Biology of human behaviour across timescales, from neurons to culture
Sapolsky, a Stanford neurobiologist and primatologist, explains human behaviour by working backwards through time from any given action. He starts one second before the behaviour with the nervous system, then seconds to minutes earlier with sensory stimuli, hours to days with hormones, days to months with neural plasticity, years with adolescence and childhood, then foetal development, genes, culture, ecology and evolution. Each chapter covers one timescale.
The method is the argument. By the end Sapolsky has shown that no single level of explanation is sufficient, and that claims about the cause of a behaviour, whether genetic, hormonal, cultural or situational, are always partial. He is particularly good on how the same biology produces opposite behaviours depending on context, showing that testosterone amplifies whatever behaviour maintains status in a given culture rather than causing aggression, and that oxytocin increases in-group warmth and out-group hostility simultaneously.
The material on us and them distinctions, on hierarchy, on the amygdala and prefrontal cortex, and on how adolescent brain development explains risk-taking is directly relevant to anyone managing people. His treatment of implicit bias, of moral reasoning as post-hoc justification, and of how quickly and unconsciously group categorisation occurs is well documented.
The final chapters on free will and criminal justice are the most contested. Sapolsky argues that free will does not exist and that the criminal justice system should be rebuilt accordingly, which many readers find unpersuasive and which is philosophy rather than neuroscience. The book is long and technically dense, with substantial appendices on basic neuroscience and endocrinology. Some cited social psychology has replicated poorly. The multi-level framework is the durable contribution.
Managers and investors who want a rigorous account of why human behaviour resists single-cause explanation, prepared for a long technical book.
Chosen by shared topics, the entries that hold them, and the people who recommended both.


