Preface

The ideal book on Herefordshire place-names has yet to be written. It would need, in its author, the knowledge of phonological laws possessed by Professor Wyld, the Celtic scholarship of the late Sir John Rhys, and Dr Horace Round's familiarity with the 11th and 12th centuries. The present writer, having no claim to rival any of these, has tried, nevertheless, to build upon their foundations.

The place-names of our county, where English, Welsh, and Norman-French influences acted and re-acted upon one another for centuries, ought to throw some valuable side-lights on our history. I have, therefore, in the scanty leisure hours of the last few years, put together the following list of practically all the early forms of Herefordshire names. The work has been laborious, but to me most interesting; and I trust it may be of interest and assistance to other students.

I have to acknowledge the kind and ready help of several scholars whom I have consulted on difficult points. In particular I am bound to mention Dr Horace Round, Mr Egerton Phillimore, and the Rev. J. B. Johnston. I have felt much diffidence in regard to the considerable number of Celtic names; but the Rev. T. Gray Jones, of Ebbw Vale, most kindly placed his intimate knowledge of Welsh at my service, and I have dealt with them as best I could, with a fearful remembrance of Dr Bradley's scathing remark that 'it would be just as reasonable to try to read Virgil by means of a French dictionary and no grammar, as to try to translate ancient British names by means of a Welsh dictionary'. Fortunately for me, no great number of Herefordshire place-names are really 'ancient British', and for most of these I have been able to find expert knowledge in the valuable notes of Dr Henry Owen and Mr Egerton Phillimore to 'Owen's Pembrokeshire'. The larger number of Welsh place-names in Herefordshire are comparatively modern, as I have shown in the Introduction. As regards either Welsh or English names I shall gratefully welcome suggestions and corrections.

A.T.B.
January 1916.

Introduction

'Much of our history that is still dark', says Dr Round, 'is written in the names that our remote forefathers gave to their English homes'. These names in the 18th century, and indeed through most even of the 19th, were looked upon as a fit subject for the wild and ignorant guess-work which has filled our books of antiquities and our county histories with many misleading theories. It is scarcely too much to say that no work on English place-names has any scientific value before 1901, when Professor Skeat introduced modern methods of investigation by his Place-names of Cambridgeshire.

The new scientific treatment of place-names is a simple application of common-sense to the subject:- to ask, first, what are the earliest forms of the name ? next, can any meaning be attached to the earliest form ? and, lastly, how has this early form developed into the present-day name ? Even when we have laboriously tried to answer these questions, it is not always possible to arrive at the true meaning of the name. Yet inability to find a satisfactory meaning is no reason for acquiescing in a guess which we know to be wrong. A collection of early forms enables us at least to reject those popular meanings which have often been handed on through centuries of false etymology. For 'folk-etymology is always with us, and the too ingenious antiquary is no modern phenomenon'. Even Leland was not the first to corrupt place-names by elaborately imaginative explanations. Giraldus Cambrensis and Robert of Gloucester had done it before him; and, before them, Henry of Huntingdon can hardly mention a place without proceeding to explain the meaning of its name.

If we can find our earliest form in an Old English Charter, it is usually possible to get at the true meaning with certainty. But, more commonly, the first occurrence of the name is in Domesday, or in the Testa de Nevill, that is, in records compiled by foreign scribes, who wrote down, as nearly as they could, the sound of the name, as they heard it from the natives. Imperfectly acquainted with English, they rarely heard correctly, and so the forms they write usually suggest rather than express the true Old English word. Thus in Herefordshire one of the Old English Hundreds was Wimundestreu, i.e. 'Wigmund's tree'. But the Norman-French scribe in Domesday writes it, in different entries, Wimundstruil, Wimstruil, and Wim Strui. Occasionally he seems to have copied his form from an earlier document, and then he is fairly correct. But more usually it is plain that he is trying to spell phonetically names from a language he does not know and cannot pronounce, and for which his alphabet had not always proper symbols. He almost always puts ch for k; initial th is usually t, and medial th is always d. He hates all gutturals, h, ch, or gh, and often boldly changes them into st - a practice which gives us the clue to many puzzling forms. He usually writes plain s for sh, or else prefixes e. In spite of these drawbacks, Domesday remains our chief storehouse of early place-names; and, as Dr H. Bradley says, 'if we understand the principles of its orthography we can often discover with certainty what the names really were'.

In the Testa de Nevill the case is far otherwise. 'It has long', says Dr Round, 'been at once the hunting-ground and the despair of the topographer'. The spelling throughout is hopelessly wild - though, finding many such entries as Solbedune for Shobdon, one wonders how much of this is due to the transcribers and editors of the badly-edited text of 1807, which is the only one in print as yet. But many of the mistakes must be attributed to sheer stupidity or carelessness on the part of the original scribe. As an example of what he usually does with English names we may take this (Kingstone) entry 'Welketon, Cobbewell, La Mare', by which he means 'Webbeton, Caldewell, la Mare', i.e. the present Webton, Coldwell, and Meer Court. Naturally when he gets to the Welsh names in Archenfield, his mistakes are even worse. We can scarcely hazard a guess as to what places he means by Trayhac, Laund', or Attelgunt, all in Archenfield !

The English settlers in Herefordshire spoke the Western variety of the Mercian dialect, which differed in many particulars even from the Eastern Mercian, and still more from the Northumbrian and Wessex dialects (though it has some few features which seem Wessex-born, or, as Mr A. J. Ellis thinks, are due to Welsh influences). Most of what we call 'Anglo-Saxon literature' is in the Wessex dialect, which is full of diphthongs. But the Mercian dispensed with these diphthongs of which the West Saxon was so fond. He said all for eall, scep for sceap (= sheep), liht for leoht (= light), and wall for weall [even these examples are sufficient to show that modern English has developed out of the Mercian dialect, which, being intermediate between the other two, helped to interpret between North and South]. He softened g into y, saying (as we see in our Herefordshire place-names) yard for geard, and yatt for geat. He said hill when the southern dialects said hull, and the Kentish hell. And he shortened the ponderous Wessex personal names into almost their modern forms.

It is probable that soon after the victory of Chester, in or about the year 615, the earliest bands of Mercians pushed to the westward of the Severn, and settled in our county. But it was during the seventeen years of Wulfhere's vigorous reign (659-675) that the English rule was firmly established in Herefordshire, and Wulfhere's brother Merewald appointed sub-regulus of the Magesaetas [for the etymology of this word see under Maund the Alphabetical List], as the new settlers to the west of the Malvern Hills now began to be called. It seems fairly certain that before the end of the century Mercia had already reached more or less its westernmost limit, and that the great work of Offa, in the next century, was one of definition and development rather than of conquest.

But this 7th century settlement of Herefordshire was, of necessity, sporadic and incomplete. Throughout the county, and beyond it to the west, the woods were particularly dense. The ley of the Mercian settler was simply a clearing in the forest, where he built for himself and his sons a tun, that is, an isolated homestead, consisting of a rude wooden or wattled house, with, around it, a few enclosures for rearing calves, and a patch of arable land. This reclaimed patch [the word 'assart' is mentioned in Domesday under Herefordshire alone] he usually called by his own name - the tun of Bacca, of Tidbriht, or of Wilmar [Bacton, Tyberton, Wilmaston]. Or, it may be, he gave to his clearing a name descriptive of the site or its surroundings - some conspicuous tree, Ethelmund's or Berthold's [Aymestrey, Bartestree]; a clay slope [Clehonger], a blind valley [Hope (there are ten in Domesday)], or a hill showing the red Herefordshire earth [Radlow]. When he found a spot suitable for dairy farming, he called it Butterley, or (in the wild scrob to the north of our county) Cheswardine. As his settlement grew, he established, in some outlying clearing, a Barton, or enclosure for the barley crop he grew there; and he built, in another part, a dwelling for his herdsman [Hardwick], or even for a friendly Welsh tenant [Welcheston]. And everywhere he had to construct, for refuge or defence, a fortified place (burh), of which there are at least thirty in the county. North, south, east, and west of such a -bury he dwelt, with comparative security, in Norton, Sutton [there are in England, it is said, 72 Suttons, and 52 Westons], Easton, or Weston. So, in four centuries from the first invasion, we get the more or less settled Herefordshire of Domesday, which, in its main features, is the Herefordshire of to-day.

For the history of the county during these long centuries the place-names are perhaps our chiefest guides. We may take it as a general rule that the -hams and -tons are older settlements than the places ending in -wood, -ley, -field, -wardine, and the like. Dr Round has shown that in Sussex and Essex - perhaps everywhere in the south - the -hams are older than the -tuns. This rule does not seem to apply in Herefordshire, where, indeed, the -hams are rare compared with the -tons, the Mercian -halh ('riverside pasture') often taking the place of -ham.

Even before the end of the 9th century the village geography of the county had taken on more or less its present form. In the rest of England, by this time, the O.E. tun had lost its earlier meaning of 'an isolated homestead', and normally implied an ancient settlement on the lines of a village community. But in Herefordshire the word retained and still has its earlier meaning [see Town in the Alphabetical List, and -ton in the Appendix]. Yet as the tun prospered and grew, here also it developed by degrees into something like the manor of later days. As we have seen, -tun is usually compounded with a personal name. This personal element in our place-names necessarily implies some sort of seignorial right or control. The man who gives his name to the settlement is not some temporary official, like the head of the imaginary mark-moot. He is owner, rather than primus inter pares. He may be little more than a farmer in a remote clearing, with cottagers who work for him on his land; or they may be free men, rendering dues and services. In either case, the lord is there, and with him the starting-point of the coming manorial system. Even in the earliest days, the type of agrarian community is such that from it the manor of the 11th century may, without any breach of continuity, have evolved.

The complete nature of the English conquest, too, is proved by these place-names. The later conquest by the Normans, complete as it also was (and Herefordshire was the most thoroughly Normanized of all the English counties), left our place-names almost untouched. The new Norman owners settled as lords in the villages they found existing, taking over, very much as they were, them, and their inhabitants, and the names by which these inhabitants called them. Even when they pushed beyond the established boundary of English Herefordshire [This boundary in the 9th and 10th centuries, and perhaps even in the 11th, was the valley of the Dore ('Straddele'). The 'Dor' is mentioned in the Anglo-Saxon Chronicle under the year 941 as one of the boundaries of Mercia; and in the Lib. Land. as forming, with the Worm, the boundary, in 1129, between the dioceses of Hereford and Llandaff. (Ecclesiastical boundaries long mark the civil or political divisions with which they originally coincided.) So sharply drawn was the distinction that, in Domesday Bacton (in the diocese of Hereford) is assessed in hides, while, only three miles distant, Ewyas Harold (in the Welsh diocese) is reckoned in 'caracutes,' like the most recent conquests from the Welsh.], and built a castle on the fringe of Ewyas, they took over both its name and the nine Welsh 'squatters,' whom they found there. In short, the Norman rule in the county had little or no influence on the place-names, beyond some occasional alteration in their spelling and pronunciation, and the turning, here and there, of a name common to several places in the county into what Dr Bradley calls a 'double-barrelled' name [It is strange and noteworthy that the place-names introduced by the Normans are everywhere so few. 'The stolid English peasant fairly extinguished the proud Norman peer'. The two chief surviving traces of Norman influence on Herefordshire names are the habit - all but universal in the county - of prefixing the definite article to the name of a place, and the equally common use of 'Court' for 'Manor-house'. Of new names given by the Normans we have very few, and this is true also of England generally. 'On the whole', says Freeman, 'really new names were confined to really new foundations'. The two Caples, The Barr, Belmont (which may be modern), the various 'Castles', The Elms, Foy, Hay, Petty France, The Golden Valley, Rowlestone, St Devereux, and the seven Valletts, with ten other -ett endings, are all that we can pronounce to be new Norman names. Norman family names have been added to Acton, Edvin, Holme, Hope, Stoke, Mansell, and Stretton]. When, therefore, we find the village names in the conquered districts so overwhelmingly English, we are driven to the conclusion that the settlements of the invading Mercians were quite different in character from the later annexations of the Normans. These latter conquerors took over already existing settlements (we may not perhaps call them 'manors,' though virtually they were so). But the English, four centuries earlier, had formed new settlements upon lines and sites entirely independent of anything they found existing in the conquered lands, whether it was some relic of Roman Coloniae, or some native Celtic village that they came upon. There must have been Welsh trefs between Wye and Severn, as in Archenfield and Ewyas, in the 5th and 6th centuries; but no traces of them remained in the middle of the 8th century. And though Mercian new-comers settled in or near the ruins of Magna, Bravonium, and Ariconium, they settled entirely on their own lines, and gave new names of their own to their new settlements.

These Mercian settlements at first - and perhaps for a couple of centuries - were bounded by the Wye. On the left bank of the river, however, they soon pushed far to the west, Radnor, Knighton, and Norton being English soil some considerable time, it would seem, before the raising of Offa's dyke. To the south of the river, in Archenfield, the place-names, for the most part, date from the 5th and 6th centuries - the age of the Welsh saints, after whom so many of the villages in Archenfield are called. When the bargain was struck which gave to the Welsh of Archenfield the local autonomy, of which the details are given in Domesday, we have no means of knowing. But the place-names afford remarkable evidence as to the totally different nature of the settlements on the opposite sides of the river, where it bounds Archenfield. On the eastern bank are such names as Brockhampton, Fawley, Brampton Abbotts, Walford, and English Bicknor [A few Welshmen are mentioned in Domesday as holding land on the left bank of the Wye. Griffin had half-a-hide in Pyon; Saissil holds Staunton-on-Wye, the English name of which implies its previous possession by an English owner; and a Welshman is undertenant of land once held by Edward, an Englishman. The fact, too, that Brismer holds Brismerfrum, to which he had given his own name, suggests a quite recent English settlement of the estate.]; on the western bank, Kilforge, Treyseck, Llanfrother, Craddock, Daffaluke, Ganarew, and similar Celtic forms. The only unmistakable Welsh name on the left bank below Hereford is Penalt in King's Caple.

Before the time of the Survey, but how long before we cannot say (though it must have been a fairly considerable time, since the English names had become established), English settlers had crossed the Wye, and established themselves among the Welsh. For we find a few definitely English place-names in the very heart of Archenfield in 1086. Wilton, Wormelow, Goodrich, Westwood, Ash, and Baysham, are clearly English settlements [about 160 years later Testa de Nevill speaks of the 'French and Welsh' of Archenfield]. And in the 12th century two Welsh names in Archenfield became curiously 'Englished', as Ballingham and Moraston. In the 13th century we have an apparently English settlement at Humfreyston; but this name, later on, was translated into Welsh and became the present Trebumfrey. So, too, the English Baysham had reverted, by Leland's day, to its earlier Welsh name of Sellack. As a whole, Archenfield remained Welsh for centuries after Domesday; and in its place-names is largely Welsh still [It is roughly true that somewhat more than half the field- and farm-names in Archenfield are Welsh; and in some districts even more. In Llangarren quite two-thirds are Welsh].

To the west of Archenfield, however, and actually between it and Wales, the district of Straddel became almost entirely English. It was the natural gateway for Welsh raiders into England, and, perhaps a century or more before the Survey, was effectively occupied by the English. From end to end it is full of English names in 1086. On its upper boundary was Cusop, and thence, down to Bacton, was a chain of English settlements. First the stronghold of Clifford (which under Earl William Fitz-Osbern was not merely a castle with a garrison, but a chartered borough, with special privileges granted to the burgesses). Then More (The Moor), Harewde, Becce (Bache), Midewde (Middlewood), Torchestone (Dorstone), Elnodestune, Edwardstone, Poscetenetune (Poston), Manetune (Monnington-in-Straddel), Brocheurdie, Beltrou, Wilvetone, Wilmestun (Wilmaston), Almundestune, Alcamestune. Nothing shows so clearly the insecure condition of this frontier district as a comparison of these names with the later place-names in the valley. Of the fifteen settlements mentioned in Domesday more than half have completely vanished, and their position cannot even be guessed at. The very name of the valley (Straddele) has all but disappeared. We have still in the valley a certain number of English names, Peterchurch, Vowchurch, and the like. But these only appeared in the 12th or early 13th century; and by far the greater number of place-names in the valley are now Welsh, dating from the 16th and 17th centuries, or even, in some cases, later. [This is well brought out in a comparison of the farm- and field-names in Dorstone, of which we happen to have a good number in an undated document of which Bishop Peter de Aquablanca (1240 - 1269) is a witness. We have, then, in Dorstone, about the middle of the 13th century, such names as these:- Selehulle, Huntehulle, Stathorne Walle, Benfelde, Huntewalle, Gulegrove, Sponne, Dudintone, Humemedue, Redewaldebrok, Stevenehus, Weleye, Timbesbroc, Marleput, Lordesleye. (One Welsh name only is included - Sakenant.) In the Dorstone of to-day most of the names are Welsh:- Cwm mills, Llanavon, Brynspard, Mynydd brith, Pwll Cam, Bedw, Llanach, Peny-y-lan, and the like.]

This action and reaction of Welsh and English influences, extending over centuries of border history, produced similar results in other districts also. We have seen that English place-names had reached as far west as Radnor even in Offa's day: Kington, Norton, Knighton, Cascob, and other like names giving evidence of vigorous English life far beyond the Dyke. But all these gains had been lost long years before the date of the Survey. Successive entries 'Wasta fuit [i.e. T.R.E.] et est in Marcha de Walis' show the devastating results of Welsh raids on these manors, which once had paid geld but now are waste and pay nothing. [An entry in the Domesday of Shropshire says that in the time of King Ethelred, i.e. about the year 1000, the three royal manors of Whittington, Maesbury, and Chirbury together rendered half a night's feorm, or about 50. At the death of the Confessor they were waste and yielded nothing.] The English manorial names, however, are entered in Domesday, and still survive; but to-day they are surrounded by a host of Welsh farm- and hamlet-names. In Kington itself there appear to be no Welsh names, but in the neighbouring Huntington quite half are Welsh, and in Brilley considerably more than half. Some at least of these Welsh names were introduced as late as the 16th century, when the break-up of the Marcher lordships, and the comparative tranquillity of the border under 'Bishop Rowland's justice', was followed by a very considerable immigration of Welshmen into Herefordshire.

Much the same thing is to be seen in the Monnow valley. The English name Crasswall, of 13th century origin, survives, but more than half the farm- and field-names in the parish are Welsh. The 13th century Michaelchurch becomes throughout the 16th and 17th centuries Llanfihangel-eskley, being re-translated into English only by 18th century antiquaries. Lower down the valley English influences were felt earlier; for Silas Taylor, in the middle of the 17th century, says that the Welsh Clodock 'hath lately taken the name of Longtown'. Yet some English names would seem, even after this time, to have been crowded out by Welsh; for 'Foscombe' and 'Burybourne', which existed in Clodock in 1540, are not there now; and to-day considerably more than half the farm-names in Longtown are Welsh. A few miles lower down the river, it was only at the end of the 17th century that the Welsh 'Pontrilas' completely supplanted the English 'Heliston'. [A good instance of this gradual re-conquest of territory by the Welsh language is to be found in Flintshire, where such typically English Domesday names as Preston, Westbury, Merton, and Bishopstree, have become the modern Prestatyn, Gwespyr, Mertyn, and Bistre. Yet not far away a group of English names was able to hold its own. For part of the lordship of Bromfield and Yale lay within the borders of the Cheshire of 1086, and English place-names were given to its townships and hamlets. Soon after the date of the Survey all this area became Welsh (the 'commote' of Merford) and remained so for centuries. Yet the English names in -ham, -tun, -ley, and the like remained almost unaltered, and still survive.]

Of the two classes into which most place-names fall, viz. those which are simply descriptive and those compounded with a personal name denoting the original owner, the English settler, as we have seen, usually preferred the latter class. The Celt, on the other hand, rarely commemorates himself - which, it may be, is an indication that there are no germs of the manorial system, no implication of seignorial right or control in his tref or commote. But the descriptive names he gives to his settlement show strangely little of Celtic originality or poetic feeling [only the Norman seems to have been impressed by the sheer beauty of a site; for he only has such place-names as Belmont, Belgrave, Belvoir, Beauchamp, Beaumont, Bewdley (i.e. Beaulieu)]. And when he does, at times, attach a personal name to a village, it is not that of the owner, but of the saint whose residence hallowed the place. Since the Celtic saints were legion, it was easy to find one in most localities. Before the Normanizing of the border there were many more place-names involving the personal names of local saints than those we have now. For the French bishops and priests, who came in the wake of the Norman lord, insisted that churches bearing the names of these Welsh saints, unknown to the church at large, should be re-dedicated. This was passionately resented by the patriotic sentiment of the Welsh, and, in their turn, they refused to accept the saints of the Roman calendar. A compromise was often made by re-dedicating the church (which usually involved re-naming the village) to the Virgin or to St Michael and All Angels. Hence we have, on the Herefordshire and Monmouthshire border, about thirty Llanfihangels and Michaelchurches, and a dozen or more Llanfairs. [Altogether in Wales and the border counties there are at least 150 Llanfairs - a proof, not of the 'Mariolatry' of the Welsh, but of their ecclesiastical patriotism.]

It only remains briefly to mention some of the changes in our place-names which have come about in the course of centuries. We cannot explain why one name-form persists rather than another, nor why one survives, perhaps, for centuries, and is then lost. But we do know that place-names, in earlier days, were in a somewhat unstable condition; and that, even in the last hundred years they had not been unchangeably fixed. We have see how 'Heliston' yielded to 'Pontrilas' after nearly two centuries of struggle, how Clodock, about 1670, 'hath lately taken the name of Longtown,' and how such names as Michaelchurch and Humfreyston have appeared alternately in Welsh and English dress. In the Alphabetical List has been included a considerable number of names which have now entirely disappeared [There are other names, of manors or chapelries, which were once important, but have dwindled to almost insignificant holdings. Webton, e.g. (in Madley) and Llanwonog (in Clodock), now only farms, were once chapelries with a priest of their own. A similar chapel at Urishay, used for centuries as a barn, has recently been restored to the service of the Church.], since these vanished names are, for historical purposes, as valuable as those which have lived on into our day. But often, when the name persists, other influences which make for change in its form are at work, above and beyond the normal development of the Old English language according to philological laws. The chiefest of these external influences is folk-etymology. If a place-name has no meaning that is apparent on the surface, an ingenious amateur antiquary will always set about to find one. The process is continually going on before our eyes, in spite of the steadying influence of the printed page. In Ewyas Harold the 'Martin Well' of the 1831 Ordnance Map is full 'St Martin's Well' before the mid-19th century, and is so still (being recently credited also with quasimiraculous healing power). A spring in Peterchurch was already named, after the church and village, 'St Peter's Well', and fitted with its legend, in the 18th century. The Dr Harford, who bought some land in Tupsley in 1684, would scarcely recognize himself in 'Hafod Road.' Britton's naively modest suggestion that 'Moneyfarthing Hill' was so called 'probably from coins found there' is scoffed at by the same well-informed persons who gravely repeat the 18th century legend of Turn-a-stone, or connect Dorstone with a Teutonic god. These well-meaning theorists have been at work upon our place-names for a thousand years; and that is why Dr Round is so fully justified in disputing the claim of the philologist to explain place-names solely by phonological laws. Mere carelessness, again, is often an important agent in changing the form of a name. Longworth (in Lugwardine) was Longford from the 12th to the 18th century, and Strongwood, which started life in Domesday as a -wardine name, became Strongford, Strongworth, and Strongward, before it grew into a -wood. The 'Lingham' of Domesday soon became, and remained for centuries, Lingeyne, and now is the meaningless Lingen. The endings -den, -don, and -ton are often confused [-ton is further confused with -stone, of which latter ending we have only two genuine instances in Herefordshire - Aylstone Hill and Langstone. (In 'Tedstone' the -stone is actually a corruption of -thorn.)], as we see in Grendon and Bicton (both originally -dene). The forms -ig and -ea were inextricably mixed up even by the 'Anglo-Saxons' themselves. Some of our -leys were originally -lows; and Colwall, Crasswall, and Eccleswall were all at first named from springs of water.

How such changes came about will easily be understood by a cursory examination of any document written before the end of the 18th century. An instance - not quite typical, but neither is it very extreme - is to be found in a Goodrich Terrier of 1722, where the same field is called in the same document, 'Hollen Duff Close', 'Holland Tuff Field', and 'Holland Dover Field'. In another document of the same date, dealing with lands at Credenhill, the same place is called on the same page 'Sheep's Court', and 'Sheep's Coat'; and another place is referred to as 'Brincourt' and 'Bringate'.

Return to top of page

URL of this page: http://www.melocki.org.uk/places/Introduction.html

Copyright notice:
All pages at http://www.melocki.org.uk
are Copyright Mel Lockie 2011.
All rights reserved.
For a detailed copyright policy see: Conditions of Use.