Strategy or Necessity? The Determinants of Marriage Between Relatives in the Barcelona Area, Sixteenth–Nineteenth Centuries

ABSTRACT

Consanguineous marriages illustrate the complex interplay between environmental and social factors in family formation. However, there remains a significant lack of quantitative studies exploring this phenomenon in preindustrial contexts. In this sense, this study examines the intensity and the structural and strategic determinants of consanguineous marriages in the Barcelona area between the 16th and 19th centuries, utilizing the unique Barcelona Historical Marriage Database. Methodologically, the isonymy method (spouses with matching surnames) was used to estimate the level of consanguinity (structural and strategic), while binary logistic regression models were constructed to analyze the social and territorial patterns of these marriages. The main findings indicate that consanguinity in the Barcelona area remained limited until the 19th century, mirroring European trends, though to a lesser degree. The nobility strategically adopted consanguineous marriages from the early modern period, while the peasantry largely avoided them, relying instead on primogeniture to preserve family assets. By the 19th century, the emerging bourgeoisie increasingly turned to consanguinity as a means of strengthening social networks and establishing class identity rather than merely preserving material heritage. Furthermore, the marriage market size and contextual factors—such as decreasing adult mortality and reduced dispensation costs—significantly influenced the rise of consanguineous unions.

1. Introduction

Marriages between relatives are a prime example of the interplay between environmental and social determinants of family formation (Bestard Camps ¹⁹⁹⁸; Goody ¹⁹⁸³). This type of marriage provides further evidence that people did not always choose their spouses on their own, but were rather influenced by their families, as part of the social group to which they belonged or which they forged (Bourdieu 1972; Ruiz Sastre ²⁰¹³). Marriage between kin has been documented as one of the most common strategies for creating social, economic, and even political alliances in the past (Berkner and Mendels 1978). The availability of spouses in the marriage market and the demographic regime may also have favored or hindered this type of marriage, preparing the ground for its emergence (García González ²⁰⁰⁰).

In Europe, the study of marital unions between relatives has mainly focused on consanguinity and, to a lesser extent, affinity, in order to analyze their level and intensity (Bittles et al. 1991). Quantitative research on this topic in the preindustrial era is scarce, and there are only a few qualitative case analyses of specific populations (Henarejos López ²⁰¹⁶). Instead, the vast majority of studies have focused on the 19th and 20th centuries to examine the well‐known boom in consanguinity (Fuster Siebert and Colantonio ^2002a; Moroni ¹⁹⁶⁷), which can be explained by several factors. First, the Catholic Church became more permissive in its dispensation of consanguineous marriages (Lanzinger 2018; Manzano Ledesma and Rueda Fernández ²⁰¹⁸). Second, horizontal ties in families were strengthened as part of one of the major transitions in the development of European kinship (Boudjaaba et al. 2016; Calvi and Blutrach‐Jelin ²⁰¹⁰; Sabean et al. ²⁰⁰⁷), while the number of same‐age relatives also increased as a result of the demographic transition (Serrano Migalln and Pressat ²⁰¹⁰). In this regard, research with a long‐term temporal perspective covering large geographical areas is almost nonexistent but essential to understand this phenomenon in all its complexity.

Most studies on marriage between relatives have relied on marriage dispensations that explicitly state the degree of kinship between the spouses (Fuster Siebert and Colantonio ^2002b). The use of this source alone has several drawbacks, including availability and accessibility. It also underestimates kinship by considering only information about the marriage itself, ignoring the familial effect that may go back many generations before the marriage in question (Esparza et al. 2015). Studying the prevalence of spouses with identical surnames, known as isonymy, presents a viable alternative. This approach leverages historical records containing individuals' names, which are widely available and often more accessible (Colantonio et al. 2003). Isonymy operates on the principle that surname transmission follows patterns akin to genetic inheritance, enabling the estimation of population parameters related to consanguinity (Crow and Mange 1965). Notably, isonymy yields results comparable to those derived from labor‐intensive family reconstructions, which are occasionally hindered by the scarcity of genealogical data (Esparza et al. 2015; Rodríguez‐Larralde et al. ²⁰⁰³; Rogers ¹⁹⁸⁷).

A further step in the study of marriage between relatives is to address its complexity by analyzing its determinants. Previous studies on these determinants have focused mainly on structural, geographical, and demographic aspects resulting from the isolation and limited size of the marriage market (Bittles 2001; Boëtsch et al. ²⁰⁰²; Spencer ²⁰⁰⁰). However, little attention has been paid to strategic factors such as social, cultural, and ideological influences (Manzano Ledesma ²⁰¹⁹). Few analyses have comprehensively considered the wide range of conditions that interact in the formation of marriages between relatives. Bras et al. (2009) provided a comprehensive and nuanced perspective on the study of marriage between relatives by showing how factors such as geographical isolation, migration processes, social belonging, and religious enculturation influenced different types of marriage between relatives in the Netherlands between 1840 and 1922. There is also a lack of quantitative research on the determinants of kinship marriage, as most studies are descriptive in nature and focus on very limited family and social groups. In Spain, there has been no quantitative study of the determinants of marriage as proposed in this paper.

In the specific case of Catalonia (Spain), research on marital consanguinity has been limited and has focused almost exclusively on the late 19th and 20th centuries (Esparza 2004; Font ²⁰⁰⁴). These studies have shown that marital consanguinity was of very little importance in this region compared to other parts of Spain. This is mirrored in 20th‐century Aragon and Valencia, which were also characterized by low levels of consanguinity (Pinto‐Cisternas et al. ¹⁹⁷⁹). There is also a lack of studies examining the factors that influenced the determinants of consanguineous marriage strategies in Catalonia. In this sense, this paper aims to fill this gap by investigating the structural and social determinants of consanguineous marriages in the Barcelona area over a long period, from the 16th to the 19th century. Consanguinity is analyzed through the lens of isonymy, looking at the surnames of the spouses in marriages performed in the Officiality of Barcelona, the main deanship of the Diocese of Barcelona, and recorded in the Barcelona Historical Marriage Database (BHMD). This database is a unique resource for carrying out this type of research due to its broad temporal scope, the type of data provided, and the large number of marriage records it contains.

2. The Barcelona Area, 16th–19th Centuries

In the early modern period, Catalonia underwent a demographic revival, overcoming the crisis of the late Middle Ages, marked by frequent outbreaks of disease and war. As a result, in the 16th century, Catalonia experienced an increase in population, from just over 250 000 to more than 400 000 inhabitants (Nadal Oller ¹⁹⁷⁸; Simón ¹⁹⁹²). This growth has been attributed to French immigration (Amengual‐Bibiloni and Pujadas‐Mora ²⁰¹⁷; Nadal Oller and Giralt ¹⁹⁶⁰) and economic expansion fueled by viticulture and textile manufacturing in rural areas (Junqueras Vies ²⁰⁰²; Lobato ¹⁹⁹⁵; Torras Elias ¹⁹⁹³). However, Catalonia's expansionary phase came to a halt in the early 17th century, followed by a demographic downturn in the second third of the century, linked to a severe agrarian crisis and exacerbated by periods of war and devastating epidemics (García Espuche ¹⁹⁹⁸).

Towards the end of the 17th century, Catalonia experienced a resurgence in population growth, which accelerated in the 18th century and was characterized by an increase in marriage rates, a decrease in the age at marriage, and, later, a reduction in adult mortality (Ferrer Alòs ²⁰²²). The population of Catalonia doubled between 1717 and 1787, doubled again between 1787 and 1860, and doubled yet again between 1860 and 1950. In figures, it increased from 402 531 inhabitants in 1717 to 829 615 in 1787 and 1 652 291 in 1860 (Cabré 2007). The demographic development of the Barcelona area mirrored that of Catalonia as a whole. Throughout the 18th century, the city of Barcelona experienced remarkable growth, with its population swelling from 35 000 to 130 000 inhabitants and finally exceeding 500 000 by the end of the 19th century. During this period, Barcelona became a magnet for migrants from other parts of Catalonia and developed into a major urban centre (López Guallar ²⁰⁰⁴).

The Barcelona area experienced a period of socioeconomic prosperity between the 16th and 19th centuries. In rural Catalonia, including the more rural parts of the Barcelona area, the peasantry achieved social advancement in the 16th and 17th centuries by creating medium‐sized estates, known in Catalan as masos, of which they were the beneficial owners and which they could pass on to their heirs. The system of undivided inheritance particularly favored these estates, preventing the division of wealth (Ferrer Alòs ²⁰¹⁵). The designated heir (the eldest son, who inherited the entire family estate) emerged as a central figure in the marriage strategies of almost all social classes (Serra Clota ²⁰¹⁶). Meanwhile, in urban centers, particularly the city of Barcelona, guilds played a pivotal role in stimulating economic growth throughout the 16th and 17th centuries (Molas 2017).

Furthermore, throughout the 18th century, there was a notable intensification in wine and rural cotton production in Catalonia, accompanied by the development of the textile industry known as Indianas (Badia‐Miró and Tello ²⁰¹⁴; Marfany ²⁰⁰⁶; Sánchez Suárez ²⁰¹³; Valls Junyent ²⁰²⁰). This set the stage for industrialization and the transition to factory systems and mechanization. The region's agricultural focus in the 19th century led to an uptick in the population's income, stimulating increased consumption and demand for manufactured goods. However, this socioeconomic transformation was not solely driven by agricultural advancements. Other contributing factors included population growth, which expanded the labor force; enhancements in distribution networks, which facilitated trade; and the availability of capital for new investments (de Motes and Bernet ¹⁹⁹⁸).

The industrialization of the Barcelona area began early, with a significant transformation of its productive landscape already in the 18th century, driven mainly by advances in the textile industry (Brea‐Martínez and Pujadas‐Mora ^2018b). This process of industrial transformation was particularly pronounced in the city of Barcelona, as well as in urban centres such as Terrassa, Sabadell, and Mataró, all of which experienced significant industrial growth (Benaul Berenguer ¹⁹⁹¹). By the 19th century, the Barcelona area had become one of the first industrialized regions in southern Europe (Rosés 1998; Thomson ²⁰⁰³), while also experiencing an early decline in fertility rates (Cabré 1999).

3. Materials and Methods

The Barcelona Historical Marriage Database (BHMD) contains the records of over 600 000 marriages performed in the Diocese of Barcelona between 1451 and 1905, taken from Llibres d'Esposalles [marriage license books]. It was developed as part of the European Research Council Advanced Grant project entitled Five Centuries of Marriages (grant agreement ID: 269796). These books serve as a fiscal archive, documenting the payment of a marriage tax imposed on couples wishing to marry in the Diocese of Barcelona, based on their socioeconomic status, in accordance with the edict of Pope Benedict XIII (Pedro Martínez Luna, 1328–1423). This tax was introduced in 1409 to finance the construction and subsequent maintenance of the Barcelona Cathedral (Carreras Candi ¹⁹¹⁴). The original books are kept in the Barcelona Cathedral's Capitular Archive. Records of tax payments prior to 1451 are aggregated without identification by name (Baucells Reig ²⁰⁰²).

The Diocese of Barcelona comprised the Officiality of Barcelona, which we call the Barcelona Area, and the deanships of Piera, Vallès, and Penedès (Figure  1 ). The Officiality of Barcelona accounts for more than 90% of the marriages recorded in the BHMD. This area includes urban centres such as Barcelona, Mataró, Sabadell, and Terrassa, as well as a group of rural parishes. It is still the most densely populated region in Catalonia. Meanwhile, the deanships of Piera, Vallès, and Penedès are almost entirely rural, with much smaller populations and occasional gaps in their marriage license books. This paper will therefore use data from the Officiality of Barcelona.

FIGURE 1.

Diocese of Barcelona. Location of the officiality and the deanships. Source: Authors' own elaboration.

Marriage tax payments were recorded in pounds, shillings, and pence, according to the Carolingian monetary system. However, the actual payment was made in the currency of the time. The tax imposed on marriages was dependent on the couple's socioeconomic status. From its introduction until 1575, the tax amount was calculated individually for each couple. From 1575 onwards, a system of seven tax categories was established, categorized according to the following social groups (Table 1): (1) the nobility; (2) military citizens; (3) honored citizens; (4) merchants, lawyers, and physicians; (5) guild masters; (6) farmers and small artisans; and (7) the poor, who were exempt and granted a marriage license by Amore Dei. The social class of the honored citizens is a specific feature of the region studied, consisting of urban aristocrats who could hold public posts. They achieved their status by monopolizing land rents and taking jurisdiction over the lands they controlled; by acquiring privileges, honorary positions, and titles; or by demonstrating a certain standard of living (Amelang 1986). This aristocratic class emerged in the Crown of Aragon in the late Middle Ages, coinciding with the political rise of major cities. During the period between 1649 and 1857, a separate tax category for merchants was introduced, resulting in eight tax groups. However, this additional category was discontinued in 1858, and the original seven groups were reinstated. The rates for each tax category remained largely stable over time. The only exception occurred in 1640, when all rates doubled.

TABLE 1.

Descriptive statistics.

	N	Proportion
Marriages between 1540 and 1880	454,267

Dependent variable
Isonymic marriages	2,258	0.005

Structural independent variables
Location	N	Proportion
Barcelona Historical city	155,700	0.34
Barcelona city incorporados 19th c	66,483	0.15
Barcelona area (not Barcelona city)	156,779	0.35
No Officiality	5,802	0.01
Location missing*	69,503	0.15

Population size (X1)	Proportion
Small	0.20
1540–1719 (0–1000 inhab.) Campoflorido Census 1717
1720–1799 (0–1500 inhab.) Floridablanca Census 1787
1800–1880 (0–2000 inhab.) Census 1887
Medium	0.12
1540–1719 (1000–5000 inhab.) Campoflorido Census 1717
1720–1799 (1500–7000 inhab.) Floridablanca Census 1787
1800–1880 (2000–15,000 inhab.) Census 1887
Large	0.68
1540–1719 (> 5000 inhab.) Ref. Campoflorido Census 1717
1720–1799 (> 7000 inhab.) Ref. Floridablanca Census 1787
1800–1880 (> 15,000 inhab.) Ref. Census 1887

Geographical isolation (X2)	Proportion
All territory
Coastal	0.71
Inland	0.29
Not Barcelona city
Coastal	0.30
Inland	0.70

Barcelona Parish (historical city) (X3)	Proportion
St Just/Mare de Déu de la Mercè/St Miquel	0.118
Sta Maria del Mar	0.260
St Pere de les Puelles	0.112
St Jaume	0.042
Sta Maria del Pi	0.419
Cathedral	0.004
St Cugat del Rec	0.044

Strategic independent variables
Socio‐occupational group (X4)		Proportion
Nobility		0.007
Higher managers	HISCLASS 1	0.007
Higher professionals	HISCLASS 2	0.03
Lower managers, lower professionals, clerical and sales personnel, and foremen	HISCLASS 3, 4 & 6	0.06
Lower clerical and sales personnel	HISCLASS 5	0.02
Skilled workers	HISCLASS 7	0.19
Lower‐skilled workers and unskilled workers	HISCLASS 9 & 11	0.38
Lower‐skilled farmworkers and unskilled farmworkers	HISCLASS 10 & 12	0.04
Farmers	HISCLASS 8	0.26
No occupation		0.03

Tax category (assimilated a < 1649)** (x5)
Min.	Max.	Mean	SD
0	240s	4.12s	5.378

	Proportion
Nobility and military citizens (TAX 1&2)	0.004
(1540–1649) 12£/2£ 8s
(1650–1880) 24£ /4£ 16s
Honored citizens (TAX 3)	0.01
(1540–1649) 1£ 4s
(1650–1880) 2£ 8s
Merchants, lawyers and physicians (TAX 4)	0.02
(1540–1649) 12s
(1650–1880) 1£ 4s
Guild masters (TAX 5)	0.03
(1540–1649) 6s
(1650–1880) 12s
Farmers and small artisans (TAX 6)	0.78
(1540–1649) 4s
(1650–1880) 8s
Poor (Amore Dei)	0.15
No tax	0.003

Ability to pay (X6)
Min.	Max.	Mean	SD
0	42,000	237.78	485.00

		Proportion
Very high	> 800 (relative units)	0.02
High	300–800 (relative units)	0.05
Medium‐high	220–300 (relative units)	0.10
Medium‐low	200–220 (relative units)	0.40
Low	170–200 (relative units)	0.17
Very low	0–170 (relative units)	0.06
No ability to pay (Amore Dei)	0 (relative units)	0.15
Ability unknown		0.04

Mixed independent variables
Abundance or rarity of the groom's surnames (X7)
		Proportion
1st Quartile groom's surnames	54 surnames
2nd Quartile groom's surname	198 surnames
3rd Quartile groom's surname	739 surnames
4th Quartile groom's surname	17,965 surnames
Groom's surnames missing		0.03

Abundance or rarity of the bride's surnames (X8)
		Proportion
1st Quartile bride's surnames	51 surnames
2nd Quartile bride's surnames	187 surnames
3rd Quartile bride's surnames	689 surnames
4th Quartile bride's surnames	16,506 surnames
Bride's surnames missing		0.08

Marital status (X9)
		Proportion
Single/Widowed	44,408	0.10
Widowed/Single	41,341	0.09
Widowed/Widowed	24,023	0.05
Single/Single	344,495	0.76

^*A sensitivity test on the occupational distribution was conducted to assess the impact of missing location data. No significant differences were observed between samples including and excluding these records.

^**£ = pounds; s = shillings.

Source: Author's own work.

Throughout the five centuries covered by the source material, the marriage license books consistently listed the names of the spouses together with the husband's first surname. The wife's first surname began to be recorded in 1643, with earlier entries usually giving the first surname of either the father or the deceased husband. The inclusion of the second surname of both spouses began in 1876, following its formal establishment in Spain under the Civil Registry Act of 1870. However, the convention of dual surnames was not systematically implemented until the beginning of the 20th century (de Prado Sagrera ²⁰¹²). The date of payment, the tax rate applied, and the husband's occupation were also meticulously recorded in the books. Depending on the historical period, additional details may have included the identity of the spouses' parents and, if widowed, of the deceased spouse. Other information, such as the spouses' marital status, the parents' occupations, and the spouses' place of birth, residence, or parish, may also have been recorded.

This study examines marriages recorded between the 16th and 19th centuries. Before 1540, less than 20% of marriages were recorded with full details, including the surnames of the spouses. Between 1540 and 1640, this figure rises to over 30%, allowing for analysis but also warranting caution about the representativeness and relevance of these data. The fact is that there are very few quality records available worldwide for this early period. From 1640 onwards, over 90% of marriages were recorded in the BHMD with the first surnames of both spouses. From 1880, however, a relaxation of tax payments led to a decline in marriage registrations despite the growing population, suggesting incomplete documentation. Therefore, this study focuses on the 454 267 marriages recorded in the Officiality of Barcelona between 1540 and 1880.

This study uses the isonymy method, where identical surnames between spouses are used as a proxy for consanguinity. The analysis is divided into two phases. The first phase is descriptive, focusing on trends in isonymy over time and the corresponding consanguinity. In the second phase, logistic models are used to estimate whether structural or socio‐strategic determinants had more or less influence on the formation of isonymic marriages.

Isonymy offers significant advantages for the study of historical populations, as it provides a means to infer consanguinity across distant generations that cannot be reconstructed (Colantonio et al. 2003). This method facilitates longitudinal comparisons within the same population and between neighboring populations with similar surname structures (Román Busto ²⁰¹⁵). However, it is not without its limitations. First, isonymy is sensitive to the origin of non‐monophyletic surnames (Mathias et al. 2000) and to the balance between patrilineal and matrilineal proportions (Lasker 1977). Second, the use of isonymy as a proxy for consanguinity may be compromised by surname mutations over time. Finally, this method fails to capture consanguinity within heteronymous, even exogamous, alliances through the female line (Collomp 2000).

In Catalonia, the system of surname inheritance became patrilineal during the 13th‐14th centuries, reinforcing the concept of family lineage or descent (Herzog 2007). Unmarried women also inherited their surname from their father, and upon marriage, it was customary to adopt the husband's surname. With the implementation of the Civil Registry in 1870, this traditional surname inheritance practice became standardized, with all men and women registered using their parents' surnames (Soler Salcedo ²⁰⁰⁸). Consequently, the pattern of surname inheritance in Catalonia, both historically and by legal mandate, enables the use of isonymy.

Isonymy makes it possible to estimate the degree of consanguinity within a population, also known as the consanguinity coefficient (F), on the basis of its surname structure. Consanguinity can be divided into two components: random consanguinity (Fr), which reflects the likelihood of two spouses sharing a surname by chance, and non‐structural or non‐random consanguinity (Fn), which refers to intentional marriages between relatives (Wrigth 1951). These coefficients are expressed as follows:

$$\mathrm{Fr}=\sum {{\mathrm{p}}_{\mathrm{i}}}^{*}{\mathrm{q}}_{\mathrm{i}}/4\mathrm{Fn}=P-\sum \left(\left({{\mathrm{p}}_{\mathrm{i}}}^{*}{\mathrm{q}}_{\mathrm{i}}\right)/\left(4\left(1-\sum {{\mathrm{p}}_{\mathrm{i}}}^{*}{\mathrm{q}}_{\mathrm{i}}\right)\right)\right)$$

where p_i and q_i represent the frequency of surname i in the population, in grooms and brides respectively, and P is the proportion of observed isonymic marriages. F, Fr, and Fn are related by the following equation:

$$\mathrm{F}=\mathrm{Fn}+\left(1-\mathrm{Fn}\right)*\mathrm{Fr}=\mathrm{Fn}+\mathrm{Fr}–{\mathrm{Fn}}^{*}\mathrm{Fr}$$

In exploring the determinants of marriages between relatives, multivariate logistic regression models are used to analyze the occurrence of spouses with matching surnames (Hosmer et al. 2013), adapted for rare events through Firth's correction (Firth 1993). In this framework, surname matching serves as the dependent variable, while the independent variables are grouped into three categories: structural, strategic, and mixed. Given that isonymic marriages are highly unbalanced in comparison to nonisonymic marriages, the maximum likelihood estimates derived from standard logistic regression may be biased—often underestimating the probability of rare events (King and Zeng 2001). To mitigate this effect, Firth (1993) proposed a penalized likelihood approach that adjusts the estimation procedure, producing more stable and reliable results in cases involving rare outcomes.

The logistic regression model is expressed as:

$$\mathrm{g}\left(\mathrm{X}\right)={\mathrm{\beta }}_0+{\mathrm{\beta }}_1{\mathrm{X}}_1+{\mathrm{\beta }}_2{\mathrm{X}}_2+{\mathrm{\beta }}_3{\mathrm{X}}_3+{\mathrm{\beta }}_4{\mathrm{X}}_4+{\mathrm{\beta }}_5{\mathrm{X}}_5+{\mathrm{\beta }}_6{\mathrm{X}}_6+{\mathrm{\beta }}_7{\mathrm{X}}_7+{\mathrm{\beta }}_8{\mathrm{X}}_8+{\mathrm{\beta }}_9{\mathrm{X}}_9+{\mathrm{\beta }}_{\mathrm{p}}{\mathrm{X}}_{\mathrm{p}}$$

where

$$\mathrm{g}\left(\mathrm{X}\right)=\ln \left[\mathrm{\pi }\left(\mathrm{X}\right)/\left(1-\mathrm{\pi }\left(\mathrm{X}\right)\right)\right]$$

The specific independent variables (X) considered for this study are as follows:

‐ Structural variables: X1: Population size; X2: Geographical isolation; X3: Barcelona parish.

‐ Strategic variables: X4: Socio‐occupational group; X5: Tax category; X6: Ability to pay.

‐ Mixed variables: X7: Abundance or rarity of the groom's surnames; X8: Abundance or rarity of the bride's surnames; X9: Marital status.

To correct for small‐sample bias in rare event data, this model is modified by applying a penalization to the likelihood function:

$${\mathrm{L}}^{*}\left(\mathrm{\beta }\right)=\mathrm{L}\left(\mathrm{\beta }\right)·\surd \mid \mathrm{I}\left(\mathrm{\beta }\right)\mid$$

where:

L*(β) is the penalized likelihood function proposed by Firth.

L(β) is the standard likelihood function.

|I(β)| is the determinant of the Fisher information matrix, which quantifies the curvature of the likelihood function.

This adjustment ensures a more accurate estimation of coefficients, particularly when dealing with infrequent outcomes such as isonymic marriages.

Structural variables relate to the size of the marriage market. These variables include geographical isolation due to inaccessibility caused by inadequate transport infrastructure, insularity or location in mountainous regions, all of which hinder the formation of exogamous marriages (Bestard Camps ¹⁹⁹²; Sutter ¹⁹⁶⁸; Zei et al. ²⁰⁰⁵).

In order to analyze the effect of geographical isolation on the size of the marriage market, the place of marriage was used. The BHMD explicitly states the town where the marriage took place from 1715 onwards. For marriages before 1715, the wife's parish was used, as it was customary for marriages to take place in that parish if the spouses came from different parishes. In cases where the wife's parish was not available, the husband's parish was used as the place of marriage.

Population size (X1) serves as a proxy for the size of the marriage market, operating under the assumption that larger populations contain larger marriage markets. Due to the length of the study period, a dynamic categorization became necessary. As a result, three population thresholds were defined to correspond to three different time periods: the 16th–17th centuries, the 18th century, and the 19th century (Table 1). Geographical isolation (X2) was considered as another structural variable, contingent upon the accessibility of major transport routes (Table 1). Finally, considering the significance of Barcelona city and the potential diversity across its parishes, these were also factored in as a structural variable (X3). Some of these were grouped together in order to mitigate the effects of the territorial changes observed during the study period (Table 1).

Strategic variables include choices made by the spouses or their families, or the conventions of their respective social groups, in order to either maintain or improve their social status (Gamella and Carrasco Muñoz ²⁰⁰⁸; Ottenheimer ¹⁹⁹⁰; Rodríguez Sánchez ¹⁹⁹¹; Rubio Pérez ²⁰¹⁰). In this way, marriage served as a means of securing family wealth by consolidating or strengthening it. Marital alliances were also used to uphold social status by establishing and reinforcing power and patronage networks (Gómez Carrasco ²⁰¹⁰; Molina Recio ²⁰²⁰; Sánchez Baena and Chaín Navarro ¹⁹⁹²).

In this study, strategic determinants can only be assessed from the husband's perspective, focusing on his socio‐occupational group, or from the couple's perspective, considering the tax paid for the marriage license. To this end, the husband's occupation (X4) was taken into consideration. For convenience, this was transformed into social class using the internationally recognized HISCLASS classification (van Leeuwen and Maas ²⁰¹¹). The original 12 social groups were condensed into eight to better align with the socioeconomic structure of the study area (Pujadas‐Mora et al. ²⁰¹⁸). Moreover, individual consideration was given to the nobility category, as marriage among relatives was a common practice within this group (Molina Recio ²⁰²¹; Orduna‐Portús ²⁰⁰⁷). Consequently, a total of nine socio‐occupational groups were identified (Table 1).

The marriage tax (X5) was divided into six categories. The nobility and the military class were combined into one category, and merchants, lawyers, and physicians also formed a single category. In order to mitigate potential confounding effects arising from the doubling of rates in 1649, this variable has been standardized to match pre‐1649 values (Table 1). There is a clear correlation between the tax category and the husband's socio‐occupational class, as the tax was primarily determined by the husband's occupation and financial capacity. Therefore, only these two variables were used in the proposed models. In addition, the concept of ability to pay (X6), as proposed by Brea‐Martínez and Pujadas‐Mora (2018a), was introduced to take account of the differences in tax obligations within occupational categories. This strategic variable captures the husband's ability to pay by considering both the fixed marital tax rate and his socio‐occupational position (Kendrick 1939). The international social ranking system HISCAM (Lambert et al. 2013) was used for this purpose, as it assigns a continuous numerical value between 1 and 100 to each occupation. For example, a servant's occupation is assigned a value of 39.90, a day labourer's occupation is assigned a value of 41.81, a farmer's occupation is assigned a value of 51.09, a merchant's occupation is assigned a value of 64.30, a ship captain's occupation is assigned a value of 75.22, and a physician's occupation is assigned a value of 99.00. Ability to pay is calculated as follows:

$${\mathrm{A}}_{\mathrm{i}}={{\mathrm{H}}_{\mathrm{i}}}^{*}{\mathrm{T}}_{\mathrm{i}}$$

where H_i is the husband's HISCAM value and T_i is the tax paid, expressed in units equivalent to pre‐1649 values to standardize the variable for time and inflation factors. Based on the distribution of records, spouses' ability to pay was classified into six categories, from lowest to highest ability (Table 1). The categorization criterion was devised to ensure a normal distribution. In addition, the category Amore Dei, which meant exemption from paying the marriage tax regardless of the husband's occupation, was considered as a proxy for circumstantial poverty or unemployment.

Three mixed variables were also considered, each of which has both structural and strategic properties. The abundance or rarity of spouses' surnames refers to the frequency of occurrence of each surname in the population, taking into account the population's surname structure and diversity. This variable can be seen as an indicator of marriage strategies, where rarer surnames may belong to smaller groups that have persisted over time or to recent migrants, potentially leading to isonymy between them due to restricted marriage markets. The influence of the abundance or rarity of the husband's surname (X7) and that of the wife's surname (X8) was analyzed separately in order to detect possible differences in behavior based on the spouse's gender, influenced by variations in the distribution of surname abundance between them. To this end, both variables were divided into quartiles based on surname frequency within each time period (Table 1). It is important to note, however, that surname abundance or rarity may also be shaped by the socio‐occupational status of the spouses, due to patterns of selective marriage. This would imply that the most common or rare surnames could vary significantly across social classes. To account for this possibility, an additional variable was constructed that measures surname rarity or abundance by period and socio‐occupational category, following the HISCLASS classification. Nevertheless, the models incorporating this new categorization did not yield substantially different results compared to those presented in the article (calculations not shown but available upon request). Moreover, this approach resulted in a significant loss of sample size.

Finally, the spouses' marital status was taken into consideration, providing insights into remarriage (X9). Second marriages were of great importance in Spain, mirroring trends across Europe since the 16th century, and served as a means of maintaining family continuity amidst family disruption (Nausia Pimoulier ²⁰⁰⁶; Tovar Pulido ²⁰¹⁹). In Catalonia, as in Spain as a whole, such marriages accounted for over 25% of all unions between the 16th and 19th centuries (García Barriga ²⁰⁰⁹). This variable was categorized according to the four possible combinations of the spouses' marital statuses (Table 1).

Given that the dataset encompasses a vast and economically diverse territory across an extended time frame, we have chosen to analyze the determinants of isonymic marriages by considering territorial, temporal, and selected strategic variables (Table 2). Initially, we analyzed the entire geographic area and the full time period collectively, omitting the strategic variables Socio‐occupational Group (X4, Model 1), Tax Category (X5, Model 2), and Ability to Pay (X6, Model 3). To better assess the impact of economic context, we then distinguished between rural areas and smaller urban centers, explicitly excluding marriages that took place in the city of Barcelona. For this last location, we again considered the strategic variables Socio‐occupational Group (X4, Model 4), Tax Category (X5, Model 5), and Ability to Pay (X6, Model 6), considering them exclusively. Finally, we analyzed the historical city of Barcelona as a distinct case, given its role as a major urban and economically dynamic center throughout the study period. Within this context, we examined marriage records from various city parishes across the entire period, incorporating the strategic variables Socio‐occupational Group (X4, Model 7), Tax Category (X5, Model 8), and Ability to Pay (X6, Model 9). Additionally, we conducted a focused analysis on marriages celebrated in the final century of the study period. On the one hand, we analyzed the records from the various parishes of the city of Barcelona across the entire study period, considering separately the strategic variables Socio‐occupational Group (X4, Model 7), Tax Category (X5, Model 8), and Ability to Pay (X6, Model 9). On the other hand, we specifically considered marriages celebrated during the 19th century, a period marked by a notable rise in consanguinity—as highlighted in the literature and referenced in the introduction of this article—alongside significant industrialization within the study area. In this case as well, the strategic variables Socio‐occupational Group (X4, Model 10), Tax Category (X5, Model 11), and Ability to Pay (X6, Model 12) were examined separately.

TABLE 2.

Descriptive models.

#	N	Structural variables	Strategic variables
Models Barcelona area and 16th‐19th centuries*.
1	454 267	X1 (Population size) X2 (Geographical isolation)	X4 (Socio‐occupational group)
2	454 267	X1 (Population size) X2 (Geographical isolation)	X5 (Category tax)
3	454 267	X1 (Population size) X2 (Geographical isolation)	X6 (Ability to pay)
Models without Barcelona city and 16th‐19th centuries*.
4	156 779	X1 (Population size) X2 (Geographical isolation)	X4 (Socio‐occupational group)
5	156 779	X1 (Population size) X2 (Geographical isolation)	X5 (Category tax)
6	156 779	X1 (Population size) X2 (Geographical isolation)	X6 (Ability to pay)
Models Barcelona historical city and 16th‐19th centuries*.
7	155 700	X3 (Barcelona parish)	X4 (Socio‐occupational group)
8	155 700	X3 (Barcelona parish)	X5 (Category tax)
9	155 700	X3 (Barcelona parish)	X6 (Ability to pay)
Models Barcelona historical city and 19th century*.
10	97 828	X3 (Barcelona parish)	X4 (Socio‐occupational group)
11	97 828	X3 (Barcelona parish)	X5 (Category tax)
12	97 828	X3 (Barcelona parish)	X6 (Ability to pay)

^*All models include the three mixed variables: X7 (Abundance or rarity of the groom's surnames), X8 (Abundance or rarity of the bride's surnames), and X9 (Marital status).

Source: Author's own work.

4. Results

4.1. Estimating Consanguinity Through Isonymy in the Barcelona Area, 16th–19th Centuries

The total number of isonymic marriages identified by the first surname of the bride and groom in the Barcelona area between 1540 and 1880 amounts to 2258, representing 0.50% of the total number of documented marriages. The frequency of spouses with the same surname showed a steady increase over the period studied, both in absolute numbers and in relative proportions (Figure  2 ). There was also a sharp increase that began in the early 19th century, coinciding with the growth in population described above, and continued until the end of the dataset in 1880. This trend closely resembles that found in studies of other European regions, where peaks in consanguinity are documented between the late 19th and early 20th centuries (Blanco Villegas et al. 2004; Montserrat et al. ²⁰⁰³; Peña et al. ²⁰⁰²; Pettener ¹⁹⁸⁵).

FIGURE 2.

Marital isonymy in the Barcelona area, 1540–1880. Source: Authors own work (BHMD).

The consanguinity coefficient, calculated on the basis of isonymy and disaggregated into a structural fraction (Fr) and a strategic fraction (Fn), shows a significant increase towards the end of the 19th century (Figure  3 ), in line with findings documented in literature from various regions of Spain, including Galicia, Andalusia, and Castile, as well as Europe and the Americas (Núñez Negrillo ²⁰¹⁵; Pettener ¹⁹⁸⁵; Varela et al. ²⁰⁰³). However, the overall level of consanguinity in the Barcelona area is relatively low compared to the aforementioned regions (Fuster Siebert and Colantonio ²⁰⁰³).

FIGURE 3.

Trends in structural and strategic consanguinity in the Barcelona area, calculated by isonymy of the first surname, 1540–1880. Source: Authors own work (BHMD).

One of the main explanations for the relatively low incidence of consanguineous marriages in the area studied could be the Catalan inheritance system, characterized by impartible inheritance and primogeniture, a practice shared with other Spanish regions such as Navarre and the Basque Country (Moreno‐Almárcegui and Zabalza Seguín ¹⁹⁹⁹). Within the social classes associated with landownership in Catalonia, consanguineous marriage strategies became less important as they were not necessary to consolidate and maintain family wealth, since the inheritance system based on a unique heir guaranteed this, unlike in equal inheritance models (Ferrer Alòs ²⁰⁰⁵). This inheritance system was fully institutionalized in the 14th century, when the Catalan Courts decided to enshrine the customary practice of primogeniture in law to prevent the fragmentation of family wealth, a principle that was maintained until the enactment of the Civil Code in the 19th century (Puig Salellas ¹⁹⁹⁶).

At the same time, the level of structural consanguinity (Fr) remained relatively stable over the almost three and a half centuries under consideration. This suggests that the observed increase in total consanguinity is almost entirely due to the strategic fraction (Fn). Thus, the rise in consanguineous marriages is primarily the result of strategic decisions by the spouses or their families, rather than mere chance encounters in the marriage market. This phenomenon occurred beyond what would be expected from the structural characteristics of the marriage market, confirming the so‐called consanguinity boom that took place in the Barcelona area towards the end of the 19th century. From this, it becomes evident that studying the social and economic determinants that explain the formation of this type of marriage is important, as outlined in the objectives of the paper.

5. The Determinants of Isonymic Marriage Formation in the Barcelona Area, 19th–19th Centuries

In order to analyze the long‐term structural and strategic determinants of isonymic marriages in the Barcelona area, 12 logistic regression models are presented (Tables 2, 3, and 4). In all of which the period in which the marriage took place is significant. Marriages from earlier periods have a lower probability of being isonymic across all models, ranging from 30% to 70%, than those that took place between 1860 and 1880, which serves as the reference category. It is important to note that the odds increase over time. This finding is consistent with the trend in isonymy and the trends in random and nonrandom consanguinity coefficients described earlier (Figures  2 and 3 ), and with the findings of the literature review on the boom of consanguinity cited earlier.

TABLE 3.

Results of marital isonymy logistic models 1–6.

# MODEL	#1	#2	#3	#4	#5	#6
N	454 267	454 267	454 267	156 779	156 779	156 779
	Exp(B)	Exp(B)	Exp(B)	Exp(B)	Exp(B)	Exp(B)
Time period
1540–1719	0.36***	0.38***	0.38***	0.32***	0.33***	0.32***
1720–1739	0.39***	0.39***	0.38***	0.35***	0.34***	0.35***
1740–1759	0.47***	0.47***	0.47***	0.40***	0.40***	0.40***
1760–1779	0.42***	0.43***	0.43***	0.34***	0.34***	0.35***
1780–1799	0.41***	0.43***	0.42***	0.36***	0.37***	0.35***
1800–1819	0.46***	0.46***	0.46***	0.44***	0.45***	0.44***
1820–1839	0.57***	0.60***	0.59***	0.62***	0.63***	0.61***
1840–1859	0.67***	0.67***	0.68***	0.69***	0.69***	0.70***
1860–1880 (ref)
Population size (X1)
Small	2.26***	2.26***	2.36***	2.44***	2.54***	2.39***
Medium	2.01***	2.09***	2.10**	2.19***	2.18***	2.08***
Large (ref)
Geographical isolation (X2)
Coastal	1.34***	1.38***	1.37***	1.36***	1.36***	1.33***
Inland (ref)
Socio‐occupational group (HISCLASS) (X4) (1)
Nobility	2.17**			1.58
HISCLASS 1	1.74**			0.67
HISCLASS 2	1.31*			1.41
HISCLASS 3, 4 & 6	1.47***			1.63***
HISCLASS 5	1.69***			1.62
HISCLASS 7	0.82**			0.77*
HISCLASS 9 & 11	0.84***			0.88
HISCLASS 10 & 12	1.15			1.03
HISCLASS 8 (ref)
Tax category (X5) (2)
TAX 1&2 (12£/2£ & 8 s)		8.73***			9.22***
TAX 3 (1£ & 4 s)		2.37***			2.47***
TAX 4 (12 s)		2.21***			2.92***
TAX 5 (6 s)		2.17***			1.72**
TAX 6 (4 s)		1.18**			1.02
Amore Dei (0) (ref)
Ability to pay (X6)
Very low			1.02			1.01
Low			1.03			0.91
Medium‐low			1.29***			1.15
Medium‐high			1.07			0.81
High			2.16***			2.09***
Very high			2.10***			2.09***
Amore Dei (ref)
Abundance or rarity of the groom's surnames (X7)
1st Quartile	4.03***	3.98***	3.86***	3.77***	3.72***	3.86***
2nd Quartile	2.33***	2.36***	2.35***	2.40***	2.44***	2.44***
3rd Quartile	1.42***	1.42***	1.36***	1.54***	1.51***	1.56***
4th Quartile (ref)
Abundance or rarity of the bride's surnames (x8)
1st Quartile	3.90***	3.88***	3.79***	3.39***	4.01***	4.05***
2nd Quartile	2.42***	2.49***	2.49***	2.75***	2.83***	2.76***
3rd Quartile	1.35***	1.37***	1.35***	1.54***	1.56***	1.55***
4th Quartile (ref)
Marital status (X9)
Single/Widowed	0.83	0.82*	0.86	0.91	0.93	0.93
Widowed/Single	0.92	0.90	0.91	0.89	0.93	0.91
Widowed/Widowed	0.57***	0.57***	0.59***	0.63**	0.68**	0.64**
Single/Single (ref)
Constant	0.00***	0.00***	0.00***	0.00***	0.00***	0.00***

Note: ***p < 0.01, **p < 0.05, *p < 0.1, 1£ (pound) = 20 s (shillings).

Abbreviations: HISCLASS 1, higher managers; HISCLASS 2, higher professionals; HISCLASS 3, 4, & 6, lower managers, lower professionals, clerical and sales personnel, and foremen; HISCLASS 5, lower clerical and sales personnel; HISCLASS 7, skilled workers; HISCLASS 9 &11, lower‐skilled workers and unskilled workers; HISCLASS 10 &12, lower‐skilled farmworkers and unskilled farmworkers; HISCLASS 8, farmers (ref).

Source: Authors' own work (BHMD).

TABLE 4.

Results of marital isonymy logistic models 7–12.

Model	#7	#8	#9	#10	#11	#12
N	155 700	155 700	155 700	97 828	97 828	97 828
	Exp(B)	Exp(B)	Exp(B)	Exp(B)	Exp(B)	Exp(B)
Time period
1540–1719	0.44***	0.46***	0.38***
1720–1739	0.40***	0.37***	0.38***
1740–1759	0.56***	0.56***	0.53***
1760–1779	0.56***	0.53***	0.48***
1780–1799	0.50***	0.47***	0.46***
1800–1819	0.45***	0.42***	0.42***	0.43***	0.41***	0.42***
1820–1839	0.48***	0.45***	0.44***	0.46***	0.43***	0.43***
1840–1859	0.67***	0.67***	0.65***	0.67***	0.69***	0.67***
1860–1880 (ref)
Barcelona parish (X3)
St Just/MD Mercè/St Miquel	1.43	1.46*	1.39	1.43	1.46	1.36
Sta Maria del Mar	1.26	1.35	1.21	1.49*	1.66**	1.45
St Pere de les Puelles	0.96	1.03	0.94	0.99	1.06	0.96
St Jaume	1.29	1.33	1.18	1.39	1.45	1.24
Sta Maria del Pi	1.09	1.17	1.05	1.03	1.11	1.00
Cathedral	3.63***	3.40***	3.51***	1.70	1.43	1,63
St Cugat del Rec (ref)
Socio‐occupational group (HISCLASS) (X4)(1)
Nobility	2.37**			2.19*
HISCLASS 1	2.51***			2.65***
HISCLASS 2	1.53*			1.66
HISCLASS 3, 4 & 6	1.46*			1.39
HISCLASS 5	1.99***			1.94**
HISCLASS 7	0.93			0.90
HISCLASS 9 & 11	0.91			0.87
HISCLASS 10 & 12	1.75**			2.92***
HISCLASS 8 (ref)
Tax category (X5)(2)
TAX 1 & 2 (12£/2£ & 8 s)		6.49***			9.68***
TAX 3 (1£ & 4 s)		1.99**			2.28**
TAX 4 (12 s)		1.84***			1.85**
TAX 5 (6 s)		2.43***			2.89***
TAX 6 (4 s)		1.25**			1.39**
Amore Dei (0) (ref)
Ability to pay (X6)
Very low			0.69			0.82
Low			1.17			1.16
Medium‐low			1.36**			1.59***
Medium‐high			1.23			1.40**
High			2.07***			2.34***
Very high			1.92***			2.06***
Amore Dei (ref)
Abundance or rarity of the groom's surnames (X7)
1st Quartile	4.59***	4.39***	4.48***	3.59***	3.37***	3.52***
2nd Quartile	2.14***	2.11***	1.99***	1.79***	1.71***	1.69***
3rd Quartile	1.45**	1.38**	1.39**	1.25	1.18	1.20
4th Quartile (ref)
Abundance or rarity of the bride's surnames (X8)
1st Quartile	3.96***	3.79***	3.89***	3.10***	2.92***	3.05***
2nd Quartile	2.02***	1.98***	1.88***	1.68***	1.60***	1.59***
3rd Quartile	1.45	1.19	1.21	1.13	1.06	1.08
4th Quartile (ref)
Marital status (X9)
Single/Widowed	0.89	0.94	0.92	0.83	0.88	0.87
Widowed/Single	1.13	1.11	1.14	1.00	1.00	1.05
Widowed/Widowed	0.60**	0.60**	0.59**	0.53**	0.54**	0.50**
Single/Single (ref)
Constant	0.00***	0.00***	0.00***	0.00***	0.00***	0.00***

Note: ***p < 0.01, **p < 0.05, *p < 0.1, 1£ (pound) = 20 s (shillings).

Abbreviations: HISCLASS 1, higher managers; HISCLASS 2, higher professionals; HISCLASS 3, 4 & 6, lower managers, lower professionals, clerical and sales personnel, and foremen; HISCLASS 5, lower clerical and sales personnel; HISCLASS 7, skilled workers; HISCLASS 9 &11, lower‐skilled workers and unskilled workers; HISCLASS 10 &12, lower‐skilled farmworkers and unskilled farmworkers; HISCLASS 8, farmers(ref).

Source: Authors' own work (BHMD).

With regard to the structural determinants of isonymy, the size of the population (X1) in which the marriage took place is statistically significant in all models including this variable (models 1, 2, 3, 4, 5, and 6). Thus, a smaller population size is positively correlated with isonymy between spouses, suggesting a more restricted marriage market. In addition, the probability of marriage between relatives was found to be lower in inland parishes than in coastal ones (X2) (models 1, 2, 3, 4, 5, and 6). It should be noted that the coastal parishes in the BHMD (with the exception of Barcelona city) are more isolated than their inland counterparts, due to limited access to transport routes and the region's main commercial centers. As a result, the difficulty for residents of coastal towns to attend fairs and markets likely hindered their interaction with potential marriage partners, thereby facilitating the formation of endogamous marriages (Ferrer Alòs ²⁰¹⁸; López ²⁰¹⁵).

With regard to the strategic variables (X4), in models 1, 7, and 10, individuals in higher socio‐occupational categories (Nobility and higher managers and professionals (HISCLASS 1 & 2)) were more likely to enter into isonymic marriages than farmers (HISCLASS 8), with nobility being twice as likely. This tendency may be related to the endogamous strategies of the higher classes, aimed at preserving their social status and preventing social decline. Moreover, for the nobility, the decline in the number of noble families over time contributed to the increased adoption of consanguineous strategies to avoid a loss of power Chacón Jiménez (2011). The Baron of Maldà, Rafael d'Amat i de Cortada, provides a vivid example of the consanguineous strategies adopted by the Catalan nobility in the 18th century (Pérez Samper ²⁰⁰³). The Amat family was the epitome of Barcelona's wealthy nobility, enjoying numerous privileges and noble titles. The baron's uncle, Manuel d'Amat i de Junyent, was a lieutenant general who would become famous as viceroy of Peru between 1761 and 1776. In 1776, Rafael married his first cousin, María de la Esperanza d'Amat i de Rocabertí, the youngest daughter of the second Marquises of Castellbell, in the parish church of Santa María del Pi in Barcelona (isonymic marriage of 21 November 1766, volume 147, page 58v; source: BHMD). Subsequently, two of the baron's children, Rafael and María Escolástica, married their first cousins on the same day, 11 December 1798, in the Barcelona Cathedral. María Escolástica d'Amat i d'Amat married her first cousin, the Marquis of Castellbell, Manuel d'Amat i de Peguera, while Rafael d'Amat i d'Amat married his first cousin Josefa Vega i d'Amat (isonymic marriages of 11 December 1798, volume 164, page 86; source: BHMD). This endogamous strategy continued for generations within Barcelona's noble class.

Regarding the category of guild masters (HISCLASS 3, 4, 5, & 6), alliances between relatives within the same guild were potentially good for acquiring trade knowledge and ensuring the continuity of family businesses (Ferrer Alòs ²⁰¹¹; Llorente ²⁰¹²). Consequently, in models 1, 7, and 10, there is a significant tendency in this category towards consanguineous marriages compared to farmers. However, in model 4, which excludes Barcelona city, there are almost no significant differences between any of the socio‐occupational categories and the reference. Conversely, model 1 shows that marriages between skilled or unskilled industrial workers (HISCLASS 7, 9, & 11) were significantly less likely to be isonymic than the reference category. Focusing only on Barcelona city (models 7 and 10), lower‐skilled or unskilled farmworkers (HISCLASS 10 & 12) show a higher propensity for marital isonymy than farmers, perhaps due more to their strong practice of endogamy than an interest in preserving wealth that they did not possess. In the other models, these groups do not differ significantly from farmers. Thus, consanguinity was less attractive for lower socio‐occupational categories than for farmers, except for one very specific case. As described for farmers, the primogeniture system in Catalonia was the main facilitator of wealth preservation, and therefore intrafamily alliances did not have this interest (Ferrer Alòs ²⁰¹⁸; Manzano Ledesma ²⁰¹⁹). In addition, possible interactions between husbands' occupations and time periods were investigated using models that included interactions of all time categories with socio‐occupational categories. The inclusion of these interactions in the regression models does not lead to any significant deviations from the models without these interactions.

Tax category (X5) emerges as the dominant strategic determinant influencing the occurrence of marriages between spouses with the same surname (models 2, 5, 8, and 11), potentially obscuring the effects of the ability to pay, as evidenced by Brea‐Martínez and Pujadas‐Mora (2018a) using the same data. This may be attributed to the construction of the ability to pay variable, which was developed according to a normal distribution to create distinct groups based on financial capacity, thus preserving the social stratification inherent in the marriage tax. Furthermore, both the marriage tax and the ability to pay categories, in addition to considering the groom's socio‐occupational status, likely accounted for other more intangible factors. Therefore, in addition to the groom's socio‐occupational position, the rate‐setter likely considered various contextual factors that enabled a more nuanced adjustment of the rate. In almost all models including it, this variable is significant with respect to Amore Dei, which denotes individuals who were exempt from paying marriage tax. For those in the highest tax categories, the probability of a surname match is up to six times higher in models 2, 5, 8, and 11 than for the exempt group. For the other tax categories, the probability of isonymic marriages is significantly higher, by a factor of two or even three, except for those paying the lowest tax rates, who have a higher probability of isonymy than the poor, but never twice as high.

When analyzing the models that include ability to pay (X6) (models 3, 6, 9, and 12), the observed differences in the probability of isonymy appear relatively modest. In models 3, 6, and 9, only marriages with a higher ability to pay show an increase in the probability of isonymy compared to those with no ability to pay. In contrast, in model 12, relating to Barcelona city in the 19th century, apart from marriages categorized as low or very low ability to pay, all others are significantly more likely to be isonymic than the reference category.

Considering the three strategic variables related to wealth and socio‐occupational level, such as the marriage tax, the ability to pay, and the socio‐occupational category of the groom (HISCLASS), the tax is the variable that best and most completely captures the existing social differentiation between unions, showing significance in almost all levels and models. In fact, although there is a table of equivalences between occupational labels and taxes to be paid, Brea‐Martínez and Pujadas‐Mora (2018a) showed that the assignment of the level of the marriage tax was not always equivalent to that shown in the aforementioned table. This suggests that tax collectors or registrars may have applied contextual knowledge, adjusting the marriage tax either upward or downward, or aligning it with the prescribed amounts based on circumstances.

Regarding the influence of the Barcelona parishes (X3) where the marriage took place, there seem to be significant differences between marriages performed in certain parishes and those performed in St Cugat del Rec. This relatively small parish, located in the heart of the historic walled city of Barcelona, has a socio‐occupational composition that reflects the average diversity of the city as a whole. In the regression models covering the entire period under study (models 7, 8, and 9), marriages performed in the Cathedral have a significantly higher probability of spouses sharing a surname, three times higher than in St Cugat del Rec. Of all the parishes in Barcelona, the Cathedral had the lowest number of marriages in the periods studied and a very high proportion of marriages between the nobility and the upper classes. For this reason, the marriages that took place in the Cathedral were much more likely to be consanguineous, as this was used as a strategy to maintain social status (Rey Castelao ²⁰²¹). Conversely, the other more popular parishes show no significant differences, both when considering the whole period under study and when considering only the 19th century.

The ability to remarry (X9) varied by sex. In the preindustrial era, it was common for widowers to seek a new wife in order to strategically secure family offspring (Blanco Carrasco ²⁰²⁰). For widows, it was culturally more difficult to re‐enter the marriage market, although less so during periods of peak mortality when the number of marriage candidates was reduced (Brodsky 1986). Spouses' marital status plays a significant role in marital isonymy across all models, but only for the category of marriages between two previously widowed spouses, which have a much lower probability of isonymy than marriages between previously unmarried spouses. This suggests that marriage between relatives was not an attractive strategy for the widowed, who tended to avoid them. An explanation for this behavior in the narrow marriage market of widows and widowers may lie in the difficulty of entering it, with possible strategic reasons playing a marginal role.

Finally, significant differences are observed across all models in the categories pertaining to the abundance or rarity of the groom's (X7) and bride's surnames (X8). Marriages between spouses whose surnames are in the more abundant surname quartiles show a higher degree of isonymy than those in the rarest surname quartile. This trend is in line with the frequency of the surnames in the population, and thus acts as a structural variable that disregards any strategic influence that may be associated with the rarest surname quartile. Thus, the surname Ferrer—the equivalent of Smith in English—is the most common surname within the analyzed population, with spouses sharing this surname showing the highest levels of marital isonymy. Isonymy involving very common surnames like Ferrer appears to be more closely linked to the population's structural composition than to consanguinity as a deliberate family strategy. By contrast, isonymy associated with rarer surnames tends to correlate more strongly with strategic consanguinity, occurring within smaller groups due to the exclusivity or relative novelty of these surnames within the population.

6. Discussion and Conclusion

Catalonia, and in particular the Barcelona area, provides an ideal context for analyzing consanguineous marriage strategies over a long study period. Its demographic trends and economic development, especially with regard to its inheritance system based on primogeniture, make this region a rich and diverse scenario in which marriage remains a key strategy for structuring society. In this regard, by analyzing marital consanguinity through isonymy data from the BHMD, this research has been able to identify the trends and determinants of such marriages in the Barcelona area.

The observed trend in marital consanguinity is similar to that found in the rest of Spain and Europe, with a continuous increase from the 17th century and a peak in the 19th century. This boom in consanguinity can be linked to several sociocultural and demographic factors. First, it was fuelled by a reduction in the cost of ecclesiastical dispensations and a loosening of the restrictions that prevented the formation of marriages between relatives from the mid‐19th century onward. Second, an increase in the size of the market for consanguineous marriage as a result of the decline in adult mortality and the greater importance of horizontal relationships made it easier to find a spouse within family circles. The strengthening of horizontal relationships marks one of the most significant transformations in kinship throughout history, beginning in the mid‐18th century. This shift was grounded in alliances, sentiment, networks of kindred, and social and familial endogamy, as Sabean et al. (2007) pointed out, and it has been previously referred to. These elements align closely with the formation of the bourgeoisie, whose wealth was not land‐based. As a result, the traditional vertical structure of families, tied to land inheritance, diminished in importance, giving way to horizontal bonds formed through assortative mating within kinship networks.

Although the trend of increasing consanguinity is the same, the levels of consanguinity derived from isonymy calculations in the Barcelona area are considerably lower than those observed in other Spanish regions. This difference is most likely due to Catalonia's adherence to a strict system of undivided inheritance, a characteristic shared with eastern and northern Spain, as opposed to the more egalitarian inheritance system prevalent in regions such as Andalusia, Castile, Extremadura, and parts of Galicia. Thus, the strategy of couples marrying to consolidate family assets is unnecessary in areas with primogeniture inheritance systems. In any case, it is worth noting that in Catalonia, consanguineous unions seem to reflect social endogamy and the desire to accumulate power rather than the imperative of safeguarding family wealth, as this was already secured by the inheritance system referred to above.

When the structural and strategic components of consanguinity in marriages in the Barcelona area from the 16th to the 19th centuries are disaggregated, the structural fraction is found to be almost completely stable over time. Therefore, the continuous increase in consanguinity and its sharp rise toward the end of the study period are due solely to the choices made by the spouses and their families, and not to the structure of the marriage market.

By analyzing the strategic determinants of isonymic marriages in the Barcelona area, as revealed by multivariate models, this study has identified different behaviors among different social groups. In particular, the marriage tax, as a reliable reflection of the husband's wealth, emerges as the strategic determinant that best describes the differences in the formation of isonymic marriages across categories. Marriages among the wealthy, who are subject to higher taxes, show a greater propensity to isonymy than those among lower‐income groups, who are exempt from such taxation. For wealthy couples, consanguinity serves as a strategic means of maintaining their privileged social position, facilitating the establishment and consolidation of power networks. Although marriages between farmers are the most common in the region under study, their frequency decreases over time due to the shift towards industrialization from the late 18th century onwards. Notably, this social class has a very low probability of isonymy, suggesting that consanguinity has limited appeal given the undivided inheritance model in Catalonia, which favors the consolidation of family property. Conversely, the ability to pay appears to be a relatively minor strategic variable in the region as a whole and in rural areas, with its impact being most pronounced in the city of Barcelona, where almost all levels of ability to pay correlate with greater isonymy compared to the poor, who show the lowest propensity to consanguineous unions. Finally, the marital status of the spouses seems to have a negative impact on the choice of partner within the family in marriages involving the widowed, probably due to the difficulties in accessing the marriage market.

With regard to the structural determinants of isonymy, which are linked to population and territorial characteristics, the results show that geographical isolation, smaller marriage markets, and the abundance of surnames in a population contribute positively to the occurrence of marital isonymy. Geographically isolated populations in terms of access to travel routes and exchange of goods and people show greater isonymy, consistent with structural endogamy tendencies. Conversely, larger populations show reduced marital isonymy due to their larger marriage markets, which facilitate the selection of spouses beyond family circles. The abundance of a surname in a population promotes isonymy in proportion to its frequency. In terms of parishes in the city of Barcelona, marriages performed in the Barcelona Cathedral show significantly higher rates of isonymy compared to other parishes, highlighting the propensity for consanguinity as a marriage strategy among influential social classes, who were the main group to marry in this venue. Marrying in a cathedral was a powerful display of status, both within one's social circle and to the broader society. The ceremony served as a public declaration of the families' prestige, influence, and social standing.

The low levels of consanguinity observed in the Barcelona area between the 16th and 19th centuries suggest the existence of alternative forms of significant marital alliances more in line with the prevailing inheritance system. In particular, sibling set exchange marriages, marriages between two widowed parents and their unmarried offspring, and marriages between relatives by marriage, especially those stemming from sororate and levirate practices, need to be studied. Such unions served to maintain family status while reducing the cost of marriage by avoiding dowries and ecclesiastical dispensations. Given their undeniable role in family strategies for maintaining social status and wealth, they deserve attention in future lines of research.

Ethics Statement

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Ruíz García, J. , Pujadas‐Mora J. M., and López‐Villanueva C.. 2025. “Strategy or Necessity? The Determinants of Marriage Between Relatives in the Barcelona Area, Sixteenth–Nineteenth Centuries.” American Journal of Human Biology 37, no. 7: e70106. 10.1002/ajhb.70106.

Data Availability Statement

Research data are not shared.