A multivariate approach to fingerprint variation in Papua New Guinea: Implications for prehistory

Abstract

Fingerprint data from three Markham Valley populations of Papua New Guinea are presented. Initial comparisons with data from elsewhere in New Guinea by conventional bivariate methods suggest hypotheses regarding prehistory, but these methods are inconclusive. Because of this shortcoming, the application of discriminant analysis to dermatoglyphic data is justified and cautiously demonstrated. Discriminant analysis is a valid approach to the parsimonious study of fingerprint relationships among human populations. Significant discriminating variables are then used with the limited published data to display historical relationships. These marker traits include the separate frequencies of thumb, middle fingers, and little finger patterning, and the complexity of the thenar eminence. With regard to prehistory, it is concluded that fingerprint data support the hypothesis of separate origins and migrations for Papuan‐ and Melanesian‐speaking peoples. Furthermore, it is suggested that these data support the linguistic hypothesis of a second Papuan migration of Trans New Guinea Phylum speakers, up the Markham Valley into the New Guinea Highlands. Finally, gene flow is indicated from Melanesian speakers of the north coast into the interior through the Sepik and Ramu River valleys, affecting both the Trans New Guinea Phylum and other unrelated Papuan groups. Moreover, there are numerous suggestions of recent reciprocal genetic exchange between neighboring Melanesians and Papuans, without obliterating the older, separate origins of these peoples. These results suggest that fingerprints have greater phylogenetic stability than other biological parameters used in the past.

A multivariate approach to fingerprint variation in Papua New Guinea: Perspectives on the evolutionary stability of dermatoglyphic markers

“…In the male analysis, the discriminations along the two axes are very close to those found in the previous three-language-group dermatoglyphic analysis (Froehlich and Giles, 1980). The discriminants are based on the scaled vectors and discriminant loadings shown in Table 7.…”

“…Thus, a total of nine villages with nearly complete sampling were used to determine whether the population structure among villages is congruent with their linguistic affiliations. The high degree of interrelatedness in these samples has been justified and discussed elsewhere (Giles et al, 1966a;Froehlich and Giles, 1980). The Austronesian-speaking Atsera live along the Markham River, about 60 miles inland from Lae (Fig.…”

Interethnic comparisons of three endogamous groups of West Bengal with respect to two sets of variables

“…al., 1974;Friedlaender, 1975;Rudan, 1978;Rothhammer et al, 1979;Jantz and Chopra, 1983;Karmakar, 1984;Reddy et. On the other hand some earlier studies suggested that dermatoglyphic traits are more advantageous polygenic marker for tracing the affinities between group.s (Rife, 1954;Froehlich and Giles, 1981;Micle and Kobyliansky, 1986) and provides greater agreement with the ethnic histories as well as geographic background (Reddy and Reddy, 1992;Sanna and Floris, 1995;Floris et al, 1998) of different groups than serological or anthropometric traits. Though quite a few others did find close congruence between delanatoglyphics and other biological data (Rothhammer et al, 1977;Malhotra, 1978;Jantzs et al, 1982;Crawford and Duggirala, 1992).…”

Dermatoglyphic variation among Finno‐Ugric speaking populations: Methodological alternatives

“…For quantitative variables, parametric multivariate procedures are frequently used, but are often considered inappropriate for nominal data. Froehlich and Giles (1981) have argued that multiple discriminants may be extended to classification data and employed them in their analysis of dermatoglyphic variation in Papua New Guinea.…”

“…Utilization of dermatoglyphics in population studies has become increasingly common as the perception grows that they may offer new insights regarding historical inference (Froehlich and Giles, 1981;Rudan et al, 1988). Methodological variation in such applications is the rule, the question of how methodological choices may influence results seldom being addressed.…”