Our maximum-likelihood analysis of the data of Serjeantson et al. (1983) (Figure 1) confirms their conclusions that no clear genetic distinction exists between the An and NAn speakers residing along the north coast of Papua New Guinea, and that population affinities are based more on geographical proximity than on linguistic similarity. There are some differences between the branching patterns of our network and their dendrogram, but it remains the case that, for the most part, populations that are geographically close to each other are relatively similar genetically, irrespective of their linguistic affiliations. Thus, for example, the An Ham are more closely related to other NAn groups in the Gogol Valley than they are to the other An populations. Similarly, the An Manam resemble their NAn neighbours in the Bogia District more than they do other An populations. Two groups on Karkar Island (An Takia and NAn Waskia) are also close to each other in the network, as are the three groups from the Siassi Islands (An Mangap and Lokep and NAn Kovai).

The situation with respect to the genetic affinities of NAn populations belonging to different phyla is less clear-cut, although, once again, geographical proximity appears to have some influence. Thus, in the Bogia District, Pay and Tani (Trans New Guinea Phylum, Adelbert Range Superstock) are more closely related to Monumbo (Torricelli Phylum) than they are to other Adelbert Range Superstock groups in the Gogol Valley (Amaimon) and on Karkar Island (Waskia). Other Adelbert Range groups in the Bogia District (Saki and Pila) are, however, also relatively distantly related to Pay and Tani.