In this study of the population statistics of Ming China I have relied almost exclusively upon figures from contemporary sources: the Veritable Records (Shih-lu) of the dynasty, the Collected Statutes (Ta-Ming Hui-tien), the Imperial Topography (Ta-Ming Yi-t'ung-chih), the Huang-yü K'ao, and a number of provincial and local gazetteers.
The first question that arises in using such works is that of the reliability of their figures and indeed of Chinese statistics in general. A great deal of scepticism has been, and still is, expressed on this point. To take a single example only: everyone who has researched in this field has experienced over and over again that where a Chinese text gives both a global total and an itemized breakdown of that total, the sum of the separate items usually amounts to something more than (or less than) the global figure, and hardly ever agrees exactly with it. A natural reaction from this unsatisfactory state of affairs is to reject all the figures concerned as unreliable. To do this, however, is tantamount to denying the possibility of quantitative investigation in Chinese economic and institutional history. Before coming to such an extreme conclusion it is at least advisable to see whether the figures are capable of rational emendation, and whether figures so emended will show a higher degree of internal consistency.