This commentary (whose title follows Morin, Reference Morin2016) focuses on definitional issues raised by the two more basic neologisms among the three Singh introduces. He claims that subjective selection is the most important process in producing the cultural manifold. The cultural manifold comprises the complete set of super-attractors. I begin with “subjective selection” and move on to “super-attractor.”

The term “subjective selection” invites two quite different readings in the context of an article on cultural evolution. First, one might think it refers to any evaluation made by an individual agent based on their subjective desiderata. Second, one might think it picks out a specific form of cultural selection, where cultural selection is understood as a population-level process analogous to natural selection. Singh’s comments indicate he is comfortable with both readings. In line with the first interpretation, he suggests that subjective selection can take place on the shortest of timescales when “a person has a goal, they invent a solution, they evaluate the extent to which that solution achieves the goal.” In line with the second interpretation, he suggests psychological processes “cumulatively give rise to population-level cultural evolutionary dynamics, including subjective selection.” The two interpretations come apart, however, in various ways (Lewens, Reference Lewens2004; Walsh, Lewens, & Ariew, Reference Walsh, Lewens and Ariew2002). For example, according to standard conceptions, natural selection relies on token entities having differential reproductive success. In other words, natural selection, along with forms of cultural selection that are meant to be analogous to it, requires parent-offspring relations between the entities undergoing selection. Singh, however, mentions “shared cultural ancestry” as an alternative to his own cultural selectionist explanation of shared similarities across cultures. It is clear enough that here he is thinking of “shared cultural ancestry” in very broad-sweep terms, which might trace all the way back to a kind of cultural “African Eve.” But, on most standard views, unless some more local shared ancestry is in play – unless, that is, there are cultural lineages to be traced – it makes no sense to think of cultural selection at work. This is where the ambiguity in the two notions of “subjective selection” becomes significant: one does not show that subjective (cultural) selection is at work solely by showing the importance of individual agents’ subjective selections.

This brings me to super-attraction. Singh’s primary goal is to explain cultural phenomena that are robust, in the sense that they develop “around the world and in societies of diverse subsistence strategies and social organization,” and also complex, in the sense that they “are not merely a collection of discrete components; rather, they relate dynamically to produce new, emergent properties or functions.”

By labelling these robust and complex cultural phenomena “super-attractors,” Singh does not merely draw attention to their existence as worthwhile explanatory targets. A super-attractor is a scaled-up cultural attractor in Sperber’s (Reference Sperber1996) sense. “Attraction,” so understood, is a notion that earns its keep because it points – via more precisely-specified “factors of attraction” – towards an understanding of the reliable recurrence of resembling cultural representations without bit-by-bit copying (Sperber, Reference Sperber and Aunger2000). Imagine two agents, a and b, who are confronted with the same type of situation and who need to respond to it in some way. Suppose a produces a solution of type S; b notices a and also produces a solution of just the same type S. This could be because b carefully copies each element of a’s solution bit-by-bit. But Sperber points out that it could also be because a and b share cognitive biases in how they approach problems, or because they face similar constraints in terms of raw materials they can use, or because the problem itself has just one very obvious functional solution, or because of many such “factors of attraction” acting individually or in concert (Buskell, Reference Buskell2017; Scott-Philipps, Blancke, & Heintz, Reference Scott-Phillips, Blancke and Heintz2018). In using the term “super-attractor,” Singh implicitly conveys – although he does not explicitly argue for it – that whatever explains the prevalence of robust and complex cultural phenomena, it is not bit-by-bit copying.

There is no contradiction in combining cultural selection and (super-)attraction in just the manner Singh alludes to. Imagine one is trying to explain the prevalence of a solution S in a population. It could be that a single individual initially discovers S alone. A few others casually observe that individual – who becomes, because of this, a cultural “parent” – and their observation inspires them to produce similar solutions through the action of commonly-held cognitive processes coming into contact with similarly-constrained resources. They, too, are then observed by others, and the process iterates until S is dominant. Here, there are reproductive cultural lineages without bit-by-bit copying, and forms of cultural selection and cultural attraction consequently work hand-in-hand (Lewens, Reference Lewens2024). But although this might be how things work sometimes, one can also imagine circumstances where there are no easily identifiable cultural “parents”; instead, S may spread through far more diffuse transmission pathways, perhaps because agents, once again with similar cognitive and material constraints, are repeatedly exposed to a whole series of different models who have developed S. At the limit, one might even imagine an exogenous environmental change that leads many different individuals’ subjective “selections,” again informed by their shared constraints, to favour S across a population. Here, there may be no form of cultural transmission – or cultural selection – at work at all.

Singh’s evidence points squarely towards the importance of subjective goals and preferences in the explanation of cultural phenomena, and in this respect his paper is a success. However, to secure the importance of subjective (cultural) selection, one needs to show that the best way to understand the repeated re-occurrence of resembling cultural phenomena must appeal to proliferating lineages of reproducing cultural items. This requires ruling out an alternative, namely, that widely-held subjective goals and preferences give rise to the repeated re-constitution of resembling cultural representations without any clear lineage formation. I take this to be one of the lessons of cultural attraction theory: if it holds for basic attractors, it holds for super-attractors also.