09 June, 2009

Chester in the seventh century: surviving infrastructure

Modern Chester was founded in around 74 AD as the Roman legionary fortress of Deva, later acquiring part of the name of the Twentieth Legion to become Deva Victrix. It clearly had a large number of impressive Roman buildings, perhaps more impressive than most Roman cities in Britain. Deva was 20% bigger than the other legionary fortresses in Roman Britain (e.g. Eboracum, modern York), and contained the enigmatic Elliptical Building, so far unique in the Empire. The purpose of the Elliptical Building remains unknown, but it must have been an impressive structure in its day. Chester also had the usual components of a legionary fortress, including a headquarters building (principia), smart houses for the commander (praetorium) and senior officers, amphitheatre, stone defensive walls and a main baths building (thermae), not to mention a large harbour and a bridge crossing the River Dee. How much of this was still standing in the seventh century, and can we tell if people were still using it for anything?



Ranulph Higden, a monk at the abbey of St Werburgh in Chester (later the cathedral) wrote a description of Chester in the mid fourteenth century. He described underground passages (the sewers), huge stones inscribed with the names of ancient men (tombstones, or perhaps also other monuments or inscriptions?), and vaulted dining rooms (perhaps parts of the main bath building [thermae]) (quoted in Mason 2001). Evidently substantial parts of the Roman infrastructure were still standing at this date, getting on for a thousand years after Roman government came to an end in Britain. Say what you like about the Romans, they built to last.


Headquarters building
In the headquarters building, the rooms along the rear of the cross-hall were refloored several times during the fourth century, and the room west of the shrine-room (aedes) was converted into a secondary shrine. Mason (2001) gives no date for these repairs, but since there were several they presumably span quite a long period of time and indicate regular use and maintenance in most of the fourth century, if not later.

Elliptical Building
In the baths suite attached to the Elliptical Building, a new doorway was inserted and the mortar bedding for its timber door sill produced a find of 24 coins dated the reign of Emperors Valens and Valentinian (364-75 AD). The hypocaust was rebuilt, and part of the suspended floor was found still intact when excavated in 1969. Gold-working crucibles were found at the north end of the building, together with a gold solidus of the Emperor Magnentius (350-353 AD).

Coins of Theodosius I (379-395) and Arcadius (395-408) have been found in Chester, but no coins of Honorius (became Emperor in 408).

Excavation in the centre of the fortress has shown that there was no extensive complex of post-Roman timber buildings as at Wroxeter in this area. Possibly such structures existed elsewhere in the fortress and have not been discovered (or not recognised), but there is no evidence for them. David Mason states that the various timber buildings identified in Chester on various digs so far are now thought to belong to the Anglo-Scandinavian town of the ninth and tenth centuries.

Not much is known about the Roman bridge, except that it was on more or less the same site as the current Old Dee Bridge, which was built in the medieval period. It is not known how long the Roman bridge stood. The location of its replacement on the same site may indicate that the Roman bridge remained standing and in use long enough for the street plan of Anglo-Scandinavian and medieval Chester to become fixed and so dictate the position of the crossing. Or it may reflect the constraints of geology/geography, for example if it happens to be the only sensible place in the vicinity to build a bridge.

Baths complex
The thermae courtyard was repaired and resurfaced throughout the fourth century.

When the thermae complex was destroyed by development in 1964, parts of the walls were still standing in situ to a height of up to 13 feet (4 m), hypocausts and mosaic floors were still intact, and large sections of collapsed roofing vaults (barrel-vaulted concrete, estimated to have stood 53 feet above floor level) lay on the floors. A layer of “dark earth” containing charcoal and bits of animal bone had accumulated to a depth of 1 foot (30 cm) over the tepidarium floor, implying a considerable period of residential occupation. It is not known when the roof vault collapsed. However, if the “vaulted dining rooms” mentioned by Ranulph Higden refer to the vaulted and decorated bath complex, then parts of the building were still standing and still roofed in the middle of the fourteenth century.

The harbour and the rest of the Dee estuary downstream of Chester has been slowly silting up since the end of the Ice Age, and the harbour is now the low-lying dry land of Roodee racecourse. In the Roman perod it was a busy harbour and may have been the base for part of the Roman naval fleet. Shipping is recorded as having trouble getting upstream to Chester in the late medieval period, according to Mason. He also says that the Anglo-Scandinavian town of the ninth and tenth centuries relied on trade, and that the street frontages were cleared of Roman rubble because they were the most valued as commercial premises. If this is correct, it implies that the harbour was still capable of taking trading shipping at a useful volume, at least for shallow-draughted ships like those used by the Norsemen, until at least the tenth century. The harbour would presumably therefore also have been similarly functional in the intervening period.

Place name

It’s worth noting that there is no trace of the Roman name Deva Victrix in the modern name of Chester or in the names Bede knew for the city in when he wrote his Ecclesiastical History in 731:

…which the English call Legacaster and the British more correctly call Caerlegion…
--Bede, Ecclesiastical History of the English People, Book II Ch. 2

Bede was writing in Latin and called the city Urbs Legionis. This translates as “City of the Legion”, as do the English and British names he quotes, so the three names are really the same name in three languages.

This contrasts with some other Roman fortresses and cities, such as Winchester, Lincoln, York, Londinium, Wroxeter and Carlisle, where elements of the Roman name can be traced in the modern name. In turn, this may indicate that whatever Chester was used for after the Romans left, and whoever was using it, the Roman name was either discarded or lost.


The coin evidence may suggest that regular Roman Army troops were still stationed in Chester, getting paid in Imperial coinage and repairing and maintaining the fortress buildings up to about the end of the fourth century, but not during the reign of Honorius in the early fifth. It is of course possible that coins of Honorius circulated in Chester and that none has yet happened to be discovered by archaeology. However, given that the usurper Emperor Constantine III invaded Continental Europe from Britain in 407, presumably with an army, it is quite plausible that he took Chester’s garrison with him.

Chester hosted a major synod in 601 (Annales Cambriae), which is probably the same event as the meeting recorded by Bede in 603 or 604 and attended by seven bishops and “many very learned men” (Bede Book II Ch. 2). The synod and the possible ecclesiastical importance of Chester was discussed in an earlier post. It also has implications for the state of the surviving infrastructure. Seven bishops probably each brought a sizeable retinue, as no doubt did Augustine of Canterbury. If Chester was the site, this implies it was:
(a) an important and prestigious place, suitable for hosting a gathering of VIPs;
(b) had enough infrastructure to cope with them and their retinues in suitable style;
(c) was sufficiently well-connected to transport networks that a large number of people could be expected to travel to it;
(d) perhaps that it had some connection with the Brittonic Christian church or an important official thereof (perhaps a Bishop of Chester, as discussed earlier.

Chester in the seventh century probably still had the following infrastructure:

  • A fortress wall (possibly not fully intact – see earlier post);

  • A functioning harbour, at least for shallow-draughted vessels;

  • A bridge crossing the River Dee;

  • Part or all of the main baths complex (thermae), still standing and still roofed;

  • Probably some or all of the similarly robust structures, such as the headquarters building and perhaps the Elliptical Buildng, were also still standing and roofed, since they had been kept in good repair up to at least 400 AD and would have taken some time before they fell down;

  • Roman streets and roads on the alignments still in use today (and maybe some others that have since been lost);

  • A population living in and/or around the fortress, probably at a low density

The location of the synod suggests that the Chester may have been a centre of ecclesiastical power, as has been suggested for Wroxeter. It may also have been a centre of secular power, or the secular power (i.e. the king or local sub-king) may have been based elsewhere, again as has been suggested for Wroxeter.

Chester’s Roman name seems to have gone out of use some time before 731, since Bede knew the city as “City of the Legion” in three languages and not as Deva, Deva Victrix, or any derivation thereof . This isn’t universal for ex-Roman cities in Britain, since Bede knew his local city (modern York) by its Roman name of Eboracum. Nor does it just reflect the limits of Bede’s knowledge, since whoever wrote Annales Cambriae also knew Chester as “City of the Legion”, not as Deva.

The absence of Chester’s Roman name by 731 may suggest either a temporary period of abandonment during which the name was forgotten, or a deliberate decision not to use the Roman name. I would lean towards the deliberate decision, since it seems unlikely that Latin-speaking literate Brittonic churchmen would not have been perfectly capable of reading the city name off milestones or inscriptions if they wanted to, even if all other records had somehow been lost. Perhaps the local population had always called the fortress “the city of the legion”, reflecting its military function, in the same sort of way as the Gaelic name for modern Fort William is An Gearasdan, “the garrison”.

Mason DJP. Roman Chester: city of the eagles. Tempus, 2001, ISBN 978- 0-7524-1922-0.
Bede. Ecclesiastical History of the English People. Translated by Leo Sherley-Price, Penguin Classics, 1968, ISBN 0-14-044565-X

Google Maps links
Fort William


Rick said...

Yes, the Romans built to last! The ruins we see are often not because the buildings simply fell into ruin but because stone was 'recycled' for later construction.

Regarding the name, I'd lean toward your guess, that 'City of the Legion' became the everyday name while the legion was still around.

I wonder how all those coins ended up in the mortar under a door sill?

Carla said...

Good question! Possibly some sort of votive offering, or maybe some unlucky individual just dropped his purse into the cement mixer. Your guess is as good as mine :-)

Robbed-out wall trenches are a standard find in excavations, where all the stone has been nicked. Quite often to build the medieval cathedral or castle next door.

Rick said...

Two dozen coins sounds like quite an offering! I picture someone swearing up a storm as they watch their money sink into the concrete mix.

Carla said...

That's the picture that came into my mind as well :-)

Rick said...

The concrete mix also swallowed the discussion! But a couple of other points that stand out are the 'enigmatic' Elliptical Building itself, and the Dee bridge. I take it that little or nothing is known about the original purpose of the building?

As for the bridge, is anything known except that there was one? I wonder if the medieval bridge might have been rebuilt on the existing piers.

Carla said...

There are quite a few theories about the purpose of the Elliptical Building - or purposes, since it was started in the 1st century, then abandoned, then rebuilt to a modified plan in the 3rd century. The one I find most convincing is that it was some sort of cross between a temple and a sort of monument to Roman power designed to overawe and impress visitors. There's a theory that Chester was originally intended to be the seat of the Roman governor of Britannia (if Agricola's campaigns in Scotland had stuck, it would have been quite central, and as an important naval base would also have been handy if there really were dreams of conquering Ireland), which could account for it being 20% bigger than usual and also account for it having this highly peculiar and presumably ceremonial building. But no-one really knows.

Re the bridge, I haven't found much about it. David Mason's book doesn't make any suggestion of the medieval bridge being built on Roman piers. He says little is known of the Roman bridge. British History Online says that bits of the Roman bridge, including pier bases and cornice fragments, were found in 2000 scattered on the river bed a few metres downstream of the medieval bridge. I haven't found any more information about that so far, but assuming the interpretation is correct and the Roman piers are in the river, they presumably aren't holding up the medieval bridge. That's not to say that there couldn't have been an earlier post-Roman bridge rebuilt on the Roman piers, if the Roman piers survived longer than the Roman spans. Depends what did for the Roman bridge, whether the river ate away the foundations of one of the piers and/or abutments and caused a collapse, or whether one or more of the arches fell in.

Rick said...

Interesting about Chester having been possibly intended as a provincial capital - I'm so accustomed (surprise) to London being the 'natural' primary city. Chester would be rather central to the whole British Isles.

What makes the Elliptical Building unique in the empire? Its elliptical ground plan? The first thing I thought of was the Colosseum, which however is not an enclosed building, needless to say. The second thing I thought is that it followed the ground plan of a pre-Roman building, but that would be really, really un-Roman.

And yes, if the Roman bridge piers are scattered in the river, that counter-indicates them holding up a later bridge. :-) And failure of a pier would be just the thing to require total replacement of the bridge.

Carla said...

I think it's the ground plan and the fact that nobody has yet managed to come up with any functional purpose for the building - it's the wrong size and shape to be an arena or a theatre or an exercise hall or a training school or a market or guest accommodation or whatever. Which probably accounts for falling back on the reliable if-all-else-fails explanation of 'ritual', or in this case 'ceremonial'. It's not easy to reconstruct a building just from the ground plan, so it may be that we just haven't understood it properly so far.

I don't think there's any evidence for a pre-Roman building on the site, or even whether there was anything much at Chester before the Romans arrived, so I don't think it's likely to have started life as a squashed roundhouse :-) Though all things are possible.

It's unlikely Britain had anything resembling a primary city before the Roman period. Some of the tribes that later became civitates had pre-Roman centres, like Camulodunum (modern Colchester) for the Trinovantes, which may have been a bit like the oppida in Gaul on a smaller scale. But those go with the tribe, and as no one tribe dominated the island there was no natural 'centre' as such. So the Romans could have invented a 'capital' at Chester as easily as London, I suppose.

I rather think a major trading centre would have evolved somewhere in the south-east, though, just because of the logistics of cross-Channel and cross-North Sea trade with the rest of the continent. Possibly there'd have been a political capital at Chester and an economic/commercial capital at or near London, rather like DC and New York, instead of both being in the same place. Interesting to speculate :-)

Rick said...

A quick google makes the Elliptical Building even more puzzling, since it was started in the early days of Roman rule, then abandoned for 150 years, then rebuilt and completed. There's speculation that it was meant as a symbolic model of the Roman world, but why in a frontier legionary fortress, even an important one?

Possibly there'd have been a political capital at Chester and an economic/commercial capital at or near London

Which quite by accident is the situation I created for my Lyonesse. The royal capital, Kelliwick*, is inland and well to the west - allowing for considerable differences of geography it rather corresponds to Wroxeter. (Also mostly coincidence!) But the main commercial city of Rosemouth corresponds to London.

* There was probably a real Kelliwic, but in Cornwall.

Carla said...

Good question, again! That's probably why no-one is really sure what it is yet. It might make some sort of sense at the first building date, since Vespasian was Emperor and had made quite a bit of his career in Britain - he might have wanted to put a flashy building in the provincial capital to celebrate his achievements there, or something.

Celliwig in Cornwall features as Arthur's court in the Triads, if I remember rightly. A pair of important cities has advantages for narrative purposes, in that you can have sub-plots (in the narrative and political sense) going on separately in each of them if you want. Why Rosemouth? Is the river the River Rose?

Rick said...

Celliwig means 'forest glade,' and I stumbled on a suggestion that Cornwall might have been a later misreading, since a couple of places in Wales had forms of the name.

You guessed correctly; Rosemouth is at the mouth of the river Rose. One of the things I wonder about is the implications for the 'future history' of Lyonesse of a national capital separate from the main cosmopolitan center.

Tangential to this, but something I've had to shake off, with incomplete success, is the presumption of kings living in 'a palace/castle.' Even as late as the 16th century this is anachronistic; royal courts were not as itinerant as in medieval times, but still relocated regularly to let palaces air out, so to speak.

Carla said...

I don't know about Celliwig specifically, but topographical names often do get duplicated, so it's entirely likely that there were several places of the same name. If it means 'forest glade' then some of the Old English -ley names or Norse -thwaite names might also be candidates, if they happen to have been translated by a bilingual local. I'd have guessed it was something to do with a hermitage from the 'Celli' element, which I've seen in Welsh place names in the form 'gell' and which apparently means a monk's cell.

It would be interesting to play with that as a concept. Some countries are much less centralised than Britain/England, e.g. France has major regional cities like Lyon and Marseille. Lyonesse might have developed more along those lines, perhaps. Has it got distinct regional identities and/or power blocs, like duchies?

As long as you relocate the court now and again or send the queen on progress you'll likely get away with it. Most of your readers will share the same assumption, after all :-) If the court does tend to stay in one or a few centres, that has implications for transport (clearly very efficient, since goods can be brought to the court) and possibly also for regional politics (while the cat's away the mice will play). Conversely, if the queen regularly shuttles between Celliwig and Rosemouth she can keep a sharper eye on the local magnates than if she were based at a single major city.

Doug said...

Since WE know that the Romans called Chester "Deva", it is difficult to see how anybody in the past would not have had some way to find out. A deliberate avoidance is almost certain. I wonder whether it might be that the Roman pronunciation of Deva (Dee-wa) was a bit close to the Brythonic word for God?

Rick said...

In fact Lyonesse does have duchies - it is synologous to Britain as a whole, not just England (the 'Saxon Pale').

And indeed my girl will be going on progresses. She's well aware that her grandfather did so far too rarely, with unfortunate effects on effective royal authority.

A minor coincidence, given the setting of your latest post: Had Catherine's father lived he would have been Henry III of Lyonesse.

Carla said...

Doug - Wasn't "Deva" named after the Brittonic name of the river Dee, which in turn meant "goddess"? No-one seems to have minded keeping the river name, but maybe a fortress was different.

It's possible for information to be lost and then rediscovered, e.g. we know when Hadrian's Wall was built and on whose orders, owing to some inscriptions, but Gildas gives a completely different and factually incorrect account of its building. That might be through deliberate choice, or it might genuinely be that he didn't know the details and/or thought they didn't matter. If he lived in the south as is usually thought, he presumably wouldn't have been able to visit the Wall and read the inscriptions even if he wanted to and even if they were still extant. That's less likely to apply to Chester since it was used for the Synod in 601.

Rick - Are Celliwig and Rosemouth both in one region, and if so, do the others have their own centres?

Rick said...

Maybe just naming the city/fortress for the river seemed ambiguous to locals, or the legion impressed them more.

Rosemouth is in the Saxon Pale (which no longer has a separate legal identity, but remains familiar usage). Kelliwick, as I spell it, is on the border between the Pale and the West Country, which is in effect a bigger Cornwall. And the duchies do indeed have ducal seats: Hampstead-on-Ryde for Ashland, Talgarth for Prydeland, and Enniskillen for Tearnac. There's also a duchy of Dunfolk in the far north, but I've not had occasion to name its ducal seat. And, though not a duchy, Halverstrand for the Isles. Which in spite of connotation are not a counterpart of Scotland's Western Isles, but mega Channel Islands.

Carla said...

My guess would be that the legion impressed them more. It would have had a tremendous impact on the local economy for the best part of three centuries, for a start.

That would perhaps put Kelliwick nearer to Gloucester or Bristol than Wroxeter. Interesting that it's a border town. What's the history behind that?

Rick said...

Kelliwick Tor is defensible and in a strategic position, defending the West Country against first the invasion of Emperor Theodosian and then the Saxons.

Arthur gives terms to the Saxons that grant them the original Pale, with the border a ways east of Kelliwick. After his death there's a generation of turmoil, then Cedric establishes himself as king and grants the Prythonic nobles essentially the same terms, except that the area around Kelliwick, later Kingshire, is naturally annexed to the Pale.

The west gate of Kelliwick, by the way, is Hammersgate (think Emrys).

My background history is surprisingly un-worked out, but you can probably guess that Theodosian is a mashup of Theodosius and Justinian. :-) I have never named syno-Rome. For that matter I never named the syno-Seine, even though most of the first book happens practically on its banks. It was enough to speak of the River, and the South Bank!

Doug said...

I thought I'd spotted an interesting coincidence that "Deva" sounded like the Brythonic word for "God", I should have suspected that that was its source! I rather like Rick's suggestion that the locals were impressed by the legion; the towns which kept a variant on the Roman name seem to be those without a resident legion (I'm open to correction, although I'm encouraged by York not being "Everchester").
But I now find (Wikipedia) that an alternative Welsh name for Chester in the 12th century was "Deverdoeu" so the "Deva" was not lost early. I wonder why the name went in two directions?

Carla said...

Rick - you only need enough history to support the story you're telling. Otherwise you'll disappear in a cloud of footnotes and never be seen again :-)

Doug - That's interesting. I wish Wikipedia had cited its source. Gerald of Wales in 1190 or so called the River Dee the Deverdoeu, which is just a form of its current Welsh name (Afon Dyfrdwy) and so entirely unsurprising. But he doesn't say the town of Chester had the same name. Here's the Project Gutenberg translation:
"The Doverdwy, called by the
English Dee, draws its source from the lake of Penmelesmere, and
runs through Chester."
I do wonder if Gerald is Wikipedia's "late twelfth century" source and the name of the river has been misapplied to the town. If this is the case, then we don't have a contradiction; the river kept its Deva name in the 12th century as it does to this day, and the town acquired its name from its Roman military function.

We may have the Vikings to thank for York not being called Everchester! The Old English name for the town was Eoforwic, which the Vikings mangled into Jorvik and Middle/Modern English then mangled into York. By analogy with London, where the early English settlement is at Aldwych, one would expect the early English settlement at York to be just outside the fortress (where there's less risk of decaying masonry falling on your head), perhaps on the site of an old vicus, and on the river for trade, i.e. more or less where the Norse settlement at Coppergate is. That would have left the way open for the Roman fortress to be called by a different name, as at London, and Eoforcaster would have been an obvious candidate, which would then likely have turned into something like Everchester or Evercaster. (I use Eoforcaster as the Old English name for the fortress in Exile, for this reason).

Bede, writing in Latin, knew York by its Roman name, Eboracum. It may be significant that the two Caerlegions (Caerleon on Usk and Chester) are both in the west, whereas the legionary fortress that kept a variant of its Roman name is in the east. Clutching at straws, perhaps if English federates had served under the Late Roman Army Command at York and then stayed on and gradually took over, they would be likely to know and retain its Latin name in some form. In the west, if the local successor elites hadn't been part of the Late Roman Army they may have had less reason to care about the Roman name. I don't insist on this! - other interpretations are possible, including random chance.

Rick said...

I'll have to remember 'Everchester' as a steal-worthy name!

On Chester, perhaps 'legion' remained in use as a term for the ex-Roman forces that remained after the official legions left.

To us the word has a connotation that mixes academic formalism (legions as contrasted to auxiliaries) with the HollyRome image. But, at least according to Edward Luttwak, The Grand Strategy of the Roman Empire, by the 4th century there had been so many detachments (vexillations) and general shuffling around that the identity of the old legions was pretty blurred. (They probably didn't look much like our mental image of 'Roman legions' by that time, either.)

So if auxiliary troops remained at Chester in the 5th century, they might cheerfully have gone on regarding themselves as 'the legion.' Who was going to tell them otherwise? And the name 'legion fort' would stick even after they eventually melted away, or whatever.

Having said that, I'd still guess that 'legion fort' was the name in common use even in Roman times, 'Dee victorious' being only an administrative name on inscriptions and official documents.

Carla said...

Rick - You're welcome :-) Agreed that legions in 400 AD didn't look much like legions of the 1st century, and that if part of a legion was still at Chester after 400 AD they wouldn't be about to stop using the name because of the Rescript of Honorius. However, even though the functional distinction between the legions and the auxiliaries got blurred, I had the impression the names were retained, i.e. the 17th Sarmatian Auxiliary Ala (or whatever) didn't start calling itself the 17th Legion even if there was little distinction in the jobs they were doing.

It's possible that some sort of military force remained at Chester after 400 and carried on calling itself the Legion, whether it was a vexillation left as a skeleton garrison when Maximus 'borrowed' the troops or a local militia or warlord. I'm inclined to think the name was already long established, perhaps back to when the Legion originally built the city in the first place, because a City and a Legion would have been so distinctive they would be obvious candidates for a place name. For what it's worth, my guess is the city was 'Deva Victrix' in official correspondence and if you were speaking or writing Latin, and 'Caerlegion' if you were a local referring to it in Brittonic. When official correspondence stopped, the official name went too but the local name stayed.

Rick said...

Speculatively, auxiliary troops still in Britain after 410 might have promoted themselves to legions, so to speak - or might have been so designated by Ambrosius.

But whether or not anything like that happened, I tend to agree with you that locals probably called it city of the legion even when 'real' legions were still stationed there. For all we know, even local people writing in Latin did so. Certainly by the time Bede came along, writing in Latin, he plainly had never heard of Deva Victrix, and instead calls it Urbs Legionis.

Carla said...

Speculatively again, the legions might have been regarded as rather old-fashioned and frumpy beside the glamorous dashing cavalry by then :-) As was the case in later centuries. The Comes Britanniarum, commanding a mobile field force, outranked the Dux Britanniarum, commanding the static garrison on Hadrian's Wall and the legion (or what was left of it) at York. I reckon relative prestige between a late empire legion and an auxiliary ala can be argued either way.

Rick said...

A whole new tangent! What do we know about cavalry in this period? Horse mobility would obviously be useful for the forces of a Dux Bellorum charged with, or asserting, defense of all Britain. And a force of cataphracts is a nice literary dodge for 'knights' in this setting. But until stirrups were available, cavalry would be totally ineffective against a shieldwall. (Well, horse archers could be effective, but I've never heard of this arm in Western Europe.)

Carla said...

A whole new tangent indeed, and one for another thread! Not a lot is known for sure, it relies on interpretation and inference from things like the Triads and poetry. We do know for sure that wealthy Anglian thanes and princes at Sutton Hoo and Lakenheath owned and rode horses and were buried with them in the late sixth to early seventh century (I posted on this a while ago), and we also know for sure that mounted warriors are shown on Pictish symbol stones from around the seventh or eighth century. It seems likely that Late Roman cavalry techniques would have continued in use if they were militarily effective and the economy could support them. But how and to what extent horses were used in early medieval warfare is mostly a matter of extrapolation and interpretation.

Turning your argument around, cavalry is ineffective against a shieldwall that stands firm. Get the shieldwall to break, through fear, ill-discipline, surprise or because the warriors think they've won and decide to charge, and it's a different matter. How about the effectiveness of light cavalry being as much psychological as physical? I should imagine it takes a lot of nerve to stand still and hold onto your spear with a bunch of horses charging right at you.

Rick said...

True about standing firm. In fact my impression is that cavalry-infantry encounters are very 'asymmetrical' that way. If the infantry holds, shock cavalry of any sort is pretty much ineffective; if the infantry breaks, the cavalry has a slaughterfest.

But my impression is that in later AS England both the English and Vikes fought on foot, even if they used horses for mobility. So if post-Roman times were cavalry dominant, the infantry shieldwall regained the upper hand at some point.

Carla said...

Asymmetric describes it well. It occurs to me that a pure cavalry force versus a pure infantry force is a recipe for stalemate, for just that reason. My guess is that military commanders in the 5th/6th/7th century weren't stupid, any more so than their counterparts now, and that the effective ones would quickly have acquired a mixed force if it were available, using allies and/or mercenaries if their own traditions were weak in the relevant area. (Any takers for that as part of the strategy underlying Penda of Mercia's long-standing alliance with Gwynedd?)

I suspect that "if available" is the key phrase in that supposition. Trained cavalry horses and trained men to ride them require a lot of investment and maintenance over a long time. An infantry shieldwall requires spears (made by the local blacksmith), shields (made by a competent carpenter) and sufficient discipline, courage and/or bloody-mindedness to stand firm. I can see how a subsistence economy - which is what post-Roman Britain mostly was, high status trade notwithstanding - can support the latter a lot more easily than the former. So I would expect that economic reality would favour the bulk of armies being infantry, regardless of any strategic or tactical considerations. With the Vikings, you have the added wrinkle of it being a right pain to transport horses by ship; or you could turn that around and say that the longship plus a long coastline and a lot of navigable rivers provided the Danes with some of the speed and mobility more traditionally provided by cavalry.

That doesn't preclude a (small) class of mounted warriors, drawn from the elite who could control sufficient resources and perhaps trailing clouds of Late Roman Imperial prestige, forming the top brass and dominating the poetry (and thus surviving traditions) out of all proportion to their actual numbers.

Also worth bearing in mind that 'warfare' in the period was likely to be mostly a small-scale affair, at least as much concerned with skirmishing and cattle raiding as with anything that looked like a set-piece battle. How about cavalry (or ships in the case of the Danes) being dominant in raiding but not so much in the occasional set-piece? Asymmetry again, but on a different level.

Rick said...

Cavalry is indeed very well to raiding, for which engaging formed up infantry hardly matters - the raiders simply don't engage them. I imagine this is a big factor in the spread of knights on the continent after the decline of the Carlovingians.

Which leaves the interesting question of why earlier fragmentation in Britain didn't lead to a similar result. I'll guess that as the West Saxon state coalesced, whatever cavalry it had earlier gave way to an alternate form of action against (local) raiders, marching on them in overwhelming force.

The alliance between Penda and Gwynedd might well have been military complementarity (I know zippo about it), but this raises an interesting side question - had the Welsh not invented the longbow yet? If so it had no broader impact till much later.

But speaking of 'much later,' there's another interesting angle to the whole cavalry thing. It is notable that the Famous English Victories of the Hundred Years' War were all won on foot. While these are most associated with the longbow, it was integrated with men-at-arms, knights in fact, on foot, forming the equivalent of a shieldwall.

Perhaps a cultural bias, that even the Norman Conquest didn't fully shake, that at crunch time the thing to do was to form up shoulder to shoulder? Much later still, think of the implications of Kipling's 'thin red line.'

Carla said...

If cavalry was more widespread among the Brittonic elites, as one could infer from the poetry, one could speculate that something similar might have happened there.

Archery is mentioned in the Battle of Maldon, and was certainly known among the Danes, so presumably the Britons and English used bows as well. But the great war bow didn't come to full development until the High Middle Ages, for whatever reason. Perhaps because to use the really powerful longbows you have to start training seriously from an early age to develop the required musculature while the body is still growing, and then you have to keep training at a high level. Possibly that may have been a problem in a subsistence economy?

For what it's worth, my guess would be that Hundred Years War tactics had more to do with effective integration with a longbow corps than with cultural memory, but who's to say? Interesting idea.

Rick said...

Perhaps the great bow was only developed when the Welsh were more or less pushed back to guerrilla warfare. In rugged country, like Wales, a few bowmen could give a patrol a very disagreeable surprise.

Tactical coordination with archers was crucial to medieval English military success, but it is still a bit striking to me how much their victories were won on foot.

Carla said...

Upland areas (like much of what is now Wales) tend to be better suited to pasture than crop-raising, and it may be that a pastoral economy has more reason than an agrarian one to develop a long-range missile weapon. Protecting livestock from predators and so on. Though that doesn't explain why the longbow was traditionally associated more with south Wales than north. It may be chance and happenstance, a regional variation that happened to find a military application - which then made it profitable and glamorous and reinforced further development.

Rick said...

Pastoral peoples are notoriously pugnacious, which could also be a factor. I didn't know that the longbow was especially associated with south Wales, only that it was Welsh.

Carla said...

Well, I may be wrong on that. I read it somewhere and can't immediately remember where. If I can find the reference to check it, I'll write it up (at some point).

Rick said...

Or you may be right. :-) All I know about the early history of the longbow is that the English picked it up from the Welsh. Which Welsh wasn't specified!

Carla said...

If I find anything definitive, I'll let you know.

Re an earlier point, the people living in the uplands of what's now Wales have probably gone in for ambush and guerilla tactics since they were called the Ordovices and Silures, if not before. The one thing that never changes about history is geography, and all that. The longbow may have been called "Welsh", but it probably originated long before anything called "Wales", north or south, existed.