We introduce archetypal analysis as a tool to describe and categorize scientists. This approach identifies typical characteristics of extreme ('archetypal') values in a multivariate data set. These positive or negative contextual attributes can be allocated to each scientists under investigation. In our application, we use a sample of seven bibliometric indicators for 29,083 economists obtained from the RePEc database and identify six archetypes. These are mainly characterized by ratios of published work and citations. We discuss applications and limitations of this approach. Finally, we assign relative shares of the identified archetypes to each economist in our sample.