Integrating disparate areas of science is a noble and often productive cause, and Kline's plea in the target article – to combine the three approaches to teaching, namely functionalist, mentalistic, and culture-based – needs serious consideration. That said, the functionalist approach proposed 22 years ago (Caro & Hauser Reference Caro and Hauser1992) in part grew out of an attempt to break away from the burden of having to show that nonhumans could attribute mental states to conspecifics (Pearson Reference Pearson1989). Mental state attribution requires establishing intent to facilitate learning in another subject, and a theory of mind, and it thereby severely constrains searches for different types of teaching in nonhumans. The focus of culture-based teaching that centers on what happens in Western classrooms is, by definition, of little help in explaining how knowledge is transmitted within the huge diversity of human societies or nonhuman species. Therefore, blending approaches and hoping that something interesting will emerge, as outlined at the start of Kline's article (see also Byrne & Rapaport Reference Byrne and Rapaport2011) is unlikely to be productive.

If one is interested, as I am, in the ecology and evolution of behavior, including teaching, one needs to establish a comparative data-base of different forms of teaching, rather than focusing on a single high-bar definition. Then, these need to be related to different ecological and social variables, and examined taxonomically. Using three criteria for demonstrating teaching – namely, teachers modify their behavior only in the presence of a naïve observer; the behavior incurs costs or no immediate benefit; and, as a result, pupils acquire knowledge or skills more rapidly or efficiently than they would otherwise, or that they would not have learned at all (Caro & Hauser Reference Caro and Hauser1992) – enormous progress has been made in documenting the incidence of different forms of teaching (Hoppitt et al. Reference Hoppitt, Brown, Kendal, Rendell, Thornton, Webster and Laland2008; Thornton & Raihani Reference Thornton and Raihani2008) and its taxonomic distribution (Thornton & Raihani Reference Thornton and Raihani2008). For example, we now know that teaching must have evolved multiple times and in several different forms (Franks & Richardson Reference Franks and Richardson2006; Raihani & Ridley Reference Raihani and Ridley2008; Thornton & McAuliffe Reference Thornton and McAuliffe2006). Furthermore, some progress has also been made in trying to understand the environmental conditions under which teaching is likely to be observed in nature (Richardson et al. Reference Richardson, Sleeman, McNamara, Houston and Franks2007; Thornton Reference Thornton2008; Thornton & Raihani Reference Thornton and Raihani2010).

I do not believe that these advances could have occurred if teaching was shackled by mentalistic or culture-based definitions, because many phenomena that are interesting (to a biologist) would have been dismissed as being unworthy of study. Kline concurs in her closing section that “researchers should not limit the study of teaching to species with forms of mind-reading or theory of mind, because it is not an a priori necessity for behavior that functions as teaching to evolve” (sect. 7, para. 3).

Can the study of teaching in different human cultures benefit from the functionalist paradigm in nonhumans? The second part of Kline's target article suggests that it can, and she supplies many empirical examples from humans in support of this proposal. Anthropologists might indeed benefit from systematically documenting the extent to which different forms of social learning are seen in various societies, and then relate these to patterns of kinship, subsistence activities, and ecological variables (e.g., Hewlett et al. Reference Hewlett, Fouts, Boyette and Hewlett2011; Tehrani & Riede Reference Tehrani and Riede2008).

Caro and Hauser (Reference Caro and Hauser1992) wrote “we are convinced that the only way to make progress in this area is to first provide a definition of teaching which can, and undoubtedly will, be modified as empirical data accumulate” (p. 152). Others have indeed expanded on the functionalist definition of teaching (see Hoppitt et al. Reference Hoppitt, Brown, Kendal, Rendell, Thornton, Webster and Laland2008; Thornton & Raihani Reference Thornton and Raihani2008), but these departures never reverted to definitions centering on intent or Western-style classroom teaching. In the target article, Kline is similarly advocating a broader approach to teaching in humans by including opportunity provisioning, stimulus or local enhancement, evaluative feedback, and direct active teaching. Once systematic qualitative, semi-quantitative, and quantitative data on these and other forms of social transmission in humans accumulate, anthropologists may be able to uncover the ecological and social drivers of various forms of teaching in the same way as their biological colleagues are currently doing (e.g., Whiten Reference Whiten2011), and even relate different forms of social transmission, including teaching, to cultural phylogenies (Steele et al. Reference Steele, Jordan and Cochrane2010).

Questions that include whether progressive teaching (where teachers adjust their behavior as pupils' skills improve) occurs with a small number of pupils who can be monitored, whether teaching is principally found in solitary activities where there are few opportunities to learn by imitation, or whether teaching usually occurs where the costs of teaching are low in terms of lost opportunities for teachers, or where the costs of mistakes are high in terms of injury for pupils, can all be asked by anthropologists just as they can by field biologists (Thornton & Raihani Reference Thornton and Raihani2010). Maybe commonalities about the ecology and even evolution of teaching in human societies and nonhuman species will emerge, maybe not. These are empirical issues. Kumbaya-style cross-disciplinary harmonizing will not yield progress, but systematic documentation of behavior based on cost-benefit analyses without explicit recourse to intent or classroom teaching just might.