《元和志》戶籍與實際戶數之比勘

嚴耕望

    三、四世紀或稍前,西域學人已視中國為世界人口最多之國度。歷代正史自《漢書•地理志》以下多記載戶口,惟此類數額,僅為地方政府向中央政府申報之戶口,地方實際戶口數額多超過此項數額,而且往往超過甚多。今以唐代《元和志》為例,證明此種情形。

    我寫此文主要是檢查元和時代各州之實際戶數,以與《元和志》所載戶數相比較,發現《元和志》所記各州戶數,大多僅為其他文獻所記各州戶數若干分之一,甚至八九分之一。又取各方鎮軍府之實際兵額與《元和志》戶籍作比較,兵數往往為戶數之若干倍。一家供數兵,此亦絕對不可能者。此外又有豐年和糴額與凶年賜粟額諸事例,皆足證明元和時代地方政府向中央申報之戶數僅為當地實際戶數若干分之一。所以我們若專據地方政府向中央申報之戶口數字談人口,則差誤必極大。

 

A Comparision of Household Register Figures with the Actual Numbers

Yen Keng-wang

New Asia Institute of Advanced Chinese Studies, Hong Kong

    As early as the third and fourth centuries or before scholars from the Western regions estimated China to be the most populous country in the world. Most geography sections of official dynastic histories, beginning with Han-shu, contained household figures. These figures were derived from statistics reported by local governments to the central government. However, the number of households contained in local registers surpassed by far the numbers reported to the central government. This paper demonstrates this phenomenon by examining the Tang dynasty Yuan-ho chih.

    The main purpose of this paper is to examine the actual household numbers in each prefecture (chou) during the Yuan-ho period. A comparison of the household figures recorded in the Yuan-ho chih indicates that the number of households reported in this source is only a fraction of those reported in other available sources, sometimes only 10 or 20 percent. A comparison of the actual number of military personnel in military commanderies with the household numbers recorded in the Yuan-ho chih yields a number of military personnel several times larger, yet this was clearly not the case. Moreover, a comparison the grain quotas during years of good harvest with those during disaster years also shows that the number of households reported by local government during the Yuan-ho period was only a fraction of the actual number at the local level. Therefore, if one relies on the household figures reported by the local government to the central government to make population estimations, one seriously underestimates the actual numbers.