2. THE TEXT

The text of the Nottinghamshire Domesday occupies some fourteen folios of volume 1, the Exchequer Domesday.[1] A further two half folios are devoted to the Roteland wapentakes of Alstoe and Martinsley, which were administratively part of the county in 1086, although later joined to the Northamptonshire hundred of Witchley to form the county of Rutland.[2] The account is divided into four unequal sections. The first deals with the boroughs of Nottingham and Derby.[3] In 1086, and probably from a much earlier period, the two counties were closely associated, sharing a sheriff and a common administration.[4] The form of the description of the boroughs, and the relationship between them, is discussed in detail in chapters 7 and 8. The second section can be broadly characterised as an incomplete shire customal to which is appended what appears to be a list of holders of liberties.[5] Its function, however, may be either directly, or incidentally, to indicate the predecessors of the Norman holders of land.[6] Third, there is a list of tenants-in-chief who held land in the county in 1086, which is by way of an index to the fourth section, the breves, that is, chapters, which form the bulk of the text.[7] It is with this latter section, the description of the estates of Nottinghamshire, that much of this study is concerned.

There are thirty breves in all. The first, as is customary in all county Domesdays, concerns the land of the king. Somewhat anomalously, however, it is followed by an account of the estates of those tenant-in-chief of comital rank - Count Alan of Brittany, Earl Hugh of Chester and the Count of Mortain. It is more normal for the clerics to follow the king, in strict order of precedence. In the Yorkshire account, for example, the three earls occupy the fifth, sixth, and seventh positions, with Earl Hugh at their head, behind the archbishop and canons of York, the bishop of Durham and his men, and the abbot of (St. Mary's) York.[8] In Nottinghamshire, however, the archbishop, the bishops of Lincoln and Bayeux, and the abbot of Peterborough are relegated to a lower position. The tenants-in-chief of the shire follow. Again, they appear in some sort of order of precedence. The great barons of the realm usually occupy a high position in all the counties in which they held land. But there are local variations which reflect the regional importance of individual lords. In Nottinghamshire, for example, the land of the clerics is immediately followed by that of Roger de Bully, the greatest landholder in the county. In Yorkshire, however, he follows Berengar de Tosny who is twenty-first in Nottinghamshire.[9] The final breve is an account of the land of the king's thanes, those Anglo-Scandinavians of modest rank who retained title to, and tenure of, their lands after the Conquest.[10]

Each breve, with the exception of headings and some place-names, is written in Carolingian miniscules.[11] The pages are divided, by prickings and rulings, into two columns each of 44 lines. Generally speaking, an attempt was made to write each section at one sitting. Breve no. 1, the land of the king, may, however, be an exception. In common with the description of terra regis in other county Domesdays, it shows signs of ad hoc and hasty compilation. Thus, entry no. 1,24, the concatenated account of some seventeen parcels of land, and the following entries which describe the same land in greater detail, reveal a difference in hand from the preceding entry concerning the manorial caput in Mansfield, which may suggest that they were a postscriptal addition to a space which had been intentionally left for the purpose. Such peculiarities, accompanied by anomalous entries and non-standard information emanating from an early stage in the enquiry or a different source,[12] suggest that the return for the king's land was not always uniform with the other breves. Indeed, it is not unlikely that the data are derived from an initial survey of the king's income and lands alone which the Exchequer Domesday scribe himself formulated in an attempt to bring it into conformity with the rest of the text.[13] Commonly, however, the scribe of the manuscript seems to be copying from an already compiled exemplar. It is thus unlikely that there was any extensive composition from diverse sources at this late stage.[14]

Emendations, corrections, and additions were, nevertheless, subsequently made, indicating that the scribe had access to a larger body of information, either an unabbreviated exemplar or separate sources used for checking. It is of vital importance to identify such changes for the stratigraphy of the text must clearly be established before the breves can be examined for traces of procedural activity and sources. Various devices were employed when material was added to the text. The most obvious are interlineation, usually, although not always, used for adding qualifying detail, and addition to the side margin. The foot margin was also used, and such additions can be detected by a line count if not by a difference in hand. Land in Tithby, for example, is enrolled below the 44th line of DB i, f.288c, and the entry is clearly postscriptal.[15] Additions to the end of the breve are more difficult to detect, but can often be identified from anomalous details of procedure such as irregularities in wapentake sequence, that is, the common order in which estates are described in every breve.[16] The three final entries of William Peverel's breve, for example, relate to land in the wapentakes of Bassetlaw and Broxtow. They are preceded, however, by estates in Bingham which, in normal circumstances, would have been enrolled after those in the other two wapentakes.[17] The entries, then, would appear to be postscriptal and, indeed, in this instance both the hand and diplomatic[18] of the text indicate that they are later additions to the text. Finally, whole entries are frequently squeezed into blank lines originally left for another purpose, or appended to the last line of existing entries. Thus, in the archbishop of York's breve, an estate in Ranskill was postscriptally entered on a blank line between two textual groups of manors,[19] and in the king's breve, Fenton was enrolled on the last line of a soke entry relating to Leverton.[20] All the postscriptal material identified, with the exception of those interlineations which are readily discernible in the manuscript and the Farley edition, are listed in Appendix 1.[21]

Wapentake rubrics are by no means general, and are only found regularly and correctly in breves nos 1, 6, 9, and 17. With only one exception, all are written on single lines in large rustic letters within the body of the text.[22] This might suggest that such information was an integral component of each breve, that is, it was entered as the text was written. The regularity of wapentake sequence found in almost every breve does indeed imply that the geographical location of each estate was known to the Domesday scribe, or to a scribe of some previous recension. But the information was clearly not thought to be vital for no rubrics at all are found in twenty two of the breves, and there is evidence to suggest that it was not always a current component of the text. In Lincolnshire, many, if not most, rubrics are evidently later additions, even when written on a blank line.[23] In Count Alan's breve, for example, wapentake and hundred rubrication is thorough, but the method of recording the information is entirely dependent on the space available in a previously drafted text.[24] Thus, a rubric was interlined or written on the last line of the preceding entry if space permitted, but was otherwise entered in the margin or a pre-existing space within the text. Rubrication, then, was clearly an afterthought and not originally considered germane to the purpose of the enquiry.

The same is probably true of the Nottinghamshire folios. Spaces, which are by definition current, often occur between the description of estates in different wapentakes, but do not contain rubrics. Moreover, blank lines are sometimes left in the text where no wapentake rubric can be intended. Berengar de Tosny, for example, held three manors which were all situated in the wapentake of Newark. The third, however, is separated from the other two by a one line space.[25] Any indication of the wapentake at this point would have been redundant for such information is usually only noted for the first estate in each division of the county. Indeed, the reason was apparently otherwise for, although the estate was in Newark Wapentake, the works of the villeins (opus villanorum) belonged to Saxilby in Lincolnshire.[26] Furthermore, spaces are almost certainly used in breve no. 5 to indicate different types of estate.[27] It is therefore possible that gaps were left in the text for one purpose and were only subsequently employed to record wapentakes. Other types of rubrication - the identification of hundred and soke - are rare and are in all cases clearly later additions to the text.[28]

In circuit 6 the basic unit of textual organisation within the breve is the manerium which is here, somewhat loosely, translated as 'manor'.[29] The caput, which is identified by the presence of an aula or hall, is described first, but the value assigned to it is that of the whole estate.[30] There then follows an account of the manorial appurtenances in separate entries, first the berewicks, then the land in soke. The form of the text, however, does not necessarily reflect the social and economic structure of estates for it is essentially a function of procedure. Thus, for example, 'intra-manorial' sokemen - those sokemen who are accounted for within the manorial caput[31]- seem to be of a similar status to those who are recorded in separate soke entries. In the absence of any parallel passages in Domesday Book itself, no direct comparison of the two classes can be made. But the fact is clear from an examination of the description of the manor of Scotter (Lincs.) in a c.1125 survey of the estates of Peterborough Abbey.[32] In 1086 there were 15 sokemen, 32 villeins, and 13 bordars in Scotter itself. There was sokeland in Scotterthorpe, recorded in a separate entry, where there were a further 8 sokemen and 4 villeins.[33] In c.1125 the two parcels of land were described as a single estate. There were 29 sokemen, 24 full villeins, 2 half villeins, and 10 bordars.[34] Clearly, the 15 intra-manorial sokemen of 1086 cannot be represented by the two half villeins of c.1125 - the latter probably represent the 4 bordars in Scotterthorpe in 1086. It is apparent, then, that there was only one class of sokemen representing both the intra- and extra-manorial sokemen of 1086. The former are only intra-manorial by virtue of proximity, and this is probably determined by the structure of local government.[35]

In the Nottinghamshire Domesday there are fewer berewicks and soke entries than in Lincolnshire. As in Derbyshire, 'manors' predominate, accounting for 61% of all entries, as opposed to 49% in the description of its more easterly neighbour.[36] However, the same basic structure is found. William Peverel's manor of Wollaton, for example, is enrolled in four consecutive paragraphs. The first relates to the centre of the estate in Wollaton itself, and it is followed by a berewick in Cossall and two parcels of sokeland in Bramcote and Sutton (Passeys). The value of the whole estate, 100 shillings in 1066 and 60 shillings in 1086, is recorded in the first entry.[37] There are, of course, as elsewhere, irregularities which arise from the exigencies of procedural convenience. Thus, in Ralf son of Hubert's breve no. 13, Leofric's manor in Barton (-in-Fabis) is separated from its berewick and soke in Clifton and the two Chilwells by another manor in Barton.[38] Geographical association may have prompted the enrolment of the second entry at this point. But it is more likely that it was, in its own way, also a dependent of Leofric's estate and was therefore entered as a manorial appurtenance.[39] Soke entries which are widely separated in the text from the manorial caput, however, may betoken the absence of a seigneurial return, that is, an account of each estate furnished by the tenant-in-chief to aid the Domesday commissioners, or accidental omission. Thus, soke of Bulwell in Watnall is separated from the manor by twenty entries, but was presumably enrolled after two manors in the same vill because the scribe did not have an account of the appurtenances of Bulwell before him. He therefore proceeded on a geographical basis, whether from a geld list or oral presentation.[40] The soke of Wysall in (King's) Thorpe and Willoughby was apparently forgotten and only subsequently added to the foot of the column with a sign to indicate its proper position.[41] Compound soke entries. in which a large number of parcels of land are described together, are also found in the Nottinghamshire folios. As elsewhere in circuit 6, they are usually, although not exclusively, associated with large estates of the terra regis[42] and, characterised by a geld total for the whole area - individual assessments are often later interlineations [43]- they appear to take their form from the administration of the estate. Thus, soke of the bishop of Lincoln's manor of Newark is described in three entries which comprise one, seven, and nine separate parcels of land in three distinct areas of the wapentake of Newark.[44] It is likely, then, that the source for this type of entry is entirely different from that of the more usual entries.[45] Finally, manor and berewick are occasionally described in a single entry, although the identification of inland is usually postscriptal.[46] This device is the norm in the Roteland and Derbyshire folios and is more a reflection of the structure of local government than significant differences in the nature or management of estates.[47]

Within the basic textual group of the manor, again in common with all the shires of the circuit, there are distinctive forms. Various calligraphic devices and diplomatic formulas are employed, in addition to, or instead of, the explicit record of relationship, to distinguish the status of individual parcels of land. The most obvious is the use of Lombardic capital letters - M for manerium, B for berewita, and S for soca [48]- which are prefixed to almost all entries (figure 1). Exceptions, often later additions to the text, may indicate uncertainty of status, if not simple omission. Roger de Bully, for example, held an estate in Fenton with sake and soke, but without a hall.[49] The absence of a marginal M may reflect this apparent contradiction. By way of contrast, it was merely omitted in the description of six manors at the end of William Peverel's breve for they are a later addition to the text.[50] Generally, however, marginal letters are consistently used, as in the Derbyshire, Huntingdonshire and Roteland folios, and the usage is therefore contrasted with those of the Yorkshire and Lincolnshire accounts in which various formulas were employed in turn.[51] Less apparent, but a no less significant, differentiating device is the treatment of initial letters and identifying place-names. All manorial entries begin with a statement 'In x, y habuit z carucatas terre ad geldum' (figure 1). The description of Walter de Aincurt's manor of Staunton, for example, begins 'M. In STANTUNE habuit Tori x. bouatas terre ad geldum'.[52] The initial I is square in form and shaded in red ink, and the place-name is written in large rustic letters and rubricated. By way of contrast, B and S entries usually begin 'In x, z carucate terre ad geldum'. Thus, soke of the manor of Staunton in Alverton, Flawborough and Dallington commences with the statement 'In Alureton et Flodberge et Dallintune vi. bouate terre ad geldum'.[53] The initial I is rustic in form and the place-name is in no way distinguish-ed from the rest of the text. These conventions of letter form are clearly used with purpose for initial I's are occasionally changed from one type to the other.[54] Deviations from the norm, then, are evidently significant and probably indicate anomalies in status. Queen Edith, for example, held a manor in Oakham with five berewicks, assessed at four carucates, and Leofnoth a second manor in the same vill which was rated at one carucate.[55] Both entries are manorial in form. The second, however, has a rustic initial I and a note is appended to it recording that the whole manor, with the berewicks, was three leagues in length and one league and eight furlongs in breadth. Leofnoth's manor, then, was clearly considered to be part of Queen Edith's estate, to which the berewicks belonged, and the form of the initial I seems to indicate its inferior status as a dependent of the larger manor. Similar examples, although less well-documented, can be found in the Nottinghamshire folios.[56] By way of contrast, William Peverel's berewick of Wollaton in Cossall appears to have undergone an upward change in status for the place-name of the entry is identical in form to that of a manorial counterpart.[57]

Figure 1: DB i, f.287a,b, the land of William Peverel.

The distinctive character of different types of entry is emphasised by the information that they provide. Manorial entries record the TRE tenant and the value of the estate in the time of King Edward and in 1086. Berewick and soke entries, however, only record this information in special circumstances, and then but rarely in the Nottinghamshire folios, for manorial appurtenances usually belonged to the lord's hall.[58] In forinsec soke entries - that is, a type of entry in which the land was held by one tenant-in-chief, while its soke was enjoyed by another - the record of a holder in 1066 identifies the soke lord. Thus, (Earl) Algar is recorded as the lord of sokeland in Willoughby (in the Wolds), but he clearly only had its soke, probably in (Upper) Broughton.[59] But this convention is more common in Lincolnshire than Nottinghamshire where an explicit statement of relationship is usually made. Walter de Aincurt, for example, held sokeland in Fiskerton, Morton and Farnsfield, but the soke is said to belong to the archbishop of York's manor of Southwell.[60] In other contexts, the record of a TRE tenant of sokeland indicates some degree of independence from the administrative machine of the manor. Soke of the royal estate of Arnold in Gonalston, for example, was in some way the right of Ernwin and four sokemen who are recorded as the tenants in 1066.[61] A separate value for a parcel of sokeland, whether intentionally or incidentally, performs much the same function. As conventional sums, the valet and valuit figures are clearly farms,[62] and therefore imply a separate unit of management.[63] Thus, twenty sokemen in Leverton in the soke of Oswaldbeck rendered 20 shillings in consuetudines and were therefore probably otherwise free of manorial exactions.[64]

All three types of entry, however, invariably contain a record of geld assessment, and commonly of teamlands, teams in demesne, population and ploughs.[65] Throughout the Nottinghamshire folios, the three main classes of peasant are recorded in a standard order of sokemen, villeins, and bordars. Slaves, where they appear, follow villeins or are relegated to the manorial stock. The amount of land held by the intra-manorial sokeman, as measured by assessment to the geld, is not universally recorded, but the information is found throughout the county survey. These characteristics of the record of population, along with the consistent use of marginal M, B, and S, are in marked contrast with the usage of the first half of the Lincolnshire text. There, it would seem, the Domesday scribe, or a predecessor, experimented with various devices to indicate the separate status of sokemen and sokeland. In manorial entries, peasants are listed in the order of villeins, bordars and sokemen, and the land of sokemen, although sometimes recorded in terms of carucates or teams, is usually not readily distinguishable from that of the villeins. The record of soke exhibits similar fluidity of diplomatic and form. Sokemen are invariably listed first, but marginal S is not used consistently until breve, no. 18 and various formulas are used in its stead. It is only from breve no. 24, f.355a, that all uncertainties are dispelled and the scribe adopts the conventions that are found in the Nottinghamshire folios.[66] The Yorkshire Domesday exhibits a similar fluidity of form which is never satisfactorily resolved.[67] The baldness of the Nottinghamshire text, then, is probably a result of its compilation after the ad hoc emergence of satisfactory formulas in the early stages of composition.[68]

Economic assets, such as woodland, meadow, pasture, fisheries, and willow beds, occur in most entries. Mills and churches, however, are found predominantly in manorial entries. This pattern may reflect a seigneurial monopoly, especially in the case of mills,[69] but it is more likely to be a function of the selective processes of the Domesday enquiry.[70]

The Roteland Domesday, although appended to the Nottinghamshire account, forms a separate section in the text.[71] The preamble to the description of the area, however, indicates that it was integrated into the administration of the shire. The wapentakes of Martinsley and Alstoe, of which it was composed, belonged to the sheriffdom of Nottingham, and half of Alstoe gelded in Broxtow and half in Thurgarton.[72] It can, then, be legitimately studied as part of the Nottinghamshire Domes-day. Its interest and importance lies in two areas. First, it contains a record of its liability to the geld in terms of twelve-carucate hundreds and is therefore of great significance to an understanding of the carucation of Nottinghamshire.[73] Second, and uniquely in the Exchequer Domesday, it is geographically arranged.[74] As such, it provides an invaluable key to the procedure of the Nottinghamshire Domesday and the structure of its local government machinery. In all other respects, the form of the Roteland text is the same as that of Nottinghamshire folios.

These, then, are the basic characteristics of the text. In the following analysis of the Nottinghamshire Domesday the importance of an understanding of the structure and stratigraphy of each breve will become apparent. It would be inappropriate, however, to discuss the form of each section of the text in the detail required before the data can be used with confidence. But the basic method is illustrated in figure 2. In each breve a record is made of all essential details of every entry - entry number, wapentake number, status as indicated by Domesday Book, TRE holders, TRW tenant - along with non-standard relationships and anomalous forms and information. Duplication and postscriptal material are carefully noted.

Figure 2: breve no. 10, the lands of William Peverel

NO	WP	STA	VILL	TRE TENANT	TRW TENANT	ANOMALOUS INFORMATION
1	4	M	Colwick	Godric	Waland	Waland holds
2.	1	2M	Sibthorpe	Leofwin	Robert
3	4	M	Gonalston, Milton	Wulfsi cilt
4	5	M	Thrumpton	Staplewin
5	5	M	Clifton	Countess Gytha
6	5		ibid.	Ulfkell	Ulfkell	deleted; equivalent to 30,25
7	5		Barton			addition to margin
8	5	S	Wilford
9	5	S	West Bridgeford
10	5	S	Normanton, Keyworth			individual assessments interlined
	6		Willoughby, Stanton			Willoughby equivalent to 30,25
11	6	S	Costock
12	5		Adbolton			addition
13	5	S	Bassingfield
14	5	S	Gamston
15	6	M	Radford	Aelfric	Wulfnoth	Wulfnoth holds 1 bov. thaneland
16	6	4M	Stapleford	Wulfsi cilt	[Robert]
				Staplewin
				Godwin
				Gladwin
17	6	M	Morton	Bovi
18	6	M	Newbound	Morcar
19	6	S	Lenton
20	6	3M	Linby	3 brothers
21	6		Papplewick			5 bov. belong to this manor; addition
						to blank line?
22	6	M	Basford	Alwin	Saxfrid
23	6	M	ibid.	Aswulf		deleted; equivalent to 52 and 30,28;
						in custody of William Peverel
24	6	M	Lenton	Wulfnoth	Wulfnoth	in custody of William Peverel
25	6	M	Toton	Haldane	Warner
26	6	S	Chilwell			equivalent to 13,5
27	6	M	Strelley	Godric	Godwin	Godwin [the priest] has one plough
						from William Peverel
28	6	M	ibid.	Brown	Ambrose	Ambrose holds from William
29	6	M	Greasley	Wulfsi
30	6	M	ibid.	Wulfsi	waste	Alric holds from William
31	6	M	Brinsley	Brown	Alric	Alric has one plough under William
32	6	M	Eastwood	Ulfketel	waste	William has charge; value TRE only
33	6	M	Newthorpe	Grimkell	waste	equivalent to 62-3?
34	6	3M	Beeston	Alfheah
				Alwin
				Ulfkell
35	6	M	Wollaton	Wulfsi cilt	Warner
36	6	B	Cossall			place-name manorial in form
37	6	S	Bramcote		waste
38	6	S	Sutton		waste	equivalent to 30,55
39	6	2M	Bilborough	Alric	Ambrose
				Wulfsi Swein
40	6	M	Nuthall	Haldane
41	6		Broxtow
42	6	S	Watnall
43	6	M	Watnall	Grimkell		no value given
44	6	M	Watnall	Siward		no value given
45	6	S	ibid.	Grim
46	6	S	ibid.	Aelmer	Jocelyn	value, of 43-6?; soke of 66; Jocelyn
					Grimkell	and Grimkell hold, 43-6?
47	6	M	Kimberley	Azor
				Grimketel
48	6	M	ibid., [Awsworth]	Alwin	waste	no value; 'William has charge' add.
						equivalent to 30,33?
49	6	M	Hucknall	2 brothers
50	6	S	Hempshill			soke of 66 and 43 or 44
51	6	2M	Basford	Alfheah	Payne
				Algot	[Saxfrid]
52	6		ibid.	Aswulf		addition and equivalent to 23 and
						30,28
53	6	M	Costock	Fredegis	Godwin	Godwin has under William
54	6	M	Rempstone	Fredegis
55	7	2M	Radcliffe	Fredegis	Fredegis	Fredegis and Wulfgeat have under
					Wulfgeat	William
56	7	M	Adbolton	Godwin the
				priest
57	7	M