114
H. G. H. NELSON
(in an attempt to use the physical structures as a basis for a random sampling frame), and reached the conclusion that in the conditions of New Territories villages, it is strictly undefinable. The local "village houses" are built in terraced rows: each "house" consists of a single "room", which, although generally divided by a partition into living and sleeping quarters, is regarded as the living accommodation proper to one married couple and their unmarried children. It is very rare for an old and widowed mother to share the same "house" as her son and daughter-in-law: I have not come across a single instance of more than one married couple, however closely related, sharing the same "house". The correspondence between room, house, and household is by no means exact: not only may one household (defined as a group of kin sharing a common budget and a single stove) be spread over a number of "houses", but where one household is occupying two adjacent "houses", an interconnecting door may be opened between them, turning them, to all intents and purposes, into a single "house" of two "rooms". It might almost indeed be more logical to regard each unit in a terrace as a "room", and define "house" as that combination of "rooms" which is occupied by a single household, even when the "rooms" are located in more than one terrace; but Chinese usage does not support this — each unit is called uk, never fong.
The Hong Kong Government surveyors who numbered all agricultural and building lots in the early years of this century evidently came up against a similar problem of definition; and there are inconsistencies in the system of numbering they adopted. Where two adjacent units were owned (and occupied?) by one man, they are numbered together as one (double-sized) house-lot. There is one row in Sheung Tsuen, however, where the living quarters of each house are separated from their kitchens by the narrow lane that gives access to them: it is clear from the pattern of ownership in 1905 (when the Block Crown Lease was drawn up), and from the way these structures have been passed down the generations since then, that each kitchen is linked with a single
* I use the Meyer-Wempe Romanisation of Cantonese terms throughout.
† Sheung Tsuen (1) in the Pak Heung (^#) where Mr. Nelson carried out his research. See p. 171 of A Gazetteer of Place Names in Hong Kong, Kowloon and the New Territories (Hong Kong Government Printer, n.d, but 1960), Ed.
Page 120
Page 121