The Baldwin effect is a hypothetical process in which a learned response to environmental change evolves a genetic basis. Modelling has shown that the Baldwin effect offers a plausible and elegant explanation for the emergence of complex behavioural traits, but there is little direct empirical evidence for its occurrence. We highlight experimental evidence of the Baldwin effect and argue that it acts preferentially on peripheral rather than on central cognitive processes. Careful scrutiny of research on taste-aversion and fear learning, language, and imitation indicates that their efficiency depends on adaptively specialised input and output processes: analogues of scanner and printer interfaces that feed information to core inference processes and structure their behavioural expression.
Sinking InWhich features of the human mind, now genetically inherited, were once the products of learning? This question relates to the Baldwin effect (see Glossary)a process in which an initially learned response to environmental change acquires a genetic basis [1][2][3][4]. This process is also known, sometimes in a misleading way [5], as organic selection [1], genetic assimilation [6], and experiential canalization [7]. Whatever it is called, Baldwin's hypothesis postulates that, at the population level, learned characteristics can 'sink in'; they can become part of what offspring inherit organically from their parents.In the following we bring empirical research from cognitive science to bear on the hypothesis that some aspects of human cognition have sunk inthat they were initially learned and are now genetically inherited. We argue that there is good empirical evidence for this hypothesis regarding peripheral cognitive mechanisms (i.e., input or perceptual systems, and output or action systems, both modulated by attention and motivation), but not for the complex and interlocking central processes, involving inference and memory, that have durable effects on the relationships between perception and action [8] (Figure 1, Key Figure).After introducing the Baldwin effect, we next focus on preparedness [9,10]. For 50 years, experimental psychologists have investigated the preparedness of mechanisms involved in learning taste aversions and fears, and how these mechanisms have been specialised by evolution to do their jobs well. We argue that this research not only provides much-needed empirical evidence of Baldwinisation but also illuminates the targets of selection. Specifically, research on preparedness suggests that, across species, Baldwinisation has had a much greater impact on peripheral than on central cognitive mechanisms [11]. There has been genetic assimilation of learning-induced changes in perception, attention, motivation, and motor controlprocesses that modulate inputs to, and outputs from, the core inference processes of the brainbut little if any change to the structure of central processors themselves. We then extend the analysis to language and imitation, human faculties that are widely thought to have been gene...