Chater and Christiansen’s social tinkering framework powerfully explains how directed, local innovations accumulate into complex cultural systems. However, the framework undertheorizes how the same cognitive and coordinative machinery systematically generates and entrenches maladaptive practices – magic, divination, taboo systems – that are socially coordinated and culturally cumulative yet instrumentally ineffective. I argue that retrospective causal reasoning under uncertainty, combined with the coordinative pressures central to social tinkering, predictably produces elaborate but materially costly cultural complexes.

Chater and Christiansen explain cultural complexity as emerging through local social tinkering. We argue that this account can be complemented by two minimal principles: sequence representation as a cognitive prerequisite for cumulative culture and cultural-systems dynamics as a formal account of how interacting traits stabilize, constrain, and reshape later tinkering. Together, they explain how simple mechanisms generate complex cultural organization.

Chater and Christiansen argue that cultural complexity arises through social “tinkering.” While compelling, this framework remains mechanistically underspecified. Integrating recent empirical and theoretical advances, this commentary proposes that cultural complexity emerges from interactions among transmission dynamics, cognitive constraints, social network structure, and environmental pressures and calls for a formalized, testable multilevel theory.

We supplement and challenge Chater and Christiansen’s (C&C) proposal of social tinkering by drawing on cognitive integration. Learning and mastering cognitive practices transforms the cognitive capacities of agents facilitating novel cognitive capacities. This entails reconfiguring C&C’s proposed six stages as loops rather than a strict ordering.

While the target paper’s proposal dovetails nicely with prominent perspectives on cultural evolution, there are also substantial gaps. In the commentary, we will argue that the target paper’s proposal does not adequately take the role of cooperative selection into account.

We reframe social tinkering from the perspective of complex dynamical systems. The progression between stages of cultural complexity reflects phase transitions in which new innovations develop in clusters of interconnected people, separated in social space from clusters that resist the new. Complete transition is rare; systems instead stabilize in semi-stable states representing a mix of the old and the new.

While interaction and problem-solving shape cultural structure, as the social tinkering framework suggests, stability and erosion too are possible and equally significant outcomes. Navigation and mushroom foraging illustrate how and why sophisticated systems can persist without formal institutions or writing. Recognising diverse trajectories challenges linear stage models and highlights cultural evolution as context-dependent, adaptive, and often intentionally stable.

The hypothesis that human culture emerges through deliberate, decentralized social tinkering rather than random variation is persuasive. However, the hypothesis is ambiguous about the stability of the resulting structures. Expanding scale may increase fragility and normative conflict, and networks of socially constructed coordination spaces may amplify imaginative disconnection from reality. Consequently, system-scale risks arise that exceed the grasp of governable tinkering.

Chater & Christiansen’s social tinkering thesis compellingly explains how cultural complexity emerges from local innovations. However, it underspecifies the role of the arts, treating them as outputs rather than inputs to tinkering. We argue the arts serve three crucial, underappreciated functions: generating variation for recombination, providing consequence-free experimentation spaces, and building the cognitive and social skills that enable sophisticated tinkering.

We are sympathetic to the target article’s emphasis on directed innovation and coordinated problem-solving as the microfoundations of emergent social order. Nevertheless, the authors’ framing of social tinkering as a middle ground between extremes of deliberate design and selection on random variation risks caricaturing reasonable scientific positions and obscuring the question of which processes (one-shot innovation, incremental tinkering, and additional processes like cultural group selection) contribute most to complex cooperative institutions.

The concept of social tinkering is useful, but insufficient to explain the evolution of human cultural complexity due to its overemphasis on cooperative interactions. The importance of conflict, violence, and inequality in the evolution of complex cultures complicates the social tinkering hypothesis. Who gets to tinker, and what determines which products of that tinkering are selected?

Anthropologists have noted, time and again, that cultural norms and institutions often (1) are “functional,” providing context-relevant benefits, (2) convergently evolve, and (3) benefit certain individuals over others. In this commentary, I consider inequality, self-interest, and coercion in complex cultural institutions. I extend Chater and Christiansen’s theory of social tinkering, taking cues from Mercier and Sperber’s argumentative theory of human reason and Singh et al.’s self-interested enforcement hypothesis, to argue that humans are well-adapted “cultural lobbyists.”

Chater & Christiansen argue that cultural complexity emerges from incremental “social tinkering.” We extend this by proposing that collective social niche construction – the self-organized modification of social environments – provides the necessary substrate for such tinkering. By examining adaptive social networks in animal societies, we show how structural complexity can accumulate through feedback loops before explicit rules or linguistic conventions emerge.

The social tinkering model convincingly explains how local coordination processes give rise to shared practices and conventions. However, it remains unclear whether such mechanisms suffice to account for normativity in the strong sense. Using contemporary language models as a comparative case, coordination alone does not appear sufficient to generate normatively binding standards. The transition to normativity likely requires additional conditions – such as public accountability, sanctions, and stable social roles – that merit more explicit integration within the framework.

Stage 2 of Chater and Christiansen’s (henceforth CC) model involves the ability to copy. Our comment concerns not the order of copying in CC’s model, but their time estimate of its origins. CC treat copying as present in apes and implied by early stone tools, and hence (presumably) had by the last common ancestor. Here we argue for a later onset.

The social-tinkering account describes how cultural complexity emerges from directed innovation and social transmission but leaves the cognitive mechanisms underspecified. We highlight how individual innovation can be decomposed into a coordinated set of cognitive processes that leverage structured representations of the environment. These shared representations also turn observation into inference, supporting both richer transmission and richer learning from others. The synergy between directed innovation and social transmission paves the way for broader social tinkering.

Although we agree with the view of cultural evolution as mainly a gradual process, we show that rapid evolution is also supported by the same criteria. Abrupt cultural changes take place through coercion, leadership, institutional interventions, or mass contagion, and these are assisted by the unique cognitive equipment of humans, especially in our globalized era that boosts fast spread.

Chater and Christensen propose a short-sighted tinkerer’s local adjustments are repurposed to support the spontaneous organization of complex cultures. We commend their conceptualization of innovation by recombination but challenge the tinkerer’s myopic focus, which contrasts with humans’ rich capacity for prospective thought. We suggest humans’ farsightedness facilitates their remarkable capacity for complex and diverse cultures.

Chater and Christiansen propose a 6-stage sequence through which social tinkering drives cultural evolution and cultural complexity. We highlight that people’s selectivity of information uptake from social others and top-down social changes results in variations in social tinkering. We argue that the process of cultural evolution and the path to cultural complexity is less linear than proposed.

Our commentators have broadly welcomed the notion of social tinkering, while also offering a range of suggestions, refinements, and potential challenges. In our response, we clarify the concept of social tinkering, sharpen our account of the six stages of cultural evolution, consider whether additional mechanisms are needed to explain the emergence of cultural complexity, and explore a range of possible extensions to the social tinkering framework. We conclude by discussing broader implications and future directions for this research program, including a reconsideration of Rogers’ paradox.