Hostname: page-component-5c6d5d7d68-wpx84 Total loading time: 0 Render date: 2024-08-08T12:18:24.895Z Has data issue: false hasContentIssue false
  • English
  • Français

Physical Anthropology and the Reconstruction of Recent Precolonial History in Africa, II: A Dermatoglyphic Survey From Kenya

Published online by Cambridge University Press:  13 May 2014

Peter Rosa
Peter Rosa*
Affiliation:
University of Stirling
Get access Rights & Permissions [Opens in a new window]

Extract

Earlier I discussed the potential of physical anthropology for the reconstruction of precolonial history in Africa. As a traditional branch of the biological sciences, physical anthropology draws on advances in genetics, population genetics, numerical taxonomy, and other biological disciplines whose paradigms define a logical framework by which the biological history of human populations can be explored. Sophisticated techniques can now be applied not only to the study of biological history, but also to the investigation of hypotheses of wider historical interest (i.e., those arising from the consideration of nonbiological sources of historical evidence). Having outlined some of the possibilities of new techniques and contrasted them to more basic techniques available to pre-war physical anthropologists, who tended to be preoccupied by the construction of racial taxonomies, I ended the paper by discussing ways by which the wider historical potential of physical anthropology could begin to be realized.

In this sequel the onus is on trying to demonstrate in a limited way some of these possibilities through an empirical study. The study in question is a biological survey of Kenyan peoples which I undertook in the early and mid-1970s. During this survey I obtained palm and finger prints from over 6,000 primary and secondary school children drawn from some 60 population units (tribes and sub-tribes). The finger and palm prints provided over 150 biological measures, which are genetically determined, to compare the population units sampled and to analyze the biological patterns of differentiation of Kenyan peoples.

Type
Research Article
Information
History in Africa , Volume 14 , 1987 , pp. 229 - 256
Copyright
Copyright © African Studies Association 1987

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

Notes

* I am indebted to Robin Law for his advice and patience during the writing of this paper.

1. Rosa, Peter, “Physical Anthropology and the Reconstruction of Recent Precolonial History in Africa,” HA, 12 (1985), 281–305Google Scholar

2. See Rosa, Peter, “An Investigation of Dermatoglyphic Variation among Ethnic Populations in Kenya,” (PhD, University of Durham, 1981)Google Scholar, Table 1.1, for a summarized account of previous biological surveys in Kenya.

3. Motulsky, A.G., “Metabolic Polymorphisms and the Role of Infectious Diseases in Human Evolution,” Human Biology, 32 (1960), 28–62Google ScholarPubMed

4. For example see Hiernaux, Jean, “The Analysis of Multivariate Biological Distances Between Human Populations: Principles and Applications to Sub-Saharan Africa” in Huizinga, J., eds., The Assessment of Population Affinities in Man(Oxford, 1972), 106–07.Google Scholar

5. The classification of dermatoglyphic features is based on features called triradii, formed by the meeting of ridges from three directions in a shape resembling a three-pointed star. (A classic example exists underneath the base of the index finger). Triradii are usually associated with loops and whorls. Finally, ridges can be counted between standard reference points (triradii and loop cores) both on fingers and palms. Ridge counts are especially useful, being continuous measures which display a high component of heredity. The best short account of dermatoglyphic features is Penrose, L.S., “Memorandum on Dermatoglyphic Nomenclature,” Birth Defeats Original Article Series, 4 (1968), 1–12.Google Scholar For a detailed listing of the specific traits used in the analyses surrounding this study, see Rosa, P.J., “Descriptive Report on a Dermatoglyphic Survey of 6,235 Schoolchildren from Kenya,” University of Durham, Department of Anthropology Occasional Paper (1983), 32.Google Scholar A full account of methods is given in Rosa, “Investigation,” ch. 4.

6. See ibid., ch 2, for a full description.

7. Ibid., Table 5.1.

8. Spear, Thomas T., Kenya's Past, (London, 1981), 16–17Google Scholar

9. Hiernaux, , “Biological Distances,” 105Google Scholar

10. Ibid.

11. There are a number of statistics which calculate genetic or phenetic distances between populations. The general principle of each statistic is to standardize the variables so that each measure carries the same weight, and to average the sum of the squared differences over all biological variables being considered. The final figure is the distance between the two populations in question. A series of distances between all the populations being examined is the matrix of distances. The methodology surrounding the calculation of phenetic distances between the Kenya population samples is described in Rosa, “Investigation,” ch. 6. The most thorough account of biological distance in general is Constandse-Westermann, T.S., Coefficients of Biological Distance, (New York, 1972).Google Scholar

12. Rosa, “Investigation,” ch. 6.

13. Rosa, , “Physical Anthropology,” 292.Google Scholar

14. Ibid., 278ff.

15. Hiernaux, , “Biological Distances,” 106.Google Scholar

16. Principal components analysis (PCA) is a multivariate statistical technique by which a series of inter-correlated variables can be reduced to a smaller number of uncorrelated components. After the first component has been extracted, subsequent components are derived from the variance that remains after each previous component has been extracted. Hence the first component is always the most important, as it accounts for the greatest share of the variance.

Each PCA solution produces a series of weights by which the scores of the original variables can be multiplied by and combined to produce new variables that reproduce the characteristics of each component. Once these new variables have been calculated, they can be used to summarize all the significant information contained in the original variables that contribute significantly to the component in question.

In the present analysis digital dermatoglyphic variables are treated separately from palmar ones, as digital variables do not correlate significantly with palmar ones. A PCA for each sex was thus performed on 148 digital dermatoglyphic variables, reducing the variance to 21 significant components (a figure consistent by sex). A separate PCA was performed on 36 palmar traits, reducing the variation to 11 components (again consistent by sex). It is these component variables that are used in the main analyses discussed in this section of the paper.

Linguistic affiliation is usually represented statistically by a nominal variable, where each linguistic group or division is given a number to identify it. For example Mijikenda might be “10,” Kalenjin “11,” and so on. Nominal variables, however, cannot be used in correlation analyses as they stand, as correlation analysis requires each variable to be at least at an interval level or measurement. To overcome this a nominal variable can be broken down into a series of dummy variables, where each category is coded “1” if true, or zero if false. Each linguistic group can thus be represented by a dummy variable where the code is “1” if a case belongs to that group, or coded “0” if it does not.

17. I.e., PCA for (1) digital traits, males; (2) digital traits, females; (3) palmar traits, males; (4) palmar traits, females.

18. For a detailed account see Rosa, Peter, “Associations Between Dermatoglyphic Variation, Topography, and Climate in Kenya”, American Journal of Physical Anthropology, 68(1985), 395–408.CrossRefGoogle ScholarPubMed

19. Spear, , Kenya's Past, 101.Google Scholar

20. Morton, R.F., “The Shungwaya Myth of Mijikenda Origins; A Problem of the Late Nineteenth-Century Kenya Coastal History.”, IJAHS, 5(1972), 397–423.Google Scholar

21. Spear, T.T., “Traditional Myths and Historians' Myths: Variations of the Singwaya theme of Mijikenda origins.”, HA, 1(1974), 67–85Google Scholar; idem., “Traditional Myths and Linguistic Analysis: Singwaya Revisited.” HA, 4(1977), 229-45.

22. Spear, , “Historians' Myths,” 75, 76.Google Scholar

23. Ibid., 76.

24. Ibid., 72.

25. Spear, , “Historians' Myths,” 71.Google Scholar

26. Spear, “Linguistic Analysis”

27. Ibid., 240.

28. A matrix of phenetic distances was subjected to principal coordinates analysis. Population scores were then obtained for each of the four most important vectors (i.e., those which accounted for the most variation). These scores were then plotted as a series of linked X,Y coordinates on the graph. Each population can thus be graphically represented as having a character line, which differs from that of another population not only in terms of distance, but also in “shape.” Similarity in shape and distance implies a very recent common origin. Similarity in shape but not distance implies a common origin at a less recent time. An absence of similarity in shape implies no detectable common origin. This technique is described in more detail in Rosa, “Investigation,” ch. 8.