Nowhere is the gap between the humanities and the sciences manifested more strongly than within anthropology. The gap is manifested as epistemological tensions over reductionism vs. holism, nature vs. nurture, and the study of micro vs. macro context. The divergent research approaches in the humanities and the sciences produce separate bodies of knowledge that are difficult to reconcile. On the one hand, cultural anthropologists study abstract symbolic systems of particular cultural groups while on the other hand, biological anthropologists may study the evolution of the shoulder joint for throwing or carbon-date plants digested by our ancestors. The common thread in both opposing poles of one discipline is an interest in the complexity of human beings and human culture, yet each subdiscipline approaches this complexity with different metatheoretical assumptions and methodological toolkits, which are often incompatible. For example, while we can create a very detailed picture of human behaviour on the level of neural networks, such analyses are very distant from the real complexity of human behaviour. The same could be said about the measures of energy expenditure or hormonal levels, often utilized by biological anthropologists. Recently, however, the rise of the theory of cultural evolution provided a much-needed framework for tackling cultural complexity with experimental methods, manifested mostly in the use of economic games in the field. In my talk, I will briefly review the work within the discipline of cultural evolution and go a step further and suggest that another potentially useful method is to combine observations, surveys, and economic games with monitoring participants’ physiological processes during usual, context- dependent behaviour. Of course, such a combination of approaches can lead to many potential pitfalls, echoed in the criticism of naive consilience put forth by the famous evolutionary biologist E. O. Wilson. To overcome these issues, I will discuss some of the principles of complexity science, including the mechanistic approach, which may help us understand how complex phenomena such as human cultures originate and why we need interdisciplinary and multi-level studies in order to properly model these phenomena. Such a framework should include the understanding of how humans evolved to possess a specific cognitive architecture; how this evolved cognitive architecture functions; how its functioning is dependent on the specific context and socio-ecological environment; and how is the socio-ecological environment evolving in interaction with the human cognitive architecture. By 21 studying human cultures in their contextual variability, mechanistic composition, and evolutionary history, the sciences and the humanities should be able to fruitfully collaborate while avoiding the previous pitfalls of excessive reductionism, progressivist evolutionism, and sweeping overgeneralizations. This point will be illustrated by the authors’ research that combines ethnographic observation with physiological and behavioural observation obtained during real-world cultural events and with data obtained through controlled laboratory studies, which together offer a strategic advantage in understanding complex phenomena such as human cultures.