Introduction
The Doctrine of Belén was settled in 1777 by splitting off from the Doctrine of Codpa, which was divided by Manuel de Abad Yllan, both under the authority of the Codpa Chiefdom (cacicazgo in Spanish) (Hidalgo & Durston, Reference Hidalgo, Durston and Hidalgo2004). It is located in the Altos de Arica (current Chilean territory), which can be divided into five distinct ecological zones: Puna (Bolivian altiplano), the sierra (2000–3500 m.a.s.l.), valley headwaters (˜2000 m.a.s.l.), the valley (< 2000 m.a.s.l.) and the Pacific coast. Ten settlements or localities were under its jurisdiction, including Belen, the seat of the doctrine’s ecclesiastical administration, Socoroma, Putre, Pachama, Parinacota, Caquena Choquelimpe, Guallatiri, Sora and Churiña. Figure 1 shows the locations of these settlements, along with others belonging to the Doctrine of Codpa: Codpa, Esquiña, Pachica, Timar, Ticnamar, Saxamar, Umagata and Livilcar (Inostroza, Reference Inostroza2017).
Figure 1. Localities of the Doctrine of Belén. Current borders of Chile, Peru and Bolivia. Source: authors’ elaboration based on Google Earth satellite photo, versión 9.140.0.4 (2021). URL: https://earth.google.com/web/@-18.79454398,-69.76675868,2146.19084742a,464229.77881908d,30y,0h,0t,0r/data=MicKJQojCiExYTNZWVo2UmtJV2hreld4OUstZlBZZ0lRYmo3UV9lM1U6AwoBMA?authuser=0.
The socio-demographic and economic features of the Altos de Arica area distinguish it from the rest of the Andean Aymara region. Around this area, the silver and quicksilver trade expanded to and from Potosí, which had a significant influence on the way of life of local indigenous people and their circulation from different areas along the Andean coastal trade route (Hidalgo, Reference Hidalgo1987; Dagnino, Reference Dagnino, Ruz Zagal and Pizarro Pizarro2016; Inostroza, Reference Inostroza2019). Local enterprises included agriculture, stockbreeding and horseback transportation. Since the Andean peoples in Altos de Arica were exempted from the Potosí mita, migrating to the Doctrine of Belén was an attractive prospect for many foreigners to escape from the pressure of the mita and potentially gain access to lands suitable for settlement (Hidalgo, Reference Hidalgo1986, Reference Hidalgo1987; Inostroza, Reference Inostroza2019).
Archaeological research has confirmed that this area has been inhabited since pre-Hispanic times, first as part of the Altiplano kingdoms (Lupaca and Pacajes) and later, under Inka rule (Santoro et al., Reference Santoro, Hidalgo and Osorio1987; Hidalgo & Focacci, Reference Hidalgo, Focacci and Hidalgo2004; Chacama Reference Chacama2005, Reference Chacama2009, among others). In the 18th century, the valley of Belén presented a diverse, multi-ethnic landscape where indigenous communal lands co-existed alongside estates owned by criollos (i.e. people of European ancestry born in America) (Marsilli & Hidalgo, Reference Marsilli, Hidalgo, Orellana and Muños1992; Hidalgo et al., Reference Hidalgo, Farías, Guzmán, Arévalo and Hidalgo2004; Inostroza, Reference Inostroza2017).
The demographic structure of the Doctrine of Belén at the turn of the 19th century has been described in detail by Hidalgo (Reference Hidalgo1986), Inostroza (Reference Inostroza2019) and Hidalgo and Inostroza (Reference Hidalgo and Inostroza2019) based on diverse documentary sources. The old Corregimiento of Arica extended over 58073 km2 with 12,690 people in the mid-18th century (Hidalgo, Reference Hidalgo1986). In 1792, the province of Arica had a total of 18,711 inhabitants, 30% of whom were located in Tacna (currently Perú) and 20.9% in the Codpa Chiefdom. Approximately 83% of the total population was indigenous, and over 70% was subject to tribute (Hidalgo & Inostroza, Reference Hidalgo and Inostroza2019). According to census data, the Doctrine of Belén’s population grew by 2.03% between 1750 and 1792 and fell by 2.17% between 1792 and 1813. Between 1750 and 1813, 50% of all men were under 19 years of age, while 50% of all women were under 24 years old and under 29 years old in 1772 and 1813, respectively.
Among women, the average age of marriage at the turn of the 19th century was 23 years. Most of them got married as single women (79%) while others as widows (21%). Also, most of the children with recorded baptism were born to legitimate marriages (85.98%), with the rest born out of wedlock (Inostroza, Reference Inostroza2019). The method for reconstructing family structure employed by Inostroza (Reference Inostroza2016) suggests that ca. 1787 the Doctrine of Belén was mainly composed of nuclear families (65%), followed by women-headed households (widows, single women or otherwise alone) (20%) and 15% by composite families (Inostroza, Reference Inostroza2016).
The isonymic method and historical documents
The concept of biodemographic structure refers to the overall number of individuals, their distribution by sex and age and the frequency of births, deaths and migrations. These factors determine gene dispersion over time and space – and among or within populations, which conditions population size (Cavalli-Sforza & Bodmer, Reference Cavalli-Sforza and Bodmer1981; Fuster, Reference Fuster, Rebato, Susanne and Chiarelli2005). In turn, the genetic structure of a population is shaped by a wide range of factors, such as age, kinship and religion, among others, which constrain partner selection and limit random mating, as well as social or physical barriers (or lack thereof), which shape migration dynamics (Fuster, Reference Fuster, Malgosa Morera, Nogués and Aluja2003).
Besides demography and genetics, populations are also structured by the geography determining the distribution of a population’s settlements, including soil conditions, weather, hydrologic and orographic characteristics, political boundaries and settlement preferences (Haggett & Ferrer, Reference Haggett and Ferrer1994). Together, these factors define a geographic hierarchical population structure where increasing complexity is observed, depending on the organizational mechanisms implemented. Over time, less-complex groups are modified by migratory movements that are subject to social, cultural and economic factors, shaping the geographic, demographic and genetic structure of the population (Haggett & Ferrer, Reference Haggett and Ferrer1994).
Consequently, the study of population structure is not a simple task. It involves careful consideration of all the elements that have contributed to said structure and their complex correlations. Thus, building an accurate model of the studied phenomena requires integrating a wide range of information sources and methodological approaches that consider as many factors as possible.
To this end, surname analysis becomes a valuable tool since surnames are a socio-cultural variable that can be easily related to biological features, enabling the study of diverse issues related to population structure (Lasker, Reference Lasker1980). Surnames can be evaluated in terms of ethnic adscription. Also, they reveal identity and biological kinship and can be traced through time, which makes surname analysis a significant methodological resource for bioanthropology and human population genetics. Assuming that two people holding the same surname belong to the same lineage, surnames can be considered equivalent to genetic markers. Therefore, their frequency and distribution in a population supply valuable quantitative data on the genetic structure of human populations.
The isonymic method and the concept of isonymy, defined as the occurrence of an identical surname in two different individuals (Rawling, Reference Rawling, Roberts and Sunderland1973) or the proportion in which this phenomenon occurs (Lasker, Reference Lasker1985), were initially proposed by Darwin (Reference Darwin1875) and later revived by Crow and Mange (Reference Crow and Mange1965). Darwin (Reference Darwin1875) suggested using the proportion of isonymous unions to estimate the frequency of first-cousin marriages. Based on the same principle, Crow and Mange (Reference Crow and Mange1965) developed and expanded the isonymic method to study consanguinity in the population as a whole – not limited to first cousins – under the assumption that all individuals that share a surname have inherited it from a common ancestor. Crow and Mange’s approach (Reference Crow and Mange1965) provided a way to overcome methodological challenges to estimate consanguinity in populations whose genetic or genealogical data were incomplete or not available.
Studies using the isonymic method to analyse population structure have extracted surname information from diverse documentary sources, such as parochial birth, marriage and death registers and census data (Roberts & Rawling, Reference Roberts and Rawling1974; Pollitzer et al., Reference Pollitzer, Smith and Williams1988; Lasker, Reference Lasker1980, Reference Lasker1985; MacRaild & Smith, Reference MacRaild and Smith2009). In the Andean region, the method has been frequently applied in the study of historical population structure. Dipierri and colleagues (Reference Dipierri, Ocampo and Russo1991) analysed the isonymic structure in the Quebrada de Humahuaca based on parochial baptismal records (1734–1810) and observed increased consanguinity over time, with intermediate values for first cousins and second cousins once removed. A similar approach has been used to study the historical populations in the Jujenean Puna, particularly in Casabindo and Cochinoca, and in San Pedro de Atacama in Northern Chile (Costa Junqueira et al., Reference Costa Junqueira, Martínez-Campos, Dipierri, Bejarano and Alfaro2000; Alfaro et al., Reference Alfaro, Dipierri, Albeck and Morales1998, Reference Alfaro, Dipierri, Albeck and Morales1999; Alfaro, Reference Alfaro2010; Peña Aguilera, Reference Peña Aguilera2016).
The documentary sources used to reconstruct historical demography in the Doctrine of Belén (Altos de Arica) between 1763 and 1820 are notable for their stable patronymics, complete marriage, baptism and death records and low migration rates (Inostroza, Reference Inostroza2019). Besides, a substantial percentage of individuals repeatedly appear in these records, indicating stable residence in the doctrine. When applying the isonymic method, such conditions constitute a significant advantage.
Given these favourable biodemographic and documentary conditions, the research objective of this study was to reconstruct historical population structure and migration patterns for the Doctrine of Belén’s population between 1750 and 1813 by analysing names and surnames through the isonymic method.
Methods
Documentary sources
Name and surname data were collected from two types of documentary sources. The first were tax rolls from the Doctrine of Belén dated 1750 (Hidalgo, Reference Hidalgo1978), 1772/1773 (Hidalgo et al., Reference Hidalgo, Farías, Guzmán, Arévalo and Hidalgo2004) and 1813 (Hidalgo et al., Reference Hidalgo, Arévalo, Marsilli and Santoro1988). The former two included the total Andean population that was tributary to the Chiefdom of Codpa. Only the ten villages or towns that constitute the Doctrine of Belén were included in this study: Belén, Socoroma, Putre, Pachama, Parinacota, Caquena, Choquelimpe, Guallatiri, Sora and Churiña. The second documentary source was parochial marriage, baptism and death records between 1763 and 1834 (Inostroza, Reference Inostroza2019). The period’s tax list (Lista de revisitas in Spanish) had a population coverage comparable to that of a census and was consistent with parochial data, especially the names and surnames of the doctrine residents and their nuclear families (Inostroza & Hidalgo, Reference Inostroza and Hidalgo2015). The census records contained information on age, sex, category, married status and, in some cases, political or religious appointment.
Surnames were grouped by similar phonetics or spelling. Children were assigned the paternal surname, in the case of nuclear families, or the maternal surname, for single mothers, since the review of parochial records revealed that this was how surnames were inherited (Inostroza, Reference Inostroza2019). The analysis considered only the first one when an individual had more than one name or surname.
Surnames as ethnic markers
The anthroponyms recorded in these documentary sources were classified according to their geographic-linguistic origin into:
-
a) Autochthonous: surnames with the phonetic or graphic influence of extant or extinct native languages, as suggested by documentary sources: catalogues, dictionaries, toponyms, electoral rolls, among others (Erdman, Reference Erdman1964; Albeck et al., Reference Albeck, Alfaro, Dipierri and Chaves2017; Inostroza, Reference Inostroza2019);
-
b) Foreign: surnames that originated in other continents;
-
c) Autochthonous/foreign: surnames that could fall into either category;
-
d) Undetermined: surnames that could not fall into any of the preceding categories.
Isonymic analysis
Based on the distribution of surnames within each period, the expected random isonymy (I ii ) was calculated assuming random mating within a population k, following Rodríguez-Larralde and Casique (Reference Rodríguez-Larralde and Casique1993):
$${I_{ii}} = \Sigma {n_{ik}}\frac{{\left( {{n_{ik}} - 1} \right)}}{{{N_i}\left( {{N_i} - 1} \right)}} = \Sigma {\left( {{n_{ik}}/{N_i}} \right)^2} - 1/{N_i}$$
Here, n ik is the absolute frequency of surname k in population i; N i is the total number of individuals in the same population (∑n ik ).
Consanguinity was estimated as an expression of the random isonymy (F ST) and calculated according to Relethford (Reference Relethford1988) using the following formula:
$${F_{{\rm{ST}}}} = {I_{ns}} / 4$$
where I ns is non-biased isonymy.
Other statistics derived from these indicators were used to describe population structure:
Surname diversity or Fisher’s α
This indicator was first proposed by Fisher et al. (Reference Fisher, Corbet and Williams1943) to estimate the abundance of animal species representative of a sample. Piazza et al. (Reference Piazza, Rendine, Zei, Moroni and Cavalli-Sforza1987) proposed its use to analyse migration rates from the distribution of surnames by estimating the existing number of surnames in a population. It measures the effective number of surnames in a population and is analogous to gene drift (Barrai et al., Reference Barrai, Formica, Scapoli, Beretta, Volinia, Barale, Ambrosino and Fontana1992). Fisher’s α was calculated according to the following formula:
$$\alpha = 1 / {I_{ns}}$$
Recent immigration or Karlin–McGregor’s ν
Since surnames are considered as multiple alleles of the same locus (Piazza et al., Reference Piazza, Rendine, Zei, Moroni and Cavalli-Sforza1987), they can be analysed using the Karlin and McGregor (Reference Karlin and McGregor1967) theory, which predicts that dead individuals are replaced by others bearing the same surname, externally and with a rate v, by mutations of surnames and by the introduction of other surnames through immigration. Since the Karlin–MacGregor v equals the sum of the mutated surnames, a rare phenomenon in societies with a regular transmission of patronymics, plus the much more frequent immigration rate, it is reasonably assumed that v equals this rate (Zei et al., Reference Zei, Guglielmino, Siri, Moroni and Cavalli-Sforza1983; Piazza et al., Reference Piazza, Rendine, Zei, Moroni and Cavalli-Sforza1987). It was calculated according to Zei et al. (Reference Zei, Guglielmino, Siri, Moroni and Cavalli-Sforza1983) with the following formula:
$$\nu = \left( {1 - {I_s}} \right) / \left[ {{I_s}\left( {{N_i} - 1} \right)} \right]$$
where I s is biased isonymy.
The actual number of migrants (Nem)
Wright’s Island Model (Reference Wright1951) allows the estimation of the definite number of migrants in a population from F ST using the following equation (Yasuda & Furusho, Reference Yasuda and Furusho1971):
$${N_{em}} = \left( {1 - {F_{{\rm{ST}}}}} \right) / 4{F_{{\rm{ST}}}}$$
The following indicators of isolation and sedentarism were also determined (Rodríguez Larralde, Reference Rodríguez-Larralde1990).
-
i. Indicator A: Percentage of the population with unique surnames, i.e. surnames held by only one individual in the population. This indicator can also be considered an indicator of migration (Rodríguez Larralde & Barrai, Reference Rodríguez-Larralde and Barrai1997, Reference Rodríguez-Larralde and Barrai1998).
-
ii. Indicator B: Percentage of the population that held the seven more frequent surnames. It is an indicator of relative isolation (Rodríguez Larralde & Barrai, Reference Rodríguez-Larralde and Barrai1997, Reference Rodríguez-Larralde and Barrai1998), as larger B values correspond to high isolation and sedentarism of a population.
The parameters I ii , F ST, ν and α were calculated using the DISTRIBU, ORMANOS and ISONIMIA programs developed and kindly shared by Dr Italo Barrai and Alvaro Rodriguez Larralde for the DBXL database processor.
Results
Geographic-linguistic origin of surnames
The 1750 document recorded 1306 individuals and 176 different surnames, of which 54% were foreign, 33.5% autochthonous, 6.8% could be either foreign or autochthonous (A/F) and 5.7% were undetermined. The 1772/1773 document listed 1717 individuals and 178 surnames, of which 59% were foreign, 30.9% were autochthonous, 6.2% were A/F and 3.9% were undetermined. The 1813 document showed a decrease in population size, recording 1445 people and 159 different surnames with similar origin distributions to those of 1772: most were foreign (59.7%), followed by autochthonous (29.6%), A/F (6.9%) and undetermined (3.1%) (Table 1). Although most individuals carried foreign surnames in all the considered records, the seven most frequent surnames in 1750 and 1772/1773 were all autochthonous. Not until 1813 did the foreign surname ‘Flores’ became one of the seven most frequent (n=42).
Table 1. Doctrine of Belén population size and absolute and relative frequencies of surname types by period
* Percentage of total unique surnames.
The contrast between the distribution of surnames by origin (i.e. mainly foreign) and the seven most frequent surnames (i.e. all or mainly autochthonous) is due to the number and origin of unique surnames. In the three periods considered, 56.9%, 63.5% and 60.0% of unique surnames had a foreign root, respectively.
Isonymic parameters
Table 2 presents the distribution of isonymic parameters, which had a slight variation over time. Population size increased between 1750 and 1772 but fell again in 1813. The number of distinct surnames was nearly the same in the first two periods, decreasing in the third one.
Table 2. Isonymic parameters of the Doctrine of Belén population by period
Random isonymy (F ST) values suggested that the population’s consanguinity rate remained stable through time. Surname diversity or abundance (α) did not vary significantly between periods. Moreover, in 1772/1773 there was an increase in population size, but α had the lowest value of the three studied periods, indicating that the increased number of individuals did not imply an inflow of new surnames. The recent immigration indicator (v), which presents its lowest value in 1772, supports the claim that this rise was not a consequence of the arrival of immigrants with new surnames to the Doctrine of Belén but was given by the people within the existing surname set.
Indicator A dropped by nearly 50% between 1750 and 1813, implying a sharp decrease in immigration to the Doctrine of Belén, while indicator B suggested significant population sedentarism across the three periods since over 25% of the population had the same seven surnames. Similarly, there was no significant variability in the effective number of migrants (N em) through time, which implies low population mobility in all periods considered.
Discussion
Given the limitations of the isonymic method previously mentioned above, the results must be considered as relative rather than absolute measures of biological affinity within and between populations when interpreting population structure based on a population’s anthroponymy. In the Doctrine of Belén, the results indicated a relative homogeneity between periods, in terms of both the population structure by geographic–linguistic origin of surnames and isonymic structure.
Consanguinity by random isonymy (F ST) and indicator B (sedentariness) values were conditioned by the people’s origin, the persistence of the traditional ayllu structure and marriage preferences within the Doctrine of Belén. According to parochial records (1763–1853), 73% of the population were natives, i.e. born within the doctrine. However, this percentage could have been even higher, considering that another administrative category (7.3%) was a neighbour (vecinos in Spanish), which included Spaniards or mestizos (mixed) that had settled in the doctrine’s towns (Inostroza, Reference Inostroza2019).
There is documentary evidence of the traditional Andean social structure known as ayllu in the area (Hidalgo et al., Reference Hidalgo, Farías, Guzmán, Arévalo and Hidalgo2004; Inostroza, Reference Inostroza2016). The Revisita de Codpa of 1773 states that the town of Belén was divided into two ayllus, Mancasaya and Aransaya (Hidalgo et al., Reference Hidalgo, Farías, Guzmán, Arévalo and Hidalgo2004; Inostroza, Reference Inostroza2016). In the most historical version of the Andean model of social and territorial organization, all the ayllu members were considered to share a common ancestor, belonging to the same lineage. In other words, an ayllu was composed of a group of nuclear families, bound to a specific territory and connected by kinship ties. They shared a common language and participated in collective labour, which favoured consanguinity and endogamy (Dipierri et al., Reference Dipierri, Alfaro, Cháves, Albeck and Rodríguez2015). From Belen’s 1787 tithe payer list, the origin of spouse was identified, indicating that 62% of all marriages were between individuals from the ayllus Mancasaya and Aransaya, and therefore, endogamous. Only 2.1% of all unions involved individuals from other doctrines (Inostroza, Reference Inostroza2016). Such precedents endorse the great degree of sedentariness among the Doctrine’s population found in this study.
In general, low values in indicator A and high values in indicator B usually represent high levels of isolation and sedentism. In the Doctrine of Belén, this was probably the result of isolation due to social factors, such as the ayllu structure, rather than geographic ones. Today, these towns are considered relatively isolated due to their high altitude (>3500 m.a.s.l.), close to the Andean massif, and routes that are mainly regional, primary, winding and unpaved. In the 1970s, the province of Parinacota was legally classified as an ‘isolated territory’ due to its extreme physical environment and historical settlement patterns, which are distinguished and asymmetric as compared to the rest of the country (Subsecretaría de Desarrollo Regional y Administrativo, 1999; Sánchez, Reference Sánchez2009). During the colonial period, and despite its significant role within the Andean trade network, isolation conditions in the doctrine was assuredly even more pronounced, thus promoting endogamy and sedentism. The indicators of evolutionary dynamics coincide with those analysed so far since both ν (recent immigration) and α (surname diversity) indicate a high degree of stability in the population.
Historical demography studies suggest that the fluctuations in population size observed in the Doctrine of Belén were the result of migration dynamics rather than natural growth (Hidalgo & Inostroza, Reference Hidalgo and Inostroza2019). In 1750, there were 1497 individuals in the doctrine’s records; by 1772/3, there were 1818; and 20 years later, 2023. Afterwards, the population size decreased; there were 1533 individuals recorded in 1804 and 1495 in 1813. These changes in population size are partially related to the behaviour of Karlin–McGregor’s ν and indicator A. Both decreased between 1750 and 1772 and increased again between 1772 and 1813. Despite such population increase between 1750 and 1772, migration was driven by familial bonds or kinship networks enhanced by compadrazgo and marital alliances (Inostroza, Reference Inostroza2019). These prevented the introduction of different surnames to the population, as observed in the number of new surnames among periods (176 vs 178). Another possible explanation is that these immigrants had the same surnames as the doctrine locals because these surnames were also highly prevalent in the neighbouring doctrines. Historical biodemography indicates that most of these immigration flows came from the Bishoprics of Charcas and La Paz, in what is now the Plurinational State of Bolivia, and that the most prevalent surnames among these foreigners were Mamani, Choque and Calle, which were already very frequent in the Doctrine of Belén (Table 1).
In terms of surname origin, autochthonous surnames represented over 30% of the total surnames recorded in each of the three documents. Meanwhile, individuals bearing these surnames decreased from 55.6% in 1750 to 49.3% in 1813, as expected from doctrines or Indian towns (Inostroza, Reference Inostroza2019). However, they do differ from those presented by other contemporary Andean populations. After a long period of acculturation, in 1786, only a third of Oruro’s inhabitants held an autochthonous surname (Sánchez Albornoz, Reference Sánchez Albornoz and Henry1974). The Humahuaca Parish expressed similar percentages between 1734 and 1810 (Dipierri et al., Reference Dipierri, Ocampo and Russo1991). In contrast, 65% of the surnames recorded in Casabindo were autochthonous in the early 19th century, representing 92% of the local population (Alfaro Reference Alfaro2010), while in Cochinoca, between 1778 and 1806, these figures were 45% and 62–66%, respectively (Peña Aguilera, Reference Peña Aguilera2016). The fact that approximately half of Belén’s population held an autochthonous surname suggests that the disruption of Andean onomastics caused by Christian evangelization in this area was not as drastic as in other Andean communities. This feature of local anthroponyms also promoted stability in the isonymic structure of the Doctrine of Belén over, at least, 63 years.
The relatively high prevalence of foreign surnames – over 50% of overall surnames – indicates that the replacement of autochthonous names with foreign appellatives took place relatively early in the Doctrine of Belén, which is evidence of a greater degree of social transformation, compared with Casabindo and Cochinoca. The observed difference with these populations from the central Jujuy’s Puna could be explained by the fact that Casabindos and Cochinocas were reduced and put in an encomienda. Indeed, they were a particular group that did not share the fate of other native populations from neighbouring areas which, as a result of differential fragmentation or ‘de-structuring’ (Lorandi, Reference Lorandi1997), had begun a process of colonial integration that led to the disappearance of native onomastics and replacement by European ones (Dipierri, Reference Dipierri2004).
There are a few comparable studies of nearby Andean populations regarding the same time frame and analogous documentary sources that ponder the magnitude of consanguinity in the Doctrine of Belén. Among them, Dipierri et al. (Reference Dipierri, Ocampo and Russo1991) estimated consanguinity by isonymy in the historical population of Humahuaca from surname data from baptismal registers (1734–1810). Costa-Junqueira and colleagues (Reference Costa Junqueira, Martínez-Campos, Dipierri, Bejarano and Alfaro2000) analysed the evolution of consanguinity and kinship in the Puna de Atacama between 1800 and 1950 by applying the isonymic method to surnames in death registers of Susques (Argentina), San Pedro de Atacama and Toconao (Chile). These studies determined the coefficient of consanguinity by random isonymy, the immigration indicator ν (Karlin–McGregor’s), the surname diversity or Fisher’s α and the isolation indicator B. Overall, the region presented an early stage of considerable population mobility, followed by progressive isolation of populations located on both sides of the Andes range. Of all the localities considered, Susques presented the highest consanguinity and sedentism levels.
Table 3 presents the values for consanguinity by random isonymy, surname diversity (α) and the recent immigration indicator (v) for these populations, which are very close to those observed in the Doctrine of Belen. As the table shows, consanguinity by random isonymy and surname diversity was nearly constant from the mid-18th century and early 19th in all the study locations. Such consanguinity values allow the identification of three groups. Susques, Toconao and Casabindo presented the highest consanguinity, followed by Cochinoca with intermediate values. The third group comprises the Doctrine of Belén, San Pedro de Atacama and Humahuaca, with significantly lower consanguinity values. As expected, surname diversity had the opposite behaviour and was highest among populations with low consanguinity, with a maximum value in San Pedro de Atacama (α=75) and minimum in Casabindo and Susques (α=23–24) (Table 3). The consanguinity levels described in this study are comparable to those found by Dipierri and colleagues (Reference Dipierri, Ocampo and Russo1991) in the Humahuaca Parish from 1734 to 1810. These authors used marital isonymy to estimate consanguinity, concluding that the native population of Humahuaca behaved as a single panmictic unit, where there are no deviations from random mating.
Table 3. Isonymy parameters in different Andean colonial populations
Differences in magnitude between the various estimates were due to the characteristics of each of the populations analysed and the methodology followed in each study. Casabindo was a closed population with no foreigners, governed and maintained in a restricted geographic area by the encomendero (Alfaro, Reference Alfaro2010), while the Humahuaca Parish population included natives, Spaniards and Mulattoes. For Humahuaca, the estimates were given by marital isonymy among effectively constituted couples, while the rest of the studies estimated consanguinity by random isonymy (Costa Junqueira et al., Reference Costa Junqueira, Martínez-Campos, Dipierri, Bejarano and Alfaro2000; Alfaro, Reference Alfaro2010; Peña Aguilera, Reference Peña Aguilera2016). Besides methodological differences, data sources also differ among the cited works. Including baptismal (Dipierri et al., Reference Dipierri, Ocampo and Russo1991) and death records (Costa Junqueira et al., Reference Costa Junqueira, Martínez-Campos, Dipierri, Bejarano and Alfaro2000), information was provided by ecclesiastical officials on one case and by population census, and tax rolls carried out by governmental representatives on the other (Alfaro, Reference Alfaro2010; Peña Aguilera, Reference Peña Aguilera2016).
The use of surnames to estimate consanguinity, isolation and migration intensity and direction enables the analysis of isonymic structure as a transversal element, indicative of other population structures (i.e. demographic and genetic) to shed light on the biodemographic, historical, socio-cultural and economic factors that shaped these structures from a multidisciplinary perspective. Despite evidence that population growth in the Doctrine of Belén was due to the immigration of ‘foreign’ individuals, isonymic analysis showed that it did not involve a change in surname patterns. Instead, the incoming population held similar surnames to the ones already part of the local pool. Therefore, it is reasonable to conclude that this population mobility, with people moving to the doctrine’s area from the (currently) Bolivian Altiplano, was an expression of traditional mobility as per Murra’s (Reference Murra and Murra1975) ‘vertical archipelago’ model. Colonial conditions might have given renewed impulse to this ancestral mobility pattern, driving people to leave their hometowns and settle in places with lighter tax burdens. Thus, this migrating population categorized as ‘foreign’ in colonial records were populations from a macro-zone connected by traditional mobility circuits. Another noteworthy aspect of this process was the relative homogeneity among surnames in the local and immigrant individuals.
In conclusion, consanguinity values by random isonymy (F ST) suggest that intra-population kinship was relatively constant through time (1750–1813) in the study population. For instance, both ν (recent immigration) and α (surname diversity) displayed high stability among the population. Moreover, certain aspects of this analysis, such as indicator B (the seven most frequent surnames) and population mobility (which was constant) indicated high sedentariness throughout the entire analysed period. Together, these features are manifestations of a highly ‘closed’ community, in a social sense, where migrations occurred according to family or kinship networks enhanced by compadrazgo and matrimonial unions that did not bring new surnames into the local population.
The present study population structure and isonymic analysis indicate that the Doctrina de Belén was a partially closed, sedentary population with high consanguinity, low diversity of surnames and particular migratory movements. This structure did not change substantially between 1750 and 1813. Indeed, the results are consistent with ethno-historical and biodemographic studies that characterized the Doctrine of Belén as a stable, rural indigenous population within the colonial framework, including limited, mostly temporary, and culturally homogenous migration dynamics. Such features were frequent among many other populations of the Andean historical and geographic setting.
Funding
This research received no specific grant from any funding agency, commercial entity or not-for-profit organization.
Conflicts of Interest
The authors have no conflicts of interest to declare.
Ethical Approval
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008.
