Empathy is a process essential for mitigation of human suffering and for the creation and long-term stability of social bonds. Although it has recently become the focus of intensive study after decades of neglect, models of empathy emerging from cognitive neuroscience, affective neuroscience, and functional imaging studies show considerable confusion about defining empathy and widespread differences of opinion about the cognitive vs. affective dimensions to empathy. Human empathy probably reflects variable admixtures of more primitive affective resonance mechanisms, melded with developmentally later-arriving theory of mind and perspective taking. This integration of primitive with more cognitive mechanisms occurs under the “supervision” of a motivated valuing of another sentient creature, a supervision that underlines intrinsic ties between empathy and attachment processes. We know little, however, about how more primitive resonance-induction mechanisms centrally involved in attachment connect developmentally to more later arriving cognitive theory of mind components. From these considerations, a basic model of affective empathy is generated as a gated resonance induction of the internal distress of another creature, with an intrinsic motivation to relieve the distress. It is “gated” in that at least four classes of poorly mapped variables determine the intensity of an empathic response to the suffering of another. Potential classes of variables affecting empathic inductions are: (1) genotypic and (2) phenotypic effects; (3) state-dependent influences (on the affective state of the empathizer); and finally (4) the perceived qualities of the suffering party. Most current models have failed to conceptualize this critical “gating” process, subsequently losing any naturalistic predictive efficacy in modeling real-world social phenomena. Contagion has been generally neglected in affective neuroscience, but it may point to poorly understood receptive processing capabilities, embedded in the distributed paralimbic and subcortical architectures for primary emotion. Thus, contagion may be a developmentally primitive emotion-induction mechanism that cognitive development largely (but not totally) supplants. Detailed differential predictions of this model are proposed.