self-consciously theorized the family as a field of investigation.
Instead they began by adopting the conventional wisdom of
sociology which, going back to Frédéric Le Play, ¹ saw a broad
change in the family from an extended form of the Middle
Ages to a nuclear form of modernity. This position main-
tained that before industrialization the family was composed
of numerous kin living together in cohesive solidarity. Only
the irresistible pressures of modernization could tear apart
these bonds. This tradition of sociological history assumed
that the family was defined by the quantity of kin relatives in
a household.

Over the past several years the thesis of a premodern ex-
tended family was challenged by a group of Cambridge
demographers led by Peter Laslett. Analysing English parish
registers from the sixteenth century on, Laslett discovered no
extended family at all. ² On the contrary, the family was amaz-
ingly stable in size, consisting of the conjugal unit with a small
number of children. There had been a significant demographic
shift from a pattern of high fertility/high mortality in the old
regime to one of low fertility/low mortality in the modern
period. Some historians could claim that the change in demo-
graphic pattern, while not affecting family size, had considera-
ble impact on the daily life of the family. Such factors as the
length of interval between births seriously affects the condi-
tion of women, and high mortality among children has deep
consequences for parents' attitudes toward their progeny. Yet
for Laslett the stability of family size over the past four hun-
dred years was the major finding, one which led him to ques-
tion the family as a suitable object for historical investigation
since it seemed impervious to change.

Laslett's conclusions were criticized effectively on several
counts. Lutz Berkner pointed out that Laslett had done his
counting in a static way, forgetting that the family has a life
cycle. ³ If families of the old regime are studied over the course
of time, it becomes clear that in a significant number of cases
in the regions studied by Berkner the grandparents live with
the conjugal unit. Hence the family begins to resemble the

-x-