An earlier version of this paper was presented at the 1981 Social Science History Association conference in Nashville, Tennessee.

¹ Much of the work of the last few decades has been synthesized in the important new study by Wrigley, E. A. and Schofield, R. S., The Population History of England, 1541-1871: A Reconstruction (Cambridge, Mass., 1981)Google Scholar. Peter Laslett's pioneering studies have generally focused on familial, rather than demographic issues, see The World We Have Lost (New York, 1965)Google Scholar, Family Life and Illicit Love in Earlier Generations (Cambridge, 1977)Google Scholar, and Household and Family in Past Time (Cambridge, 1972)Google Scholar. Research articles by members of the Cambridge Group can usually be found in the two journals Local Population Studies and Population Studies. An important methodological guide to parish register research is edited by Wrigley, E. A., An Introduction to English Historical Demography (London, 1966)Google Scholar. Another collection edited by Wrigley, E. A., Identifying People in the Past (London, 1973)Google Scholar, addresses the complex problems involved in record linkage.

² Although much amended by more recent work, the pioneering study of medieval demography is Russell's, Josiah CoxBritish Medieval Population (Albuquerque, NM, 1948)Google Scholar. J. Hajnal first identified the unique characteristics of early modern marriage in “European Marriage Patterns in Perspective,” in Population in History, ed. Glass, D. V. and Eversley, D. E. C. (London, 1965), pp. 101–43Google Scholar. Richard Smith has recently postulated—through reanalysis of the poll taxes of 1377, 1379, and 1381—that this unusual pattern might have existed as early as the fourteenth century: “Some Reflections on the Evidence for the Origins of the ‘European Marriage Pattern’ in England,” in The Sociology of the Family: New Directions for Britain, ed. Harris, C. (Sociological Review Monograph, 28) (Totowa, NJ, 1979), pp. 74–112Google Scholar.

³ Homans, George C., English Villagers of the Thirteenth Century (Cambridge, Mass., 1941)CrossRefGoogle Scholar.

⁴ For example, see Keith Wrightson's thoughtful critique, “Medieval Villagers in Perspective,” Peasant Studies, 7 (1978), pp. 203–17Google Scholar.

⁵ See, for example, the description of manorial records by Macfarlane, Alan in Reconstructing Historical Communities (Cambridge, 1977), pp. 46–57Google Scholar. Macfarlane identifies these records as major sources for studying the rural life of late medieval England, but he focuses solely upon their early modern characteristics (describing, for example, an almost exclusive focus in manorial courts upon property transfers). He also states, without an adequate recognition of the extraordinary depth of medieval materials, that medievalists concentrate upon these records because of the absence of other materials that facilitate the work of early modernists (p. 49).

⁶ Raftis, J. Ambrose, Tenure and Mobility (Toronto, 1964)Google Scholar and Warboys (Toronto, 1974)Google Scholar; DeWindt, Edwin Brezette, Land and People in Holywell-cum-Needingworth (Toronto, 1972)Google Scholar; Britton, Edward, The Community of the Vill (Toronto, 1977)Google Scholar.

⁷ Hilton, R. H., The English Peasantry in the Later Middle Ages (Oxford, 1975)Google Scholar; Razi, Zvi, Life, Marriage and Death in a Medieval Parish (Cambridge, 1980)Google Scholar. It is worth noting that Hilton's approach to the medieval peasantry (in terms of both sources and geography) ranges more widely than that adopted by Raftis. Razi's study, however, shares with the published works of the ‘Toronto School’ a focus on the manorial records of a single community.

⁸ For the best methodological critique, see Wrightson, “Medieval Villagers.” Most of Wrightson's reservations are echoed by Razi, in “The Toronto School's Reconstitution of Medieval Peasant Society: A Critical View,” Past and Present, 85 (1979), 141–57CrossRefGoogle Scholar. Razi has also leveled sharp criticisms of the techniques and conclusions of Raftis and his colleagues in “Family, Land and the Village Community in Later Medieval England,” Past and Present, 93 (1981), 3–36CrossRefGoogle Scholar. Richard Smith of the Cambridge Group has avoided partisanship in this fray; see his dissertation, “English Peasant Life-Cycles and Socio-Economic Networks: A Quantitative Geo-graphical Case Study,” (Ph.D. thesis, Cambridge Univ. 1974)Google Scholar.

⁹ For a concise introduction to English manorial records, see Titow, J. Z., English Rural Society 1200-1350 (London, 1969)Google Scholar.

¹⁰ Both Hilton and Raftis began their academic careers with such studies, see Hilton, R. H., The Economic Development of Some Leicestershire Estates in the 14th and 15th Centuries (London, 1947)Google Scholar, and Raftis, J. Ambrose, The Estates of Ramsey Abbey (Ibronto, 1957)Google Scholar. Two notable recent studies are Harvey, Barbara, Westminster Abbey and its Estates in the Middle Ages (Oxford, 1977)Google Scholar and Dyer, Christopher, Lords and Peasants in a Changing Society (Cambridge, 1980)Google Scholar.

¹¹ The most comprehensive introduction to village court rolls is Raftis's pioneering study, Tenure and Mobility.

¹² According to legal theorists, manorial courts were, in fact, broken into four discrete parts: the court baron (for free tenants), the court customary (for servile tenants), the view of frankpledge (to oversee tithing organization), and the court leet (to adjudicate major infractions of the peace). In actual practice, these legal jurisdictions were frequently combined. On most manors, the courts baron and customary were merged into the three-weekly curia sessions, while the frank-pledge and leet jurisdictions were merged into semi-annual or annual meetings. The records of these latter sessions usually include information on the administration of tithings (with young men either entering tithings or being fined for non-entry), on infractions of the peace (focusing on incidents that involved the entire community because of the raising of the hue and cry), and on commercial activities (bakers, butchers, and other petty salespeople were often liable only for annual fines paid at the frankpledge court). It should be noted that the suggestions raised below—on the historical use of both regular court and frankpledge-leet sessions—apply only to rural court rolls. The courts convened in towns or large, market villages could cover different sorts of information. Although some of the techniques proposed below might be applicable to such archival materials, they will not be specifically considered in this essay.

¹³ Information on ages, marriages or deaths can occasionally be inferred from other sources (boys usually entered tithing groups at 14 years of age; some marriages were mentioned, long after the actual weddings, in inheritance cases or land disputes; and in rare instances, wills or testaments written by peasants have survived). Such information, like data on family relationships, is not consistently provided. It should be noted that data on female peasants are especially limited. Because the village courts were a public forum, males predominate among persons cited in the records. Women not only appeared much less frequently, but also are more difficult to trace because they usually changed their names upon marriage.

¹⁴ Bennett, Judith MacKenzie, “Gender, Family and Community: A Comparative Study of the English Peasantry, 1287-1349,” (Ph.D. thesis, Univ. of Toronto 1981), pp. 46–48Google Scholar.

¹⁵ Razi, , Life, p. 10Google Scholar.

¹⁶ The rule of four to five inheritance and marriage entries per year is a very flexible guide. Razi found this rate for recorded marriages in Halesowen, , Life, pp. 45-50, 152–53Google Scholar. It is, however, impossible to postulate the average occurrence of inheritances or marriages in any given manorial environment because we lack some basic demographic data; we do not know (1) the exact size of the population under study (population estimates from court rolls are always exceedingly tenuous and cannot be made until after all data collection is completed), and (2) the number of deaths or marriages that would escape court attention (because they concerned non-tenants). Hence, medievalists cannot apply strict tests such as those proposed by D. E. C. Eversley for selecting parish registers (see “Exploitation of Anglican Parish Registers by Aggregative Analysis,” in English Historical Demography, ed. Wrigley, , pp. 54–56Google Scholar).

¹⁷ Brewing citations must be eliminated because the number of women involved in ale production varied from village to village, and this variance dramatically affects ratios of female appearances. Consider, for example, the data for the pre-plague records of Brigstock (549 courts), Iver in Buckinghamshire (71 courts) and the Ramsey Abbey manor of Houghton (32 courts). In these records, women respectively accounted for 22 per cent, 14 per cent and 13 per cent of all court citations. But this wide range is narrowed considerably by eliminating brewing citations; without ale fines, women respectively accounted for 12 per cent, 11 per cent and 9 per cent of all appearances. See Bennett, “Gender,” pp. 194, 276, 330, and 331.

¹⁸ See Smith's studies of Redgrave and Rickinghall in “Life-Cycles,” and Razi's study of Halesowen in Life.

¹⁹ Razi's study (Life), for example, is based upon such a fine series of court records that the lack of ancillary sources scarcely matters.

²⁰ The best guide to the process of data extraction is Macfarlane, 's Reconstructing pp. 81–112Google Scholar. Although his elaborate suggestions might not be within the reach of those who lack research grants and research assistants, his proposals can be modified to suit the needs of individual reseach projects. Hence, the duplication of all cards (by using carbon copies) might not be feasible, but researchers should, nevertheless, construct (during the actual process of data collection) indices of important citations. For example, researchers should keep separate lists of officers, of tradespeople, of placenames, and of extraordinary name citations (especially those in which aliases are indicated). Researchers should also construct systems of colored cards and tabs to flag data that are especially relevant for the projected analyses. Although the immense bulk of the data collected make computer analysis attractive, the initial stages of research are best done by hand. The early processes of data analysis—separating individuals from names and families from surnames—are too complex and idiosyncratic to be amenable to standardized programming. At subsequent stages, however, the data (once sorted into individuals and families) could be analyzed with the aid of computers.

²¹ In the interests of clarity, some of the issues involved in individual identifications are discussed below under the techniques of family reconstruction. Hence, the initial process of alphabetization requires, of course, the judicious sorting out of Latin and English surnames. Although the reconstruction of individual histories must precede family reconstitutions, the two processes are inextricably intertwined. In many cases, individual histories will be revised and altered in the light of family reconstitutions, as, for example, the separate careers of fathers and namesake sons become more distinct.

²² Wrightson, “Medieval Villagers,” and Razi, “A Critical View.”

²³ The occurrence of persons identified without some sort of surname is so rare that it is not even discussed in most studies (see, for example, Razi, , Life, pp. 3-4, 11–26Google Scholar). The exact incidence of citations without surnames from a roll of courts held in Iver (Buckinghamshire) between 1332 and 1376 was 78 (0.57 per cent) of 13, 761 citations. Of these 78 cases, 3 persons were identified by unusual personal names that sufficed for identification, 3 were identified by two personal names (John son of Waba), and the remaining 72 were identified by official position (29 cases), occupation (3 cases) or ecclesiastical position (40 cases). These data were extracted from the roll deposited at the Buckinghamshire Archaeological Society, 128/53. Until quite recently, most non-medievalists have assumed that surnames were a fairly modern phenomenon. In the first edition of The World We Have Lost (London, 1965), Peter Laslett claimed that “Not so long before the reign of Elizabeth most Englishmen were without even surnames,” p. 44. He wisely deleted this unfounded assertion from the second edition (1971).

²⁴ Even Razi, who has made much of the instability of surnames, concedes that the problem is largely confined to the pre-plague period, see Life, p. 3. Richard McKinley has traced the growing stability of surnames during the thirteenth and fourteenth centuries in Norfolk and Suffolk Surnames in the Middle Ages (London, 1975), pp. 3–30Google Scholar.

²⁵ The Reeve-Sprot example is hypothetical, but typical of most cases. For examples of the care that must go into tracing these aliases, see Razi, , Life, pp. 11–24Google Scholar. It is worth noting that instances of one individual using two forenames have never been reported; the problem of aliases applies solely to surnames.

²⁶ In the roll of Iver courts held between 1332 and 1376, only 45 male forenames (including 17 used only once or twice) and 38 female forenames (including 16 used once or twice) were employed in 13,761 citations.

²⁷ In some cases, these additional modifiers will not suffice for clear historical reconstruction. The Iver roll, for example, boasts 10 different methods of citing members of the large Norgent surname group with the forename William (William Norgent without additional modifier/senior/junior/son of Richard Norgent/son of Christina Norgent/shepherd/de Northwode/son of Lawrence Norgent de Northwode/de Thorney/son of William Norgent de Thorney). Although several of these different citations can be linked because of additional data, complete reconstructions cannot be confidently completed. Fortunately, such complications are rare, but they do illustrate the unhappy fact that full use of court roll data is impossible. Even the most careful and complete methods will yield name data for which confident linkages cannot be made.

²⁸ In the pre-plague courts of Brigstock, for example, males averaged over 20 appearances each, and some individuals accumulated enormously large files of court appearances. Henry Kroyl senior merited 210 court citations; his son Henry Kroyl junior tallied 586 court entries. Bennett, , “Gender,” pp. 137 and 194Google Scholar.

²⁹ Buckinghamshire Archaeological Society, 128/53, folio 47 (court for 6/6/1349) and folio 78 (court for 6/9/1360).

³⁰ This process of checking all surname variations will require the use of the usual research tools (Medieval Latin dictionaries, Middle English dictionaries, the appropriate volumes of the English Placename Society). Other useful references are: Withycombe, E. G., The Oxford Dictionary of English Christian Names (London, 1947)Google Scholar; Reaney, P. H., A Dictionary of British Surnames (London, 1976)Google Scholar; Bardsley, C. W., A Dictionary of English and Welsh Surnames (London, 1901)Google Scholar; Ekwall, Eilert, The Concise Dictionary of English Placenames (Oxford, 1940)Google Scholar; Lofvenberg, M. T., Studies in Middle English Local Surnames (Lund, 1942)Google Scholar; Thuresson, B., Middle English Occupational Terms (Lund, 1950)Google Scholar. For the development of methodologies that use surname evidence for demographic analyses, see McClure, Peter, “Patterns of Migration in the Late Middle Ages; The Evidence of English Place-Name Surnames,” Economic History Review, second series, 33 (1979), 167–82CrossRefGoogle Scholar, and Robert McKinley, Norfolk and Suffolk Surnames.

³¹ During the five decades (1332-1376) covered by the Iver roll, 127 of the 299 frequently-used surnames (42.5 per cent) were preceded by a preposition at least once. But only 21 of these surnames continued to be preceded by prepositions into the 1370s. It is interesting to note that these 21 surnames occurred frequently in the rolls, indicating that neither familiarity nor repeated use caused abandonment of prepositions. Rather, it appears that prepositions continued to be used for those surnames whose origins were still recent and immediate. For each of the three placename surnames that still employed “de” into the 1370s, the original individual from the locality identified by the surname was still living at the time.

³² Some researchers might object to wholesale elimination of surnames for which some linkages are known. For example, some scholars might not want to eliminate all data under the Reeve surname because some persons so identified were explicitly related. It should be remembered that surnames like Reeve are inherently untrustworthy as guides to families—even if some relationships are known. Therefore, the surname cannot be elevated to the “presumptive family” category discussed below. Data from such surnames can, of course, be tagged for later examination of specific known relationships. In other words, the surname group as a whole cannot be used, but specific known relationships can be pulled for later analyses.

³³ Hence, in the case of Iver, all three placename surnames preceded by “de” into the 1370s were sufficiently stable (and from sufficiently distant places, rather than adjacent localities) to merit treatment as distinct surnames.

³⁴ Because manorial records provide valuable data on many aspects of medieval rural life, medievalists must accept a reasonable method that will allow scholars to integrate familial factors into their social, economic, and legal analyses without actually undertaking the arduous (and tentative) process of family reconstitution.

³⁵ Smith, Richard M., “Kin and Neighbors in a Thirteenth-Century Suffolk Community,” Journal of Family History, 4 (1979), 252–54CrossRefGoogle Scholar.

³⁶ The necessity of reevaluating personal histories in the light of family reconstitutions is discussed by Razi, , Life, pp. 21–25Google Scholar. We must, of course, consider researchers who use only presumptive families in their analyses and consequently miss this opportunity to reassess personal histories. Such scholars should take care to discern the need for closer evaluation of individual files at the level of personal identification, and should partially reconstruct the families of all questionable cases. For example, names passed on through generations of a family will usually require partial family reconstitution to establish individual identifications.

³⁷ See, for example, the genealogies presented by Razi, , Life, p. 18 and p. 20Google Scholar. Almost no cousins or affinal relatives are traced in these tables.

³⁸ Both Razi (“A Critical View”) and Wrightson have criticized the studies of Raftis, Britton, and Edwin DeWindt for their rather cavalier socio-economic groupings based on official activities. They have argued that these studies place too much socio-economic importance on official duties and ignore change over time (because a surname that appeared for many decades could be highly ranked based on the official activities of only one person). The first criticism is overdrawn; official activity clearly constitutes an easily retrievable and relatively accurate index of socio-economic status. Anne DeWindt's careful study of a wide variety of factors contributing to socio-economic status (a study that merited praise by Razi, p. 149) proves that officeholding was closely correlated to success in other areas (as judged by landholdings, number of animals, pledging activities, etc.). See DeWindt, Anne, “Peasant Power Structures in Fourteenth-Century King's Ripton,” Mediaeval Studies, 38 (1976), pp. 236–67CrossRefGoogle Scholar. Other studies confirm the general accuracy of officeholding patterns as guides to overall status, see Britton, , Community, pp. 70–93Google Scholar. The second problem is more serious and can best be resolved by employing two levels of categorization; the first level assigning a rank to a particular surname (or presumptive family) and the second, more precise level assigning an individual rank based on either the individual's official activities or (in the case of females) the activities of fathers and husbands. These matters are discussed more fully in Bennett, “Gender,” in which new categories (to measure brewing activity and longevity) are also employed.

³⁹ For example, Razi, (Life, pp. 24–26)Google Scholar estimates the population trend in Halesowen between 1271 and 1395 based upon a “census-like enumeration of the villagers mentioned in the court records.” But Razi fails (1) to consider that changing court procedures over this century could have resulted in different recording practices and (2) to prove that different numbers of extant courts for different periods do not skew these trends. At best, such indirect estimates are only tenuous; see Bennett, “Gender,” pp. 50-54 for estimates that do include adjustments for source differentials.

⁴⁰ Razi, , Life, p. 92Google Scholar.

⁴¹ This phenomenon is common to all manorial records. For examples, see Bennett, , “Gender,” pp. 194, 276, and 331Google Scholar.

⁴² Britton, Edward, “The Peasant Family in Fourteenth-Century England,” Peasant Studies Newsletter, 5 (1976), 2–7Google Scholar. Britton6's discussion is based on the proper assumptions, but it must be noted that his results might have been skewed by inaccurate preliminary analyses.

⁴³ For an example of the usefulness of such microcosmic analyses, see Judith M. Bennett, “The Tie That Binds: Peasant Marriages and Peasant Families in Late Medieval England,” forthcoming in the Journal of Interdisciplinary History. This essay analyzes the social ramifications of a single marriage in a fourteenth-century village.