THE DIALECTS OF HONG KONG BOAT PEOPLE
49
of any living resident and they have no consensus on their own provenance. In chatting with my informants on this subject I found some agreement that Tung Kun District was their source plus much speculation and guesses ranging from 'some place up north' to 'maybe Fukien Province'. The northern origins are of course common to all Han Chinese and reflect no special knowledge on the part of the informant. The possibility of Fukien Province seems completely unsupported by the linguistic evidence, but in view of the fact that many Boat People are Swatow, Hoklo, or other obviously Fukien types2, it is more than possible that Fukien individuals have been absorbed by the Kau Sai group from time to time. However, there is evidence to indicate that some area reasonably close to Tung Kun District may well be the origins of this community.
Concerning the Boat People, certain assumptions have been made elsewhere which do not seem valid or which should at least be held in abeyance until making a number of the studies of the type I will describe here. First, the Boat People, or sometimes those referred to specifically as the Tanka, are often treated as a homogeneous group which represents the remnants of the earliest inhabitants of the South China regions, assumed to descend from the non-Han tribes and to have been assimilated and acculturated as the Han peoples moved into this area. It is difficult to refute this point except with cultural and linguistic data which support Ward's (1965) point that the boat people's descent is probably neither more nor less non-Han than that of most other Cantonese speaking inhabitants of Kwangtung.1 It would be reasonable to assume that some Yao or other southern barbarian blood may still flow in local veins but probably to about equal degree in the Boat People as in the average resident of Kwangtung Province. With nothing very concrete to go on we would be in the same position if we discussed the amount of Pict blood in today's inhabitants of the British Isles.
When we do not have complete historical evidence for origins of a group it is possible to get information from other sources, such as archeology, anthropology, and linguistics. However, with all these fields our results will be more reliable if we are dealing with an overall picture of structured data rather than extracted