Franz Boas and Native American biological variability

Human Biology, Jun 1995 by Jantz, R L

Der Gesichtspunkt, welcher mich bei der Sammlung des Materials leitete, war wesentlich auf die Thatsache begrundet, dass die durch Messungen zum Ausdruck gelangenden Unterschiede zwischen Menschen-Varietaten so gering sind, dass nur bei grossem Beobachtungsmaterial sichere Schlusse gezogen werden konnen. [The point of view which has guided my collection of this material was based essentially on the fact that differences between human varieties expressed through measurements are so small that reliable conclusions can only be drawn with large numbers of observations.]

In contrast to Hrdlicka, who despised statistics (Howells 1992), Boas understood them and their necessity in understanding small differences in variable populations. Despite the computational burden, Boas recognized that the statistical approach was essential.

Another factor that must have motivated Boas was the rapid pace of change and assimilation that Amerindian populations were undergoing by the late nineteenth century. Boas realized that the data he collected would soon not be possible to obtain. Today it is even more obvious how farsighted Boas was in his efforts to preserve information. Whether it was Boas's intention to collect data to be used by future generations is not clear. If so, he took few steps to ensure their availability. If not, then he overestimated his ability to deal with the large quantities of data resulting from these projects.

The design of large-scale anthropometric projects such as these required a number of methodological considerations, such as measurement selection, sample selection and documentation, interobserver variation, and training of observers. It is essential to understand Boas's philosophy on these matters and the choices he made if his data are to be properly understood.

Measurement Selection

Boas's standard data sheet contains 12 measurements: (1) height standing, (2) height of shoulder, (3) height of point of second finger (this actually refers to digit 3), (4) finger reach (span), (5) height sitting, (6) width of shoulders, (7) length of head, (8) breadth of head, (9) height of face, (10) breadth of face, (11) height of nose, and (12) breadth of nose. An example of this standard data sheet is given by Jantz et al. (1992). This data sheet was used in the World's Columbian Exposition project, the most ambitious of Boas's anthropometric surveys, but its basic design was established earlier (Boas 1892).

The measurements are evenly divided between head and face dimensions and body dimensions. The measurements are described by Boas (1892) and by Sullivan (1920). Boas intended shoulder height and finger height to be used to calculate arm length (Boas 1892; Sullivan 1920). The body measurements are biased toward longitudinal dimensions; shoulder width is the only transverse dimension included. Consequently, there is considerable redundancy among the measurements.

The head and face dimensions would today be considered standard dimensions. We have been unsuccessful in locating a reference to a measurement authority on which Boas relied in setting up his measurements. Because Boas's early training in physical anthropology took place under Rudolph Virchow, the measurement definitions are presumably those of the developing German school. Boas was present in Berlin in 1885 when nine visiting Bella Coolas were measured by Virchow (Cole 1985), an event that must have impressed upon him the importance of anthropometry.