The authors make a very good point in demonstrating the fundamental nature of knowledge representation in humans. It has older evolutionary origin than that of belief representation, and that explains why nonhuman primates can do the first but fail to do the second. But is it not a contradiction to argue, on the one hand, that knowledge representation, in so far as it can be seen as a basic cognitive competence, is not distinctive of the human species and, on the other, that what we normally see as the most distinctive characteristic of the human species, which is the capacity to accumulate cultural knowledge, originates in that very same competence? If this is so, one could legitimately wonder why cumulative cultural knowledge is not much more widespread among nonhuman primates than what seems to be the case (Tennie, Call, & Tomasello, Reference Tennie, Call and Tomasello2009; Whiten, Reference Whiten2017).

The authors only mention the accumulation of cultural knowledge at the end of the paper, in section 6.2.1, and they do not elaborate the reasons why they confidently state that “Although the ability to represent others' beliefs may indeed turn out to be unique to humans and critically important for some purposes, it does not seem to underwrite humans' capacity for the accumulation of cultural knowledge.” However, this is undoubtedly a key question for all the sciences of human behavior. A priori, one could plainly state that knowledge representation, rather than belief representation, is instrumental to the accumulation of cultural knowledge for the very simple reason that it is “knowledge” what we accumulate, not “beliefs.” Does that mean that understanding beliefs is irrelevant in the process of social learning that leads to the accumulation of cultural knowledge?

Belief representation, the authors concede, is relevant for predicting other people's behavior, but it is knowledge, and not belief, “that allows us to represent others as reliable guides to the actual world” (6.1). This is undoubtedly true in a rather obvious sense; but it can also be misleading, for it glosses over the process of social learning as it takes place in all known human societies and that enables any apprentice to acquire knowledge from his or her teacher (Sterelny, Reference Sterelny2012). Let me illustrate this with a very simple example. If I want to know how a computer works, I may ask a computer scientist about it. Quite obviously, I am interested in the computer scientist's knowledge about computers, not about her beliefs. But the point I wish to make is that I shall only have access to that knowledge if I am able to understand her beliefs (Salazar, Reference Salazar2018, pp. 37–62).

There is ample evidence that the process of social learning among humans is not simply learning from others, but it is normally conducted within some form of pre-existing social bond (Boyd & Richerson, Reference Boyd and Richerson1985; Cavalli-Sforza & Feldman, Reference Cavalli-Sforza and Feldman1981; Henrich, Reference Henrich2015; Kline, Reference Kline2015; Nielsen, Reference Nielsen2008; Zuidema, Reference Zuidema2002). More specifically, when social learning entails the transmission of socially shared forms of knowledge, what we normally define as “culture,” social learning can only take place when some culturally significant form of social relationship links teacher and apprentice. For the majority of human societies, these social relationships are normally kinship relationships and, more specifically, family relationships, for it is from those that the first and most elementary parts of one's cultural knowledge are to be acquired (Demps, Zorondo-Rodríguez, García, & Reyes-García, Reference Demps, Zorondo-Rodríguez, García and Reyes-García2012; Hewlett & Cavalli-Sforza, Reference Hewlett and Cavalli-Sforza1986; McElreath & Strimling, Reference McElreath and Strimling2008). This basic nucleus of kinship relations will later be supplemented by other kinds of relationships in different ways. WEIRD (western, educated, industrialized, rich, and democratic) societies are somewhat unique in the sense that they have reduced the social relationship between teacher and learner to the (relatively) impersonal bond created in institutional schooling. However, even when there is some form of selectivity (Bentley & O'Brien, Reference Bentley and O'Brien2011), cultural knowledge is very rarely transmitted between anonymous individuals (cf. Osiurak & Reynaud, Reference Osiurak and Reynaud2020).

But why should that be the case? One might be tempted to argue that those networks of social relationships provide a sort of external framework within which “real” knowledge can circulate, but they do not really affect the nature of that knowledge in any substantial way and, crucially, do not transform it into “mere beliefs.” Let me show why this cannot be a valid assumption by going back to the simple example of the teacher – computer scientist. The knowledge I am likely to obtain from her will certainly be a partial knowledge about how the computer actually works – otherwise, I would become a computer scientist myself. But, given my ignorance about computers, there is no way I can have access to that knowledge if I have not previously understood what she believes to be the case about the computer and, specifically, if I do not trust her (Csibra & Gergely, Reference Csibra, Gergely, Munakata and Johnson (2006; see Hewlett, Fouts, & Boyette, Reference Hewlett, Fouts and Boyette2011). In other words, before getting knowledge from any teacher, I have to believe in that teacher and share her intentionality, so that my knowledge becomes a “dialogic cognitive representation” (Tomasello, Carpenter, Call, Behne, & Mol, Reference Tomasello, Carpenter, Call, Behne and Mol2005). In order to acquire the objective knowledge about the world that will enable me to make use of my computer, I have to understand what goes on in the mind of the computer scientist that is teaching me, that is, understand her beliefs so I can end up thinking “through her mind” (Veissière, Constant, Ramstead, Friston, & Kirmayer, Reference Veissière, Constant, Ramstead, Friston and Kirmayer2020). This is what identifies cultural learning as a specific form of social learning (Tomasello, Kruger, & Ratner, Reference Tomasello, Kruger and Ratner1993).

To conclude, from an objective point of view, the accumulation of cultural knowledge does effectively entail knowledge representation. But subjectively, that accumulation is only possible through belief representation. Culture is knowledge acquired from a subject, not from the world, hence only humans' capacity to understand other minds as “subjects in the world” permits its assimilation. And it is by understanding another subject's beliefs that I can assimilate her knowledge and, consequently, I can add up her knowledge to mine, that is, accumulate.