February 5, 2015
Abstract #1: World wide analysis of genetic and linguistic relationship of human population Author: Chen, Jiangtian, Merritt Ruhlen, and Robert R. Sokal. publish on August 1995
This journal is about different language around the same population on the globe with genetic distances. They had analysis less genetic system (11) than before but have more population
(130). Distance were associated with each genetics distance matrix, because the number of the population were different among the studies. Language and genetics were found in both overall genetic distance and systems. There are also two different type of linguistic matrices. One with higher taxonomic structure and the other one without the structure. The subSaharan Africans separates from nonAfricans in the basal and the Eurasiatic. “Both types of trees indicate a pattern of group of east Asian, Arctic population, and Australian natives separating from
Caucasian and African population.” This show that UPGMA and “neighborjoining tree” came up with linguistic and genetic data of the three group separating from Caucasian and African but the split pattern is not supported by it. It also shows that linguistic is different from people who are at least partially reflected by the genetic differences among them. However, there are existence of other possible higher linguistic structures still needs to be tested.
To keep proportion of missing data low, they had to include the genetic system shared from the study of 130 localities from the whole world. “This resulted in 11 genetic systems, yielding 33 gene frequencies of which 2 are shared by the two ABO systems. The final percentage of missing values in the data matrix was 22.4%.” This mean the numbers of sampled differed among the systems studied, all pairs were calculated separately for each system. If i read it correctly the equation they had use was [d.sub.ij] = [summation of] [absolute value of][X.sub.ik]
[X.sub.jk] where k = 1 to n/2, (1). N is the number of alleles in the genetic system. I and j is the two populations under study, and [X.sub.ik] is the gene frequency of the allele.
According to Ruhlen, the languages belong to 17 language phyla and 1 language isolate. Some of the languages were represented by more than one locality. The distribution of the languages we studied among the phyla are: AfroAsiatic, Altaic, Amerind, Australian, Caucasian, Chukchi
Kamchatkan, ElamoDravidian, Eskimo Aleut, IndoHittite , IndoPacific, Khoisan,
NigerKordofanian, NiloSaharan, NaDene, SinoTibetan, UralicYukaghir , and language isolate is Basque. The team estimated linguistic distances for the populations both within each phylum and among all 18 phyla, including the language isolate. An overall linguistic distance matrix for 130 localities was assembled based on the distance data provided by Ruhlen.
Geographic Data this show the distances were calculated as greatcircle by using the geographic coordinates of the localities.
There are not any chart available but it does describe what is going on. “Nine of 11 systems are significant at p = 0.004 (1/250), which is the lowest possible permutation probability in this study. All significant correlations are positive, as expected… Columns 4 and 5 (Table 1) show the matrix correlations of geographic distances with equalbranchlength and unequalbranchlength linguistic distances, respectively… For the equalbranchlength linguistic matrix, only three systems (ABO, Duffy, and Rhesus) show significant positive correlations, with the first two systems significant at p = 0,004… For unequalbranchlength linguistic distances, four systems (ABO, Duffy, Kell, and Rhesus) show significant positive correlations, all at p = 0.004. For the overall genetic distance matrix the