Earlier, two strategies were sketched for applying DIT to patterns of cultural change: one that directly models them as resulting from basic social-learning mechanisms, and another that focuses on the interrelations between the evolution of learning mechanisms and patterns of cultural change. Then, some grounds were given for being cautious about the explanatory merits of models that implement the first strategy. These do, of course, not entail that advocates of DIT should invest exclusively in the second strategy. In this section, we show one way of implementing the second strategy that is complementary to the first strategy.

This implementation starts with some compelling motivational narratives given by DI theorists. There, attention is drawn to a vital characteristic of human cultures, namely that the great diversity of cultural items, such as hunting practices, technologies and dietary restrictions, have enabled human beings to thrive in virtually every corner of the globe, including inhospitable areas like arctic coasts and inland deserts. Moreover, the motivation continues, these cultural adaptations cannot be the result of individual inventiveness. As Henrich and McElreath (2003) describe vividly in their presentation of the basic structure of DIT, neither advanced technology nor prolonged exposure to the local environment helped Robert Burke’s 1860 expedition to cross the Australian desert. Richerson and Boyd (2005: 46–48) use the similar example of crossing El Camino del Diablo between California and Mexico, the unwelcoming home of the Papago Indians. From these examples, DI theorists conclude that human beings can thrive in various environments, not because they are exceedingly clever adapters individually, but because they have acquired suitable cultural repertoires over many (human) generations.

These motivational narratives set up a particular explanatory project, which shall be called the ‘cumulative-adaptation’ project here. Its explanandum is the acquisition, by groups of humans, of local cultural adaptations such as hunting practices and tool traditions. Its explanans is characterized both positively and negatively: it involves the gradual accumulation of an adaptive cultural repertoire over many generations and it does not involve individual inventions of optimal cultural items.

This project is not marginal to DIT. It specifies the “Why not baboons?” question alluded to earlier, by clarifying in which respects and to what extent human beings are unlike baboons. Earlier, part of the answer was given: human beings are capable of cultural learning, which is characterised by particular transmission mechanisms (‘forces’ or ‘biases’) and which can arise as the result of (multi-level) natural selection on human beings. By this type of learning, human beings are capable of imitating a clever trick after they have observed a conspecific employing it.

However, the narratives just given go beyond noticing this imitative capacity. They point out that humans are capable of retaining clever tricks and gradually modifying them to a level of ingenuity that no individual inventor could ever hope to achieve.14 To illustrate how this offers a more elaborate view on how human beings are unlike baboons, consider cycling. We transmit the cultural trait of cycling over the generations by cultural learning—and substantial time and patience on the teacher’s part, and willingness to cope with frustration and occasional pain on the student’s part. Baboons are incapable of this.15 This means that, even if—which is conceivable—the trick of riding a bicycle is within the grasp of a socially isolated individual learner, the trick would die with its baboon inventor, but is likely to outlive its human inventor. In addition, human beings are unlike baboons in being able to design bicycles, which have thousands of integrated parts and embody technological knowledge of huge complexity. In combination, these two uniquely human features characterize how we are unlike baboons, and thus specify one of the central questions posed by DI theorists.

The cumulative-adaptation project set up by the motivational narrative not only specifies the “Why not baboons” question. It also strongly resembles what Peter Godfrey-Smith (2001) has called explanatory adaptationism, both in its explanandum and in its rejection of a particular candidate explanation. In ‘biological’ explanatory adaptationism, the ‘Big Question’ is to explain the adaptedness of life forms to their environments without appealing to the perfect wisdom of an individual Designer of the life form. The ‘Big Answer’ lies in the cumulative changes wrought by the operation of natural selection on life forms. In DIT, the Big Question is the adaptedness of (groups of) human beings to their local environments, and the Big Answer is supposed to lie in the cultural transmission and gradual accumulation of cognitive capital over many generations, without appealing to the wisdom of an individual designer.

Given the resemblance of this project to explanatory adaptationism, part of the answer to the “Why not baboons?” question might lie in natural-selection explanations on the level of cultural items, as well as on the genetically inherited capacities for cultural learning. The former natural-selection explanations would involve populations of cultural items (say, arrowheads) that are replicated through production, use, maintenance and other human activities. To give a semi-specific example: the shape of arrowheads may change over generations, under the influence of learning mechanisms such as success and prestige bias, and despite the influence of other mechanisms, such as conformity bias. Moreover, small variations are introduced by individual intent or accident. Some of these may be detrimental to the successful use of the item; others may increase its usefulness for present purposes, or may play a role in alternative uses. As a result, generations of arrowheads might arise as adaptations, showing the gradual refinement and diversification that is characteristic of human cultural traditions (Richerson and Boyd 2005: 115).

Not every suite of social learning mechanisms facilitates this pattern of cumulative cultural change. If human learning would, for instance, be dominated by conformity bias, small innovations in design or use would never spread, no matter how advantageous they would be.16 Without success and prestige bias, there would be no retention of the cultural repertoire over multiple generations; without imitation errors and (limited) individual inventiveness, variation in cultural traits would quickly disappear. Yet that certain varieties of these learning mechanisms, in combination, allow adaptive technologies to accumulate might be pivotal in explaining how these (combinations of) varieties evolved by natural selection: the evolutionary benefits of the former reside in enabling the latter. As Richerson and Boyd (2005:7) put it: “the human cultural system arose as an adaptation because it can evolve fancy adaptations to changing environments”.

In this ‘dual-selectionist’ interpretation, DIT involves two coupled selection processes: one that acts on populations of human beings and explains how they have evolved specific learning mechanisms; another that operates on populations of cultural items and explains how they could be optimized to a great diversity of environments. The first part of this explanation cannot be divorced from the second: human beings evolved their unique suite of social learning mechanisms because it enables, under certain conditions and in some cases, the gradual accumulation of cultural capital. Conversely, an explanation of this cultural accumulation cannot be divorced from an explanation of social learning mechanisms.17 This interrelation resembles that between more specific examples of gene-culture co-evolution, such as that between the spread of lactose tolerance and of dairying practices.

Boyd and Richerson occasionally hint at these interrelated selection processes, e.g.: “The single most important adaptive feature of culture is that it allows the gradual, cumulative assembly of adaptations over many generations, adaptations that no single individual could invent on their own.” (Boyd and Richerson 2000, p.424). Here, both the learning mechanisms and the products of these mechanisms are referred to as adaptations—and the “generations” referred to could be of either human beings or cultural items.18 Elsewhere, an intricate interaction between the evolution of complex cultural traditions and the evolution of ever more sophisticated learning mechanisms is suggested: “As the evolving traditions become too complex to imitate easily, they will begin to drive the evolution of still more-sophisticated imitation.” (Richerson and Boyd 2005: 139).

As said at the end of the previous section, cumulative cultural change is a complicated and diffuse phenomenon. This section makes clear which aspect of this phenomenon could be particularly relevant as an explanandum for DIT. Moreover, its emphasis on the role of social learning mechanisms in acquiring cultural traits appears to provide DIT with the conceptual equipment for giving this explanation. Still, making explicit this explanatory project reveals a hiatus in current modelling efforts. To see why, consider again Henrich’s (2004) tool-loss model, discussed in the previous section. Contrary to the motivational narratives summarized at the start of this section, Henrich’s model assumes unlimited individual inventiveness: every individual draws a skill-level value from the interval [0,∞), albeit with a probability distribution that depends partly on the skill-level value of the individual’s mentor and that is skewed towards a decrease in skill level. Therefore, even if Henrich’s model is taken as explaining cumulative cultural change (rather than the loss of cultural complexity), it does not feature the kind of cumulative change that is presented in the motivational narratives.19

This interpretation therefore identifies a considerable burden of proof resting on future DI work. Perhaps this burden cannot be shouldered. In culture, the Big Answer of explanatory adaptationism might be false, just as the Big Question of the cumulative-adaptation project might be mistaken. However, a philosophical analysis of a theory that advises its advocates against endorsing a possibly false hypothesis is clearly too risk-aversive. Engaging in the cultural-adaptation project seems a risk worth taking: it might increase DIT’s explanatory value, and would give it a pivotal role in an evolutionary synthesis of the social sciences (Mesoudi et al. 2006; Mesoudi 2011b: Ch.10).