¹ See Laslett, and Wall, , Household and Family in Past Time (Cambridge University Press 1972).CrossRefGoogle Scholar

² This is brought out in Jack Goody's contribution to the volume already named, under the title ‘The Evolution of the Family’.

³ This is particularly so for research units, which, like the Cambridge Group for the History of Population and Social Structure, have responsibility for collecting records of this kind. Principles for the analysis of evidence from listings of inhabitants, adapted primarily for handworking, were originally set out in Laslett's, Peter, ‘The Study of Social Structure from Lists of Inhabitants’, in Wrigley, E. A., ed., An Introduction to English Historical Demography (London: Weidenfeld and Nicolson, 1965). See especially p. 183 for the conventions of transcription, and for the prompting of a document from ancillary records such as parish registers.Google Scholar

⁴ See Barnes, J. A., ‘Genealogies’ in Epstein, A. L., ed., The Craft of Social Anthropology (London: Tavistock Publications, 1967).Google Scholar

⁵ It must be obvious that few researchers would go to great length simply to make certain of one particular relationship, and we have found in practice that explicatory diagrams are only used when the structure of the domestic group as a whole is complex and obscure. The explication is often a piece of inference, or even guess work. Nevertheless the principle of complete ideographic translatability requires in strictness that wherever a word has to appear in the domestic group diagram, its meaning should be made clear in an accompanying explicatory diagram. Such a rule has the further merit of forcing the researcher to make up his mind on every relationship appearing in his data.

⁶ See Hombert, Marcel and Préaux, Claire, ‘Recherches sur le recensement dans l'Egypte romaine’, Papyrologica Lugdona-Batava, V (Leyden, 1952).Google Scholar

⁷ Or, as well, that of ‘dependent’, a complication we shall ignore.

⁸ The distinguished French demographer Louis Henry solves this problem by counting each widowed head coming at the beginning of the name block not only as the head but also as a noyau (a CFU) in himself or herself. See his Manuel de démographic historique (Paris, 1967), p. 44.Google Scholar This solution is attractve in many ways, but we cannot adopt it for our purposes. It is not consistent with the principles that at least two persons involved in or arising from a conjugal link be present to form a CFU and that the conjugal link be taken as the point of reference. It lays even greater stress on primary position in a listing as an indicator of headship. The consequence of the different definitions is this: Henry's system makes all households where widowed persons are named first in the list into ménages multiples (multiple family households), where we would count them as extended family households, except when the widowed person had an accompanying unmarried child. If a widowed person comes later in the block, no noyau is reckoned. This is inconsistent with the principle of the conjugal link as the point of reference, and makes households with initial widowed persons into multiple households, besides laying so much stress on coming first as an indication of headship.

⁹ Frérèches are composed of two or more married siblings of either sex with accompanying spouses with or without the presence of unmarried siblings. If only one of the siblings is married, the household is classified as 4c, and if none, as 2a (see subtable to Table 1).

¹⁰ Where the data contain ages, Table 1 can be supplemented by a further table headed ‘Composition of Households by Age of Head’, bringing out differences over the life cycle.

¹¹ See section 1.3 of Laslett’s introductory chapter to Household and Family in Past Time.

¹² For the astonishing wealth of such materials for Tokugawa Japan, see especially Akira Hayami in Chapter 18 of the volume already cited. The Italian materials are only just coming into light, though in 1971 a research group at Florence at work on Italian demographic history made it clear that very many Italian parishes, especially apparently in the Florence area, had continuous series of annual or periodical nominal registrations. These are in the form of libri status animarum, and for the single parish of Fiesole near Florence, such lists exist for the years 1638, 1683, 1707–36, 1746 and 1751–1906. See the article by Schifini d'Andrea in Population, 26, No. 3. (1971), pp. 573–80. No series of this kind has yet been discovered for England, but there are two for Northern France in the later eighteenth century which are described in Laslett's article in Annales de Démographie Historique for 1968.CrossRefGoogle ScholarPubMed

¹³ See Hammel's paper, ‘The Zadruga as Process’ in the volume edited by Laslett and Wall, cited earlier.

¹⁴ When the requirements of specificity in the original source, or of prior semantic analysis of the terminology have been met, and if non-genealogical relationships do not pose a problem, even more useful analytical languages of notation can be employed. Although originally developed in analysis of household structure, such notations have been used most frequently in formal semantic analysis of kinship terminologies. See Hammel, E. A., ed., ‘Formal Semantic Analysis’, American Anthropologist, Vol. 67, No. 5, Part 2 (1965),Google Scholar and Romney, A. K. and D'Andrade, R. G., ‘Cognitive Aspects of English Kin Terms’, American Anthropologist, Vol. 66, No. 3, Part 2 (1964), pp. 146–70.CrossRefGoogle Scholar