Domesday Survey: Introduction III

Pages 90-94

A History of the County of Middlesex: Volume 1, Physique, Archaeology, Domesday, Ecclesiastical Organization, the Jews, Religious Houses, Education of Working Classes To 1870, Private Education From Sixteenth Century. Originally published by Victoria County History, London, 1969.

This free content was digitised by double rekeying. All rights reserved.



The social classes mentioned in the Middlesex Survey are francigene, Anglici, milites, villani, bordarii, cotarii, and servi. In addition there are recorded a number of homines, and a few burgenses appear as holders of arable land. There are no references to other classes such as the liberi homines and sochemanni of the eastern counties or the radmen of the west. There are no references to servientes or, at the other end of the scale, to coliberti (or burs). The structure of Middlesex rural society seems therefore to be free from the complexities met with in many other counties. The terminology of Domesday is often, however, deceptively simple and as Sir Frank Stenton has said: 'No line of research will ever give to Domesday Book the precision of a well drafted medieval survey. But the vagueness that baffles the modern enquirer is itself a significant fact, for it reflects a society on which historical forces had been playing for many generations to the blurring of class distinctions and the confusion of personal relationships.' (fn. 1)

On five manors, Fulham (17), Hillingdon (56), Ruislip (74), Isleworth (80), and Tottenham (96), men styled francig' or francigene are listed among the tenants. As a class they defy definition. Round comments that, 'beyond the fact that they were Frenchmen by birth, it is not easy to say of whom this class was composed'. (fn. 2) In the same paragraph he noted, from the Worcester evidence, that probably many were 'serjeants' of various kinds whose services were rewarded by land. In Kent there is a reference to francigene milites, (fn. 3) and in Shropshire one of the complaints of the English burgesses related to tenures formerly held by them, and held in 1086 by francigene burgenses. (fn. 4) It is also of interest to note, as Freeman pointed out, (fn. 5) that the words francigene or franci, used in the sense of Frenchmen, apply to all William's followers wherever they came from. There is nothing to be learnt from their position among the various classes enumerated in Middlesex. Usually they follow after the peasant classes, but at Tottenham (96) they are placed between the cotarii and the servi and at Greenford (43) a francigena is placed between the bordarii and th c ii. At Hillingdon two of these men hold 1½ hide, at Ruislip there are four on 3 hides 1 virgate, and at Tottenham there are two on 1 hide 3 virgates. Some have men under them. Thus at Fulham francigene et quidam burgenses Lundon' have under them inter villanos et bordarios xxxi, at Hillingdon sub istis manent iii homines, and at Isleworth francigene et quidam Anglicus have under them inter villanos et bordarios xii. In this last entry both the francigene and the Anglicus are said to be milites probati.

Milites, who were all probably trained knights, are entered on nine Middlesex estates. At Westminster (36) reference to 25 houses belonging to the abbot's knights and other men provides one of the clearest examples of household knights to be found in Domesday. In 'Ticheham' (58) three of Earl Roger's tenants were knights. There, as in other places, they had men under them. At Harrow (4) three knights and a priest had seven men and at Stanwell (76) there were two knights on 2½ hides et sub eis vi bordarii manent. It can be seen from the Isleworth entry that not all knights were Frenchmen and this conclusion is borne out by other evidence. In Wiltshire the nephew of Bishop Harman is referred to as a 'miles by command of the king', (fn. 6) and Stenton has remarked on the king's command to Lanfranc to make knights of his 'drengs'. (fn. 7) It is clear none the less that although not all francigene were milites, most milites were francigene.

The phrase inter franc' et villanos occurs in connexion with the ownership of the plough-teams in a number of places. The extension of franc' in these cases is of considerable importance. Maitland (fn. 8) and Vinogradoff (fn. 9) extended it to francus and used the phrase inter francos et villanos to support the view that the villani were not free. The phrase in this latter form is used in Shropshire (fn. 10) where the word francos has been rendered by its translator 'freemen'. (fn. 11) But the words francus and francigene occur elsewhere in the same county, and there is doubt concerning the correct rendering, (fn. 12) although it is highly probable that here as in Devon, (fn. 13) where it is contrasted with Anglici, francus means Frenchman. In Middlesex the phrase inter franc' et villanos occurs only on estates where there are either milites or francigene and franc' should therefore probably be extended to francigene. Whether the abbreviation stands for francigene or franci the word clearly means Frenchmen.

In Middlesex, as in other counties, the most numerous class of peasants is that of the villani and their description as 'the backbone of the rural community' (fn. 14) is well demonstrated here. The number recorded in Middlesex is 1,141, (fn. 15) considerably fewer than in neighbouring counties (fn. 16) but comparable both in proportion to the size and assessment of this county. It has long been recognized that the Domesday villanus cannot be equated with the villein of later times and that the word as used in Domesday carries with it the vaguer connotation of the Old English word tunesman, the inhabitant of a village, and for which the term 'villager' would seem to be a more appropriate modern rendering. (fn. 17)

The Middlesex Survey is notable for the information it records concerning the size of the peasant holdings. (fn. 18) In most entries the size of individual holdings is given and the compilers state the number of men with tenements of the same size, beginning with the largest. For instance at Enfield (73) the list is: Ibi unus villanus de i hida et iii villani quisque de dimidia hida Presbyter i virgata, et vii villani quisque i virgata, xxxvi villani quisque dimidia virgata. Some entries are less specific, recording only how much land a number of men held between them and leaving the size of the individual holdings in doubt. The entry for West Bedfont (78), for example, reads: Ibi ii villani de iiii hidis et ii villani de ii virgatis et ii villani de i virgata and does not state how the holdings were divided between each pair of men. The largest tenement held by a villanus in Middlesex is 2 hides; there are two who are specifically stated to have this large holding. Only 18 villani hold less than ½ virgate. (fn. 19)

Apart from the information about holdings the Middlesex Domesday does not throw much light on the social standing of the villani and it gives no information about their services. Payment of rents by them is not usual, two villani each holding ½ hide on the wife of Brien's tenement at Stepney (8) pay respectively 4s. and 8s. de domo sua, but other payments are confined to the lower economic ranks of the peasantry. Willesden (22), a 15-hide manor belonging to the Canons of St. Paul's was farmed by the 25 villani there, and on this manor there was no demesne. On some other holdings belonging to St. Paul's there was no demesne and the land is said to be held under the canons (sub canonicis) by the villani.

In this county, as in others, the bordarii and cotarii form a considerable part of the peasant population. The number of bordars in Middlesex, exclusive of those at Fulham (17), Isleworth (80), and Stepney (85) where the numbers are not given in the text, is 342, (fn. 20) slightly less than one-third of the number of villani and slightly fewer than the number of cottars. Although there is little to be added to what has already been written about this class of peasants who derived their name from their borda or cottage, the references to bordars in the Middlesex Survey deserve close examination since the Middlesex Survey is the main source of information about the tenements of the men in this class in 1086 no less than those of the villani. The Middlesex evidence throws considerable light on the differences between the three classes, villani, bordarii, and cotarii. The bordars' holdings are given in Appendix II. There are only 9 who have no holdings; 54 have an average of fewer than 5 acres each; (fn. 21) 98 have 5 acres each; 48 have more than 5 but fewer than 10 acres; 55 have 10 acres each; 14 have between 10 and 15 acres; 34 have 15 acres (i.e. ½ virgate) each; and 7 have an average of more than 15 acres each. Four bordars on the manor of Dawley (57) and 6 bordars at Stepney (13) are each recorded as holding 5 acres between them and it is not unlikely that the word 'quisque' has been omitted in the text. (fn. 22) It is evident that 55 of the bordars in Middlesex were, economically, in substantially the same position as the half-virgate holders among the villani. The most common holding is 5 acres, but it is also clear that those with more than 5 acres are more numerous than those with fewer. (fn. 23) This helps to explain why the compilers of the Survey frequently associate the bordars with the villani, sometimes stating how many virgates the two classes together occupied, (fn. 24) and why it is stated that there are 3 ploughs among the villani on an estate where there are 14 bordars on 1½ hide (9) but no villani. Occasionally, however, the bordars seem to be classed with the cottars rather than the villani. (fn. 25)

The Middlesex Survey shows that in economic position the bordars as a class stood below the villani and above the cottars, although the classes overlap. The analysis of the cottars' holdings given in Appendix II shows that their holdings range from nothing to just over 5 acres. (fn. 26) More than half can be regarded as landless men. Out of 464 cottars there are 243 with no holding and 49 with only their gardens or allotments attached to the cottages. There are 67 with approximately 1 acre or less; 8 who have an average of 2 acres; 93 who have from 2 to 5 acres each; and 4 only with 5 acres or more. The cottars recorded on the Terra Regis, it has been suggested, (fn. 27) are squatters or 'at any rate scattered tenants with infinitesimal holdings', and they are placed under the king because they have no other lord. Although a few cottars in Middlesex seem to be economically of the same standing as the bordars, the evidence suggests that most must have been living in much humbler circumstances. It suggests that they were in fact 'cottagers' in the modern sense of the word with even fewer acres than the bordars, or with no land apart from their cottages and the common land. Some cottars are required to make annual money payments. Thirty unattached cottars who are listed under the king's lands (2) pay 14s. 10½d. a year (qui reddunt per annum xiiii solidos et x denarios et i obolum) and two more at Holborn pay to the King's sheriff 20d. a year. Another ten at Bishopsgate (32) hold jointly 9 acres for which they are required to pay 18s. 6d. a year, and it is not quite clear whether the 46 cottars on 1 hide at Stepney (6) are responsible for the whole of the annual payment of 30s. or whether it is laid upon the whole of the peasant population on the manor. In all probability it was exacted only from the cottars.

Divergent answers have been given to the question whether the bordars and cottars are two different classes. The cottars and 'coscets' do not, like the villani and bordarii, occur in all counties. They are most numerous in Kent, Sussex, Surrey, Middlesex, Wiltshire, Dorset, Somerset, Berkshire, Hertfordshire, and Cambridgeshire, and less numerous in Buckinghamshire, Devon, Herefordshire, Worcestershire, Shropshire, and Yorkshire. (fn. 28) In some of these counties, of which Middlesex is one, bordars and cottars are recorded in the same entries, but in others they are not. For instance, in Berkshire bordars only are found on the estates of the king, Abingdon Abbey, Earl Hugh, and William fitz Ansculf; cottars only are found on the small estates of the Bishop of Durham and Bishop Osbern; and bordars and cottars are found on the estates of the Bishop of Winchester, Walter Giffard (Gifard), and Henry de Ferrers (Ferieres), but on none of these last three estates are the two classes of peasant mentioned together in the same entry. Sussex is another county in which bordars and cottars do not appear together. In Surrey 'a marked local distribution' has been noted. (fn. 29) It was once argued that the terms bordarii and cotarii are synonymous and interchangeable. (fn. 30) The study of the surveys of Shropshire, Wiltshire, Worcestershire, and Herefordshire has led commentators to the opposite conclusion. (fn. 31) The Middlesex evidence is here of fundamental importance for it contradicts decisively the view that the two classes are identical and shows that the bordars as a class were superior to the cottars and that some of them must have been almost indistinguishable from the less prosperous villani. The points of similarity between the bordars and the villani must not, however, be overstressed, for too little is known about the services rendered by these classes. (fn. 32)

It has long been recognized that the servi of Domesday were comparatively few in the east and increased in number towards the west. (fn. 33) In Middlesex there were only 112. Some of the entries in the county support the view that the servi were associated with the demesne plough-teams. Instances in which there are two servi to each demesne plough-team occur at Shepperton (42), Feltham (62), Kempton (63), West Bedfont (78), Hanwell (44), Harmondsworth (48), Stanmore, (90), and Tottenham (96). With these may possibly be included Colham (55) where there were three demesne plough-teams and eight servi, and Dawley (57), which in the Survey is associated with Colham, (fn. 34) where there were no servi but one plough-team in demesne. Stanwell (76) contained three demesne plough-teams and eight servi and at Bedfont (77), the greater part of which had been (fuit) a berewick in Stanwell, there was one demesne plough-team but no servi are recorded. The remaining entries recording servi show a considerable variation in the proportion of them to plough-teams. (fn. 35)

On several estates priests were numbered among the tenants, while at Laleham (88) Estrild, described as a nun, held an 8-hide estate of Robert Blund.

Table 4
Classes of Tenants in Middlesex

Edmonton Elthorne Gore Hounslow Ossulstone Spelthorne Totals
francigene 2 7 .. .. .. .. 9
anglici .. 3 .. 1 .. .. 4
milites .. 7 3 .. .. 4 14
burgenses .. .. .. .. .. 46 46
homines .. 3 7 .. 24 .. 34
villani 139 244 148 140 323 147 1,141
bordarii 49 82 15 4 107 85 342
cotarii 61 66 15 6 283 33 464
servi 14 42 5 .. 16 35 112
Totals 265 454 193 151 753 350 2,166

Not included above:

Fulham (17) 'Inter francigen' et quosdam burgenses Lundon'.

Sub eis manent inter villanos et bordarios xxxi'.

Westminster (36) 'et xxv domus militum abbatis et aliorum hominum' Isleworth (80) 'Francig' et . . . Sub eis manent inter villanos et bordarios xii' Stepney (85) 'et bord' de dimidia hida et dimidia virgata'.


  • 1. F. M. Stenton, Anglo-Saxon England, 470.
  • 2. V.C.H. Worcs. i. 273. Stenton, Eng. Feudalism, 144, considers the term to be a 'vague word' which will cover serjeants and household officers as well as knights and their squires.
  • 3. Dom. Bk. (Rec. Com.), i. 12.
  • 4. Ibid. 252.
  • 5. E. A. Freeman, Norman Conquest, v. 766.
  • 6. V.C.H. Wilts. ii. 82, 121 (no. 39).
  • 7. Stenton, Eng. Feudalism, 145-6.
  • 8. Maitland, Dom. Bk. 46, n. 1.
  • 9. Vinogradoff, Eng. Society, 452, n. 4.
  • 10. Dom. Bk. i. 252b (entry for Sudtone).
  • 11. V.C.H. Salop. i. 313.
  • 12. Ibid. 311 (Dom. Bk. i. 252): lidberie francus is translated 'Frenchman', with a note that more probably it should be translated 'a freeman'. On f. 254b in three succeeding entries, the first two, Lestone and Arculum, use francigena and the third, Uptune, francus, all obviously with the same meaning and translated in the V.C.H. text (pp. 320-1) as 'Frenchmen'.
  • 13. Dom. Bk. i. 114 (Aisse).
  • 14. V.C.H. Worcs. i. 274.
  • 15. This figure excludes the entries for Fulham (17) and Isleworth (80) where villani and bordarii are lumped together. A figure of 1,163 is given in Dom. Geog. S.E. Eng., ed. Darby, 118. This includes 22 villani obtained by the arbitrary division of 31 villani et bordarii (Fulham (17)) into 16 villani and 15 bordars, and 12 villani et bordarii (Isleworth (80)) into 6 villani and 6 bordars.
  • 16. e.g. in Herts. there were 1,830, in Berks. 2,623, in Essex 4,087, and in Surr. 2,363. The figures are from H. Ellis, Introd. to Domesday Bk. vol. ii; they conflict with those given in Dom. Geog. S.E. Eng., ed. Darby.
  • 17. V.C.H. Hunts. i. 324. See also Maitland, Dom. Bk. 59. V. H. Galbraith, E.H.R. lxv. 4, translating an entry relating to Halton Hundred (Exon Domesday, f. 19b) renders terra villanorum 'villager's land'. In the translation on pp. 119-29 the word villanus, following accepted usage, is translated 'villein'.
  • 18. The most recent discussion of peasant tenements is that of Lennard, Rural Eng. 339-92, where full use is made of the Mdx. evidence.
  • 19. These figures include those which are specifically stated and those which are inferred. For a complete analysis of villani holdings see App. II.
  • 20. This figure conflicts with that of 364 given in Dom. Geog. S.E. Eng., ed. Darby, 118. See above, n. 71.
  • 21. Only average figures can be given for some bordar and cottar holdings since in many instances they are inferred.
  • 22. The text should probably be amended to 'ibi iiii villani quisque de i virgata et iiii bordarii quisque de v acris' which would give the bordars tenements more in keeping with holdings of men in this class in estates in Harlington, Colham, Hillingdon, and 'Ticheham', which belonged to the fief of Earl Roger and were adjacent to Dawley. Since the text may be correct as it stands the four bordars are shown in the table as sharing the 5a between them.
  • 23. It may be noted that in the only entry (Hailet, f. 140) which records the size of the peasant holdings in Herts. one bordar has 10a.
  • 24. e.g. Haggerston (83): 'Ibi iii villani et vii bordarii qui tenent hanc terram'.
  • 25. e.g. Islington (28): 'Ibi iiii villani qui tenent sub canonicis hanc terram, et iiii bordarii et xiii cotarii'.
  • 26. See above, n. 77.
  • 27. Vinogradoff, Eng. Society, 238.
  • 28. Maitland, Dom. Bk. 39.
  • 29. Round, V.C.H. Surr. i. 292, remarks that Mr. Malden, translator of the text, 'has also observed in the Survey "a marked local distribution" in the classes of bordars (bordarii) and cottars (cotarii) which he is unable to explain. In the hundreds of Godalming, Wallington, and Emleybridge "cotarii" he has found "are nearly universal to the exclusion of the bordarii", and on no estate in the county has he found both classes mentioned. "The only rule", however, he considers, "seems to be that there should be no cotarii on the royal demesne".'
  • 30. F. Seebohm, English Village Community, 77, 95-97; A. Ballard, Domesday Inq. 152.
  • 31. V.C.H. Salop. i. 301; V.C.H. Wilts. ii. 54; V.C.H. Worcs. i. 274, 276; V.C.H. Herefs. i. 288.
  • 32. The Herefs. peasants, who worked one day a week and so resembled the cottar of the 'Rectitudines' (translated in Eng. Hist. Docs., ed. Harmer, ii. 813 sqq.) who had 5a and worked for his lord each Monday throughout the year, are entered not as cotarii but as bordarii.
  • 33. Maitland, Dom. Bk. 23; Vinogradoff, Eng. Society, 464.
  • 34. See p. 124.
  • 35. In the translation on pp. 119-29 the word servus, following accepted usage, is translated 'serf'.