here are two different theoretical frameworks accounting for the origin of mirror neurons (MNs) in primates. According to the ‘genetic’ hypothesis, MNs’ matching response properties was acquired through natural selection and stored in the species’ genome. This ‘mirroring’ adaptation is evolutionarily selected to support a wide range of social functions so that individuals can use their own motor (or affect) representations to understand other’s actions (or emotions). In contrast, ‘associative’ theories argue that the mirroring mechanism comes from individual’s post-natal sensorimotor experience. Among the associative theories, associative sequence learning (ASL) hypothesis tries to elucidate the ontogeny of MNs with an extreme empiricism. By using a paradigm called S-R procedure, ASL hypothesis collected various empirical evidences and became the most influential associative theory. Although the evidence that MNs are malleable to sensorimotor experience can be well explained by ASL, it does not fully rule out the possibility that gene biases the development of MNs. In the history of evolution, it seems that MNs are most likely developed in those highly socialized animals lived in a socio-ecological niche required to communicate, cooperate and compete with inter-group members. In the early development of new-born primates, the brain responds differently to perceptual event with social significance (lipsmacking) than non-social one (tongue protrusion). Evidence suggests that these individual differences of brain responses to social action in early age can predict the later engagement of social interaction. Finally, the empirical evidence from ASL is primarily based on the stimulus-response compatibility paradigm, which is not ecological valid to test the emergence of MNs. Thus, the ontogeny of MNs is still in debate and requires further clarification.

在灵长类镜像神经元系统的研究中，关于镜像神经元系统的产生主要存在两种观点，基因假说认为镜像神经元系统的出现基于一定的演化压力，服务于动作理解这一社会性功能；而联想学习假说则认为镜像神经元系统是感觉运动学习的产物，并不服务于动作理解。在联想学习的阵营中，联想序列学习主张一种极端的后天论立场，并通过刺激-反应范式提供大量的实证证据。但是，联想序列学习虽然能够解释大部分镜像神经元的研究，却无法否定其发生机制中的基因成分。比较镜像神经元系统在多个物种中的演化历史，探索个体早期发展对社会性信息的偏好均显示出联想序列学习存在局限，而联想序列学习提供的证据也不具有生态学效度。因此，对镜像神经元的基因基础如何影响其产生、发展与演化的探讨仍将是一个极具意义和生命力的主题。