Wansdyke home I What's New I Sitemap I Bibliography I Vortigern Studies l POLLS I LINKS l Sitemaster I FAQs
Vortigern Studies Index
The Welsh Marches contain a large number of linear earthworks of various sizes and lengths. Offas Dyke is the best known and longest of these, but other well known examples include the Rowe Ditch in Herefordshire and the one I will be discussing here, Wats Dyke. It is usually thought that the name is not of any great antiquity, being first documented (as Clauwdd Wade) in 1431, although the lateness of the date may be a result of the vagaries of survival of Welsh medieval documents. There is certainly a strong hint that it was already known as such by the late twelfth century (see below). Long standing confusion over the separate identity of Offas and Wats Dykes is shown by a number of placenames along the line of Wats Dyke, including Bryn Offa (Wrexham) and Clawdd Offa (Sychdyn, Flintshire). At Ruabon, the two earthworks are less than a kilometre apart, but they are clearly separate entities. Wats Dyke runs, more-or-less continuously, between Basingwerk (Flints.) and south of Maesbrook (Salop.), a distance of some 65 km.
The origins and meaning of the name
The name of the dyke incorporates either the Old English personal name Wat (which is unlikely given the early spellings) or the name of the folk hero Wade (Old English Wada). There is no reason to suspect that it derives from Old English wet (wet), as has been suggested, as the early forms of the name insist on an initial wa-. There is a Wat attested as King of the South Saxons in a charter dated 692, but it is highly improbable that he is the same individual commemorated here. Nor does it appear likely that it derives from the name of a member of the Mercian royal dynasty, which contains no individuals with a name containing the element Wad-. Perhaps there is more to be gained through a comparison with the name Wada.
Wada is familiar in Old English poetry as the ruler of the Hęlsinga in the poem Widsith, as part of an enumeration of famous rulers of European peoples; the Hęlsinga are listed between the Swęfe of Swabia and the Myrgingas, living between the Eider and the Elbe, but appear to be the people of Hälsingland in eastern Sweden. He is found in other Germanic tales, not as a human but as a sea giant, whose magic boat was used by Heorrenda and Hild to flee from Hilds father to Heodens court. Both Heoden and Hagena, Hilds father, are mentioned a line before Wada in Widsith, suggesting that the story was well known to the composer of the poem. The hero Weland (the smith of Germanic mythology) was portrayed as his son and Widia, an historical character mentioned in Jordanes de Origine et Actibus Getarum xxxiv as Vidigioia, his grandson. His popularity as a folk hero continued until at least the fifteenth century, when his magic boat is mentioned by Chaucer in The Merchants Tale and Troilus and Criseyde. In Britain, his memory has been preserved in Wats Dyke, Wades Causeway (a Roman road between Malton and Whitby, North Yorkshire), a pair of standing stones, each known as Wades Stone, at Barnaby and Goldsborough (North Yorks.) and the placename Wades Gap along Hadrians Wall.
Perhaps most interesting of medieval references to Wada is in Walter Maps de Nugis Curialium II.17 (Courtiers Gossip), written between 1182 and 1193. Under the name Gado, he is portrayed as the son of a king of the Vandals who befriended Offa, King of Mercia. As a champion of the just, he was temporarily able to frighten off the Romans who were preparing to invade Offas kingdom. However, whilst he was away resolving a conflict in India, the Romans took the opportunity to land in England. The Indian crisis averted, Gado had hoped to return to his homeland, but his magic boat brought him instead to England, where he visited the Roman army in the hope of preventing the Emperor from attacking Offa. When his mission failed, he and Offa attacked the Romans, defeating them in the streets of Colchester.
Whilst this story is clearly unhistorical (for instance, there was no Roman Empire, Holy or otherwise, during the reign of Offa), it is curious that the eponyms of the two great boundary Dykes of western Mercia were associated by a twelfth-century writer. If nothing more, it suggests that Wats Dyke was already known by that name early in the twelfth century, although direct references have not survived. It may also indicate that, by then, the two Dykes were seen as somehow sharing origins, as Sir Cyril Fox believed, although this is perhaps reading more into a late story than is justified. Nevertheless, might we speculate that the Romans of Maps story are the eighth-century Welsh Cymru, citizens?
There is a further piece of evidence about the origin of the name. Watling Street, the road connecting the channel ports with Chester, derives from the Old English name Wat. The name is first recorded in Bedes Historia Ecclesiastica (i.7) as Uęclingacęster (for *Uętlingacęster), given as the Old English name of Verulamium, St Albans. All manuscripts of Bede have the incorrect initial Uęc-, so the corruption of the name must go back to a very early copy of the Historia, if not to Bede himself. Nevertheless, the Old English town name may be analysed as meaning Roman city of the Watlingas, the Watlingas being the people of little Wat. Watling Street shares the same derivation and is the main road passing through Verulamium. Who the Watlingas were or where they lived is not at all clear and it would be dangerous to assume that they were the people who lived around Verulamium. Once again, it is tempting to go beyond the evidence and suggest that the Watlingas were the people who lived around Wats Dyke and that the road gained its name because it led into their territory from the Anglo-Saxon heartlands of the South East. This could mean either that the Dyke was so called because it lay in or on the border of their territory or that they took their name from a Dyke that already existed by the early eighth century.
The course of Wats Dyke
Sir Cyril Foxs pioneering survey concluded that Wats Dyke consisted of three separate stretches of earthwork, with gaps between them. In these places, he surmised, formerly dense forest or deep ravines made a formal barrier unnecessary. Recent work has demolished this view. Not only does Wats Dyke emerge as a single, almost continuous earthwork, but it extends at least 4 km farther south than Fox believed.
The northern end of Wats Dyke is generally believed to start at Basingwerk, on the Flintshire coast of the Dee Estuary; the placename suggests that there was once a fortification of some kind here, although no trace now survives. The First Edition Ordnance Survey maps clearly mark its course, though, while Thomas Pennant was able to describe it in 1773. The Dyke first becomes visible to the northeast of Holywell, at a site known as Coed Strand, where the line has been proven by excavation. Here, it turns more to the south then to the southeast, to run into the Nant-y-Fflint, where it can be seen at Fron Dudur. At the mouth of Nant-y-Fflint, it turns to a more southerly alignment via Coed Llys, then to a south-easterly alignment via Sychdyn and Mynydd Isa towards Hope, where it meets the Afon Alun, which it follows to Bryn Alun. Crossing the river, it runs on a south-south-westerly alignment and is then lost under the railway as it passes through the suburbs of Wrexham at Plas Coch and Bryn Offa before becoming visible on a south-south-easterly alignment to the south-west of the town. It crosses the Afon Clywedog, passing along the west side of Erddig Park, where a Norman motte-and-bailey castle overlies it. Running due south again, it is very well preserved as it passes Middle Sontley. As it crosses the B5426, it turns more to the south-southwest to Pentre-clawdd, then resumes its southerly course past Ruabon to Wynnstay Park. The Park was originally known as Watstay until changed by Sir John Wynn, who died in 1719. There is then a gap of about 6 km, as the Dyke crosses the formerly boggy ground in the Dee Valley; it is unclear if the line was marked by an earthwork in this area. On the north side of the Dee, Fox believed that Afon Eitha formed the boundary, while to the south, Afon Ceiriog and Morlas Brook served to define it.
Beyond this, it runs along a more south-westerly direction from Esgob Mill through Henlle Park, Prys Henlle, Bryn-y-Castell and Gobowen, passing under the level crossing, via Pentrer Clawdd to the hillfort known as Old Oswestry (where the older name Yr Hźn Ddinas Welsh for The Old Fort is to be preferred). After this, the Dyke turns more towards the south south-east, running east of Oswestry via Mile Oak and Ball Mill to Maesbury, where the earthwork ends, although Sir Cyril Fox believed that its termination lay a little farther north, at Pentre-coed. However, his survey provides evidence that the River Morda had been artificially straightened at this point, and the Dyke can now be shown to stretch another 4 km south of Maesbury. This is the earliest post-Roman hydraulic engineering in Britain. Traces of earthwork south of this suggest that the Dyke faded into the boggy ground where the Rivers Morda, Vyrnwy and Severn meet.
Intriguingly, Wats Dyke forms the boundary between those parts of Cheshire that were hidated in 1066 and those that were not. Hidation was part of the process of taxation in Anglo-Saxon England and involved an assessment of how many families could be supported by a given area of arable within a community (according to Bede, Historia Ecclesiastica i.25, a hide was sufficient for one family; by the time of Domesday Book in 1086, it had been more-or-less standardised at 120 statute acres, although originally, it seems to have varied a great deal according to the productivity of the land). It was something that occurred early in the Saxon period (the so-called Tribal Hidage, a document of early seventh-century date, already assesses individual English kingdoms in terms of hides) and the coincidence of the line of Wats Dyke with the limit of hidation in north-west Mercia is an important observation. Equally important is the fact that township boundaries largely ignore the Dyke and that the boundary of Cheshire in 1066 did not follow it, either. Nevertheless, those parts of Cheshire that lay to its east contain a high proportion of Old English placenames (albeit, often with a later Welsh disguise). The area west of the Dyke, by contrast, is much more Welsh in both placenames and the recorded tax arrangements in 1066. As a boundary, its human impact was therefore immense and long lived.
The form of the Dyke
The Dyke consists for most of its length of a bank and ditch with no berm between them. The ditch is always to the west and the bank to the east. In a number of places (such as Fron Tydyr, east of Holywell), there is no bank, because the hill slope renders it unnecessary, although a ditch was dug, producing a lynchet-like profile today. The bank is a little over 8.1 m wide at the base on average, with extremes of 6.4 and 12.2 m, while it stands today up to about 2 m high in places and would originally have been somewhat higher. It has a turf facing and is mainly composed of soil strengthened with turf layers. At Maes-y-Clawdd (Oswestry, Salop.), the bank was 10.2 m wide at the base, with a turf core and a slight berm (of 0.7 m) between it and the ditch. A small quantity of residual Romano-British pottery was found incorporated into the bank. At Pentre Wern (Oswestry, Salop.), the ditch was 8 m wide and up to 4 m deep, and the bank some 7 m wide, while at Mynydd Isa (Flints.), it was 6.1 m wide. At Wrexham General Station, the bank was found to be over 6 m wide, with hints of a robbed out facing on the west side. At Selattyn (Yewtree, Salop.), the ditch was at least 1.2 m deep. There is no indication that it ever had a palisade or wall on top (as some stretches of Offas Dyke appear to have done), although the published section from Sychdyn (Flints.) appears to show the remains of a timber palisade between the bank and ditch, suggesting a feature resembling a Roman turf-and-timber fort rampart.
At Rhydyn Hall, south of Hope (Flints.), there was a hint of a cobble feature at the base of the rampart, perhaps a kerb of some kind, although at Pentre Wern (Salop.), a similar feature was interpreted as an accidental dump of gravel. The ditch is an average 5.7 m wide and was found to be 2.4 m deep at Sychdyn (Flints.), with a cleaning slot at the bottom. At Mynydd Isa (Flints.), it was 3.7 m wide, with a flat bottom 1.8 m wide, and 1.2 m deep, while at Maes-y-Clawdd (Salop.), it was 7.4 m wide and 2.3 m deep, with no sign of a subsequent cleaning. Fox considered Wats Dyke to be, on average, about two-thirds the height and width of Offas Dyke.
The question of entrances through the Dyke does not appear to have been settled. There are two of places where Fox thought that the present gaps might have been original: north-west of Clawdd Offa in Hope (Flints.) and south-east of Henlle Hall (Salop.). However, at neither of these is there any trace of an early trackway that would have used them. Moreover, at Henlle Hall, excavations demonstrated that the ditch continues across the gap. The question of how gates through the Dyke were protected (if at all) does not seem to have been settled by fieldwork.
The Dyke also incorporates a number of other earthworks. Some, such as Yr Ddinas Hźn, were already in existence; at least one, though, the work that gave its name to Basingwerk, seems to have been Mercian in origin. Although Fox concluded that they were not used as part of the defensive scheme, it is worth considering their possible role. A number of hillforts, for instance, were reoccupied during the late or sub-Roman periods, and the incorporation of reoccupied hillforts into the line of Wats Dyke might then strengthen a case for its early date (see below). It would also make Wats Dyke very different from any of the other sub-Roman and Saxon earthworks of Britain, which exist as simple bank and ditch structures without garrisons or mustering points.
Wats Dyke presents a much more uniform appearance than Offas Dyke, which suggests that it was built in a short time by a strong central authority. This may mean that the work of construction was undertaken by a single group of workers (which is just feasible) or that a great deal of planning went into its building and that the individual foremen had clear instructions (which fits better with our understanding of how Mercian Dykes were built). There is also a clear logic in its siting and overall strategy. It runs from the Dee Estuary in a course broken only at the River Dee, to the River Morda, a tributary of the River Vyrnwy (itself a tributary of the River Severn), a distance of 65 km, and although it runs through lowlands, it uses knolls and ridges to give a clear view to the west. In this, it is very different from Offas Dyke. This occupies the lower slopes of the Welsh uplands and presents a formidable barrier to those approaching from the west; Wats Dyke, by contrast, looks as if it is sited to allow those manning it to keep an eye on the uplands and to restrict movement through the lowlands for those approaching from the west.
The date of the Dyke
Wats Dyke is generally supposed to be earlier than Offas Dyke. The arguments are based largely on the geography of the two earthworks; as Wats Dyke is rather to the east of the Offas Dyke, it is supposed to represent an earlier boundary between Mercia and Powys than that implied by Offas Dyke. As the conquest is generally seen as a single process, with the Saxons pushing inexorably ever westwards, then a more easterly Dyke must, by its very location, be earlier. The fallacy of the logic in this view is clear. However, the relative locations of the two earthworks mean that it is essential to attempt an assessment of their relationship (if any). Offas Dyke is generally acknowledged to have been built during the documented campaigns of Offa, King of Mercia 757-796, against the Welsh, which appear to have been at their most intense during the 780s. Although no single date can be suggested for the construction of Offas Dyke - and it is inherently unlikely that it was built in a single year anyway - it is likely to have been in existence by the early 790s, if the attribution to Offa is correct.
Sir Cyril Fox considered the two Dykes to be so similar that they ought to be of roughly contemporary date. He considered that the details of design, construction and overall strategy so similar that both should be considered part of a Mercian school of dyke-builders. Other dykes that need to be considered as part of this group are the so-called Whitford Dyke (which Fox considered a northern extension of Offas Dyke), Wantyn Dyke (Radnor), Rowe Ditch (Heres.) and the so-called Short Dykes of east Radnorshire and Shropshire. Fox makes the point that all these (with the exception of the Whitford Dyke) are highly localised in conception and fall into two main types: those that cross ridges and terminate on their lower slopes, and those that cross valleys and terminate on the hill slopes that define them. He thought that their construction was perhaps related to the position of the Mercian capital at Tamworth, which was probably a foundation of Penda (King of Mercia c 632-655) and that they related to a phase of defence somewhat earlier than Offas Dyke but nevertheless within the same general tradition.
At Mynydd Isa (Flints.), W J Varley found an annular clay loom-weight in one of the ditch fills. The loom-weight was broken and had been discarded into the ditch, apparently deliberately. It lay on top of a patch of burnt puddled clay, interpreted by the excavator as the base of a hearth, which in turn lay above a primary ditch fill consisting of two humic deposits separated by a thin band of grey clay. This primary fill was only 50 mm thick, suggesting a short period of formation before the construction of the hearth, estimated by the excavator as three decades or less. Unfortunately, thermoluminescent and thermoremanent magnetic dating techniques had not been developed when the excavation at Mynydd Isa took place in 1957; they would have enabled direct dating of the hearth. However, the presence of the broken loom-weight on top of a hearth constructed within perhaps thirty years of the Dyke is an important and often overlooked dating tool. The form of the weight is Middle Saxon (i.e. c 650-800). This is the best date available for the construction of Wats Dyke, but although it indicates a Middle Saxon origin, it is not accurate enough to satisfy our historical curiosity.
Until recently, the consensus has been that Wats Dyke is somehow intermediate in date between the Short Dykes of the seventh century, which were designed to counter small-scale military activity, and the late eighth-century Offas Dyke, which is imperial in scale (indeed, it is the most massive archaeological monument of the British Isles). It is usually assumed that its builder was Ęthelbald, King of Mercia 716-757, and the greatest English king of the early eighth century. Alfred Palmer thought the placename evidence for the confusion of the two Dykes pointed to Offa as builder of both, although this suggestion does not seem to have met with much favour. David Hill has made a case for it being later than Offas Dyke, as a replacement for an unusable and incomplete northern section, perhaps built in the reign of Cnwulf (King of Mercia 796-821) or as late as the 850s. Nick Higham has suggested that it might be seventh-century, as early as c 642. David Hills colleague in the Offas Dyke Project, Margaret Worthington, has recently supported a radical redating of the Dyke, suggested by a single radiocarbon date.
A scientific date for Wats Dyke?
An article by Keith Nurse in History Today (August 1999), called New dating for Wats Dyke, first brought to a wider public attention a radiocarbon date from excavations conducted in February 1997 at Maes-y-Clawdd, Oswestry (Shropshire). Whilst the articles raises some important points, its general conclusion that it revolutionises our understanding of the date of the earthwork is over-optimistic and archaeologically naļve. The date is undoubtedly important (it probably - but not certainly - confirms the post-Roman date of the earthwork), but it is neither unexpected nor earth-shattering in its implications. The problem is that too many conclusions have been drawn from a single, unsupported date that is not even of material from the Dyke itself.
The radiocarbon date
The date, as supplied by the radiocarbon laboratory of Queens University, Belfast on 7 July 1997 (sample number UB-4158), is 1571 ± 69 years BP (Before Present). Present in terms of radiocarbon dates means 1950; the ± symbol gives us an indication of the sampling errors in determining the age, expressed as a statistical standard deviation. Unfortunately, radiocarbon years are not the same as calendar years, and need to be calibrated against samples of known age. The calibrated date can be calculated as AD 486 ± 75; what this means is that there is a 68% chance that the true calendar age of the carbonised wood from the hearth falls into the period AD 411 to 561; increase the range to two standard deviations (as most archaeologists will do routinely) and there is a 95% chance that the age falls into the range AD 268 to 630.
In technical language, the dates are:
The radiocarbon date therefore tells us that a hearth constructed on the ground surface before Wats Dyke was built contained wood that came from a plant that was alive some time between the mid-third century and the earlier seventh century. What it does not tell us is that Wats Dyke was built in the fifth century, as some have suggested, even less that it was built AD c 446.
Firstly, the date is not a date in the everyday sense of the word. We can state that the Emperor Septimius Severus died on 4 September 211 and know that on that day, the earth was in a particular position in the sky, that the moon and planets appeared in certain constellations and so on. That is a true historical date. A radiocarbon determination, on the other hand, is a statistical approximation to the age of a sample. It tells us how old it is relative to the date at which the determination was made. This is known as a relative date, because calculating it depends on knowing the starting point (in other words, the historical date of the radiocarbon test).
Worse, working out the age of a sample is done by calculating the relative proportions of two isotopes of carbon: C12, which is stable, and C14, which is radioactive and decays at a known rate (currently thought to be about 1% every 83 years). Both types of carbon are absorbed from the environment by living organisms and once formation of the cellular structures ceases (usually at death), the proportion of radioactive C14 in a sample begins to decrease. The proportion of the two isotopes in the environment is in the region of 670,000,000,000 to 1 (in European numbers, this is six hundred and seventy thousand million to one; in American numbers, six hundred and seventy trillion to one), so the measurement of proportions in samples is far from precise and because of the decay of the radioactive atoms, the proportion is always lower than this. Moreover, we cannot count all of the atoms, so an approximation is made based on the weight of the sample and either the evidence of radioactive decay measured by means such as a Geiger counter (a technique known as beta counting) or spectroscopic analysis of charged carbon ions in a particle accelerator (known as accelerator mass spectrometry). The reliability of the age calculated depends on how many atoms of C14 were present in the original sample (the older the sample, the fewer the radioactive atoms and the greater the difficulty of spotting them), the size of the sample (the smaller the sample, the fewer the atoms available to count) and the length of time available in beta counting (the shorter the time, the fewer the blips recorded by the Geiger counter). An assessment of the reliability is given in the margin of error figure that accompanies the dates; the quoted figure is a statistical standard deviation.
These techniques give an assessment of the age of the sample based on how far the proportion of radioactive C14 has declined from its assumed starting point of one atom in about 670,000,000,000. However, this would work only if the proportion of C14 in the environment has always remained at the same level. We now know that, for a variety of reasons, this is not the case. The proportion has varied considerably over time, so that the starting ratios of the two isotopes will also have varied. A technique is needed to determine changes in the proportions of the two isotopes over time. Such a technique has been available since the late 1960s, when it was found that by radiocarbon dating growth rings from trees, which can be assigned to a particular calendar year, a consistent pattern emerges that shows times when there has been more C14 in the environment and other times when there has been less. Of course, the assessments of radiocarbon dates of tree rings is also subject to a margin of error, so the margin of error for the age of a sample will increase when the age is calibrated.
Then there is the question of exactly what is being dated. If it is part of a human being (an Egyptian mummy, say), we will get a determination of the date of death for that individual. If it is a piece of timber, we will get a date for the formation of the particular group of growth rings being tested. This is where problems can occur. Structural timbers are often reused many times over the centuries, so a radiocarbon determination of timber needs to take this into account. Wood used in a hearth (which is what was dated at Maes-y-Clawdd) may derive from recently felled timber gathered specifically as firewood, or it may have arrived on the site as rubbish from a building that was already centuries old.
There is finally the problem that a single date is no date: the potential for contamination, reuse of old materials, laboratory error and so on is so great that a single radiocarbon determination is next to useless, especially if it gives a date that appears to contradict what might be expected. It is good practice to date a number of samples whenever possible, ensuring that a good range of materials is represented, minimising the risks outlined. This is why the Shroud of Turin was subjected to a number of separate tests by different laboratories and why we can be certain that the shroud was made between the mid thirteenth and the late fourteenth centuries: numerous samples, all giving the same general age, can be combined to increase the accuracy of the determination, and determinations from different laboratories reduce the effects of experimental error, laboratory contamination and so on.
Implications for Wats Dyke
It is interesting to look at the alleged early date of Wats Dyke as supported by the radiocarbon determination. What we have is a radiocarbon date from a hearth built on the ground surface beneath the Dyke at Maes-y-Clawdd. Wood used in the hearth was growing between the late third and early seventh century, but it most certainly does not date Wats Dyke. There is the basic archaeological principle of the terminus post quem: the date after which. The Dyke was constructed after the hearth was in use; the hearth was in use after the tree that supplied its fuel had been felled. That is what the radiocarbon date tells us; this is why archaeological scientists regard dates from carbonised wood as among the least informative for chronology. It has nothing to do with the construction of the Dyke, but rather tells us that the Dyke must be later, and there is no reason why it could not be many centuries later.
Nevertheless, even if we disregard the radiocarbon date as irrelevant to the Dyke, H R Hannafords suggestions are intriguing. He compares it with the Wansdyke, a linear earthwork south of the western part of the River Thames, which is of certainly post-Roman date, although the usual attribution of the eastern portion to the fifth century rests on poor evidence, while it is not clear why we should regard the western portion as a separate and later monument, as this seems to rest on an early twentieth-century hypothesis that has not been tested adequately. The placename attribution of the Dyke to the mythical hero Wada might indicate an early origin: the numerous Grims Ditches and Wansdyke itself are named after mythological beings, not historical personages. How far can we press this and is there any other evidence to consider?
David Hill has recently suggested that Offas Dyke was built not de mari usque ad mare (from sea to sea), as stated by Asser (Vita Ęlfredi Regis xx), but as a defence against the Kingdom of Powys, rather than the whole of Wales. Indeed, Wales is an entirely anachronistic concept; the Britannia of Asser was composed of eight or so separate states that sometimes formed alliances with each other or with Anglo-Saxon kingdoms, and sometimes fought each other or Anglo-Saxon kingdoms. The political history of the seventh to ninth centuries is much too complex to reduce to simple ethnic lines. David Hill has demonstrated that Offas Dyke makes more sense in terms of the relationship between two individual states than it does in terms of the relationship between Angle and Briton.
If we are to understand the purpose of Wats Dyke, we must also understand the context of its construction: it is not simply an earthwork set in an empty landscape, but an expression of a human and political geography. The major difference is that although we can use the known date of Offas Dyke to set it into its political context, we do not have a date for Wats Dyke and it is the political context that may help us to suggest a date for it.
In view of the radiocarbon date from the hearth beneath Wats Dyke at Maes-y-Clawdd, we need to understand the politics of the region from the later fifth century onwards to suggest periods at which the construction of the Dyke might be plausible. While we need to exercise extreme caution in dealing with the documentary sources for the period, as they are, with only one exception, of much later date, there is a growing body of archaeological evidence that allows us to understand something of the centres of political power at this time, even if we cannot produce a list of rulers and their relationships.
The political history of north-east Powys
The political history of Britain during the centuries following the collapse of Roman rule is very obscure. It is not even clear at what date any of the kingdoms that are recognisable later actually emerged as political units. Much of the material written about history of this period is wishful thinking at best, fantasy at worst. The simple truth is that there are no contemporary materials for writing any sort of narrative history of any of the Welsh states before the ninth century and the later sources are all of dubious authority. None can be shown to contain other than retrospective attempts at writing history.
The largely legendary material about the origins of Powys is full of contradictions. Our earliest text is the Historia Brittonum, which recounts a confusing legend about the tyrannus Benlli and Cadell Ddyrnllug (Chapters 32-5). The genealogies make Cadell an ancestor of the kings of Powys, but there is confusion about his relationship (or lack thereof) with Vortigern, and he is given three different fathers. His putative great-great-grandson, Brochwel Ysgithrog, is the first ruler of the region known to real history, if the traditions identifying him with the Brocmail named by Bede as leader of the Britons at the Battle of Chester (Historia Ecclesiastica ii.2) are correct. However, there are problems with the identification, as the Annales Cambrie and Annals of Tigernach record that his grandson Selyf ap Cynan Garwyn was the king of the Britons killed there. Even if we assume Selyf to have been young at the time (say, no more than 20), as he is not recorded as having any descendants, his grandfather would have been between 60 and 80 at the time, too old to command a seventh-century war band. The genealogies record the descent of later kings of Powys from his brother, Eiludd, but there are still problems with dates. Only four generations are recorded between Eiludd and his descendant Cadell, recorded as dying in 808, making an impossibly long generation length of around forty years. The rulers of Powys emerge from the obscurity of the genealogies only at the time of the Battle of Chester, c 616 and other sources suggest that the political situation was much more complicated than would appear from the genealogies alone. The position of Powys at the time of the disastrous Battle of the Winwęd in 655 is unrecorded.
Unlike the Anglo-Saxons, who preserved king lists complete with reign lengths, the Britons were more concerned with genealogies. This reflects their obsession with the legitimacy of their kingsperhaps because they did not usually follow the patrilineal descent patterns of the English, but had a kinship-based system of inheritance in which distant cousins might compete for power. The situation for Powys is worse than for many Welsh kingdoms. Although the genealogies record the descendants of Cadell Ddyrnllug in some detail, there are problems. The main confusion is over the position of Cyndrwyn and his son Cynddylan, who are known from the later poetry ascribed to Llywarch Hen as the rulers of a place called Pengwern. The poetry associated Cynddylan with Powys and many historians have seen the Cyndrwynin as a rival dynasty to the Cadelling, who briefly gained the ascendancy during the early seventh century. Seen in this way, the absence of the Cyndrwynin from the later genealogies is unsurprising, as the Cadelling were eventually triumphant.
However, there is an alternative way of looking at the data. A possibly historical fragment links Cynddylan (as an ally, possibly of Gwynedd) with the Battle of Maes Cogwy in 642, while bardic tradition links Cyndrwyn with the Battle of Chester. Cyndrwyn cannot have been king of Powys if Selyf ap Cynan Garwyn was also king of Powys: he must have been the ruler of another kingdom. The fifth stanza of the Marwnad Cynddylan appears to implicate him in the death of mab Pyd, the son of Pyd (Danger), perhaps Penda, son of Pybba. Elsewhere, Cynddylan and his brother Morfael are said to have fought at Cair Luitcoit, Lichfield. As Pengwern appears to be near the Wrekin and all the other places mentioned in the poems are located in northern Shropshire, it is tempting to assume that Cyndrwyn and his son were rulers of the Wreocensęte of the Tribal Hidage, generally assumed to be the people of Shropshire and Cheshire. Nick Higham has argued persuasively that territorial names with the -sęte ending were British sub-kingdoms within Mercia. The date at which the Wreocensęte were completely subsumed within Mercia is not known, but is usually assumed to have occurred during the third quarter of the seventh century, and there is reason to suspect that, if the Cyndrwynin dynasty were its rulers, it disappeared following the disaster of Winwęd. They would be more than simply allies, they would be a confederated people; however, it is significant that Cynddylans sister Heledd fled to Gwynedd, a state that had been allied to Mercia. The effect of Winwęd was to deprive Gwynedds Mercian allies of their kingdom and Heledd would have fled the new and hostile regime.
The Marwnad Cynddylan also represents Dogfeiling - north Wales east of Afon Clwyd - as being part of Gwynedd at the time of its composition, alleged to be after the fall of Cynddylans kingdom to the English. At other times, Dogfeiling is said to be part of Powys. Moreover, the poet of Marwnad Cynddylan calls the anonymous king of Gwynedd to whom it was addressed Cadelling ffrau, terror of the Cadelling. This suggests that this part of North Wales was taken over by the kingdom of Gwynedd around the middle of the seventh century following the defeat of the Cadelling. Had the Cyndrwynin been the dominant dynasty of Powys during the decades before Winwęd, it becomes difficult to explain the loss of Dogfeiling to Gwynedd.
In this context, the claims of the Pillar of Eliseg at Valle Crucis become important. The inscription claims to have been set up by Concenn filius Cattell, Cyngen ap Cadell, recorded by the Annales Cambrie as dying in 854. According to the inscription, Cyngens great grandfather, Eliseg filius Guollauc, Elise ap Gwylog, recovered the Cadelling inheritance of Powys; the difficulties of the inscription, recorded in two different versions and now illegible, make it unclear whether he recovered it for nine (years?) from the power of the English (per viiii e potestate Anglorum) or with the help of the English (per vim et potestate Anglorum). Elise is undated, but we can calculate that Cyngen, his great grandson is unlikely to have been born much before 790; Cyngens father Cadell, who died in 808, would then not have been born before about 750, Cadells father Brochwel not before about 710 and Brochwels father Elise not before about 670. We have also seen how Selyf ap Cynan Garwyn may have been born around 590; his brother Eiludd would also have been born around the same time. Eiludds son Beli would not have been born much later than about 630; Belis son Gwylog would not have been born much later than about 670. Gwylogs son Elise would then have been born no later than about 710, placing the range of possible birth dates c 670×710. This suggests that Elise was at his prime during the first quarter of the eighth century, one of the main periods suggested for the construction of Wats Dyke.
What was the achievement of Elise, then? The implication of the Pillar is that Powys has not been in the control of the Cadelling and that Elise has recovered it, either from the control of the English (which is the usual interpretation), or that he has recovered it from an unnamed enemy from with the help of the English. This latter interpretation is worth exploring. During the third quarter of the seventh century, as we have seen, an unnamed king of Gwynedd (either Cadfael ap Cynfeddw or Cadwaladr ap Cadwallon) appears to have annexed Dogfeiling. Might the whole of Powys also have been brought under its control? That would explain the obscurity of the history of Powys during this period: it had effectively ceased to exist as an independent state. If Elise recovered his inheritance from Gwynedd (presumably from Cadwaladrs son Idwal Iwrch), it would not be surprising if it were with Mercian help, in which case there would be little point in building Wats Dyke. A date early in the reign of Ęthelbald is therefore improbable in the light of British politics.
The political history of north-west Mercia
The origins of the Mercian kingdom have always been obscure and are not relevant to the present discussion, since if Wats Dyke is of Mercian origin, it can only date from after the acquisition of Cheshire during the seventh century. During the second quarter of that century, two brothers, Penda and Eowa, established a strong kingdom based in the Trent valley. Bede (Historia Ecclesiastica ii.20) believed that Penda became king after the defeat of Eadwine, king of Deira, in 733, but the author of the Historia Brittonum (Chapter 65) seems to have believed that Penda did not become king until after the death of his brother at the Battle of Maes Cogwy/Maserfelth in 642. Both authors may have been trying to rationalise a much more confused situation: the roughly contemporary Tribal Hidage represents Mercia as a complex of tribal groups, some with apparently British names.
By the 640s, at any rate, Penda was in control of most of the English midlands. Earlier in his career, he had been allied with Cadwallon of Gwynedd and this alliance continued after Cadwallons death with his successor Cadfael; at the Battle of the Winwęd in 655, he was supported by thirty duces regii (royal leaders), sub-kings who included Ęthelhere of East Anglia and thilwald of Deira (Bede Historia Ecclesiastica iii.24) and possibly Cynddylan of Pengwern. Penda seems to have been recognised as some sort of de facto overlord in southern England, perhaps even over some of the Britons, although he was almost certainly the junior partner in the alliance with Cadwallon. Shropshire and Cheshire seem to have lain outside his kingdom proper. The Battle of Maes Cogwy/Maserfelth was fought in disputed territory (Bede Historia Ecclesiastica iii.9), perhaps in south Lancashire, between the Mercians (with their British allies) and Bernicians; thirteen years later, the Battle of the Winwęd was also in disputed territory near Leeds.
It has long been thought that Cheshire and Shropshire were the territory of the Wreocensętna, which ought to mean that north-western Mercia was effectively a British buffer zone, protecting the core territories of the Trent valley from attacks by unfriendly British states. The failure of Cynddylans family to retain his kingdom (as is implied in the Marwnad Cynddylan, The Lament for Cynddylan) may indicate that he was among the allies of Penda at the Battle of the Winwęd in 655 and that he lost his kingdom and his life there. This may have been the end of the semi-independence of the Wreocensęte. His family fled to Pendas former allies in Gwynedd. At the same time, as we have seen, Gwynedd seems to have acquired Dogfeiling in north-east Wales.
This was a turbulent time in Mercia. Pendas defeat at Winwęd was complete: not only was he killed, but his kingdom was partitioned between Oswiu of Bernicia and Peada, Pendas son and Oswius son-in-law, who was little more than a Northumbrian puppet. Peada was murdered within a year and in 658, an uprising of Mercian nobles ended Oswius direct rule and placed Wulfhere, Peadas brother, on the throne. Little is known of the events of Wulfheres reign; under his rule, Mercia seems to have recovered as the dominant power in southern England, but his campaign against Ecgfrith of Northumbria in 674 (Eddius Stephanus Vita Wilfridi xx) ended in defeat. He was succeeded by another brother, Ęthelręd I, in 675, the details of whose reign are also obscure.
However, it is in Ęthelręds reign that St Johns church in Chester is said to have been founded, in 689. Although the source for this statement is late (the fourteenth-century Chronicle of St Werburgs) and there are problems with identifying the bishop said to have been associated with its foundation, it has been widely believed. If we can believe the statement (and there is now archaeological evidence that may back it up), it shows that Chester was under Mercian control by the late 680s. This suggests that Cheshire was incorporated into Mercia either by Oswiu, following the Winwęd débācle, or as a result of an unknown campaign by Wulfhere or Ęthelręd I between c 655 and 689. This is therefore the earliest date at which Wats Dyke could have been built by the Mercian state. Unfortunately, we do not know whether Wulfhere and Ęthelręd I continued their fathers policy of alliance with Gwynedd; the reigns of Cadwaladr ap Cadwallon (after 655 - 682) and his son Idwal Iwrch (682 - c 712) are among the most obscure in the history of the kingdom. Nevertheless, given what appears to have been happening in Powys early in the eighth century, it appears very unlikely that Ęthelbald was responsible for Wats Dyke, despite the current consensus.
Wats Dyke is a neglected monument of early medieval archaeology in Britain. Long overshadowed by its larger neighbour, Offas Dyke (and, indeed, too often confused with it), it is nevertheless an impressive human achievement. At 65 km in length, its volume can be calculated as about 500,000 m3 of soil and turf (at least, this is the approximate volume of material removed from the ditch). If we allow, say, an average of twenty minutes to excavate a cubic metre of soil and to pile it up in a structured bank, this would give us a total of 166,666 man-hours or about 20,000 man-days to build the Dyke. One hundred men could then build the Dyke in a single year, allowing for time off to worship on Sundays and other Obligation Days. This seems a reasonable estimation.
However, the question that remains is in which year? We cannot at present date the work with any accuracy. The name of the Dyke is not helpful, although it would tend to suggest that the Mercians who named it were unsure of its origin and that they regarded it as belonging to an heroic age in the past; folklore of the twelfth century seems to have linked its mythical builder with Offa. Its position, on the other hand, may yield some important clues. Firstly, its northern end on the Dee Estuary and its completeness in this area makes it look as if it were designed to protect Chester, as Sir Cyril Fox noted. That being the case, it must belong to a period when Chester was of some strategic importance.
Can we determine when that might have been? The character of occupation of Roman Deva (Chester) changed dramatically after c 360, when the supply of new coins all but dried up. This means either that any people in government employ were no longer being paid and that taxes were no longer being collected there (as the second point is unthinkable, we can probably discount this suggestion), or that the placed ceased to have an official function (any remaining troops or civil servants being transferred elsewhere). The place seems to have become derelict fairly rapidly, although there are hints that someone of consequence was living in the Roman enclosure. In particular, the abandoned amphitheatre seems to have been the focus of sub-Roman activity. This activity remains small scale until the reign of Offa, when the first major new building since the fourth century seems to have been built. By the later eighth century, we can see an urban community establishing itself in the shell of the old Roman fortress. In other words, between c 360 and c 780, we know of nothing at Chester worth protecting, although future discoveries may well change this view.
There is also significant negative evidence. If Wats Dyke were late or sub-Roman, the main place in need of defence in the Marches would be Wroxeter, Roman Viroconium Cornoviorum, a city that continued to flourish almost up to the end of the sixth century. A defensive dyke lying entirely to the north-west of the city makes no strategic sense in terms of what we can see of the sub-Roman politics of the upper Severn valley. A date between c 400 and c 600 can be ruled out with some confidence.
The story therefore ends on a somewhat negative note. Wats Dyke is almost certainly not sub-Roman despite the statements being made on the back of a single radiocarbon date; Sir Cyril Foxs observation that it fits into a pattern of Mercian dyke building has much to recommend it. However, we cannot be so confident in ascribing it to Ęthelbald (716-757), as there is little reason to believe that there was anything at Chester worth defending early in the eighth century. Sir Frank Stenton considered the Welsh raids on Mercia during the reign of Cnręd (704-709) evidence that Wats Dyke did not yet exist, making it a construction of Ęthelbald (716-757). However, the documentary source for these raids (Felixs Vita Sancti Guthlaci) does not mentions from which Welsh kingdom the raids had come: Wats Dyke affords protection only against raids from Gwynedd, not from any other kingdom.
On the other hand, the context of the acquisition of Cheshire by the Anglo-Saxons, probably at some time around the middle of the seventh century, might prove an attractive option. Penda had been allied in turn with Cadwallon and Cadfael (kings of Gwynedd c 625-634 and c 634-c 655), so north-west Mercia was secure until his death. He was supplanted, briefly, in Mercia by Oswiu (King of Bernicia 642-670), an enemy of Gwynedd; when his son Wulfhere was restored to the Mercian throne in 658, he seems not to have returned to Pendas alliance with Gwynedd. Instead, it was probably during his reign that Cheshire became Mercian, by a process that is not known but which presumably involved military action. This is also the most likely date for the hidation of the county, as discussed above. Might we push the date of Wats Dyke back to his reign or that of his brother and successor, Ęthelręd I (675-704)? In this case, it is not Chester that Wats Dyke was intended to protect, but Cheshire, the newly acquired territory.
 Palmer, A N Offa's and Wat's
dykes. Y Cymmrodor 12 (1897), 65-86,
Wats Dyke: a North Welsh linear boundary is copyright ©2001 Keith Fitzpatrick-Matthews, used with permission.
Comments to: Keith Matthews
Wansdyke Project 21 are copyright © Robert Vermaat 1999-2007.