The problems of incompleteness and inconsistency of the data in ethnob
otanical research are considered. Some inconsistency, such as that see
n in the identification of plants, is unavoidable, but if care is take
n, it can be reduced step by step. Some inconsistency, however, like v
ariation in the ways plants are used, is intrinsic to ethnobotanical d
ata, and must be recognized as such. Inconsistency does not mean usele
ssness so long as we recognize the characteristics of ethnobotanical d
ata and use, suitable methods for handling them.