Abandoned Forts?

[Titus Juventius Tertius]

So I've been thinking lately. We all associate weapons, armor, shields, etc with finds from abandoned forts (defined as no longer in use today). Now often I see, "occupied by such and such legion in the such and such year". My question is, how reliable is this information?

I pose this scenario to you. Let's say Legio II (just picking a number) leaves a fort they built and have been known to occupy. Let's say a known Auxilia unit replaces them there. Now theoretically let's go ahead and now have Legio III replace those troops and it not be known.

Now everything we thought we knew about the fort and every item we hear about excavated there is false (or at least under a different assumption).

What brought this up was the fact that I look at how much work was put into these forts. They couldn't have been outright abandoned after the first unit left them. I would think they'd still be used even a 100 years later (look at current military bases or forts that last 50+ years even in hostile territory). One could dig up the fort thinking the unit in BC occupied it, when a different unknown legion could have occupied it in AD. Which means the items found may belong to the later unit and it throws our whole understanding of events, times, etc off.

Does this make any sense? Has anyone thought about this before? Is this a concern at all?

[Jona Lendering]

Does this make any sense? Has anyone thought about this before? Is this a concern at all?

Yes, it makes sense. What you're aiming at is sometimes called "the positivist fallacy": we need evidence for what happened in the past, but much is lost, and we do not know how much is lost. Now there are, if I may quote the well-known American poet D.H. Rumsfeld, known unknowns and unknown unknowns. The first category includes such things as "the vicus was occupied so the fort was occupied, but we do not know by which unit", and the second category includes your case: we have no evidence that there is something we do not know. This is, in archaeology, the subject of "reconstruction theory".

I could not resist making a joke about Rumsfeld's well-known quote about the unknowns, which can indeed be read as a poem. But in fact, he was absolutely right.

[D B Campbell]

They couldn't have been outright abandoned after the first unit left them. ... One could dig up the fort thinking the unit in BC occupied it, when a different unknown legion could have occupied it in AD. Which means the items found may belong to the later unit and it throws our whole understanding of events, times, etc off.

Interesting question, T.J. But our knowledge is not usually as sophisticated as you think!

Archaeologists usually have a good reason for believing that a particular Roman fort was "abandoned" (I prefer "vacated", which raises fewer assumptions). All excavated fort sites show evidence of phasing, by which I mean the usual round of care and maintenance as buildings are refurbished and roadways resurfaced. Occasionally, a site-wide destruction layer indicates a major break in the phasing (i.e. for some reason, the entire fort has been "destroyed", or -- perhaps better -- "dismantled"), and in many cases occupation continues above the destruction layer, often with buildings in a different layout, even a different size of fort. But in some cases, there is no continuation, so the site has been "abandoned" (or vacated).

Most of the equipment finds -- and our colleague Dr Bishop will (I hope) keep me right here -- derive from this moment, when a departing garrison is jettisoning anything that they can get away with not carting away. Interestingly, besides broken hinges and rusty buckles, this often includes dinnerware, which seems to have been readily available at the next stop. There may even have been a ritual element in the departure -- the Romans were such a superstitious lot that we should expect nothing less.

Of course, eventually all fort sites must end. But in the majority of cases where a fort can be demonstrated to have been demolished and rebuilt, we assume that the reason is linked with a change of garrison. Why else would you knock down a fort and build a new one? (That's not a rhetorical question, by the way, but a subject for an entire thread of its own!! ) Often, we can demonstrate a change of garrison, which ties in neatly, and -- before you jump to the wrong conclusions! -- the original garrison can often be picked up subsequently in the archaeological record elsewhere.

So -- with apologies for the lecture -- we rarely know exactly which unit was garrisoning any particular fort in any given year, so misplacing the odd cohort would have no major repercussions. Does that about cover it? (No, I didn't think so. :wink: )

Edit: Oops ... I was posting at the same time as Jona -- now I must go and read what he wrote!

[Matthew Amt]

So I've been thinking lately. We all associate weapons, armor, shields, etc with finds from abandoned forts (defined as no longer in use today). Now often I see, "occupied by such and such legion in the such and such year". My question is, how reliable is this information?

Well, troops that built walls and buildings often included inscribed stones indicating their unit, and sometimes a date I think. Or we might find items with an owner's name and unit marked on them, or altars with inscriptions. Obviously those aren't solid *proof*, just good evidence, but usually that's how a conscientious writer will phrase it. Hadrian's Wall is one exception, because it was built by legions but manned by auxiliaries, so "centurial" stones from the wall naturally don't tell us anything about the garrison. There might also be references in literature to where a unit was based at a particular time, and it's really handy if such a reference happens to agree with archeological finds from that site, but I have a feeling that's pretty rare.

I pose this scenario to you. Let's say Legio II (just picking a number) leaves a fort they built and have been known to occupy. Let's say a known Auxilia unit replaces them there. Now theoretically let's go ahead and now have Legio III replace those troops and it not be known.

Well, to be picky, legions tended to be based a few miles back from the frontier, in large fortresses, while auxiliaries were in smaller forts right along the border. But I get what you're getting at. A lot depends on how long a unit was in one place, and of course the extent of any excavations (generally a surprisingly small percentage of the entire site!), or stray finds. But I have a feeling it would be difficult for a legion to be posted in a fort for a generation or two without leaving SOME kind of evidence of its identification. As others have pointed out, vacating a fort usually involved the dumping of a lot of old stuff, in effect creating a "high water mark" for us to find. So as I understand it, you tend to get numerous sites with good finds in areas where the army arrived, stayed a while, then moved out to go forwards or backwards, effectively moving the whole frontier. Anyplace they moved out of could have a good dump, or stuff under a datable destruction layer. Areas that were calm and involved no significant changes of garrison would turn up a lot less in the way of cool archeological finds, even if they were occupied longer.

Now everything we thought we knew about the fort and every item we hear about excavated there is false (or at least under a different assumption).

Oh, we get that already! For instance, the best information on lorica segmentata comes from Corbridge--an AUXILIARY site. Hmmm.....

What brought this up was the fact that I look at how much work was put into these forts. They couldn't have been outright abandoned after the first unit left them. I would think they'd still be used even a 100 years later (look at current military bases or forts that last 50+ years even in hostile territory). One could dig up the fort thinking the unit in BC occupied it, when a different unknown legion could have occupied it in AD. Which means the items found may belong to the later unit and it throws our whole understanding of events, times, etc off.

Well, pottery and coin finds will help date the various archeological strata, so if we believe a particular unit moved out of a site at a particular time, it follows that artifacts dating later than that must be from a different unit. But VERY few artifacts can be ascribed to a particular unit, really, in fact we usually can't look at a sword or scabbard or piece of armor and be sure if it was used by a legionary or an auxiliary (without an inscription, of course). It's really looking like different types of infantry units used a lot of the same gear (though perhaps with different tendencies one way or the other). I think that the general movements of the legions at least are reasonably documented, so it's not too likely that a change of garrison could happen without our having some inkling of which legion was leaving and which one arriving.

In short, it's true that specifics can be vague or shaky, but even if we aren't certain of the exact progression of garrisons, it's not likely to shake the foundations of our historical knowledge for that site, or the region, or military equipment in general. If you keep in mind that an awful lot of our knowledge is already generalities, exact identifications and chronologies are not that big an issue. New digs and finds can only add data.

Vale,

Matthew

[D B Campbell]

... troops that built walls and buildings often included inscribed stones indicating their unit, and sometimes a date I think.

This is indeed the main source of evidence for assigning units to forts. Whenever the army built something, they were very keen on advertising the fact, usually as a dedication to the emperor, which is where the date comes in. (I can think of only one building inscription that carries a consular date -- maybe there are others? --, which pins it down to a particular year; normally we're limited to emperor, sometimes fine-tuned by the particular titulature of the emperor, which -- as you know -- changed from year to year.)

... or altars with inscriptions.

Another good example, Matt. Only the unit in garrison would dedicate an altar -- there's some evidence to suggest that the unit should be dedicating a new altar every year, but we have nowhere near enough actual examples to prove this.

Hadrian's Wall is one exception, because it was built by legions but manned by auxiliaries, so "centurial" stones from the wall naturally don't tell us anything about the garrison.

The Antonine Wall, also, was built by the legions. (And the Raetian wall, but everybody forgets it. ) But, remember -- the forts on every frontier were largely manned and maintained by auxiliaries.

There might also be references in literature to where a unit was based at a particular time, and it's really handy if such a reference happens to agree with archeological finds from that site, but I have a feeling that's pretty rare.

Compared to epigraphy (i.e. inscriptions), yes! (You're basically limited to Tacitus -- ironically, from memory, none of the forts and garrisons he mentions are corroborated by other evidence. Fortresses are another matter, because as a good senator, he likes to mention the legions.)

... legions tended to be based a few miles back from the frontier, in large fortresses, while auxiliaries were in smaller forts right along the border.

You're not alone in believing this generalisation, Matt. I think it probably derives from Luttwak and his diagrams. It's really only true of Britain (but think of Inchtuthil, almost the northernmost Roman fortification in the Flavian empire). Elsewhere, the fortresses take their places right up on the line, keeping an eye on their auxiliary cousins. (Think of Nijmegen, Vetera, Neuss, Cologne, Bonn, Mainz, ... Carnuntum, right in the thick of it for Marcus Aurelius' Marcomannic Wars, Vindobona, Aquincum, ...)

As others have pointed out, vacating a fort usually involved the dumping of a lot of old stuff, in effect creating a "high water mark" for us to find. So as I understand it, you tend to get numerous sites with good finds in areas where the army arrived, stayed a while, then moved out to go forwards or backwards, effectively moving the whole frontier. Anyplace they moved out of could have a good dump, or stuff under a datable destruction layer. Areas that were calm and involved no significant changes of garrison would turn up a lot less in the way of cool archeological finds, even if they were occupied longer.

Actually, I don't think there's that much of a difference. Think of Vindolanda, where a change of garrison has generated a vast amount of dumped writing tablets. Then think of (e.g.) Eining, where a stable long-lived garrison has generated a huge amount of inscriptional evidence.

For instance, the best information on lorica segmentata comes from Corbridge--an AUXILIARY site.

Is it?! :wink: [size=85:1rvbpcor](Off to have dinner, now.)[/size]

[PhilusEstilius]

I do think that I have to agree with Duncan as far as Corbridge is concerned, it did have connection with Auxiliaries but only from the fact it was a major Roman town near to Hadrians' Wall that supplied troops from it's Fabrica. Indeed I do have artifacts from the area that would confirm this.

[Gluteus Maximus]

Slightly off topic, but pertinent to some of the content of this discussion: has it been unequivocally proven that Auxillia did NOT use Lorica Segmentata? I mean, I have always 'known' this to be the case, but how was this conclusion reached? I have often pondered the case of the Corbridge Lorica but shoved the conundrum to the back of my mind.

[Matthew Amt]

Ah! Thanks, Duncan and Brian for the clarification on Corbridge. Actually that sounds familiar, now that you've mentioned it, so I probably already knew it and forgot. (Frantic tap-dance to save reputation, ha!)

Slightly off topic, but pertinent to some of the content of this discussion: has it been unequivocally proven that Auxillia did NOT use Lorica Segmentata? I mean, I have always 'known' this to be the case, but how was this conclusion reached? I have often pondered the case of the Corbridge Lorica but shoved the conundrum to the back of my mind.

No proof at all! It's probably a generalization grown from the images on Trajan's Column, which is *believed* to be pretty consistent in showing legionaries in segmentata and auxiliaries in hamata. We think. Usually. But the archeological record is a lot fuzzier, with segmentata bits coming from what are believed to be auxiliary sites. Then we argue whether that means auxiliaries wearing segmentata, or contingents of legionaries in auxiliary forts! I'm not sure it's quite safe enough to claim that they all wore the same stuff, because we run across things like the bit from Tacitus where troops are pulling gear out of the armory without regard for whether it was meant for auxiliaries or legionaries or Praetorians (something like that). Not a perfectly clear passage, of course, but it sets off alarm bells.

Valete,

Matthew

[D B Campbell]

... has it been unequivocally proven that Auxillia did NOT use Lorica Segmentata?

That's a difficult one to prove. It has always seemed most likely that they didn't, but then ...

I have often pondered the case of the Corbridge Lorica but shoved the conundrum to the back of my mind.

Thank you for reminding me. When Matt mentioned Corbridge, I rather myopically thought of the 2nd/3rd C town with its military compound, and forgot the Flavian/Trajanic forts which originally stood there. The presence of the Corbridge hoard -- Dr Bishop is the expert here -- in an "auxiliary" fort still presents a problem: (a) were there legionaries amongst the garrison? (b) did legionaries send their broken equipment to Corbridge for recycling? © were there auxiliaries armed in "legionary" fashion at Corbridge? (d) ... [fill in your own suggestion here!]

[Praefectusclassis]

A nice example of the puzzling question about what we don't know and related to TJ's question is the one about the size of the Imperial army. There are two references - a vague one from Tacitus about the year 23 AD - and a very precise number from the sixth century historian Jordanes about the army just before or just after (debated, of course) the 3rd century crisis. We have a reasonable idea for the number of legions for most of the time - give or take a legion that is missing for a few years - but we can only count auxiliary units mentioned in diplomas, inscriptions and papyri and get some idea of their totals. Most of the units known to us can be placed in a certain province for some time, but there are huge holes in the available evidence, especially before diplomas get fairly common in the Flavian era. How does one start to describe the movements of the Roman army if you only have a basic idea of its makeup (which is why the new Ospreys about the Order of Battle for the Roman army for certain eras are a bit silly imho...).

[D B Campbell]

... which is why the new Ospreys about the Order of Battle for the Roman army for certain eras are a bit silly imho...

Agreed, Jasper. I was originally approached to do the first one a couple of years ago: I said it couldn't be done. Sure enough, the Roman ones bear no resemblance whatsoever to the blueprint followed for other periods.

[Gaius Julius Caesar]

Just out of curiosity, Chesters Fort Museum is crammed with alters, mile stones, and stele.
Is there a record of the different units that were based there from this evidence, aside from (I think) the XXth Legion, which I think was more from wall milestones?

[PhilusEstilius]

Byron. There is of course a bit of a problem with what is in Chesters Museum for the John Clayton memorial collection covers many sites both on the Stanegate line and also the Wall forts. There are a couple of grave stele around that may give a little more knowledge about Chesters however these have been hidden in the vaults of the Antquities Museum for years now.

[D B Campbell]

Chesters Fort Museum is crammed with alters, mile stones, and stele.

Used to be my favourite museum -- fantastic old fashioned wooden cabinets and row upon row of sculptures -- but I think they recently "renovated" the place.

I bet Budge's Account of the Roman antiquities preserved in the museum at Chesters is well worth twenty quid.

[PhilusEstilius]

Duncan. I used to work at Chesters as a seasonal custodian many years ago and I used to have to clean those cases every morning, there has been renovation but thank heaven they have not changed the building.