Marriages among people with disabilities in 19th-century Sweden: marital age and spouse’s characteristics

ABSTRACT While marrying was an expected event in 19th-century Western society and has been subject to much historical research, there are few studies on how disabilities influenced people’s marriage patterns and spouse selection. The aim of this analysis is to contribute clarification on this issue by examining with whom disabled men and women married and the marital age and socio-demographic characteristics of them and their spouses. In total, 188 disabled individuals born in the first half of the 19th century and who married in the Sundsvall region, Sweden, are studied. The results reveal that disabled men and women did not marry each other, and they entered into marriage at a slightly higher age than the average, although there was usually no marked age gap between them and their spouse. Endogamous patterns were primarily found regarding the socio-spatial background of the two spouses. This analysis is one of the few studies identifying the marriages among a comparatively large number of disabled people using demographic data. Their participation in the partner pool highlight their agency historically and emphasize that disability did not lead to distance from social life in past society.


Introduction
Marriage was one major goal of young people in the past, and it has been subject to considerable historical research (e.g. Alter, 1991;Hajnal, 1965;Lundh & Kurosu, 2014;Reher, 1998;van Leeuwen & Maas, 2019). Having a partner and family still signify the transition to adulthood and constitute one key to be recognized as a 'real' man or woman. Historically, knowledge on the lives and possibilities of disabled people is limited, especially regarding their role as possible partners in marriage. One reason for this is because they constitute a minority in populations rarely recognized in sources, except from occasionally in records on poor relief or from institutions or asylums (Anderson & Carden-Coyne, 2007;De Veirman, 2015;Förhammar & Nelson, 2004;Haage, 2017;Stiker, 1999). Not recognizing their agency and possibilities, such records and studies suggest that they led passive or dependent lives being faced with marginalization in the labor and marriage markets or from society at large.
The life and marriage of Stina, from the Sundsvall region in Sweden that we studied, tell a slightly different story. According to the parish registers, she was born in 1814 in the parish of Häggdånger nearby this region to which her family moved when she was 17 years old, and she was recorded as being weak sighted. Stina took up employment as a maidservant (piga) and held her domestic position moving between different farm households and parishes in the region. In 1842 and at the age of 28, Stina was employed by a farmer in the parish of Hässjö named Per, a recent widower. One year later, she married her master at the age of 29, while he was 53. Stina exemplifies one disabled woman who married, but to a widower substantially older than herself. We might wonder whether such a large age gap and widowers were typically found among other disabled women who married and what these patterns looked like among disabled men. This study answers these questions and others regarding the demographic and socio-economic features of disabled men and women and the spouses they married.

Aims and objectives
This analysis aims to contribute to the debate over disabled people's agency and possibilities in history by investigating 188 disabled men and women who married in the 19thcentury in the Sundsvall region of Sweden. First, we analyze whether they tended to marry each other, their age at marriage, and whether disability resulted in any marked age gap between the spouses. Second, their socio-economic and geographical backgrounds are examined to see if any socio-spatial similarities governed partner selection. Parish registers digitized by the Demographic Data Base (DDB) of Umeå University are key to revealing the marriages of persons with disabilities. We use descriptive statistics to identify their participation in social life and society as represented by marital unions. To add some depth to the quantitative results, a few individual cases are highlighted from the parish registers.

Background and rationale of the study and its theoretical framing
The present study is one in a series of analyses comprising a comparatively large number of cases to obtain life-course data on people with disabilities in the past. In a previous analysis, we used the DDB registers to run Cox proportional hazard regressions of more than 25,000 cases and found that disability decreased people's marital chances to a significant degree in 19th-century Sweden . In another study applying sequence analysis of some 8,800 individuals during young adulthood, we showed that disability impeded the trajectories toward occupation, marriage, and childbearing . Our quantitative life-course findings both support and differentiate research based on a smaller number of cases suggesting that impairments have historically limited people's participation in work and social life. However, in large datasets allowing aggregated analyses of events like marrying, less typical patterns tend to fade away in favor of more general results that are statistically significant. Moreover, studies by us and other scholars discussed below show that neither work nor marriage was impossible if disability interfered with life, which stresses the agency or possibilities of disabled people in society even though they married to a lower extent than others. One major rationale for the present analysis is to take these notions further by highlighting disabled individuals who became partners in marriage.
Even though this study takes its point of departure in individuals having disabilities, our examination of their marriage patterns is influenced by developments in disability studies taking the context into consideration. Since the 1990s, disability history has increasingly moved its focus away from the individual/medical model to the social model for studying disability (Longmore & Umansky, 2001;Oliver, 1996Oliver, , 2013. While the former model recognizes disabled people as suffering from a timeless pathological condition, the latter model situates disability in broader economic, political, and sociocultural contexts like gender, class, and ethnicity (Anderson & Carden-Coyne, 2007;Kudlick, 2003). Consequently, disability outcomes are shaped by structural factors and attitudes in the time-space environment, not primarily by the impairment itself or the individual suffering from it. This makes disabled spouses most interesting to examine, and not just to stress the fact that they did marry. An analysis of their marriages and partners provides clues to how the surrounding society looked upon disabilities. This constitutes another major rationale for our study and includes wider circles of a past people than just the disabled spouses we target for analysis.
Labeling theories refer to the surroundings as well in suggesting that human marginalization originates from behaviors or attributes regarded as deviant by the environment (Becker, 1966;Goffman, 1963;Lemert, 1967;Susman, 1994;Vikström, 2008). If persons having disabilities are perceived accordingly by others, they become subject to normative attitudes and stigmatization that can lead to segregation from society. Labeling mechanisms probably add an explanation to many findings on how disability impedes people's possibilities, including our life-course comparisons between non-disabled and disabled individuals showing that the latter experienced significant disadvantages concerning work, marriage, and survival (Haage, Häggström Lundevaller, & Vikström, 2016;Haage et al., 2017;Vikström, Häggström Lundevaller, Junkka, & Haage, 2019). However, focusing on disabled people who married holds some benefits beside making their opportunities become more visible than if being consistently compared with non-disabled 'averages', which also has normative implications. An investigation of their marriage patterns and spouses will suggest whether disability was associated with social distance between groups of people due to a possible labeling. If disability narrowed the circle of friends and social networks and participation in society, it would imply a smaller supply of potential partners who in addition could be old or poor or have health issues, for example. Disability could also delay marriage if contracting a spouse was difficult. Furthermore, spouses with impairments might have married each other, as this could be kind of a match if they primarily interacted with disabled peers. Such outcomes would suggest that disability increased the social distance between people or caused marginalization in the partner pool, although they obviously socialized by marrying.

Historical research about marriage and from a disability perspective
For a long time, marriage was considered the only adequate form for intimate relationships and reproduction, and it has been the subject to a vast body of historical research. Because it transferred social and economic interests across generations and between families, marriage was one means to regulate power and relationships between individuals and groups in past society. This section discusses marriage patterns in the 19th century and from disability perspectives after having outlined some key prerequisites for marrying and the legislation surrounding marriage in Sweden.

Prerequisites for marriage and legal restrictions
Beside the demographic makeup of the partner pool and laws about what constituted eligible spouses, the labor market and economic production worked to structure the prerequisites for marrying in 19th-century society. As in most other areas in North Western Europe, the major occupations among young people in Sweden consisted of positions as farmhands, cottagers, or apprentices and increasingly as unskilled labors in industry, while women were devoted to the domestic sector working as maidservants (Harnesk, 1990;Matović, 1984;Vikström, 2010). Because the salaries were low and the opportunity to inherit land or property limited, it took time for young people to gather the material resources needed to establish a household through marriage. Consequently, age at marriage was high, and in Sweden men married at the average age of about 27 and women at about 25 (Gaunt, 1983;Lundh, 2007;Nilsson & Tedebrand, 2005). Impairment could imply marriage difficulties if it impeded the capability to work, which was key to affording a family, especially for men who were regarded as the major providers (Harnesk, 1990;Horrell & Humphries, 1995;Janssens, 1997). That insufficient or uncertain access to material resources limits people's attraction in the partner pool has been confirmed in research on past as well as present times (Oppenheimer, 1994(Oppenheimer, , 1997. This helps explain the high age at marriage obvious in the few historical works on the marriage age of disabled people (De Veirman, 2015;Olsson, 1999) as discussed below.
One of the primary issues that the church regulated was marriage; spouses too young were not allowed to marry, nor if they were too closely related by blood. In 19th-century Sweden men had to be at least 21 years old and women 17 years old to be married (Hafström, 1975). Since the 18th century, authorities have paid attention to the benefits of having a healthy population, and this promoted the Swedish government's interest in impediments for marriage for other reasons than age and kinship, and the common practice during the 19th century was to prevent marriages between those having mental disabilities (sinnessvag, sinnesslö) (Hafström, 1975). However, already the church law of 1686 stipulated that people who had incurable and contagious diseases were not allowed to contract a spouse, and the law of 1757 prohibited those suffering from epilepsy or idiocy from marrying. The fear of the church and authorities was that the offspring of such unions could inherit dysfunctions from their parents and become a burden to the community (Engwall & Larsson, 2012;Förhammar & Nelson, 2004;Olsson, 2010). Parents shared some of this fear as they sometimes discouraged the marriage of their disabled adult children and also worried about how society would perceive these marriages. Similar worry could make the disabled themselves refrain from close relationships and marriages (Kudlick, 2008).

Endogamous and exogamous marriage patterns
Historical research shows that marriage patterns depended on multiple factors beyond the couple marrying and their ability or decision to realize or afford it (Alter, 1991;Hajnal, 1965;Kalmijn, 1998;Lundh & Kurosu, 2014;Reher, 1998;van Leeuwen & Maas, 2019). The socio-economic position of the family and its property and inheritance that could be transferred through marital unions was one major key to the choice of spouse. Beside socio-cultural factors such as language and religious denomination, marriage patterns depended on demographic factors such as the gender and age distribution of potential spouses and their geographical origin. Shorter (1975) proposed that these factors were losing ground while love became more essential for partner selection, thus leading to the smaller age gaps between spouses he was observing during the urban-industrial processes of the 19th and 20th-century Western world. However, Shorter's theory has been contested. Studying 19th-century Belgium, sociologists Van de Putte and Matthijs (2001) conclude that restrictions regarding the legal marital age created a pool of candidates having similar ages that caused a decreasing age gap. They further find that over time this decrease particularly characterized the cultural middle classes, while among the economic middle and lower working classes this age gap was modest already at the beginning of the 19th century. The age gap persisted among the elite groups because their pool of potential partners was relatively small and remained small because spouses in general did not marry outside their social group. In Sweden as well, wealth and social position played a part when families from affluent strata chose among potential partners for their children and could result in substantial age differences between younger brides and older grooms. Such difference was rare among couples from the lower social strata because they were less governed by inheritance of land or economic concerns (Dribe & Lundh, 2005Gaunt, 1983;Nilsson & Tedebrand, 2005). However, farmers could seek a daughter-in-law having a dowry benefiting their sons, and journeymen could look for a daughter or widow of a master artisan to obtain his title (Holmlund, 2003).
Scholars researching partner selection and marriage often use the concepts of endogamy and exogamy to capture the 'distance' in terms of dis(similarities) between spouses. While the former concept concerns 'within-group' or 'equal' patterns, exogamy refers to the reverse such as marrying across groups (De Veirman, 2015, pp. 288-295;Matović, 1984, pp. 27-28;Vikström, 2003, pp. 185-209;Kalmijn, 1998;van Leeuwen & Maas, 2019;van Leeuwen et al., 2005). Age endogamy means that the two spouses marrying show about the same age at marriage. This is regarded to reflect a higher level of equality between male and female spouses in society than age exogamy, which is more typical for marriages out of economic concerns in patriarchal structures (Beekink, Liefbroer, & van Poppel, 1998;Van de Putte et al., 2009;Van Poppel, Liefbroer, Vermunt, & Smeenk, 2001). While social endogamy is when the spouses originate from the same social strata, spatial/ geographical endogamy refers to marriages between partners from similar areas (Norberg & Åkerman, 1973;Vikström, 2003, pp. 185-209). Historically, endogamous marriage patterns have dominated in North Western Europe because past society was stratified by social position that impeded interaction across social groups and because young adults primarily migrated shorter distances, and such circumstances shaped the pool of partners available to meet and marry (Dribe & Lundh, 2005Lundh & Kurosu, 2014;. That marriages in the Sundsvall region during the 19th century were similarly made up by social endogamy has been shown by sociologist van  in addition to Swedish historians (Nilsson & Tedebrand, 2005;Norberg & Åkerman, 1973;Vikström, 2003). Some scholars contend that endogamy is the outcome because joint socio-spatial origins lead people to share common norms and values that promote a partner preference for each other (Dribe & Lundh, 2014;Hollingshead, 1950).

Marriages from disability perspectives
Although disability history has gained increasing interest since the 1990s, there are only a few works on the marriages of people with disabilities, while their attainment in education and the labor market have been more studied. Whether disability promotes endogamous or exogamous marriage patterns is hardly examined at all. De Veirman's thesis (2015) is one exception using demographic analysis to investigate the life courses of 284 persons being deaf or who had hearing difficulties in 18th-and 19th-century Flanders, out of whom only 48 married. De Veirman discusses how authorities considered the marriages of these individuals to be inappropriate or prohibited by law because disabilities were believed to be inherited and allowing their reproduction through marriage would risk the health of the population stock, which was a concern among many governments across the Western world. In Flanders, individuals having hearing disabilities married to a significantly lower extent and later in life (the average mean age was about 35 for men and 28 for women) compared to their hearing siblings, who De Veirman (2015) used as controls. The latter did not show such a large spouse age gap as did their siblings having hearing difficulties who were also less inclined to move up the social ladder through exogamous marriage. These results suggest that hearing disability limited the pool of partners to choose from. From a comparative study of the marriage of people with hearing disabilities in Flanders and the Sundsvall region, it appears that their spouses were usually not disabled (De Veirman, . This notion is echoed by historian Iain Hutchison (2007) in his study on disabled people in 19th-century Scotland, whose marriages tended to be organized so that the disabled partner could find support through their non-disabled spouse. Addressing pension files of war veterans, Blackie (2011Blackie ( , 2014) stresses the marriageability of this specific group of disabled men. Soldiers who became injured in the Revolutionary War in the US (1775-1783) married and headed households to a similar degree as did non-disabled veterans. These results do not depict disabled men as passive persons, Blackie argues, but 'force us to rethink popular ideas about disability and dependency ' (2014, p. 29). In another veteran study of soldiers from the First World War (1914)(1915)(1916)(1917)(1918), Joanna Bourke (2016) draws upon popular media images in the UK. She concludes that the marriages of impaired veterans were encouraged in British society to restore their broken bodies and manifest their masculinity through reproduction. For them, a nondisabled and respectable wife was regarded as the best match due to her nurturing and empathetic constitution by gender.
The past lives and marriages of disabled women are less examined than for men because most historical sources are gender-biased in their reporting. However, addressing narrative documents such as autobiographical accounts, scholars have gained insights into women's experiences of disability and gendered perceptions of it, occasionally pertaining to partnership (Burch & Rembis, 2014;Nielsen, 2009). Kudlick (2008) uncovers mainstream views of disability, womanhood, and marriage that blind women in 19thcentury France and the US were faced with. These views deemed them as unfit to marry while their male counterparts were encouraged to marry a non-disabled woman to enjoy the caring support that was linked to the female sex at the time. Kudlick further shows that these disabled women took on the view as being unworthy of a husband; having him help them cope with their disabling condition and domestic duties was unthinkable. Consequently, disability could affect partnerships differently in 19th-century society depending on the gender of the individual.
In a Swedish historical context, we have conducted life-course studies on the marital opportunities of disabled and non-disabled layers in a population of about 25,000 people in the 19th-century Sundsvall region Vikström et al., 2017). Disabilities made people less prone to marry, albeit with variations. While this was particularly true for those having mental disabilities, whose marital chances were about 80% lower compared to the non-disabled group, other impairments had lesser negative effects, but still being about 50% lower. For women, sensory limitations did not impede these chances as much as did physical disabilities, while the reverse held for men. This indicates that bodily defects particularly afflicted the female gender. The disability gap in marital opportunities was also slightly wider among women than men and for reasons that appear to not be directly linked to the impairment itself. First, in the study area the sawmill industry generated an increasing demand for a male workforce and a male surplus due to the influx of young men looking for employment (Vikström, 2003). This benefitted women's marriage in general, as did the fact that men are more inclined to remarry (Gaunt, 1983;Lundh, 2007;Nilsson & Tedebrand, 2005). However, this was less the case if the woman had a disability. Gendered expectations structuring the type of job and labor market probably contributed to the disadvantage of their disability in marriage, and this likely applied for men as well. The male-breadwinner ideal (Horrell & Humphries, 1995;Janssens, 1997) recognized men as the chief provider conducting manual work in agricultural or handicraft production or increasingly in manufacturing at the sawmills. Women were not assumed to take up jobs in the sawmill industry but to do household work or find employment in the service or catering sectors. Disabilities might not have hindered them entirely from performing such work tasks and thereby maintain some of their feminine appeal on the marriage market, while men were more affected if disabilities jeopardized their capability to perform hard labor and become a breadwinner.
Having studied about 70 disabled persons who married in the Swedish town of Linköping in 1714-1870, historian Olsson (1999) finds that they could marry if they held a job and had audiovisual or physical disabilities, while mentally disabled spouses were few. The average marriage age was rather high, about 30 for men and even 31 for women. Disabled persons did not marry each other and primarily showed socially endogamous patterns according to Olsson. Historian Sköld (2003) has researched marriages among people infected with smallpox, which could lead to disability and facial defects, and found that smallpox-infected persons tended to marry those afflicted by the same disease.

Area, data and methods
The source in this study was digitized parish registers from the 19th century stored at the DDB at Umeå University, Sweden. These consist of original records for birth, baptism, marriage, migration, death, burial and the catechetical examination records from 13 selected parishes in the Sundsvall region about 400 km north of Stockholm (Figure 1). In eight of the parishes, the economic structure was mainly agricultural during the entire 19th century, while the town of Sundsvall was the only urban setting. Along the coast, another four agricultural parishes came to depend on the expanding sawmill industry, which rapidly established itself from the 1860s onward. The regional population increase was most obvious due to declining mortality and the arrival of migrants (Edvinsson, 1992;Vikström, 2003). In 1840 there were 18,793 inhabitants, and that number increased to 46,418 in 1880 (Alm Stenflo, 1994).

Data from digitized parish registers and documentation of impairments
Because all parish registers are linked by the DDB they provide demographic data summarized at the individual level from different parish records (Vikström, Edvinsson, & Brändström, 2006;Westberg, Engberg, & Edvinsson, 2016). Those registers showing betrothals and marriages are key to this study because they provide details about the spouses. The catechetical examination records are similarly important and unique in an international comparison (Nilsdotter Jeub, 1993). In these records the ministers annually reported about the parishioners' knowledge of the catechism and their reading ability. There is also information on whether parishioners were poor or bore illegitimate children, which we made some account of in this study. Most of all, we made use of the ministers' remarks regarding their impairments (lytesmarkeringar) which have been largely under-researched (Drugge, 1988;Rogers & Nelson, 2003). We used these remarks to identify and categorize disabilities as shown in Table 1 and further discussed in previous work by us (Haage, 2017, pp. 77-94).
Defining disability is problematic because it is socio-culturally constructed depending on time and space, and the type of source used or person asked can matter (Grönvik, 2009;Kaplan, 1999;Mont, 2007). Scholars show that disability definitions have been applied arbitrarily and even misused historically, for example, by authorities using legislation or confinement to control citizens regarded as deviant or dangerous to society (Eggeby, 1993;Foucault, 1961Foucault, /2006Schweik, 2009). Although the ministers' impairment remarks cannot be completely comprehended today, they indicate parishioners who were distinguished from able ones, probably governed by perceptions of (dis)abilities at the time. To gain information of the health status of the entire population, Statistics Sweden also instructed all ministers to document impairments and with increasing consistency through guidelines, especially those from 1860 (SFS64, 1859), according to which ministers were to account for blindness, deafness, idiocy, insanity, and epilepsy. However, for many decades the ministers had already reported such information and whether someone was crippled, for example. Although this documentation is not clear-cut, we find it useful because it tells about who the people with disabilities were and about a comparatively large number of them. To avoid misinterpretations and translation issues, we primarily made use of the same concepts the ministers did but with no intention to offend, although many of these terms are considered derogatory today.
With some modifications to fulfill this study's purpose, the sample under analysis originated from the comprehensive dataset we used in a previous study of the marriage propensity by disability in the 19th-century Sundsvall region . First, this sample targets individuals who had impairments reported before they married and who married below the age of 50. Second, only those who married for the first time are accounted for, while their spouses may have married for their second or even third time. Third, we chose those born in 1800-1850 because the DDB digitization of the registers ends in the middle of the 1890s. This enabled us to follow people born in 1850 at least into their forties to look for marriages in the region. After these considerations, our sample resulted in 188 marriages of disabled people (119 men and 69 women) and their spouses. The male surplus is not because the disabled men were more prone to marry than their female counterparts , but because the ministers seem to have reported men's incapability to perform work if impaired more thoroughly than for women.

Methods and categorizations
Methodologically, we used descriptive statistics across gender and occasionally by distinguishing between disability types and social groups. Figure 2 shows the percentage distribution by disability type among the disabled men and women who married. About 75% of them had sensory or physical disabilities (i.e. groups of blind, deaf mute, Mental dysfunctions identified in adulthood having fully developed intellect as a child, e.g. insane, feeble-minded, and crazy Multiple disabled Combination of two or more of above disabilities Source: Digitized parish registers, the Sundsvall region, Demographic Data Base (DDB), Umeå University, Sweden Note: Concerning the distinction made between cases labeled 'idiots' and 'insane', see Drugge (1988), (Haage, 2017) and Statistics Sweden (1907). and crippled individuals). While cripples constituted the largest group, they were more pronounced among the men than women (45% vs. 36%). Some gender differences were further found among spouses having hearing or visual disabilities, while the distribution of men and women labeled 'idiots' and 'insane' was similar, some of whom appeared in the group with multiple disabilities if having physical and/or sensory impairments as well. Average estimations identified the marriage age and age gap between spouses, while cross-tabulations helped uncover whether couples shared geographical or socioeconomic backgrounds. The latter was measured in two ways. While the socio-economic origin was based on the father's occupation at the time the spouses married, the socioeconomic status refers to the occupational status of the spouses themselves. If a father had died before the wedding took place, his most recent work was used to define the socioeconomic origin. If the father was unknown from the registers, the mother's occupation was used if reported. Because some of the spouses had migrated to the Sundsvall region without their parents, the occupations of fathers or mothers were not always accessible in the DDB registers. This explains why we also took the occupations the spouses held themselves into account at the time of marriage or just before. Their socio-economic origin (father's occupation) and status (own occupation) were categorized according to the social groups in Table 2, which correspond to the HISCO model (van Leeuwen, Maas, & Miles, 2002). The DDB classification we used did not completely match the commonly used classification schemes in historical studies (SOCPO and HISCLASS), but there are many similarities (Edvinsson & Broström, 2012 two different social strata are exogamous. If spouses were born in the same parish, their geographical background was regarded as endogamous. Some individuals from the sample were selected to exemplify the results in a narrative way. They were not necessarily chosen to represent typical patterns, nor is there space to detail all demographic data regarding the selection at every occasion.

Results: marriage patterns and spouse characteristics
We first examine the age at marriage and the age gap between disabled spouses and their partners. A high marriage age or a profound age gap between the spouses would suggest that high levels of inequality were involved in the marriages. We next compare the social and geographical backgrounds to look for more (dis)similarities between the spouses. If exogamous patterns primarily characterized the couples, this would indicate that disability was associated with social distance in the partner pool because research shows that a strong preference for socio-spatial similarities governed partners resulting in endogamous marriages (cf. section 2.2). Comparing the spouses' characteristics accordingly would tell if disability implied disadvantages in marriage or had little impact on the couples concerned. First, however, did people with disabilities frequently marry each other? The answer is no, as exogamous patterns predominated in this respect (about 95%). Hence, they did not have to choose a spouse among each other as one possible consequence of disadvantage in the marriage market. This is less surprising considering that the entire group of disabled people, as identified from the parish registers in the area under study, comprised just about 1.5% of the young adult population (15-34 years old) originally selected for our disability studies . To meet and marry a peer of this minority group was not particularly likely if not being subject to high levels of marginalization. Among all the 188 cases, only 10 married another disabled spouse (5.4%), i.e. six men (out of 119) and four women (out of 69). Eric Magnus and Märta Greta married in 1873, for example, both of whom were deaf mute. Another couple was Lisa, who was blind, who married Pehr in 1839, who was labeled an 'idiot'. Obviously, legal restrictions regarding mental disabilities did not stop his marriage. Table 3 presents the mean and median ages at first marriage among the disabled men and women and their spouses, the latter of whom did not necessarily marry for the first time. The marriage ages were slightly higher among the disabled spouses compared to average measures in Sweden (Lundh, 1997) of men and women (27 and 25 years, respectively) and in the Sundsvall region (29 and 27 years, respectively) (Alm Stenflo, 1994;Vikström, 2003). Those we examined were slightly younger than the disabled spouses in Olsson's Linköping study (1999) and considerably younger than the deaf spouses in Flanders that De Veirman studied (2015) whose average age was about 35 for men and 31 for women. The age difference between the spouses introduced in this study was pronounced, as Stina was 24 years younger than Per (29 vs. 53 years old). Table 4 shows that such an age gap was not typical for disabled women who married, nor among disabled men. The latter married on average a wife who was almost 2.16 years younger, while husbands to disabled women were on average 1.34 years older. These gaps were not exceptionally wide nor as profound as among couples in Flanders involving deaf spouses that De Veirman examined (2015). That the disabled women married widowers to a higher extent than disabled men married widows (13% vs. 6%, respectively, Table 4) is a gendered pattern typically found in marriage studies (Gaunt, 1983;Lundh, 2007).

Marriage age and age gap between the spouses
Beyond average measures, a few outliers may always be found. Brita is one such case having married at the age of 16 and recognized as being weak-sighted. She was born in 1823 and the second child to a farmer; her older brother had died as an infant. When Brita was only one year old her father died, and her mother remarried a man who took over the farm. Three more siblings were born before Brita's stepfather died in 1836. In 1839, she married Olof who moved into Brita's household where her mother and younger brothers were living. Perhaps Brita's early wedding was encouraged by her mother to find a man able to run the farm and provide for herself and her children. Olof may have regarded Brita as a good match because he accessed a farm through her. In total, however, only five of the disabled women married before their 22nd birthdays and only four of the disabled men did so. These four young men worked as farmer, farmhand, shoemaker, or unskilled laborer, which suggests that they had an income to support a wife and family as discussed in the next section.  Source: Digitized parish registers, the Sundsvall region, Demographic Data Base (DDB), Umeå University, Sweden. Note: The age gap is calculated from the marriage age of the disabled spouse minus the age of his/her spouse. The negative figure for the mean age of disabled women is generated because their spouses were on average 1.34 years older.
One possible reason affecting age at marriage was unwed women's pregnancies. This could postpone marriage or ruin the marital chances, but it could also rush the wedding of couples having a relationship. About 25% of the disabled women bore a child within six months after the wedding date. This means that they were likely aware of their pregnancy before marrying, which may have advanced the decision to marry. That another 33% of the disabled women bore illegitimate offspring helps explain their high average marriage age because these children probably added some difficulties to contracting a spouse. However, these lone and disabled mothers eventually married, and about one in three married the father of the child. Accessing a male provider this way may in fact have been one survival strategy for some of them to find subsistence for themselves and their children.

Occupational structure and socio-economic background of the spouses
This section analyzes social exogamy to gain information on the distance between spouses and the extent to which the disabled people had jobs indicated by occupations in the parish registers. In all 188 couple cases but one, the ministers reported that both spouses or one of them, usually the man, held an occupation at marriage or just before. Figure 3 shows the distribution of occupations by social group among the disabled spouses and their partners, while Table 5 shows the most frequent occupations, which allocated the majority to the lower social strata where farmhands, cottagers, unskilled laborers, and maidservants are found. Table 5 further reveals that the percentage of cottagers was higher among disabled male spouses (18%) than among husbands to  Source: Digitized parish registers, the Sundsvall region, Demographic Data Base (DDB), Umeå University, Sweden Explanation: See Table 2.
Upper strata: Large-scale business entrepreneurs and higher civil officials Middle strata: Small-scale entrepreneurs in trade and industry, master artisans and craftsmen, farmers and tenant farmers Lower strata: Skilled laborers, craftsmen and artisans below the rank of master, unskilled laborers in trade and industry, farmhands, crofters and maidservants No occupation: Individuals reported as not holding any occupation at marriage disabled women (11%), while the latter men were unskilled laborers (24%) to a larger degree than the former (16%). These results reflect the notion that disabilities tended to impede people from finding jobs in industrial production, while this was less obvious within the agricultural sector or in handicrafts (Bengtsson, 2012;Oliver, 1996). Figure 3 shows no major distribution differences by disability for either gender. Disabled or not, almost 30% of the female spouses did not hold any occupation when marrying, while this held true for only about 7% of the disabled men and for about 4% of the grooms to disabled brides. That fewer and less detailed occupations are found among the female than male spouses is not surprising because parish registers reflect gendered ideals of their time by recognizing men's work more than women's ( Van de Putte & Miles, 2005;Vikström, 2010). However, these registers show that disabled people of both genders found jobs, and this indicates their working capability and suggests why they could marry.
Endogamous unions were to be expected according to previous research and because the above results show that most spouses belonged to the lower social strata. While father's occupation defined the socio-economic origin by group among the spouses as explained in the method section, their socio-economic status was based on the occupation they held themselves at marriage. Table 6 shows to what extent disabled spouses married a partner from the same social strata as themselves. Unfortunately, missing Table 5. The five most common occupations held at marriage by the disabled men and women (born 1800-1850) and by the spouses they married in the Sundsvall region.  (100) Source: Digitized parish registers, the Sundsvall region, Demographic Data Base (DDB), Umeå University, Sweden. Note: The percentages are based on the cases within gender and disability category. Table 6. Distribution of social endogamy/exogamy regarding the marriage patterns based on the socio-economic background of the disabled men and women (born 1800-1850) and of the spouses they married in the Sundsvall region. Source: Digitized parish registers, the Sundsvall region, Demographic Data Base (DDB), Umeå University, Sweden. Note: The percentages are based on the cases within each gender and disability category. Socio-economic origin refers to fathers' occupations of the individuals when marrying or just before, while socio-economic status refers to the occupation the spouses held themselves when marrying or just before. Unknown cases appear when occupational data are missing in the parish registers or are incomplete concerning the fathers, one of the spouses or both, thus preventing comparisons from being made.
occupational data generate a great many cases that cannot be compared, but when possible, social endogamy predominates the partner selection. Exogamous unions were few but revealed that disabled spouses could marry both downwards and upwards, and that particularly men did the latter. Disabled women tended to marry downwards primarily because their father's occupation due to older age was of higher status than that of their husband. Tables 7 and 8 differentiate exogamous and endogamous patterns accounting for the socio-economic origin (father's occupation). Table 7 shows that every second disabled man from families in the upper/middle strata contracted a woman from the same strata (51%), while 22% married a woman from the lower strata. Similarly, about half of the disabled grooms from the lower strata married brides sharing their social origin (52%), while 25% contracted a woman whose father belonged to the upper/middle strata (27%). The disabled women demonstrated a different pattern according to Table 8 that adds to explaining their exogamous patterns seen in Table 6. Only one woman in three (33%) from the upper/middle strata married a spouse matching their father's social standing. However, and despite the missing cases, it was primarily similarities in socio-economic background that characterized couples who married, and this was also the case when one spouse was disabled.
In 1860, for example, Olof married Lisa. Olof was noted to be weak-sighted since he was 15 years old. At marriage, he held a position as a farmhand while Lisa was maidservant. Their own occupational status allocated them to the lower strata, as did that of their Table 7. Socio-economic origin (father's occupation) at marriage of the disabled men (born 1800-1850) and of the spouses they married in the Sundsvall region.  fathers. Olof's father was a boatswain, while Lisa's father was a cottager. This makes them representative of endogamous patterns. Eric and Juliana exemplify a more complex social background. Eric was a tailor and a had a stammer since his youth. In 1844, he married Juliana, a maidservant. Their occupational status allocated them to the lower strata. However, Eric's father was a cottager while Juliana came from the upper/middle strata having a father who was a farmer.

Geographical background of the spouses
This section examines whether the disabled men and women were born in the same parish as their partner, making some account of adjacent parishes. According to Figure 4, most disabled spouses married someone born in the same parish as themselves situated in the Sundsvall region, or a parish nearby. This 'close distance' between the spouses characterized well over 60% of the men and slightly more than 50% of the women and suggests that they took part in the local partner pool and were familiar with each other for a long time. The men showed a higher preference than the women for a partner born in the same parish (40% vs. 30%, respectively).
In only four cases did both spouses originate from parishes beyond the Sundsvall region. Erik and Maria were born in the province of Värmland, near the Norwegian border. Erik was born in 1834 in the parish of Sunne. In 1860, he moved to Hässjö parish in the region and took up employment in a sawmill. The minister later reported Erik's loss of one arm, but Erik kept on his work and at the age of 37 he married Maria, a maidservant born in a parish nearby his parish of birth. Their wedding was held shortly upon her arrival in Hässjö in 1870.  Source: Digitized parish registers, the Sundsvall region, Demographic Data Base (DDB), Umeå University, Sweden Explanation: Same: Spouses born in the same parish NeighP1: Spouses born in neighboring parishes with one parish border between NeighP2: Spouses born in neighboring parishes with two parish borders between NeighP3-4: Spouses born in neighboring parishes with 3-4 parish borders between Sweden: Spouses born in parishes in Sweden with more than 4 parish borders between Abroad: One of the spouses born in another country than Sweden Unknown: Birth parish not known for one of the spouses or both.
In about 30% of the couples, the spouses were born in parishes farther away from each other. One such common union was when one of them was born in Stockholm and the other in the Sundsvall region (10 couples out of 45). Even if these unions manifest exogamy, some exhibited more endogamous patterns, as Gustava and Nils Olof illustrate. Gustava departed from an orphanage in Stockholm when she was nine years old and came to the Sundsvall region reported as having hearing difficulties, where a childless cottager couple in Njurunda parish cared for her. When her foster father died, Gustava married Nils Olof, who was born and lived in Njurunda, and they then took over the cottage.
About 4% of the disabled men and women married someone born in another country, three of whom were from Finland, two were from Norway, and one was from Germany. However, only one disabled spouse was from abroad, and his name was Adolf, born in Finland in 1807. Since the age of four, he had spent his life in the Sundsvall region (Tuna parish) where the minister reported him as being crippled and his father to be in miserable conditions having died in 1813. The following years, Adolf lived with his mother and siblings in the parishes of Tuna and Indal. They were poor according to the parish registers, and in 1816 the mother gave birth to an illegitimate son. In 1822, when Adolf was 15 years old, he took up work as a farmhand. For many years he held this position serving different farmers. Adolf married rather late in life being 40 years old, to Sigri Cajsa, a maidservant, originating from a parish in the Sundsvall region.

Concluding remarks
There is little knowledge on the marriages of disabled people in the past, not only because they are overlooked in historical sources, but also because disability is often perceived to make people less possible as partners. We know from previous research including our own studies that disability significantly impeded people's marriage as a possible consequence of social marginalization in society, yet some of the disabled married. This study examined who they were and who they married in order to find further clues toor even questionhow disability distanced people from social integration in past society. Digitized parish registers were used to investigate 188 marriages of disabled persons and their spouses in the 19th-century Sundsvall region, Sweden. Below, we discuss the major findings and a few limitations coupled with our analysis that should be made up for in future research.
First, we find that disabled individuals who married rarely married each other. This lack of endogamous pattern suggests that the spouses we study participated in the regular partner pool and were not faced with exclusion from society. That primarily men and women with sensory or physical disabilities were present among the 188 spouses, while substantially fewer were labeled mentally disabled, indicates that disability type and attitudes towards it mattered for marriage. Second, in general and even if disability resulted in a slightly higher marriage age, it did not imply any marked age gap between the spouses. This finding also speaks in favor of the disabled spouses taking part in the regular marriage market. Third, having examined the socio-spatial backgrounds of the couples, similarities in terms of close distance tend to have governed the choice of partner. According to historical research such endogamous patterns are typical for spouse selection, and disability did not make those we studied differ substantially. Fourth, we find only a couple of gender differences worth mentioning regarding the marriage patterns.
Disability tended to make women marry later than average women did, while this was less obvious among the men, and disabled women moved slightly more downwards on the social ladder when marrying than did their male counterparts.
It should be borne in mind that our results are based on a more limited number of disabled women who married than men (69 women vs. 119 men). Nevertheless, there is suggestive evidence that reject rather than confirm marginalization in the partner pool, because the disabled spouses primarily demonstrate typical patterns regarding age at marriage and partner choice and not the reverse. Although they constitute a special selection among people with disabilities in having married, they extend their levels of agency and possibilities that have been largely neglected in historical research or proven difficult to trace. Our marriage results further stress the recognition of disabled people as companions 'for life' among potential peers in the partner pool and surroundings. Other than one of the spouses having a disability, the couples resembled others concerning their demographic and socio-spatial characteristics at marriage.
Unfortunately, there is sometimes incomplete data in the parish registers regarding the occupational background, nor do the registers fully detail the impairments or how severe these were. However, these registers are exceptional in documenting disabilities and provide demographic data on the parishioners having them. Unlike our previous studies, the present study did not conduct advanced statistical analyses or case-control comparisons, which would have helped us to assess the results beyond references to other studies and the predominant picture of endogamous patterns they paint. Partly coupled with this limitation is that our findings are based on a positive selection only showing disabled persons who married. Most likely they were capable of providing for themselves and were regarded as possible partners by the surrounding society in contrast to their disabled peers who never married.
However, our sample of spouses can also be viewed as a strength because they belong to the heterogeneous group of disabled individuals whose marriages have been underresearched. One major rationale for this study was to contribute to the debate of the agency of disabled people in history, and this sample enabled us to recognize their opportunities more than their obstacles. Even if not being representative for the whole group of disabled people, our results are of empirical and theoretical relevance in differentiating the human experiences of disability. While disability tends to substantially lower the chances to socialize through marriage, our study constitutes an important reminder that disability is not necessarily associated with marginalization from social life and society. This is not to say that marriage brought joy or social recognition to all, disabled or not, and complex family ties and socioeconomic circumstances played part in the marriages we examined as indicated by the individual cases highlighted above. Recall young Brita, for example, whose wedding was most likely settled by her widowed mother, or the marriages of the women having both disabilities and illegitimate children. Perhaps marriage was one means for some men with disabilities, or for their parents, to access free help for them from a supportive wife. An even closer look at the individual and familial level of analysis would show how vulnerabilities as well as survival strategies governed the marriages and partner selection among people with disabilities, and how gender influenced these patterns. Yet another big issue to clarify is how their family life developed concerning reproduction. The focus and findings of this study will encourage such research and a further differentiation of stereotyped views that persist still today regarding the possibilities of people with disabilities.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
This study is part of a project headed by Lotta Vikström that has received funding from the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation programme [Grant Agreement No. 647125

Data availability statement
For more information about the availability of the data used in this article (19th-century Swedish parish registers digitized by the Demographic Data Base, DDB), visit the below websites at the Centre for Demographic and Ageing Research (CEDAR), Umeå University, Sweden: https://www.umu.se/en/centre-for-demographic-and-ageing-research/databases/parish-regis ters-databases/ https://www.umu.se/en/centre-for-demographic-and-ageing-research/search-tools/