We commend Burkart et al. for their impressive review of current theory and research on the evolution of general intelligence. Although their analysis yields a persuasive account of the empirical data regarding general intelligence across species, it also motivates increased attention to key components of human cognition that, in our view, will prove to be essential for a complete understanding of the evolution of cognition.

A comprehensive theory of general intelligence and its evolution must somehow account not only for the commonalities evident across species, as targeted by Burkart et al., but also for the obvious wide divide between humans and nonhuman primates. Although we agree that the “cultural intelligence approach” endorsed by the authors will be at the heart of a successful theory, we believe their analysis also suggests directions for further elucidation that would specifically address what is special about human cognition.

Specifically, and consistent with the cultural intelligence approach, we propose that language and theory of mind (ToM) together can go far toward explaining superior human general intelligence. Language and ToM arguably have the markings of primary modules, at least as characterized by Burkart et al. (sect. 1.2.3, Table 2): Both appear rapidly and dependably early in development across all cultures without explicit instruction and even in the most disadvantaged and deficient environments (e.g., Wellman et al. Reference Wellman, Cross and Watson2001).

Others can better address the role of language in human intelligence (see Dunbar Reference Dunbar, Hurford, Studdert-Kennedy and Knight1998; Smith Reference Smith, Carruthers and Smith1996), so we focus here on the contribution of ToM, defined as the universal propensity of humans to understand and explain their own and others' behavior in terms of internal mental states and processes such as beliefs, desires, goals, and intentions (e.g., Wellman Reference Wellman1990). We note here that language and ToM are so ontogenetically and symbiotically intertwined that they are difficult to discuss separately. Indeed, some have made a strong case for their co-evolution: “The evidence at hand suggests that language and theory of mind evolved … in constant interaction, serving one primary adaptive goal: to improve social coordination” (Malle Reference Malle, Givon and Malle2002, p. 280).

Nevertheless, a conceptual case can be made for ToM alone being especially foundational for human general intelligence. As a developing domain-specific causal framework that supports advanced meta- and social cognition, such as recognition of the possibility of false beliefs and variation in knowledge states between individuals, ToM contributes essentially to social learning as characterized by the cultural intelligence approach. We offer here a few examples of phenomena illustrating the crucial role of ToM in social learning and, by extension, the evolution of human general intelligence. For a recent collection of relevant research and commentary on these issues, see Legare and Harris (Reference Legare and Harris2016), and especially Tomasello (Reference Tomasello2016).

Even children under age 2 engage in observational imitative learning that depends on attention to the actor's intended action or goal. For instance, Meltzoff (Reference Meltzoff1995) demonstrated that when shown an actor who tried and failed to achieve a goal, such as hanging a string of beads on a peg, 18-month-old children imitated what the actor tried to do rather than what the actor actually did (dropping the string of beads). Carpenter et al. (Reference Carpenter, Nagell and Tomasello1998) found that young children shown adult demonstrations of action sequences containing both accidental and intentional actions more often reproduced the intentional actions, suggesting that even early observational learning is contingent on an appreciation of another's intentional state.

Similarly, research on children's early attunement to potential knowledge in others reveals the advantages for a social learner afforded by an emerging ToM. Even 2-year-olds attempting to locate a desired object efficiently seek help from informed rather than ignorant adults (O'Neill Reference O'Neill1996), demonstrating an incipient appreciation of mental states that will eventually help them when they seek to obtain information rather than objects. Recent studies of children's reaction to testimony suggest that, although children generally assume what they are told is truthful (e.g., Harris Reference Harris2012), they are more willing to believe an expert or experienced source. Moreover, as children advance from early to middle childhood, their evaluation of others' testimony increasingly recognizes the possibility of deception or distortion (Mills Reference Mills2013). On the positive side, children's growing sophistication eventually includes recognizing the specifically pedagogical motives of adult teachers, as described in Gergely and Csibra's (Reference Gergely, Csibra, Enfield and Levinson2006; Csibra & Gergely Reference Csibra and Gergely2009) proposed theory of natural pedagogy. Children's particular “readiness” to learn through social instruction (and also to teach others; see Flynn & Whiten Reference Flynn and Whiten2008) hinges on a recognition of the internal content of others' minds. From our own research, we can add that young children's reflections on their own learning, as assessed in transcripts of at-home conversations, are in fact particularly focused on their teachers and what their teachers know as opposed to other aspects of learning events, such as how or when learning occurred (Bartsch et al. Reference Bartsch, Horvath and Estes2003).

Early childhood is replete with phenomena suggestive of the central role of ToM in social learning. In addition to these examples, the acquisition of belief understanding, widely recognized as a hallmark of developing ToM and first evident around 4 years of age in children's explicit predictions and explanations of action (e.g., Wellman et al. Reference Wellman, Cross and Watson2001), can also be viewed as foundational to the most advanced human achievements, such as those that occur in the collaborative activity of scientific research and the uniquely human creation of formal methods to facilitate learning (e.g., Meltzoff et al. Reference Meltzoff, Kuhl, Movellan and Sejnowski2009). At a fundamental level, our greatest accomplishments must rest on our basic capacity to imagine and recognize the variations and vulnerabilities of human cognitive states. With this in mind, future researchers guided by Burkart et al.'s final recommendation to explore further the relationship between intelligence and socio-cognitive abilities would do well to direct those efforts toward theory of mind.