Chater and Christiansen’s target article makes a welcome and overdue argument. Against five decades of cultural evolutionary theorizing built on analogies with Darwinian natural selection – blind variation, selective retention, and individual transmission – they argue that cultural change is intelligent, directed, joint, and coordinative. Their reconceptualization of culture as creation rather than replication, and their systemic vision of culture as spontaneous order, mark a substantive reorientation of the field.
But the framework rests on a metaphor whose explanatory weight the authors do not fully bear. They borrow François Jacob’s (Reference Jacob1977) image of evolution as tinkering: an opportunistic, local, recombinative process operating on whatever materials lie to hand. The metaphor is powerful, but in Jacob’s original setting, it is not free-standing. Biological tinkering is intelligible because it sits inside a theory – the modern evolutionary synthesis – that specifies the criterion by which some recombinations are retained and others discarded: inclusive fitness. The bricolage of evolution is not a random walk through the space of biological forms; it is constrained at every step by differential reproduction. Remove that constraint and Jacob’s metaphor collapses into mere shuffling. It is fitness that makes the bricolage interpretable.
In the social tinkering account, the analogous criterion is absent. Cultural tinkering is described as directed, agents as intelligently choosing, and innovations as solutions to the problem of the moment. But what counts as a problem, what counts as a solution, and which solutions persist while others vanish are taken as given. The authors offer a cognitive psychology of how tinkering proceeds (joint reasoning, virtual bargaining, shared intentionality, and focal-point coordination) but leave open the prior question of what selects among its outputs. Cultural tinkering, in their telling, is Jacob without the synthesis.
Two explanatory tasks become tractable only once a fitness criterion is restored. First, persistence: among the countless innovations agents produce, which become entrenched and which fade? The authors invoke entrenchment of precedents, but this leaves open why some precedents entrench while others die in a single generation. Solutions persist when they reliably enhance the fitness of those who adopt them – either directly, by improving access to resources, mates, or alliances, or indirectly, by becoming necessary conditions for participation in cooperative networks. Without that constraint, persistence is left as a brute fact.
Second, convergence. Cultural systems converge on structurally similar arrangements wherever conditions are similar: subsistence ecology predicts kinship structure, pathogen prevalence predicts conformity and in-group orientation, resource defensibility predicts inheritance regimes, ecological precarity predicts ascetic values, and resource abundance predicts the cultural elaboration of romantic love (Baumard et al., Reference Baumard, Huillery, Hyafil and Safra2022; Baumard & André, Reference Baumard and André2025). These regularities are systematic, replicable, and well documented. Pure tinkering cannot generate them: clever local recombination has no reason to converge on the same outputs across unrelated societies. An adaptationist account does: shared evolved goals, pursued under similar ecological constraints, converge on similar solutions.
Importing fitness into a theory of cultural change does not require the strong modularist commitments Chater and Christiansen rightly criticize (§4.2). Their concern with rigid Pleistocene modules rests on a confusion between what evolved psychology specifies and how it operates. What evolution specifies are goals – the functional outcomes that promote fitness: harm avoidance, resource acquisition, mate attraction, parental investment, alliance formation, status, and reputation. What evolution does not specify, and could not have specified, are the particular means by which those goals are pursued. Means are read from the local environment in real time.
Across the living world, this is what evolved adaptations look like. The vertebrate immune system does not store fixed antibodies; it generates variants and amplifies the ones that bind, with binding success serving as the local proxy for the underlying fitness goal of pathogen neutralization. A satin bowerbird does not execute a bower template; it builds, evaluates, and rearranges decorations using whatever its environment offers in pursuit of the fitness goal of mate attraction. Trees do not unfold from blueprints; they grow by responding to local light, gravity, and obstacles in pursuit of the fitness goal of photosynthetic capture. In each case, evolution specifies the end and a generative system, leaving the means to be assembled from local materials.
Human behavior shows the same architecture, only more so. The fitness goal of social standing is pursued through hunting prowess in one ecology and through scientific publication in another. The goal of mate attraction is pursued through skillful flute-playing in one society and through high-status credentials in another; the goal of harm avoidance is through fear of snakes in ancestral environments and fear of cars and electrical outlets in modern ones. None of these means is evolutionarily ancient. All of them recruit ancient motivational systems whose ends are conserved while their means are reassembled to fit the local ecology. The flexibility Chater and Christiansen attribute to social tinkering is not in tension with an adaptationist explanation – it is what an adaptationist explanation predicts, once one recognizes that adaptations specify goals and leave means open.
This is the deeper revision the social tinkering framework still owes us. Tinkering describes a dynamic – cumulative, coordinative, and opportunistic recombination of means. Fitness, channeled through local ecologies, supplies the goals and the criterion that selects among the means generated. Without it, social tinkering describes how cultural innovation proceeds but cannot say which innovations survive. With it, Jacob’s metaphor is preserved in full: not just the bricolage, but what the bricolage is for.
You have
Access
Chater and Christiansen’s target article makes a welcome and overdue argument. Against five decades of cultural evolutionary theorizing built on analogies with Darwinian natural selection – blind variation, selective retention, and individual transmission – they argue that cultural change is intelligent, directed, joint, and coordinative. Their reconceptualization of culture as creation rather than replication, and their systemic vision of culture as spontaneous order, mark a substantive reorientation of the field.
But the framework rests on a metaphor whose explanatory weight the authors do not fully bear. They borrow François Jacob’s (Reference Jacob1977) image of evolution as tinkering: an opportunistic, local, recombinative process operating on whatever materials lie to hand. The metaphor is powerful, but in Jacob’s original setting, it is not free-standing. Biological tinkering is intelligible because it sits inside a theory – the modern evolutionary synthesis – that specifies the criterion by which some recombinations are retained and others discarded: inclusive fitness. The bricolage of evolution is not a random walk through the space of biological forms; it is constrained at every step by differential reproduction. Remove that constraint and Jacob’s metaphor collapses into mere shuffling. It is fitness that makes the bricolage interpretable.
In the social tinkering account, the analogous criterion is absent. Cultural tinkering is described as directed, agents as intelligently choosing, and innovations as solutions to the problem of the moment. But what counts as a problem, what counts as a solution, and which solutions persist while others vanish are taken as given. The authors offer a cognitive psychology of how tinkering proceeds (joint reasoning, virtual bargaining, shared intentionality, and focal-point coordination) but leave open the prior question of what selects among its outputs. Cultural tinkering, in their telling, is Jacob without the synthesis.
Two explanatory tasks become tractable only once a fitness criterion is restored. First, persistence: among the countless innovations agents produce, which become entrenched and which fade? The authors invoke entrenchment of precedents, but this leaves open why some precedents entrench while others die in a single generation. Solutions persist when they reliably enhance the fitness of those who adopt them – either directly, by improving access to resources, mates, or alliances, or indirectly, by becoming necessary conditions for participation in cooperative networks. Without that constraint, persistence is left as a brute fact.
Second, convergence. Cultural systems converge on structurally similar arrangements wherever conditions are similar: subsistence ecology predicts kinship structure, pathogen prevalence predicts conformity and in-group orientation, resource defensibility predicts inheritance regimes, ecological precarity predicts ascetic values, and resource abundance predicts the cultural elaboration of romantic love (Baumard et al., Reference Baumard, Huillery, Hyafil and Safra2022; Baumard & André, Reference Baumard and André2025). These regularities are systematic, replicable, and well documented. Pure tinkering cannot generate them: clever local recombination has no reason to converge on the same outputs across unrelated societies. An adaptationist account does: shared evolved goals, pursued under similar ecological constraints, converge on similar solutions.
Importing fitness into a theory of cultural change does not require the strong modularist commitments Chater and Christiansen rightly criticize (§4.2). Their concern with rigid Pleistocene modules rests on a confusion between what evolved psychology specifies and how it operates. What evolution specifies are goals – the functional outcomes that promote fitness: harm avoidance, resource acquisition, mate attraction, parental investment, alliance formation, status, and reputation. What evolution does not specify, and could not have specified, are the particular means by which those goals are pursued. Means are read from the local environment in real time.
Across the living world, this is what evolved adaptations look like. The vertebrate immune system does not store fixed antibodies; it generates variants and amplifies the ones that bind, with binding success serving as the local proxy for the underlying fitness goal of pathogen neutralization. A satin bowerbird does not execute a bower template; it builds, evaluates, and rearranges decorations using whatever its environment offers in pursuit of the fitness goal of mate attraction. Trees do not unfold from blueprints; they grow by responding to local light, gravity, and obstacles in pursuit of the fitness goal of photosynthetic capture. In each case, evolution specifies the end and a generative system, leaving the means to be assembled from local materials.
Human behavior shows the same architecture, only more so. The fitness goal of social standing is pursued through hunting prowess in one ecology and through scientific publication in another. The goal of mate attraction is pursued through skillful flute-playing in one society and through high-status credentials in another; the goal of harm avoidance is through fear of snakes in ancestral environments and fear of cars and electrical outlets in modern ones. None of these means is evolutionarily ancient. All of them recruit ancient motivational systems whose ends are conserved while their means are reassembled to fit the local ecology. The flexibility Chater and Christiansen attribute to social tinkering is not in tension with an adaptationist explanation – it is what an adaptationist explanation predicts, once one recognizes that adaptations specify goals and leave means open.
This is the deeper revision the social tinkering framework still owes us. Tinkering describes a dynamic – cumulative, coordinative, and opportunistic recombination of means. Fitness, channeled through local ecologies, supplies the goals and the criterion that selects among the means generated. Without it, social tinkering describes how cultural innovation proceeds but cannot say which innovations survive. With it, Jacob’s metaphor is preserved in full: not just the bricolage, but what the bricolage is for.
Financial support
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
Competing interests
None.