|
The History of the Database
by Josef Ehmer
In
At that time most
historians in German speaking areas were not aware that in 1972 quantitative
research in family history and especially the statistical analysis of census
records was already underway on an international level. In this very year Peter
Laslett and Richard Wall published the volume
'Household and Family in Past Time' (Cambridge 1972), a work which was to exert
a profound influence on further research involving substantive issues,
conceptualizations and methods. Laslett and his
colleagues based their analysis on a collection of English census records. From
1964 they had collected these listings in the 'Cambridge Group for the History of
Population and Social Structure'. Their volume was an attempt to use this
type of source as a basis for broadly based international and historical
comparison. Census records seemed to be an excellent tool for 'statistical
comparisons of household size and composition between cultures and across
centuries', as John Hajnal put it a decade later
(John Hajnal, Two Kinds of Preindustrial
Household Formation Systems, in: Richard Wall (ed.), Family Forms in Historic
Europe, Cambridge 1983, 99). It was on the basis of such apparently hard facts
that the debate about the History of
Family was to progress.
1972 was also the year when
the American social historian Lutz K. Berkner
published an article dealing with the methodical problems of the cross
sectional analysis of census records (Lutz K. Berkner,
The Stem Family and the Developmental Cycle of the Peasant Household: an
Eighteenth-Century Austrian Example, in: American Historical Review 77, 1972,
398-418). The explosion in the number of analyses of census records in the
succeeding years was to some extent accompanied by critical discussions of both
sources and methods (See especially Lutz K. Berkner,
The Use and Misuse of Census Data for the Historical Analysis of Family
Structure: a Review of Household and Family in Past Time, in: Journal of
Interdisciplinary History 7, 1975, 721-738; and the special volume 'Historische Familienforschung und
Demographie' in the first year of 'Geschichte und Gesellschaft', 1975). Yet, these discussions had remarkably
little impact on the actual process of empirical research. In these years of
euphoria about the contribution quantitative historical research based on
census records could bring to our understanding of the past societies the
methodological reflections influenced research practice only to a limited
extent.
All historians interested
at this time in large quantities of data faced a hard job. The most important
tools of the historian remained pencil and paper, and all those who considered
electronic data processing were confronted with huge and unwieldy machines to
make and read punched cards. 1972 was also the year when the American computer
scientist Alan Kay created the term 'personal computer': His utopian vision,
that every man would own his own computer amused the scientific community (Klemens Polatschek, Wer druckt, lebt
in der Vergangenheit, in:
Die Zeit, 26.6.1992, 54). After all, Edward Shorter
had published his book 'The Historian and the Computer' (Toronto 1971) only a
year earlier, in which he volunteered a few suggestions as how to change the
diffused information of historical sources into the neat, concise and
rectangular ordered numerical series, which seemed necessary to set computer
and programs in motion.
All these impulses from
family history as practiced at the international level found a reactive
audience among like-minded historians in
In the later 1970s Mitterauer succeeded in extending his research on family
history in a number of ways. Between 1977 and 1981 a project on 'A Comparative
Perspective on European Family Structure' subsidized by the 'Volkswagen
Foundation' in
At the same time a
long-term co-operation was opened up between the Viennese research group, the
Max Planck-Institute for History in Göttingen, the
Cambridge Group for the History of Population and Social Structure and also
with individual historians involved with family history in various European
countries. In Vienna within the program on 'Family and Social Change' the
research on family history was extended towards new sources and methods,
particularly oral history and, later, written autobiographies (On oral history
cf. the various works of Reinhard Sieder;
and for selection of autobiographies of people who in this project were
encouraged to write their life stories; see the book-series 'Damit es nicht
verloren geht', Böhlau Verlag, Vienna).
In the field of
quantitative research some new developments were taking place. At the Institute
for Advanced Studies in
Despite this innovation,
data processing still suffered from methodical shortcomings, which makes it
difficult to analyze these samples today. First there is the problem of the way
in which the populations were sampled. In the beginning of quantitative
research in family history the opinion prevailed that original census records
very seldom survived in a complete form, and that every source of this type
guaranteed an addition to knowledge. The result was that everything that could
be found in the Archives of various European cities was enthusiastically made
computer readable. As the project was already underway, it became clear that in
fact such census records could be found in large numbers all over
The size and the extent of
the samples also lacked uniform criteria. Villages, market-towns and smaller
towns were usually recorded completely. For large cities such as
The chosen format of data
entry created another problem. Every record representing an 'individual'
includes an item on the household role of that person. This input convention
seemed appropriate because the position in the household of each person was
accorded a higher value than were other characteristics. This rule had the
effect that every person was assigned a position in the household at the time
the data were recorded, regardless of whether such a role was mentioned
explicitly in the sources. Usually this was the case, but sometimes the
position in household represented only the interpretation of the researcher.
As a last example of
methodological problems, reference should be made to the methods applied to the
classification of occupations. The object was to differentiate between sectors
of the economy as well as between modes of production, especially between small
scale craft production and wage labor. The great temporal and regional spread
of the database however makes such a standardized system somewhat problematic:
A particular occupation judged typical of craft production in
These cautionary remarks
are not intended to discourage the use of the existing database. Previous
experience has shown beyond all doubt that the existing samples, despite their
shortcomings because of their temporal and spatial distribution, offer enormous
potential for resolving various questions concerning social structure and
family history from a comparative - but also from a local historical - perspective.
The potential of this information has not in any way been thoroughly exhausted.
The use of these samples simply demands the same critical approach as does the
use of any other source.
Finally it needs to be put
on record that it is because of the merit and hard work of Annemarie Steidl and Heinz Berger that it has proved possible to
publish this database. They checked, cleaned, systematized and documented the
data in such a way that it could be made available. The time, energy and
enthusiasm they spent on this project were far from being rewarded by the
financial compensation they received. They in particular deserve our thanks and
appreciation for the successful completion of this edition of the database.
For all questions or suggestions please send an e-mail to
annemarie.steidl@univie.ac.at
or heinrich.berger@univie.ac.at