptor be traced in the documents and historical writings in a similar manner as limitanei and comitatenses and vexillationes? If so, again, are they used in a context which might support or cast doubt on this hypothesis? ? It is one thing, I think, to argue that these units in the ND are specifically naval units as they stand but one would also need to examine the origin of that term to qualify it.

Thirdly, the diocese of Britannia, already holding a maritime command whose ultimate expression is the Comes Litoris Saxonici, and which also holds one of the few attested Barcarii units (possibly two), bears only one known 'milites' unit - the Praepositus militum Tungrecanorum stationed at Dover. If McAndrew's thesis were accurate, one would expect a similar scattering of units along the Saxon Shore as well as up along the region about Lancaster after its re-fortification in the style of the 'Saxon Shore' forts. It doesn’t. The numerus is the main unit descriptor rather than militum. Although, interestingly enough, in Gaul there are two units possibly derived from Pevensey, the milites Anderetianorum and the classis Anderetianorum possibly marking a re-deployment and/or subsequent split which the Praepositus numeri Abulcorum unit then replaced.

I think, to sum up, I am wondering if this hypothesis can be broadened out and if anyone here at RAT can shed light on the use of the militum descriptor outside its use in the ND?

Any thoughts or references would be appreciated!

[Flavivs Aetivs]

Wow, looking at this, it makes a lot of sense, if it's not outright revolutionary.

It also kind of screws up Pseudocomitatenses units. A lot of those are drawn from "Milites" garrisons.

[Nathan Ross]

can the rise of the term milites/militum as a unit descriptor be traced in the documents and historical writings in a similar manner as limitanei and comitatenses and vexillationes?

I don't think they can - it's such a general term that I'm not sure we can draw anything much from it.

The article seems quite well argued, although the premise is shaky. There's not quite enough emphasis, I think, on those units and commands listed in the ND that do pertain to ships - specifically the various legion and other commands on the Danube. For example, there's the Praefectus legionis quartaedecimaegeminae militum liburnariorum, Praefectus legionis secundae Italicae militum liburnariorum, Praefectus legionis primae Noricorum militum liburnariorum, and two praefecti classis in Pannonia and Noricum, the Praefectus ripae legionis primae Ioviae... et secundae Herculiae musculorum Scythicorum et classis and Milites navclarii in Scythica, Milites tertii navclarii and Milites navclarii Altinenses in Moesia, and the Praefectus classis Ratianensis in Dacia. If 'milites' were naval troops, why would they need the 'navclarii' suffix?

The legion commands suggest, I think, that the bulk of the riverine fleets of the 4th-5th centuries were manned and controlled by the legions of the frontier garrisons. There's a 4th-century tile stamp of XXII Primigenia from Mainz showing galleys of the lusoria type, which might indicate that the Rhine fleet(s) were also under legion control; if we had a full list of the Rhine garrisons from a few decades before the date of the western ND, I suspect we might have seen something like a Praefectus legionis duoetvicensimae Primigeniae militum liburnariorum!

[Longovicium]

Apologies for not replying sooner but got hit with manflu last Friday and only just shaking it off now!

Nathan, I think you are right about the use of milites in terms of ascribing any technical sense to it. However, your post above about the legion troops and the naval commands associated with them made me go back to the ND. Luke Ueda-Sarson's website carries some interesting breakdowns of these latter. He reports that some 38 milites units are specifically legion troops and many of them have naval titles or descriptors. Those who don't - according to McAndrews - either fill gaps in the known record of naval units or potentially could.

The remaining milites listed in the ND - some 23 in total - are auxiliary units and are demonstrably not legion troops brigaded out to specific naval operations. However, of these 23 troop units, all but 3 are stationed either under the command of the Dux Scythiae or the Dux Moesiae Secundae. Both these frontier commands also list legion troops below the register of the auxiliary milites. That's a very interesting grouping of units under 2 closely related frontier commands quite distinct from the remaining legionary milites identified by Ueda-Sarson. 

My first thought was that these auxiliary milites - some of which have clear naval titles - were raised in response to a specific localised emergency on an ad-hoc basis to supplement the legion naval units and then remained on the lists as such. Speculation, obviously, but I do wonder on why those 21 milites exist as distinct from the legion naval/marine units under only 2 frontier commands closely associated together geographically and sited along the critical Danube frontier. The obvious historical event is the post-Adrianople period and the re-ordering of the frontier later . . .

Again, speculation, though.

As for the reason why the milites descriptor arose at all here in the ND, McAndrews points to Vegetius' use of the term in a naval context to describe what we today call marines. Is it possible that with the establishment of legion troops tasked with specific riverine and lacustral remits, it was adopted to distinguish them (along with other suffixes and titles) from main-line infantry legion troops? While still retaining a 'generic' quality, the descriptor, when applied to units stationed at specific river and lake and estuary fortifications took on a slightly less ambiguous meaning and stood to define specifically marine and/or naval units? Again, speculation and perhaps impossible to verify.

If one takes a step back from the ND with MacAndrews and Ueda-Sarson to hand, it does open up the possibility that late Roman naval capability (certainly in terms of its non-maritime aspect) was more developed and organised than I had perhaps appreciated in the past.

[Nathan Ross]

Glad you've escaped the dreaded manflu! Plenty of that about at present.

Lot's of speculation, yes, and I think if the text of the ND is all we have to go on we're unlikely to squeeze any new meaning out of it. In as much as riverine defences are always going to involve boats or ships to some degree, these 'milites' may have had something to do with that. But I don't think we can sort of retro-fit them as marine contingents!

McAndrews points to Vegetius' use of the term in a naval context to describe what we today call marines. 

This had always been the case - inscriptions from the principiate list (for example) milites classis praetoriae Misenensis. A miles was a soldier, and it didn't matter whether he was on land or water! But the word is used for almost every other military application too.

I agree that the 'milites' of the ND would fit best as a sort of locally-raised force, maybe a kind of militia in fact. Alternatively they could be barbarian troops, like the band of foederati that did so badly trying to protect a town in Noricum (I think) in the life of St Severinus.*

[*Edit - "Comagenis... very strictly guarded by the barbarians established within, who had entered into a league with the Romans" (Vita San.Sev 1) - the same account has Mamertinus the tribune with a small number of 'milites' holding Favianis.]

it does open up the possibility that late Roman naval capability (certainly in terms of its non-maritime aspect) was more developed and organised than I had perhaps appreciated in the past.

There certainly was a significant riverine naval capacity - there's that note in the Theodosian Code about several hundred (I think) lusoriae on the Danube*. We could probably imagine the same sort of thing on the Rhine, and on Lake Como too.

But we'd have to ask what the purpose of a river navy actually was - the Romans weren't intending to fight battles on the Danube or the Rhine. The sort of small craft involved - if we can take the Mainz ships as typical - would have been useful for scouting and patrolling, keeping the shipping routes open and attacking small bands of marauders. Even a hundred vessels like that could be manned and operated by a couple of thousand men. Once any larger barbarian incursion managed to cross the river, the ships would be of little use. All the more unnecessary, then, to turn all the frontier milites garrisons into marines, I think!

[*Edit 2 - 100 lusoriae in Moesia and 135 in Scythia: Co.Theo 7.17.4, Jan AD412)

[Flavivs Aetivs]

[*Edit - "Comagenis... very strictly guarded by the barbarians established within, who had entered into a league with the Romans" (Vita San.Sev 1) - the same account has Mamertinus the tribune with a small number of 'milites' holding Favianis.]

These events date to 454 or 455 (ish) and I believe that the revolt in Comagenis was tied to the death of Aetius or maybe Valentinian III, because the death of the leader in Barbarian Culture technically nullified a treaty. The guards of Comagenis were probably Foederati.

Mamertinus' unit (along with Batavis) appears to have been an understrength Limitanei unit. Most of the Rhine and Upper-Middle Danube garrisons were still intact until 450, then a lot of them got wiped out by Attila.

I think it's possible that a lot of the stationings may have been similar to what we see with V Macedonica in Egypt: a central unit (in Thebes, numbering slightly less than 400 men in 399 AD) with several centuries split off in the surrounding towns (Hierapolis, Antinoupolis, and Antipolis). Hence the small garrisons like that at Batavis and Favianis.

[Nathan Ross]

The guards of Comagenis were probably Foederati.

I think so - the 'league' mentioned in the translation would be a foedus. This suggests, perhaps, that relatively small bodies of barbarians could enter Roman service and be allocated to the defence of a single town or area. Origin of our 'milites', maybe?

Mamertinus' unit (along with Batavis) appears to have been an understrength Limitanei unit.

Yes, probably. Although the ND gives us Praefectus Legionis Liburnariorum Primorum Noricum Fafianae - presumably the same place! So in c.400, the town was held by a naval detachment of legionaries from I Noricorum. 50 or so years later the garrison had changed to a small group of 'milites' under a tribune, poorly armed and fighting on land...

[Flavivs Aetivs]

Exactly, hence my theory that what was happening on the limes was similar to V Macedonica in Thebes.

[Longovicium]

Nice call out regarding the lusoriae numbers in the Theodosian Code, Nathan - did you know I have just obtained Pharr's translation and have it on my desk now?!

I see that Title 17 (page 175 of Pharr's translation) mentions judiciarae and agrarienses vessels renovated from old craft - presumably lusoriae. These latter are defined as reconnaissance craft and inshore patrol craft respectively. Souter's ' A Glossary of Later Latin' defines the agrariensis as a ship or boat but from its designation, it is clearly an inshore patrol craft. The listing above this one is for 'agraria' - a kind of military watch, while the descriptor itself is derived from the Latin word for fields - 'agralis'. This vessel then is distinct from the lusoriae in that it is tasked with close inshore patrols along river-banks.

D'Amato's latest work 'Imperial Roman Warships 193-565 AD' makes a different case for the former craft. He argues that the converted judiciarae are in fact official government ships used to convey a high functionary. Souter agrees here. This would contradict Pharr's understanding of them as being reconnaissance craft.

Interesting that the Theodosian Code preserves an order for the refit and construction of patrol craft along the Moesian and Scythia riverine borders precisely where the main non-legion 'milites' units are raised under the Dux Scythiae and the Dux Moesiae. I think you are right, however, about wondering on the use of a naval presence along the Rhine and Danube frontiers. I think Haywood in his 'Dark Age Naval Power' sum it up quite nicely when he suggests that later Roman naval capability should be seen more in terms of amphibious operations than proper naval capability. Roman naval actions focus more on patrol, transport and scouting than actual ship-to-ship combat.

[Nathan Ross]

judiciarae and agrarienses vessels renovated from old craft - presumably lusoriae.

Yes, I think the title of that piece is de lusoriis danuvii, so these are all military vessels (lusoriae) of some sort.

Agrarienses is interesting: if the 'fields' thing has any relevance than might these be nothing more than converted agricultural transports - maybe grain barges or something?

I don't know anything like enough about Latin etymology to work out judiciarae, but I could imagine that the word may have connotations of investigation, judgement or reconnaisance that would fit a scouting vessel, rather than being related to a magisterial function?

Although there are references in earlier inscriptions, I think, and an Egyptian papyrus of AD402, to governors having their own river vessels... I've mentioned P. Vindobona Boswinkel 14 before, I think - a note about repairs to (military?) galleys on the Nile dated c.AD324.

[Longovicium]

The navis agrariensis is an intriguing class of ship. As Souter lists 'agraria' directly prior to this word and defines it as 'a 'kind' of military watch' while prior to 'agraria' he lists 'agralis': ' . . . with regard to fields', I suspect the ship-type is less to do with transport than it is to do with surveillance or inshore patrol. I am not familiar with the agraria listing. What sort of military watch is Souter referring to here, I wonder?

D'Amato speculates on the judiciariae vessels and sees one of the Mainz vessels as a possible example. Vessel 3 is beamier and was fitted with a cabin - cubiculum. Souter lists several late Latin words derived from judicator/iudicator and all are to do with judgement, governor, magistrate, or belonging to the governor. The vessel then would seem to be one tasked to transport official civilian dignitaries upriver - superioris - and downriver - inferioris instead of using the road transport. It is amusing that the Theodosian Code specifically states that such vessels should be designated from the refurbished lusoriae and not the newly-built ones in Moesia but that in Scythia, iudiciariae must be brand-new.

I do wonder if these are not specific ship-types though but rather ship designations? The lusoriae class could easily be re-fitted or re-tasked into different roles such as close inshore patrol and also fast dispatch of official dignitaries up and down the Danube.

[Nathan Ross]

Souter lists 'agraria' directly prior to this word and defines it as 'a 'kind' of military watch'

Odd - all the definitions I can find (albeit online!) relate the word to agriculture. I've never heard of a 'military watch' (presumably a period of time) called anything like that.

I do wonder if these are not specific ship-types though but rather ship designations? The lusoriae class could easily be re-fitted or re-tasked into different roles such as close inshore patrol and also fast dispatch of official dignitaries up and down the Danube.

Yes, you're probably right. Although if I were a dignitary I wouldn't fancy travelling anywhere very far in the Mainz 1 ship (the fast galley) - I was taking a look at the stern of the reconstruction in the Schiffmuseum and there's barely room for a single person to squat uncomfortably in front of the helmsman!

[Renatus]

Souter lists 'agraria' directly prior to this word and defines it as 'a 'kind' of military watch'

Odd - all the definitions I can find (albeit online!) relate the word to agriculture. I've never heard of a 'military watch' (presumably a period of time) called anything like that.

Vegetius 1. 3 .3 has 'in agrariis ', which Milner translates as 'on outpost-duty'.

[Nathan Ross]

Vegetius 1. 3 .3 has 'in agrariis '

Ah, so he does! Although the passage is about the problems of conscripting city dwellers rather than countrymen, so the general idea of being out in the fields, in an agricultural rather than urban setting, probably still stands.

It doesn't help much with our boats, all the same - but I'd say the likelihood is that agrarienses derives in some way from a connection to agriculture. What the boat actually did in a military context remains obscure. I'd guess either transport or inshore (close to the fields?) patrols, perhaps.

[Renatus]

Vegetius 1. 3 .3 has 'in agrariis '

Ah, so he does! Although the passage is about the problems of conscripting city dwellers rather than countrymen, so the general idea of being out in the fields, in an agricultural rather than urban setting, probably still stands.

Lewis & Short refers to 'agrariae stationes ' in Ammianus 14. 3 and defines the expression as 'outposts'. I don't doubt that the etymology derives from the notion that outposts would, by definition, be out in the country and remote from urban settlements. If Vegetius and Ammianus are indeed referring to outposts, I guess that that might justify agraria being defined as an outpost or 'a kind of military watch'.