Abstract
Anthropologists, conservation biologists, and psychologists have generally found a long-term (secular) decline of ethnobotanical knowledge among indigenous people. To estimate such knowledge loss, researchers have typically relied on a single cross-sectional data set to (a) measure knowledge among people ofdifferent ages, (b) compare measures between ages, and (c) infer a loss of knowledge if the old knew more than the young. We improve on the approach by simultaneously controlling for cohort effects and age effects - the first refers to the effect of the birth period and the second refers to the effect of the life cycle (or aging). Failure to simultaneously control for both effects may produce the misleading impression that the old know more than the young, and the conclusion that the difference reflects a secular loss of knowledge when in fact it may reflect different positions in the life cycle. We use data collected during 2005 from a native Amazonian society offoragers-farmers in Bolivia (Tsimane') to estimate secular changes in knowledge. Participants included 269 women and 287 men (age ≥20) born 1920-1985. We equate knowledge with theoretical knowledge of useful plants and use cultural consensus to measure knowledge. Multiple regressions were used with knowledge as an outcome and age, birth decade, schooling, and sex as explanatory variables. We find no significant secular change in knowledge in the main analysis, but results were sensitive to (a) the definition and domain of ethnobotanical knowledge and (b) the sample. In the sensitivity analysis we find evidence of a secular increase in knowledge, consistent with the view that knowledge is dynamic and changes.