Raiding and Dykes

Offa’s Dyke


As we lack detailed written descriptions of raiding in early medieval sources, let us examine periods of incessant small-scale military forays from other periods and places. By understanding how raiding worked, it may be possible to see how early medieval dykes functioned as a deterrent. Obviously, we have no records from prehistoric societies, but archaeologists have speculated that cattle raids plagued Bronze-Age British society, usually in the autumn when the harvest had been gathered in. Unsurprisingly, there is evidence of raiding from the Viking period from sources like the Orkneyinga Saga; interestingly Viking raiders usually went elsewhere when they faced fortifications, suggesting that dykes would have acted as a deterrent. In the later medieval period, there is evidence that rulers avoided decisive pitched battles and that raiding enriched one side while destroying the other side in a conflict. The raids that characterized life on the Anglo-Welsh and Anglo-Scottish borders in the later medieval period are well attested in written records, as well as in physical remains, like the Pele towers.

The incessant raiding in nineteenth-century East Africa was slightly ritualized, though forays in search of women, cattle and slaves could destroy kingdoms. In World War I, the British Army started systematically raiding enemy trenches, which lead to retaliatory raids along the Western Front. The objective of these raids was to demoralize the enemy, gather intelligence and ensure that all the attackers returned safely to their own lines. Halsall’s anthropological study of warfare in societies with a similar technology to early medieval Britain in Sudan, South America, New Guinea and the Maoris of New Zealand suggests that there was a great deal of raiding that involved the theft of goods rather than mass invasions to steal land. Even though these raids were often ritualized, in certain locations they could lead to a large number of fatalities. These studies confirm that raiding has often been an integral part of war. It was sometimes ritualized, usually involved the theft of animals or people, raiders often retreated if they encountered resistance and raids could demoralize or devastate the community under attack.

What these studies of other periods and places suggest is that raiding is often carried out by nomadic societies against more settled agrarian communities (for example, Berber raids on the Roman Empire, Mongol raids on China and Tuareg raids on their neighbours to the south). As well as the largest early medieval dykes, a great many earlier hill forts lay along the Anglo-Welsh border. This is possibly because this is the interface between higher land to the west (which supported a more pastoral economy) and the lowlands (mainly arable society) to the east. The dykes of the Anglo-Welsh border and southern Wales were probably designed to prevent attacks by highland raiders on settled communities to the south and east.


While it is dangerous to assume that dykes from other countries or even British dykes from different periods fulfilled similar purposes, a study of a phenomenon that treats it in isolation is flawed. It is impossible to go into the same level of detail, especially with foreign dykes, as was undertaken on early medieval dykes, so only those directly relevant are discussed.

Earlier and Later British Walls and Dykes

As we obviously have no written records, we are probably even less likely to understand the purposes of prehistoric dykes than those of an early medieval date. Witness Sauer’s study of Aves Ditch that contains fifteen pages discussing the issue, with numerous comparisons with other earthworks, but can only tentatively conclude that it was possibly a tribal boundary. Even then, he adds a question mark to the statement. Some earthworks may have been trackways or cattle droveways; others probably demarked land divisions, or were at the edge of wasteland to delimit a group’s cultivated territory; some look like they fulfilled a defensive role, while many appear to be territorial boundary markers. It is perhaps significant that while there are few finds from excavations of prehistoric dykes, they have produced some contemporary pottery sherds and metal objects; certainly more than early medieval dykes. Pottery finds from early medieval dykes are invariably prehistoric or Roman pottery sherds sealed under the bank, or residual material incorporated into it. As already mentioned, the pollen evidence does suggest that prehistoric dykes cut through more intensively cultivated areas. If the early medieval dykes were built in thinly inhabited contested borderlands that would explain the lack of contemporary pottery finds.

Like many early medieval dykes, prehistoric cross dykes are short (sometimes 100m or less in length) and tend to bisect ridges, usually with a single bank and a single ditch, making them difficult to distinguish from some early medieval route-blocking dykes. Perhaps the cross-ridge or cross-valley dykes are prehistoric dykes that fulfilled similar purposes to early medieval dykes in preventing cattle raids.

The Roman frontier works of northern Britain were highly visible features in the medieval landscape, so may have been an inspiration to early medieval dyke builders. Gildas and Bede mention them (though they misdate them to the end of Roman rule). These frontier works may have possibly been an inspiration to early medieval dyke builders, though they probably had little idea of how the Roman frontier works functioned. Roman writers described them as dividing the Romans from the barbarians to keep the latter out, but it is likely that they also controlled trade with the tribes to the north. The blocking of minor gateways in the late second century and early fourth probably represents an attempt to funnel trade through the more important crossing points. It is unlikely there was enough trade to make this the Romans’ primary stimulus for building the walls and forts, even if this role occupied much of a garrison’s time during periods of peace.

There are some obvious differences with the early medieval dykes. Hadrian’s Wall and the Antonine Wall have features not found on early medieval dykes: forts; gateways; a wall or palisade; and clear evidence of a resident garrison. Early medieval dykes are rarely contiguous with administrative boundaries like county or parish boundaries, suggesting that they were not located at borders. Similarly, the Romans set their frontier works back from the frontiers, though they also built signal stations and forts garrisoned with scouts in front of them to give advance warning of attack. Roman frontiers were not sharply defined lines on the ground, but zones. The walls, if properly manned, certainly could stop small groups of people crossing into Roman Britain and they do command good views to the north, but they do not seem designed to repel large-scale attacks. The wall was not wide enough to use as a fighting platform, therefore mobile troops would have to destroy any attackers held up at the frontier. The frontier works probably functioned in a manner closer to the Berlin Wall than the Maginot Line. The Romans would gather troops from the forts on or around the walls and engage large-scale invasion forces in open country, where the superior discipline, cavalry and heavy projectile weapons of their more mobile troops would be used to devastating effect. Perhaps some early medieval dykes worked in a comparable way.

While Romans walls may have influenced early medieval dykes, the latter in turn could have influenced later earthworks. As we have already discussed, the end of dykes probably coincided with the rise of the burh and the arrival of Viking raiders, whom the burh walls were designed to keep out (internal dykes are not much good at stopping seaborne raiders). The Vikings did build some dykes in England, but these were short features designed to defend a tongue of land, like the one on Danby Rigg in the North Yorkshire Moors, or the bank between the Thames and the Kennet recorded by Asser. Prior to the rise of the burh, the only earthworks of a comparable design to the dykes in early medieval Britain were the ramparts of hill forts. In the early medieval period in lowland Scotland these were often new constructions, while in lowland Britain they were reoccupied Iron-Age structures. Like the burhs, both Iron-Age and early medieval hill forts have palisades and gateways, of which archaeologists have found abundant evidence. This adds credence to the supposition that early medieval dykes did not originally have palisades or gateways, as the numerous excavations of these earthworks would surely have uncovered some evidence of them.


British dykes are not unique. There are similar prehistoric and medieval earthworks across Europe, with examples in Ukraine, Hungary, Apulia in Italy, Sweden (Götavirke) and Spain (the 2.5-mile/4km long El Muro near Teverga) and Romania. The nearest is a series of long south-facing earthworks in Ireland, which match in scale some in Britain; they run from Bundoran on the west coast to near Armagh, effectively dividing Ulster from the south. The largest are the Dane’s Cast, Black Pig’s Dyke and the Dorsey, but differentiating among them, especially the first two, is difficult as locals use the two names interchangeably and all three lie on a similar alignment.

Contact between Ireland and northern Britain may have influenced dyke building on either side of the Irish Sea in the early medieval period. Earthworks found across areas traditionally thought of as the Anglo-Saxon homelands (Denmark and northern Germany) may also have provided the inspiration for Anglo-Saxon dyke builders (and raiders). One European earthwork that British archaeologists have drawn parallels with since Pitt Rivers took a (borrowed) spade to it over a century ago is the Danevirke. This south-facing earthwork runs for 18.6 miles (30km) along the base of Jutland, blocking access into Denmark from Germany. It was built in at least seven phases and though the Royal Frankish Annals attribute it to King Godfred in 808, dendrochronology suggests that the earliest phases of building occurred shortly after 737. Interestingly, the Royal Frankish Annals also claim that it ran from sea to sea, a statement that is as inaccurate as Asser’s assertion that Offa’s Dyke performed the same feat, as the ends of the Danevirke lie on rivers. Even with a twelfth-century rebuild that clad the front in stone, the evidence for a wooden palisade in the earlier phases is obvious, suggesting that if the British dykes had been furnished with one, we would have found evidence of it by now. The Danevirke had a main gateway where the Hærvej, or army road that runs along the spine of Jutland from Germany, crossed the earthwork. The road is also called the Cattle Road, the Oxen Road, the King’s Road and the Main Road, suggesting that cattle under the control or protection of the king (possibly as tribute) walked along it in and out of Denmark. Though the references are frustratingly terse, early medieval battles were fought at the earthwork; the Danish army used to muster along the earthwork during times of international uncertainty up to the nineteenth century.

To the north of the Danevirke, there are at least twenty-eight earthworks in Jutland, many of which cut routeways, as well as six tree barriers built across narrow belts of sea. The most elaborate is the Olgerdiget: this was a 7.5-mile (12km) long stockade made up of large poles, though a 1.2-mile section (2km) has a ditch (1.6m deep by 4m wide) with a bank that is dated to 219 by dendrochronology. It was not garrisoned, but possibly patrolled with defenders mobilized in time of war, and seems to mark the dividing line between the Jutes and the Angles. This means that the Anglo-Saxons and Jutes had a history of building dykes before they gained control of England. Interestingly, this study found no records of dykes in Brittany, a place where so many other aspects of British culture were imported in the fifth and sixth centuries. This suggests that the building of a dyke to combat raiding was initiated in early medieval Britain by Germanic incomers (Anglo-Saxons), rather than being a part of native British culture, though of course prehistoric British dykes might also have been an inspiration. Unfortunately, like their British counterparts, few contemporary sources survive describing many of these European earthworks, so it is difficult to ascertain their original purpose. Even when records do survive (both inscriptions and texts), such as those associated with the 87-mile (140km) long dyke that the Bulgars built in Thrace against the Byzantine Empire, which clearly suggest that it had a military purpose, scholars claim it was purely symbolic. Other scholars (myself included) are less convinced that we should ignore such clear primary evidence. Earthworks were replaced by forts that had a clear military function as the main type of Bulgarian defence, which possibly parallels the change from dykes to burhs or forts that King Alfred built in ninth-century England.

Asian Dykes

With the Great Wall of China, there survives documentary evidence which tells why the Chinese built it and how (this study used evidence from China to calculate the labour needed to build linear earthworks). The earliest walls were anonymous, practical structures built when the Chinese Empire was weak or their diplomacy particularly unsuccessful to counter raiding by nomads to the north. Many of the dykes of southern Wales, like Tor Clawdd and Bedd Eiddil, seem to block access to the coastal plains from the mountains (where people lived a more pastoral and possibly nomadic lifestyle), while Offa’s Dyke and Wat’s Dyke possibly fulfilled the same purpose of keeping Welsh raiders out of Mercia. Perhaps Offa’s Dyke represents a breakdown of diplomacy with the Welsh during his reign; relations had been much closer when the earlier Mercian King Penda was a close ally of the Welsh King Cadwallon. Alternatively, like the later Chinese walls, whose remains we see today on tourist posters and which were often symbolic rather than anti-raiding defences, Offa’s Dyke merely reflected Offa’s imperial pretensions. These two very different functions, though not mutually exclusive if these structures were multi-functional, do highlight the danger of cherry-picking analogous examples from other countries or periods.


Linear earthworks in other countries and periods have controlled trade and delimited territory, but often were designed to protect areas from raiders. While it is possible to make analogies with early medieval British dykes, we should be cautious, as people can build similar structures in response to dissimilar circumstances. Such comparisons do suggest that gateways and palisades would leave obvious traces on early medieval dykes in Britain and therefore we can possibly dismiss the suggestion that they were ever created.

2 thoughts on “Raiding and Dykes

  1. Pingback: Raiding and Dykes – Glyn Hnutu-healh: History, Alchemy, and Me

Comments are closed.