Extending the record

Before attempting to generalize, or even begin to formulate a research agenda on the possible causes of small stature in the pre-industrial era, it is essential to inspect additional evidence.

study of pre-industrial heights from military records is rather short (or non-existent) in all countries.⁵ Skeletal data, in which stature is inferred from long bone (femur) lengths, provide information on the more distant past. Before examining the results, however, it is worth discussing the methodology of using skeletons to estimate heights.

Valid use of skeletons requires knowledge of sex and age at death.⁶ For all parts of the skeleton, the female elements typically have smaller size and lighter construction. These differences become pronounced in the skull and the pelvis beginning in late adolescence, but before these ages, sex cannot be accurately determined. The skeletal features of robust adult females may resemble those of small and light adult males, but using multivariate analysis, it is possible to estimate sex of adults with a high degree of accuracy in well-preserved remains.

Over a person's lifetime, the skeleton undergoes sequential chronological change. The changes are pronounced in dentition among children and in fusion of the epiphyses (or ‘growth plates') in various bones of children and in young adults. Among older adults, there are systematic changes in the shape of the pubic symphysis that are used as a guide to age. Above age 50, however, these techniques (and others) become increasingly unreliable, and for this reason, older adults are often grouped together for analytical purposes. In estimating the height of adults, it is the fusion of epiphyses in the long bones that is the key for estimating adult height. Once fusion has occurred, growth ceases and length does not change even with increasing age.

Excavated burials may not reflect the once-living population if burials were geographically dispersed, excavation was incomplete or quality of preservation of the bones was poor. Therefore, in this line of work it is useful to compare results from a particular site with those from sites where these problems are thought to be minimal.

Table 9.2 gives details from individual studies found in a search of the literature in physical anthropology from sources easily accessible in the United States. Although some results were available for other parts of Europe, only northern Europe had studies that span the entire period from the early Middle Ages to the present. My first effort leaves a gap for the fifteenth and sixteenth centuries, which additional search may fill.

Table 9.3 summarizes results by era for those studies that give evidence for time periods as small as two or three centuries. Studies reporting results for burials during ‘Medieval Era' or the ‘Middle Ages' are lumped together in the middle of the table. It is remarkable if not stunning that the average heights during the early and late Middle Ages exceeded those observed for the eve of industrialization by several centimetres. While one should always devote some attention to the issues of representativeness and sampling in skeletal data, the large number of studies covering several northern countries suggests that the results cannot be dismissed as a statistical fluke or aberration. It is conceivable that all of the estimated heights for

Table 9.2 Average heights in northern Europe estimated from adult male skeletons

‘ Simple average across seven combinations of sites and dates.

^b Sex of the 102 skeletons was unknown but based on written evidence, a sex ratio of 50—50 was assumed. Since men are on average 10 cm taller than women, the overall average height for the sample was adjusted upward by half this amount (5 cm) to estimate equivalent average height for men.

* Not available or missing information.

the Middle Ages were biased upward by some as yet undiscovered process of selection, but one would then wonder why that selection process ceased to be a factor in the centuries immediately prior to industrialization.

Generalizing about pre-industrial height trends will be difficult without more evidence from the fifteenth through the eighteenth centuries. It seems reasonable to suggest, at least tentatively, that net nutritional conditions of the past millennium reached a low point in Europe prior to the onset of industrialization. Between the Middle Ages and the twentieth century, heights were U-shaped with a minimum

Table 9.3 Summary of adult male height trends in northern Europe

attained sometime between 1450 and 1750, when historical heights become widely available from military records.

Taking the evidence at face value indicates that average heights fell from an average of 173.4 cm in the early Middle Ages to a low of 165.8 cm during the seventeenth and eighteenth centuries. This decline of 7.6 cm exceeds by a factor of two any fluctuations observed during industrialization. Recovery to levels achieved a millennium ago was not attained until the early twentieth century. Both the extraordinary level relative to recent times and the U-shaped time trend are remarkable phenomena worthy of considerable study.

Some people may claim that genetic factors are responsible for the tall statures observed during the Middle Ages, pointing to the fact that northern Europeans are taller, even today, than those from more southern European countries (Schmidt, Jorgensen, and Michaelsen 1995). But the southern Europeans of the modern period, who tend to be poorer, are catching up, and in any event, studies of children around the globe indicate that children who grow up under similarly good environmental conditions have about the same heights (Malcolm 1974; Martorell and Habicht 1986). If genetic factors were relevant, presumably they had little or no effect on the trend within areas surrounded by the North Sea and the Baltic. Thus, I seek environmental explanations for northern Europe's U-shaped trend in stature.

In suggesting candidates for further study, it is relevant to recall that average height measures a population's history of net nutrition—diet minus claims on the diet, made by work and by disease. Urbanization and growing population density, which occurred during industrialization, increased exposure to disease. Could the diet have been poorer and work more arduous in pre-industrial times, by enough to offset the benefits of lower population density? And if the diet was poorer and the work was more arduous, why was net nutrition so good before the sixteenth century?

The data at hand confront conventional wisdom about changes in living standards since the Middle Ages, and lead one to ask: Why did net nutrition decline sometime between the Middle Ages and the pre-industrial period? Why did heights generally improve during the nineteenth century, albeit with interruptions in some countries, when some factors adverse to heights (urbanization, inequality, and business cycles) were getting worse?

As I do not have convincing answers to these questions, I look forward to additional research.

4.2