Brain imaging studies show that social pain (like social rejection, embarrassment, and guilt) activates brain regions characteristic of painful bodily experiences. The brain regions that are activated by both evoked social and physical pain include the anterior cingulate cortex, the right prefrontal lobe, the insula, amygdala, and somatosensory cortex. Even deep brain structures, such as the brainstem periaqueductal gray (PAG), are known to be evoked by mother–infant separation, marked by intense and repeated distress cries. These functions are highly conserved among mammals and, perhaps, birds (Eisenberger & Lieberman Reference Eisenberger and Lieberman2004; Nelson & Panksepp Reference Nelson and Panksepp1998).
This evidence contradicts Carruthers' hypothesis that we learn about ourselves by turning our social mindreading capacities upon ourselves. No doubt we do learn about ourselves based upon what we have learned about others. After all, we constantly transfer knowledge between different domains of reference. However, it is simply not the case that all of our introspective self-knowledge is of this kind. Children acquire “theory of mind” abilities in about the fourth year of life. But long before that time we can observe, pain and pleasure perception, the distress of abandonment, anticipatory fear and joy, and a wide spectrum of social and imaginary emotional experiences.
Carruthers could maintain that such emotional experiences are not true cases of “metacognition” and “introspection.” It is possible to define such terms in very limited ways, but there is no doubt that emotional feelings express propositional attitudes: They are about something, namely the well-being of the self. Thus, hunger, thirst, air-hunger, social distress, fear of rejection by the mother, peer envy, and numerous other infant emotions are by no means simple “reflexes.” They are socially contingent, though not explicitly deliberated, reactions to real-world events that are critical to the infant's survival. This crucial self-related information has extraordinary breadth of conservation among mammals, suggesting an evolutionary history of some 200 million years (Baars Reference Baars2005).
Pain is not the only kind of introspective experience humans have with minimal social input, but it is perhaps the most compelling. Metacognitive self-report (“introspection”) has been used for two centuries in psychophysics. It is a well-established methodology that converges extremely well with other empirical evidence, such as brain recording methods (Baars & Gage Reference Baars and Gage2007).
Science is a constructive enterprise, but it is tightly constrained by evidence. That is why, like other human activities such as farming and tax accounting, it is not merely constructed, but also bound by considerations of accuracy and predictability. That is true for humans, but it is equally true for animals, who must survive real-world challenges in environments in which errors lead to extinction. Brain evolution is not separate from the ability to observe and know the real world. On the contrary, when we are given truthful feedback about the world, humans and other animals become quite reality-based. There is no contradiction between constructivism and realism.
