Chapter 2

THE NATURE OF THE DATA

2:1 Historical records in general.

In the first chapter we reviewed some of the approaches to the analysis of communities. In order to make the discussion more concrete we will now turn to the specific discipline of history and, within that, to a small part of the data, namely that emerging from English historical communities. This is an illustration of a methodology which, it is believed, will apply to other bodies of data of interest to investigators in other disciplines. It was stressed earlier that studies of communities in the past were only one of many types of community study, but that methods developed in one area of 'community studies' are of interest to all those who study bounded units of various kinds.

Massive deposits of records relating to 'communities' in the past exist for most continents. Findings arising from the growing interest in social history, historical demography and economic history, suggest that such records are even more plentiful and extend over longer periods of time than one might, even ten years ago, have imagined. There is no single reference work which describes-the local records for civilizations throughout the world, though there is a useful bibliography in Hollingsworth (1969). This is not the place to remedy this omission, nor are we qualified to do so. Yet a short and necessarily impressionistic account of how English records compare with those for other countries is needed. The criteria by which we will judge the records are as follows. We will be considering the records of a country, or larger area, as a whole, and not the often exceptionally rich deposits relating to a particular family or social group. We will primarily be interested in the records concerning ordinary people, not those for the wealthy and powerful which often survive even where there is no other documentation. We will only consider sets of documents which survive for a hundred years or more at a stretch, not the occasional exceptional survey or tax assessment or inquisition. We will only be describing cases where at least three different types of record, which in England might mean manor courts, parish registers and wills, bear on a particular geographical or social area. The method we will describe requires the overlap of different sources. Finally, we will only consider the situation before the middle of the nineteenth century. It should be stressed that this is a preliminary sketch which omits many areas through ignorance and probably oversimplifies the position elsewhere.

By the criteria outlined above, England as a whole starts to be historically visible from the end of the thirteenth century and continues so until the present. The Celtic fringe only begins to be documented in detail from the second half of the seventeenth century. Within Western Europe as a whole, it would appear that a number of areas are suitable for the type of analysis we envisage. The Italian records are particularly impressive for the fourteenth and fifteenth centuries (Herlihy in Wrigley 1973: 47; Klapisch in Laslett 1972: ch. 10), and those for Spain may turn out to be very good from the same or slightly later period (Linz in Lorwin & Price 1972: ch. 5). The records for France have been extensively studied, particularly the excellent parish registers of the seventeenth century and onwards (Goubert 1960; Goubert 1972). The German records, though they seem to start a little later, are immensely detailed in some areas from the seventeenth century (Berkner 1972; Knodel 1970). Similarly, in Sweden the records start a good deal later than those for England but are better when they do begin (Schofield 1974). The records for the Low Countries and Holland are also likely to prove good (Van der Woude in Laslett 1972: ch. 12).

The East European countries have a varied set of records; in Hungary, Serbia, Yugoslavia and elsewhere there are often deposits that go back at least to the seventeenth century (Hammel in Laslett 1972: ch. 14), but those for Russia do not appear to start in comparable detail until the nineteenth century (Kahan in Lorwin & Price 1972: 361). The North American records do not begin until the seventeenth century, but from then are equal in variability to those described below, while the records for certain parts of Canada are superior to English records (Legare et al. 1972). South American material in the early period mainly consists of records made by colonial governments, but may turn out to be good (TePaske in Lorwin & Price 1972: ch. 10). Likewise, in Africa it is now becoming clear that there are more historical records than earlier writers-imagined, though it would be difficult to argue that they fulfil the criteria listed above before the middle of the nineteenth century.

A major civilization which has produced a vast number of historical records is India. It is clear that many classes of local records exist for parts of India from the nineteenth century onwards, and these are beginning to be exploited by historians and social anthropologists (David Morris in Lorwin & Price 1972: ch. 13; Kessinger 1970; Kessinger 1974; Kessinger 1976). The Singhalese records are also detailed from at least the start of the nineteenth century (Leach 1961; Obeyesekere 1967).

We end this superficial survey by looking at two particularly intriguing cases. So far, for duration and multiplicity of sources, it has been difficult to find any area which matches England. The two main competitors may be China and Japan. At present, very little is known concerning the historical records of Mainland China, yet vast deposits of genealogies, gazetteers, diaries and other accounts did once exist and covered about a thousand years. Many must have been destroyed in the various upheavals and the present political climate in China makes it impossible to be sure, but China may prove to be the most richly recorded of all civilizations. The other possibility is Japan where the long feudal period from the twelfth to nineteenth century has led to the amassing of large numbers of records of all kinds from censuses, genealogies and 'religious investigation registers' to tax surveys and diaries. These remarkable collections are widely dispersed and hence difficult to use, but their potential for demographic and social history is now widely recognized (Yanamura & Hanley in Lorwin & Price 1972: ch. 13; Hayami & Uchida in Laslett 1972: ch. 18; Hall 1958; Hanley 1974).

It is particularly difficult to undertake a survey of the existence of records throughout the world since it is frequently the case that the documents are not known to have survived until interest is shown in the possibility of their existence. In general it is true that the same combination of circumstances that lead to an archival revolution in England and Western Europe is affecting societies throughout the world. The result of these changes is that it has become clear that archives exist to be analysed for some parts of the world from before the fifteenth century, that in most of Western Europe and North America such records exist from the seventeenth century and for India and Russia from the nineteenth century. The number of records is beyond computation. A general method for extracting the maximum amount of information from them. is needed. Thus, while this book is particularly concerned with English records for the fifteenth to eighteenth century, it is believed that comparable analyses could be carried out in many other parts of the world.

The records of English communities start effectively in the thirteenth century. It is from that century 'that the first major source which bears on particular communities year by year, namely manorial records, commences (Elton 1969: 129-30). Before that date, if we are interested in the local community, it is impossible to catch more than unconnected glimpses of individuals. For three centuries from the thirteenth the records produced by the manorial administration provide almost all the information we have for activity at the local level. The sixteenth century witnessed an explosion of documentation. From the middle of that century it is possible to combine many different sources in such a way that we are able to gain a three dimensional picture of particular individuals and particular communities. Specific classes of document emerge and then disappear and records vary enormously from county to county and even village to village, yet the following assertions seem plausible. Firstly, very extensive records exist for many parishes in England from the mid sixteenth century up to the present. Secondly, in some respects, the records are better for the late sixteenth and seventeenth centuries than for any subsequent period, including the recent past. Thirdly, although the records for a parish during the nineteenth century appear different from those for the same community some two centuries before, they are similar in intention and in structure. For example, both sets deal with land ownership, demographic events, disputes, administration of the church and the poor rate. Fourthly, although the survival of records varies from village to village and region to region, the problems and techniques of analysis are uniform. Although the legal differences between different areas need to be understood, whether between the see of York and Canterbury in testamentary dispositions, or between manor and manor in the transfer of customary land, needs to be understood, the records are basically the same in structure from region to region.

If the largely undocumented assertions stated above are true, it would seem fair to sample from the immense records of a nation which, even at the end of the thirteenth century, comprised several million inhabitants. The following discussion will largely be based on the records for only two English counties, Essex and Cumbria. Within those counties only the records for two parishes are examined in detail. Within those parishes only the period 1500-1750 is covered in any depth. Preliminary comparisons with earlier and later periods and with studies made of other villages in other counties, suggest that the techniques are applicable to eighteenth or nineteenth century material and to any other English county where records survive.

2:2 Problems in delimiting the study.

Boundaries are necessary but always, to some extent, arbitrary. The analysis of a spatially delimited area is, as we have seen, subject to many criticisms. Yet there is one consideration which explains why so many disciplines have used the 'community study' approach. This is the belief that one should attempt to study the 'totality' of human life, the interconnections and complexities which seem to emerge most fully within a small geographical area. Yet even the most ardent devotees of this method will be aware that it is only a tool; the approach contributes some answers to some questions. Without such a method many of the major problems asked by social scientists cannot be solved. Yet such an approach is unlikely, on its own, to contribute a full answer to any one question. The most important findings are always presented by subject, for example studies of myth, magic, capitalism, slavery, the family, or numerous other general themes. One layer of evidence, however, will come from community studies. Without this dimension, the infinite complexity of evidence from a demarcated area, we are left with an impoverished picture of man. Fully aware of the dangers and limitations of this method, yet resolved to delimit the use of the voluminous records to a specific area and a specific period, we face the problem of choosing the sample.

A demographer or sociologist might try to choose a 'random sample' of the population using the appropriate statistical techniques. Other disciplines, with defective information at their disposal, are forced to select on the basis of less scientific criteria. Social anthropologists, for example, tend to let practical considerations of distance and terrain weigh heavily while historians searching for a community to study intensively are likely to be influenced by the nature of the surviving records as well as more idiosyncratic criteria such as the distance from their place of residence. In each of these disciplines, as we saw in the last chapter, it has been found convenient for any 'total' study to select an area that holds not more than 2,000 individuals at any one time. The obvious unit for an English or French historian is the parish or group of parishes. This was the unit of ecclesiastical and civil administration for many purposes and hence many of the records with which the investigator will be dealing come to him organized on the parish basis. Manorial jurisdictions often cut across the parish, but can usually be subsumed within a study based on parish boundaries. It will often be necessary to move outside the particular parish in order to follow individuals through their lives or to study a group of parishes in order to obtain a large enough sample. Yet the core of the study will probably be this unit.

The degree of usefulness of a community study will often depend very considerably on the original choice of area. It is therefore worth stating some rather obvious yet basic points concerning such a choice. There are many criteria which affect a decision as to the size and location of the community to be studied. Among these are the type of problem to be studied, the availability of time and money, the availability of documents. There are now some guidelines as to the type of community one should select when undertaking historical demography and these seem a reasonable start for the more general approach suggested here (Wrigley 1966: 104-6). A parish of not less than 800 persons and not more than 1,500 is ideal for many purposes. Anything smaller than this makes it difficult to assemble meaningful statistics, anything larger becomes unmanageable unless one is interested in only one problem or a limited source. In order to obtain some idea of change a period of at least a hundred years needs to be studied, even if most of one's attention is concentrated on a shorter period.

It is clearly necessary to choose a community with as complete a series of major records as possible and these records need to be appropriate for the investigator's particular interests. If detailed work on household budgeting or agricultural economics is to be attempted, there must be probate inventories; if crime in the local area is of central interest, the Quarter Sessions and Assize records must survive for the period under investigation. Although it is a considerable oversimplification, it could be argued that for any study of a local area in England in the period up to 1841 an area should be chosen which has at least six out of eight of the following major sources for a substantial proportion of the period under study: parish registers, wills, inventories, manorial or other surveys, manor court rolls, ecclesiastical court records, Quarter Sessions Records, deeds. If one were particularly interested in the eighteenth or nineteenth centuries the checklist would have to be modified.

If the two criteria of size and survival of critical documents are applied, the prospective investigator will have his choice narrowed down to certain English counties and certain parishes within these counties. Less than one quarter of the total of parishes will be left to choose from. The final choice of a particular community can be made on the basis of finding what may be termed a special source. For example, the parish of Earls Colne in Essex whose records will be used to illustrate this volume was chosen because one of the inhabitants kept a most detailed diary during the seventeenth century. As well as this unique source there is an unusually good Elizabethan map. Such 'special sources' are not as rare as one might imagine. There are quite large numbers of parishes with a particularly good set of court rolls, a listing of inhabitants before the nineteenth century, an early map. It is clearly far better to choose a well documented parish or set of parishes, even though they cannot be proved to be typical, than to choose a 'random' but poorly documented and uninteresting place.

2:3 The convergence of records.

As Marc Bloch noted, 'the deeper the research, the more the light of the evidence must converge from sources of many different kinds' (1954: 67). The basic aim of the approach adopted here is to gather together and analyse all the records which relate to a certain set of individuals, in the past. One strategy is to delimit this set to those who happened to live at some time in a specific area. In order to do this the records of nearby areas will also have to be searched. The very great collections of records generated at the national level, which are deposited in the central archives and bear on the selected individuals, will need to be examined. The approach dictates that every single document concerning the selected area and individuals must be transcribed or copied in full. It will be shown that the effect of bringing in further records is not merely additive; each extra record illuminates all the previously assembled ones. To omit or abbreviate records because, at one point of time, they appear to have no great significance is the counsel of practicality. In the long term, however, it is folly. Time and man's negligence have selectively destroyed very large parts of the past; other parts were never committed to paper. What remains is infinitely precious and each tiny shred needs to be carefully used.

The reason why one should wish to use all the records has been alluded to on various occasions. We are interested in human beings and their activities in as many contexts as possible. We shall see that the data about the English past is of such a quality that it is possible to build up profiles of specific individuals which are, in many respects, as full as those which we could construct for living individuals. The essence of the approach is the necessity that several different records bear on a particular individual at different points in his or her life. This concept will be familiar to those who have studied the 'family reconstitution' technique which links births, marriages and deaths in order to build up demographic profiles of specific individuals. Extended to incorporate the tens, and often hundreds of other references to individuals, this approach is the basis of the method outlined below. An attempt is being made to practice 'individual reconstitution', using every record which survives for selected individuals. The intellectual rewards already gained by 'family reconstitution' are well known, yet they are only a fraction of the rewards to be obtained from 'total reconstitution'. But just as the returns are greater with the total approach, so the investment in time and labour is far greater.

2:4 An inventory of some major records.

[Figure 2:1. A selection of major sources for the study of a community.]

France is justly noted for its great regional histories and for intensive demographic studies of particular parishes. England on the other hand appears to have one of the most thriving traditions of local history. There are now innumerable 'village studies' of particular communities and many of these illustrate clearly what documents are available and some of the uses to which they may be put. Among the best older style parish histories are those by Cowper (1899), Horsfall Turner (1893), Lucas (1931), Millican (1937), Pearson (1930), Rushton and Witney (1934), examples of more recent studies are Hey (1974), Hoskins (1957), Parker (1976), Spufford (1974). This thriving branch of history has, since the last war, produced a growing number of guides to documents. These show, frequently with extracts from records, what the sources consist of, where they are to be found, and how they may be used. The best bibliographical guide to the sources for local history is Stephens (1973), which may be supplemented by the earlier guides by Hoskins (1959, 1970). Those who are interested in carrying out a parish study will naturally refer to these and other guides. None of these provide a simple checklist of the major records one might expect to find relating to an English parish in the past. It would therefore seem helpful to produce such a list, fully recognizing that it cannot be exhaustive and that no particular parish will have all these records: see figure 2:1. The other purpose of producing this brief inventory is to give some idea of the wealth of historical records relating to English communities in the past. In the following sections of this chapter we will merely be selecting twelve major types of document out of this inventory. It is only fair to point out, however, that in terms of sheer volume, these twelve are likely to contain over two-thirds of the material relating to any given parish over the period 1500-1841.

2:5 Twelve specific classes of data.

[Figure 2:2 Twelve sample records.]

The records we shall use to illustrate our approach are taken from two English parishes during the period from the sixteenth to nineteenth centuries. One is Earls Colne, near Colchester in Essex with an approximate population of 1,20O in the middle of the period, the other is Kirkby Lonsdale in Cumbria with a late seventeenth-century population of about 2,500. The specific records chosen are believed to be representative of classes of data which can be found in many different societies at different points in time. Since this may at first slight seem difficult to believe, it is worth suggesting, in a very concentrated and oversimplified way, some of the major disciplines which have conventionally been interested in such classes of record and often produced their own data of these types. The main features can be illustrated most simply by a figure: see figure 2:2.

It will be obvious that the allocation of disciplinary interests is, to a large extent, arbitrary and overlapping. Most of the disciplines , and others not mentioned, are interested in most of the types of information listed above. The table is useful, however, since it shows that, although the particular records are drawn from the past and have been traditionally the preserve of local historians, they are likely to be of much wider interest. This can be illustrated in greater detail if we take each in turn.

2:6 Parish register.

[Figure 2:3. An Essex parish register.]

Vital registration information for communities in the past is contained in many classes of record, among them registers of baptisms, marriages and burials of the Anglican Church. In theory, these registers commence for all English parishes in 1538 and are replaced by Civil Registration in 1837. In practice, many parishes in England do not have registers for the first few years, but a large number commence some time in the sixteenth century. These registers have been very extensively used during the last fifteen years, especially by those interested in historical demography who wish to work out sophisticated measures of fertility, nuptiality and mortality. The conventions concerning how to use such documents are well described and their value is widely appreciated. The methodology is based on the work of Henry, Fleury, Goubert and other French demographers and is best described for English readers by Eversley and Wrigley (Wrigley 1966: chs. 3, 4)-. One example of their format, both in the original and after direct transcription, may be taken from the seventeenth-century register of Earls Colne in Essex: see figure 2:3. If we combine the burial register with the other registers of baptisms and marriages, it is possible to work out very precise mortality rates, specifying age, sex, marital status and other variables. Earls Colne with its population of about 1,200 at the time the document above was produced, would witness an average of between 20 and 25 burials a year. Marriage and baptism registers also exist and each entry in them contains an average of 2 names, for example a child and a parent or a man and his bride. We would expect to find about 100 names recorded each year, on average, in this parish register. These registers exist, with some small gaps, from 1558. If we continue until the mid nineteenth century (using Civil Registers for the last few years), and count each occurrence of each name as one, we will have accumulated roughly 30,000 very brief references to individuals. To extract each entry and then link it up with other entries relating to the same person (family reconstitution), would take approximately 1,500 man-hours, or one year working 30 hours a week, 50 weeks a year. Wrigley states that, very roughly, to reconstitute a parish of 1,000 persons for three centuries might take 1,500 hours (Wrigley 1966: 97). Parish registers are only one of many sources bearing on mortality; for example, wills and manorial records often give explicit or implicit information about people's deaths. Bringing in other records also enriches mortality statistics very greatly by providing the social and economic context within which deaths occur.

2:7 Manorial rental.

[Figure 2:4. A manorial record.]

There is massive documentation of property holding in England from at least thirteenth century onwards The most voluminous records of ownership are those produced by manorial officials. The two major classes of manorial records are the periodic surveys and rentals at specific points in time, and the continuing registration of transfers of land and other property in the manor court rolls. Rentals and surveys exist in large quantities from the fourteenth to the nineteenth centuries and are sometimes accompanied by maps. These descriptions of property ownership often enable one to piece together landholding patterns in particular parishes every 20 or 30 years. For example, during the period 1395-1677 there are 14 surviving rentals of one kind or another for the manor of Earls Colne, or an average of one every 20 years. Their accessibility and detailed nature has meant that they have been frequently used and described by historians. An extract from one such rental, made in 1638, will give some indication of their nature: see figure 2:4. This is a moderately informative example of the class; some of the others are much more detailed, giving the names of previous owners, the measurements of the property, the names of adjoining properties and their owners. Roughly speaking, each rental contains descriptions of some 300 pieces of immovable property. Thus there would be roughly 5,000 small descriptions in the rentals for Earls Colne up to 1677. Rentals and surveys continue in abundance until the end of the nineteenth century, so that a total of 7,000 such individual descriptions for the whole period would not seem excessive. Earls Colne manor covered slightly over half the parish of Earls Colne, but only represents approximately half the total documentation on landholdings since the other manor, Colne Priory, held considerable lands outside the parish. If we combine the rentals, even allowing for slightly worse documentation for Colne Priory, a rough estimate of 12,000 descriptions up to the end of the nineteenth century would seem likely.

Given the very large number of-detailed descriptions, it is not surprising that even those who are most impressed by their value should not have been able to make full use of this class of document. (Spufford 1974: 126n). Yet England is less well served in this respect than parts of France with its superb series of land surveys. In terms of the information they contain, each record is probably equivalent to two or three parish register entries. Judged solely by quantity, therefore, this class is roughly equivalent to the parish registers. Obviously the kind of information is very different from that in the Anglican registers. Furthermore, it has to be treated with extreme caution. The rentals cannot be trusted as an index of overall property holding for many people held land in a number of manors. There was also a great deal of subletting which is frequently invisible in rentals. Furthermore, the rentals normally record owners rather than occupiers and it is extremely difficult to be sure where a person is actually living. Despite these difficulties these periodic rentals and maps do provide the basis for geographical and economic analysis of communities in the past, as well as providing a great deal of incidental information on social relationships.

2:8 Manor court transfer.

[Figure 2:5 A manor court transfer and transcript]

The second major property record is the manor court roll in which the transfers of certain classes of property were written down. These documents have been extensively used by English medieval historians, partly because they are one of the few sources for local studies in that period. They are thought to have declined in quality during the sixteenth century and for this and other reasons have been little used by historians of later periods. Yet in many parts of England they continue to register land and other property transfers with almost unabated vigour into the eighteenth century. Their uses and nature have been described in a number of studies (Elton 1969: 128-34; Stephens 1973: 42) One example of a property transfer from this most voluminous of community records may be given: see figure 2:5. In our sample parish of Earls Colne, manor court rolls exist from 1400 to 1931 for the manor of Earls Colne,

and for the period from 1489 to 1931 for the manor of Colne Priory. Allowing for some gaps caused by loss of documents and counting the manors separately, there are rolls for about seven hundred years. There are an average of about 10 transfers a year in the Earls Colne rolls and about 5 a year in those for Colne Priory. A rough estimate would be that between 5,000 and 7,000 transfers are recorded for this parish in these rolls. An average of between four and six persons are referred to in each transfer so the total of names alone is roughly similar to the 30,000 estimated for the parish register. But it will be seen that the incidental information about relationships and about property is very much more complex than that in parish registers. The sheer bulk of information makes them almost impossible to use. Maitland, one of the greatest authorities on this source, recognized the difficulty of using them while also emphasizing their very great value. He write, 'How best to garner the great mass of information contained in the manorial rolls so as to render it available for students of legal history is a great question ... A few sets of rolls completely printed beginning in the thirteenth and ending in, let us say, the sixteenth century, would be of inestimable value, especially if they began with surveys or "extents" and ended with maps' (1889: xi).

Manor court rolls not only give far more detailed information about the ownership of land and houses than any, other source, they also provide invaluable demographic and social data. They often fill in family relationships, give death dates, highlight neighbourly and other bonds. If a method could be developed to use them-in the ways which Maitland envisaged they should be used, a new dimension could be added to economic and social history. Yet, as can be seen from the illustration, they are complicated and sometimes ambiguous. To understand them and use them requires considerable knowledge of the legal background within which they were created.

2:9 View of frankpledge.

[Figure 2:6. A view of frankpledge and translation]

Property registration of the kind illustrated above was the concern of only one part of the manorial jurisdiction, the 'court baron' as it was known. In theory, the law enforcement side of the manorial administration was entirely separate and was known as the 'court leet' with 'view of frankpledge'. It was an added prerogative which might or might not have been granted by the king to a particular tenant in chief. For this reason, not all villages which have manors have court leet records. Even those that do have such records do not always have extensive lists of those who were sworn into 'frankpledge' or 'views of frankpledge'. Where such views were held they can be an invaluable source for the social and demographic investigator (Stephens' 1973: 42-3). In theory, they list all the owners and inhabitants over a certain age who lived within the jurisdiction of the leet. If people made excuses for not appearing ('essoin') or defaulted, in other words did not appear, their names were also recorded. A view for one of the Earls Colne manors will illustrate their nature: see figure 2:6. It is clear from this example that although the document does not provide a great deal of information about individuals, very large numbers of names appear. Once we know exactly who did and did not appear, this class of record could provide important information in our attempt to build up a picture of individuals in the past. Since they are so detailed, it is perhaps surprising that they never seem to have been used very extensively by historians. Once again, it is probably the practical problem of size which is largely responsible. The views were supposed to be held and recorded twice a year. In practice, in Earls Colne, they seem to have been recorded about once a year. In the manor of Earls Colne there are such views over most of the period 1400-1700 and in Colne Priory manor sporadically between 1490 and about 1700. When a view was held, between 50 and 100 names were usually recorded. Allowing for the loss of records, this parish produced roughly between 20,000 and 30,000 names in the views of frankpledge up to the end of the seventeenth century. Again, this is comparable to the whole parish register and the problem of linking in these names is clearly formidable.

2:10 Court leet.

[Figure 2:7. A court leet presentment and translation]

The 'view of frankpledge' was part of the general peace-keeping machinery delegated to certain lords. Those appearing had to take an oath that they would keep the peace and abide by the customs of the manor. Those who did not abide by their oath could be presented at the 'court leet', a judicial court which had competence to deal with petty crime. Since the 'court baron' also had jurisdiction in certain not totally dissimilar matters, and the records are written together on the same parchment roll, it is often difficult to separate the cases. The nature of these records can be illustrated by a short set of presentments in the Earls Colne manor court against one individual: see figure 2:7. It will be seen that such cases can provide us with material on interpersonal relations which give flesh to the economic and demographic records which we have previously been describing. It is therefore again somewhat surprising that they have been so little used by historians of the early modern period in England. As with other manorial records, part of the problem seems to be sheer size. In Earls Colne, the court leet remained active until the end of the first quarter of the seventeenth century. On average, between 20 and 30 cases a year were presented in the two courts combined. Allowing for the loss of a good number of original court rolls, it would seem reasonable to estimate that between 3.000 and 4.000 court leet cases of one kind or another survive for this parish, mainly in the period to 1620. Although less in size and complexity than the records previously discussed, they still constitute a large amount of data. Preliminary work suggests that court leet records are available for a good number of English parishes in the period up to the middle of the seventeenth century.

2:11 Ecclesiastical court.

[Figure 2:8. An ecclesiastical court presentment.]

There were numerous other courts with jurisdictions which complemented that of the court leet. Those which were most similar in their detailed supervision of village behaviour were the various church courts. It was believed by many in England until at least the middle of the seventeenth century that the Anglican Church had the right to supervise morals, for example to control sexual, marital and other affairs. The records of the ecclesiastical courts are therefore a valuable source for the study of topics such as social control, the nature of interpersonal tensions, sexual misdemeanours, witchcraft and sorcery. There were many levels of court, but the one with the most extensive records in the county of Essex is that of the Archdeacon. These archdeaconry records, as well as those of the other ecclesiastical courts, have been widely described and we now know a good deal concerning the administrative procedure (Marchant 1969; Owen 1970; Stephens 1973: ch. 8). One example of a presentment relating to a villager from Earls Colne will indicate something of their nature: see figure 2:8. A combination of abbreviated latin, a fairly intricate process and the sheer bulk of the documents, has meant that much less use has been made of this source than one might otherwise have expected.

The case illustrated above is not unrepresentative in the amount of detail it gives or in its format. It is difficult to generalize for the quality and survival of ecclesiastical records varies enormously both temporally and from jurisdiction to jurisdiction in England. The records for the courts which covered the parish of Earls Colne are excellent, particularly during the period 1570-1640. During those years an average of about 20 cases a year were brought to the various ecclesiastical courts relating to inhabitants of Earls Colne. Each 'case' normally led to several appearances as the trial proceeded. For the whole of this seventy year period we are dealing with between 1,200 and 1,400 'cases', or something over three thousand separate records if we treat each appearance as one instance. For the short period when they exist in quantity, these documents are the most intimate record of everyday life and tensions that we possess. If we bear in mind that Earls Colne only constituted a little over 1% of the population of one county, it will be clear that the problem of analysing this source on a regional basis are very great.

2:12 Quarter Sessions.

[Figure 2:9. A quarter sessions presentment.]

One jurisdiction which overlapped with that of the court leet and ecclesiastical courts was that of the Justices of Peace at the Quarter Sessions. The documents concerning the business of the Justices contains information relating to many topics including petty theft, bastardy, maintenance of the poor, highways and bridges. Considerable work has been done on this class of records in the early modern period and their content and nature has been thoroughly described. The records have already been used to study topics such as theft, witchcraft, riot, treatment of the poor (Guide to Essex Record Office 1969: 1-49; Stephens 1973: 55). Since the business could consist of many different types and could be brought to court in a number of ways, it is difficult to select a representative record. One example of a Quarter Sessions presentment for Earls Colne will give some indication of one kind of business: see figure 2:9. The records again vary from county to county both in quality and in their survival. Those for the county of Essex commence very early, in 1556, but in many counties they are not available until a century later. For Essex they continue in large quantities until the middle of the nineteenth century. Although they survive for a longer period than do the ecclesiastical court records, they are less dense for specific years. Thus for Earls Colne during the period 1560-1710 there are, on average, between one and two Earls Colne cases a year surviving in the Quarter Sessions. During this hundred and fifty years, approximately 250 separate 'records' in all survive relating to this one parish. If the period were extended to the middle of the nineteenth century, we would still be dealing with no more than 500 separate references. The depositions, that is statements of witnesses and of the accused, are often extremely detailed and complex. An example of this type of document, which occasionally occurs in ecclesiastical and Quarter Sessions cases, may be examined under the next court.

2:13 Assize.

[Figure 2:10. An assize deposition and transcript]

As well as the local courts, whether of church or county, there were numerous central courts, civil and criminal, with jurisdiction over the villagers of Essex and Cumbria. One of the most important of these were the Assize courts. The records of these courts are now being extensively used by legal historians and there are a number of good guides to both their quantity and the procedure which produced the final documents (Cockburn 1972; Guide 1963). One category of record within the general class may be given, namely a deposition. This illustrates a general type of document which is produced by most courts. The particular deposition is one among a large set relating to coin clipping, burglary and highway robbery in the Kirkby Lonsdale region: see figure 2:10. Much more extensive depositions than this are contained in the records of other central courts such as Star Chamber and Chancery which have not, as yet, been systematically used by investigators of particular parishes.

2:14 Hearth Tax.

[Figure 2:11. A Hearth Tax.]

One of the classes of records most used by historians is that created by the need for revenue on the part of the State. Many studies have been made using medieval and early modern taxation records, Lay Subsidies, Hearth Tax and Land Taxes. The nature of these records is now well known and there are numerous guides to their use (Stephens 1973: 34). We may take as a representative of this general class the seventeenth-century Hearth Taxes, periodically levied during the reign of Charles II. Once again these documents vary from county to county, both in the degree to which they survive and in the way they were assessed. An example of this type of record for Earls Colne forms figure 2:11. In all, the document partially illustrated below, lists 187 names. There are four other surviving Hearth Taxes which cover Earls Colne during the period 1660 -75. Combined with the one above, they give us a little over 700 names. Earlier tax documents for this parish provide a total of about 600 names. By far the most concentrated source are the Land Tax assessments, taken yearly at the end of the eighteenth century and surviving in Essex for most of the period 1780-1830. These record both owner and occupier categories separately, even though the people are fairly frequently the same, this source alone produces over 10,000 names. The degree to which we can estimate wealth structure in a parish, or the facts that we can infer about named individuals from such records requires much more investigation. Yet the records have many values other than as a source for economic history. For example, the Hearth Tax for some areas of England appear to have been taken in house order, so that we can move along with the tax assessor and reconstruct the size and location of houses. Where occupiers are given, they provide a vital check on manorial records, which frequently only give owners.

2:15 Will.

[Figure 2:12. A will and transcript]

Last wills and testaments, in which individuals bequeathed their movable property and sometimes immovable also, were little used by historians until a few years ago. Genealogists, however, were fully aware of their value. There has been a rapidly growing interest in these records and there is now an extensive literature on what they contain, where they are to be found and some of the difficulties in using them. Among the best general introductions are those of Camps and Spufford (Camps 1963; Spufford 1974: ch. 13). Though they varied considerably within a common form, wills were fairly standardized documents in the past. It is therefore not too difficult to give some idea of their content by citing one example from Earls Colne: see figure 2:12.

In some respects wills are the most complex and interesting of all local records since they combine information about property, religion, literacy, interpersonal relationships and a number of other topics. The number of wills and their quality is again very varied both in time and space. They appear to survive for most of England in considerable quantities from the early sixteenth century and continue in large numbers until the later eighteenth century. Some areas of England, for example some of the northern counties, have much higher ratios of wills to population than do others. Essex is not particularly favoured, yet the parish of Earls Colne has produced about 350 wills of persons residing in, or owning property in, the two manors during the period 1500 -1800. On average, roughly 10 people are mentioned in each will, so that this constitutes references to over 3,000 persons. Only future analysis will establish who failed to make wills and in what ways the surviving wills are biased. What is certain is that this class of documents contains a very great deal of demographic, economic, social and other information which blends well with the other sources already described.

2:16 Probate inventory.

[Figure 2:13. A probate inventory and transcript]

The other major probate records, which are often filed with wills, are inventories of possessions. By ecclesiastical law the administrators of estates were required to present an inventory of the movable possessions belonging to a man at his death. These goods were listed and valued by neighbours and the inventory was shown at the ecclesiastical court. This type of record has been widely used, especially by historians of agriculture, and their content and format is well known (Hoskins 1954: ch. 8; Stephens 1973: 36-7, 111). The survival of these inventories varies from county to county. Unfortunately most of the Essex have disappeared, except for small areas or a few selected individuals. Very large numbers of inventories have survived for the Archdeaconry of Richmond, within which our sample parish of Kirkby Lonsdale was situated. Consequently there are over 2,000 such inventories for Kirkby over the period 1550-1750. Each inventory averages between 20 and 30 items of property, so that one is dealing with about 50,000 items in all. On average, each inventory contains between 10 and 15 names, including that of the deceased and 4 witnesses. There are therefore between 20,000 and 30,000 names in all. One example of this source may be given: see figure 2:13.

A great deal of work needs to be done in order to establish what exactly was omitted in inventories, how accurate the valuations were, the bias in inventory making, the degree to which the values changed with the point in the life cycle at which they were made. There can be little doubt as to the historical value of this class, however. They not only provide information about the physical setting, that is the housing, furniture, agriculture, nutrition, prices, but also material and social relationships. The lists of debtors and creditors often appended to inventories are invaluable as an index of interpersonal relationships. Where inventories have survived, however, their volume is once again an obstacle to their proper exploitation.

2:17 Listing.

[Figure 2:14. A listing.]

Formal censuses which list the inhabitants of England exist at ten-year intervals from 1841. Before 1841 there are a considerable number of listings for specific towns and villages, taken for many different purposes. These early listings have been used and described in a number of works (Laslett in Wrigley 1966: ch. 5; Laslett 1972 : ch. 4; Styles 1951-2). Where such early listings exist, they provide unique information concerning who was actually present at a point in time. One example can be given, for the chapelry of Lupton within Kirkby Lonsdale in 1695. The list was made to provide the basis for a new tax on baptisms, marriages and burials, and bachelors over the age of twenty-five: see figure 2:14. This particular listing was taken for all the nine chapelries of Kirkby Lonsdale, and covers the whole population of over 2,000 persons. Some of the other chapelry listings are more detailed, giving name and sex of children, for instance.

2:18 Summary and conclusion.

Anyone who is familiar with historical documents for England will be aware that only a very small proportion of the available records have been selected from the past. Yet even these will show that there is a varied and rich set of material, differing from parish to parish but often extending back for four or five hundred years. The sheer size of the data in a well recorded parish is somewhat overawing. If we guess that the twelve types of records described above represent roughly two thirds of all the material for an 'average' parish, we may make some general estimates of the total amount of data for an area which averaged about 1,200 persons over a period of some four hundred years, between 1450 and 1850. All the records would produce references to about 200,000 separate names to be linked into specific individuals. Between one-sixth and one-seventh of these names would come from the registers of baptisms, marriages and burials. Total reconstitution from all sources would, at this level, take much longer than 'family reconstitution' based on one source. There would also be the added labour of dealing with houses and land. The records contain between 20,000 and 30,000 descriptions of immovable property, similar to those indicated in the manorial rental and transfer illustrated above. To index and link these is a very considerable task. The major source for movable items of property, the probate inventories, are missing for the sample Essex village. A study of the Cumbria parish of Kirkby Lonsdale, to which we have previously alluded, would need to deal with between 40,000 and 60,000 items of property mentioned in this source. Finally, there is the class of presentments for offences in the various courts. The most numerous of these, as we have seen, were the cases in the ecclesiastical courts and courts leet. Between 6,000 and 8,000 such 'cases' probably survive for Earls Colne over the whole period in all types of court. See figure 2:15 for a detailed breakdown of these figures.

The sheer size of the documentation explains why many of those who have attempted to undertake a 'total' study have failed. It explains the great collections of notebooks in which antiquarians painfully tried to transcribe the records for particular communities. Their energies were usually drained by the process, and they had little time to reorder or analyse their findings.

Whereas an anthropologist deals with 1,000 or so persons over one or two years, the investigator who tries to chart a parish's history through time may be dealing with very much larger numbers of persons. The task clearly requires a new methodology. The following chapters are an attempt to lay down some general guidelines. The problem is a fairly obvious one. Given the kind of data which has been presented in this chapter, how are we to analyse the documents so that they can be most efficiently integrated and linked? For those who feel that it is not only an obvious but also an easily soluble problem, the best plan would be to stop at the end of this chapter and to sit down with a pencil to see whether they could devise a way of doing it. They will probably find, as we have done, that the task is by no means trivial. Other readers may still not be convinced that the data is of the quality to make it worth even attempting such an analysis. Chapters 4 and 6 will outline some features of the data, what it contains, its shortcomings, and its value when compared to the material collected by contemporary sociologists and social anthropologists. Chapter 5 provides a very brief survey of some of the questions which may be asked of such data once it has been indexed.

[Figure 2:15. Numbers of records for Earls Colne.]