Superficially the reconstruction of the eleventh-century landscape is the easiest task for the historian of the period. Domesday Book appears to furnish a mass of detailed evidence. The apparent record of settlement is unparalleled in most areas before the Hundred Rolls of the late thirteenth century, and the general survey of estate structure and stocking is unique in the Middle Ages. Comparable descriptions of the minutiae of resources and management are confined to individual estates and manors. However, the unique nature of the record is at the very root of the problem of interpretation. Unlike, for example, the extents of inquisitions post mortem, the context of the Domesday data is rarely understood. Nevertheless, by the judicious use of other types of evidence, critical criteria can be developed. We have already seen that the record of churches is very deceptive. Archaeological and structural evidence demonstrates the existence of many foundations in the mid eleventh century which do not appear in the text. This is not surprising if Domesday Book is primarily a record of demesnal estates and the dues which they enjoyed. We can hardly expect it to provide us with a complete list of pre-Conquest churches. But such a realisation opens up new possibilities of interpretation. The record of a church, or its omission, enables us to evaluate something of its status in the eleventh century which would otherwise be impossible from the available sources. By no means a complete or objective survey,  Domesday Book nevertheless tells us much about what it does record.[1] The significance of identifying place-names is a problem of the same kind, although considerably more complex to interpret. It has long been recognised that many settlements which independent evidence shows to have been in existence in the eleventh century, do not appear by name.[2] Straightforward oversight on the part of the Domesday commissioners is sometimes responsible, but the phenomenon is so common that this is clearly not always a sufficient explanation. Both the procedures of the inquiry and the sources employed introduce distortions into the survey in this context as in many others. The identifying names of the Domesday Book must be examined in these terms before we can fully understand what evidence they provide for the elucidation of settlement structures.

            Three hundred and thirteen place-names are recorded in the Notting-hamshire folios. Occasionally it is immediately possible to discern some characteristics of settlements to which they refer, for adjacent vills with the same name are sometimes given differentiating epithets. Morton, for example, is distinguished from 'the other Morton' and North Morton, Ordsall from South Ordsall, North from South Muskham.[3] Some settlements were apparently divided in 1086 although not subsequently: two Chilwells and two Thistletons are represented in the later records as single settlements,[4] but nevertheless it seems likely that the usage implies two separate nuclei, if not necessarily two nucleated villages. Most divisions of this kind, however, are probably not made explicit in the text. There were, for example, four estates called Leverton at the time of Domesday held by the archbishop of York, Roger de Bully, the king, and Count Alan.[5] By 1316, North and South Leverton constituted separate vills, the one held  by  Master  Lewis  de  Beaumont and Adam de Everingham, and the other by the king and Thomas Latimer.[6] Since the four fees apparently descend from the four Domesday estates, the very same division between North and South Leverton may have been in existence in 1086.[7]

            The failure to distinguish different settlements is not confined to those with the same name. Some entries identified by one name include settlements subsequently known by a different name. According to Domes-day Book, William Peverel held a manor in Toton with half a church.[8] In the thirteenth century the vill was situated in the parish of Attenborough, half of the advowson of the church of which belonged to the Peverel fee. The other half was appurtenant to a holding which descended from an estate with half a church, which was identified as Chilwell and East Chilwell in Domesday Book.[9] It seems likely that Attenborough was in existence in 1086, but its description is subsumed in the entries for Toton and the Chilwells. In both cases the identify-ing place-names are evidently those of estates. The Southwell entry, of course, affords the most obvious example of the use of an estate name in the Nottinghamshire text. The one name describes the estate centre and twelve unnamed berewicks which belonged to it.[10] The dependent settlements can be identified from independent evidence as Normanton, Upton, Fiskerton, Farnsfield, Gibsmere, Bleasby, Goverton, Halloughton, Hallam, Kirklington, and Morton.[11] But many entries may be of this type, although the fact is but rarely explicit. Such names were probably taken from seigneurial sources and, where they have been extensively used in, for example, breves which take their form the tenant-in-chief's return, the identifying names probably indicate estate and, by implication, settlement nuclei, with some degree of precision. As estate names, however, they may encompass several settlements other than the ones named.

            Not all Domesday place-names, however, are of this type. The names of local government units were extensively employed in the North. Thus, in Lincolnshire, the hundred name is frequently used. An estate in Long Sutton, for example, is identified as Tydd because in was situated in the hundred of that name.[12] It has been suggested that all Domesday names in the West Riding of Yorkshire were those of townships.[13] Such names were presumably taken from an official source, almost certainly a geld list, and imply that relatively little use was made of seigneurial returns in those breves in which they occur. Clearly, as names of local government units, they do not necessarily identify either estate or settlement nuclei. In Nottinghamshire, it has not proved possible to determine the nature of the unit of local government at its lowest level, the vill. However, the same type of process may have been at work. The official part of the survey was derived from a geld list and its nomenclature may have been adopted in some of the breves.

            It is not always possible, then, to determine whether a place-name is that of an estate or of a local government unit. However, it is clear that, at best, it will only identify a nucleus of some kind. In fact, Domesday Book affords very few clues to the actual form of  settlements. In the past, it has been assumed that mast Anglo-Saxon settlements were nucleated, that is, of the classic Midlands pattern of peasant tofts clustered around a church and manor house within an open field system. Modern research has cast much doubt on this simple picture. In Devon, for example, there is a landscape of scattered farms set within their own fields which is evidently an ancient pattern of settlement. Domesday Book, however, gives the superficial impression that nucleation was the norm.[14] Disposed settlement patterns may, it seems, be widespread.[15] It would be misleading, though, to see it as the basic form of Anglo-Saxon settlement for it is likely that there are both centrifugal and centripetal forces in the dynamics of settlements in most periods. The nature of society probably establishes the dominance of one type over another at any one time. Low population, for example, and relatively weak bonds of lordship, may be reflected in a more dispersed pattern. By way of contrast, high population and advanced manorialisation may imply some greater degree of nucleation.[16] The two alternatives, however, are not mutually exclusive. Eadwy's 956 grant of Southwell to York attests to some degree of dispersion of settlement for 'the cottages' are recorded as one of the boundary marks of the estate. But presumably there was a sizeable royal hall, a natural focus for nucleation, within Southwell itself.[17] To a greater or lesser extent, this may have  been true for many settlements in 1086. There is little evidence in Nottinghamshire, however, to elucidate the relative importance of either form. In some circuit 6 entries, anomalous descriptions of sokemen and holdings suggest dispersion of settlement. In Lincolnshire, for example, sokemen are occasionally said to hold tofts rather than the usual carucates and bovates.[18] But no such suggestive descriptions appear in the Nottinghamshire text. Domesday Book in itself, then, can tell us little about the forms of settlement.

            Nevertheless, the survey can be used in conjunction with later evidence to identify settlement and estate nuclei. The sources of the twelfth century and later often reflect settlement structure with some degree of accuracy. When it is possible to reconstruct the descent of estates from 1086 into the thirteenth century, such evidence can often fill out the terse formulations of Domesday Book. Thus, as we have already seen, the later history of Leverton tends to suggest that two of the Domesday fees were situated in North Leverton and two in South Leverton.[19] The sites of churches are often useful in this connection. There was frequently a close relationship between church and manor in the eleventh century, and the site of the one in the later Middle Ages can often indicate the hall of the other.[20] But this method is retrogressive and must therefore be used with considerable caution. We cannot automatically assume a continuity of site between the eleventh and thirteenth centuries, and churches, although private in Domesday, may have entirely different origins which are reflected in their  location.[21] At best, the method can be predictive, and corroborative evidence, such as archaeological and topographical analysis, must be employed to verify its conclusions.[22]

            Many of the problems of settlement form would be easier to resolve if we had a greater understanding of estate structure. But, here again, there are formidable problems of interpretation. The matter devolves upon the relationship between Domesday Book manors, tenurial units, and economic entities. It is clear from early-twelfth century surveys of the estates of Peterborough and Burton that the three types of organisation often coincided,[23] but it is equally clear that this was not universal. As we have seen,[24] the Domesday manor embraces many different types of estate, but in its essentials, it was not an economic unit. Although not necessarily endowed with sake and soke, it was primarily concerned with delegated tribute. Its identifying feature was thus the point at which such dues were rendered. It was the hall which was its essential physical manifestation. Thus, in Eaton there were 10 thanes, each with his hall, and there were therefore 10 separate manors, while in Epperston and Woodborough, Wulfric and Alsi held 3 carucates and 4 bovates, but there was only one hall  and  consequently  a  single manor.[25] The Domesday manor, then, was first and foremost a legal concept and as such was subject to changes of form which probably amounted to little more than a redistribution of dues. Hence, in the Nottinghamshire text we read that so many carucates were held 'pro manerio', 'for' or 'as a manor',[26] and in Lincolnshire there is evidence that parcels of sokeland were converted into manors between 1066 and 1086. Thus, Scottlethorpe had been held in soke, but it was deraigned as a chief manor in 1086.[27] Indeed, there was probably more change in the structure of manors than is always apparent. The boundaries of the estates in Sutton in Nottinghamshire and Barton and Barrow in Lincolnshire both underwent considerable changes between the later tenth century and Domesday Book. The one lost land, while the other apparently gained several parcels of sokeland.[28] This fluidity can hardly imply a generally well defined internal identity. No doubt the lord's demesne functioned as an economic unit over a long period of time, although the existence of a portable hall in the bishop of Durham's manor of West Aukland in County Durham in the late twelfth century suggests that a permanent establishment was not indispensable.[29] But in so far as soke was concerned with tribute, it follows that the manor of the text is not necessarily an economic unit as such.

            Nor was it always the basic tenurial nexus. The record of a  tenant or a separate value in some sokeland entries often suggests the existence of estates within the manor. In Gonalston, for example, Ernwin the priest and 4 sokemen held 5 bovates of sokeland belonging to the king's manor of Arnold.[30] This type of information is not common in Nottinghamshire. But the commissioners were probably not anxious to record it for many such sub-tenancies in Lincolnshire only come to light in the Clamores as opposed to the body of the text.[31] Other types of evidence, however, suggest that it may have been of widespread occurrence, and point to some fluidity of management within the structure of the manor. The tenurial context and status of forinsec sokeland in relation to the soke centre have already been discussed.[32] The detachment of parcels of sokeland from the parent manor in the same breve may attest to some similar degree of separate management. Superficially, such entries appear to have been inadvertently omitted in the account of the parent manor, and were enrolled in their appropriate contexts on the basis of the form of a geographically arranged source. This may in many cases be an adequate explanation of the phenomenon, but  this type of displacement may also reflect tenurial and/or economic arrangements within the manor in the actual exploitation of the land. For example, Roger de Bully held a manor in West Markham, and the account of the estate is followed by that of two parcels of land in the same place which were soke of the manors of Tuxford, Grove, Eaton, and Drayton.[33] In two cases the soke centres were geographically remote, and it seems more likely that the land in question was actually exploited by the manor in West Markham. The Domesday manor, then, is not necessarily a basic tenurial unit. As a soke nexus, however remote, it may encompass several estates which were managed by someone other than the lord of the soke centre. To what extent these estates represent economic units in the exploitation of land is unclear. Many may have functioned in the same way as the manors of the abbeys of Burton and Peterborough. But evidence is essentially wanting, and at present little is understood about their relationship with settlements and field systems and the communities of which they may have formed a part. There are grounds, however, for believing that lordship was beginning to express itself in terms of economic manorialisation in the classic sense. The record of demesne teams in most manorial entries implies the presence of demesne which was cultivated by the villeins. It is unlikely that the latter were personally unfree in the eleventh century, but they were evidently closely associated with the estate.[34] Almost all manorial entries record their presence, while they are rare on sokeland, and, unlike the sokemen, their liability to the geld is never recorded. It was almost certainly discharged within the hundred by the tenant of the manor. Ecclesiastical surveys of estates in Nottinghamshire, Derbyshire, and Lincolnshire indicate that many owed onerous day works.[35] In a very real sense they belonged to the manor. By way of contrast, the services of the sokemen, both intra- and extra-manorial, were comparatively light, and it is probable that manorialisation of their land was not advanced. There are, thus, only nine entries in the Nottinghamshire Domesday in which demesne teams are recorded.[36] However, this may  not be significant in the light of the comparatively low number of sokeland entries. Seigneurial encroachment onto intra-manorial soke is effectively hidden by the form of the text. But if there was still personal freedom within the estate, it would appear that many commodities formerly enjoyed by the community were becoming more closely identified with the lord's hall. The record of manorial appurtenances such as woodland, meadow, pasture etc. is unlikely to be a reliable guide to the extent of economic resources. Rather it reflects those from which the lord derived a direct or indirect income. Thus, in Lincolnshire, there is only a handful of references to fen and marshland, despite the fact that it was a valuable resource.[37] It would seem that it was generally intercommoned by groups of communities, and only in exceptional circumstances had it been appropriated by individuals.[38] It is, then, but rarely noted in Domesday Book. Much woodland is recorded in the Nottinghamshire text attached to almost every manor north of the Trent. Already it would seem that each holding had its own share, and in some way it was attached to the estate. Whether, as in Lincolnshire, there had ever been any communal interest in the waste is not clear, but by 1086 it seems that much belonged in some way to the lord's demesne.[39]

            As a class, the large soke is not different from the normal manor in kind, but extent. It is characterised by a large number of parcels of sokeland, often at some distance from the caput, which owed service to a central hall. On this account, the type can hardly have  operated  as  aconventional manor. In fact, as we have seen,[40] it is an estate in which the essential unity is a tributary nexus. The form is commonly found throughout England in various guises such as soke, lathe, shire, or multiple estate, and there is no reason to see it as a particularly Danish institution.[41] By the time of Domesday Book, many were clearly very fragmented, but their break-up had given rise to many of the smaller manors in the county. The area of the former estate can, however, sometimes be identified from the pattern of interlocking appurtenances in the surrounding estates. Nevertheless, considerable care is needed in interpretation. First, the institution was still vital in the eleventh century. It cannot be assumed, then, that all or indeed any characteristics of a particular example are necessarily ancient. In Lincolnshire, for example, part of the soke of Greetham in 1086 was situated on land which had only recently been reclaimed from the sea.[42] Adjustments in the distribution of sokeland were probably also more common than is always immediately apparent.[43] Once estates were held by book, they could be divided or amalgamated at will to suit the particular requirements of the lord. Any number of imponderables may lie behind the form of any particular Domesday estate. Second, royal sokes were sometimes administered in groups. In Derbyshire, Darley, Matlock (Bridge), Wirksworth, Ashbourne, and Parwich were farmed by a single reeve, as were Bakewell, Ashford and Hope, for only one value is given for each group.[44] Such arrangements may not always be so apparent  and may therefore conceal the structure of individual estates. Royal manors are sometimes attached to some other type of organisation. Part of the soke of Grantham, for example, was in some way appended to the wapentake of Aswardhurn.[45] Just quite what this means is uncertain, but in Huntingdonshire some parcels of land were in the soke of the hundred of Leightonstone, although they were administered from the royal manor of Alconbury.[46] As such, it appears that they did not belong to the manor, but had become associated with the royal estate through forfeiture, commendation, or whatever. The soke of any particular manor may, then, include lands of varying status which were appended to it for administrative convenience.

            This is not the place to review the Nottinghamshire evidence in any detail. In the light of the importance of the subject, however, a number of comments can be made on the accounts of the large royal estates in the north of the county. The Domesday manor of Mansfield encompassed at least three elements which were almost certainly independent in origin. First, there was the manor of Grimston with soke in Grimston - probably in fact located in Ompton - Kirton, Willoughby and Walesby, Besthorpe, and Carlton, and possibly Farnsfield.[47] The account, however, is duplicated in the entries relating to the manor of Mansfield where Grimston is called a berewick and its land soke of the same estate. But the connection was probably only a temporary expedient. Grimston forms a geographically discrete estate which interlocks with the appurtenances of the manor of Laxton (figure 13) - the two estates  had  clearly  constituted one organisation at an earlier period - and was peripheral to Mansfield and subsequently farmed as a separate estate.[48] Second, there was Mansfield itself with soke in some fifteen settlements.[49] The whole formed a fairly tight unit in the west of the wapentake of Bassetlaw. Its appurtenances may interlock with many of the manors of Roger de Bully in the same wapentake. Third, there was what was known in the twelfth and thirteenth century as the soke of Oswaldbeck.[50] It had a nucleus in the parish of South Wheatley, and was also separately farmed.[51] None of its elements appears in the summary of the land of Mansfield, and, with one exception, each entry is given a separate value.[52] According to the account in the Hundred Rolls, the whole of the wapentake of Oswaldbeck had originally constituted one estate,[53] and, indeed, the soke interlocks with the manor of Laneham and the estates of Roger de Bully (figure 10). It is likely, then, that there were originally three separate manors, but, like the king's estates in Derbyshire, they were probably administered together under the manor of Mansfield. Thus, the soke recorded in Domesday Book is composite and therefore affords no evidence that the manor had formerly encompassed the whole of the north of the shire.[54] The other royal estates north of the Trent are less problematic. Bothamsall and Dunham are apparently independent of Mansfield and its Domesday satellites,  but  nevertheless interlock with surrounding manors.[55] Finally, Arnold, the smallest of the royal manors, may have been a later formation in the form in which it appears in 1086.[56] The description of all its appurtenances are later additions to the text, and much of the land may have been appended to the manor fairly recently, for the honey render of the estate had risen between 1066 and 1086. This is unlikely to indicate an increase in the value, but may suggest that land has been added to it, and its traditional render had been transferred to the king's hall in Arnold.

[1] D. R. Roffe, 'Domesday Book and the Local Historian', The Nottinghamshire Historian 37, (1986), 3-5.

[2] See, for example, P. H. Sawyer, 'Introduction: Early Medieval Eng-lish Settlement', Medieval Settlement, ed. P. H. Sawyer, London 1976, 2.

[3] Notts. DB, 1,11; 13. 9,34. 30,42. 1,5; 12. 9,19; 23. 30,56. 5,2; 5. 8,2. 12,11-13. 30,7; 46.

[4] Notts. DB, 10,26. 13,4-5. 30,52. Rutland DB, ELc, 7-8.

[5] Notts. DB, 1,32. 2,10. 5,4. 9,130.

[6] FA iv, 106.

[7] CI i, 225; ii, 231; xvi, 355; FA iv, 151.

[8] Notts. DB, 10,25.

[9] Notts. DB, 13,4; 5. R. Thoroton, The Antiquities of Nottinghamshire ii, reprint Wakefield 1972, 178.

[10] Notts. DB, 5,1.

[11] VCH Notts. i, 218.

[12] Lincs. DB, 14/97; 57/51; 73/6; D. R. Roffe, 'The Lincolnshire Hundred', Landscape History 3, (1981), 31.

[13] D. Michelmore, M. L. Faull, S. Moorhouse, West Yorkshire: an Archaeological Survey to AD 1500, Wakefield 1981, 232.

[14] W. G. Hoskins, Fieldwork in Local History London 1967, 18, 40-3.

[15] T. Rowley, Villages in the English Landscape, London 1978, 91-103.

[16] Rowley, Villages, 91.

[17] S659. The charter is the earliest firmly dated reference to Southwell. However, the eleventh-century document known as 'The Resting Places of the Saints' records that St Eadburh, probably the abbess of Repton in c.700 (D. H. Farmer, The Oxford Dictionary of Saints, Oxford 1978, 118), was buried at 'Southwell-on-Trent' (D. W. Rollason, 'Lists of Saints' Resting Places in Anglo-Saxon England', Anglo-Saxon England 7, (1978), 89). Dr Rollason has suggested that the information for the places identified by a topographical feature was derived from a source drawn up in the late ninth century (62-3). It is  possible,  then,  that there was a foundation at Southwell, and no doubt a considerable settlement nucleus, before the grant of the estate to York.

[18] See, for example, Lincs. DB, 34/6.

[19] See above.

[20] F. Barlow,  The English Church, 1000-1066, 2nd ed. London 1979, 184.

[21] The church of Barton-on-Humber was a private institution in 1086, but had its origins in an early monastic foundation (Lincs. DB, 24/13; D. R. Roffe, 'Pre-Conquest Estates and Parish Boundaries: a Discussion with Examples from Lincolnshire', Studies in Late Saxon Settlement, ed. M. L. Faull, Oxford 1984, 120-2; P. Everson, 'The Pre-Conquest Estate of Ęt Bearuwe in Lindsey', ibid., 123-7).

[22] See, for example, Sleaford, ed. C. M. Mahany, D. R. Roffe, Stamford 1979. An analysis of estate nuclei and churches suggested that the twelfth-century 'new town' of Sleaford was a major pre-Conquest estate centre, and subsequent excavation in the vicinity of the church revealed Anglo-Saxon structures.

[23] Chronicon Petroburgense, ed. T. Stapleton, London 1849, 157-83; C. G. O. Bridgeman, 'The Burton Abbey Twelfth-Century Surveys', Collections for a History of Staffordshire, William Salt Archaeological Society 1916, 212-47.

[24] See chapter 5.

[25] Notts. DB, 9,20. 14,5.

[26] Notts. DB, 5,6. 10,64-6. 12,12. 16,12. 20,8. 30,55. The formula is common in the Yorkshire folios where it is the normal device for indicating the status of a manor before the scribe resorted to the use of marginal M's. In the rest of the circuit, however, it appears to have a specific purpose in indicating a vague legal relationship.

[27] Lincs. DB, 57/14. See also Lincs. DB, 14/78; TMS, 52.

[28] G. T. Davies, 'The Anglo-Saxon Boundaries of Sutton and Scrooby, Nottinghamshire', TTS 87, (1983), 13-22; Roffe, 'Estates and Parish Boundaries', 120; Everson, 'Ęt Bearuwe', 123-7.

[29] Boldon Book, ed. D. Austin, Chichester 1982, 37.

[30] Notts. DB, 30,49.

[31] See, for example, Lincs. DB, 13/7; 69/15; 1/28; 73/6.

[32] See chapter 5.

[33] Notts. DB, 9,28-30.

[34] Lincs. DB, xxvii-viii.

[35] Chronicon, 157-83; 'Burton Surveys', 212-47.

[36] Notts. DB, 5,18. 10, 9; 10; 46. 1,15-16. 13,4. 17,11; 13.

[37] Lincs. DB, 1/3, 4; 7/43, 50; 11/3, 4; 12/81; 24/13, 14, 54, 57-8; 26/52; 63/27; 67/4.

[38] H. E. Hallam, Settlement and Society: a Study in the Early Agrarian History of South Lincolnshire, Cambridge 1965, 162-6.

[39] Roffe, 'Domesday Book and the Local Historian', 3-5.

[40] See chapter 5.

[41] J. E. A. Jolliffe, 'A Survey of Fiscal Tenements', EcHR 6, (1935-5), 157-71; W. E. Kapelle, The Norman Conquest of the North, London 1979, 50-87.

[42] A. E. B. Owen, 'Halfdic: a Lindsey Name', Journal of the English Place-Name Society, (1972), 45-56.

[43] Roffe, 'Estates and Parish Boundaries', 120.

[44] Derbys. DB, 1,11-15; 27-9.

[45] Lincs. DB, 1/15.

[46] Hunts. DB, 1,9.  19,15-22.  D25.  See  also  soke  of  Normancross Hundred.

[47] Notts. DB, 1,17-22. Soke of Grimston 'in the same place' as the manorial caput, appears to be duplicated in a further entry where it is identified as Ompton (Notts. DB, 1,18; 24).

[48] Notts. DB, 1,24; 27. 12,1-10. D. Crook, 'The Community of Mansfield from Domesday Book to the Reign of Edward III', TTS 88, (1984), 14-16.

[49] Notts. DB, 1,23-30.

[50] Notts. DB, 1,31-44.

[51] PNN, 43; Crook, 'Mansfield', 14-16.

[52] Notts. DB, 1,24.

[53] RH ii, 25, 300-1.

[54] M. W. Bishop, 'Multiple Estates in Late Anglo-Saxon Nottingham-shire', TTS 85, (1981), 37-47.

[55] Notts. DB, 1,1-16.

[56] Notts. DB, 1,45-50.