University of Massachusetts Amherst

From the SelectedWorks of Joel M. Halpern

2012

Patriarchy in the Balkan Temporal and Cross Cultural Approaches Joel Halpern Karl Kaser Richard A. Wagner

This work is licensed under a Creative Commons CC_BY-NC-ND International License.

Available at: https://works.bepress.com/joel_halpern/133/ Studies on South East Europe

Karl Kaser (Ed.) Household and Family in the Balkans Two Decades of Historical Family Research at University of Graz

LIT Household and Family in the Balkans Two Decades of Historical Family Research at University of Graz

edited by

Karl Kaser

LIT Gedruckt mit Unterstiitzung der Karl-Franzens-Universitat Graz und des Landes Steiermark

Das Land Steiermark UNI -ยป Wissenschaft GRAZ

Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the Internet at http://dnb.d-nb.de. ISBN 978-3-643-50406-7

VERLAG GmbH & Co. KG LIT VERLAG Dr.w.Hopf Wien2012 Berlin 2012 Krotenthallergasse 10/8 Verlagskontakt: A-1080Wien Fresnostr. 2 Tel.+43(0)1-4095661 D-48159MUnster Fax+43 (0)1-409 56 97 Tel.+49 (0)2 51-620 320 e-Mail: [email protected] Fax+49 (0)2 51-23 19 72 http://www.lit-verlag.at e-Mail: [email protected] http://www.lit-verlag.de

Auslieferung: Deutschland: LIT Verlag Fresnostr. 2, D-48159 MUnster Tel. +49 (0)251-6203222, Fax +49 (0)251-9226099, e-Mail: [email protected] Osterreich: Medienlogistik Pichler-OBZ, e-Mail: [email protected] Schweiz: B + M Buch- und Medienvertrieb, e-Mail: [email protected] JOEL M. HALPERN, KARL KASER AND RICHARD A. WAGNER

PATRIARCHY IN THE BALKANS: TEMPORAL AND CROSS-CULTURAL APPROACHES

ABSTRACT

The article deals with the history, distribution and ideology of Balkan patriarchy, the Balkan agnatic kinship system, and with the most significant patriarchal family struc- tures. The demographic analyses are based primarily on the Serbian state census of 1863 (the first available census listing women) and the Federal Yugoslav census data for 1948, 1953, and 1961, coupled with archival and field data for the Central Serbian village of Orasac for 1818-1975. Balkan patriarchy has much in common with similar well-documented systems in Asia and the Middle East. The Balkan situation was dif- ferentiated, however, in that this system existed both within and outside formal state structures. The patriarchal ideology shaped kinship and family patterns, as well as coresidential patterns within households. These patterns reproduced patriarchal struc- tures, but the full reproduction of the system was constrained by economic circum- stances.

INTRODUCTION

This article is the product of a joint American-Austrian research project on patriarchal and family structures in the Balkans. Two American cultural anthropologists and an Austrian historian, who considers himself an anthropological historian, propose that by combining historical-demographical data and historical-anthropological ap- proaches it is possible to examine the evolution of family and kinship structures in the Balkans. 48 JOEL M. HALPERN, KARL KASER AND RICHARD A. WAGNER

At present, the rationale for patriarchal structures is being questioned, especially by processes of modernisation underway since World War II. Historically the Balkan patriarchal pattern encompassed most of the Balkan Peninsula. It was present in its regional complexity from Southern Croatia and Bosnia-Herzegovina to Serbia, West- ern and Central Bulgaria, Montenegro, Albania, Macedonia and Northern Greece. As is evident, these regional forms crossed national borders. The northern and western transitional zone of patriarchal patterns which ran through Croatia; the coastal parts of

Map 1: Distribution of patrilineages and singular family structures around 1850 and at present

Distribution at about 1850 Singular distribution at about 1850 Present distribution

Source: Adapted from Kaser 1995: 268.

Dalmatia, Northern Croatia and parts of Eastern Slavonia were outside the pattern's distribution. Today this patriarchal pattern is still strong in regions populated by Alba- nians, especially Kosovo, Western Macedonia and Northern Albania (see Map 1). This article is divided into three parts. The first deals with the history, distribution and ideology of Balkan patriarchy. The second part deals with the agnatic kinship system, and the third with the most significant patriarchal family structures. The de- PATRIARCHY DM THE BALKANS: TEMPORAL AND CROSS-CULTURAL APPROACHES 49 mographic analyses are based primarily on the Serbian state census of 1863 (the first available census listing women and the Federal Yugoslav census data for 1948, 1953, and 1961, coupled with archival and field data for the central Serbian village of Orasac for 1818-1975. In giving primacy to these two data bases it is necessary to stress that the evidence cited here by no means exhaust historical demographic resources for the region. For example, from an anthropological perspective, Hammel and his associates have undertaken extensive analysis of Croatian records, especially those from Slavo- nia. The work of the demographer Botev on Bulgaria and the Balkans generally (Bo- tev 1988: 1990) deals with fertility decline while the monograph of the social historian Todorova (1993) analyses the family structures in the Bulgarian area. Many of these analyses, however, are more strongly focused on demographic factors such as the evo- lution of fertility and mortality patterns but these directly affect patriarchy. Balkan patriarchy can be defined as a complex of hierarchal values embedded in a social structural system defined by both gender and age. This structuring is further linked to a system of values orienting both family life and broader social units. Balkan patriarchy achieves its historical form through the classically complex and interlock- ing systems of patrilinearity, patrilocality, and a patriarchally-oriented common law. Such supports not only divide and ascribe position by gender, but also allocate to males the predominant role in society. An obvious corollary to this defined structure is the formal subordination of women within the context of an overtly 'protective' family and household environment. Ideally the father has unquestioned authority over his sons and the older brother over the younger. Its broader manifestations are many, but are perhaps best exempli- fied by the supremacy of male moral authority reinforced through both traditional and formalized state law codes (Hasluck 1954; Kanuni 1989; Whitaker 1976; 1981). As a system, Balkan patriarchy has much in common with similar well- documented systems in Asia and the Middle East. The Balkan situation is differentiat- ed, however in that this system existed both within and outside formal state structures. For example the particular forms in which this pervasive concept of patriarchy has been expressed are sensitive to the differences between agricultural and pastoral econ- omies. And, indeed the same peoples have alternated from one system to another; the former being associated with state-level structures and the latter based on pastoral tribal organizations forming the baseline model in this article.

HISTORY, DISTRIBUTION AND IDEOLOGY OF BALKAN PATRIARCHY

In researching this topic one can focus on processes of development or in distribution. Maria Todorova's recent comprehensive analysis (Todorova 1993: 133-158) takes a view at variance with that of Kaser. She stresses the demographic aspects of the com- plex family systems and implicitly that of the patrilineage. Her reaction to the sugges- tion that the Balkan patriarchal family can be viewed as an archaic survival is to re- strict the existence of this pattern. Todorova sees the pattern as a response to changing economic and political processes in the Ottoman Empire at the end of the 18th and the beginning of the 19th centuries. 50 JOEL M. HALPERN, KARL KASER AND RICHARD A. WAGNER

Instead of using primarily a demographic approach Kaser stresses a cultural one. There are many indications that the Balkan family pattern is indeed of archaic origin and that its existence is connected with a pastoral economy and a general patriarchal pattern (Kaser 1994a: 45-68). A major problem in analysis is that archival population data are lacking for non-state . Therefore these two approaches deal with dif- ferent kinds of data, oral traditional accounts and population registers. Sometimes there are possibilities of comparing the two types of data and when that has been done they have been found to be mutually supporting, each providing a separate type of information (Halpern & Kerewsky-Halpern 1980; Wagner 1983). Following Kaser's approach, this pastoral socio-cultural legacy can be called the 'Illyrian heritage'. Patrilinearity, one of the cornerstones of Balkan patriarchy and the Balkan family pattern, has to be considered in its full historic depth, one which pre- dates both Roman and Greek civilizations. Patrilineal structures cannot have emerged first at the beginning of the 19th century coincident with the written documentation of a pre-existing Balkan family pattern (Kaser 1992: 1-39). Both the Balkan joint family and the patrilineage emerged first as results of pasto- ral economies and the patriarchal influence of Illyrian cultural legacy. (In part, the comparable culture of the Central Balkans is an autonomous development.) After the Roman conquest of the Illyrian lands these features were preserved by Albanian and Vlach nomads. They were later joined by Slavic groups who followed them into the uplands. What we have here is a phenomenon within limits of an adaptive strategy based on both ecological factors and predatory expansion. The idea of a relationship between pastoralism and the existence of both the joint family household and the patrilineage is not new. Todorova describes the highest con- centration of joint family households in Western Bulgaria in regions with a large area of meadows and a developed pastoral economy (Todorova 1990: 18-19). Earlier, Mosely stated that, in general, the joint family had shown a greater viability in the mountainous regions of the Balkans than in the plains (Mosely 1976a: 31). Filipovic notes, the "appearance and persistence of the zadruga as an institution originated in connection with livestock herding" (Filipovic 1976: 273). While Mitterauer states that the distribution of the joint family households is basically confined to mountainous, remote regions where a money economy and forms of wage work played a lesser role, he also suggests that a pastoral economy might have promoted the emergence of com- plex family structures (Mitterauer 1980: 67-69). The Ottoman conquest of the Balkans from the 14th to the 16th centuries was gen- erally accompanied by massive migrations of the Balkan peoples in a variety of direc- tions. Reconstruction of the migration movements is difficult, but the main direction was from south to north following the pattern of conquest. Pastoralists or semi- pastoralists, recently settled, rediscovered their former survival strategies. The moun- tain regions became repopulated (Cvijic 1922: 127-181). Generally, the Ottoman ad- ministration did not absorb the mountain dwellers, and so they independently devel- oped appropriate social structures and concomitant survival strategies based on the patrilineage and patriarchal joint family. The joint family, like the lineage of which it was a part, was never static but un- derwent fissioning following the dynamics of the life course and family cycles. The tribal lineages constructed of these joint families were reinforced by their focus on shared sentiment and ritual. Thus the Balkan joint family became the basic unit for PATRIARCHY IN THE BALKANS: TEMPORAL AND CROSS-CULTURAL APPROACHES 51 patrilineal tribal lineages that developed from the 14th century onward. This system was flexible enough to adapt to the bilaterally based kindred of Vlachs and Sara- katsans. At the same time, this plasticity enabled the individual household to create cyclical alternations of nuclear and joint family households depending on fertility, fission and fusion (Halpern & Anderson 1970: 83-97). In this way these units also functioned for settled agriculturalists. What characterized patriarchal Balkan social structure, as the pioneering works of Cvijic illustrated, was the constant interrelationship between becoming settled farmers and/or pastoralists. Until the 19th century this was a reversible process. This ended with the spread of industrialisation, urbanization, and the modern state. It is thus much more logical to assign the origin of the Balkan joint family to the goat- and sheep- keeping families of the mountains than to see it as a result of conditions in the plains. But the fact is that many joint families resided in the valleys and plains. How then did patriarchal joint family and patrilineage emerge in the plains? For centuries pastoral families of the mountainous regions migrated into the plains where they settled. In the generally chaotic situation caused by the Ottoman conquest not only did Slavic families flee to the mountains, but others, especially those of the Vlachs, left their mountainous homelands and settled in Ottoman-occupied territories. The valleys of Serbia, Bosnia, and, especially along the borders between the Ottoman and the Habsburg empires, were favoured sites. They were attracted by open land and by the privileges held out to them by the Ot- toman administrators who wished to have the land resettled as a defence strategy (Hammel 1972: 345-346). Another reason for leaving the mountainous regions was tension that arose due to population increases and the limited resources available to a pastoral society. The carrying capacity of the land was obviously much greater in the valleys where the focus was on labour intensive agriculture. Ancestor worship, warfare and vengeance, and agnatic dyadic relationships were the core elements of patriarchal ideology. This agnatic lineage structure and its patri- archal ideology are associated with Eurasian pastoral societies, as is ancestor worship. This was the ritual annual religious feast (in several South Slavic languages called slava, in Albanian feshta) in which a patron saint was venerated. In the early decades of the 20th century this was the most important religious feast of the year and one that was held at home, not in the church, a very important factor in putting the focus on lineage identity. The core of this event was then the pre-Christian worship of ancestors of the patrilineage, for whom a Christian patron saint was substituted (Kaser 1993: 93- 122). Protection in a hostile political and economic environment was one of the main aims. Predatory warfare reflecting these conditions was a main cultural focus, as is well-documented in traditional oral epics and explicitly linked by Albert Lord and others to the warrior tales of classic Greece (e.g. Homer's Iliad and Odyssey). The rifle was both symbol and tool for accomplishing these ends; hostile conflicts with other tribal units were frequent. Adherence to oral-traditional legal codes, which subse- quently achieved written form, were stressed, and masculine pride tied to lineage sta- tus were strongly held values. Infringement of these codes was severely punished. Indeed, these agnatic structures were well-adapted to forming vengeance units (Durham 1928; Boehm 1984). The idea of lineage vengeance persisted even after coming under state control and emigration. Thus The New York Times has carried 52 JOEL M. HALPERN, KARL KASER AND RICHARD A. WAGNER reports of vengeance killings in the Bronx (New York City) where there is a signifi- cant Albanian population. Interestingly, the Albanian minority has not been preyed on by gangs of other ethnic groups because of their reputation for exacting vengeance. Another specific of Balkan patriarchy becomes clear if we focus on kin relation- ships at the household level, as manifested in particular dyadic relationships in an age- specific context. Through this, we can more specifically understand the occurrence of particular processes.

Table 1 Relation to household head, Orasac 1863 Relational Term Frequency Percent Father 1 .1 Mother 21 1.9 Father's Brother 2 .2 ****Head**** 131 12.1 Wife 112 10.3 Brother 62 5.7 Brother's Wife 28 2.6 Sister 26 2.4 Father's Brother's Daughter I .1 Son 257 23.7 Daughter-in-law 49 4.5 Daughter 200 18.5 Brother's Son 48 4.4 Brother's Son's Wife 1 .1 Brother's Daughter 46 4.2 Step- son 2 .2 Step-daughter 1 .1 Grand-son 56 5.2 Grand-daughter 39 3.6 Total 1083 100.0

Source: Balkan Family Project, University of Graz

If we define the critical structural features of patriarchy as embodied in the father- son and brother-brother dyads then we can measure both the frequency of these rela- tionships and their time span. Further, one of the key factors in the decline of patriar- chy and the emergence of the nuclear family is the growing prominence of the hus- band-wife relationship as a counterpoint to the prevalence of agnatic ties. This ap- proach also avoids the need to focus on household size which can sometimes more reflect high fertility rather than organizational complexity. This then is an example of the way in which historical reconstruction of patriarchal forms can be correlated with historical demographic analyses. PATRIARCHY IN THE BALKANS: TEMPORAL AND CROSS-CULTURAL APPROACHES

Table 2 Relation to household head, Orasac 1961 Relational Term Frequency Percent Grandmother 10 .5 Father 3 .1 Mother 78 3.9 Father's Brother's Wife 2 .1 Wife's Father 1 .0 Wife's Mother 10 .5 Stepmother 1 .0 Aunt 1 .0 Foster Father 1 .0 ****Male Head**** 412 20.4 Wife 363 18.0 ****Female Head**** 38 1.9 Brother 11 .5 Brother's Wife 8 .4 Sister 11 .5 Wife's Brother 1 .0 Wife's Sister 3 .1 Son 345 17.1 Daughter-in-law 155 7.7 Son-in-law 12 .6 Daughter 199 9.9 Brother's Son 3 .1 Brothers' Sons Wife 1 .0 Brother's Daughter 2 .1 Adopted Daughter 1 .0 Grandson 165 8.2 Grandson's Wife 9 .4 Granddaughter 139 6.9 Grandniece 1 .0 Great Grandson 12 .6 Great Granddaughter 11 .5 Servant 2 .1 Female Servant 1 .0 Roomer 1 .0 Uncertain 2 .1 Total 2015 100.0

Note: In 1961 8.8% of individuals are listed under other categories. Source: Balkan Family Project, University of Graz. 54 JOEL M. HALPERN, KARL KASER AND RICHARD A. WAGNER

Table 3 Selected kinship for nine communities by relation to household head, Yugoslavia 1961 (number of cases and percent of total population) Bobo- Lekenik Orasac Veleste Zupca La- Slano Bukovi- vac buniste ca Head 318 491 450 393 248 475 303 164 26.2 30.1 22.3 14.4 21.4 17.4 22.6 22.4 Wife 252 358 362 330 206 396 220 111 20.8 21.9 18.0 12.1 17.8 14.5 16.4 15.1 Son 188 286 345 697 336 722 270 207 15.5 17.5 17.1 25.5 29.0 26.5 20.1 28.2 Daughter 86 214 199 434 284 450 191 169 7.1 13.1 9.9 15.9 24.5 16.5 14.2 23.1 Brother 5 3 11 95 4 24 21 10 .4 .2 .5 3.5 .3 .9 1.6 1.4 Sister 3 5 11 13 2 4 26 15 .2 .3 .5 .5 .2 .1 1.9 2.0 Mother 86 45 78 71 9 57 36 13 7.1 2.8 3.9 2.6 .8 2.1 2.7 1.8 Daugh- 87 50 155 173 31 204 53 8 ter- 7.2 3.1 7.7 6.3 2.7 7.5 4.0 1.1 in-law Grandson 60 53 165 204 14 178 56 9 5.0 3.2 8.2 7.5 1.2 6.5 4.2 1.2 Grand- 48 41 139 156 14 144 59 10 daughter 4.0 2.5 6.9 5.7 1.2 5.3 4.4 1.4

1 = Bobovac/Croatia/Croat/Catholic. 2 = Lekenik/Croatia/Croat/Catholic. 3 = Orasac/Serbia/Serb/Orthodox. 4 = Velesta/Macedonia/Albanian Muslim. 5 = Zupca/Bosnia/Muslim. 6 = Labuniste/Macedonia/Muslim-Orthodox. 7 = Slano/Croat/Croatia/Catholic. 8 = Bukovica/Montenegro/Montenegrin/Orthodox. 9 = Arandjelovac/Serbia/Serb/Orthodox.

Source: Balkan Family Project, University of Graz.

The analyses that follow are based on the relationship to household head as defined in census lists, hi the Serbian village of Orasac, for the years for which we have com- plete census data (i.e., 1863, 1928, 1958, 1953, and 1961), the frequency of the term 'son' was exceeded by that of 'wife' for all years until 1961. Partly this was due to the number of children in households, so that in 1863 the most frequent term was that of PATRIARCHY IN THE BALKANS: TEMPORAL AND CROSS-CULTURAL APPROACHES 55

'son' followed by that of 'daughter'; 'wife' was less than half of that (257 as opposed to 200 out of a total population of 1083 persons). As fertility began to decline, the number of sons became almost double the number of daughters in 1928 (302 versus 171); a disparity that continued until 1961 (345 to 199). The proportionately larger number of sons is due to the continuing presence of married sons and is documented by the continuing significant presence of the classifi- catory kin term snaja, which refers both to daughter-in-law and sister-in-law. As the population aged and there were proportionately fewer children, it is the continuation of the pattern of inmarrying sons that contributes to this disparity along with the decline in household size. In the hundred-year period between 1863 and 1961, the total popu- lation of the village approximately doubled from 1083 to 2015. The number of female in-laws also almost exactly doubled from 78 to 163 during the same period. These figures document the persistence of the agnatic ideology over the course of a century despite a precipitate decline in household size. In 1863 approximately one- third (32 percent) of all households with ten or more members contained half (49 per- cent) of the total population. By 1961 households with ten or more members repre- sented only 1.3 percent of all households, or 3.1 percent of the total population of the village. During this same period the average household size declined by almost half, from 8.3 members to 4.5. The persistence of the agnatic ideology, however, does conceal a tremendous struc- tural change in complex households from a predominant pattern of lateral extension to one of lineal extension. Here again the relative frequency of kin terms depicts a clear picture of the transformation - a transformation best seen in the drastic decline in the number of brothers within households. In 1863 there were 62. The number had halved to some 32 in 1928 despite an approximate 50 percent increase in population (to 1585). By 1961 there were only eleven, a seven-fold decline over a century. At the same time the number of grandsons approximately tripled from 56 to 165 while the population approximately doubled. By examining the nationally distinct regions in 1961 of what was then Yugoslavia, we can see the temporal changes played out in Orasac displayed across contrasting economic and environmental settings. In Table 3 nine communities are contrasted (Bobovac, Lekenik, and Slano in Croatia [Catholic]; Orasac and Arandjelovac in Ser- bia [Orthodox]; Velesta in Macedonia [Albanian Muslim]; Zupca in Bosnia [Bosnian Muslim]; Labunista in Macedonia [mixed Muslim-Orthodox]; and Bukovica in Mon- tenegro [Orthodox]. The Albanian village of Velesta in Macedonia most closely ap- proximates the model of 1863 Orasac as far as fraternal household composition is concerned, with by far the highest number and proportion of brothers. The most con- servative patrilinear household structures, in addition to Orasac in Serbia and Velesta in Macedonia, are the nearby village of Labunista (near Velesta) and the somewhat isolated village of Bobovac in Croatia. The predominantly nuclear household structure of the now relatively large market, industrial and tourist town of Arandjelovac, with a diverse population, is clearly evident. Arandjelovac had the highest proportion of nu- clear family household members, household head, wife, son, or daughter, and lowest for each daughter-in-law and grandchildren. Another measure indicating the persistence of patriarchal ideology is constancy in age at marriage coupled with age patterns in childbearing. In data from Orasac, mean age at marriage for women born between 1850 and 1939 has varied from a low of 19.4 56 JOEL M. HALPERN, KARL KASER AND RICHARD A. WAGNER years (women born 1870-1879) to a high of 23.4 years (1890-1899). Women born between 1930 and 1939 had a mean age at marriage of 21.2 years. At the same time, mean age at first birth varied from a high of 24.9 years (women born 1850-1859) to a low of 21.3 years (women born 1910-1919). Thus most women still spent their entire adulthood in a married relationship. Undoubtedly, part of the tendency toward early childbearing relates to the im- portance of children for establishing adult status. The continued ideological im- portance of sons however can be best demonstrated not by early births, but rather by the preponderance of males over females in last births. In Orasac, males account for 60 percent of all last births to women whose marriages survived until age 45 or until 1975. This measure reached a high of 71 percent for women born 1910-1919, but was still over 63 percent for women born between 1930 and 1939. Based on other analyses, these figures apparently do not reflect disproportionate non-registration of female births - a conclusion consistent with Vukmanovic's general assertion that under-registration is not sex-specific (1971: 72). Nearly all deviation from expected norms can be removed by excluding last births. With last births re- moved, the sex ratio for live births of women whose marriages survived until age 45 is 50.5:49.4. This would seem to indicate the tendency to try once again for another son after the birth of a daughter. More importantly, it raises questions about the social importance of mean completed family size as a measure of ideological concepts. As an alternative measure, we suggest the number of sons born to Orasani mothers. Today, as was true a century ago, most women strive to have one or two surviving sons (Wagner 1992).

THE PATRILINEAGE

Many of the complex family structures in the Balkans can be traced to tribal lineage systems or more generally, to large kinship agglomerations in the mountainous Dinar- ic regions. One of the basic structures of Eurasian pastoral societies has been an agnat- ic kinship ideology centred upon a named male ancestor. His sons were regarded as founders of sub-lineages, and their sons of smaller segmentary lineages. Such lineages in the Balkans did not, of course, simply evolve freely as a natural adaptation to a highland ecology. Rather, the mountainous areas were refuge zones beyond the ex- panding Ottomans. Thus the coming into prominence of these lineage structures re- flected a local adaptation to the absence of larger state structures. That these lineages were also functional in an upland ecology was another important conditioning factor. Tribal autonomy, however, whether in the Balkans or elsewhere as in the Middle East and Central Asia, was never absolute but always relative, sensitive to shifting political and ecological factors. Under conditions of resettlement in valley areas, a patrilineage could become synonymous with all the households of a village, or a specific part of a village. Sometimes the households of a patrilineage could be spread over several vil- lages of a region (Cvijic 1992: 115-126). The most frequent Balkan type, however, was that forming a quarter of a village (mahalla). Examples have been described for PATRIARCHY IN THE BALKANS: TEMPORAL AND CROSS-CULTURAL APPROACHES 57 both rural Central Serbia (Halpern 1956: 38-39, 312-328; Halpern 1967) and in Cen- tral Bulgaria (Sanders 1949: 232). Conditioned by different degrees of generational depth, patrilineages were seg- mented minimal and maximal lineages. Each unit had different competencies in deci- sion-making. The minimal lineage consisted of up to three generations. In other cases, a threefold subdivision was common. Serbian ethnographers distinguish, for example, between the household (kucd) as minimal lineage; the broader extended family, in- cluding first and second cousins (familja), as middle lineage; and the brotherhood (bratstvo) as maximal lineage, consisting of as many as thirteen or fourteen genera- tions of agnatically related kin (Djurdjev 1954: 165-220; 1965-1966: 187-195; Pulaha 1975: 121-145; 1976: 173-179). Hammel sums up some of these relationships and indicates their regional variation and bilateral meanings. Bratstvo is the most inclusive group, literally meaning 'a brotherhood'. Porodica comes from the root rod, referring variously to kinship, gen- der, lineage and birth. In Montenegro, porodica can sometimes refer to the maximal agnatic group, while in much of Serbia it denotes the nuclear family. On the other hand, familija refers to the nuclear family in Montenegro and the specific agnatic line- age in Serbia. Rod often has a particular meaning for women, as when the say "idem u rod' (I am going to visit my kin). But both sexes might say "mi smo rod" (we are kin). The identical root, rodbina, can at times be employed as equivalent tofamilija (i.e., the Serbian agnatic group), but can also indicate a kindred comprised of consanguineal relatives connected through both agnatic and uterine connections (Hammel 1968: 27). The patrilineage was not only an ideological unit but also, by definition, an exoga- mous corporate group. In 19th-century Montenegro, as in present-day Northern Alba- nia this rule went hand-in-hand with economic structure. Lineage males cooperatively owned common territory (mainly woods and pastures). Exchange of women with their dowries was perhaps the key factor that bound lineages together, and was likewise integral to the process of household formation. Within this patrilocal framework, the formal subordination of women needs to be considered in a setting where the role of mother's brother is important. These affinal ties provide both balance and serve as a way of integrating the lineage system. The wife never forgot her family ties, and these links provide an important theme in the oral-traditional folk poetry. This is best illustrated in the saga of a wounded young man and the demands of the mountain spirit when a sister provides all, symbolized by her hair, but a wife refuses to part with a part of the trousseau (a necklace) given by her father, thus reaffirming this emotive tie and giving it affective primacy over her marital bond (Halpern & Kerewsky Halpern 1986: 94). This bridging ideology sym- bolized by the tale helps us understand that patriarchal societies cannot operate com- posed of isolates but are linked by the role of women. Two types of social structures evolved in this area of the Balkans. In Northern Al- bania and Montenegro, tribal organization emerged. Affirming this similarity in social structure of the Montenegrins and Albanians provides an important insight into the ways in which Balkan societies have been structured. There were some 30 named Montenegrin (pleme) in the second half of the 19th century, and more than 60 identified Albanian tribes (fis) at the beginning of the 20th century (Durham 1928: 13- 52). Recent historical investigations contend that both the Montenegrin and the Alba- nian tribes emerged during and after the Ottoman conquest and were not the direct 58 JOEL M. HALPERN. KARL KASER AND RICHARD A. WAGNER continuation of older structures (Djurdjev 1954: 165-220; 1965-1966: 187-195; Pulaha 1975: 121-145; 1976: 173-179). Political authority in these regions resided at the level of the . The position of the chieftain was an official rank, its competencies were fixed, and its authority stressed. Such status tended to reside within a certain family or lineage. It was characteristic of both regions that lineages controlled defined territories of summer and winter pastures recognized in common law. This legal basis made these structures very resistant to the efforts of the modern state to destroy these self- governing units. In Northern Albania, having survived almost a half century of op- pressive communist rule, most of the traditional patrilineages and their territories still exist today (Kaser 1992b: 179-202). By contrast, in the south (as in Southern Albania and Greece) there developed an alternative organization characterized by the leading role of the lineage segment in political, religious, and economic activities. This second pattern also was reflected in Serbia, Western Bulgaria, Bosnia-Herzegovina, Macedonia and, to a lesser extent, in Croatia, where the authority of the organized state prevailed. This southern pattern was utilized by the Habsburgs in forming military units to guard their frontiers. De- scendants of these latter folk constituted much of the Serbian minority in Croatia up to August 1995 and were at the heart of the former autonomous Serb mini-state of Rraj- ina( 1991-1995). The more isolated these pastoral societies were, the greater their tendency to con- struct segmentary lineage systems. The Albanian mountain dwellers in Southern Al- bania and Northern Greece, through intensive segmentation, lost their tribal organiza- tion during the time of the Ottoman occupation (15th century). These lineage seg- ments then became political, economic, religious and predatory units. These patriline- ages had cycles of fission, parallel to the cyclical division of households and like household fissioning were conditioned by outside influences. A Balkan patrilineage used a number of important symbols that bound lineage members together (Simic 1991: 28). There was, for example, the common lineage name derived from the male founder and carried by each male lineage member. The Mihajlovici, for example, were the lineage derived from the common ancestor Mihaj- lo. Many lineages had origin myths, orally transmitted from generation to generation which were essentially charters of identity (Hahn 1854: 183-210; Tomic 1902: 357- 497; Drobnjakovic 1923). Individuals from the oldest extant Balkan lineages can orally trace their ancestors some thirteen and fourteen generations back to the Ottoman period. However, in such cases even where the oral tradition remains strong, one gets the names of only the directly connecting series of ancestors to the exclusion of collateral kin. The feast of the patron saint (slava orfeshta), thought to be the protector of the lin- eage was elevated by Serbs to be a vital part of their cultural heritage. An old Serbian folk saying links the existence of the slava to the presence of Serbs: "Gdje je slava, tu je Srbin" (Where there is the slava, there are Serbs) (Halpern & Kerewsky-Halpern 1986: 110). It was the most important religious feast of the year, lasting three days or longer, and held close to the calendar day of the patron saint. This event can be inter- preted as a Christianized form of a pre-Christian ritual celebration of the lineage an- cestor and provided the extant patrilineage with a sacred, religious identity (Kaser 1993: 93-122; Todorova 1993a: 123-129). This sacred identity was also reinforced by PATRIARCHY IN THE BALKANS: TEMPORAL AND CROSS-CULTURAL APPROACHES 59 the fact that a common slava could also be regarded as a reminder of distant kinship even when the actual ties had been lost (Hammel 1968: 21). Each household celebrated this event independently and invited affinal kin. This ritual reinforcement of both agnatic and affinal ties on the household level is a vital part of symbolizing the way in which these societies were constituted. Affines are here symbolized as honoured guests and by their presence reinforce the solidarity of the agnatic group. This festival also provides a way to reach out to non-kin as guests, in- cluding them in the interactive circle of primary social relationships (Halpern 1967; Halpern & Kerewsky-Halpern 1986: 110). If we consider the relationship to household head by lineage for Orasac for 1863 and 1961 there are a number of important factors which reflect the difference between achieved reality and the ideal prototype. In two lineages in 1863 (composed of 30 and 31 individuals respectively) there are no grandchildren.

THE PATRIARCHAL HOUSEHOLD

The Balkan extended family, often incorrectly generalized as the zadruga (see Todo- rova 1990), is considered here as a special family pattern that came into existence autonomously and differs from other, albeit related, East European complex family structures formed by feudal institutions and, in some cases, focused on the mainte- nance of imperial frontiers. The geographical distribution of this Balkan family pattern is similar to the patriar- chal pattern. The patriarchal Balkan family pattern is characterized by: (1) vertically and/or horizontally extended family-household units with emphasis on shared econo- my and physical security. This structure was influenced by initial high fertility and high mortality; (2) members of the household act with the subsequent shift to a re- duced fertility and increased longevity as a corporate economic unit - working, own- ing, and consuming cooperatively; (3) social system embedded in clearly defined pat- rilineal and patrilocal structures originating in Post-Roman Balkan kingdoms; (4) the prototypical environment was both mountainous and pastoral, an historical adaptation to the Ottoman conquest. Thus patrilineal complexity, corporateness and autonomy are the cornerstones of the archetypical pattern of the patriarchal Balkan family in areas where centralized bureaucratic control was weak or absent. Croats, Serbs, Bulgarians, Macedonians, Albanians, Greeks, Vlachs, Karakacani and Sarakatsani - peoples of different origins and religions - all shared this pattern. Among some groups people of different religions would live in the same household as in this century among Albanians in Kosovo (Markovic 1974). Concerning marital patterns before and after the demographic transition (since the second half of the 19th century with a peak after World War II), this pattern was characterized by both early and universal marriage. Northern Greece marks one transition into a Mediterranean system; Croatia, another point, into a West European pattern (Kaser & Halpern 1994: 108-109); while the classic zadruga and associated lineage patterns occurred in an intermediate area. 60 JOEL M. HALPERN, KARL KASER AND RICHARD A. WAGNER

In analysing the Balkan patriarchal family one must be constantly aware of the ide- ologically ideal pattern as well as the achieved realities revealed in field investigations and census lists. The system operated under cyclical rules of household formation and dissolution, but these cycles altered in response to altering patterns of modernisation. The ideal prototype with its associated value patterns enjoyed greater stability than the alternating everyday life realities. The ideal structure included co-resident married siblings, early age at marriage along with an articulated differentiation of agricultural tasks and household administration (among males) also with household work cared out among women, markedly different from most West or Central European house- hold types. This pattern was useful for providing a supply of soldiers as in the case of the Habsburg military frontier settlements (Kaser 1992a: 31-32) facing those areas under direct Ottoman control. The achieved reality is well illustrated in some of the first complete census records for the Balkans as in the Serbian state census for 1863. These records show that at any one time a significant portion of the population lived in nuclear family households. This also meant that, in terms of individual life experiences, much of the population experienced life in a nuclear family household. During the course of an individual's life, membership in this more restricted kin group was a common experience for many if not most of the population. Further, figures for average household size can be mis- leading. Many of these households were large because of the number of children. Thus it was often that household size was not due to household complexity. Fig. 2 shows the household typology of Orasac in 1863 according the Laslett model (for a detailed comparison of villages in Central Serbia for 1863 see Halpern 1972).

Figure 2: Household Typology, OraSac 1863

50

40

30

20

10

Unknown Solitary Incomplete Nuclear Extended Multiple

Source: Balkan Family Project, University of Graz.

This quantitative analysis can be supplemented with an example of individual life- course experience. This approach enables us to examine what proportion of the life experience of a given individual was gained while residing in a nuclear family house- PATRIARCHY IN THE BALKANS: TEMPORAL AND CROSS-CULTURAL APPROACHES 61 hold or an extended family household in its various forms. Such data can to a degree be reconstructed from records but is best, of course, obtained through interviews. One such example is analysed in Halpern and Kerewsky-Halpern (1986: 36-38) for a man born in 1894 (based on interviews conducted in the early 1960s). His natal household was that of a large laterally extended zadruga of three generations (i.e., three component nuclear households of approximately eighteen individuals). The unit persisted until about 1909 when there was a division into constituent nuclear family units after the death of the household elder. The exact number in the household at any given time varied due to births and deaths of minor children, a large proportion of which did not reach maturity. Ego's resulting nuclear family persisted for approxi- mately two years (1909-1911). Then his mother died and two years later his sister. Two years later, the father having remained a widower, ego and his brother married before going off to fight in the Balkan Wars. The brothers and their growing families remained together after the World War I until 1937 when they divided. Ego then be- came part of his own linearly extended household composed of four generations which lasted until his father's death in 1945 when ego succeeded to headship. He headed a more truncated four-generation household until his death in 1967. By that time the household had been greatly diminished by the urban migration of his grandchildren. Thus in the course of his life Ego went through many if not all of the typologies identified by Laslett. Micro-historical demographic data alone, even though they pro- vide much more information about family processes than macro-analyses, still requires specific ethnographic case studies for a full understanding of complex social process- es. In discussing patriarchy, its associated structural forms, and their historical devel- opment we cannot fully understand the nature of this institution until we begin to con- nect data derived from both historical records and reconstructed household structures and relationships. Such analyses seem to constitute a necessary prerequisite to study- ing the behaviours manifested within these structures.

SUMMARY

This article analysed the cultural background and functioning of Balkan patriarchy. This has been a twofold process. On the one hand, the patriarchal ideology shaped kinship and family patterns. On the other, these patterns reproduced patriarchal struc- tures, but the reproduction of this system was constrained by economic circumstances. The full understanding of this process is dependent on historical and anthropological perspectives usefully implemented by the collaboration of the family historian and the sociocultural anthropologist. The formal structure of the family as given in archival census documents and parish registers does not provide sufficient context. Field ob- servations, where pertinent, and autobiographical accounts supplement archival case history data. The Croatian and Serbian complex households in their diverse manifestation are well documented in this issue by Grandits, Gruber, Mitterauer and Vekaric. They dis- cuss the diverse systems of the military and civil Croatia and include an ideologically analogous example in Central Serbia. This limited comparison must be seen within the full range of regional complexities. 62 JOEL M. HALPERN, KARL KASER AND RICHARD A. WAGNER

We need to stress here that in evaluating Balkan patriarchy one needs to take into account both the classic tribal- and lineage-based structures in Albania, Montenegro and Herzegovina and those cultures founded by immigrants from these regions (as in Orasac where ancestor worship was celebrated), as well as those regions longest in close contact with urban societies on the Dalmatian coast where certain key elements of the Balkan patriarchal order were present but in attenuated form. Because of the complexities of ecological and economic differences and varying cultural, historical and economic factors, pertinent generalizations need to be strictly qualified. Above all, we need to continually return to focus on the individual life course for we must understand how family structures are conceptualized by the individual. For these ab- stract structures have never existed apart from the individual biographies which have given social process form and meaning.

ACKNOWLEDGEMENTS

The research project 'Patriarchal Social Structures in the Balkans' is located at the Centre for Southeast European Studies at the Institute for History at the Karl- Franzens-University Graz/Austria. This project uses Joel M. Halpern's anthropologi- cal and demographic Balkan data. The Austrian Science Fund provides financial sup- port.

REFERENCES

Boehm, Christopher. 1984. Blood Revenge. The of Feuding in Montenegro and Other Tribal Societies. Kansas. Botev, Nikolai. 1988. "Features of the Fertility Decline in the Balkan Countries Since the End of the XIX Century." Etudes Balkaniques 24: 87-98. . 1990. "Nuptuality in the Course of the Demographic Transition: The Experiences of the Bal- kan Countries." Population Studies 44: 107-126. Capo, Jasna. 1990. Economic and Demographic History of Peasant Households on a Croatian Estate, 1756-1848. Dissertation, University of California, Berkeley. Cvijic, Jovan. 1992. Balkansko poluostrvo. Belgrade. Djurdjev, Branislav. 1954. "Iz istorije Crne Gore, brdskih i malisorskih plemena." Radovi Naucnog druStva Bosne i Hercegovine 2: 165-220. . 1965-66. "Postanak brdskih, crnogorskih i hercegovackih plemena". Zgodovinski casopis 19/20: 187-195. . 1992. "Prevalence of the Formal Zadruga in the Village of Vilovo, Vojvodina, Serbia". Jour- nal of Family History 17: 319-329. Drobnjakovic, Borivoje. 1923. Jasenica, Antropogeografska ispitivanja. Belgrade. Durham, Mary E. 1928. Some Tribal Origins, Laws and Customs of the Balkans. London. Filipovic, Milenko S. 1976. "Zadruga (Kucna Zadruga)". Pp. 268-279 in Communal Families in the Balkans: The Zadruga, edited by Robert F. Byrnes. South Bend: University of Notre Dame Press. PATRIARCHY IN THE BALKANS: TEMPORAL AND CROSS-CULTURAL APPROACHES 63

. 1982. Among the People, Selected Writings, edited by Eugene A. Hammel et al. Ann Arbour: Papers in Slavic Philology 3, Department of Slavic Languages, University of Michigan. I lahn. Johann G. 1854. Albanesische Studien. Jena: Friedrich Mauke. Halpern, Joel M. 1956. Social and Cultural Change in a Serbian Village. New Haven: Human Relations Area Files. . 1967. A Serbian Village. New York: Harper and Row. . 1972. 'Town and Countryside in Serbia in the Nineteenth Century: Social and Household Structure as Reflected in the Census of 1863". Pp. 401-427 in Household and Family in Past Time, ed- ited by P. Laslett and R. Wall. Cambridge: Cambridge University Press. Halpern, Joel M. and David Anderson. 1970. "The Zadruga, A Century of Change". Anthropologica N.S.12: 83-97. Halpern, Joel M. and Eugene A. Hammel. 1969. "Observations on the Intellectual History of Ethnology and Other Social Sciences in Yugoslavia". Comparative Studies in Society and History 11: 17-26. Halpern, Joel M. and Barbara Kerewsky-Halpern. 1980. "Yugoslav Oral Genealogies and Official Rec- ords: An Approach to Their Combined Use". Salt Lake City, World Conference on Records, Utah Ge- nealogical Society 1, ser.526: 175-204. . 1986. A Serbian Village in Historical Perspective. Prospect Heights: Waveland Press. Halpern, Joel M. and David A. Kideckel. 1983. "Anthropology of Eastern Europe". Annual Review of Anthropology, 12: 377-402. Hammel, Eugene A. 1968. Alternative Social Structures and Ritual Relations in the Balkans. Englewood Cliffs: Prentice-Hall. . 1972. "The Zadruga as Process." Household and Family in Past Time, edited by P. Laslett and R. Wall. Cambridge: Cambridge University Press: 335-373. . 1980. "Household Structure in 14th Century Macedonia." Journal of Family History 5:242- 273. Hasluck, Margaret. 1954. The Unwritten Law in Albania. Cambridge: Cambridge University Press. Kanuni i Lek Dukagjinit. 1989. Mbledhur dhe kodifikuar nga Shtjefen K. Gjeovi, edited by Akademia e Shkencave, Tirana. Kaser, Karl. 1992a. "The Origins of Balkan Patriarchy." Modem Greek Studies Yearbook 8:1-39. . Hirten, Helden, Stammeskampfer. Ursprung und Gegenwart des balkanischen Patriarchats. Vienna: Bdhlau. . 1993. "Ahnenkult und Patriarchalismus auf dem Balkan." Historische Anthropologie 1:93- 122. . 1994. "The Balkan Joint Family Household: Seeking Its Origins." Continuity and Change 9: 45-68. . 7995. Familie und Verwandtschaft auf dem Balkan. Analyse einer untergehenden Kultur. Vienna: Bohlau. Kaser, Karl and Joel M. Halpern. 1994. "Contemporary Research on the Balkan Family, Anthropological and Historical Approaches." Septieme Congres International d'Etudes du Sud-est Europeen (Thessalo- nique, 29 aout - 4 septembre 1994), Rapports, Athens. Kerwesky-Halpern, Barbara. 1981. "Genealogy as Genre in Rural Serbia." Pp. 301-321 in Oral Tradi- tional Literatures (Festschrift in honor of Albert B. Lord), edited by John Foley Columbus: Slavica Publications. Kerwesky-Halpern, Barbara and Joel M. Halpern, eds., 1977. Selected Papers on a Serbian Village: Social Structures as Reflected by History, Demography, and Oral Tradition. University of Massachu- setts, Department of Anthropology, Research Report No. 17. 64 JOEL M. HALPERN, KARL KASER AND RICHARD A. WAGNER

Mitterauer, Michael. 1980. "Komplexe Familienformen in sozialhistorischer Sicht." Ethnologia Europaea 12:47-86. Markovic, Milovan. 1974. "Relativno duze odrzavanje porodidnih zadruga u Albanaca na Kosovu." Soci- ologijaSela 12:95-100. Mosely, Philip E. 1976a. "The Distribution of the Zadruga Within Southeastern Europe." Pp 58-69 in Communal Families in the Balkans: The Zadruga, edited by Robert F. Byrnes Notre Dame: University of Notre Dame Press. . 1976b. "Adaption for Survival." Pp. 31-57 in Communal Families in the Balkans: The Zadruga, edited by Robert F. Byrnes. Notre Dame: University of Notre Dame Press. Pulaha, Selami. 1975. "Mbi gjallrimin e lidhjeve farefisnore dhe krijmin e fiseve ne Shqiprin e veriut ne shekujt XV-XVI." Studime Historike: 121-145. . 1976. "Formation des regions de self-government dans les Malessies du sandjak de Shkodra aux XV-XVIIe siecles". Studio Albanica:ll3-n9. Sanders, Irwin T. 1949. Balkan Village. Lexington: University of Kentucky Press. Simic, Andrei. 1991. "Obstacles to the Development of a Yugoslav National Consciousness: Ethnic Iden- tity and Folk Culture in the Balkans." Journal of Mediterranean Studies 1:18-36. Stoinanovich, Traian. 1980. "Family and Household in the Western Balkans 1500-1870". Pp. 189-203 in Memorial Lutfi Barkan. Bibliotheque de I'institut Franfaise d 'etudes Anatoliennes d 'Istanbul XXV-III. Paris. Todorova, Maria. 1983. "Population Structure, Marriage Patterns. Family and Household (According to Ottoman Documentary Material from North-Eastern Bulgaria in the 60s of the 19th Century)". Etudes Balkaniques 19: 59-72. . 1989. "Recent Research on Household and Family in the Balkans (15-19th Century)." Pp. 11-22 in Von der Pruth-Ebene bis mm Gipfel des Ida. Festschrift zum 70. Geburtstag von Emanuel Tur- czynski, edited by Gerhard Grimm. Munich. . 1990. "Myth-Making in European Family History: The Zadruga Revisited." East European Politics and Societies 4: 30-76. . 1993. Balkan Familv Structure and the European Pattern: Demographic Developments in Otto- man Bulgaria. Washington DC: American University Press. . 1993a. "Slava und Zadruga". Historische Anthropologie 1:123-129. Tomic, Svetozar. 1902. "Drobnjak. Antropogeografska ispitivanja". Srpski Etnografski zbornik 4: 357- 497. Vucinich, Wayne. 1976. "A Zadruga in Bileca Rudine." Pp. 162-186 in Communal Families in the Bal- kans: The Zadruga, edited by Robert F. Byrnes. Notre Dame: University of Notre Dame Press. . 1975. A Study in Social Survival: Katun in the Bileca Rudine. Denver: University of Denver. Vukmanovic, Cedomir. 1971. "Kvalitet statistiCkih podataka o vitalnim dogadjajima" StanovniStva 9: 57- 74. Wagner, Richard A. 1983. "Different Views of Historical Reality: Oral and Written Recollections in a Serbian village". Southeastern Europe 10: 75-188. . 1992. Children and Change in OraSac, 1870-1975: A Serbian Perspective on the Fertility De- cline. University of Massachusetts, Program in Soviet and East European Studies, Occasional Paper Series. No. 22, Amherst. Whitaker, Ian. 1976. "Familial Roles in the Extended Patrilineal Kin Group in Northern Albania." Pp. 195-203 in Mediterranean Family Structures, edited by J.G. Peristiany. Cambridge: Cambridge Univer- sity Press. . 1981. "A Sack for Carrying Things: The Traditional Role of Women in Northern Albanian Soci- ety". Anthropological Quarterly 54: 146-156.