There is disagreement in the literature about the exact nature of the phenomenon of empathy. There are emotional, cognitive, and conditioning views, applying in varying degrees across species. An adequate description of the ultimate and proximate mechanism can integrate these views. Proximately, the perception of an object's state activates the subject's corresponding representations, which in turn activate somatic and autonomic responses. This mechanism supports basic behaviors (e.g., alarm, social facilitation, vicariousness of emotions, mother-infant responsiveness, and the modeling of competitors and predators) that are crucial for the reproductive success of animals living in groups. The Perception-Action Model (PAM), together with an understanding of how representations change with experience, can explain the major empirical effects in the literature (similarity, familiarity, past experience, explicit teaching, and salience). It can also predict a variety of empathy disorders. The interaction between the PAM and prefrontal functioning can also explain different levels of empathy across species and age groups. This view can advance our evolutionary understanding of empathy beyond inclusive fitness and reciprocal altruism and can explain different levels of empathy across individuals, species, stages of development, and situations.

The Perception-Action Model (PAM) is a cogent theory of how organisms get information about others' experiences. However, such a stimulus-driven mechanism does not handle well the complex choices that humans face about how to respond to this information. Hyperbolic reward discounting permits a reward-driven mechanism for both how aversive empathic experiences can compete for attention and how pleasurable empathic experiences are constrained.

A primary function of empathy is to help individuals form and maintain social bonds. Empathy should thus occur only when individuals seek to solidify social bonds, and not in response to any opportunity to process others' emotions. Empathy should also involve only certain types of emotion – specifically, emotions that facilitate social bonds – and not any and all types of emotion.

Research only alluded to by Preston & de Waal (P&deW) indicates the disproportionate involvement of some brain regions in the perception and experience of certain emotions. This suggests that the neural substrate of primitive emotional contagion has some emotion-specific aspects, even if cognitively sophisticated forms of empathy do not. Goals for future research include determining the ways in which empathy is emotion-specific and dependent on overt or covert perception.

Empathy can be viewed as an intervening variable to explain complex webs of causation between multiple factors and the resulting responses. The mediating role of emotion, implicit in the concept of an intervening variable, can be at the basis of the flexibility of empathic responses. Knowledge of the underlying neurophysiological mechanisms is needed for empathy to be considered as a biologically functional intervening variable.

A model positing that perception of another's affective state automatically generates matching emotional and instrumental responses predicts more than has ever been observed. Reflexive empathicness would produce emotional exhaustion, inhibitory strain, and debilitate everyday functioning. Self-regulation of empathic responses involves, not only reactive inhibition, but agentic proactive control. Pervasive inhumanities involve selective disengagement of empathic restraints through dissociative psychosocial mechanisms.

In recent years, explanations of primate cognition highlighted clever arguments, rather than different ability. In the target article, definitions unify, explanations rely on basic nervous system functioning, theory is built on data that fit, and the emphasis is on evolutionary continuities. This commentary describes complexities inherent in the development of empathy that are not accounted for in Preston & de Waal's theory.

Empathy is likely more widely distributed among animals than many researchers realize or perhaps are willing to admit. Studies of social carnivores, other group-living animals, and communication via different modalities will help us learn more about the evolutionary roots and behavioral, sensory, and cognitive underpinnings of empathy, including what it means to have a sense of self. There are also important implications for debates about animal well-being.

We question whether empathy is mediated by a unitary circuit. We argue that recent neuroimaging data indicate dissociable neural responses for different facial expressions as well as for representing others' mental states (Theory of Mind, TOM). We also argue that the general empathy disorder considered characteristic of autism and psychopathy is not general but specific for each disorder.

Cooperative peer play emerges in the second year of life. How applicable is Preston & de Waal's (P&deW's) model to the empathic processes in cooperative play? Empathic responses during peer play are more general than they propose, and more dependent on mental state understanding. Moreover, peer play forces children to reason about others' feelings, possibly serving as a unique mechanism for empathy development.

Preston & de Waal are understandably cautious in applying their model to autism. They emphasise multiple cognitive impairments in autism, including prefrontal-executive, cerebellar-attention, and amygdala-emotion recognition deficits. Further empirical examination of imitation ability in autism may reveal deficits in the neural and cognitive basis of perception-action mapping that have a specific relation to the empathic deficit.

A sequential, hierarchical stage model of empathy can account for a comprehensive range of empathic behaviors. We provide an illustrative table, “Stages of Empathy,” to demonstrate how increasingly complex empathic behaviors emerge at each stage, beginning with the infant's “automatic empathy” and ending with the advanced adult's “coconstruction of empathetic reality.”

Under Preston & de Waal's proposed model, empathy might be regarded as everything that determines the quality of a social relationship. Although the authors provide a useful heuristic for understanding relationships, clinical research efforts with a somewhat narrower focus have provided some additional insights into this topic, which might lead to testable hypotheses regarding the neurobiology of empathy.

To date, a wide range of interdisciplinary scholarship has done little to clarify either the why or the how of empathy. Preston & de Waal (P&deW) attempt to remedy this, although it remains unclear whether empathy consists of two discrete processes, or whether a perceptual and motor component are joined in some sort of behavioral inevitability. Although it is appealing to offer a neuroanatomy of empathy, the present level of neuropsychology may not support such reductionism.

Preston & de Waal minimized differences among constructs such as empathy, sympathy, and personal distress. However, such distinctions have been shown to relate differently to altruistic behavior. Moreover, although the authors discussed the role of regulation in empathy, they did not consider the possibility that sometimes empathy is not well-regulated and likely leads to personal distress rather than sympathy.

Within the perception-action framework, the underlying mechanisms of empathy and its related processes of moral behavior need to be investigated. fMRI studies have shown different frontal cortex activation patterns during automatic processing and judgment tasks when stimuli have moral content. Clinical neuropsychological studies reveal different patterns of empathic alterations after dorsolateral versus orbital frontal cortex damage, related to deficient cognitive and emotional processing. These processing streams represent different neural levels and mechanisms underlying empathy.

Empathy is the phenomenal experience of mirroring ourselves into others. It can be explained in terms of simulations of actions, sensations, and emotions which constitute a shared manifold for intersubjectivity. Simulation, in turn, can be sustained at the subpersonal level by a series of neural mirror matching systems.

We argue that cognitive empathy and other instances of mental state attribution are a byproduct of self-awareness. Evidence is brought to bear on this proposition from comparative psychology, early child development, neuropsychology, and abnormal behavior.

The wealth of important and convergent evidence discussed in the target article contrasts with the poorly conceived theory put forward to explain it. The simulation theory does a better job of explaining how automatic “mirroring” mechanisms might work together with high-level cognitive processes. It also explains what the authors' PAM theory merely stipulates.

Attempts to integrate diverse phenomena in terms of common processes are much needed in psychology, but definitional precision is a necessary preliminary to explanation. It is also preferable to use caution in juxtaposing concepts from different realms of discourse.