[1] Limitations of space prevent inclusion of much relevant information on the history and culture of the study region (especially on the complex history of the Military Border), on technical details of the analysis, as well as full citation of the sources (especially those in Croatian). Interested readers should consult Hammel 1993c, which may be obtained from the Department of Demography, University of California, Berkeley, CA 94720.

² This research was supported by initial grants from the Joint Committee on Eastern Europe of the Social Science Research Council and the American Council of Learned Societies in 1983-84, the Wenner-Gren Foundation for Anthropological Research in 1985, the Center for Slavic and East European Studies at the University of California, Berkeley, in 1985-86. Major funding was provided by Grant No. BNS 84-18760 of the National Science Foundation in 1985-89, continuing as DBS 91-20159 and NICHD RO1 HD29512. Facilities in the U.S. were provided by the Department of Demography and the Quantitative Anthropology Laboratory at the University of California, Berkeley. The research was made possible by the hospitality and collegiality of the Institute for the Study of Folklore, Zagreb (now the Institute for Ethnology and Folkloristics), especially Dr. Dunja Rihtman, Olga Supek, and Jasna Capo, by the Archive of Croatia, especially Mr. Josip Kolanovic, by Profs. Vladimir Stipetic, Alica Wertheimer-Baletic, and Jakov Gelo of the Faculty of Economics of the University of Zagreb, Professors Igor Karaman and Mirko Valentic of the Faculty of History, Dr. Stjepan Krivosic, the Franciscan Monastery of Cernik, local civil registry offices and police stations in the Cernik region, and by the kindness of Prof. and Mrs. Vlado Ivir of Zagreb. I am indebted to Ruth Deuel and Carl Mason for programming assistance, to Jasna Capo, Dubravka Mindjek, Andrew Ruppenstein, Marija Olujic and Rada Bozic for research assistance and to Ronald Lee, Kenneth Wachter, and Patrick Galloway for analytic advice and commentary. I am especially obliged to Dr. Jasna Capo for her continuing cooperation in the analysis. None of the institutions or persons named are responsible for any errors of fact or interpretation. Preliminary reports of the research are to be found in Hammel 1990a, 1992, 1993c.

[3]The failure to find generally valid economic explanations for fertility behavior has led to a recognition of the importance of cultural factors, in more extreme arguments to calls for substitution of the cultural for the economic. Cf. Coale and Watkins 1986; Knodel and van de Walle 1979; Lesthaeghe 1983; Levine 1986; van de Walle and Knodel 1979; Tilly 1986; Watkins 1987, inter alios on the results of the European Fertility Project, Cleland and Wilson 1987 for a critique of microeconomic explanations, and Hammel 1990b on the use of cultural explanation.

[4] For a view of this war, see Hammel 1993a, 1993b. For an introduction to the history, ethnography, and economy of the area, useful initial sources are Tomasevich 1955; Moacanin and Valentic 1981. Capo's dissertation (1990, 1991) contains an excellent review of Slavonian feudal institutions. On the Military Border see especially Rothenberg 1960, 1966.

[5] Having been myself schooled in the importance of ethnic differences during extended ethnographic research in the Balkans, I expected on beginning the present study to find ethnic explanations paramount. The contrary outcome, except for the anticipated importance of the political status of military colonists was a surprise.

[6] See Andorka 1971, 1972, 1976, 1979a, 1979b, 1984, 1986; Andorka and Balazs-Kovacs 1986; Andorka and Farago 1983; Demeny 1968; Hölbling 1845; Lengyel-Cook and Repetto 1982; Vassary 1989.

[7] Hammel 1968, 1972, 1976, 1980a, 1980b.

[8] The size of a session of land depended on the time period, the place, and the quality of the land. See Capo (1990, 1991) for a discussion of these measures and their uncertain definitions.

[9] For details see Tomasevich 1955: 84-88. Landless peasants received no land in this distribution, and landed serfs received only the allotments that had been held in formal tenure (see below), not land they had held under nonfeudal sharecropping or other arrangements. Amelioration of their contractual relations was not achieved until 1876. For details see Karaman 1981.

[10] The Military Border was almost entirely lacking in activities other than subsistence farming, military construction, military activity, religion, and administration. However, the Gradiska Regiment that borders the estate of Cernik was the most "industrialized" of all the regiments; in 1857 it had 57 priests (about average for a regiment), 475 artisans (the highest number in any regiment), and 49 merchants (the next to highest after the 53 in Brod, the next regiment to the east), comprising nine percent of the population. The commercial activity centered in Nova Gradiska was right next to Cernik town, and both were on the trade route from Pozega to Bosnia and from Zagreb and Sisak to Zemun and Belgrade.

[11] Unfortunately, it is virtually impossible to determine from the published sources the involvement of specific regiments, especially the Slavonian ones, in military conflicts.

[12] See Capo (1990) for a detailed account. Land in the civil region was of four types: allodial land (the landlord's feudal holding), urbarial land (the land allocated to serfs in feudal tenure), industrial land (land taken from the commons by serfs and placed in agricultural production), and the common lands, which were available to all for wood gathering, pasturage, etc. In the military zone serfs held land as though it were allodial, as direct feudal tenants of the Crown, even though its allotment and distribution were closely controlled by the military authorities.

[13] Andorka, op. cit., personal communication.

[14] Druztva gospodarskoga hervatsko-slavonskoga 1850. Several of the accounts refer to the pressures generated within joint households when a less populous segment had to work hard to support a more populous segment. The resistance of adults to support their brothers' children was an important force in household division, as were the usual conflicts between the diligent and the lazy, the selfish and the altruistic.

[15] Biskupija Djakovacko-srijemska 1873a-b, 1874; Pavlinovic 1875; Kornfeind 1907; Biskupija bosanska i srijemska 1930a-c.

[16] Guci 1880; Krsnjavi 1882; V. Bogisic 1874; Valjevec 1901; Mirkovic 1903; Lulek 1914; Lukic 1921, 1926; Buturac 1941a-e.

[17] Lovretic and Juric 1897:376. See also Ilic Oriovcanin 1846, Jancula 1980, Jovic 1962, Matic 1970, Stojanovic 1852.

[18] Viz. Reljkovic's references to female beauty two centuries earlier. Women's "trinkets", usually necklaces of gold ducats, were the form in which family liquid wealth was kept, and were usually transmitted as dowry directly from mothers to daughters. Only males inherited real property and livestock. It is of some interest that a major cultural contribution of early modernization in this area was to transfer control of women's bodies from their husbands' lineages to the state, with pious churchmen continuing to cluck in the background.

[19] The history of inheritance in this region is unclear. The classic system of zadruga inheritance is agnatic and per stirpes, or division by successive generations of male heirs (po kolenima, "by knees", "by generations"). Under per stirpes inheritance the negative externalities generated across ramages do not occur, since the division of property is executed conceptually, step by step, in each generation. Per stirpes division is still the traditional form among Orthodox Serbs, and this factor may account for their more persistent high fertility in the Border (see later comments in the text). The alternative form of inheritance is per capita (po glavi, po muskoj glavi -- "by the head, by the male head,") among surviving heirs. Per capita inheritance to survivors at the time of division does generate externalities. There were efforts to legislate per stirpes inheritance in the late 19th century, perhaps to strengthen the traditional zadrugal system by diminishing internal pressures for fission that resulted from differential fertility of the constitutent nuclear families. That suggests that an original per stirpes inheritance may have been replaced by a per capita system, and that there were then attempts to restore the more ancient tradition, but one can only speculate. The debate is confused by issues having to do with the preservation of ancient Slavic tradition and pressures to liberalize the economic system (see Hammel 1980a).

[20] Pirc carried out his most detailed survey for the village of Otok near Vinkovci, in Eastern Slavonia. His technique is really quite remarkable for its time, in computing duration-standardized completed fertility rates for marriages that differed in duration across the different exogamic-endogamic categories. There are a few arithmetic errors in his results, which I have recomputed from the original data given. My report here omits the marriages in which both partners came from the old Military Border, since there are only 44 of those out of 1440 total marriages. Such marriages had fertility between that of the locally exogamous and the externally endogamous marriages, namely 3.4 children.

[21] Some of the earlier ethnographic sources claim that men had no ken of matters concerning birth and that abortion and contraception were practiced without their knowledge, being the province particularly of old women, especially midwives. It is certainly possible that male knowledge of and participation in fertility control increased over time. Sremec's account indicates that at least by the early years of this century a good husband was expected to practice coitus interruptus and that adult males applied moral, verbal, and physical pressure inter se to control fertility in joint households. I cite my favorite South Slavic proverb, "Da je brat dobar, i Bog be ga imao" ("If brothers were so good, then God would have had one.").

[22] See Hammel 1985 for analysis of short term fluctuations in the Military Border 1830-48 and Capo 1986 for the parish of Cernik 1760-1860. Other data are for 6 parishes of central Slavonia 1760-1860, gathered by Capo for Hammel, and reanalyzed in cooperation with Patrick Galloway (viz. Galloway 1987, 1988). Although this population was composed almost entirely of subsistence agriculturalists, price data on basic food commodities still serve as a reasonable indicator of food supply. It is as yet difficult to make precise comparisons between military and civil Croatia and between these areas and others in Europe. However, the elasticity of births in Croatia, with respect to economic indicators and net of endogenous demographic effects, seems among the highest in Europe. At the same time, nuptiality responses are correspondingly weak. The control mechanism thus seems not to have been a marriage valve but a birth valve. A full analysis of the comparative data will be published separately. See Komlos 1985, 1986 for a more general picture of immiseration as reflected in nutritiion in the Hapsburg Monarchy.

[23] Although the border was administratively merged with the rest of Croatia in 1881, the regimental districts of the former border region were clearly distinguished in the immediately subsequent censuses and can be tracked to 1910. I am indebted to Andrew Ruppenstein for organizing and recording these data. The western area comprises 5 regiments from Lika, headquartered at Gospic, around the shoulder of Bosnia to and including the first Banska Regiment, headquartered at Glina. The central regiments are the second Banska, headquartered at Petrinja, and then the four stretching from the Durdevacka through the Krizevacka and Gradiska to the Brodska. The eastern zone consists of one regiment, the Petrovaradinksa, which is in Srem. The five central and the eastern regiments lie on a great arc that runs from the Drava, south around the curve of the Sava, and to the confluence of these waters at the meeting of the Sava and Danube at Belgrade. Technically, Slavonia extends from the western border of Srem only to the Ilova river. But formerly the region of the Durdevacka and Krizevacka regiments was considered as part of Slavonia. This great arc, of course, surrounds and faces the central Slavonian massif and the Pannonian plain, including Baranya, within which fertility control seems to have appeared at an early date.