EDWARD DUTTON finds that societies tend towards social immobility
The Son Also Rises: Surnames and the History of Social Mobility
Gregory Clark, Princeton University Press, 2014, hb., 366pp
In his ground-breaking 2007 book A Farewell to Alms, Gregory Clark, Davis economist at the University of California, marshalled a large body of data on historical breeding patterns. He fearlessly presented his controversial conclusion: until the Industrial Revolution, Europe was undergoing eugenic fertility. If England was divided into the richer 50% and poorer 50% then the richer group had around double the number of surviving children than did the poorer group. As intelligence (and personality factors, such as Conscientiousness) is significantly positively associated with socio-economic status and is strongly genetic (80% for intelligence), Clark argued that the Industrial Revolution happened when it did because England reached a tipping point of high intelligence and was politically stable. Since the Industrial Revolution, this pattern of fertility has actually gone into reverse.
With The Son Also Rises, Clark is in the difficult situation of attempting to improve on a brilliant book. I’m not sure he succeeds in that regard, but The Son Also Rises is certainly highly original and a fascinating read. Clark attempts to trace social mobility over time in assorted countries, including England, the USA, Sweden, Japan, Korea, China, and India. He achieves this by focusing on surnames and, in particular, unusual ones. In general, Clark argues, social mobility is assumed to be relatively high because the correlation between, for example, a father’s education level and his son’s is relatively low. However, the problem with this is that ‘education’ – or, for that matter, occupational status, or income – are only partial measures of social status. People can make trade-offs between these measures. They might have a higher income than their father but a lower occupational status.
In addition, because the factors that underpin social status are strongly genetic, differences between father and son are subject to random fluctuation. There is a difference, for example, between genotypic intelligence (the genetic limits of a person’s intelligence) and phenotypic intelligence (a person’s actual intelligence, as partly caused by environment). Accordingly, Clark argues that measuring social mobility between two or even three generations and doing so on one measure is subject to error. A superior strategy is to assess the ‘underlying social competence’ of a particular family over many generations. This can be accomplished by tracing the history of the surname, because people tend to mate with those of similar social status to themselves.
Here, The Son Also Rises starts to become rather intriguing. In England, for example, Clark argues that you can tell the historical social status of a family by the kind of surname they have. High status surnames are those which are Norman (such as those that end in ‘ville’) or those which are ‘locative’ – the names of places. People with these surnames are descended either from Normans, who took the name of their feudal manor, or from wealthy Saxons who moved around a lot. Surnames which refer to a profession – Bailey, Cooper, Thatcher – are middle ranking, while low ranking surnames tend to end in ‘son,’ be the name of the father, or refer to the part of the village a person lived in, such as ‘Hill.’ Assessing the data between the Medieval period and 2012, Clark finds that across the period, and even now, those with Norman or locative names are over-represented among proxies for high social status. These include Oxbridge graduates, barristers, and physicians. Those with low status surnames are under-represented.
Over this period, Clark emphasises, there has been a gradual ‘regression to mean.’ This means that those at the representation extremes have moved closer to the average over this period, so that Normans are less over-represented than they used to be and Johnsons are less under-represented than they used to be. Clark provides a number of explanations for this. Something akin to regression to mean happens in genetics. Because of the large number of genes involved, it sometimes (though rarely) happens that children are significantly more (or less) ‘socially competent’ than their parents. They then marry a person with comparable abilities and move up (or down) the social hierarchy. Alternatively, a person with high genotypic ‘ability’ marries a person with high phenotypic ‘ability.’ Finally, as pre-modern fertility was predicted by high social status, Norman surnames would have to spread downwards through the population.
Overall, Clark finds that social status across history is around 0.75 heritable, its heritability is the same now as it was in medieval England, and sudden shocks that might raise social status very quickly – such as winning the Lottery – wear off within a few generations. Interestingly, he finds the same results in countries that are far more egalitarian than Britain, such as Sweden. Dividing between the surnames of the higher nobility, the surnames of the untitled nobility, Latin surnames (historically adopted by the highly educated non-noble), geographical surnames (e.g. Berg) and the lowest status surnames (those ending in ‘son’), Clark finds a similar pattern to England. Those with noble and Latin surnames are still over-represented at the top universities, amongst the richest, and in the top professions, and the heritability of social status is about 0.75. Looking at multiracial and multi-religious societies, such as the USA, Clark concludes that racial and religious differences in socio-economic status are as he would predict, though he is careful to only imply his views on the reasons for this rather than overtly state them. Religious and racial boundaries appear to be fluid at the edges, meaning, for example, that lowest status English have tended to drop down into the ‘traveller’ category.
Clark writes in a very personal style, referring to his middle-class Glasgow childhood and giving us many illustrations, including of his childhood street. For this reason, I think that relatively complex concepts – such as regression to the mean, genotype, phenotype and so on – should have been explained in considerably more depth. Even the nature of a correlation coefficient appears to cause confusion amongst graduates not trained in statistics. As such, if, and this appears to be the case, Clark is evangelical about his findings, it might help to illustrate them with metaphors, which, considering his habitual use of Hemingway puns, should be no problem for him. In the same way, Clark does not examine in any depth the factors that underpin differences in social status. He refers to them as ‘social competence,’ but what he is really talking about is intelligence and personality. It is true that he looked at intelligence in his previous book, but in a book aimed at a non-specialist audience (many of them possibly sceptical of the whole concept of genetic abilities) it would help to look at these issues in some depth. Indeed, at one point Clark argues, in a footnote, that IQ does not strongly predict indices of social status. This is odd, because he has already emphasised that individual measures introduce error and we should thus focus on many measures.
Clark argues, in his conclusion, that, in that social status is strongly heritable, a meritocracy will be unequal. As such, society should not reward social competence and should instead attempt to create social equality. He does not examine the potential problems with this thesis, such as dysgenics, which are examined, to some extent, in his previous book. Finally, there are many nuances to his general thesis that he could have examined. For example, the English nobility has historically been more stratified than that of other European countries. English people have, until relatively recently, had a strong sense of social class identity that has been less obvious in countries like Sweden. Could this have influenced marriage patterns in any way between the two countries and thus social mobility? Clark’s thesis is very persuasive but it would be improved by acknowledging and refuting all potential difficulties with it. There is certainly space to do this because Clark provides us with various not entirely necessary appendices, such as a guide to discovering the ‘status of your surname lineage.’
Overall, this is a well written and thought-provoking book. Clark had an extremely hard act to follow and even if this is not quite a Farewell to Alms it is certainly worth the read. I look forward to his next book – and his next Hemingway pun!
Dr EDWARD DUTTON‘s book Religion and Intelligence: An Evolutionary Analysis is published by the Ulster Institute for Social Research (2014)