Seasonality of Marriages in the Sajkaska Region (North Serbia),1869 to 2011

Abstract

This article studies monthly distribution of marriages in Sajkaska region (North Serbia) for the period 1869 to 2011. The data were derived from 64,175 marriages that are found in marriage registers. For the purpose of the analysis, the entire period is divided into seven smaller periods. The main finding of this article is that seasonality of marriage changed along with the system of production. Also, adherence of religion played an important role in the past. At the beginning of the period of our analysis, the region was inhabited by families practicing an inefficient extensive agriculture, and seasonality of marriage was highly present: in only two months (November and October), more than a half of all marriages were concluded. In the second part of the twentieth century, the share of agricultural population decreased and the seasonal pattern of marriage changed.

Keywords

Serbia Sajkaska region historical demography marriages seasonality

Introduction

Marriage is the event in which biological and sociocultural components interact, influencing the evolution of a population’s genetic patrimony.¹ Also, marriage occurs as the need of spouses to create family and it is socially and economically influenced.² On the other hand, the wedding period of the year is influenced by various factors. Natural conditions, mainly in agricultural areas, were among key factors impacting the date of wedding ceremony in the past. Moreover, religious practice was a crucial factor in seasonality of marriage due to restrictions to marriages imposed by religious holidays and periods preceding religious holidays. The study of marriage is important since it is frequently the prerequisite for population reproduction and a means to find out about customs and culture of a nation. Similarly, reconstruction of marriage patterns within isolated communities may usefully indicate the level of isolation and provide better understanding of biological and genetic structure of human population.³

Therefore, marriage is studied using crude marriage rate (CMR), age-specific marriage rate, geographical origin of spouses, social status of spouses,⁴ relation between education and marriage,⁵ education-related marriage and fertility,⁶ as well as education and age at first marriage.⁷ However, seasonality of marriage has been increasingly observed lately as well as the factors influencing it. Certain research data connect seasonality of marriage with seasonality of births.^8,9

The main objective of this article is to present seasonality of marriage pattern in Sajkaska region and detect seasonality changes for the period 1869 to 2011. Sajkaska region is located in the north part of Serbia (Vojvodina Province; Figure 1), and it comprised fourteen villages with more than 60,000 inhabitants. Today, the region does not represent any administrative entity (only a geographical region) and is divided among four municipalities. This region has important historical background (Figure 2).¹⁰ Historical circumstances provided that the Sajkaska region over the last one and a half centuries was a part of a number of states, which were distinctive, both in the size of their territories and in the form of their system of government. The region was created by the Austrians in 1763 as a part of Military border to serve against Turkish advance until the year 1873. The name of Sajkaska region means “the land of sajkasi.” Sajkasi was a special military unit of Austrian army. They used long, narrow boats, known as “sajka.”

Figure 1.

Geographical location of Sajkaska region.

Figure 2.

Historical background of Sajkaska region. Source: After Hammel, (Military border in 1815).

Theoretical Background

Seasonality of marriage is of interest for researchers because it is controlled by different environmental conditions, geographical locations, socioeconomic, demographic, and historical background (such as different legislation regulating the property relations between spouses). Fast factor related to legislation acts do not have big influence on seasonality marriage but has important role in marriage due to different property relations. For example, up until the World War II (WWII), the Hungarian law, that is, the Law Chapter XXXI from 1894 governed the property relations between spouses in Vojvodina. This regulation allowed for entering into marital property, that is, marital agreements. Under these agreements, spouses were able to modify or completely exclude legal property regime. The common property regime was a legal regime, except in the case of nobility and persons with university education engaged in a lucrative profession, in which case the system of separation of property applied. The freedom of contract was such that spouses could “temporarily or provisionally exclude the application of common property rules. Moreover, the parties were able to revoke the established property community at any time upon mutual agreement.”¹¹ Family legislation in Serbia, until after the WWII (until the adoption of the Family Law of 2005), did not allow conclusion of marital property agreements, that is, any modifications to the legal property regime. Only the new Family Law of the Republic of Serbia of 2005 reinstated the institute of marital agreement into positive legislation (in Serbia), but retained, at the same time, those provisions which stipulated the right of spouses to regulate their property relations within the legal common property regime.

Several papers have analyzed seasonality of marriage in different communities and regions. Sanna and Danubio¹² researched seasonality of marriages in four pastoral and four agricultural villages in Sardinia during nineteenth century. As a result, they concluded that marriage seasonality in these communities followed monthly distribution of marriage pattern similarly to the communities in other parts of Italy as well as in Europe. Similar study was carried out by Coppa et al.¹³ They researched seasonality of marriage in ecological contexts in rural area of Central-Southern Italy regarding monthly distribution of marriages during four centuries (1500–1871) and found similar seasonality pattern of marriages to agricultural patterns in France and Spain. Also, this study showed that religious factors strongly affected the timing of weddings only during Lent, and monthly distribution of recorded marriages was subject to change over time. Seasonal variation in marriages was also recorded in Andora. Since 1940 traditional system of production has been replaced by service industries and such changes have influenced marriage seasonality.¹⁴ Dribe and Van de Putte¹⁵ studied marriage seasonality related to industrial revolution in southern Sweden. Their analysis showed that seasonality of marriage pattern seriously changed in 1685 to 1894 which was a period of transition from agricultural to early stage of industrialization. Impact of economic and religious factors on marriage was recorded by Broderick¹⁶ researching marriage in Trinity Bay, Newfoundland. Furthermore, he perceived that the time of wedding for fishing communities differed from the rest of the population. Cressy¹⁷ explored seasonality of marriages in New and Old England in the period from 1540 until 1799. He found seasonal fluctuation of marriages with peak in autumn and early summer, and he concluded that religious factor has strongest effect of timing of marriage in France than in England. Seasonality changes of marriage were also analyzed in Cerdanya Valley (divided by Franco-Spanish border), and results show divergence over time between France and Spanish patterns of marriages. The months with most clearly difference are March, May, July, and August; and in these four months, higher number of marriages were recorder in French Serdanya than in Spanish Serdanya. This divergence is the result of cultural differentiation between populations which once were homogeneous.¹⁸

Data and Method

According to the 2011 census, Sajkaska region had about 64,696 inhabitants. From the beginning of the observed period, demographic development of Sajkaska region was accompanied with both population growth and decline. During the first decade of the twenty-first century, the number of total population started to decline, the main factor for population change being low fertility rate.

The analysis of marriage seasonality comprised 64,175 recorded marriages in the span of 142 years. Monthly data were collected from both parish and civil registers: 8,835 marriages were recorded in parishes from 1869 to 1895 (until October) and 55,861 marriages were from civil registers in the period from 1895 (from November) to 2011. Absolute number of marriages as well as CMR has been analyzed. CMR was calculated on monthly level as the number of marriages recorded in a given month per 1,000 inhabitants in the same month. In order to research temporal changes, the data were split into seven periods (Table 1). Henry’s seasonality coefficient (H) was used for measuring seasonality. Henry’s coefficient is most frequently used for the study of seasonal changes of marriages and it is calculated according to the following equation (e.g., in the period 1989–2011):

Table 1.

Descriptive Statistics of Sajkaska Region.

Period	Population (Average Number)^a	Marriages
Period	Population (Average Number)^a	n	Number of Recorded Marriages	Mean	CV
1869–1888	37,870	20	6,648	332	22.4
1889–1908	52,240	20	9,079	454	21.0
1909–1928	58,683	20	11,633	582	39.5
1929–1948	56,371	20	9,456	473	30.6
1949–1968	55,765	20	10,213	511	9.4
1969–1988	60,534	20	9,447	472	9.1
1989–2011	64,718	23	7,699	335	10.7

Source: Number of total population: Slobodan Ćurčić, Naselja Bačke, geografske karakteristike (Novi Sad, Serbia: Matica Srpska, 2007); A KereskedelemÜgyi magyar kir. Miniszter rendeletéből & A magyar kir. Központi statisztikai hivatal, Magyar statisztikai közlemények 1900: Népszámlálása, a népesséeg általános leirása községenkint, 1 kötet (Budapest, Hungary: Pesti könyvnyomda-részvénytársaság, 1902); A KereskedelemÜgyi magyar kir. Miniszter rendeletéből & A magyar kir. Központi statisztikai hivatal, Magyar statisztikai közlemények 1910: Népszámlálása, a népesség főbb adatai községek és népesebb puszták, telepek szerint, 42 kötet (Budapest, Hungary: Az athenaeum irodalmi és nyomdai r.-társulat nyomása, 1912); Popis stanovništva, domaćinstava i stanova u 2011. godine u Republici Srbiji, Uporedni pregled broja stanovnika 1948-2011, knjiga 20. Republički zavod za statistiku (Beograd, Srbija: 2014); Protokol venčanih 1869-oktobar 1895 (Parish register of marriages 1869-October 1895); Matične knjige venčanih novembar 1895-2011 (Civil register of marriages November 1895-2011).

Note: n = number of observed years; CV = coefficient of variation is calculated using total number of marriages in year.

^aAverage number of population is estimated using linear interpolation between censuses.

H_{m, 1989 - 2011} = (\frac{\sum N_{m, t, 1989 - 2011}}{\sum_{1}^{12} N_{m, t, 1989 - 2011}}) \times 1, 200,

where N_m is the number of recorded marriages in a given month. The number of marriages was expressed as a total value of 1,200 over the year, corrected according to the number of days in each month (28.5 in February), so the expected number of marriages in a given month would be 100 (it implies the lack of seasonality).^19

–22 Cluster analysis (k-means clustering) was applied for detecting differences or similarity between months. Cluster analysis is the convenient statistical method for identifying groups of observations that are very similar, and k-means algorithm had recognized ability to analyze large data set.

Table 1 shows mean population number, mean value, and coefficient of variation (CV) for the analyzed number of registered marriages in the observed periods. Based on the CV, it may be confirmed that the data of the number of marriages are significant, because its value is under 50.0 percent for the entire period.

Results and Discussion

Decreased Trend of Marriages

CMR oscillated over time. Mean CMR in the second half of the nineteenth century was 8.3 per mille, with maximum values recorded for the years 1872 (12.8 per mille) and 1879 (11.9 per mille). In the first half of the twentieth century, CMR was 9.3 per mille. This period also covers the period of the World War I and WWII, when maximum values were recorded exactly in the first postwar years (1919 [20.6 per mille], 1920 [16.2 per mille], 1946 [14.7 per mille], 1947 [12.2 per mille], and 1948 [13.4 per mille]) as shown in Figure 3.

Figure 3.

Crude marriage rate (per mille) in Sajkaska region, 1869 to 2011.

Mean CMR in the second half of the twentieth century and in the first decade of the twenty-first century was 7.2 per mille. Maximum values were recorded in the 1950s, that is, in 1952 (10.9 per mille) and 1960 (10.4 per mille). Conversely, minimum values occurred during war time, for example, in 1915 CMR was only 2.1 per mille, and then in 1916 it was 3.7 per mille. Lower CMRs were registered for 1942 (4.7 per mille), 1944 (4.4 per mille), and 1945 (4.0 per mille). Low CMRs (under 6 per mille) were also recorded for the first decade of the twenty-first century.

The analysis for this period indicated that since mid-1980s the CMR remained under 7 per mille. In the first decade of the twenty-first century, the rate remained under 6 per mille. Main reasons for declining CMR at the beginning of new millennium can be found in new social conception of marriage as union between two people. This new perception of marriage is reflected through highest age at first marriage (changing age timing of first marriage).

Figure 4 shows mean monthly values of CMR presumably without seasonality factor, that is, the absence of all factors influencing seasonality of marriage. In such case, the value of CMR would be the same for each month of the year as the graph shows. For the first two periods (1869–1888 and 1889–1908), average CMR on monthly level would be 0.7 per mille. For the third period 1909 to 1928, it would be 0.8 per mille, and in the fourth period 1929 to 1948, monthly average CMR would be 0.7 per mille. During the last three periods, theoretical value of average CMR would be in the range from 0.8 per mille (1949–1968) and 0.7 per mille (1969–1988) to 0.4 per mille (1989–2011). However, analyses in this article show that CMR varies for each month over time, as result of various social and natural factors.

Figure 4.

Average monthly crude marriage rate (per mille): theoretical assumption in the absence of seasonality in Sajkaska region.

Seasonal Patterns of Marriages in Sajkaska Region

Monthly pattern of H coefficient during the second half of the nineteenth and beginning of twentieth century reveals a model with highest values in certain winter and autumn months, exceeding 100, whereas in other months its value was below 100. The first period (1869–1888) shows maximum H coefficient in January, October, and November. Second period (1889–1908) indicated slight changes with highest value of H coefficient in February, October, and November. Next two periods (1909–1928 and 1929–1948) show similar pattern of H coefficient with maximum in February, May, and November. In the period from 1949 to 1968, highest values of H coefficient were found in February, October, and November. Next two periods (1969–1988 and 1989–2011) show quite different seasonality model. The period 1969 to 1988 shows highest coefficient in May, August, September, October, and November. Similar model (with the exception of November) was detected in the last period (1989–2011).

For cluster analysis, monthly CMRs were used and data from seven periods were observed. For each period, two clusters were formed by the analysis: cluster 1 and cluster 2. With regard to CMR, the months extracted to cluster 1 were the months with higher CMR with regard to those in cluster 2. In the first period, the months extracted to cluster 1 were those with higher CMR with regard to other months of the same year (extracted to cluster 2). The number of marriages in October and November made about 58.8 percent of the total number of marriages for the period 1869 to 1988. February and November were extracted into cluster 1 for the second period due to the highest CMR, whereas other months were grouped into cluster 2. The number of marriages for these two months was 35.3 percent of the total marriages for the period 1889 to 1908. For the third period, from 1909 to 1928, the analysis extracted February, May, and November into cluster 1. Higher CMR was recorded for these months compared to other months (cluster 2) and about 41.9 percent of the total number of marriages for this period. During the fourth period (1929–1948), February and November were grouped in cluster 1 with about 29 percent of total recorded marriages. In the period from 1949 to 1968, February, September, October, and November were gathered in cluster 1, and in these four months, about 41 percent of total marriages were celebrated. In sixth time span from 1969 to 1988, May, August, September, October, and November were extracted to cluster 1 with 52.3 percent of total marriages. In the seventh period (1969–2011), both clusters gathered the same number of months. May, June, July, August, September, and October were grouped in cluster 1 with 66.1 percent of celebrated marriages. Other six month have created cluster 2 (Table 3). CMR oscillations within cluster 1 and cluster 2 are illustrated in Figure 5, which clearly indicates the difference between mean CMR values for the clusters and the trend of marriage pattern variations between months in a cluster. Furthermore, Figure 5 clearly indicates the trend of CMR oscillations between defined clusters. During the first two observed period (second half of the nineteenth and beginning of twentieth century), the CMR within cluster 1 shows growth and oscillations with time, whereas the CMR within cluster 2 shows almost linear trend, without high oscillations. In the next five periods, CMR for months grouped into cluster 1 shows higher oscillations over time, whereas CMR for months within cluster 2 shows smaller changes in the observed period, with the exception of war period, when there was a decrease in CMR. Conversely, the last three periods (1949–1968, 1969–1988, and 1989–2011) show the similar linear decreasing trend of mean CMR for both clusters.

Table 2.

Distribution of H Coefficient by Period in Sajkaska Region.

Month	1869–1888	1889–1908	1909–1928	H Coefficient 1929–1948	1949–1968	1969–1988	1989–2011
January	129.24	114.86	95.32	80.20	92.59	78.50	53.46
February	95.67	167.20	185.06	164.09	113.15	68.34	60.48
March	2.35	60.14	81.39	93.40	86.83	75.58	57.98
April	32.31	48.38	58.80	76.52	86.95	86.25	72.48
May	68.23	104.42	127.29	111.80	100.58	124.48	129.99
June	34.12	66.22	85.10	72.97	76.49	86.12	102.40
July	33.39	67.01	85.31	73.35	76.73	86.76	104.27
August	39.17	57.89	63.03	80.58	88.24	134.14	154.62
September	83.75	80.89	85.00	95.30	107.39	129.95	184.39
October	249.64	123.98	95.42	109.90	121.37	124.10	117.21
November	432.13	256.55	190.84	178.81	149.93	115.47	94.45
December	0.00	52.47	47.45	63.07	99.76	90.31	68.27

Table 3.

Results of Cluster Analysis.

Month	1869–1888	1889–1908	1909–1928	1929–1948	1949–1968	1969–1988	1989–2011
January	2	2	2	2	2	2	2
February	2	1	1	1	1	2	2
March	2	2	2	2	2	2	2
April	2	2	2	2	2	2	2
May	2	2	1	2	2	1	1
June	2	2	2	2	2	2	1
July	2	2	2	2	2	2	1
August	2	2	2	2	2	1	1
September	2	2	2	2	1	1	1
October	1	2	2	2	1	1	1
November	1	1	1	1	1	1	2
December	2	2	2	2	2	2	2

Figure 5.

Trends in mean value of crude marriage rate for each cluster.

Although the results of cluster analysis and H coefficient may not be completely compared since H coefficient indicates the deviation in the number of marriages per each month, based on absolute value of marriages in a particular month with regard to the number of marriages during one or more years, cluster analysis classifies a certain phenomenon (in this case CMR) with regard to its characteristics in each month of the year.

The timing of marriage peak, as well as marriage frequency, varied over time. During the first three periods, first minimum varied from March to April and depends from time of Easter and second minimum was in December. In the fourth period, minimum was in December, and the period from 1949 to 1968 shows minimum in June. In last two periods, minimum celebrated marriages were in winter months: February (1969–1988) and January (1989–2011).

The low average CMR during the whole observed period was also found in March and April as well as in summer months: June, July, and August (except in the last period; Figure 6). The analysis of CMR pointed out that for the period 1869 to 1888 not a single marriage occurred in December. Similar situation was observed for March within the same period, when only a few marriages occurred.

Figure 6.

Average monthly crude marriage rate (CMR per mille) in Sajkaska region.

Changes in seasonal pattern of marriages over time can be related to religious factors, economic activities (mostly in agriculture), and lifestyles. During first three periods, the religious factors had the strongest effects on marriage seasonality, the minimum number of recorded marriages was found during Lent and Advent. Also, the effect of moving holiday (Easter) had influence on CMR in February (in particular year, Lent begins in this month). In this period, occupations of population also had big influence on seasonality of marriages. Most of the inhabitants of Sajkaska region worked in agriculture, and marriages usually occurred during the period of the year which is characterized by the decreased activity in the agriculture. Since the middle twentieth century, majority of the population began to move in nonagricultural sector and according to the data from the beginning of twenty-first century, less than 30 percent of economically active population of Sajkaska region were agricultural workers,²³ which had the impact on the shift in seasonal pattern of marriages. The last two periods (1969–1988 and 1989–2011) showed significant changes compared to previous time span, pointing to increasing value of H coefficient in August (Table 2), which indicated the increase in the number of marriages that may have been related to summer holidays. Presently, most of the couples plan their marriage ceremony during holidays because significant role of the wedding date selection is assigned to the presence of relatives and friends at the ceremony. As part of this change, marriage become less a community celebration event, but affirmation of shared values and of personal and familial ceremony.²⁴

Similar seasonal changes in marriage rate pattern were recorded in research of other authors. Seasonal changes in marriage rate in Abruzzo region (Italy) for the period 1500 to 1871 indicated that the highest frequency of marriage occurred in November, and the lowest in March, April, July, and August.²⁵ Seasonal changes in CMR during the nineteenth century were studied in agricultural and pastoral communities in Sardinia. Maximum values of marriage numbers in agricultural communities occurred in February and in autumn months, whereas maximum values of marriage numbers in pastoral communities occurred in the period July to November. Marriage research in Andora pointed out that during the seventeenth century, most of the marriages occurred in February and October. In the eighteenth and nineteenth century, there was a shift in seasonal marriage pattern: maximum number of marriages occurred in summer months (June and July) and minimum number in March.²⁶ The results of seasonality of marriage in England and the Netherlands indicated that the key factor of shift in seasonality of marriage was the change in economic activity of the population.^27,28

Changes in seasonal pattern of marriages in the listed regions may not be completely compared with the perceived changes in Sajkaska region, since the same periods of time are not covered, but it may be concluded that religion had huge impact on seasonality of marriage in all the regions, regardless of their geographical position up to the twentieth century. During the twentieth century and first decade of twenty-first century, influence of religious customs on seasonality of marriage decreases and the system of production increases.

Conclusion

The analysis proved a varied trend of CMR. From the 1869 until middle of twentieth century, with slight fluctuation, CMR increased. The decrease in marriage trend dates back to the mid-1950s. Another striking feature of population marriage pattern in Sajkaska region is seasonality of marriage, with a changing trend of seasonality patterns over time. In the second half of the nineteenth and the first half of the twentieth century, seasonality of marriage in Sajkaska region follows the seasonality of marriage pattern of agricultural communities, whereas from the mid-twentieth century, seasonal pattern gradually changes. Based on the changes in seasonality of marriage pattern, two models of marriage are observed in Sajkaska region:

First marriage model, typical for traditional agricultural societies, where religion and agricultural work highly influence the annual marriage dynamics.

Second marriage model, the feature of modern societies where apart from the economic factor, seasonality of marriage is caused by numerous modern trends and lifestyles.

According to results of cluster analysis and H coefficient, it can be concluded that a reduced visibility of marriage seasonality is followed by the lower and lower number of marriages concluded. Reducing impact of religious factors, timing of marriage during the year became an individual choice and depends only on personal preferences and aspirations. In the course of modernization, cohabitation and postponement of marriages will be more and more frequent events during the twenty-first century. That probably implies further changes in marriage behavior.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This article is part of the project No. 114-451-1135/2014-01 funded by the Provincial Secretariat for Science and Technological Development of Vojvodina Province, the Republic of Serbia.