The Son Also Rises Page 24

  FIGURE 13.1. Distribution of Protestants in Ireland, by county, 1911.

  To understand what drove these exchanges, it is useful to divide Ireland into the six counties with the greatest share of Protestants—Armagh, Antrim, Down, Fermanagh, Londonderry, and Tyrone, which together became Northern Ireland—and the rest. Figure 13.1 shows the Protestant share of the population of Ireland by county. Figure 13.2 shows the share of the population of each surname type by religious affiliation and age group (ages 0–29 and 30 and over) in each region in 1911.

  Figure 13.2 reveals that surnames were changing to those of the predominant religion of each region. In the south, the overwhelming majority of the Irish surnames were still Catholic, whereas in the six northern counties, only two-thirds of Irish surnames remained Catholic. In counterbalance, in the northern counties, 93 percent of Scottish surnames remained non-Catholic, while in the south one-third of the Scottish surnames were held by Catholics.

  Figure 13.2 also suggests that this process had been under way for generations. When we divide people in the census into two age groups, those age 30 and older, and those younger, we see that the transition of Irish surnames toward Protestantism, and of Scottish surnames toward Catholicism, is only modestly greater for the younger cohort.

  The two groups, Catholics and Protestants, were socially differentiated. As figure 13.3 illustrates, throughout the country, Catholics were less literate than Protestants. And as figure 13.4 shows, Protestants were more likely to be found in skilled occupations all over Ireland. However, as these figures also show, the transitions from one religion to another helped perpetuate differences in social status between the two religions. Catholics with Scottish surnames had much lower social status than Protestants with the same surnames, even controlling for regional differences: they were more likely to be illiterate and to hold less-skilled occupations. Similarly, Protestants with Irish surnames had much higher status than Catholics with the same surnames: they were more likely to be literate and to hold more skilled occupations.

  FIGURE 13.2. Percentages of Irish and Scottish surnames held by Catholics, 1911.

  FIGURE 13.3. Percentages of Irish and Scottish surnames held by illiterate males age 16 and over, 1911.

  FIGURE 13.4. Ratio of skilled to unskilled males age 18 and over, 1911, by surname type and religion.

  Thus Irish history shows how even communities as mutually antagonistic as Irish Catholics and Protestants can undergo not only a significant movement of people from one group to the other but also a movement that solidifies the positions of the respective elites and underclasses within of the two groups. Those at the bottom of the social scale in the Protestant surname group were much more likely to have transferred their religious affiliation to Catholicism sometime in the previous three hundred years. Those at the top of the social scale in the Irish surname group were much more likely to identify as Protestant by 1911.8

  The Origins of the Modern Jewish Population

  Suppose Jewish populations interacted with the surrounding populations in Europe, North Africa, and the Middle East in the same way as the Protestant minority in Ireland did with the majority Catholic population. Then Jewish economic advantage later might be explained by selective flows between the local and Jewish populations.

  There are, however, some puzzles about the origin of the modern Ashkenazi Jewish population, descended from communities concentrated in Slavic Eastern Europe in the early nineteenth century. In 1170 the Sephardic and Mizrahi Jews of the Middle East, North Africa, and southern Europe represented the great majority of the known Jewish population. By the early twentieth century, the Sephardic and Mizrahi population had grown little and was still only about one million. There is clear information on their origins and lineage: Sephardic and Mizrahi Jews are plausibly all the descendants of the Jewish population of 1490 CE, with some local admixtures and losses from conversions to other religions (as in Spain after 1492). In contrast, the Ashkenazi population, which was estimated in medieval times to account for 4 percent or less of all Jews, had come to constitute the overwhelming majority of Jews by the early twentieth century, numbering as many as eight million.9 Despite the Holocaust, it is estimated that the Ashkenazi now constitute more than 80 percent of the world Jewish population.

  How the Ashkenazi emerged by the nineteenth century as the bulk of the world Jewish population is an intriguing historical mystery. It is ironic that a community renowned for its early embrace of literacy has no written records to show its own origin or migration to Eastern Europe. As one scholar of the origins of Yiddish, Robert King, notes: “The legacy of early Jewish life in the Slavic East was very largely the bones of its dead.”10

  It used to be accepted without controversy that the Ashkenazi were an offshoot of the Sephardic community that migrated from Italy to western Germany in the Middle Ages. Later, this group was supposedly driven east under the pressure of persecution in Germany following the onset of the Black Death in 1347. This is the so-called Rhineland hypothesis. However, it would require extraordinary rates of population growth in Eastern Europe between the Middle Ages and the nineteenth century to produce the eight million Ashkenazim observed in 1900: consistently more than 1.5 percent per year, or 50 percent per generation—much faster than the general growth of population in Eastern Europe. This rate is not implausible: in societies such as preindustrial England, elite groups did have much faster population growth rates than the general population. In England, for example, the richest lineages had increases in family sizes of 50–100 percent per generation.11 A common origin of Ashkenazi and Sephardic Jews from a subgroup of no more than 10 percent of a larger parent population in the Roman Empire would fit with the hypothesis on the biological transmission of status advanced here.

  Some scholars have questioned, however, whether the Ashkenazi population growth rate required by the Rhineland hypothesis is consistent with other evidence on the demography of preindustrial Jewish communities.12 An alternative proposal is that the Ashkenazim were descended from a mass conversion of the Khazars, originally from the Caucasus, to Judaism in the ninth century. After the collapse of the Khazarian Empire in the 960s CE, the remnants of the Khazars supposedly migrated to Eastern Europe, bringing Judaism with them. This would contradict the hypothesis offered here on the source and nature of elite populations. A whole population converted to Judaism should not become elite, since religious precepts in themselves are irrelevant to social competence except insofar as they lead to selective recruitment to a religious affiliation.

  However, the genetic evidence seems to support the Rhineland hypothesis on the origins of the Ashkenazi.13 A recent survey of the genetic evidence on Jewish populations, including evidence from the whole genome, the Y chromosome, and mitochondrial DNA, concludes that Ashkenazi are indeed closely related to other Jewish populations. Further, they are most closely related genetically to the Italian, Greek, and Turkish Jewish populations, a finding that fits with the conventional story of the migration of the Ashkenazi to northern Europe.

  The genetic evidence also suggests that the Ashkenazi population had a very small number of both male and female founders and so must have undergone a rapid expansion to reach its present size. Four women, for example, account for 40 percent of the mitochondrial DNA of the Ashkenazi, with a modest number of others contributing the rest. The evidence from the Y chromosome also suggests that only 5 to 8 percent of the Ashkenazi genetic material comes from admixture of European males.14 Thus the Ashkenazi may represent an elite within an elite, a finding that would be entirely consistent with the hypothesis on social mobility advanced here. And the limited admixture of European DNA from the Ashkenazis’ long sojourn in Eastern Europe is consistent with the proposition that a group retains its elite status over the long run either by practicing endogamy or by selectively losing its lower-status members.15

  Minorities in Muslim Societies

  Evidence for the role of selective affinity with minority populations can also
potentially explain the emergence of Christians, Jews, and Parsis as elites in predominantly Muslim societies. This mechanism is laid out for Egypt in an interesting study of Coptic Christians by Mohamed Saleh.16 Muslim societies had two characteristics. First, subject populations were not forced to convert to Islam: Muslim societies were, at their inception, tolerant of religious minorities.17 But under Islamic law, non-Muslim males were subject to a head tax called the jizya(h). This was a fee for permission to practice another religion, designed as an inducement to convert.

  The head tax was sometimes levied at variable rates. Thus Abu Yusuf, the chief justice of Baghdad in the eighth century, in his treatise on taxation and public finance (Kitab al-kharaj), stipulated that the jizya should be forty-eight dirhams for the richest men, twenty-four for those of moderate wealth, and twelve for craftsmen and laborers. But such rates would still have made the tax much more burdensome on the poorest laborers than on wealthier members of religious minorities.18

  Saleh shows that in Egypt, Coptic Christians, who formed the vast majority of Egyptian society on the eve of the Arab Muslim conquest, selectively converted to Islam in the centuries following the Arab conquest of 641 CE. He finds evidence that under the pressure of the jizya, the poorest Copts converted to Islam: the conversion rate was greater in areas where heavier taxes were imposed. Moreover, in areas where the conversion rate was highest, the remaining Coptic population was more elite by the nineteenth century. In the Byzantine Empire, Copts had the lowest status in society, below the Jews and upper-class Greek Orthodox Christians; in Muslim Egypt, the remaining Copts became, like these two other minorities, an elite. In the nineteenth century, in both urban and rural areas, Copts had higher occupational status than Muslims despite being a political minority.

  The situation in Egypt is echoed in other Muslim societies. In Iran, for example, an analysis of the 1966 census found that the high-income capital, Tehran, with about 10 percent of the country’s population, contained two-thirds of Armenian Christians and Jews, and half of Zoroastrians and Assyrian Christians. The explanation was that after modernization began in 1921, many of the early physicians, engineers, mechanics and teachers of foreign languages with Western training came from minority groups. Tehran was in the vanguard of modernization and presented a high demand for such professionals. Thus minorities were attracted to the city.19

  However, minorities in Iran even in 1966 constituted only 1.2 percent of the population, large numbers of Jews, Zoroastrians, and Christians having previously emigrated because of Shia Islam’s intolerance toward minorities. In Lebanon, Syria, Jordan, and Iraq, Christian minorities all constituted elites after the Muslim conquests, presumably because of a similar pattern of conversion to Islam under the jizya system.

  Once created, minorities in Islamic societies seem to have maintained their high status over more than a millennium through high rates of marital endogamy. A study of the ABO blood groups of Iran concludes that the Jewish, Armenian, Assyrian, and Zoroastrian minorities were genetically isolated from the rest of the Iranian population for long periods.20 These groups account for such a small share of the population, however, that although it can be concluded that they gained few members from Muslim population groups, the possibility cannot be ruled out that they lost members to assimilation with Muslim groups.

  All these experiences of the creation of elite subgroups and the persistence of elites are consistent with the simple model of social mobility outlined in chapter 6. Elites and underclasses are formed by the selective affiliation to a religious identity of some upper or lower share of the distribution of abilities within the population. In Islamic societies, the practice of imposing taxes on religious minorities tended to recruit to Islam the lowest socioeconomic strata of conquered societies.21 Elites and underclasses have maintained themselves over periods as long as 1,300 years because of very high rates of endogamy, which preserves the initial advantage of elites from regression to the mean by preventing intermarriage with less advantaged populations.

  Gypsies and Travellers in England

  Do such explanations also hold for the major English underclass for the last four hundred years, the Gypsy or Traveller community? This community has long been at the bottom of the socioeconomic ladder in England. The U.K. Equality and Human Rights Commission notes that they are “one of the most deprived groups in Britain. Life expectancy for Gypsy and Traveller men and women is 10 years lower than the national average. Gypsy and Traveller mothers are 20 times more likely than the rest of the population to have experienced the death of a child. In 2003, less than a quarter of Gypsy and Traveller children obtained five GCSEs and A*–C grades, compared to more than half of students nationwide.”22

  Table 13.2 reports the result of a survey of nearly three hundred adult British and Irish Gypsies/Travellers in 2006. About half the people surveyed no longer traveled, and about a quarter traveled only in the summer. Even so, only two-thirds of the Travellers had ever attended school, and their average age on leaving school was thirteen. Nearly three-fifths of them smoked. Half reported a chronic cough, and more than a quarter reported suffering from anxiety or depression. Women on average had given birth to 4.3 children, and some were still of childbearing age.

  A comparison group was composed, somewhat mysteriously, of poor English whites, Pakistanis, and blacks of Caribbean origin. The outcomes for the comparison group were, systematically, significantly better than for the Traveller population. The comparison group had more schooling and better health. They also, notably, had much lower fertility.

  The mythology of the Gypsy/Traveller community is that they are the descendants of the Roma (Romany), with origins in India. However, there is plenty of evidence to suggest that the community is almost entirely of native British origin. Consider, for example, figure 13.5, which shows two Traveller children at the 2011 eviction of Travellers from an illegally occupied site at Dale Farm, Essex. These do not look like people of Indian origin.

  Evidence from the surnames of Gypsies and Travellers in England suggests that they are indeed of domestic origin. Other immigrant groups in England, such as the Jewish population that emigrated to England in the seventeenth century and later, exhibit an unusual distribution of surnames by frequency. Figure 13.6 compares surname frequency in the general population to that among people with characteristically Jewish first names, such as Solomon and Golda, who were born in England between 1910 and 1914. The most common surnames in the native population in 1881 were very infrequent among Jews. Half the Jewish population had surnames held by fewer than two hundred people in 1881.

  TABLE 13.2. Characteristics of Travellers and comparison disadvantaged group, England

  Status

  Travellers

  Comparison group

  Average age

  38.1

  38.4

  Ever attended school (%)

  66

  88

  Average age on leaving school

  12.6

  16.4

  Current smoker (%)

  58

  22

  Average number of children born (for women)

  4.3

  1.8

  Reports anxiety/depression (%)

  28

  16

  Chronic cough (%)

  49

  17

  Gypsy or Traveller families tend not to have similarly distinctive surnames (figure 13.7). These families can be identified in the 1891 census as those whose housing is described in such terms as “in canvas tent,” “in caravan,” and “on the common” and their occupations as “travelling hawker” or “showman.” Apart from one peculiarity, surname frequency in this group echoes that of the general population. That peculiarity is the unusually high frequency of the surname Smith. This is the most common surname in England: in 1891 it appeared at a frequency of 1.4 percent among the general population. Among Travellers, its frequency was 7.7 percent. Travellers do not seem to have had unusual surnames of non-Englis
h origin: their surnames are a representative sample of rare, intermediate, and common English surnames.

  FIGURE 13.5. Young Travellers look on as bailiffs evict Travellers from a settlement without planning permission at Dale Farm, Essex, 2011.

  FIGURE 13.6. Surname frequencies among general population, 1881, and Jewish population 1910–14, England.

  These findings suggest that the Gypsy and Traveller population of England is not descended from some exotic band of imported Roma, an underclass many generations old, but is almost entirely indigenous and more recent in origin.23 It is likely that among the indigenous English population, by random chance, some families ended up at the margins of society as traveling harvest workers, hawkers, basket makers, and showmen. But this marginal group, perhaps even drawing inspiration from the few genuine Roma they encountered, adopted a romanticized version of the Gypsy lifestyle and a creation myth of their own.24

  FIGURE 13.7. Surname frequencies among general population, 1881, and Gypsy and Traveller population 1891, England.

  By the nineteenth century, the first names of Gypsy and Traveller children were sometimes colorful. For boys, favored names allegedly included Goliath, Belcher, Dangerfield, Gilderoy, Nelson, Neptune, and Vandlo. Favored girls’ names included Britannia, Cinderella, Dotia, Gentilia, Fairnette, Freedom, Mizelli, Ocean, Reservoir, Sinfai, and Vancy.25 Are these timeless Gypsy names, passed down by Roma forefathers? Not likely. If we look at the extensive records of baptisms in England 1538–1837, we find that almost all these supposed Gypsy and Traveller surnames first appear only in the late eighteenth and nineteenth centuries. Thus the first recorded Cinderella baptism was in 1798, the first Goliath 1817, the first Ocean 1797, the first Freedom 1803, and the first Gilderoy 1785.26 (The late appearance of Cinderella is not surprising, since the Cinderella story, based on a French tale, was first published in English in 1729.)