Using the International Bibliography of Historical Demography (IBHD) as an indicator of its tendencies

Serge NEKRASSOFF (1996)

At the 1985 IUSSP Conference held in Florence, Professor A. Perrenoud successfully scrutinized the IBHD in order to detect the main streams of research in progress in this speciality . Among his conclusions, he expressed the wish that a computerized IBHD should be used as a tool for the analysis of tendencies. This project is almost completed. For the last years, all items have been systematically stored in a data base. ISIS processing programs are able to tabulate , classify - without tedious enumerations- any parameter required for bibliometry considered either as a science or as a tool.

In 1995, IBHD inner structure and content provide new research performances. Somewhat different from the previous introduction this one is aimed at giving a rough idea of these and at suggesting some paths to all users, whatever their environment might be : a specialized department, a library or an information center.

Some performances as an indicator

The IBDH published more than ten thousand bibliographical items from 1978 on to 1991 . Of course, we do not claim that the data already collected are exhaustive. However when put together, we might assume they build up a fairly designed sample and they encompass years enough as to reduce a short-lived trend to its proper proportions .

The IBDH avoids value judgments and as such has no critical dimension. But it tends toward an analytical one viz. providing numerous descriptive explanations targeting the users' research.

For documentary purposes, an elementary level of data homogeneity is ensured by a set of instructions recommended to each contributor. These should make the computerized processes easier and more efficient. All contributors are trained in historical demography which should in itself guarantee a high level of scholarship.

The coding of each item according to their themes provides an outstanding tool for analysis giving access to a large array of further choices. Cross-checking according to period and country, ready for use either in each issue of the IBHD or in the data base, adds to the flexibility of a research design.

Truly, at its beginning, the yearly IBDH had not been conceived as an indicator. Hence the usefullness of some critical advices about its limits and its range. Let us have a look at its structure.

IBHD items are primarily collected by about thirty benevolent contributors being part of a worldwide network. They are given instructions dealing only with the formal shape of a bibliographical description. For anything else, they do as one pleases. The editor has thus to deal with a whole set of contributions; some are restrictive in their choice, others are reluctant whenever a selection is needed. Coding is open to discrepancies as well; some lay aside secondary entries whereas others resent as a shortcoming the fact that there are only six auxiliary entries. When two contributors describe the same item, it may happen that the coding differs. It then falls to the editor to arbitrate. Finally, until now, the IBDH includes only printed items leaving out many theses and students' reports about on going research. The best may subsequently be published, but a few years later, thereby providing a valuable stet forward the field of historical demography.

All previous critical comments are not to be considered as obstacles but as routine duties incumbent to most bibliographical inquiries. We have to keep in mind these limitations when using the IBHD as an indicator.

Redundancies and gaps

Thus we had first of all to ask appropriate questions fitted to the IBHD as a whole. Follo-wing examples intend to show how to avoid pitfalls or at least how to reduce uncertainty.

The field devoted to "geographical area" outlines the space step by step : continent, country, region. It should not be assimilated with the place where the item is issued. When asking the data base about the trend of interest toward Chinese demography, it is not relevant to list only items printed in China.

Chronological codes are modelled on classical European textbooks : Ancient, Medieval and Modern History. Though easy to understand, such an habit is not quite fitted to far away countries. Taking centuries into account would yield more accurate results in spite of the recent tendency to encompass slow and long-term evolutions.

Some interesting topics escape even careful scruting. For instance, a few pages devoted to population hidden in a village monograph; or marriage customs made obvious through genealogies. There are no special codes for such inquiries in spite of their virtual significance for demographers. Moreover, it would make sense to know whether those special studies which were frequent earlier are becoming less popular.

Similarly, even if there is a field devoted to methods, in fact it is often omitted to fill it and it would be unappropriate to ask our data base about the frequency of L. Henry's method of family reconstitution.

Starting from codes, investigations are made more intricate by the fact that IBHD uses both English and French. This complexity should be reduced by the compilation of a thesaurus designed for historians and demographers. Instead of translating usual dictionaries, it would be based on words currently used in titles of books and papers or at least in their English and French equivalent. That would become a useful tool for bibliographical research.

Coding according to themes has been primarily designed for dissemination on printed lists. The prevalence of one main code was considered as necessary whereas the alphabetical order of authors has to be limited. Such a process lays stress systematically on one side of each item, leaving others in the background, i.e. a numerical mention among auxiliary entries. Computerized procedures provide a remedy to that kind of ill-deserved priority. Hereafter we shall make use of this.

European hegemony: the weight of tradition or a matter of sources ?

Table 1 demonstrates a numerical prevalence of items dealing with European populations : there are 9,103 (= 71%) among 12,865 mentions of continents. On every single year, the proportion remains constant. The last year (1991) shows a tendency to weaken Europe's supremacy with "only" two thirds of the mentions.

Inside Europe, western countries play a leading role still reenforced by the growing contributions of Italian and Spanish scholars who have been given a strong implusion by new societies, namely SIDES and ADEH.

Focusing so much research on Europe might be related to the fact that a majority of IBDH contributors live in Europe : in 1979, they were 18 in a total of 34; in 1991, the disproportion is still more obvious : 19 out of 31. Roughly, one small continent is covered by half of the contributors. The network should be filled out elsewhere. Such an imbalance might also be due to other causes. Seminal is the fact that historical demography had its roots in western Europe where it grew thanks to typical sources, unknown elsewhere. Administrative traditions (censuses, parish registers) going back to the medieval period, the number of research centers and of historical periodicals disseminated in every region, stimulated the study of various sources outside a microcosm of highly specialized demographers. On the other hand, in most L.D.C.'s demographers have to deal with immediate problems. Until now, studies dealing with post World War I topics were kept outside the scope of IBDH. If this period were to be extended, it would increase interest for historical demography in former colonial countries.

A second but highly valuable place is held by North America which, in many aspects extended the field of European investigations especially on migration.

Overall the so-called developped countries take up nine tenths of bibliographical output. Continental Asia, South America, Africa and Oceania lag behind and, what is more, many of their items either have an European author or deal with immigrants from Europe. Let us finally notice an almost complete lack of studies about muslim countries.

Continuity

Constancy through the years is a striking feature in most figures. We were looking for an evolution or a trend in the long term. Continuity is dominant.

First the annual total of items has been about 800 since 1981. A peak of 901 items is reached in 1991 . It testifies to the health of historical demography and does not live up to predictions of its shortness of breath.
The chronological distribution is no surprise since this depends on the frequency and quality of available sources. Up to now, the 18th and 19th centuries have the advantage and this situation will last untill new methods prove efficient in the use of new types of sources.

The relative constancy to be noted in the distribution of items by coded themes calls for more comment (see table 4, Annual distribution by main codes). It is not obvious that historical demography would be confined to the same themes in the same proportions.

From 1979 on to 1991 studies devoted to the Spatial distribution of population (coded 21 to 28) were dominant with an annual average of about one fifth of the total. Only once -in 1981- was this numerical supremacy barely challenged by the themes related to Methodology (codes 101-107). Lower down, classical subjects are fairly well graduated: Mortality (14 %), Nuptiality, family, households (12 %), Population growth (10,5 %), Structure of past populations (10 %).The chapter on Methodology is filled at the same rate (ca 10 %). This suggests that further thought is continuously given by researchers to their speciality. It augurs well for needful innovative dynamism.

Lower again in the scale come Interrelations between economic (coded 81 to 84) or or social (coded 91 to 94) variables and demographic ones, with respectively 4 and 3 %. But this does not mean that they are neglected. As their titles suggested it, those are chapters connected with historical demography and, as such, they deal with external but influential factors. Interrelations are often taken into account by auxiliary entries. Their role is also obvious in items under the headings Differential mortality (coded 45) or Differential fertility by economic or social group (coded 53), etc. The same is true for chapter 7 : Structure of past populations and subpopulations (codes 71-77). Instead of being under-registered, items dealing with economic and social history are significantly taken into account in different but sensitive fields since they contribute to the explaining of meaningful demographic evolutions.

Finally the chapter on fertility may look somewhat neglected. But this standpoint needs correction since fertility is dealt with elsewhere, under the heading Demographic transition (coded 35).

Remarks about main entries apply also to secondary or auxiliary entries, with one exception. Reasons for this were given when commenting on the interrelations between demographic and socio-economic factors (see also table 3).
Discrepancies between regions.

Are there research themes which become a kind of regional speciality ? The IBHD provides some response on the frequencies of themes probably related to cultural features for instance religious denomination (roman catholic or protestant), attitudes toward death, birth control, family traditions, political ethic, behaviour in particular circumstances (migration, epidemics). More than 10.000 items constitute a solid basis to draw certain conclusions. However a scholar's productivity may depend on one single influence emanating either from a school of thought, or a strong personality, or even from a contributor's choice. In order to prevent such a bias in table 5, we selected themes and countries where significant numbers were available. Each percentage expressis the proportion between the total of items on a country and the number of items dealing with a particular theme within the country.

International migration (code 26) is the most frequent theme in the IBHD. It occurs in 35 % of items on the U.S., a pluri-ethnical country entirely settled by migrants and where even nowadays ethnical troubles still exist. South America and Canada share the same concerns. The significant weight of the theme of migration is to be related to the importance given to the theme Ethnic, racial and linguistic characteristics (code 73). The former USSR and Rumania bring plenty of reading material on Geographical aspects of past settlement (code 22), Estimates of past population (code 34), Ethnic, racial and linguistic characteristics (code 73).These again are sensitive focuses of attention for pluri-ethnic countries.

It may be noted, on an entirely different subject, that Rumanian scholars are silent about Birth control, a taboo under Ceaucescu's reign.

Western countries display less contrasts. There maximal frequencies scarcely reach beyond 20 %. The distribution of items according to main chapters is usually well balanced, even if the theme Spatial distribution of population is rather well represented, as in non-European continents. On the whole, historical demography does not suffer from exclusiveness.

As Professor A. Perrenoud already noticed ten years ago, each European country tills its own field of research independently. Hence differences prevail and until now they prevent any overall view of historical demography for the continent. The same applies for the U.S. and Canada.

A keener analysis, based upon associated keywords, should bring to the fore favourite debates such as the ancient Daces' settlement in Rumania or the Great Famine in Ireland. This method might be applied cautiously even now, while awaiting a thesaurus able to test its reliability.

Conclusion : losing momentum or preparing a new impulse ?

We have just surveyed the role IBDH has played as an indicator of tendencies. Its computerization has made possible smooth and easy advance through scrutiny of the whole corpus of items. As a tool this is not yet quite satisfactory but we are expecting better results from a thesaurus specially designed for historical demography and made compatible with ISIS's research routines.

Because they acknowledge the constancy of the number and distribution of items, some readers might be prone to perceive that historical demography already reached its cruising speed and consider it as a forewarning of a decline.

Several signs induce a more optimistic reaction. In itself the total amount of items collected every year is a sign of lasting success. For instance, Spain recently joined the group of leading countries not only by the mere quantity of its contributions but by their high quality and diversity. Outside Europe research about LDC's is more frequent and at the time this paper is written, the International Commission on Historical Demography is overcrowded with more than 80 reports to be delivered at the Montreal Conference (Aug. - Sept. 1995). Finally a lot of bibliographical data still await our attention; unexpected discoveries require further exploration through Internet and recent technologies.

The attention given to the economic and social environment means that historical demography is not confined in a merely descriptive role devoted only to measuring classical quantitative variables : fertility, nuptiality, mortality. On the contrary, it is about to build an identity of its own. Nowadays it intervenes as a way of understanding socio-economic environment of past and present civlizations. It clarifies man's responses to change, whether conscious or no, either sudden or barely perceived in the long run. In this way historical demography may become influential and play a meaningful role.

Historical demography might not be blamed for ignoring the present time. Researchers scrutinize IBHD on current problems : migration, segregation, national minorities...A look at this 1992-1993 issue shows the attention paid to the environment. The time when man believed he could rule Nature is gone. More than ever demographic variables are to be related to ecological conditions.

Another stimulating way forward is the discovery of new methods to be applied where traditional ones prove ill-fitted or not suitable.That would lead to the use of new sources and would introduce investigations in other periods and other countries.

Historical demography stands at a crossroad of variegated disciplines. Hence its rigorousness. It requires expertise in historical criticism, in demography, in statistics, in computer science among others. Hence it is swept along in the main streams of to-day, when new technologies bring within reach a whole display of methods and large quantities of data. The time is ripe for an encounter or a partnership between disciplines, not for segregation. By its next issue, the IBHD will benefit from recent progress in processing, circulating and collecting data.

Serge NEKRASSOFF (Nov. 1996)

Notes

1. A. PERRENOUD, Où va la démographie historique ? Analyse du contenu de la Bibliographie internationale, in Annales de la démographie historique, Paris, 1986, pp.251-272.

2. The CDS-ISIS program, concieved by UNESCO, is available free of charge for any scientific institution.

3. For any further information dealing with the building of the IBHD data base, see the previous IBHD introductions.

4. We have deliberately excluded the items of the first issue of the IBDH. In a way, this was a first attempt and thus did not present the characteristics of the following issues yet (number of contributors, of items, ...).

5. For exemple, the publication of the proceedings of a conference on a particular theme can give momentarily a higher importance to the code associated with this theme.

6. 1979 557 notices 1984 828 notices 1989 621 notices
1980 577 notices 1985 745 notices 1990 685 notices
1981 747 notices 1986 821 notices 1991 901 notices
1982 822 notices 1987 850 notices 1992/93 1601 notices
1983 725 notices 1988 863 notices

7. Lowest total reached 270 items collected about Japan.

8. In this issue (1993), items dealing with South America are already significantly more numerous

TABE 1 : Geographical descriptors for each continent

	Citations	%
World	424	3,4
Africa	352	2,7
America North and Central	1721	13,4
South America	373	2,9
Asia	653	5,1
Europe	9103	70,7
Oceania	239	1,8