2026-04-16T08:34:48Z https://datacatalogue.cessda.eu/oai-pmh/v0/oai

85ccccedaa4894f2224064cb779bb02da2984aacefcb83d9d474877a551f7cbe 2025-12-08T02:04:02Z language:enopenaire_data

thakbong56: The ROC census on Taiwan and Penghu in 1956 thakbong56: The ROC census on Taiwan and Penghu in 1956 doi:10.17026/DANS-ZRS-3JXKthakbong56easy-dataset:55398 O. Streiter 2011-10-31 DANS Data Station Social Sciences and Humanities 2013-09-02 Social SciencesCensus conducted by the Republic of China on Taiwan and Penghu in 1956 Visionary research allows for applications that few people had in mind at the time the research was conducted. When Chen and Fried worked in 1964 through millions of ten-year-old census sheets and produced a book filled on 969 pages only with numbers, who but the researchers might not have thought of a waste of time and money? Fifty years later, using a relational database and an intelligent interface, the data have been digitised by the authors of this paper into a data set, called "thakbong56: The ROC census on Taiwan and Penghu in 1956".The interest in the geographic distribution of names in Taiwan, today, as in 1956, relates on the one hand to the emic use of names as an idealised compass in search for an identity. Through one character, a name suggests to its bearer up to thousands of years of history.Driven by the interest in names as collective traits, Chen and Fried sampled in 1964 one quarter of the 9.311.312 personal census cards of 1956. The cards had been stored after 1956 in 361 packages, lumping village and city communities together into townships and city districts. Data on Kinmen and Mazu are not reported.The administrative classification implied by the physical packages formed the first dimension of the 3-dimensional table that Chen and Fried assembled. The 1027 attested Chinese names formed the second, and the ethno-linguistic classification, which distinguishes Fukien, Kwangtung, Others, Aborigines and Mainlanders, the third. Each of the 1.855.540 table cells reports one quarter of the people with name X, living in 1956 in place Y, and being ethno-linguistically classified as Z. The values of two dimensions, for example, CHEN2 (陳) and Taipei, define a probability distribution for the third dimension, for example, [Fukien: 76 per cent, Kwangtung: 2 per cent, Mainlander: 21 per cent].What is logically a tree-dimensional table is restricted to one of three logically possible views, when printed on paper. The probability distribution for CHEN2 and Kwangtung, for example, is hidden in 361 pages. Thus, not one, but three volumes a 1000 pages would have been required to represent all views. Probably as a compensation, Chen and Fried produced a Volume II that shows in maps the spatial distribution of some names under certain ethno-linguistic parameters.The data itself reflects very much the time in which it has been gathered, the technical approach and the prevailing ideology. The storage of papers after the census caused first the unnecessarily broad administrative classification. Then, in 1964 no computer was used. Inconsistencies showed up during the digitisation in margins and higher level tables and have been silently corrected, assuming the more elementary value to be the correct oneii. Finally, social tensions and strong ideologies in the time of the census left their imprint on the data.First of all, the ethnic classification foresees neither ethnic changes across generations nor mixed ethnic ancestry. In Taiwan, however, Indigenous groups, Holo and Hakka have been constantly crossing ethnic boundaries through adoption, inter-ethnic marriage or acculturation. Second, the language of the people was used as prevailing criterion of the census, and levelled out ethnic variations in a period, the KMT hoped to reconquer the Chinese mainland through the ‘Chinese nation’. Holo-speaking Hakka and Pepo communities have, to different degrees and depending on the local administration, been classified as Fukien, as the comparison with older, Japanese census data shows. https://ssh.datastations.nl/oai doi:10.17026/DANS-ZRS-3JXK 2025-12-03T01:00:44Z ddi:codebook:2_5