ARTICLE /ARTICLE
Marriage patterns in the Western French Pyrenees during the 1800-1899 period: data from the village of Béost (Ossau Valley, Béarn)*
Comportements matrimoniaux dans l’ouest des Pyrénées françaises durant la période 1800-1899 : données du village de Béost (vallée d’Ossau, Béarn)
J.P. Dugène · F. Bauduer
Revised: 20 February 2012, Accepted: 12 November 2012
© Société d’anthropologie de Paris et Springer-Verlag France 2012
AbstractThe Western Pyrenees are made up of isolated val- leys that were associated until recently with low marriage mobility. We chose to study the Ossau Valley (Béarn, France), which has one of the least mobile populations in this area.
Marriage patterns in the village of Béost were analyzed over the 1800-1899 period. Data on 238 marriages obtained from parish, civil and notarial registers were used to assess the fol- lowing variables: seasonality, endogamy rate, isonymic mar- riages and the consanguinity coefficient calculated from iso- nymy and from ecclesiastical dispensations. A clear seasonal marriage pattern emerged, linked to religious and local social parameters. Levels of endogamy were 57.5% in the village and 92.4% in the valley as a whole. The consanguinity coeffi- cients calculated from isonymy (Ft) and ecclesiastical dispen- sations (α) were 0.0031 and 0.0040 respectively. The negative non-random component of F (Fn) indicated that Béost resi- dents avoided contracting potentially consanguine isonymic marriages. These biodemographic markers were compared to those observed among various other 19thcentury mountain populations in Western Europe.
KeywordsPyrenees · Béarn · Marriages · Consanguinity · Isonymy
RésuméLes Pyrénées de l’ouest sont composées de vallées isolées qui ont été associées jusqu’à récemment à une mobi-
lité matrimoniale réduite. Nous avons choisi d’étudier la val- lée d’Ossau (Béarn, France), l’une des populations les plus
« fermées » de cette zone. Les comportements matrimoniaux ont été analysés dans le village de Béost entre 1800 et 1899.
Les données concernant 238 mariages obtenues à partir de registres paroissiaux, d’état civil et notariés, ont été utilisées pour évaluer les variables suivantes : répartition mensuelle des mariages, taux d’endogamie, proportion d’unions iso- nymes, et coefficient de consanguinité (à partir des taux d’isonymie et de dispenses ecclésiastiques pour apparente- ment). Nous avons noté une saisonnalité des mariages en relation avec des paramètres religieux et sociaux locaux.
Les taux d’endogamie étaient respectivement de 57,5 % pour le village et de 92,4 % pour la vallée. Les coefficients de consanguinité ont été estimés à 0,0031 à partir du taux d’isonymie (Ft) et à 0,0040 selon le taux de dispenses ecclé- siastiques pour apparentement (α). La valeur négative du taux de consanguinité non lié au hasard (Fn) indique que les habitants de Béost évitaient les unions isonymes suscep- tibles d’être consanguines. Ces paramètres matrimoniaux ont été comparés avec ceux décrits dans d’autres populations de vallées de montagne d’Europe occidentale au 19esiècle.
Mots clésPyrénées · Béarn · Mariages · Consangunité · Isonymie
The endogamy rate is an important variable in population genetics [1]. Patterns in marriage may be used to study this parameter. Mountain populations are interesting models where reproductive isolation is usually linked to the geo- graphical environment [2-5]. Socio-professional parameters have also been shown as important factors [6,7]. Several studies have been conducted on marital characteristics among some European mountain populations, especially in the Western part of the Alps and the Pyrenees [3, 6-17].
These suggest a remarkable level of geographical endogamy in comparison with non mountainous zones, as well as
J.P. Dugène
Maison d’Ossau, F-64260 Arudy, France F. Bauduer (*)
Unité d’Anthropologie biologique, Laboratoire MRGM, EA 4576, Université Bordeaux Segalen,
F-33076 Bordeaux cedex, France e-mail : bauduer.frederic@neuf.fr
*This paper was presented in part at the annual meeting of the Société d’Anthropologie de Paris, Bordeaux, 25-27 January 2012 DOI 10.1007/s13219-012-0075-1
several characteristics (such as seasonality) that are related to local environmental or sociological factors. The Pyrenees are made up of isolated mountain valleys, with ancient autochthonous populations living in their Western extremity, especially the Basques who could represent the remnants of a pre-Neolithic European group [1, 18]. The Ossau Valley is thought to be home to one of the most isolated populations in the Western Pyrenees [9]. Interestingly, no biodemographic studies have been published for this valley so far, except the work of Fresel-Losey conducted in the village of Bilhères [19]. In the Béarn, as in the neighbouring Basque Provinces, household structure is ethnologically specific. Traditionally, there was a system of indivisible inheritance; it was custom- ary for the eldest child, regardless of gender, to be the sole heir and preferably not to marry a first-born person [10, 14, 20-23]. The 19th century is a very interesting period for anthropological study because it precedes the so-called
“demographic transition” in this part of Western Europe, and a number of reliable sources of data are available from this time [24]. This period is associated with a high rate of consanguinity, especially in isolated mountainous zones [2-8]. Several papers have been published on this topic, on Alpine valleys especially, but the Pyrenean zone has been less extensively researched [2-17]. The aim of this study is to investigate the marriage pattern in a western Pyrenean village during the 19thcentury, in order to analyse several biodemographic indicators such as age at first mar- riage, endogamy rate, consanguinity coefficient and isony- mic marriages, and to compare our data with those obtained from other European mountain populations. This study should enable us to distinguish between regionally specific traits and general characteristics observed elsewhere.
Material and methods Population studied
Béost is a French village located in the Ossau Valley, in the south-western part of the Pyrénées-Atlantiques Départe- ment, in the Aquitaine administrative region (altitude:
534 m, latitude: 42°59’35’’ north, longitude: 0°24’58’’
west, 210 inhabitants in 2006). The Ossau Valley covers 6161.9 km2and borders the Ouzom Valley to the east (vil- lages of Arbéost and Ferrières) and Spain to the south. Sev- eral villages lie about 20 km distant from Béost (Fig. 1).
Laruns, covering 2500 km2, is the third largest village in France in terms of surface area. The Ossau Valley may be considered as geographically isolated with no significant migratory changes until the beginning of the 20thcentury.
In the 19thcentury, the population’s occupations were almost entirely centred on pastoral activities (sheep) in association with subsistence agriculture. The population of Béost lives
in the village itself and in the nearby hamlet of Bagès (about 350 m distant as the crow flies, and 1500 m by road). The neighbouring villages are Louvie-Soubiron to the north, Lar- uns to the west, Eaux-Bonnes to the south and Arbéost (Ouzom Valley) to the east. A large portion of the village’s Fig. 1 Distribution of the villages in the Ossau Valley (distance in km to Béost given in brackets) / Distribution des villages au sein de la vallée d’Ossau (distance en km avec Béost indiquée entre parenthèses). [A: Arudy (14.0), A.B.: Aste Béon (4.5), Be.:
Bescat (17.0), B.: Bielle (9.0), b.: Bilhères (11.0), Bu.: Buzy (19.0), C.: Castet (14.0), E.B.: Eaux Bonnes (4.0), G.B.: Gère Belesten (5.0), I.: Izeste (12.0), Laruns (1.8), L.: Lys (23.0), L.J.:
Louvie Juzon (12.0), L.S.: Louvie Soubiron (1.4), S.C.: Sainte Colome (16.0), S.: Sévignacq Meyracq (15.0), Ouzom Valley:
Arbéost (35) and Ferrières (38)]
mountain pastures lies within the Ouzom Valley. The name Béostappears for the first time in 1355 (15thcentury manu- script - Archives of the Pyrénées-AtlantiquesDépartement).
In 1385, under the reign of Gaston Fébus (the local ruler), twenty five families were registered in Béost itself and seven in Bagès. The surnames of some individuals who lived in the village during this period are known and most of them still exist today. Because herding was the main occupation of spouses, with the men farming and women responsible for running the household, we did not categorize our sample by socio-professional class. As all the residents were Roman Catholics, weddings were systematically recorded in the parish register by the local priest.
Sources of data
All marriages registered in Béost from 1800 (January 1) to 1899 (December 31) were collected. To ensure their exhaus- tiveness, we collected data from several complementary sources:
•
Parish registers and dispensations for consanguinity in the case of unions between cousins up to the fourth degree, when available (these dispensations have been mandatory for Catholics since the Council of Trent, 1545-1563),•
Official civil registers (from the Béost city hall and the archives of the Pyrenees Atlantiques Département (ADPA),•
Notarial recordsDuring the 19thcentury, a fire destroyed the city hall and its records for the 1874-1883 period. Data for the missing decade were obtained from the ADPA and from parish and notarial registers. There was a perfect match between the data when these various sources were compared, except for one couple. In this case, the notarised premarital agreement was not followed by a wedding and this case was therefore not included in the study. Overall, our files included at least the surnames of the spouses, their dates of birth (in 90%
of cases), their places of birth and residence, and the wed- ding date. In order to avoid problems of duplication due to wrongly registered multiple surnames corresponding to a single lineage (because of spelling mistakes), all the spelling versions of surnames within families (including in birth reg- isters) were listed using Excel®. Signatures were also exam- ined when available (about 93% of men and 99% of women in the Béarn were able to sign their names by the end of the 19th century [19]). In contentious situations, the surname spelling version chosen was either the most common or the oldest. Demographic data for the period were also collected from the official municipal population censuses periodically performed in every French city since the year 1801 follow- ing their introduction by Lucien Bonaparte.
Biodemographic parameters
The marriage rate (i.e. the number of marriages per thousand in a population in a given year) was determined. We col- lected the age of spouses at marriage. Seasonality over the total study period was analyzed by recording the number of marriages per month. The number of marriages was expressed as N per 1200 and corrected for the number of days in each month in order to obtain an expected number of marriages per month of 100 with no seasonality (monthly index) [25]. The percentage of endogamous marriages (village and valley endogamy) according to the place of res- idence and the straight-line distances between the birth and residence parishes of the spouses were determined. The con- sanguinity coefficient was estimated in two different ways.
First, we assessed the percentage of isonymous mar- riages, the total consanguinity coefficient from isonymy (Ft), random (Fr) and non-random (Fn) mating according to the method developed by Crow and Mange [26] and adapted by Crow [27]:
Ft = Fn + (1–Fn) Fr
Fn = (P -∑piqi) / 4 (1 -∑piqi) Fr =∑piqi / 4
Because marriage registers were used to represent the population,∑piqi can be calculated as:
∑piqi =∑mifi/Nm2
Abbreviations: pi: proportion of the male (pi) and the female (qi) population with a given surname, P: isonymic rate, mi and fi: numbers of male and female spouses respec- tively with surname i, Nm: number of marriages (squared because male and female numbers are equivalent).
Second, based on the percentage of consanguineous mar- riages associated with the degree of kinship (first, second, third or fourth cousins) indicated in the parish registers, we calculated the consanguinity coefficient from ecclesiastical dispensations with :
α=ΣipiFi[28]
where pirepresents the frequency of consanguineous mar- riages that would result in offspring with a consanguinity coefficient Fi. Fi varied from 1/8 (union between uncle/
aunt and niece/nephew) to 1/1024 (union between two fourth degree cousins) [28].
Results
Demographic data
Between 1800 and 1899, the population of Béost varied between 314 and 410 according to the data from 17 censuses (Table 1). After a marked increase at the beginning of the period, the population gradually declined, as observed in all the villages of this Pyrenean zone [19]. Three hundred
and thirty four marriages including at least one spouse origi- nating from Béost or Bagès were registered during the study period (1800-1899). Two hundred and thirty eight weddings were celebrated in Béost (71.25%) and 96 (28.75%), where one spouse was from Béost, elsewhere in the Ossau Valley.
The marriage rate did not parallel the population demogra- phy. The number of marriages per 10-year period ranged from 20 to 32, except during the wartime period of 1870- 1879 and in 1890-1899 (Table 1). 137 marriages were between spouses both originating from Béost or Bagès. All
of the spouses living in Béost were born in the village (the place of birth was not available for all individuals from other villages). Therefore, Parish (or village) endogamy, consider- ing both birth and residence, was 57.5 %. The age of the spouses ranged from 16 to 66 years for men and 16 to 47 years for women. Regarding first marriages and given that the age was not recorded in 22 cases, the mean age of the spouses was 31 years and 3 months for men and 25 years and 10 months for women (Fig. 2). In forty-seven marriages, the woman was older than the man. There were eleven
Table 1 Distribution of marriages in the village of Béost and in surrounding places from 1800 to 1899 involving at least one spouse from Béost (per 10-year period) /Distribution des mariages dans le village de Béost et dans les communes avoisinantes concernant au moins un conjoint de Béost entre 1800 et 1899 (par décennie).
Decades 1800
1809
1810 1819
1820 1829
1830 1839
1840 1849
1850 1859
1860 1869
1870 1879
1880 1889
1890 1899
Béost total population* 314 No census 404 410 390 326 378 373 358 323
Marriages in Béost 26 33 22 26 20 31 26 16 22 16
Marriage rate in Béost (per annum)
7.28‰ NA 5.44‰ 6.40‰ 5.39‰ 8.73‰ 6.82‰ 4.30‰ 6.23‰ 5.00‰ Marriages elsewhere
in the Ossau Valley including a spouse originating from Béost
11 12 6 9 12 10 10 3 12 11
NA: not assessable (no population census available during this period); *at the beginning of each given decade.
Fig. 2 Mean age of spouses at first marriage in Béost from 1800 to 1899 (10-year periods) [N = 202] /Age moyen des époux au premier mariage à Béost entre 1800 et 1899 (périodes de 10 ans) [N = 202]
second marriages of widowers at 49, 59, 40, 49, 42, 66, 39, 55, 40, 52, and 45 years of age and three second marriages of widows at 45, 38, and 38 years of age.
Seven families were involved in 97 marriages between people living in an area of 330 m. The marital mobility pat- terns are shown in table 2. Between 1800 and 1899, 14 migrants (i.e. external to the Ossau Valley) came to Béost for matrimonial reasons. All of them came from a place less than 40 km distant except one bridegroom from Lesperon in the Landesdépartement(150 km). The valley endogamy was 92.4%.
Distribution of marriages
The marriage rate ranged from 4.3 to 8.7‰ (Table 1). A particular seasonal pattern of marriage was found (Fig. 3):
the frequency of marriages, as estimated with the monthly index, was highest in November (monthly index: 337), in October (187), i.e. after the‘Saint Michel fair’(September 29) during which the sale of animal or agricultural products earned money for marriage contracts, and in February (152).
The lowest frequencies were associated on the one hand with religious restrictions in this Catholic zone, i.e. March/Lent period (19), May/month consecrated to Mary (29), Decem- ber/Advent period (44.3), and with herding activity, i.e.
August (19) (summer migration of men with their herds in the mountains). This pattern is specific to this village and differs from that of other mountain communities in which spring was traditionally the preferred period for weddings [25,29,30]. The religious ceremony was usually celebrated
on the same day as the registry office wedding or during the following week.
Consanguinity coefficient from patronymic data We obtained 132 different surnames from the 238 marriages celebrated in Béost. Their distribution is shown in table 3.
Sixty (45.4%) of the surnames were unique, whereas the commonest surname was found in 23 individuals. The fre- quency of isonymous marriages was 2.20%. The consanguin- ity coefficients were respectively: Ft = 0.003141, Fr = 0.003928 and Fn = - 0.0007899. Seven families had the high- est number of marriages (97 i.e. 40.7%) within the village of Béost. Four of these families occupied 7 of the 31 ‘stem houses’registered in 1385 (census ordered by Gaston Fébus).
Consanguinity coefficient from ecclesiastical dispensations
Of 145 marriages for which detailed parish records were available, 38 (26.2%) were associated with an ecclesiastical dispensation for consanguinity. From these data, the coeffi- cient of consanguinity (α) was calculated at 0.0040. The dis- tribution of these dispensations was not homogeneous throughout the study period (35 out of 109 marriages from the end of 1849 compared with 2 out of 41 between 1830 and 1849). The end of 1849 was the date of arrival of a new priest who seems to have adopted a stricter policy than his predecessor. The majority of these dispensations were given for unions between third or fourth degree cousins.
Table 2 Matrimonial migration to or from Béost during the 1800-1899 period /Migrations matrimoniales vers ou en provenance de Béost entre 1800 et 1899.
Villages Arrivals in Béost Departures from Béost Total migration pattern
M F T M F T M F T
Aste-Béon 2 7 9 2 2 4 0 + 5 + 5
Gère-Bélesten 3 3 6 4 1 5 - 1 + 2 + 1
Castet 0 0 0 0 1 1 0 - 1 - 1
Izeste 0 0 0 2 1 3 - 2 - 1 - 3
Sainte Colome 0 0 0 3 0 3 - 3 0 - 3
Louvie-Juzon 0 0 0 3 0 3 - 3 0 - 3
Louvie-Soubiron 3 21 24 11 16 27 - 8 + 5 - 3
Aas, Assouste, Eaux-Bonnes 3 8 11 4 16 20 - 1 - 8 - 9
Laruns 19 2 21 7 23 30 + 12 - 21 - 9
Total (within the Ossau Valley) 30 41 71 36 60 96 - 6 - 19 - 25
Other villages in thePyrenees Atlantiques
2 4 6 0 0 0 + 2 + 4 + 6
Villages in theHautes Pyrenees 6 8 14 1 1 2 + 5 + 7 + 7
Villages in theLandes 1 0 1 0 0 0 + 1 0 + 1
Abbreviations: M: male, F: female, T: total; Villages in the Ossau Valley are given in bold.
Discussion
Patterns of marriage are interesting indicators in population biology. Mountain populations offer one of the best models of geographical reproductive isolation.
The results obtained in this study on the Pyrenean village of Béost during the 1800-1899 period need to be compared
with those from other European mountain communities (Table 4). First of all, the stringent methodology used for data collection must be underlined. The combination of var- ious sources strengthens the reliability of our results and enables us to calculate the main biodemographic indicators.
The marriage rate we found is low (Table 1) compared to the rest of France (16‰in 1825-1870 [30]) but quite similar to that of Alpine populations (reviewed in [31]). This repre- sents a general characteristic of mountain populations during the 19thcentury.
A study performed by Marquer [9] on 1051 individuals from the Béarn Province in the 1960’s showed that the place of birth was nearly always the place of family origin. This finding is confirmed in our study. The percentage of birth endogamy is also identical to that of residence endogamy in the Non Valley [4] and in the Upper Sole Valley [3] in the Eastern Italian Alps. Marquer [9] also found that the Ossau Valley had the highest village and valley endogamy index in the Béarn Province. Village endogamy in Béost was 57.5%. This percentage does not seem particularly high when compared with other mountainous places in the same period. For instance, in Törbel, a village in the Swiss Alps, about 80% of all marriages were endogamous to the village until the 1940s [32], and 79.9% in the French Dauphiné region during the 19thcentury [16]. Our result is comparable to those reported for seven parishes in the Non Valley by Gueresi et al. [4]. However, at 92.4%, the level of valley Fig. 3 Number of marriages per month (plotted on the Y axis) and monthly index (in brackets) in Béost (1800-1899). The sum of monthly indexes is 1200 (seematerials and methods) /Nombre de mariages par mois (indiqué en ordonnées) et index mensuel (entre parenthèses) à Béost (1800-1899). La somme des index mensuels est 1200 (voir matériels et méthodes)
Table 3 Distribution of the 132 surnames among the 238 mar- riages celebrated in the village of Béost between 1800 and 1899 /Distribution des 132 noms de famille parmi les 238 mariages célébrés dans le village de Béost entre 1800 et 1899.
Surname frequency Number of names 1 (i.e. unique names) 60 (45.4 %)
2 22
3 15
4 9
5 3
6 2
7 3
8 2
9 3
10 2
11 2
12+ 9
endogamy is among the highest ever published [16]. Regard- ing this finding, we are not aware of any genetic particularity related to a founding effect in the Ossau valley similar to that which appears in various neighbouring Basque zones [18].
In the Western and Central Pyrenees, the endogamy index began to decrease significantly from the First or the Second World War depending on the geographical zones [13]. In the 19th century, more than 50% of men and about 45% of women never left their villages [23]. In contrast, Viazzo showed intense temporary mobility, sometimes between long distances, in his study on Alpine populations from the 16thto the 19thcentury [7]. Therefore, the traditional image of mountain communities as passive and closed should not be seen as applying everywhere.
Here, the Ft value fits relatively better with theα value than in the other published series (Table 4). The high per- centage of isonymy in this population explains the relatively high value of the random consanguinity component (Fr).
This is mainly influenced by the fact that relatively few sur- names exist in a small-sized population [33]. Furthermore, the expected number of isonymic marriages is greater than the observed number, as reflected in the negative values of Fn. The negative Fn indicates that Béost residents avoided contracting isonymic marriages that could be consanguine- ous. This practice has been observed in other 19thcentury mountain populations, for instance in the village of Törbel in the Swiss Alps [32] or in the Salazar valley in the Spanish Province of Navarre [34]. Our consanguinity coefficient Ft,
determined using isonymy frequency, is 0.0031, which is slightly lower than the range in the selected series shown in Table 4 but higher than in the population of Törbel in 1825 (0.0029) [32] or Salazar Valley during the 1601-1981 period (0.0022) [34]. The consanguinity coefficient (α) calculated from ecclesiastical dispensations is slightly higher, at 0.0040. There are many biases to consider when calculating the consanguinity coefficient from either isonymy or eccle- siastical dispensations. With the first approach, illegitimacy and adoption must be taken into account as well as mis- spelled surnames and polyphyletism [35]. In addition, in the Pyrenean inheritance system, naming is more “house- attached”than“family-attached”[14,19,21,22], so that indi- viduals may change their name (for instance when a male who does not inherit marries a female heir and may adopt the name of her house). Theαconsanguinity coefficient esti- mated from ecclesiastical dispensations is mainly influenced by consanguineous marriages between first cousins. In addi- tion, the rate of dispensation is closely related to the ecclesi- astical policy, as clearly shown here. Thus, in the parish records from Béost, there was a significant increase in dis- pensations after 1859 with the arrival of the new priest.
Overall, the ecclesiastical dispensations method tends to underestimate consanguinity [36] while the isonymy method overestimates this parameter [4]. Moreover, Bideau et al showed that consanguinity did not occur in the entire popu- lation of the Valserine Valley (Jura, France) but in a sub- group of families in which spouse selection and mobility Table 4 Biodemographic indicators on marriage during the 19thcentury among some European mountain populations /Indicateurs biodémographiques matrimoniaux concernant quelques populations européennes de montagne au 19esiècle.
Place Period Consanguineous
marriages
Consanguinity coefficientα
Ft Fr Reference
Ossau Valley (Pyrenees, Béarn)
1800-1899 26.2% 0.0040 0.0031 0.0039 This study
Non Valley
(Eastern Italian Alps)
1825-1923 5.8-23.1% 0.0019-0.0045 0.0073-0.019 0.0029-0.020 [4]
Upper Sole Valley (Eastern Italian Alps)
1825-1923 13.0-40.3 % 0.0023-0.0096 0.0068-0.039 0.0077-0.031 [3]
Fersina Valley (Eastern Italian Alps)
1800-1914
Italian speakers 9.6-25.5% 0.0018-0.0053 0.0059-0.020 NR [44]
German speakers 27.5-51.5% 0.0047-0.0092 0.013-0.024 NR
Roveto Valley (Central Apennines)
1876-1900 9.0-35.2% 0.0016-0.0087 0.014-0.053 NR [12]
Orozco Valley (Pyrenees, Basque Country)
1880-1904 59.7% NR 0.0023 NR [45]
Gredos Valleys (Central Spain)
1875-1894 NR 0.0007-0.0026 0.010-0.019 0.0047-0.010 [39]
NR: not reported.
were different. These consanguineous lineages have demon- strated a high prevalence of Rendu-Osler disease, a domi- nant inherited condition [36]. This point cannot be assessed here because of our sample size. In the majority of European rural populations, the timing of marriages during the year is mainly influenced by the work calendar and by climate and religious concerns (for Catholics, the month of May consecrated to Mary and the Lent and Advent periods) [25,29]. This is the case here, with a specific pattern related to the local fair at the end of September, which was condu- cive to the writing of marriage contracts.
Regarding marriage, a particular social, family and spatial system prevailed in the Western Pyrenean zone for more than ten centuries, which aimed to conserve heritage down the generations [14,19,22,37]. This finding contrasts with the pattern of inheritance based predominantly on historical and political factors, according to the data collected by Cole and Wolf for a study in an Alpine valley [38]. In the Pyre- nees, a family’s entire inheritance was given to the first-born child (male or female), who had to marry the youngest child of another family. The idea was to avoid a multiplicity of houses and to preserve sufficient working capacity. This tra- ditional rule known as‘droit d’ainesse intégral’was abol- ished in 1804 by a French law that instituted strict equality between all heirs. However, this ancestral system has per- sisted until now in the Basque area and in some neighbour- ing Pyrenean areas including the Béarn [37]. As observed in our Pyrenean population, patterns of land division have been reported to influence spouse selection processes, as observed in other European populations [39]. It is well known that marriage practices in Europe evolved to prevent the frag- mentation of land [40]. Marriage was a contract based on conditions fixed by the parents of the spouses. Marrying possibilities developed as way of regulating the population (an average of two marriages per family was usually permit- ted). The historically very high childhood mortality rates decreased with progress in lifestyles and medicine, and the increasing numbers of potentially marriageable individuals became a major factor in the destabilisation of this ancestral pattern. Two adaptive social strategies appeared: marriage at an older age and emigration, especially to America [21].
Thus, our mean age at first marriage is higher for bride- grooms in comparison with 400 French villages and towns [41] or Alpine valleys [31], at 31.2 versus 25.8 versus 27.5 years of age, respectively, whereas no significant differences are observed for brides. Obviously, migration patterns among non-heirs, an important element in our local biode- mography, cannot be assessed here through our marriage- based approach. A significant point is that until 1930, fertil- ity and celibacy rates in this area were higher than the French average [42]. A second cultural trait is to be considered here.
The local language is theGascon Béarnais(a dialect derived from the Romance languages), which was a further factor of
geographical isolation. Interestingly, according to some spe- cialists, the Ossau valley could represent a distinct linguistic entity [43]. A similar biolinguistic correlation has been shown in the Alps between the German and Italian speakers of the Fersina Valley [44].
Conclusion
Overall, this study brings new data on biodemographic indi- cators from Pyrenean populations during the 19thcentury.
The reliability of our data was strengthened by combining several sources. The study provides evidence, correlating with a work conducted during the 1970’s [9], of one of the highest levels of valley endogamy in Europe. In contrast, the consanguinity coefficient was within the range of other European mountain populations. Specific social and cultural factors, such as the age of bridegrooms or seasonality, account for the particular marriage pattern. Finally, the pres- ervation of houses, lands and cattle down the generations was the key regulator of this system.
Acknowledgments :This study was supported by the‘Sang 64’Association. Thanks to the two anonymous reviewers for their valuable contributions.
References
1. Cavalli-Sforza LL, Menozzi P, Piazza A (1994) The history and geography of human genes. Princeton University Press, Prince- ton, 413 p
2. Pettener D (1985) Consanguineous marriages in the Upper Bolo- gna Appennine (1565-1980): microgeographic variations, pedi- gree structure and correlation of inbreeding secular trend with changes in population size. Hum Biol 57:267–88
3. Martuzzi Veronesi F, Gueresi P, Pettener D (1996) Biodemo- graphic analysis of italian Alpine communities (Upper Sole Val- ley, 1725-1923). In: Biondi G, Mascie Taylor CGN (eds) Biode- mography in Anthropological studies. Rivista di Antropologia 74:55-75
4. Gueresi P, Pettener D, Martuzzi Veronesi F (2001) Marriage behaviour in the Alpine Non Valley from 1825 to 1923. Ann Hum Biol 28:157–71
5. Fuster V, Colantonio SE (2002) Consanguinity in Spain: Socio- economic, demographic, and geographic influences. Hum Biol 74:301–315
6. Abelson A (1978) Population structure in the Western Pyrenees:
social class, migration and the frequency of consanguineous mar- riages 1850 to 1910. Ann Hum Biol 6:165–78
7. Viazzo PP (1989) Upland communities: environment, population and social structure in the Alps since the sixteenth century. Cam- bridge University Press, Cambridge, 325 p
8. Fernet P, Jacquard A, Jakobi L (1975) Mariages et filiations dans la vallée pyrénéenne de l’Ouzom depuis 1744. Population 30:187–96
9. Marquer P (1979). Endogamie, exogamie et variations de la stat- ure et de l’indice céphalique dans la population béarnaise (Pyrénées-Atlantiques). Bull Mem Soc Anthropol Paris 6:333–42 10. Bonnain R (1986) Le mariage dans les Pyrénées centrales 1769-1836. In: Augustins G, Bonnain R, Peron Y, Sautter G (ed) Les Baronnies des Pyrénées, tome II. Editions EHESS, Paris, pp 123-56
11. Darlu P, Jakobi L, Hochez J (1994) Mating choice between myth and reality: the example of the Ouzom valley in Béarn (France) since the XVIIIe century. Bull Mem Soc Anthropol Paris 6:199–213
12. Danubio ME, Di Marcello S, Gruppioni G, Pettener D (1995) Analisi biodemografica di comunità isolate della Val Roveto (Aq) negli ultimi due secoli. Rivista di Antropologia 73:291–304 13. Bas M, Nogues RM, Aluja MP, Sevin A, Larrouy G (1996) Mar- ital mobility in population of the high Aran valley. Int J Anthro- pol 11:91–100
14. Fauve-Chamoux A, Ochiai E (1998) House and the stem family in Eurasian perspectives. Nichibunken, Kyoto, 468 p
15. Gonzalez-Martin A, Toja D (2002) Inbreeding, isonymy and kin- structured migration in the Principality of Andorra. Hum Biol 74:587–600
16. Prost M, Boetsch G (2005) Biodemography and matrimonial migrations in Alpine ecosystem. The mountains populations of Dauphiné Province (16th-19thcenturies). Anthropo 10:1–18 17. Riegler A, Marroni F, Pattaro C, Gueresi P, Pramstaller P (2008)
Isolation and marriage patterns in four South Tyrolean villages (Italy) during the nineteenth century. J Biol Sci 40:787–91 18. Bauduer F, Feingold J, Lacombe D (2005) The Basques: review
of population genetics and mendelian disorders. Hum Biol 77:609–27
19. Fresel-Lozey M (1969) Histoire démographique d’un village en Béarn aux XVIIIe-XIXe siècle: Bilhères d’Ossau. Bière, Bordeaux, 291 p
20. Le Play F (1875) L’organisation de la famille selon le vrai mod- èle signalé par l’histoire de toutes les races et de tous les temps.
Mame, Tours, 510 p
21. Etchelecou A (1991) Transition démographique et système cou- tumier dans les Pyrénées occidentales. Institut National d’Etudes Démographiques/Presses Universitaires de France, Paris, 260 p 22. Goy J (1982) Contribution à l’histoire de la coutume: maisons,
système d’alliance et système successoral en Béarn et en Bigorre au XIXesiècle. In : Köpeczi B, Le Goff J (ed) Objet et méthode de l’histoire de la culture. CNRS Editions, Paris, pp 229-41 23. Arrizabalaga MP (2005) Pyrenean marriage in the nineteenth cen-
tury; the French Basque case in marriage choices and class boundaries: social endogamy in history. Van Leeuwen MHD, Maas I, Miles A. (eds), International Review of Social History supplements. Cambridge University Press, Cambridge, 295 p 24. Dupâquier J, Le Mee R (1988) Histoire de la population fran-
çaise. Tome III: 1798-1914. Presses Universitaires de France, Paris, 601 p
25. Coppa A, Di Donato L, Vecchi F, Danubio ME (2001) Seasonal- ity of marriages and ecological contexts in rural communities of central-southern Italy (Abruzzo), 1500-1871. Coll Antropol 25:403–12
26. Crow JF, Mange AP (1965) Measurement of inbreeding from the frequency of marriages between persons of the same surname.
Eugen Q 12:199–203
27. Crow JF (1980) The estimation of inbreeding from isonymy.
Hum Biol 52:1-14
28. Cavalli-Sforza LL, Bodmer WF (1999) The genetics of human populations. Dover, New York, 965 p
29. Lucchetti E, Manfredini M, Boetsch G, Bley D, Aluja P, Pena J, Revello D, Melleri R, Sevin A (1996) Changes in marriage sea- sonality among some European rural populations. Int J Anthropol 11:73–81
30. Chasteland JC, Pressat R (1962) La nuptialité des générations françaises depuis un siècle. Population 2:215–40
31. Girotti M (2010) Étude biodémographique de deux populations des Alpes occidentales: Chiomonte (haute vallée de Susa–Italie) et l’Argentière-la-Bessée (Hautes-Alpes – France). PhD Thesis.
Université de la Méditerranée, Marseille
32. Ellis WS, Starmer WT (1978) Inbreeding as measured by iso- nymy, pedigrees, and population size in Törbel, Switzerland.
Am J Hum Genet 30:366–76
33. Pinto-Cisternas J, Pineda L, Barrai I (1985) Estimation of inbreeding by isonymy in Iberoamerican populations: an exten- sion of the method of Crow and Mange. Am J Hum Genet 37:373–85
34. Toja DI (1987) Estructura matrimonial de las poblaciones de dos valles pirenaicos. PhD Thesis. University of Barcelona, Barcelona
35. Colantonio SE, Lasker GW, Kaplan BA, Fuster V (2003) Use of surname models in human population biology: a review of recent developments. Hum Biol 75:785–807
36. Bideau A, Brunet G, Heyer E, Plauchu H (1994) La consangui- nité, révélateur de la structure de la population. L’exemple de la vallée de la Valserine du XVIIIesiècle à nos jours. Population 1:145–60
37. Arrizabalaga MP (2005) Succession strategies in the Pyrenees in the nineteenth century: the Basque case. The History of the Fam- ily 10:271–92
38. Cole JW, Wolf ER (1974) The hidden frontier, ecology and eth- nicity in an Alpine valley. Academic Press, New York, 350 p 39. Fuster V, Colantonio SE (2006) Isonymic analysis of population
structure in Gredos, Spain. Coll Antropol 30:199–203
40. Alvarez L, Mendoza C, Nogues RM, Aluja MP, Santos C (2010) Biodemographic and genetic structure of Zamora Province (Spain): insights from surname analysis. Hum Ecol (DOI:
10.1007/s10745-010-9363-x)
41. Henry L, Houdaille J (1979) Célibat et âge au mariage aux XVIIIe et XIXe siècles en France. II Age au premier mariage.
Population 2:403–42
42. Coale AJ, Treadway R (1986) A summary of the changing distri- bution of overall fertility, marital fertility and the proportion of married in the provinces of Europe. In: Coale AJ, Watkins SC (eds) The decline of fertility in Europe, Princeton University Press, Princeton, pp 31-181
43. Seguy J (1974) Atlas linguistique et ethnographique de la Gas- cogne. CNRS Editions, 455 cartes
44. Pettener D, Gueresi P, Martuzzi Veronesi F (1994) Struttura bio- demografica della valle del Fersina (Valle dei Mocheni) dal 1800 al 1914. Bollettino di Demografia Storica 20:131–41
45. Peña JA, de Mateo L, Basabe JM (1985) Isonimia marital en el valle de Orozco (Vizcaya) 1880-1980. Actas IV Congreso Espa- ñol de Antropologia Biologica 1:113-20
