William James in his 1872 publication entitled “Are We Automata?” makes the point that “a succession of feelings is not one and the same thing with a feeling of succession, but a wholly different thing” (James Reference James1872, p. 5 ). James insists that: “The latter feeling requires a self-transcendency of each item, so that each not only is in relation, but knows its relation, to the other.” James concludes that this “self-transcendency of data” constitutes what consciousness is, in his own words “what constitutes the conscious form” (p. 6 ). It is also what makes us nonreducible to automata.
If we agree with James that we are more than machines, an engineer look at self-consciousness (e.g., metacognition and introspection) and the consciousness of others (mindreading and theories of mind) is insufficient and does not capture the essence of the phenomena.
At the core of these issues, there is more than the rational attribution of mental states to the self or to others (theories of mind). There are also feelings and emotions, what make us sentient and incommensurate to machines. In the social domain, “self-transcendency” corresponds first and foremost to the feeling of our relation to others, not just to a propositional attitude turned toward others or onto the self (metacognition).
Children in their social-affective development tell us that upstream to mindreading and metacognizing, there is the affective urge to see the self as seen and evaluated by others – the deep-seated affective propensity to have “others in mind” (Rochat Reference Rochat2009).
The origins of a tendency to attribute attitudes (i.e., adopt a propositional attitude stance) toward others as well as toward the self are rooted in the basic propensity to know where we stand emotionally in relation to others: whether we feel affective distance or closeness, a sense of affiliation and approval, or on the contrary a sense of disapproval and rejection from others.
There is now much evidence showing that long before children display signs of explicit “theorizing” about the mental states of others, as well as their own, whether for example they have an appreciation of their own knowledge or ignorance, they have already developed astute implicit intuitions of whether they can more or less “trust” someone, or “trust” themselves in what they believe and know.
By 2 months, infants expect someone to behave in a certain way in certain social contexts. They show distress and emotional withdrawal toward an adult who abruptly and for no apparent reasons adopts a frozen facial expression. They also show resistance in recovering positive affects following a still-face episode (Tronick Reference Tronick, Nadel and Muir2005; Tronick et al. Reference Tronick, Als, Adamson, Wise and Brazelton1978). Infants are indeed highly sensitive to the attention received from others, constantly gauging how much others are engaged toward them (Reddy Reference Reddy2003). They are also quick to develop a preference for the interactive contingency style of their mother, generalizing such preference when encountering strangers (Bigelow & Rochat Reference Bigelow and Rochat2006).
These facts demonstrate that at the roots of mindreading and metacognition, there is a complex pragmatics of intentional communicative exchanges that is not just rational, but ontologically affective and emotional. The first nonverbal language of children is well documented, expressed from at least 2 months of age in the context of face-to-face interactions, made of mutual gazes, precise contingency of exchanges and turn taking, with compulsive “motherese” and other affective markers from adults.
The syntax of this early nonverbal grammar (e.g., when I smile, the other should not scream or cry) and the semantics that children derive from it (e.g., if I smile and the other looks away, something is wrong) is primarily emotional and affective. These rich and reciprocal communicative exchanges quickly determine a proto-rationality that is expressed in the epistemic as well as affective trust that children from the earliest age place in others, as well as eventually into themselves.
Topal et al. (Reference Topal, Gergely, Miklosi, Erdohegyi and Csibra2008) provide further evidence of the deep affective and emotional roots of mindreading and metacognition. These authors show that the famous Piagetian stage 4A-not-B error of object permanence found in 10-month-olds and younger infants, depends in part on the presence or absence of subtle communicative and emotional cues (i.e., eye contact, motherese, affective attunement, or social contingency) from the experimenter hiding the object. Infants are inclined to interpret the adult as wanting to teach them something (a hiding game), while having a good, intimate social time of sustained mutual attention. Topal et al. show that 10-month-olds' perseverative search errors can be induced by pragmatic misinterpretation of an adult's intentions.
By the time children objectify themselves in mirrors passing the famous mark test (Amsterdam Reference Amsterdam1972), they also begin to express secondary emotions such as embarrassment, shame, or pride (Lewis Reference Lewis1992). Early on, infants demonstrate a complex appreciation about their public appearance, long before they express rational theories of mind and metacognition. Affectively, children first feel how others like or dislike them (affective reading), gauging also how they think they might be liked or disliked by others (meta-affectivity). Both seem to co-emerge in development, necessarily codependent.
From a developmental vantage point, affective reading and meta-affectivity are ontologically linked, representing two sides of the same coin. Findings on early social and emotional development demonstrate that there is not much empirical grounding for conceptualizing them as separate processes.
In short, let us not forget that we are born in need of social attention; and it is, I would suggest, in this primary motivational context that early social expectations develop. These expectations, co-constructed in the history of interactions with others, become by 4 to 5 years of age explicit “theories” about self and others' propositional attitudes. This primary motivational context represents an invaluable source of information that cannot be dismissed or overlooked, particularly if one attempts at capturing “how we know our own mind.”
Dealing with the question outside of any motivational context, as done by Carruthers in the target article, is a disembodied, abstract exercise. By analogy, it is like playing chess or moving armies on a map instead of being physically and emotionally engaged on the battlefield. The battlefield is riskier, but closer to what it is all about.
