This study reviews Gergely and Watson's (1996) social biofeedback theory of parental affect-mirroring, furthered by Fonagy, Gergely, Jurist, and Target (2002), especially the “markedness” hypothesis. These authors hypothesize that a salient facial expression and the singsong prosody of markedness called “motherese” are necessary for successful emotion sensitization and symbolization. The revision investigates whether this is accomplished through “internalization” mechanisms requiring markedness of the mirroring, or solely through social biofeedback processes. The article argues that the infant contingency-detection mechanism (similar to that of adult biofeedback training) mediates the functions of sensitization, representation, and symbolization of emotions through its processes of covariance–invariance detection, maximization, and contingent control of the parental mirroring, whether the mirroring is marked or not. The review argues that a caregiver producing a covariant–invariant mirroring can help bring the infant's emotional somatosensations to consciousness along with its implicit dispositional content without using motherese. It considers the clinical implications of the new model and speculates about parental difficulties centered on sharing troubling emotions. Finally, it discusses how the model may be a mediating mechanism in the change process in the therapy of adults through the promotion of sensitive emotions, their awareness, and symbolization.