Chater and Christiansen offer a compelling account of how cultural structure can grow out of repeated human interactions and problem‑solving. We enthusiastically share their interest in explaining culture without appealing to overarching plans or fixed trajectories, highlighting how small‑scale actions can give rise to large‑scale patterns. Our comments are intended as a constructive extension of this framework, focusing not on its core motivation but on how its staged structure may steer interpretation towards progressive cumulation rather than towards the full range of cultural trajectories. The progression from individual learning to norms, language, institutions, writing, and science can – even if not intended as such – easily be perceived as a single path forward, nurturing implicit empirical expectations.

One such expectation might be that, once language is in place, social tinkering will accelerate and augment cultural complexification. Chater and Christiansen hint at this possibility when they describe feedback between large brains, language, and cultural innovation. This expectation, however, sits uneasily with the evolutionary timing of these capacities. Both large brains and some form of language appear to predate our species (Dediu & Levinson, Reference Dediu and Levinson2018), yet no other hominin, and not every modern human group, brought forth the cultural systems described in the later stages of the model. Instead, while cognitive constraints may be species‑wide, the emergence of innovations rather depends on culturally specific conditions and ecologies (Bender, Reference Bender2020; Bender et al., Reference Bender, Straffon, Gatewood and Beller2024).

Numerous stable societies are highly sophisticated and locally adaptive in the absence of writing or formal institutions (Bender & Beller, Reference Bender and Beller2019; Henrich, Reference Henrich2015). In these cases, stability is not a “failure to progress” but a genuine and recurring cultural outcome. This applies both to contemporary societies that maintain complex cultural repertoires and to extinct Homo species whose material cultures were structured yet showed little cumulative change over long periods (d’Errico et al., Reference d’Errico, Doyon, Colagé, Queffelec, Le Vraux, Giacobini, Vandermeersch and Maureille2018; d’Errico & Stringer, Reference d’Errico and Stringer2011; Foley & Gamble, Reference Foley and Gamble2009). Cultural complexity can also decrease and accumulated knowledge can disappear, as has happened for various artisanal techniques, from Roman concrete and Damascus steel blades to stained glass (Dell’Acqua, Reference Dell’Acqua, Carson Pastan and Kurmann-Schwarz2019; Elsen, Cizer, & Snellings, Reference Elsen, Cizer and Snellings2013; Verhoeven, Reference Verhoeven1987) or for (folk-)biological knowledge around the world (Atran, Medin, & Ross, Reference Atran, Medin and Ross2004; Wolff, Medin, & Pankratz, Reference Wolff, Medin and Pankratz1999). One reason for such erosion is the uneven distribution of expertise, skills, and tinkering within populations (Gatewood, Reference Gatewood1983, Reference Gatewood2012). Accounts of cultural complexification can therefore not afford to neglect long‑term stability, diversity, and even de-complexification and loss as common outcomes of cultural evolution.

One mechanism that may favour stability over progress is cultural exaptation (signalled in Section 2.4). Rather than arising as the next step in a sequence, many cultural changes emerge through the reuse and reorganisation of existing practices to meet local demands (Colagè & d’Errico, Reference Colagè and d’Errico2020; d’Errico & Banks, Reference d’Errico and Banks2015). Cultural accumulation is evident across deep evolutionary time, but it depends on shifts in transmission strategies and remains uneven across domains and periods instead of being universal or continuous (Colagè & d’Errico, Reference Colagè and d’Errico2025). Features such as writing and formal teaching, too, emerge as context‑dependent reorganisations, not as general endpoints of cultural development (Straffon et al., Reference Straffon, Papa, Øhrn and Bender2023). With its emphasis on directional growth, the six‑stage sequence leaves less room for understanding how reuse and stabilisation shape cultural outcomes.

Consider navigation. Conceptual systems of wayfinding differ markedly across populations despite shared cognitive capacities (Fernandez-Velasco & Spiers, Reference Fernandez-Velasco and Spiers2024), reflecting the stabilisation of locally effective practices rather than differences in general cognition. In many WEIRD contexts, navigation problems are routinely offloaded to external technologies, allowing systems to change rapidly (and cognitive skills to erode) as tools change. In contrast, traditional Pacific Island voyaging depends on learned attention to stars, swells, winds, birds, and other environmental cues, developed through repeated practice in structured social settings (Gladwin, Reference Gladwin1970; Lewis, Reference Lewis1994). Once such practices reliably solve recurrent problems and conditions remain stable (Fogarty, Reference Fogarty2018), further elaboration is neither required nor selected for.

This may be even more pronounced in domains where mistakes carry high costs, such as foraging for hazardous food. Correctly identifying edible mushrooms, for instance, requires fine‑grained perceptual discrimination and risk appraisal – skills learned through culturally structured heuristics and shared practices (Kaaronen, Reference Kaaronen2020). Across populations with broadly similar cognitive capacities, mushroom foraging traditions stabilise around different solutions to the same problem, with some fostering detailed discrimination and others favouring generalised avoidance (Bender & Oterhals, Reference Bender and Oterhals2025; Papa, Reference Papa2025). Where these practices reliably manage risk, they are maintained and transmitted with high fidelity, leaving little pressure for continued modification.

Social tinkering alone does not explain why some cultural systems accumulate while others stabilise or simplify. Social learning often preserves existing solutions rather than transforming them, and adaptive change may favour conservatism or simplification rather than elaboration (Boyd, Richerson, & Henrich, Reference Boyd, Richerson and Henrich2011; Boyd & Richerson, Reference Boyd and Richerson1988). In other domains, such as art, cultural conventions may lack a clear utilitarian target altogether, leaving no obvious direction of cumulative change (Straffon & Tennie, Reference Straffon and Tennie2025).

Finally, the six-stage model misrepresents the dynamics illustrated above by treating distinct processes as successive phases. Individual learning, culturally organised apprenticeship, norm stabilisation, institutionalisation, and external symbolic storage operate at different levels and timescales and are not components of a single developmental trajectory. As the cases of navigation and mushroom foraging show, complex and stable systems can emerge through socially organised learning and transmission without progressing towards formal institutions or writing. Collapsing these heterogeneous processes into stages makes cultural change appear more linear and uniform than it is in reality.

Cultural evolution can proceed along multiple pathways shaped by local conditions and learning arrangements, rather than as movement towards ever more complex forms. Accumulation is only one possible outcome of this process. Other cultural trajectories are equally central to the explanation. Beginning from observed diversity, instead of complexity, helps preserve what is most compelling about the social tinkering approach while better accounting for the uneven patterning of cultural change. Reframed this way, social tinkering may offer a richer account of how cultural systems change, but also of why a great deal of cultural structure remains remarkably stable over time.