02 April, 2006

Roman water infrastructure in post-Roman Britain. Part 2: Post-Roman survival

Part 1 listed some examples of sophisticated Roman water engineering in British towns and fortresses. How much of this infrastructure survived into the post-Roman period and for how long?

There is evidence that at least some water supply systems were maintained very late in, or after, the Roman period. For example, a timber-lined water trench was constructed in York some time in the fourth century AD (Ottaway 2004), and in St Albans (Roman Verulamium) a new water main had been laid later than the fourth-century Roman levels, probably some time in the fifth century (I can’t find a link online and haven’t got the book to hand, but this Amazon review mentions it). The Roman sewer system excavated at Church Street, York, is considered to have been used until the late 4th or early 5th century, on the basis of pottery finds (Buckland 1976).

Even when maintenance had ceased, some of the infrastructure may well have remained operational for a considerable time. Complex pumping devices such as the London bucket-chain system (see part 1) would probably have broken down fairly quickly without regular maintenance and repair. Wells, however, could have been used until they became contaminated or blocked; it seems logical to me that common sense prompts people to continue using existing infrastructure as long as it still works and is still useful. There is a suggestion that the seventh-century monastery founded by St Hild at Whitby on the Yorkshire coast may have utilised the well from a Roman signal station on the same site (Mundahl-Harris 1997), though no evidence is quoted for or against. (No definite trace of a Roman signal station on Whitby cliffs has been found, apart from a few Roman coins from an excavation in the 1930s, but as the two known signal stations either side are out of sight of each other, I think it very likely that there was one in between, and it may well have been at Whitby).

Sophisticated aqueducts such as those serving Lugdunum in Gaul (modern Lyon; see part 1) would fail as soon as one of the closed siphons leaked or an arch collapsed. There is less to go wrong with a simple gravity-fed aqueduct consisting of a pipe in a trench laid along natural contours. Such systems could have functioned for a long period until they fractured, collapsed or silted up. Wooden water pipes (bored-out logs, like the ones discovered at Colchester) supplying spring water to Vindolanda fort on Hadrian’s Wall were still working almost 2000 years after they were laid, efficiently enough to flood the archaeological excavation that discovered them.

One item of documentary evidence can be added to the archaeological evidence. In his Life of St Cuthbert, Bede describes a visit by St Cuthbert and the then queen of Northumbria to Lugubalia (modern Carlisle) in 685 AD, where the local citizens showed them the town walls and ‘a remarkable fountain, formerly built by the Romans’ (Chapter 27).

This suggests to me that Carlisle probably had a gravity-fed aqueduct that was still working and still serving a street fountain (or possibly a fountain in what had once been a private house or a baths complex), and that the local inhabitants knew about it and were proud of it. Perhaps they still used it.

I think it is also noteworthy that Bede and/or his informant knew who built the fountain, which suggests that some knowledge of Britain’s Roman heritage was still extant when Bede was writing in the 730s. It is sometimes claimed, usually on the basis of the poem The Ruin (from the 10th-century Exeter Book manuscript), that the early English (‘Anglo-Saxons’) were ignorant of the Roman past and superstitiously believed that Roman buildings had been made by giants. Very possibly some people did; but Bede’s mention of the fountain at Carlisle indicates that this was not always the case. It can be argued that knowledge of the Romans and their architecture was reintroduced from Continental records by St Augustine’s missionaries from the Roman Catholic church and that this was the source of Bede’s knowledge. However, it should be noted that Bede did not have accurate information on the building of the Antonine Wall and Hadrian’s Wall; in his Ecclesiastical History he follows Gildas’ erroneous assertion that the Britons built the Antonine Wall and that the Romans then built Hadrian’s Wall and left immediately, leaving the Britons to man it (Book I Ch. 12). It seems unlikely that records reintroduced from the Continent by the Roman church would have known the origin of a fountain in Carlisle but not known who built the two major defensive structures in northern Britain. It seems to me that some continuity of local tradition is a simpler explanation than loss and reintroduction of information.

I interpret the information as follows: Towns and military sites in Roman Britain were provided with engineered water supply systems, which were maintained on at least some sites up to or after the end of the Roman period. Some of these systems, gravity-fed underground aqueducts in particular, were sufficiently simple and robust that they could have continued to function for a considerable time even after maintenance had ceased. The underground water pipes at Vindolanda were still working after nearly 2000 years. Bede’s mention of the fountain in Carlisle is consistent with a working gravity-fed aqueduct there in 685 AD. This was probably unusual, as Bede thought it was worthy of mention and the locals thought it was worth showing off to visiting dignitaries, but it was not necessarily unique; the Vindolanda water pipes show that much longer survival was possible. The origin of the fountain was known to Bede and/or Bede’s informant, although there is no indication as to whether anyone still understood how it worked. It is possible, and may even be probable, that similar systems survived in other Roman towns and forts, and that they continued to be used as a source of fresh water by populations living on those sites for a considerable period, perhaps centuries. (How large those populations were, and how long they stayed, is a separate question). It’s often assumed that a major reason for people to continue living in or near Roman towns and forts was the presence of surviving defensive fortifications. I wonder if the presence of a surviving water supply may also have played some role?

What do you think?

Bede, Life of St Cuthbert. Translation available online.
Bede, Ecclesiastical History of the English People. Translated by Leo Sherley-Price, Penguin Classics, 1968.
Buckland PC, The environmental evidence from the Church Street Roman sewer system. York Archaeological Trust, 1976.
Mundahl-Harris S, St Hilda and her times. Caedmon Press, Whitby, 1997.
Ottaway P. Roman York. Tempus Publishing, 2004.


Rick said...

"Active" systems such as waterwheels would fail as soon as maintenance lapsed. Raised wooden structures probably wouldn't last very long. Stone archways could last a very long time indeed, since the Romans really over-engineered those things. Doesn't Segovia still get part of its water supply from the Roman aqueduct?

The two big killers of aqueducts would be enemy action and silt. If you are besieging a town, breaking its aqueducts is an obvious temptation. Silt is the insidious killer. The Romans built settling tanks to slow it down, but how long an aqueduct would last without active effort to desilt it probably depends on the source of the water, i.e., how much silt it carries in the first place.

On the other hand, the Anglo-Saxons weren't stupid, and you don't have to know how to build an aqueduct to figure out to send guys with shovels to clear it out.

Memories of the Romans - now there's an interesting question, that I am wholly unqualified to answer. In the earlier period, I imagine the amount of memory varied widely, according to the local circumstances.

At least in the once more Romanized regions, the Britons must have continued to think of themselves as Romans long after 410 - the name Ambrosius Aurelianus is surely indicative. But if, in a particular locality, the Anglo-Saxons killed or drove out the native elite, knowledge of the Romans was probably lost with them - the peasants likely didn't know much, and the new overlords likely weren't asking them.

Flip side, though, I seem to recall that there are indications that the transition was fairly peaceful in some regions. Doesn't the West Saxon king list begin with suspiciously British-sounding names? If there was some social continuity, some knowledge of the Roman past was more likely to survive. For that matter, aren't there some indications that "Saxons" were already in Britain as Roman auxiliaries? If so, their descendents might well preserve some tradition of their service.

I didn't realize "The Ruin" was so late. Or is that only when it survived in manuscript? If the poem itself is late, it may be more a romanticization of heroic ancestors than a reflection of what Anglo-Saxons at any period actually knew or believed.

Gabriele C. said...

Isn't Vindolanda teh coolest place evah? *grin*

I knew I picked a good one when I decided to make it the main fort in Storm over Hadrian's Wall (I need a better title for that one since the wall doesn't begin to get built until the end of the book). I'm so going to sneak the hospital water pipes in, after all, I do have scenes taking place there. :)

Thanks for sharing that info, Carla.

Bernita said...

All makes sense to me, Carla.
But then, you always do.

Wonder if the usual attack technique was poisoning the well/water source - rather than destroying it - dead animal type of thing - since one might wish to utilize the resource later.

Carla said...

I don't know about Segovia but it sounds quite possible. I remember reading somewhere that the celebrated Trevi fountain in Rome gets its water from a Roman aqueduct.

If the aqueduct was a simple buried pipe it might be quite hard for enemies to find. The aqueducts I've found descriptions of in Britain from excavation reports have been much smaller than the big Continental aqueducts, more like a big pipe than a 6-foot high stone/concrete channel that you could walk through. If that was the norm, perhaps because the volume of water required wasn't as great as in Continental cities or because there were alternative supplies like wells, they might have been less of a target for enemy action. Does beg the question of how they were cleaned out, doesn't it? Upland streams often don't contain very much silt, and also if the water flows fast enough the pipe will stay clear by itself. I should think that must be how the Vindolanda pipes worked; certainly nobody had cleaned them out for many centuries and they were still carrying water. According to the DynoRod engineer who unblocked our drains at our previous house, the number one prime cause of blockage isn't mud/grit etc, which he said tends to wash through, but fat, because it's sticky.

The degree of continuity, or otherwise, from Roman administration to the early English kingdoms seems to be a subject of fierce academic debate. with much heat on both sides but not an awful lot of light. I agree with you that it would likely have been extremely variable from one locale to another. (This is one of the things I pick up on in the 7th-C-set books).

The Ruin is in the Exeter Book, along with the riddles, and the manuscript is usually stated to be 10th-C. The poem might be much older; I daresay a linguistic expert could tell you the likely date from the language patterns, but I don't know it. Though at any date it could have been a poetic contrivance - 'even the work of giants is brought low' sort of thing. It always reminds me of the Ozymandias King of Kings poem.

Gabriele - if I were setting a scene in Vindolanda at the appropriate time I'd be so tempted to include a plumber joke as an aside. "[sucks teeth] What Brittunculi son of a donkey put this in, guv'nor? It'll never last."

And I'd stick with the title as is; at least people have heard of Hadrian's Wall!

Bernita - that would certainly be a possible course of action, providing the enemy could find the source and get to it. Possibly also has the advantage that the garrison may not know you've done it until they start falling sick, by which time it's too late for them to use an alternative source even if one exists.

I don't know whether sieges were a big part of military activity in Britain? One would imagine so because the later forts are clearly built to be defended, with projecting towers to enfilade along the walls (early forts are designed more as bases, from which the army goes out to fight). The Multangular Towers at York are supposed to have had artillery platforms on the top. But the only armies that had siege equipment were surely other Roman ones? Maybe sieges were a feature of the civil wars resulting from the Empire coming to put down assorted British army usurpers (there were certainly a lot of them). In which case, the Empire army would probably not want to trash the infrastructure too badly, since they'd be trying to take the province back under central control.

Gabriele C. said...

Hehe, Carla, maybe something about Horatius sitting on the loo and in vain looking for a nice sponge on stick and muttering about the moss he has to use in that bloody place of a barbarian country, and it was all his brother's fault anyway. :)

Bernita said...

Finding the source would not be a problem if it lay outside the fortifications of the burgh or fort.
Preliminary reconnaissance or some judicious interrogation of traders, travellers or unfortunate peasants would divulve it, I would think.

Rick said...

True that a buried pipe would be very hard to find - even the big walk-through channels, when underground, wouldn't be noticeable once grass grew over the filled-in trench. They had access points at regular intervals, but even these aren't obtrusive; hardly anyone but an aqueduct engineer would know what to look for.

The comparison of "The Ruin" to Ozymandias is well-taken, and it is a small step from "there were giants in those times" to a sense of literal giants - think of the outsized grave claimed to be Arthur's at Glastonbury.

You're surely right that only the Romans had siege equipment, but that doesn't preclude sieges by others - strictly speaking, the purpose of siege equipment is to shorten a siege by allowing an assault. If you don't have it, the alternative is to starve the place out, all the more reason to destroy the water supply. To be sure, holding an army together long enough to starve a place out was very difficult, with great risk that disease or disaffection would produce a debacle for the besiegers.

Gabriele C. said...

Well, the Goths did lay siege to Rome quite successfully. In 408 AD they got paid off, in 410 they got inside. *grin*

Arthur's grave in Glastonbury was a very convenient find at a time when King Henry II of England needed some mythical ancestor that connected his lineage to the land. He might well have had a hand in that discovery - probably a hand full of gold or holding a charte donating the abbey more land.

Geoffrey de Monmouth's Historia Regum Brittaniae was written as sort of pro-Plantagenet propaganda. "...the creation of an English historical mythology in service and interest of the Norman conquerers who were not rooted in the indigenous traditions - it was necessary to establish a concept of history that made the Norman rulers the fulfillers of British history, and at the same times served as a concept of national identification by setting up an oppostion to the kingdom of France and the French epics."

Mertens, Volker. Artus. In: V. Mertens and U. Müller (eds). Epische Stoffe des Mittelalters; p. 292 (my translation) - and not the only thesis I found about that matter.

I see another blogpost coming, but give me a few days. :-)

Carla said...

Good point about siege equipment, Rick. I was thinking about the possible reasons why fort design might have changed and not writing terribly clearly. There's some evidence of sieges involving non-Roman armies; Bede mentions that Osric (briefly king of Deira in 633 AD) besieged Catwallaun of Gwynedd in a strong city. The (legendary?) Mount Badon is called a siege, and I think there's a tradition of a siege of Din Eidyn (modern Edinburgh) some time in the late 6th C. Unlikely that any of those involved siege equipment.

There are lots of traditions involving giants in English and Brittonic folklore - e.g. Robin Hood's Stride is a rock formation in Derbyshire that only a giant could stride across. Arthur fights giants in Culhwch and Olwen, Cai is sometimes described as a giant, the Wrekin is supposed to be a heap of rocks thrown down by a giant, the ice-carved Hole of Horcum in Yorkshire is named after a local folklore giant called Horcum, and Cadair Idris in Wales is supposed to be named for a giant called Idris. I think there's an outsized grave slab in Herefordshire that's supposed to be the grave of Arthur's son Amr. And so on.

Arthur was indeed a convenient target for Norman-Plantagenet (and later Tudor) 'spin', which is probably one of the reasons why the dominance of the legend annoys me :-)

Alex Bordessa said...

I can't remember which Alfred Duggan book it's in (I'm sure it's one of his - the sense of humour is so like him) but there was a scene about a siege. The sieger and the besieged shouted rude words at each other until quite satisfied they'd insulted each other enough. It was funny, honestly, the way Duggan told it ...

Yes! Carla you mentioned Verulam :-)

Rick said...

I seem to recall that in fact the Goths slipped into Rome via the aqueduct channels.

Arthur has often been politically useful! Arthuriana in modern times dates more or less back to Tennyson, I think, who also had an implicit political message, I imagine, of British nationalism.

But in spite of Henry VII naming his first son after him, Arthur went into something of a decline in Renaissance and early-modern times. I don't think he figured much in Elizabethan literature - The Faerie Queen has some kind of Arthurian connection, IIRC (I've never read it), but Shakespeare and his stage contemporaries didn't make much if any use of him. All in all there's a gap in major Arthuriana from Malory till Tennyson.

Carla said...

There's a scene in Count Belisarius by Robert Graves where the Goths try to get in via the aqueducts but are forestalled by Belisarius, and another where Belisarius' men take Naples by means of the aqueducts.

Could it be Conscience of the King, Alex? I've not read that one but the scene sounds as if it might fit Cerdic.

Alex Bordessa said...

You're right Carla. It could well be Cerdic. It's been a couple of years since I read Conscience ..., so I've a good excuse to go back to check :-)

Gabriele C. said...

There are several accounts how the Goths got into Rome in 410AD. King Alaric offering noble young Goths as hostages who then opened a gate is one, a poor old widow sick of war opening it another. A traitor showing them the way through an aquaeduct or the canalization would fit, all the sources agree on one thing: it must have been treason, because how else could the Goths get inside a city as strong as Rome. Those sources tent to forget that the defenses were in decline after the emperor moved to Ravenna.

I'm going to have my own version of it. ;-)

Gabriele C. said...


Argh, I want an edit function.

Carla said...

I thought you might deal with Alaric and Rome at some point, Gabriele :-) I take it that would be in Towards the Kingdom of Tolosa? The nice thing about there being several versions is that you can pick whatever suits your story best, or make up something different :-)

Gabriele C. said...

Yes, it's Towards the Kingdom of Tolosa. I've had the Visigoths on my brain since 2003 thanks to a History Channel report, and some emotive German ballad about Alaric being buried under the Busento river which was led into a different channel for some time. No proof for that one, but it's fun. If it's true, the Goths would have needed a Roman engineer, or someone with appropriate training, to achieve it.

TtKoT begins in 408 at the border of Norium (Vindobona) when the Goths negotiate with Stilicho who then is assassinated by Honorius, which leads to the first siege of Rome in summer 408; and goes on until the murder of King Athaulf near Toulouse in 415.

Sarah Cuthbertson said...

I've enjoyed reading these posts and everyone's comments. Roman water engineering fascinates me. Just a wee boast: I was labouring in the next door trench at Vindolanda the summer those pipes were found. I was teasing away at a piece of timber door that had (possibly) been re-used as flooring. Being a rank amateur, I was surprised that Andy Birley let me near something so fragile but I noticed him keeping a beady eye on me!

Rick said...

I seem to recall a tradition (or chronicler's claim) that Alaric was buried in a riverbed, and after it was diverted back all the workers were killed so that no one would know just where he lay.

The Goths were pretty Romanized by this time - Stilicho an obvious example. In fact, I've been told that by this time they were only "Germanic tribes" in the vaguest sense, due to absorbing deserters, runaway peasants, and the like. The Alamanni (sp?) are indicative; whatever the correct spelling, it is recognizably "all-men."

If a Goth could be chief minister of the west, there might have been an engineer or two among them as well.

Gabriele C. said...

Definitely, Rick. My Goths, at least the 'upper class', are quite Romanized. King Alareiks had spent some time in Byzantium, I bet he spoke Latin and Greek pretty well and was literate. So does my fictive character Alamir.

Stilicho was a Vandal, btw. :)

Carla said...

That must have been quite a thrill, Sarah.

Rick/Gabriele - I hadn't heard that tradition about Alaric being buried under a diverted river, but it's a good story (I wonder if it might originate from a natural shift in a river that changed course to flow over the (reputed?) site of his grave?).