Abstract

Evolutionary theory has proven to be an influential model for examining change in technological systems. However, the application of an evolutionary framework often requires macro-scale data that is informative of technological change at the population level over time. Archaeological and cultural data repositories present an emerging source of valuable information to study the evolution of technology but are commonly viewed with skepticism due to uncertainty about data aggregation, lack of data structure, and sampling biases. Here, a case study that uses mined data on American automobiles is used to highlight a novel Bayesian approach for estimating rates of technological origination and extinction from occurrence data. Results from this analysis are compared with preliminary findings on the rate of diversification in both modern and ancient technologies to explore commonalities in patterns of technological change. This research concludes that data science has much to offer social sciences for future quantitative analysis of technology and culture.