THROUGH Domesday we can see that there was in 1086 a resurgence of English
and Norman settlement along almost the entire length of the Welsh frontier.
Forests and underbrush were being cleared, and long-abandoned manors were being
brought back into productivity. This medieval frontier boom was not a spontaneous
process, however, but appears to have been inspired and directed from above. The
frontier prosperity of 1086 gives every appearance of having been, for the most
part, an artificial development. It was also a selective process. While some
areas were the scene of intensive activity, others were neglected or even
abandoned. The favored lands were not always distinguished by their fertility or
pleasant location. Their primary value lay in their strategic importance or their
usefulness in consolidating or improving the holdings of those Norman lords who
had emerged as the dynamic force along the frontier.
It was in the direct interests of these lords to populate and develop their
estates as quickly as possible. This development would have served the double
purpose of increasing their revenues and at the same time providing enough
fighting men to render the task of frontier defense much easier. In view of such
considerations, it is not surprising that such men as the riding men, bovarii,
and hospites inhabited the border. What is surprising is that there were so few
The radmanni, radchenistri, bovarii, and hospites of Domesday Book numbered only slightly over twelve hundred men. It is clear
The Domesday Frontier 63
that these groups were far too few to have accomplished the extensive
reconstruction which was underway in 1086.
An immense task had faced the border shires after peace had been secured in the
area. The raids of the Welsh and the forays of Edric the Wild had left large
numbers of estates in ruins. In Herefordshire, for example, Domesday reported
that fifty-two vills completely lacked the teams or men to till their land in
1066. They were, in Domesday's terminology, wasta. Fifteen other
estates were inhabited but had suffered extensive damage and were considered as
partly waste.1 This picture must have worsened in the period after
1066, when the attacks of Edric caused widespread devastation in the northern
part of the shire. How great this destruction might have been is difficult to
say, but it is certain that additional vills were laid waste, especially in those
areas lying near Wigmore and Richard's Castle.
A great deal of progress had been made by 1086. The number of completely waste
estates had been reduced to thirty-four. This meant that at least eighteen
estates, and perhaps a goodly number more, had been returned to full production.
This process was more striking than the statistics would indicate since the waste
estates tended to be concentrated in a relatively restricted area. The heavily
devastated northwest of the shire had completely recovered, and there were no
waste estates in the vicinity of Wigmore. Recovery was almost complete in the
area lying between the Dore and Wye rivers. Some waste still existed, but by far
the greater number of vills had been returned to cultivation.
Domesday makes it clear that a great amount of capital and a large number
of men were succeeding in repairing the damage the shire had suffered. Both men
and capital, however, were curiously concentrated in two favored areas-the
northwest and the region between the Dore and the Wye.3 If this
process of recovery had been unconscious or undirected, one would expect that the
more fertile and protected lands in the eastern part of the shire would have been
favored. They were not. This concentration of men and capital suggests that
1The Domesday Geography of Midland England, ed. H. C.
Darby, fig. 32, p. 95.
2Ibid., fig. 33, p. 96.
3Ibid., fig. 34, p. 97. For a similar treatment of other shires, see
the other works of this series.
64 The Normans in South Wales
the process of recovery, in Herefordshire at least, represented a conscious
effort on the part of those in authority.
This realization raises some serious questions. Where and how did the Normans
acquire the manpower and plough teams which were used in this redevelopment
program? We have seen how the frontier protected the riding men and attracted the
hospites, but neither of these groups were large enough to have contributed
significantly to the work that was being accomplished. Again, why did the Norman
lords employ this curiously selective approach? Prosperity was not general in
Domesday Herefordshire. To the contrary, a comparison of the waste vills of 1066
with those of 1086 shows that the greater part of the shire was still distinctly
underprivileged. Not only had there been no progress made toward repairing the
damage of 1066, but in some areas the number of waste estates had actually
increased. This brings us to a third question. From where had these newly waste
estates come? It is possible to explain many as the possible result of Welsh
raiding.4 Others, such as those in the relatively well-protected east
of the shire cannot be so explained. It is not at all certain, or even probable,
that these estates, or even a significant proportion of them, were laid waste as
the result of military activity. But even if this were the case, why were these
vills, especially those in the hands of the king -immediate overlord of
Herefordshire since 1075- left out of the program of redevelopment that was
These three questions are not unrelated. The obvious inference to be drawn is
that men and capital were being drained from some regions of the shire in order
to be concentrated in others. This suspicion is confirmed by the distribution of
slaves as recorded by Domesday Book.5 In the extreme east
of the shire, servile elements formed about 25 per cent of the total population.
In the fertile central plains, this figure decreased sharply to between 10 and 15
per cent. In the far western part of the shire, along the border but still under
direct Norman control, this trend was reversed, and slaves formed between 15 and
20 per cent of the frontier population.
It seems likely that this abnormal concentration of slaves was
4Ibid , p. 98.
5Ibid., fig. 152, p. 429. These figures avoid the question of whether
the slaves mentioned in Domesday Book represented the heads of
families as did other statuses mentioned. If they should be so regarded, another
problem arises. There would be a tendency for unmarried slaves to be concentrated
in the dangerous frontier region. This fact would alter the relative percentages
The Domesday Frontier 65
caused by their importation, probably from other parts of the shire. The newly
waste vills which had appeared by 1086 had not been devastated but abandoned, and
their inhabitants used to develop other estates. The slaves were the first to go,
since they were completely under the control of their masters and lacked the
recourse to customary rights which was open to other classes of society. Other
groups must have followed after the slaves had accomplished the first and most
difficult labor. Indeed, the very presence of the slaves would have encouraged
such immigration. Where slaves, or such men as the bovarii, were available
to till the demesne, other classes were spared many burdensome duties which
otherwise fell on them. It may well be that reduction of such obligations was
offered as an inducement to immigration. In any event, later evidence indicates
that the villeins of the border were not generally liable to many of the works
demanded on estates elsewhere in England. Even their comparatively light duties
were replaced by money payments at quite an early dated
It is clear that a general and consistent policy was being pursued in 1086 -one
designed to redistribute the population of the shire and to direct immigrants
into certain areas which had been given a high priority in the process of
redevelopment. It is likely that surplus population was being drawn from all the
surrounding region. Some areas, however, were contributing an inordinately large
number of such immigrants, so great as to reduce their own productiveness. The
distribution of waste vills in 1086 shows that the main area so exploited
consisted of those estates which the king held in the vicinity of Radnor. It is
not surprising that the king allowed or encouraged this, since it was in his
direct interest to do so. The security of the interior of the shire and the
efficiency of its system of border defense demanded that the border lords be able
to develop their frontier estates as quickly as possible. Royal grants of estates
along the border were meaningless unless the king provided his vassals the means
to defend them against the Welsh. He therefore drew the necessary manpower from
his estates in the vicinity, and sacrificed his immediate interests in favor of
longer range considerations.
The evidence of Domesday Book is tangled and often inconclusive
6W. J. Slack, "The Shropshire Ploughmen of Domesday Book," The
Transactions of the Shropshire Archaeological Society, L ( 1939 ), p. 34.
7An analogous situation apparently existed in Yorkshire. See T. A. M.
Bishop "The Norman Settlement of Yorkshire," Studies in Medieval History
Presented to Frederick Maurice Powicke, eds. H. W. Hunt et al., pp.
66 The Normans in South Wales
and the work that remains to be done to analyze it is immense. The bulk of the
evidence, however, points to the conclusion that there existed in 1086 a
conscious and effective program on the part of the government to develop the
region lying along the border and to encourage immigration into this
strategically important area. From Chester to Chepstow, the Welsh frontier was
being transformed into an intricate defensive network.
The central portion of this line was under the command of Roger of Montgomery,
earl of Shrewsbury and of Arundel. He was the scion of one of the greatest of the
noble families of Normandy and his family connections were rather mixed. King
William, William Fitz-Osbern, and Ralph of Mortimer were all his cousins. His
brothers, on the other hand, were turbulent and disorderly and had brought no
credit to the family name. one of them, William, had plotted the assassination of
Duke William and had killed Fitz-Osbern's father in the attempt.8
Roger himself was a firm supporter of Duke William and enjoyed his favor and
confidence.9 Roger's possessions included Montgomery and L'Hiemois,
and, through marriage, Bellêum;me, Alençon, and Séez. He was
active in support of the invasion of England and distinguished himself at the
battle of Hastings. In 1071 he was granted the earldom of Shrewsbury on the
western frontier of England. A strong base had already been established in
Chester to the north, while to the south William Fitz-Osbern had by this time
completed the pacification and organization of his earldom of Hereford. Roger was
obviously expected to close up the last gap in a strong western border defense
In accordance with the royal frontier policy at the time, Shrewsbury was given
what amounted to the status of a county palatine. With certain minor exceptions
the entire county of Shropshire was held by Roger as tenant in chief and he
enjoyed a wide grant of power. He distributed much of his lands to his followers,
and the brunt of the defense of the border naturally fell on them. Prominent
among this number was the sheriff. Roger was himself the vicecomes of Oximin in
Normandy and was apparently fully convinced of the value of a strong and loyal
sheriff. He accordingly endowed the office with over seventy manors within
Shropshire and some as far afield as Sussex.
8See chapter II above.
9Wace, Maistre Wace's Roman de Rou et des Ducs de Normandie,
nach den Handschriften, ed. H. Andresen, 11. 4415 ff., p. 207.
The Domesday Frontier 67
The first occupant of this important position was Warin the Bald, whose
allegiance had been heightened by his marriage to Roger's niece, Amiera.10 If the
praise which Orderic Vitalis heaps upon him was valid, Warin amply justified
Roger's faith in him.11 By the time of Domesday, however, he
was dead, and his position, both as sheriff and as Amiera's husband, was occupied
by a certain Rainault.12 R. W. Eyton, one of the foremost historians
of Shropshire, identified his place of origin as Bailleul-en-Gouffern, which he
had held as a fief, "... under Roger de Montgomery, when he was called to fill
the more important position of Sheriff of Shropshire."13
His new holdings were extensive and somewhat scattered, as we have said. They
were concentrated, however, in the hundred of Mersete in the northwestern part of
Shropshire. Over 80 per cent of the hundred lay in the sheriff's
hands,14 and he had pushed forward the construction of fortifications
in the area.15 The most important of these was the massive work
located at Oswestry, which became the key to the frontier defense of northern
Shropshire. From this base of power Rainault was able to extend his power
westward into Wales, although he did not succeed in establishing any settlements
there. Domesday records that two Welsh commotes paid Rainault an annual render of
eighty shillings and eight cows as ferm.16
Immediately to the south of Rainault's holdings in Mersete lay the hundred of
Ruesset and the estates of the two Fitz-Corbet brothers, Roger and
Robert.17 Although they held estates scattered over Conodovre and
Rinlau hundreds, the main body of their holdings lay in Ruesset and Witentreu,
immediately to the south.18 These lands had
10R. W. Eyton, Antiquities of Shropshire, VII, 203.
11Orderic Vitalis, "Historiae Ecclesiasticae libri XIII in partes tres
divisi," Patrologia Latina, ed. J. P. Migne, vol. CLXXXVIII, col.
12Eyton, Antiquities of Shropshire, VII, 205-206. Eyton
holds that Rainault was merely exercising the post during the minority of Warin's
son, Hugh, to whom Rainault transferred the office and its lands sometime after
13Eyton, Antiquities of Shropshire, VII, 206.
14J. Tait, "Introduction to the Shropshire Domesday Book," The
Victoria Histories of the Counties of England: Shropshire, I, 296.
16Domesday Book: or The Great Survey of England by William the
Conqueror A.D. MLXXXVI, fol. 253b.
16Ibid., fol. 255. The territories of "Chenlei" and "Derniou" probably
represent Cynllaith and Edeyrnion.
17For this important family of border barons, see the Duchess of
Cleveland, The Battle Abbey Roll: with Some Account of the Norman
Lineages, I, 219-223.
18See the Domesday Map of Shropshire in The Victoria Histories
of the Counties of England: Shropshire, I, facing p. 309.
68 The Normans in South Wales
suffered extensively in the disorders which had followed the death of Edward the
Confessor. Many, if not most, of the manors of the area showed a considerable
decrease in value between 1066 and their acquisition by the
Fitz-Corbets.19 This valuation had risen remarkably by 1086 however.
Domesday does not mention that any fortifications had been raised in Ruesset.
Roger seems to have taken steps to remedy this lack, for Cause Castle was soon
erected on a high hill near his estate of Alretone.20 It seems hardly
likely, however, that this was the first fortification in the area.
Domesday reports that Alretone was occupied by five knights, all vassals
of Roger.21 It is probable that this concentration of military tenants
was part of Roger's preparations for the erection of his great castle. A similar
concentration of knights was found at the manor of Wrdine (Worthen).22
It seems more than likely that some sort of fortification also existed here.
To the south of the Fitz-Corbets' holdings, in the hundred of Witentreu, lay the
key to the defense of the middle march. Here Earl Roger had constructed his great
castle of Montgomery just west of Offa's Dyke. The castle itself was surrounded
by over fifty hides of waste land, but not too far away lay eight of Roger
Fitz-Corbet's manors. Some agriculture was practiced on these estates, and it may
well be that these supplied the immediate needs of the castle's garrison. Three
fisheries, an animal trap, and some woods for pannage was the Domesday
summary of the economy of this rather desolate, but strategically vital
South of Witentreu lay the mountainous hundred of Rinlau. The Domesday manors of
this region were restricted to some few which lay in the valley which the river
Clun has cut deeply into the highlands of the forest of Clun. Here too were
located the estates of Robert of Sai, or Picot as he was apparently always called
in England.24 Picot's home lay in Argentan, in Roger Montgomery's
vicecounty of L'Hiemois. It seems probable that, like Rainault, Picot had
19Domesday Book, fols. 255-256.
20 Tait, "Introduction to the Shropshire Domesday Book," p. 297, and
Eyton, Antiquities of Shropshire, VII, 6.
21Domesday Book, fol. 253b.
22Ibid., fol. 255b.
23Ibid., fol. 254.
24Tait, "Introduction to the Shropshire Domesday Book," p. 297, and
Eyton, Antiquities of Shropshire, XI, 225.
The Domesday Frontier 69
been brought to the border at the command of his lord.25 Despite their
isolated location, Picot's thirteen manors were well-to-do and yielded about
twenty-five pounds annually.26 They had been worth a good deal more in
1066 but had obviously suffered considerably in the aftermath of the Conquest.
Under Picot's management, however, their value had risen greatly by 1086. The
center of this compact group of estates lay at the manor of Clun, a vill lying at
the fork of the river. A large castle was built here subsequent to the Domesday
survey. It seems probable that some fortifications already existed, for there was
a group of military vassals established here quite similar to those found at
Alretone and Wrdine. This settlement at Clun represented the deepest penetration
of Norman settlement into the Welsh highlands.
To the south of the estates of Picot lay the Herefordshire hundred of Hezetre.
The river Teme ran through this hundred in a sharp arc, cutting across an
eastward tongue of the Welsh uplands and forming an extensive pocket of fertile
soil. The Teme, together with the Lugg, reached deeply into the Welsh highlands
and formed a major route of access from the central plateau of Wales into the
heart of Herefordshire. A high and rocky ridge lay along the Teme. A natural
cleft ran across the breadth of this ridge, isolating its eastern end and making
it a position of great defensive strength. A mound was thrown up on this eastern
end, and a timber fortification was built at the command of William Fitz-Osbern.
This small fort, replaced by a stone structure early in the next century, became
the strong point for the defense of northwestern Herefordshire.
Under the name of Wigmore Castle, it also formed the heart of the holdings of
Ralph of Mortimer, a cousin of the earl of Shrewsbury.27 Near the
castle lay a flourishing borough, which paid its lord an annual render of seven
pounds. This was not an insignificant sum, considering the liberal terms and low
rents the burgesses enjoyed. Near the town and along the banks of the Teme were
located a number of manors. These all lay in Ralph's hands, and gave a general
air of prosperity and growth. Just to the west lay several manors which had
25The Duchess of Cleveland, however, identifies Picot with Wace's "Cil
de Saie," one of the original Conquerors. The Battle Abbey Roll,
26Domesday Book, fol. 258.
27For Wigmore, see G. T. Clark, "Wigmore," Archaeologia
Cambrensis, Series IV, Vol. V (1874), pp. 97-109.
70 The Normans in South Wales
been held by English thegns under Edward the Confessor, but were now
apparently deserted. There is no indication that any effort had yet been made to
bring them back under cultivation, and it seems likely that they were mainly used
From these waste manors, a great swath of abandoned estates swept southward along
what is now the English-Welsh border, centering on a large group of royal manors
around Radnor. Domesday Book mentions no activity at all in the area
except for the fact that Osbern Fitz-Richard claimed eleven manors "on the
marches of Wales" which he used only for hunting.23
Just to the east of this deserted region lay the hundred of Elsedune, where we
can see the basic processes of frontier development well underway. A number of
lords held estates in this hundred, but the primary landholder in the area was
Roger of Lacy, who held five Elsedune estates from the king and one from the
church of St. Guthlac.29 Progress and intense activity were apparent
on all of these manors. The first thing that strikes our attention is the
frequent mention of hospites. These settlers were found on no less than three of
these six estates. New lands were being opened up throughout the area. In the
adjacent hundred of Stradford lay the estates of Wibelai and Fernehalle, also in
Roger's hands. In the passages describing these estates are found two of the four
Domesday entries mentioning assarts.30 This process of agricultural
expansion was not restricted to Roger of Lacy's estates. In describing the manor
of Wrdeslege, held by Gruffydd ap Maredudd, Domesday states that the manor "was
waste and still is, except for three acres of land which have been recently
cultivated here."31 It is also clear that labor was in high demand in
Elsedune. Roger's manor of Elmelie was being tilled by men from another village,
who paid a considerable sum for the privilege. Elsedune Hundred displayed many
characteristics of what can best be described as a frontier boom.
Elsedune also shows the early stages of the process which made the marches of
Wales a land of castles. Domesday Book records no massive
fortifications in the hundred, but does mention the existence of
28Domesday Book, fol. 186b.
29Ibid., fols. 184b and 182.
30 Ibid., fol. 184b. It is interesting to note that the other two
entries also occur in Herefordshire; on the estates of Marcle (fol. 179b) and
Leominster (fol. 180).
31Ibid., fol. 187b.
The Domesday Frontier 71
two peculiar establishments each known as a "domus defensabilis." One of these
was located at Herdeslege and was held from Roger of Lacy by a certain Robert.
The land on which it was situated paid no geld, paid no customary dues, and
formed an independent region, not a part of Elsedune or any other hundred. It lay
in the heart of a deep forest and its population consisted of two serfs and a
Welshman.33 After the time of Domesday, this small establishment was
to grow into the forbidding pile of Eardisley Castle.33 The second
domus defensabilis was held by Gilbert Fitz-Turold at Walelege, a manor
which had been given to him by William Fitz-Osbern. Only two of the four hides
here were geldable. There appears to have been no agricultural activity, and like
Herdeslege it was situated in a dense forest.34 No definite record can
be found of the later history of this holding, but it seems likely that Walelege
is identical with Lenmore Mount, locally reputed to have been the site of a
strong castle.35 It is easy to see in these two small establishments
the earliest stages of the process which produced the permanent fortresses that
eventually secured the Welsh frontier.
Just to the south of Elsedune Hundred lies the valley of the Wye which here
descends from the alpine meadows of Brecknockshire and passes into the heart of
Herefordshire. This was a prime invasion route and it was guarded by the great
castle of Clifford, constructed by William Fitz-Osbern. At the time of Domesday,
the castle was in the hands of Ralph of Todeny, brother-in-law of the late earl.
Clifford was virtually a sovereign realm. Like Herdeslege, it belonged to no
hundred nor did it pay any customary dues. All the tenants of the area held
directly from Ralph.36 These tenants numbered four, named by
Domesday Book as Gilbert, Roger, Herbert, and Drew. It is not
difficult to identify Gilbert Fitz-Turold, Roger of Lacy, and Drew Fitz-Pons,
three of the greatest landholders of the shire. Between the three of them,
excluding Herbert, they held nine ploughs in demesne and had established the
nucleus of a borough which at the time claimed sixteen burgesses.
Ralph himself appears to have taken little interest in Clifford. He
32Ibid., fol. 182b.
33R. H. Warner, "Eardisley and Its Castle," The Transactions of
the Woolhope Naturalists' Field Club for 1903, pp. 256-262.
34Domesday Book, fol. 187.
35I. C. Gould, "Ancient Earthworks," The Victoria Histories of
the Counties of England: Hereford, I, 226-227.
36Domesday Book, fol. 183.
72 The Normans in South Wales
held nothing there but one plough and apparently had no men to work it. There was
a mill in the castle, but it was not subject to him and was probably not built at
his order. As a matter of fact, Ralph had farmed the entire castle to Gilbert,
the sheriff of Herefordshire, for an annual render of sixty shillings. Of all of
Ralph's extensive holdings,37 he had chosen to fix his seat at a manor
he held in Hertfordshire, rather than in the castle which formed the obvious
caput of his most extensive holdings. As well as we can judge from Domesday
Book, Ralph of Todeny took neither interest nor initiative in the
development of this major marcher fortress. The credit for improving and
garrisoning Clifford, and for guarding this important approach to Herefordshire,
must remain with the small group of barons who held of Ralph.
Stretching southeasterly from Clifford flows the river Dore, along a path roughly
parallel to, and between, the Wye and the Welsh border. For the first fifteen
miles of its length it runs through the narrow valley it has cut in the uplands
which tower as much as three hundred feet on either side of the stream. This was
the "Golden Valley" into which the English had been pushing since before the time
of the Conquest. It was now entirely in Norman hands, but lay exposed to attack
from the uplands which stretched along its western bank. The upper valley was
controlled by the same Gilbert who held Clifford. The land was almost completely
undeveloped and offered Gilbert little revenue except the hawk and two dogs he
annually received from the eight Welshmen who were settled on the manor of
Bach.33 Subsequent entries in Domesday Book credit the sheriff with
holdings totaling fifty-six hides in this fertile valley.39
Immediately to the south of Gilbert's manors lay the lands which William
Fitz-Osbern had granted to Hugh L'Asne.40 The center of these holdings
was the castle of Snodhill, commanding the northern entrance to the
valley.41 A minor castle was located nearby at Urishay
37He held estates in Berkeshire, Herefordshire, Gloucestershire, and
38Domesday Book, fol. 187. Also see G. Marshall, "The
Norman Occupation of the Lands of the Golden Valley, Ewyas, and Clifford and
Their Motte and Bailey Castles," The Transactions of the Woolhope
Naturalists' Field Club for 1936-1938, pp. 141-158.
39Domesday Book, fol. 187.
40Ibid., fol. 180b.
41J. H. Round, "Introduction to the Herefordshire Domesday," The
Victoria Histories of the Counties of England: Hereford, I, 276.
The Domesday Frontier 73
and acted as an outlier to the major fortress.42 Three of Hugh's five
Golden Valley estates were waste, but the remaining two showed considerable
activity. South of these manors lay the holdings of a number of other lords,
prominent among them being Roger of Lacy and an Englishman, Alfred of
It is difficult to derive from Domesday Book any complete picture of
the state of affairs in the Golden Valley. Certain things can be conjectured.
Frenchmen and riding men were both common in the valley, and must have made it
attractive to the developers because they were able to insure not only
agricultural development, but at least some defensive strength. Border were
attracted by the promise of steady work. one manor records two ploughs in
demesne, six bordarii and nothing else.43 The industrious
landholder could expect excellent profits from well-developed estates. One of
Alfred of Marlborough's estates made an annual render of three
pounds.44 The Golden Valley was fertile, underdeveloped, and
strategically important. The men and capital moving into the area would soon plug
this gap in Herefordshire's western defenses.
The southern entrance to the Golden Valley was guarded by the castle of Ewyas
Harold, perched on a rocky ridge overlooking Dulas Brook.45 It
dominated not only the confluence of the Dulas and the Dore, but also the point
where they joined the Monnow about a mile south. This position was of the
greatest strategic importance in that, in addition to guarding the valley of the
Dore, Ewyas Harold also protected the valley of the Monnow, one of the great
invasion routes leading from the highlands of the Black Mountains into the heart
of Herefordshire. So important was this position that it is not surprising to
note that a castle had existed there before the Conquest, built by the early
colony of Herefordshire Normans.46
One of the tenants at Ewyas Harold was the same Roger of Lacy whom we have
already encountered at Elsedune, Clifford, and the upper Dore Valley. He seems to
have been just as active here as he was elsewhere. At any rate, Domesday
Book reports that, like Rainault in Shropshire, he had managed to
establish control over a Welsh
42 Gould, "Ancient Earthworks," pp. 244 and 254.
43Domesday Book, fol. 184.
44 Ibid., fol. 186.
45 G. T. Clark, "The Castle of Ewyas Harold," Archaeologia
Cambrensis, Series IV, Vol. VIII ( 1877) pp. 116-124.
43Domesday Book, fol. 186. The castle had been
refortified at William Fitz-Osbern's Command.
74 The Normans in South Wales
commote just to the west of his holdings. The degree of this control is
uncertain; Domesday merely states that he was to receive an annual render of
fifteen sestiers of honey and fifteen swine, and was to have the right of pleas
over the inhabitants of the area "when the men are there." This last phrase is
vague, but it may refer to the semi-nomadic customs of the free tribesmen. In any
event, it is important to note that Roger had made no attempt to plant a colony
here. As we have seen, he had no men to spare even in Elsedune. He seems simply
to have attempted to gain power and revenue in the area. It may well be that he
was simply assuming the position of the local tywysog.
To the south of Ewyas Harold and the commute of Ewyas Lacy lay yet another Welsh
commote under Norman control. This region, Erging. or Archenfield, had been taken
long before by the Anglo-Saxons. Domesday Book is vague and
incomplete about its status under the Normans. It is clear, however, that no
attempt had been made to absorb or displace the Welsh who inhabited the area.
They were all regarded as king's men and had certain customary dues and
obligations. Beyond this nothing is known of the state of affairs within the
area. A strong castle, Goodrich, lay on its borders, and had apparently been
there in 1066.47 Although its original purpose had probably been to
pacify the Welsh of Archenfield, it also commanded a vital ford on the old Roman
road leading to Monmouth Castle.
Monmouth and Strigoil completed the castle system which stretched clear across
the Welsh peninsula. The two were quite similar in a number of important
respects: both had been constructed by William Fitz-Osbern, both had, or were
soon to have, flourishing boroughs, and both commanded important fords in the
Roman road system which linked England with southern Wales. Although both
locations were important links in the Herefordshire defense system, Strigoil (or
Chepstow) is the more deserving of interest, for it was from here that the
Normans made their first permanent penetration of Wales.
At the beginning of the Domesday entries for Gloucestershire there is a compact
series describing the settlements which the Nor-
47Domesday Book, fol. 181. Rev. Prebendary Seaton,
"History of Goodrich," The Transactions of the Woolhope Naturalists' Field
Club for 1902-1902, pp. 212-225
The Domesday Frontier 75
mans had made in lower Gwent.48 On the eastern edge of this region lay
the fortress and town of Chepstow, on the western the small colony which William
of Scohies had established at Caerleon. The northern limit of settlement was
probably defined by those highlands now known as Went Wood. The northernmost
place name mentioned is that of Lamecare, which is most likely to be identified
with Llanvair Discoed.49 The southern limit lay along the marshes
which stretched south of the line now followed by the railway between Chepstow
and Newport. More than likely, this now-fertile region was uninhabited, since the
place names of the area are, without exception, non-Celtic. It should be noted
that the general area of Norman settlement did not form the spreading wave
pattern which might be expected from an undirected advance into a frontier
region. It rather formed an elongated westerly thrust, the axis of which was
formed by the old Roman road which linked Chepstow and Caerleon.50
Within this long and narrow area, the patterns of landholding were sharply
differentiated. Generally speaking, the lands of the western portion lay in the
hands of the king, and the eastern portion, consisting of lands lying close to
Chepstow, was under individual Norman lords. On the royal lands Welsh
arrangements were continued with little change. Under Welsh tribal law,
homesteads were formed into groups of twelve for the purpose of making food
renders to the tywysog. Domesday Book describes what appears to be a
direct continuation of the system under the overlordship of the Norman king. The
entry notes that four "praepositi," Waswick, Elmui, Bleius, and Idhel, were
responsible for groups of thirteen, fourteen, thirteen, and fourteen vills
respectively. Not only was the pre-Conquest Welsh system retained, but it would
seem that the native Welsh maers were continued in office. In addition to this
Welsh group, King William held a few hardiwicks, or hamlets. These estates,
located at the eastern end of the royal holdings in Gwent, were organized more
along the usual manorial lines. Finally, a number of individuals held land from
the king in the area. These lands were held freely and without
48Domesday Book, fol. 162.
49C. S. Taylor, "The Norman Settlement of Gloucestershire," The
Transactions of the Bristol and Gloucestershire Archaeological Society for
1917, p. 82.
50This road is now followed, along most of its course, by the A48
76 The Normans in South Wales
dues. These men represented a legacy from William Fitz-Osbern's policy of
conciliating individuals of dubious loyalty by granting them land, on liberal
terms, on the frontier.51
The western end of Norman holdings in Gwent was anchored by the castle of
Caerleon, located on the western bank of the Usk at the point where the old Roman
road forded the stream. This location was one of the most strategically important
in southeastern Wales and long before had been the site of the legionary fortress
of Isca. The sea was readily accessible from the castle, and a series of Roman
roads converged upon the site. These roads, still the best available to the
Normans, led west into the vale of Glamorgan, north along the valley of the Usk,
and east to Chepstow.52 The settlement which supported the castle was
still quite small in 1086. William of Scohies, one of the more important
landholders of Herefordshire, held an estate of eight carucates to the west of
the river, and these were held of him by Turstin Fitz-Rolf. There were two serfs
and one plough operating on the demesne lands. In addition, three Welshmen had
three ploughs and were allowed to continue their Welsh customs (leges Walensi
viventes). This estate, despite its small size, was valued at the
considerable sum of forty shillings. Of lands lying to the north and west of this
tiny settlement Domesday Book is silent. Apparently, at Caerleon one
had reached the very edge of the frontier.
Domesday Book has afforded us an almost uniquely detailed picture of
a medieval frontier in operation. Although the account is in many ways incomplete
and fragmentary, it is possible to observe many of the processes which we
associate with frontier development. Along the frontier an abundance of land was
coupled with a lack of manpower. The underdeveloped resources of the region
awaited the energies of the exploiter. At the same time, the government was
directly interested in having this strategically important area developed and
populated as quickly as possible. Under these conditions it is not surprising to
see that a frontier boom was underway along the Welsh border in 1086. Land values
were rising, population was increasing, and the military security of the region
was being assured through the construction of extensive fortifications. Moreover,
all of the classes who participated in this process were receiving substantial
rewards for their industry.
51Domesday Book, fol. 162a.
52R. E. M. Wheeler, Prehistoric and Roman Wales, pp.
The Domesday Frontier 77
These rewards were not equal, however, and the greatest were reserved for those
barons who took up the task of frontier defense. Along the frontier the nobility
gained an independence and freedom denied to the barons of the more settled and
secure regions of England. The holdings of the border barons were concentrated
into relatively compact units, rather than being widely scattered as elsewhere in
England. This was true all along the frontier, but was most apparent in
Shropshire. This appears to have been the result of a conscious action by the
Normans, since such concentrations of holdings were not characteristic of the
pre-Conquest Anglo-Saxon system.53 The military advantages of such a
system are obvious, but it should also be noted that this arrangement afforded
the border barons advantages which William was loath to grant his other
In both England and Normandy William made constant effort to limit the building
of private castles by his nobility. Along the border, however, the construction
of such castles was in the royal interests, since a border defense line could be
built in this way with a minimum of royal expenditure. Thus we see numerous
examples of castle-building underway along the border. The details of this
process are reasonably clear. A certain area was set aside and freed of normal
dues. A domus defensabilis was then constructed to defend the area while initial
development was made. A group of military vassals were moved in as soon as the
estate could support them. In the course of time a timber fort was constructed
out of the materials which lay closest to hand. Finally, masonry replaced the
timber, and the castle was complete. With his compact holdings, and his own
castle complete with private army, the border baron was a man of considerable
power, and might, if he so wish, defy even the king.
Thus it is clear that the nobility found power and some degree of independence
along the frontier. Even greater opportunities awaited them across the border,
and Domesday Book indicates that at least some of them were taking
advantage of them. From their bases of power in England, the marcher lords were
slowly bringing Welsh border districts under their personal control. They were
not attempting to colonize these areas, but only to control and exploit them.
Colonization eventually followed, but it is clear that the leading edge of the
frontier consisted of the border barons and their followers, who were striving to
extend the peculiar feudal system of the Welsh
53Tait, "Introduction to the Shropshire Domesday Book," p. 299.
78 The Normans in South Wales
marches westward into Wales proper. Herein lay a great difference between the
Welsh frontier process and the American version with which we are perhaps more
familiar. In the American frontier, the dynamic element of colonization and
exploitation was the individual-trapper, trader, miner, or farmer. As a
consequence the line of settlement moved westward at a faster pace than did
political organization.54 In medieval Wales, however, this order of
precedence was reversed. The feudal noble, the embodiment of the political system
of the time, provided the active force for westward expansion, and the progress
of political organization constantly tended to outstrip that of actual
settlement. The results of this peculiarity were to be of some consequence in the
development of Norman South Wales.
54Land speculation, of course, represents something of an exception to
this rule. The normal political institutions of the time-civil law,
representative government, and the like-were generally late arrivals to a given