See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/223790466

Marriage practices and ethnic differentiation: The case of Spanish

Article  in  The History of the Family · January 2005

CITATIONS READS
41 819

2 authors, including:

Juan F Gamella
University of Granada
95 PUBLICATIONS   668 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

A paper on reproductive histories of Romanian Roma women in Spain View project

EVICT Project View project

All content following this page was uploaded by Juan F Gamella on 13 November 2017.

The user has requested enhancement of the downloaded file.

Marriage practices and ethnic differentiation: The case of Spanish

Abstract

Spanish Gypsies [gitanos] are one of the main minorities and distinctive cultural groups in Spain. They
constitute approximately 1.5% of the total population (500,000–700,000) and live in all regions and provinces.
Here we present a model of the differential nuptiality patterns of gitanos, showing the importance of their marriage
and kinship practices as a source of their physical and cultural reproduction. We analyze genealogical
reconstructions based on archival research and oral history conducted in 23 Andalusian towns and villages, where
vital events are recorded in the civil registers. Our data concern over 9000 Gypsies born between 1850 and 1994.
Four main interrelated patterns are considered here: (1) the different form of establishing marriage based on
cohabitation and the exigency of the virginity of the bride, which ideally has to be proved in a public ceremony; (2)
the generalization of marriage and low levels of celibacy; (3) the early age at marriage; and (4) the importance of
ethnic endogamy and homogamy, which often results in a high level of consanguineous marriages repeated
generationally. We close by considering some of the potential demographic and cultural consequences of these
elements of Gypsy marriage patterns.
D 2005 Published by Elsevier Inc.

Keywords: Minorities; Gypsies; Marital patterns; Spain

T Corresponding author.
E-mail address: [email protected] (J.F. Gamella).

1081-602X/$ - see front matter D 2005 Published by Elsevier Inc.

1. Introduction

1.1. The Spanish Gypsies

The gitanos or calé of Spain and southern France are an ethnic minority related to other Roma/Gypsy
groups of Europe and America. All Roma populations, notwithstanding a common remote origin, have
adapted to the surrounding cultures, and today show traits of bfamilialQ resemblance within a gamut of
sociocultural variation (Acton, 1979; Fraser, 1992).
The gitanos come from the first migratory waves of Roma into western Europe, which ended in the
second half of the 15th century. Their customs are the product of a long coexistence with local Spanish
populations. Life in common has been marked by conflict and persecution, segregation and
discrimination, but also cooperation and hybridization (see Gómez Alfaro, 1998, 1999; Leblon, 1985;
San Román, 1986, 1997). Today most gitanos are proud of their ethnic identity, although they consider
themselves autochthonous Spaniards. Gitanos speak the languages and dialects of the regions where
they live and have lost most of their old trades and occupations. They have, however, developed other
differences to construct and vindicate their shared identity (Gay Blasco, 1999). Moreover, Spanish
Gypsies have contributed much to Spanish culture and folklore. Perhaps in no other part of Europe has
such a cultural fusion occurred as in Spain, especially in Andalusia, where many of the symbols and
practices that identify the region to the world (such as Flamenco singing and dancing) have a crucial
gitano component (Leblon, 1990; Pasqualino, 1998).
Almost all Spanish Gypsies are sedentary; they have been living in the same towns and counties for
generations and often have a strong attachment to their places of birth or residence, defining themselves
as Andalusians, Catalans, or even Sevillanos and Granadinos. In the last century, gitanos have followed
migration patterns similar to other Spaniards. Thus, we find gitanos from Extremadura and western
Andalusia in Madrid, from Old Castile in the Basque Country, and from eastern Andalusia in Catalonia,
the Levant, and the Balearic Islands. In our ethnographic fieldwork we also found many Gypsy families
who migrated in the 1960s to Germany, Switzerland, and France as guest workers but returned years
later to the Spanish towns and cities of their ancestors.1
Informed estimates of the size of the calé population put it in the range of 500,000 to 700,000,
around 1.5% of the total Spanish population (Fresno, 1994). Although Gypsies live in all regions of
Spain, around 40% are in Andalusia. Madrid, Catalonia, and the regions of Levant (Valencia and
Murcia) also have large gitano settlements. Some gitanos live in the periphery of major cities such
as Seville, Granada, Málaga, Madrid, and Barcelona. But they also manifest an important presence
in localities of different size and structure, from the industrial towns of the Basque Country to the
tourist resorts of the Mediterranean and the agrarian regions of Andalusia, Extremadura, and
Murcia.
Since 1977, when the new political context brought about democracy and decentralization of the
Spanish state, the gitanos have finally achieved full status as citizens, at least legally. Moreover, there
have been clear improvements in their access to health care, education, and housing but not without

1
Marriage is also a common reason for moving. Although post-marital residences of new Gypsy couples tend to be virilocal or patrilocal,
there are many variations according to social and economic opportunities as well as to the space available in the houses of the couple’s parents
and other relatives.
 E. Martin, J.F. Gamella / The History of the Family 10 (2005) 45–63 47

resistance and conflict with groups of the dominant majority. Large sectors of the minority, however, still
suffer from segregation and marginalization, particularly visible nowadays in the slums that have been
constructed, often by public initiative, on the outskirts of many Spanish cities (see Gay Blasco, 1999;
San Román, 1997). Still remarkable is the contrast between the exaltation of some Gypsies in art and
entertainment and the prejudice and discrimination that most Gypsies suffer in their daily interactions
with Payos (non-Gypsies).
Today the life and traditions (old and reinvented) of the gitanos are being rapidly transformed.
Changes, however, affect differently the underprivileged families in shantytowns; middle-class gitano
families of professionals, clerks, or successful merchants; and flamenco musicians. Hence, the Gypsy
population is increasingly heterogeneous, even polarized between the new middle class and a gitano
underclass affected by poverty and social exclusion (Gamella, 2004). The latter suffer from lack of
occupational training, and have a new set of social problems to overcome, such as drug dealing, drug
dependence, and AIDS. Even if these new problems affect only a limited number of Gypsies, their
impact is great, since they not only threaten life and health, but also provide new opportunities for
prejudice and discrimination. For example, in many jails, Gypsy women account for over a fourth of all
Spanish women incarcerated (Barañi, 2001), and the stereotype of Gypsies as violent, thieving, and drug
dealing remains powerful in the dominant mentality (Gamella & Sánchez Muros, 1998). In this article
we will concentrate on an aspect of gitano shared identity that is usually overlooked: the institutions that
produce and reproduce them physically and socially.

1.2. The Gypsy marriage system

The marriage system of Spanish Gypsies is a key factor in their permanence as a distinct group,
and in differentiating them from the dominant majority. These interrelated patterns concern central
elements of marriage institutions: the recurrent practices and expectations concerning when, how, and
who to wed. Hence, they depend on the pressures and mechanisms that bsubtly influenced the
proportion marrying and the age at marriageQ (Macfarlane, 1986, p. 276). Gypsies have tended to
make marriage a necessary transition to adulthood and have pursued strategies universalizing
marriage, which has kept celibacy low. They have also stressed the value of virgin brides, and
confirmation of virginity in public ceremonies both as an ethnic marker and a form of social control of
their women. And they have preferred that both brides and grooms marry at an early age, and have
favored endogamous and homogamous matches, which have resulted in a high incidence of
consanguineous unions.
Ethnic differences in marriage patterns derive from different conceptions of person, kin, and body,
resulting in specific gendered moralities and gender roles (see Gay Blasco, 1999). This system in
turn supports a group of strategies decisive for the demographic configuration of the minority and
their resistance to assimilation in the face of rejection, discrimination, and accelerated social change.
Marriage is also the key to understanding the condition of Gypsy women; the gender system that
prevails among Gypsies is crucially determined by matrimonial practices and the roles, beliefs, and
values that sustain it.
Marriage patterns have been neglected in the study of gitanos and in Gypsy studies in general,
including the application of anthropological models that stress descent, especially through the male line
as the key element of their social organization. Perhaps, as well, the reliance on male informants and the
lack of genealogical or demographic investigations has helped to obscure marriage as a research subject,
48 E. Martin, J.F. Gamella / The History of the Family 10 (2005) 45–63

except in its symbolic and ceremonial aspects (see Ardévol & del Pino, 1986; Gay Blasco, 1999;
Pasqualino, 1998; San Román, 1997).

1.3. Objectives

In this article we will use data from an ongoing research project on the demographic history of
Gypsies in the Granada province of southern Spain, to further our understanding of this marriage
pattern and its consequences. Although the data are preliminary, our results show the importance of
micro-demographic studies to understanding bhiddenQ populations (Lambert, 1990). We will
concentrate on three crucial dimensions of marriage patterns: (a) The universal character of gitano
marriage that produces a low rate of celibacy among Gypsies, especially Gypsy women. (b) The low
age of marriage both for males and females. Ideally, Gypsy women should marry in their teens soon
after menarche, and their bridegrooms should be a few years older. (c) The preference for endogamous
and homogamous marriages that result in a high rate of consanguineous unions maintained through
generations.

2. Methods and sources

At least since 1783, the ethnic affiliation of gitanos has not appeared in Spanish official records or
censuses. Spanish Gypsies are thus indistinguishable from any other citizens in civil or parochial
registers, or in public documents in general. Hence, official data do not allow the direct demographic
study of this minority or comparison with other groups, unlike some countries of central and eastern
Europe where Gypsies have been recognized as a nationality or ethnic minority and their ascription
recorded in official records. That approach has permitted interesting studies on their demographic
structure (see Meszaros & Foti, 1996; Daroczi, 1994; Gheorge, 1994; Hooz, 1987; Kalibova, 1994;
Mann, 1992; Srb, 1998).
Moreover, even some of the most important experts in the field have assumed that gitanos have never
been interested in registering births, deaths, or marriages. This would make the demographic archival
research of any period after the 18th century unfeasible or worthless (see San Román, 1997). This
assumption has often coincided with characterizing gitanos as nomads, itinerants, or at least as a group
that has low and uncertain territorial and local links. Our research contradicts both assumptions, at least
for Andalusia. In the villages and towns of that region, Gypsy families have been inscribing the births
and deaths of their kin at least since in the early 1870s when the civil registration record (Registro Civil)
began. In addition, many Gypsy relatives from one village can be traced in the Registros of neighboring
towns and villages.

2.1. The area and population studied

This research was carried out in 23 localities in the province of Granada in the south of Spain.
Localities sampled ranged in population from 1200 to 22,000 and included various types of settlements,
from rural to urban, and from more isolated to close to major cities of the province.
We believe that our data concern the whole Gypsy population of the localities studied. We have
considered as Gypsies those who identified themselves as such, and those who were identified as
 E. Martin, J.F. Gamella / The History of the Family 10 (2005) 45–63 49

Gypsies by their families and neighbors. Self-ascription was therefore the key identifying factor for our
project. In all cases these two criteria coincided with descent of others also identified as Gypsies,
although interethnic marriages are not unknown in this region as in the rest of Spain.2 The relative
magnitude of the Gypsy population was uniformly high in the region studied, ranging from 4% to 30%
of the total population in the various pueblos (settlements) see Gamella, 1996).

2.2. Methodological strategy

Our demographic study was based on previous ethnographic research in which we established rapport
with, and knowledge of Gypsy residents and of the living conditions of the Gypsy minority in the area.
With the help of local Gypsy and non-Gypsy residents, we reviewed carefully the principal local census
[Padrón Municipal de Habitantes] and established a list of the local Gypsy population in each of the
villages and towns where we worked. In this form the large database we created was used for the first
study of the basic structure of the Gypsy population of the area (see Fig. 1, and Martı́n Carrasco-Muñoz,
1999). Our data were immediately codified to maintain anonymity; we have never used real names, nor
identified Gypsy citizens to third parties.
Ten contiguous localities with important Gypsy populations were selected for genealogical
reconstruction. Starting with data from the 1991 census, the civil registers were searched from the
present to the earliest period recorded (1870–1875). Three books of the Spanish civil registration records
were consulted; those concerning births, marriages, and death.3 The genealogical approach served as a
web on which we could reconstruct of the Gypsy population in each locality. To limit the problems
associated with the use of genealogies (see Bideau & Poulain, 1984; Dupâquier, 1993; Hollingsworth,
1976; Jette & Charbonneau, 1984), we also checked the register from 1870 forward, looking for indirect
references or clues that might confirm the affiliation of our subjects. In recording Gypsies the phrase
Castellano Nuevo (which replaced the forbidden term gitano after 1783) was often used. Allusions were
made to trades and occupations associated with Gypsies, such as esquilador (shearer), chalán (horse
dealer), canastero (basket maker), as well as streets and barrios where Gypsies lived (such as Cañada de
los Gitanos). Thus, we were able to create ascendant and descendant genealogies that merge at some
point and others that do not. In this way we have been able to collect information about individuals who
disappeared from a particular locality long ago either by death or migration.
We have traced genealogical connections from available sources for the whole local Gypsy
population, not just for specific families, using birth, death and marriage records. Therefore, this is a
demographic reconstruction of a population that resided in a specific region during a specific period. It
includes patrilineal and matrilineal links of equal depth, since both mothers’ and fathers’ names appear
equally in Spanish records. Under-registration of unimportant people (Post, van Poppel, van Imhoff, &

2
Mixed marriages have been recorded in Andalusia since the 17th century. The historian Gómez Alfaro concluded that by the middle of the
18th century about 5% of Gypsies married non-Gypsies. These rates may have increased and are greater today. The children of mixed marriages
used to identify and to be identified with the extended family they grew up in, more likely to be their father’s family. But we know of many
exceptions to this rule. Thus, in Andalusia many gitanos have non-Gypsy ancestors and vice versa although we do not know the exact rates
among the population we have studied.
3
The gap between death dates may be due to a change in regulations concerning death registrations. Until 1975, events had to be registered in
the place of occurrence (mostly hospitals located in cities). Post and her collaborators (1977) have noted a similar problem in the Netherlands. In
many pueblos ecclesiastical records were damaged or destroyed during the Spanish Civil War.
50 E. Martin, J.F. Gamella / The History of the Family 10 (2005) 45–63

80 +
WOMEN
75-79 MEN

70-74
65-69
60-64
55-59
50-54
45-49
40-44
35-39
30-34
25-29
20-24
15-19
10--14
5--9
0-4
8 6 4 2 0 2 4 6 8

80 +
WOMEN
75-79 MEN
70-74
65-69
60-64
55-59
50-54
45-49
40-44
35-39
30-34
25-29
20-24
15-19
10_14
5_9
0-4
8 6 4 2 0 2 4 6 8

Fig. 1. Age pyramid of Gypsy population, Granada, and Andalusian population, 1991.

Kruse, 1997) was not a problem, since the goal of collecting the genealogies had been to include every
individual. Children who died at an early age were also recorded, as well as unmarried males and
females. By including the entire local population extracted from civil registration records, some of the
problems of using genealogies were avoided (see Henry, 1956, p. 16, 1967, p. x).
We were aware of some of the problems posed by genealogical reconstitution methods. Thus, we
sought independent evidence to confirm the relevance of our data and the links established in our
genealogical networks both within and beyond the records themselves.
Available data from all localities and sources were then carefully contrasted to trace relatives and
relationships of the Gypsy people identified in each locality. We connected ascendants and descendants
 E. Martin, J.F. Gamella / The History of the Family 10 (2005) 45–63 51

Table 1
Percent distribution of women by age and marital status in groups of birth cohort (1,607 Gypsy women, Granada, Spain)
1900–1909 1910–1919 1920–1929 1930–1939 1940–1949 1950–1959 1960–1969
N 74 87 136 166 217 541 389
Married women 77.03 68.97 70.59 67.46 61.75 61.83 64.78
Single women 1.35 0.00 0.73 0.00 0.00 0.77 0.00
Dead before age 20 14.86 18.39 13.24 13.25 17.05 12.20 2.83
No data 6.76 12.64 15.44 19.29 21.20 25.20 32.39

in growing genealogies, which sometimes included hundreds of individuals and confirmed data about
each person in every kin network. The triangulation process confirmed that data about the local Gypsy
population from the civil registers were usually accurate. Our reconstruction has included more than
8,500 Gypsies born between 1875 and 1999.
Often the same or similar names repeated themselves in several generations. Occassionally the same
person was designated by different names at different times and annotations. To establish individual
identities we used information from relatives interviewed during our ethnographic fieldwork. Moreover,
we have also used the many cross references to parents and grandparents of individuals in our sample (as
is the Spanish custom, each had two family names: one from the father’s and the other from the mother’s
line, respectively), as well as the exact recorded dates of birth, death, and marriage.
This has been a lengthy and laborious task, and in the collection of archival data some of our students
and colleagues have helped. The triangulation of data case by case and the establishment of kin
networks, however, were carried out by the authors themselves.

3. Results

Preliminary analysis of two data sets permit us (a) to analyze the population structure found in the
census of 1991;4 and (b) to establish variables concerning nuptiality using data from our reconstruction
of the local Gypsy population based on civil registers from 1875 to 1999.

3.1. Celibacy rates

To estimate rates of celibacy data from two different sources were combined. Table 1 shows the
percentage of single women in our reconstruction. Although it is provisional since we are still adding
data from new localities, it illustrates our point: The percentage of single women was low-even zero for
some decades. Most single women appear in the groups for which we lack information. They are women
for whom we only have birth dates, who probably migrated before or when they married. In our
population the proportion of women who migrated is high and may have increased throughout the
second half of the 20th century when migration patterns increased greatly in the area for both Gypsies
and non-Gypsies. The Gypsies show a preference for virilocal or patrilocal postmarital arrangements,

4
The more recent data from the Census of 2001 is not being entered.
52 E. Martin, J.F. Gamella / The History of the Family 10 (2005) 45–63

which may contribute to the number of women who left the area of study at the time of marriage and
before the birth of their first child. We are collecting a random sample of these women who lived in other
parts of the province of Granada because we think that in their new settlement they married. Thus far,
their nuptiality patterns do not seem to differ significantly from those who remained in the area of study.
We have also found some of the missing women either in recently researched localities different from
their natal ones, or in Granada’s metropolitan area. In any event, most of them were married.
Table 2 and Fig. 1 examine the structure of the Gypsy population residing in the area in recent times.
This concerns 23 pueblos of different sizes and structures, including some localities in the metropolitan
area of Granada and in the neighboring province of Almerı́a. Over 6,300 Gypsies lived in this area in
1991. The results confirm some of the claims of our model. Thus the portion of men over 50 who had
remained single was 3.4 percent, and for women, less than two in a thousand.

3.2. Age at marriage

Gitanos have an elaborate system for establishing and celebrating matrimony (casamientos), which
often includes a public ritual in which the bride’s virginity is proven. This ceremony, in which hundreds
of relatives may participate, occupies a central position in the boda gitana (Gypsy wedding), one of the
high points in Gypsy culture and Gypsy blaw.Q There were, however, other ways of establishing a
socially accepted bond between a Gypsy couple. Agreed elopement, followed by acceptance of both
families and cohabitation, may have been the most common way of establishing a socially acceptable
gitano marriage in the period considered. In most cases, according to Gypsy blaw,Q cohabitation implied

Table 2
Proportion of Gypsy population of 23 Andalusian localities by marital status, 1991 (N=6397)
Age group Men Women
Single (%) Married (%) Widowed (%) Single (%) Married (%) Widowed (%)
0– 4 100.00 0.00 0.00 100.00 0.00 0.00
5–9 100.00 0.00 0.00 100.00 0.00 0.00
10–14 100.00 0.00 0.00 99.75 0.25 0.00
15–19 97.14 2.86 0.00 85.07 14.93 0.00
20–24 76.65 23.35 0.00 57.28 42.41 0.31
25–29 47.35 52.65 0.00 25.67 73.56 0.77
30–34 24.91 73.21 1.89 8.37 89.54 2.09
35–39 12.50 87.50 0.00 3.37 93.26 3.37
40–44 8.43 89.76 1.81 2.47 93.21 4.32
45–49 2.42 95.16 2.42 0.00 96.26 3.74
50–54 4.81 93.27 1.92 0.00 94.74 5.26
55–59 2.30 96.55 1.15 0.00 83.05 16.95
60–64 3.53 90.59 5.88 0.00 79.79 20.21
65–69 1.27 89.87 8.86 0.00 63.51 36.49
70–74 4.17 89.58 6.25 1.52 54.55 43.94
75–79 5.26 63.16 31.58 0.00 41.30 58.70
80–84 8.33 58.33 33.33 0.00 29.63 70.37
85 and over 0.00 60.00 40.00 0.00 13.33 86.67
N: 1942 1309 43 1583 1346 174
 E. Martin, J.F. Gamella / The History of the Family 10 (2005) 45–63 53

marriage (see Gamella, 2000), and the Gypsy community considered children born to such couples to be
legitimate. A commitment thus established had to be respected, and implicit were a set of rights and
duties that relatives from both sides could strictly enforce.
Over the centuries, the peculiarities of Gypsy unions have been a main cause for their customs to be
censured by the dominant culture. Phrases like bnot keeping matrimonyQ (in the year 1594), bnot
following the Law of the (Catholic) ChurchQ (in 1610), bhave not been married and have not asked for a
dispensationQ (in 1619) are widespread in the anti-gitano decrees of the ancien régime. On the other hand,
bnormalizationQ of their marriages was a crucial factor in Gypsies’ access to a legitimate identity. For
example, owning a Church-issued marriage certificate was enough proof for a family to be freed after the
Causa General of 1749. This was a genocidal attempt throughout Spain in which thousands of gitano
men, women, and children were arrested and sent to jail (see Gómez Alfaro, 1998, 1999, 1993; Leblon,
1985; and Sanchez Ortega, 1977, 1986, for the history of measures against the Spanish Gypsies).
For the dominant majority and the state, however, only religious (Catholic) or civil contracts produced
a legitimate marriage. Hence, the civil register records many Gypsy children who were born to couples
whose marriages did not exist for official purposes. Often there is no trace of the matrimonial contract in
public documents until years after the couple had been living together and had several children. Table 3
shows the percentage of couples who had formalized their marriages according to Spanish law sometime
during their lives, and among them, the portion who did so before the birth of their first child (second
column). It is not uncommon to find that marriage was formalized just days before the death of one of
the spouses, usually the husband. The proportion of Gypsy couples who formalized their situation in
recent decades has increased, an indication of the bnormalizationQ of this population.
Gypsy and payo (non-Gypsy) perceptions of a proper wedding have somehow converged in the last
decades. The Catholic Church maintains considerable leverage and power in family matters, and during
most of the 20th century, Spanish family law followed Catholic canonic law. Therefore, there is a
methodological problem in obtaining reliable data about the dates of Gypsies’ weddings when often their
definition of marriage has not coincided with that of the dominant majority encoded in civil law.

3.3. Age at birth of first child: Gypsy mothers in our sample

The age at which Gypsy women had their first child has been analyzed as a proxy variable for age at
first marriage. Since cohabitation of parents normally started months before that birth, the data we
present clearly underestimate the age at marriage.

Table 3
Percentage of Gypsy couples legalizing their marital status in groups of birth cohort (Gypsy population. Granada, Spain)
Year of birth Legalized marriages (%) Portion of legalized marriages
birth of first child (%)
1900–1909 44.90 9.09
1910–1919 39.66 13.04
1920–1929 56.70 9.09
1930–1939 74.31 20.99
1940–1949 88.06 55.08
1950–1959 88.55 76.19
1960–1969 80.56 94.58
54 E. Martin, J.F. Gamella / The History of the Family 10 (2005) 45–63

Table 4
Percent distribution of mothers by age at birth of first child in groups of birth cohort (Gypsy population, Granada, Spain)
Age of mother at first child 1900–1909 1910–1919 1920–1929 1930–1939 1940–1949 1950–1959 1960–1969
b15 8.51 6.90 4.26 1.92 1.80 2.23 2.16
15–19 42.64 44.83 37.23 43.27 58.56 59.92 54.04
20–24 29.79 32.76 37.23 39.42 30.11 30.11 30.81
25–29 10.64 8.62 13.83 9.62 6.69 6.69 11.35
30–34 6.38 3.45 4.26 3.85 0.90 0.74 1.62
N35 2.04 3.45 3.19 1.92 0.90 0.30 0.00

Throughout the whole period, most Gypsy women had their first child in their teens (see Table 4). The
age group 15–19 is the largest birth cohort followed by age group 20–24. Also note the proportion of
gitanas having children before age 15, and how this tendency has diminished since 1930.
Table 5 compares Gypsy women with the Spanish population at large. Data in contrasting columns are
not equivalent: In the first two columns, age at first child is used as a proxy variable for gitanas’ age at
first marriage. In the third column, mean age of Spanish women at first marriage is used (Reher, 1996).
At least 1 year should be subtracted from the Gypsy data. In any case, the contrast with the larger
population is clear, and in both cases these patterns have been consistent over time. Both Reher and
Rowland have pointed this out for the whole Iberian Peninsula (Reher, 1991, 1996; Rowland, 1987). The
mean age of first marriage of Spanish women remained in the range of 24 to 26.5 years from 1887 to
1979; it increased in the last two decades as nuptiality rates have fallen. For gitanas, the mean age at first
child has remained in the age range of 19 to 20. Therefore, a gap averaging about five years has
separated the minority and the dominant majority.
It may be argued that the region we have investigated is among the poorest and most backward in
Spain, and this may affect marital patterns. But mean age at marriage for Andalusian women during the
20th century has been over 25, except in 1920, when it fell to 24.93 (IEA, 1999). The singulate mean age
at marriage (SMAM) for women in this area in 1887 was 23.32 (Reher, Pombo, & Nogueras, 1993).
During the 20th century, it has remained within in the age range of 24–26 (IEA, 1999).

Table 5
Comparison of median and mean age at first child between Gypsy women in Granada and Spanish women, 1870–1979
Years Gypsies, Gypsies, Spanish,
median age mean age mean age at marriage
1870–1879 19 20.13
1880–1889 21 20.27 24.19
1890–1899 20 20.57
1900–1909 19 19.21 24.47
1910–1919 20 19.31 25.1
1920–1929 20 20.89 25.71
1930–1939 20 20.28 25.83
1940–1949 19 19.35 26.65
1950–1959 19 19.19 26.43
1960–1969 19 19.87
1970–1979 18 18.85
Source for Spanish data: Reher, 1996.
 E. Martin, J.F. Gamella / The History of the Family 10 (2005) 45–63 55

3.4. Consanguineous marriages

Among the Gypsies studied, consanguineous marriages are more frequent and show a different
structure than those observed for the Spanish population at large and for most Spanish regions and
provinces (see Calderón, Peña, Morales, & Guevara, 1993; Calderón, Peña, Delgado, & Morales, 1998;
Fuster, Morales, Mesa, & Martı́n, 1996; Gómez Gómez, 1989; Pinto-Cisternas, Zei, & Moroni, 1979;
Valls, 1982).
Two individuals are said to be consanguineous when they share one or more close ancestors. The
progeny of consanguineous parents are inbred (Cavalli-Sforza & Bodmer, 1971, p. 341). Using our
incomplete data we will provide some estimates of the frequencies of the various types of
consanguineous unions, the rates of total consanguinity, and the mean levels of inbreeding in the
population studied. These are crucial aspects of the genetic and demographic structure of any population.
The rate and patterns of consanguineous unions are dependent, among other factors, on demographic
variables, such as density of population, mortality and birth rates, and age at marriage. On the other
hand, inbreeding can influence age at marriage, age at first livebirth, and fertility rates, as well as infant
morbidity and mortality (see Bittles, 1994; Bittles, Mason, Greene, & Appaji Rao, 1991).
The search in our population was limited to the 1267 marriages in which names of the parents of both
spouses are known, and thus, a genealogical depth of three to seven generations in patrilateral and
matrilateral lines can be established (see Cazen and Cazen, 1996). These unions occurred between 1903
and 1999, although most took place in recent decades. Of these 1267 marriages, 366 were
consanguineous. Four of every five of these consanguineous marriages (83.3%) happened after 1950,
a period when inbreeding decreased rapidly throughout Spain. At present, consanguineous marriages are
still common among Gypsies. A fourth of all inbred unions in our sample took place after 1990.
Table 6 presents the frequency of every type of consanguineous marriage, considering the main
kinship links between spouses. Also indicated is the corresponding coefficient of inbreeding ( F),5 as
well as the number of multiple consanguineous unions for each type of marriage. Uncle–niece marriages
have been rare among gitanos; we found only one union of this type in the 1267 marriages considered.
The two most common types of consanguineous unions are those between second cousins ( F=0.0156),
of which we found 89 cases (7% of all marriages), and between first cousins ( F=0.0625), 87 cases, or
6.9% of all marriages considered. Matings among first cousins once removed and third cousins were also
relatively frequent (2% of all marriages).
The 366 consanguineous marriages were characterized by an average inbreeding coefficient of
F=0.0276. For the 1267 marriages, the rate of total consanguinity was 28.9%, and the resulting mean
coefficient of consanguinity, 0.00797. If these results could be extrapolated to the whole gitano
population, the rates of consanguinity would be high by Spanish and European standards.
In a population like this, where uncle–niece marriages are rare, the correlation between the
consanguinity rates and inbreeding values greatly depends on the trends of the ratio between the closer
unions–first cousins/second cousins (Calderón et al., 1993, p. 757)–but also of the presence of multiple
consanguinity (Calderón et al., 1998). In our sample, first-cousin marriages ( F=0.0625) account for
5
The coefficient of relatedness or inbreeding, F, is the probability that an individual receives two genes that are identical by descent at a given
locus (Cavalli-Sforza & Bodmer, 1971, p. 343). Although usually measured from marriages rather than from the birth of offspring, it provides a
statistical measure of bthe proportion of gene loci at which an individual is homozygous.Q Its average values indicate the mean level of
inbreeding in populations (Bittles, 1994, p. 579). Here we assume that the genealogical links that have been established from archival and
ethnographic research reflect procreative and genetic relations, but this topic requires further study and scrutiny.
56 E. Martin, J.F. Gamella / The History of the Family 10 (2005) 45–63

Table 6
Distribution of consanguineous marriages by main type of kin relationship in 1267 gitano marriages (1903–1999)
Marriage type of kinship F N % of Mc % of Mt McM % McM
of Mc
Uncle–niece 0.125 1 0.3 0.1 0 0.0
First cousins 0.0625 87 23.8 6.9 19 21.8
First cousins once removed 0.0312 45 12.3 3.6 13 28.9
Half-first cousins 0.0312 10 2.7 0.8 6 60.0
Second cousins 0.0156 89 24.3 7.0 20 22.5
Half-first cousins once removed 0.0156 6 1.6 0.5 3 50.0
Second cousins once removed 0.0078 44 12.0 3.5 7 15.9
Half-second cousins 0.0078 4 1.1 0.3 2 50.0
Third cousins 0.0039 25 6.8 2.0 10 40.0
Other (less than third cousins) b0.0039 55 15.0 4.3 14 25.5
Total 366 100.0 28.9 94 25.7
F: coefficient of inbreeding; Mc: Consanguineous marriages; Mt: Total number of marriages; McM: Multiple consanguineous
marriages.

55.6% of the total rate of inbreeding. Unions of first cousins once removed and half-first cousins
( F=0.0312) account for 17.3%. In total, these three types of unions account for almost a quarter of the
whole value of F (72.9%).
But the complexity of relationships is not fully depicted by these data. There were 17 different types
of kinship relationships among these 366 unions; in 25.7% of them we were able to trace pedigrees that
showed other links of multiple consanguinity of great complexity and variety. Since our analysis is
partial, results in Table 6 underestimate the importance of multiple consanguinity in this population.

3.5. Multiple consanguineous marriages

A crucial trait of gitano nuptiality is the frequency of multiple consanguineous marriages in which
couples share various degrees of kinship and are linked by various lines to several common ancestors.
The biomedical consequences of consanguineous matings may be greater if consanguinity has
accumulated in previous generations within a family. This has been noted in different contexts (for
Spain, see Calderón et al., 1998).
In a first review of the 366 consanguineous unions in our sample, we identified 94 (25.7% of all
consanguineous marriages) as multiple. Three were unions between double first cousins ( F=0.125); in
one case, the spouses were also second cousins ( F=0.14).6 There were also six double second cousins
( F=0.0312) and three cases in which couples were first cousins once removed and second cousins
(0.047). In 23 cases, three or more independent consanguineous relationships between the spouses were
found. Some kindred gitanos appear to inbreed more repeatedly than others.
The level of inbreeding does not seem to correlate with the size of the local Gypsy population. This is
a crucial aspect of gitano consanguinity. In fact, it is difficult to show the complexity of relationships
using standard indices. And the contribution of multiple consanguinity to the genetic structure of the
population may be underestimated. To illustrate the intricacy and complexity of some of these forms of

6
In these cases, two brothers married two sisters, a type of match also found among Gypsies in other countries (see Okely, 1983).
 E. Martin, J.F. Gamella / The History of the Family 10 (2005) 45–63 57

1 2

3 4 5 6

7 8 9 10 11 12

13 14 15

16 17

Fig. 2. Case I. Gypsy consanguineous marriage, Granada, 1997.

inbreeding, we have selected several examples from the reconstructed genealogies. They are not
exceptional in their complexity; in our database, there are dozens like them.
In Figs. 2 and 3, we show some of the consanguineous relationships found in a Gypsy kin
network in which inbreeding has been sustained for generations. Three recent marriages have been
analyzed and the connections between the spouses and their closest common ancestors have been
graphed in a simplified form. The complexity of these patterns is indicated by the difficulties of
graphing the maze of relationships in a two-dimensional diagram. Every person in the graph has a
different number. Males are represented by triangles, women by circles. Horizontal lines connect
siblings and spouses.

1 2

3 4 5 6 7 8 9

10 11 12 13 14

16 17 18 19

15 20 21

22

Fig. 3. Case II. Two Gypsy consanguineous marriages. Granada, Spain, 1987 and 1993.
58 E. Martin, J.F. Gamella / The History of the Family 10 (2005) 45–63

3.6. Case 1. The marriage of Luis and Estrella in 1997

The first case concerns Luis, a gitano born in 1974 in a pueblo about 60 km from Granada. At age 23,
he married Estrella, a girl of 14, who was his neighbor and kinswoman. (Pseudonyms are being used
through the cases.) The following year, in 1998, they had their first daughter.
In Fig. 2, the various links between Luis (number 16) and Estrella (17) are depicted as they appeared
in our genealogical reconstruction. Seven different consanguineous links have been located between this
couple. They are first cousins once removed, second cousins, double third cousins, and also triple fourth
cousins. Their parents and grandparents also made consanguineous unions. In this family the
grandchildren of two brothers married among themselves, and several of their children did the same,
in time developing a pattern of inbreeding of considerable complexity.
Surprisingly, the resulting coefficient of inbreeding for Luis and Estrella’s daughter is smaller than
that obtained between first cousins ( F=0.0576). In such cases, F values may be insufficient in measuring
genetic proximity, especially in populations where founder effects may obtain (see Kalaydjieva,
Gresham, & Calafell, 2001).

3.7. Case 2. The two marriages of Esteban

In 1987, Esteban, a 20-year-old Gypsy male, married his cousin Remedios (age 15), who lived in
another town. Soon they had three children (born in 1988, 1989, and 1990). In 1993, Esteban left his
wife and moved in with another girl, Malena, then age 16, who was also his cousin and lived nearby.
From this union Malena bore three children in 1994, 1996, and 1997.
Between Esteban (number 20 in Fig. 3) and Remedios (15) at least four different consanguineous
links exist. They are half first cousins (having a grandparent in common, number 4), and three times
second cousins ( F=0.0781). The parents of Remedios (10 and 11 on the graph) are also first cousins.
Esteban and Malena (21) are linked in many ways to their common ancestors. We have been able to
trace a total of 11 independent consanguineous links. They are cross first cousins (Malena is daughter of
the brother of the mother of Esteban). They are also second cousins, four times second cousins once
removed, double third cousins, double third cousins once removed, and third cousins twice removed.7
The inbreeding coefficient of their children was at least F=0.0114, close to that of double first cousins
and uncle–niece marriages ( F=0.0125).
Both wives of Esteban, Remedios (15) and Malena (22) are also first cousins once removed and second
cousins once removed. Parents and grandparents of these young Gypsies also came from inbred unions.
Thus, we are describing kinship in which consanguineous matches have been common for generations. It
would be difficult to find such a level of sustained inbreeding in other European populations.

4. Discussion

We are considering here a form of ethnic difference with profound demographic and perhaps
genetic consequences. Nuptiality has always played a crucial role in demographic theory. The

7
Indirectly this marriage shows another relevant factor in gitano marriage patterns: the fusion of generations due to the long genetic periods
of Gypsy women who continue bearing children even after their daughters have begun parenting.
 E. Martin, J.F. Gamella / The History of the Family 10 (2005) 45–63 59

Gypsies seem to engage in marriage practices that differ greatly from those hypothesized for western
European populations by Hajnal (1965). For centuries high levels of celibacy and late age at
marriage appear to have been fertility control mechanisms in Europe. Spain did not differ from other
western European countries in this respect, notwithstanding local and regional variations (see Livi-
Bacci, 1968a,b; Reher, 1996).
Our research confirms that gitanos have marriage patterns different from their non-Gypsy neighbors.
Contrary to what was expected, these differences have grown larger in recent years. Somehow the
majority and the minority appear to have adapted differently to accelerating change, at least in
demographic terms. In some crucial aspects, differences result today in opposites, which some decades
ago might have been perceived as gradations in a continuum, at least in relation to the poorer sector of
the Spanish population. Many of these marital strategies and the resulting aggregate traits they produce
are relatively common in Africa and Asia (see, for example, Bittles, 1994; Bittles et al., 1991; Goody,
1990).
In the last four decades, Gypsy fertility may have remained high, but improvements in nutrition,
health, and living conditions may have reduced infant mortality resulting in growth in the gitano
population. There are signs, however, that the fertility rates of gitanas decreased in the 1990s.
The patterns of nuptiality described in this article are interconnected, as has been shown in other areas
of the world. For instance, consanguinity has been associated with greater gross fertility, due in part to
younger maternal age at marriage and at first birth (Bittles, 1994; Bittles et al., 1991).

4.1. Consanguinity

The level of inbreeding in this Gypsy population seems very high by European standards (Bittles,
1994; Bittles et al., 1991). But comparisons between our results and those from other Spanish and
European populations can only be made in broad terms. Important differences in methodology, period,
and region must be considered.
Our results are coherent with those of the only available study that has provided valid data on
Gypsy inbreeding in Spain. A national study of tens of thousands of births found about 30 percent
of consanguineous unions among the parents of newborn Gypsy children, compared to about 1
percent among parents of non-Gypsies (Martı́nez-Frı́as, 1993; Martinez-Frias & Bermejo, 1992).8
Our results are also generally consistent with the increasing data on population and genetic structure
of European Gypsies, mostly relating to issues of public health (Assal, Susanszky, & Czeizel, 1991;
Thomas, Doucette, Thomas & Stoeckle, 1987) and incidence of genetic diseases (see Gresham et al.,
2001; Kalaydjieva et al., 2001, for recent reviews).9
In Spain in the first half of the 20th century, consanguineous marriages have been more prevalent than
in other European countries. Using Vatican records of Catholic dispensations granted in the whole of
Spain, Pinto-Cisternas et al. (1979, p. 63) found that total consanguinity rates fluctuated from 5.8% in

8
It was a case-controlled study of birth defects from the Spanish System on Congenital Malformations (ECEMC) that included data from 84
maternity hospitals throughout the country. Nearly 906,000 births from 1976 to 1991 were scrutinized. In 1.6% of cases the mothers identified
themselves as Gypsies. The study focused on the prevalence of multiple congenital anomalies and autosomal recessive syndromes at birth. The
latter syndrome was bapproximately seven times more frequent in Gypsies than in non-Gypsies,Q a situation that reflected the higher rate of
consanguinity in this minority (Martı́nez-Frias & Bermejo, 1992, p. 634).
9
Our data also coheres to Cohn’s speculations on U.S. Gypsies. For that group he estimated bthe average coefficient of inbreeding F to be at
least 0.0118 but probably closer to 0.016. In the general population, the coefficient is well below 0.001Q (Cohn, 1973, p. 72).
60 E. Martin, J.F. Gamella / The History of the Family 10 (2005) 45–63

1911–1914 to 4.12% in 1940–1943.10 These authors also noted the heterogeneity of the country and its
inhabitants with bremarkable differences not only among provinces or regions, but within some
provincesQ (p. 55). These authors also noted bthe continuous decrease of first-cousin marriages as
opposed to the continuous increase of second-cousin marriagesQ (p. 68).
Consanguinity may have been especially high in some rural areas and among isolated populations,
such as those living on small islands and in landlocked valleys. Bestard (1991) offered one of the best
examples of an isolated Spanish population maintaining high levels of local endogamy and
consanguinity over time. In his study of the Balearic island of Formentera, he found that almost
half of the marriages celebrated from 1872 to 1888 involved relatives. Relatively few, however, were
between first cousins—only 8 of 235, or 3.4%. In local terms, autochthonous people were appropriate
spouses. That implied kin marrying kin because, as natives said: bHere we all are related.Q But
marrying first cousins was morally wrong, as bthey are too closeQ (massa familia) (see Bestard, 1991,
pp. 163–169).
Levels of inbreeding, however, decreased rapidly throughout Spain in the 1960s and 1970s to rates
comparable to those of other western European countries (Calderón et al., 1993, 1998; Fuster et al.,
1996; Gómez Gómez, 1989). In recent decades, rates of inbreeding may have declined even further.
On the other hand, consanguinity has remained high among gitanos in the face of urbanization,
migration, and industrialization. In fact, the decrease in infant mortality derived from better living
standards and health care may have increased the number of those who could find a close relative to
marry.
The Gypsy preference for marrying within the extended family is not a remnant from the past
maintained by the blind pursue of binding traditions. It is more often the response to ethnic segregation
and discrimination, and a cultural choice with social and economic advantages. For instance,
consanguineous unions may be seen as breducing the risks and hidden uncertainties regarding health
or other unfavorable family characteristicsQ (Bittles, 1994, p. 567).
Among gitanos, however, marriage transactions, such as dowry or bridewealth, rarely matter much,
and the maintenance of family property is not usually a major consideration. Marrying kin reserves the
girls of the family (and their work and reproductive capacity) for the men of the family. It also increases
the compatibility between spouses, and between them and their in-laws. Thus, in case of patrilocal
residence, consanguineous marriages may benefit the status and relative autonomy of the bride (Khlat,
1989). In any case, although the motives for these marital patterns are a crucial subject for research, they
lie beyond the objectives of this article.
The high rates of consanguinity among Spanish Gypsies may have important consequences both
social and biological. On one hand, they increase the solidarity of Gypsies along family lines, but on the
other, they also tend to isolate kin groups from one another and from society at large. And these patterns
may also have important consequences for genetic and congenital defects.
In the clinical studies available, gitanos appear to have up to seven times more autosomal recessive
syndromes than other Spaniards (see Martı́nez-Frı́as, 1993). We have found important instances of
congenital defects in the population studied, such as a high frequency of congenital deafness and
albinism in some of the reconstructed genealogies. This is an important result to consider in future
preventive efforts and in education and genetic counseling programs. Any such project should be

10
The average values of the inbreeding coefficient oscillated from a maximum of 0.00211 in 1920–1924 to 0.00142 in 1940–1943 (Pinto-
Cisternas et al., 1979, p.63).
 E. Martin, J.F. Gamella / The History of the Family 10 (2005) 45–63 61

tailored to the specific needs and understandings of this population, avoiding any form of
stigmatization. Moreover, we should bkeep the levels of expressed genetic defect in perspectiveQ and
address the diverse forms in which consanguinity and other marital patterns act bas a covariable with
other sociodemographic parameters,Q such as maternal education, poverty, exclusion, and discrimination
(Bittles, 1994, p. 578).

Acknowledgments

We want to acknowledge the help of some of our students and collaborators in the collection of data
and extend our thanks especially to Alejandro Quesada, Mercedes Alba, and Mari Luz Flores for their
help and kind assistance.

References

Acton, T. A. (1979). Academic success and political failure: A review of modern social science writing in English on gypsies.
Ethnic and Racial Studies, 2, 231 – 252.
Ardévol, E., & Dolores del Pino, M. (1986). Antropologı́a urbana de los gitanos de Granada. Granada7 Ayuntamiento de
Granada.
Assal, S., Susanszky, E., & Czeizel, A. (1991). High consanguinity rate in Hungarian gypsies communities. Acta Paediatrica
Hungarica, 31, 299 – 304.
Barañi, E. (2001). Mujeres gitanas y sistema penal. Madrid7 Metyel.
Bestard, J. (1991). What’s in a relative? Household and family in formentera. New York7 Berg.
Bideau, A., & Poulain, M. (1984). De la genealogie a la demographie historique: Genealogie ascendante et analyse
demographique. Annales de Demographie Historique, 55 – 68.
Bittles, A. H. (1994). The role and significance of consanguinity as a demographic variable. Population and Development
Review, 20, 561 – 584.
Bittles, A. H., Mason, W. M., Greene, J., & Appaji Rao, N. (1991). Reproductive behavior and health in consanguineous
marriages. Science, 252, 789 – 794.
Calderón, R., Peña, J. A., Delgado, J., & Morales, B. (1998). Multiple kinship in two Spanish regions: New model relating
multiple and simple consanguinity. Human Biology, 70, 535 – 561.
Calderón, R., Peña, J. A., Morales, B., & Guevara, J. I. (1993). Inbreeding patterns in the Basque Country (Alava Province,
1831–1980). Human Biology, 65, 743 – 770.
Cavalli-Sforzza, L. L., & Bodmer, W. F. (1971). The genetics of human populations. San Francisco7 W.H. Freeman.
Cazen, M. -H., & Cazen, P. (1996). Comment mesure la profundeur gènèaloguique d’une ascendance? Population, 1, 117 – 141.
Cohn, W. (1973). The gypsies. Reading, MA7 Addison-Wesley.
Daroczi, A. (1994). La situation des Tziganes dans Hongrie. In C. Auzias (Ed.), Les Familles Roms d’Europe de l’Est (pp. 37 – 42).
Paris7 IDEF.
Dupâquier, J. (1993). Généalogie et demographie historique. Annales de Démographie Historique, 391 – 395.
Fraser, A. M. (1992). The gypsies. Oxford7 Blackwell.
Fresno, J. M. (1994). Análisis socioantropológico sobre la situación actual de la comunidad gitana en España. Madrid:
Secretariado General Gitano. Documentos Técnicos, n.2.
Fuster, V., Morales, B., Mesa, M. S., & Martı́n, J. (1996). Inbreeding patterns in the Gredos Mountain Range (Spain). Human
Biology, 68, 75 – 93.
Gamella, J. F. (1996). La población gitana en Andalucı́a. Un estudio preliminar sobre sus condiciones de vida. Sevilla7
Consejeria de Trabajo y Asuntos Sociales. Junta de Andalucı́a.
Gamella, J. F. (2000). Mujeres gitanas: matrimonio y género en la cultura gitana de Andalucı́a. Sevilla7 Consejeria de Asuntos
Sociales.
62 E. Martin, J.F. Gamella / The History of the Family 10 (2005) 45–63

Gamella, J. F. (2004). Exclusión social y diferencia étnica: el caso de los gitanos. In J. F. Tezanos (Ed.), Tendencias en
desigualdad y exclusión social (pp. 603 – 647). Madrid7 Sistema.
Gamella, J. F., & Sanchez-Muros, P. (1998). La imagen infantil de los gitanos. Estereotipos y prejuicios dominantes en escuelas
multiétnicas. Valencia7 Bancaixa.
Gay Blasco, P. (1999). Gypsies in Madrid. Sex, gender and the performance of identity. Oxford7 Berg.
Gheorge, N. (1994). Les enfants Tziganes en Roumanie. In C. Auzias (Ed.), Les Familles Roms d’Europe de l’Est (pp. 24 – 27).
Paris7 IDEF.
Gómez Alfaro, A. (1993). La gran redada de gitanos: La prisión general de gitanos en 1749. Madrid7 Presencia Gitana.
Gómez Alfaro, A. (1998). Gitanos de Priego de Cordoba: Entre el acoso y la superviviencia. Legajos, 1, 61 – 72.
Gómez Alfaro, A. (1999). Tipologı́as. Matrimonios mixtos y mestizajes gitanos en los censos históricoa andaluces. Demófilo,
30, 30 – 52.
Gómez Gómez, P. (1989). Consanguinity: Geographical variation and temporal evolution in the north of the Iberian peninsula,
1918–1968. International Journal of Anthropology, 4, 119 – 124.
Goody, J. (1990). The Oriental, the Ancient and the Primitive: Systems of marriage and the family in the preindustrial societies
of Eurasia. Cambridge7 Cambridge University Press.
Gresham, D., Morar, B., Underhill, P. A., Passarino, G., Linn, A. A., Angelicheva, D., Calafell, F., et al. (2001). Origins and
divergence of the Roma (Gypsies). American Journal of Human Genetics, 69, 1314 – 1331.
Hajnal, J. (1965). European marriage patterns in perspective. In D. V. Glass, & D. E. C. Eversley (Eds.), Population in History:
Essays in historical demography (pp. 101 – 143). Chicago7 Aldine.
Henry, L. (1956). Anciennes familles genevoises. Etude démographique: XVI–XX siécle. Paris: P.U.F. INED, Cahier 26.
Henry, L. (1967). Manuel de demographie historique. Paris7 Librarie Droz.
Hollingsworth, T. H. (1976). Genealogy and historical demography. Annales de Demographie Historique.
Hooz, I. (1987). A Cigany népesség számának alakulása [Changes in the Gypsy population]. Statisztikai Szemle, 65, 220 – 233.
IEA (Instituto de Estadı́stica de Andalucı́a) (1999). Un siglo de demografı́a en Andalucı́a: la población desde 1900. Sevilla7
Junta de Andalucı́a.
Jette, R., & Charbonneau, H. (1984). Généalogies descendantes et analyse demographique. Annales de Demographie
Historique, 45 – 54.
Kalaydjieva, L., Gresham, D., & Calafell, F. (2001). Genetic studies of the Roma (Gypsies): A review. BMC Medical Genetics,
2, 5.
Kalibova, K. (1994). Les Tsiganes en République Tcheque et en République Slovaque. Espace Populations Sociétés, 3,
359 – 362.
Khlat, M. (1989). Les marriages consanguins à Beyrouth. Paris7 Institut National d’Etudes Démographiques.
Lambert, E. Y. (Ed.). (1990). The collection and interpretation of data from hidden populations. Rockville, MD: National
Institute on Drug Abuse. Research Monograph, 98.
Leblon, B. (1985). Les Gitans d’Espagne. Paris7 Presses Universitaires de France.
Leblon, B. (1990). Musiques tsiganes et flamenco. Paris7 L’Harmattan.
Livi-Bacci, M. (1968a). Fertility and nuptiality changes in Spain from the late 18th to the early 20th century. Part I. Population
Studies, 22, 83 – 103.
Livi-Bacci, Massimo (1968b). Fertility and nuptiality changes in Spain from the late 18th to the early 20th century. Part II.
Population Studies, 22, 211 – 235.
Macfarlane, A. (1986). Marriage and love in England, 1300–1840. Oxford7 Basil Blackwell.
Mann, A. B. (1992). Vyvoj Romskej rodiny na priklade troch spisskych obci [Reconstitution of Gypsy families using data from
three communes. Development of the Romany family (using the example of 3 communes)]. Demografie, 34, 118 – 130.
Martı́n Carrasco-Muñoz, E. (1999). La estructura de la población gitana andaluza: notas introductorias. Demófilo, 30,
89 – 105.
Martı́nez-Frı́as, M. L. (1993). Malformaciones Congénitas en la Población Gitana. Estudio epidemiológico en un grupo de la
población gitana. Madrid: Documentos 38/93. Real Patronato de Prevención y de Atención a Personas con Minusvalı́a.
Martinez-Frias, M. L., & Bermejo, E. (1992). Prevalence of congenital anomaly syndrome in a Spanish Gypsy population.
Journal of Medical Genetics, 29, 483 – 486.
Meszaros, A., & Foti, J. (1996). A Cigany Nepesseg Jellemzoi Magyarorszagon. Statisztikai Szemle, 74, 908 – 929.
Okely, J. (1983). The Traveller-Gypsies. Cambridge7 Cambridge University Press.
Pasqualino, C. (1998). Dire le chant. Les Gitanes Flamencos d’Andalousie. Paris7 CNRS Editions.
 E. Martin, J.F. Gamella / The History of the Family 10 (2005) 45–63 63

Pinto-Cisternas, J., Zei, G., & Moroni, A. (1979). Consanguinity in Spain, 1911–1943: General methodology, behavior of
demographic variables and regional differences. Social Biology, 26, 55 – 71.
Post, W., van Poppel, F., van Imhoff, E., & Kruse, E. (1997). Reconstructing the extended kin network in the Netherlands with
genealogical data: Methods, problems, and results. Population Studies, 51, 263 – 278.
Reher, D. S. (1991). Marriage patterns in Spain, 1887–1930. Journal of Family History, 16, 7 – 30.
Reher, D. S. (1996). La familia en España: Pasado y Presente. Madrid7 Alianza Editorial.
Reher, D.S, Nieves Pombo, M., & Nogueras, B. (1993). España a la luz del Censo de 1887. Madrid7 INE.
Rowland, R. (1987). Mortalidad, movimientos migratorios y edad de acceso al matrimonio en la Penı́nsula Ibérica. Boletı́n de la
Asociación de Demografia Histórica, 5, 41 – 63.
San Román, T. (1986). Entre la marginación y el racismo. Reflexiones sobre la vida de los gitanos. Madrid7 Alianza
Universidad.
San Román, T. (1997). La diferencia inquietante: viejas y nuevas estrategias culturales de los gitanos. Madrid7 Siglo XXI.
Sanchez Ortega, M. H. (1977). Documentación selecta sobre la situación de los gitanos españoles en el siglo XVIII. Madrid7
Editora Nacional.
Sanchez Ortega, M. H. (1986). Evolución y contexto histórico de los gitanos españoles. In T. San Roman (Ed.), Entre la
marginación y el racismo. Reflexiones sobre la vida de los gitanos. Madrid7 Alianza Editorial.
Srb, V. (1998). Zmeny v reprodukci Ceskoslovenskych Romu 1970–1980 [Changes in reproduction of the Czechoslovak
Romas, 1970–1980]. Demografie, 30, 305 – 309.
Thomas, J. D., Doucette, M. M., Thomas, D. C., & Stoeckle, J. D. (1987). Disease, lifestyle, and consanguinity in 58 American
Gypsies. Lancet, 15, 377 – 379.
Valls, A. (1982). Antropologı́a de la consanguinidad. Madrid7 Editorial de la Universidad Complutense.

View publication stats