In their target article, Heintz & Scott-Phillips (H&S-P) effectively confirm that the central problem for the emergence of domain-general, open-ended communication (including language) is its evolutionary stability. Their account is one of the rare few that takes this challenge seriously and does not presuppose human communication to be special. H&S-P rightly observe that “all evolved communication systems should be tied to narrow domains of statistical mutual benefit” (target article, sect. 5, para. 4), with human communication being no exception, as it also adheres to this evolutionary, behavioural–ecological constraint. The crucial exception, they propose, is that for humans this narrow domain is the domain of informative intentions – and it is these informative intentions that can in turn be domain general (“virtual domain generality”). This solution has a main advantage of offering an elegant and productive connection between the evolutionary considerations and the explanatory apparatus of cognitive pragmatics. It also has a main downside of pushing the real problem one level deeper: Why is it that the domain of informative intentions happens to afford statistical mutual benefit to humans (including prehistoric hominins), if it does not for anyone else (including our ape cousins)?

Here, H&S-P's answer – “expression can be unleashed in partner choice social ecologies” (target article, sect. 1, para. 3) – is perhaps plausible, but not new and not demonstrated to be superior to existing alternatives. Although the authors do not expressly refer to the biological markets theory by this name (BMT, Noë & Hammerstein, Reference Noë and Hammerstein1995), their solution is essentially a BMT account combined with the idea, originally developed by Dessalles (Reference Dessalles, Hurford, Studdert-Kennedy and Knight1998), of individual reputations hinging on being a relevant communicator. In sum, language and early forms of prelinguistic communication become a device for the honest signalling of one's usefulness in a competitive market of potential cooperative partners. Both those proposals need additional assumptions to make them work. As one example, scaling up BMT to account for the evolution of cooperation in hominins would presuppose starting from already very advanced cognitive–normative capacities for tracking the relative prices of numerous kinds of commodities across different timeframes (Witteveen, Reference Witteveen2021). In turn, reputations based on relevance are vulnerable to the winner-take-all effect, that is, the entire group converging on a single or a few most relevant speakers (contemporarily best illustrated by social media celebrities), unless additional constraints are in place, such as constraints on the accessibility of social links (Dessalles, Reference Dessalles2020). Even more importantly, H&S-P claim that “to a degree that surpasses that of other great apes, [the human] social ecology generates many opportunities for win-win cooperation, and risks of exploitation” (target article, sect. 6, para. 2) – but they stop short of making a convincing argument why and how this should be true of the evolutionary history of our species. That is, what specific “opportunities for win-win cooperation” would distinguish the ancestral evolutionary ecologies of Homo from those of Pan, in a way leading to lineage-specific pressures on cooperative partner choice? The target article somewhat generically mentions “animal hunting, building shelter, maintaining a fire, alloparenting” (target article, sect. 6, para. 6), and then pedagogy, but without developing convincing links to existing evolutionary accounts focusing on such interdependence (e.g., Tomasello, Melis, Tennie, Wyman, & Herrmann, Reference Tomasello, Melis, Tennie, Wyman and Herrmann2012 or Bickerton & Szathmáry, Reference Bickerton and Szathmáry2011; cf. also Beecher, Reference Beecher2021). Finally, H&S-P do not consider probably the most plausible account of the evolutionary emergence of cooperation and prosociality in humans: the cooperative breeding hypothesis (CBH; e.g., Burkart et al., Reference Burkart, Allon, Amici, Fichtel, Finkenwirth, Heschl and Meulman2014). First, CBH explains a very early start of the ontogenetic development of human cooperative dispositions (including dispositions for cooperative communication), something with which interdependence-based accounts struggle (Tomasello & Gonzalez-Cabrera, Reference Tomasello and Gonzalez-Cabrera2017). Second, it has a rare degree of empirical support from extant primates, because cooperative breeding is known to underlie a suite of cooperative behaviours in callitrichid monkeys, including teaching, foodsharing, and joint vigilance (Burkart, Hrdy, & Van Schaik, Reference Burkart, Hrdy and Van Schaik2009).

What we see as a particular advantage of the target article is the choice of its guiding metaphor: unleashing expression (otherwise leashed to narrow domains of statistical mutual benefit). This is an apt metaphor whose heuristic value should be fully appreciated. Standard accounts of language origins typically work with, perhaps more intuitive, metaphors of transformation or even accretive development; that is, they focus on the communicative–cognitive skills that need to be added to (or reworked from) generalised ape cognition, or on the accretive increases in the complexity of communication systems. Reframing the question from developing into unleashing expression prioritises the ultimate-level perspective of behavioural ecology and signalling theory over the proximate-level perspective of implementation, and it foregrounds a behavioural–ecological “leash” as the main factor that is responsible for the rarity of language-like systems in nature. This approach is correct, because selection could not possibly promote the cognitive skills for language-like communication without first removing the evolutionary constraints for its emergence. In similar spirit, we have previously proposed a different metaphor, of domain-general communication resting on a “platform of trust,” without which such domain-general communication must collapse (Wacewicz & Żywiczyński, Reference Wacewicz and Żywiczyński2018). However, an additional advantage of “unleashed expression” is the implication that expression is something not inherently missing but rather very much latently present in nonhuman apes. This seems to be in line with much recent primatological research, which suggests that “the most important limitation to the evolution of human-like language was indeed the motivation to share information, rather than the cognitive ability to do so” (van Schaik, Reference van Schaik2016, p. 423). Finally, we strongly support “expression” as an explanatory target that extends beyond language to other forms of rich communication and activities such as teaching. We agree with the authors that despite their superficially very different manifestations, language, pantomiming, declarative pointing, or teaching all form a natural kind, as all of them are types of domain-general “information donation,” subject to the same behavioural–ecological constraints.