Kline provides a commendable synthesis, valuable for designing future research on learning and teaching both cross-culturally and across species. Kline ties teaching and learning to the cognitive, social, and cultural evolution of humans. However, although Kline claims to have incorporated “all known teaching mechanisms in humans and other animals into a cohesive theoretical framework” (sect. 3, para. 2), the role of social play is virtually ignored.

Play only seemingly serves no function and has no apparent, immediate benefits, but the incredible phylogenetic depth of play strongly suggests that play is a functional and adaptive behavior (Burghardt Reference Burghardt2005). It is well appreciated that play facilitates learning in humans, and it is thought that learning is a key explanation and function of play across the animal kingdom (Pellegrini Reference Pellegrini2011; Pellis & Pellis Reference Pellis and Pellis2009).

Many lines of evidence point to social play as a prime building block for the evolution of both learning and teaching. Play, in human and nonhuman infants, provides opportunities for the acquisition of relevant stimuli, for the diffusion of social and environmental knowledge, and for the development of fundamental cognitive and communicative capacities (Power Reference Power2000). Play reduces spatial separation between individuals (Macaca tonkeana, Palagi et al. Reference Palagi, Dall'Olio, Demuru and Stanyon2014) and increases cooperation and empathic proclivity (Pan paniscus, Demuru & Palagi Reference Demuru and Palagi2012; Theropithecus gelada, Palagi et al. Reference Palagi, Leone, Mancini and Ferrari2009). Play renders individuals more competent in the emotional domain (Pellis & Pellis Reference Pellis and Pellis2009), thus creating favorable conditions for learning and teaching (Ciani et al. Reference Ciani, Dall'Olio, Stanyon and Palagi2012). In brief, play cannot be separated from learning, and is a ready-built platform from which to launch teaching.

What differentiates play from other forms of teaching and learning is the positive reinforcement arising from its pleasurable nature (Lewis & Barton Reference Lewis and Barton2006; Rilling et al. Reference Rilling, Scholz, Preuss, Glasser, Errangi and Behrens2011). Play between adults and infants, in particular, is a behavioral setting in which a “teacher” and a “learner” can be clearly recognized. While benefits always outweigh costs for immature learners, costs for adult teachers are generally high and strongly dependent upon the type of play and on the social environment. In order for teaching to evolve, there should also be benefits for the teacher. For example, teachers may be related to their pupils, or pupils may be future allies or mates (Mancini & Palagi Reference Mancini and Palagi2009). In species characterized by a complex sociality, such as human and nonhuman primates, play between adults and immature subjects can be an important bridge to broaden and strengthen social networks of the adults (Palagi et al. Reference Palagi, Cordoni and Borgognini Tarli2004; Reference Palagi, Paoli and Borgognini Tarli2006).

Social play is a multifaceted behavior with various typologies (Cordoni & Palagi Reference Cordoni and Palagi2011; Palagi Reference Palagi, Coplan and Bowker2014; Palagi & Cordoni Reference Palagi and Cordoni2012), which come in succession in early stages of life and then integrate into one another. Information gathered and lessons learned by playing can later be recruited and tinkered to cope with many other aspects of life. Long-standing data show that the longer the period of development and learning, the more pervasive the social play (Fagen Reference Fagen, Pereira and Fairbanks1993).

Play between mother and offspring is probably both ontogenetically and phylogenetically the first means of teaching and learning. In primates, the role of mothers in the playful exchanges with newborns is fundamental. For example, great ape mothers are responsible for initiating and ending play sessions with offspring in their very first months of life. Later, infants spontaneously become more interactive in initiating new playful interactions with their mothers (Hoff et al. Reference Hoff, Nadler and Maple1981; van Lawick-Goodall Reference van Lawick-Goodall1968). Gradually, newborns are introduced into the social network of the mother (Berman Reference Berman1982). During this period, the mother actively monitors her infant's play sessions and often modifies their content (Power Reference Power2000). The direct, active interventions of the mother decrease as the infant learns to self-regulate and acquires social and emotional competence (Govindarajulu et al. Reference Govindarajulu, Hunte, Vermeer and Horrocks1993; Pellis & Pellis Reference Pellis and Pellis2009). The provision of positive or negative reinforcements by the mother is a form of evaluative feedback (sensu Kline) because it leads to the appropriate management of play by infants.

Over time, the infant's sphere of play extends outward to other kin adults and, in tolerant social systems, even to non-kin adults (Ciani et al. Reference Ciani, Dall'Olio, Stanyon and Palagi2012). In despotic societies, cultural and social knowledge gained by playing is therefore mainly vertical, limited to close kin, whereas in tolerant societies, transmission can also spread horizontally, involving unrelated individuals. Moreover, because unrelated subjects can use different play schemes and modules compared to related subjects, social tolerance enhances the diffusion of behavioral and cultural innovation (Fagen Reference Fagen, Pereira and Fairbanks1993; Huffman et al. Reference Huffman, Leca, Nahallage, Nakagawa, Nakamichi and Sugiura2010). Hence, play is an engine fostering a positive feedback linking tolerance to teaching and learning.

According to Kline, teaching includes specific communicative capacities as behavioral markers or ostensive cues. Similarly, play requires intensive, constant exchanges of ostensive signals and can be actively stimulated. Nonhuman primate research is rich in examples of the capacity of mothers and adults to adapt their communicative schemes to infants. When addressing infants of other females, macaque females (Macaca mulatta) use specific vocalizations to communicate their benign intent, a form of metacommunication (“vocal motherese”; Whitham et al. Reference Whitham, Gerard and Maestripieri2007). Gorillas (Gorilla gorilla) use a higher rate of repetitions and sequences of tactile gestures (Luef & Liebal Reference Luef and Liebal2012) when playing with infants. This form of “nonvocal motherese” sustains the development of infant gorillas in the learning process of nonvocal signals, and meets the criteria defining direct active teaching (sensu Kline).

As Kline notes, “the prevalence of teaching in humans and other animals is a contentious issue” (sect. 4.1, para. 1). However, the existence of play in all human societies and across mammalian species is not. This lends credence to our hypothesis that play may be one of the most basal building blocks from which human learning and teaching evolved.