In an ambitious review, Kline aims to integrate literature across multiple species and theoretical perspectives. She develops a comprehensive taxonomy of teaching behaviors, arguing for distinctions to be made based on the adaptive value of the teaching behavior. I am sympathetic to her goal of developing a framework that can be used across sub-disciplines. However, focusing solely on the teacher – or on the first act of teaching between the teacher and learner – defines teaching as a rather one-sided process. I argue that a broader inspection of teaching requires a more thoughtful discussion of learning, for three reasons.

First, although Kline explicitly defines teaching according to the three different theoretical perspectives, and again according to the six different adaptive problems, she fails to focus on the to-be-learned behavior. Indeed, in some places, the learned behavior is situation-specific, whereas in others the information to be learned allows the learner to generalize to new contexts. For example, Kline highlights ants' ability to signal the location of food to a naïve learner through tandem running (Franks & Richardson Reference Franks and Richardson2006; Richardson et al. Reference Richardson, Sleeman, McNamara, Houston and Franks2007). This example certainly fulfills Kline's definition of teaching, as well as the teaching definition widely used in studies of animal behavior (Caro & Hauser Reference Caro and Hauser1992). However, guiding a naïve learner to a food destination only helps that learner return to that specific destination – and not to other destinations in general. Such an act might be considered teaching, but it is also consistent with more general prosocial behavior directed towards conspecifics (Tomasello Reference Tomasello2009), which often is directed towards a specific instance.

For an act to be considered teaching, learning must be present, and the information to-be-learned must be generic and representational. Such a definition of learning is consistent with how learning and teaching have been described in developmental science (e.g., Csibra & Gergely Reference Csibra and Gergely2009; Gelman et al. Reference Gelman, Ware, Manczak and Graham2013; Strauss et al. Reference Strauss, Ziv and Stein2002), allowing teaching behaviors to be uniquely distinguishable from more general prosocial helping interactions (e.g., Warneken Reference Warneken2013), as well as learning behaviors to be distinguishable from imitation (Lyons et al. Reference Lyons, Young and Keil2007). Based on this framework, several of the adaptive behaviors mentioned in Kline's review might not be classified as teaching (or for that matter, learning).

Second, Kline's model of teaching focuses solely on instances in which the teacher motivates the exchange of information – through ostensive cues (Csibra & Gergely Reference Csibra, Gergely, Munakata and Johnson2006; Reference Csibra and Gergely2011) or stimulus enhancement. What may distinguish human teaching from other species is the learner's ability to signal to the teacher that an intervention is necessary – either through nonverbal signals such as joint attention (e.g., Butterworth & Jarrett Reference Butterworth and Jarrett1991; Tomasello Reference Tomasello, Moore and Dunham1995; Tomasello et al. Reference Tomasello, Kruger and Ratner1993), motoric gestures such as pointing (e.g., Bates Reference Bates1976), or early question-asking abilities (e.g., Chouinard Reference Chouinard2007). Such requests for intervention have implications for cultural and interspecies variation, as well as for the speed of the transmission of information. Indeed, it is plausible that learner-motivated teaching may look very different from a teacher-directed interaction.

Moreover, the interaction between the teacher and learner does not end after the first exchange. Kline's taxonomy certainly allows for this possibility, but I would argue that the follow-up exchanges are the most interesting when exploring the relationship between teaching and learning. Kline rightly notes that even after a teaching exchange, there is likely to be a wide range of possible inferences available to the learner (e.g., Boyer Reference Boyer1998; Sperber Reference Sperber1996). True teaching is contingent teaching – that is, teaching that is adaptive to the learner's changing knowledge state.

Finally, although Kline touches on this somewhat in her discussion of the psychology of teaching, a large body of developmental research suggests that the type of teacher matters for learning (e.g., Harris Reference Harris2012; Harris & Corriveau Reference Harris and Corriveau2011; Sobel & Kushnir Reference Sobel and Kushnir2013). A consideration of teacher qualities should go beyond questions of honesty, to focusing on the learners' understanding of the knowledge or expertise of the teacher (Keil et al. Reference Keil, Stein, Webb, Billings and Rozenblit2008; Mills & Keil Reference Mills and Keil2004) as well as their recognition that the teacher is a member of their cultural group (Corriveau et al. Reference Corriveau, Min, Kurkul, Einav and Robinson2014; Richerson & Boyd Reference Richerson and Boyd2005). Moreover, most of the examples presented by Kline involve a more senior member teaching a naïve learner, although knowledge is also transmitted horizontally, as well as vertically (e.g., Flynn & Whiten Reference Flynn and Whiten2008, Reference Flynn and Whiten2012). More research is needed to determine how children's learning from and teaching to peers might be different from their interacting with adults (Wood et al. Reference Wood, Kendal and Flynn2012).

Overall, although Kline has done a commendable job in integrating literature from animal behavior and developmental science, more thought is needed in placing teaching behaviors in the context of learning.