R2.1. Contextual variability and peer culture responsiveness

Commentators highlighted that peer cultures are far more pervasive, contextually diverse, and culturally variable than we initially emphasized. Children construct peer cultures within the varied social, economic, and ecological landscapes they inhabit, responding to the specific affordances and constraints of each. A comprehensive study of children’s peer cultures should thus include a wide range of cultural settings, industrialized societies among them.

Greenfield draws attention to the contextual features that contribute to children’s cultural innovations. Her long-term research among Zinacantec Maya communities suggests that children’s creative expression in weaving, once highly constrained by traditional conformity in textile design, expanded dramatically with increased schooling and participation in market economies. She argues that, rather than emerging in all environments, children’s innovative capacities are culturally variable. We agree that not all environments are conducive (or tolerant of) cultural innovations, and indeed that in some contexts, innovations may be actively discouraged, such as among the Maya, where in the early 1970s, to be “different” was to be “bad.” We see this not as a recession of peer culture, but as a reflection of its responsiveness to socioecological variables. Just like adult cultures, peer cultures mirror the affordances and constraints of their environments.

It’s also possible that, in settings where conformity is highly valued, children experiment with alternative practices and ideas within the safety of their peer groups, even if those differences remain invisible or undervalued by adults. These small departures can persist within peer cultures until broader conditions make them legible to the wider community. In the Mayan example, we wonder whether, prior to market integration and schooling changes, children developed novel weaving designs during play activities outside of the purview of adults, even if these were not integrated into formal textile production. If they did, this would provide evidence that peer cultures not only mirror the present but also anticipate future shifts, preserving and refining emergent forms of behavior that may later be recognized as innovations. Importantly, Greenfield points out that it was Mayan adolescents who were the first to use the term “different” in a positive way. We note that across cultures, adolescents often seem most open to adopting market-based innovations. For example, Bolivian Tsimané acquire a majority of their market-based skills in adolescence (Schniter et al., Reference Schniter, Gurven, Kaplan, Wilcox and Hooper2015). Further, in a recent study, Lew-Levy and colleagues found that Congolese BaYaka and Bantu adolescents showed shifting norms regarding the sharing of money with increased market integration. Together, these findings suggest that adolescents and their peer cultures may play a disproportionate role in adapting to market integration, a pattern that would benefit from further study.

Kong and Ding further suggest that, in contexts where conformity is prized, children’s innovations may take on more subtle forms, expressed through style, language, or humor rather than overt departures from adult norms. Specifically, in East Asian peer cultures with high parental and institutional control, children frequently develop more covert and coded forms of peer culture, shifting their creativity into digital and semi-private “third spaces.” These hidden forms of play and communication demonstrate that even in constrained settings, peer cultures still emerge. We would like to point out another example of peer culture emergence in the face of high institutional control. From the 1880s until World War II, Navajo children were forcibly sent to boarding schools where English was compulsory, and children who spoke Navajo among themselves were punished (Schaengold, Reference Schaengold2004; Velupillai, Reference Velupillai2015). Yet, as linguist Schaengold demonstrates (Reference Schaengold2004), in secret, children at these boarding schools continued to speak Navajo with each other, preserving their language to the best of their abilities in the face of severe pressure. Having been sent to boarding school from a young age, most children’s knowledge of Navajo was incomplete. So, they often substituted unknown Navajo words with English ones, creating a “code” – a mix of Navajo and English – which also became a means of communicating back home with other boarding school survivors. Between the 1950s and 1970s, Bilingual Navajo stabilized as a language distinct from Standard Navajo and is now the main Navajo language spoken by many. Here, peer cultures helped retain and even refine a huge swath of cultural knowledge despite grave threats from adults. Understanding how children manufacture physical or virtual third spaces and how peer cultures covertly arise, adapt, and are transmitted within these spaces offers a promising direction for future research.

Lancy complements this discussion through an assessment of street children’s cultures, offering a case study where children’s autonomy is not restricted but instead expanded by necessity. Without adult or institutional oversight, these groups teach one another through apprenticeship, building shared repertoires of skill and moral understanding. Far from marginal exceptions, such groups reveal children’s impressive organizational and pedagogical capacities when completely unregulated by adults, demonstrating how peer cultures can generate complex cultural forms related to education, cooperation, and creativity even under extreme constraints. Further, through their migration patterns, street children transport with them new ideas and material culture between urban and rural settings. Future research should investigate how children’s migration – in the presence and absence of adult caretakers – shapes the circulation and adoption of novel peer culture forms.

Drawing on their work with Ethiopian Koygu transitional hunter-gatherers, Nyangatom pastoralists, and Central African Aka hunter-gatherers, Bira and Hewlett describe how culturally constructed niches shape peer culture diversity. Where cultural schemas reflect hierarchical social values, such as among the Nyangatom, teaching from adults is more widespread. In contrast, among the egalitarian Koygu, adults minimize their interventions into children’s learning. These conditions may suppress peer group transmission among the former and amplify it among the latter. Echoing the example of high-control East Asian communities above, we would like to note that even in the face of non-supportive cultural schema, peer cultures can covertly emerge. For example, among Masai pastoralists, Tian (Reference Xu, Hewlett, Rawlings, Scaff, Abels, Tian, Badiani, Ndlovu, Rosun and Oppong2025) reports that community elders regarded play as an unworthy topic of inquiry, with some noting that they themselves had not engaged in play during childhood. Nonetheless, Masai children guided Tian in observing how and when they played away from adult supervision. Thus, while we agree that cultural schemas contribute to diversity in peer cultures, we propose that they also influence the extent to which peer cultures are overt or concealed, thereby shaping their visibility and our capacity to study them.

Bira and Hewlett also argue that the size of settlement patterns shapes who children can interact with. They show that larger settlements among the Koygu offer more opportunities for peer group interaction. We additionally note that, in larger settlements, children have more opportunities to assort with same-aged peers, allowing them to play competitive games, which are rarer in mixed-aged groups due to large variations in skill level (Draper, Reference Draper1976; Edwards & Whiting, Reference Edwards and Whiting1988; Lancy, Reference Lancy and Smith1984). Further, in larger settlements children are more likely to segregate into same-gender playgroups (Lew-Levy et al., Reference Lew-Levy, Boyette, Crittenden, Hewlett and Lamb2020). As demographic shifts occur, it’s possible that in some communities, children’s social experiences diverge quite dramatically from their parents. In these cases, it could be that children’s peer group play becomes the site of novel social norms (such as a trend toward hierarchy or gendered segregation), which may have downstream effects for the next generation’s societal schema (Cannon, Reference Cannon2023). Takada adds to this discussion, arguing that mixed-age groups also allow for younger children to observe and imitate older children. He asks: What happens when same-aged groups begin to outnumber mixed-aged ones? This is an important question for the study of both peer culture and cultural evolution more broadly, especially in the face of increasingly formalized institutions that segregate children almost exclusively into same-aged peer groups. Future work should examine how these demographic and institutional shifts reshape the cultural ecology of childhood, influence the balance between cooperation, competition, and hierarchy in peer groups, and how peer cultures are produced and transmitted in these contexts.

Finally, Burdett and Fong point out that our target article primarily focused on children in small-scale societies and that not enough attention was paid to children in urban settings (a point also echoed by Miu et al.). Our rationale for this choice was to highlight instances of peer cultures in communities understudied in child development (Amir & McAuliffe, Reference Amir and McAuliffe2020; Henrich et al., Reference Henrich, Heine and Norenzayan2010; Nielsen et al., Reference Nielsen, Haun, Kärtner and Legare2017). We acknowledge, however, that this emphasis underrepresents how peer cultures manifest in industrialized environments and, as Burdett and Fong observe, may inadvertently give the impression that peer cultures are less prevalent there. A more holistic study of children’s peer cultures should sample broadly across cultural settings, including industrialized ones. The commentaries addressed above go some way in correcting this bias.

R2.2. Communication, social regulation, and cognitive development

Several commentaries focused on distinct communication and social regulation systems at the heart of peer culture, which function to enforce norms, generate knowledge, and negotiate meaning. In turn, such peer interaction shapes children’s developing social cognition and moral reasoning. Together, these commentaries highlight peer groups not merely as sites of cultural transmission and innovation, but as developmental engines in which cognition and culture evolve together.

Yucel highlights norm enforcement strategies – such as protest, tattling, and gossip – as key mechanisms of moral regulation and identity formation in peer cultures. In her view, these strategies do more than police behavior – they also construct the moral landscape of the group, signaling what kinds of actions, alliances, and identities are valued. We agree that this rich social ecosystem deserves closer empirical attention, especially the study of how peer groups influence children’s developing moral reasoning capacities and emerging sense of identity. An informative study, for instance, could consider how the content of peer gossip varies across cultures and ages, revealing which underlying social values are most salient in peer cultures at different stages of development.

Building on the idea that social dynamics shape what and how children learn, Langenhoff et al. highlight disagreement as a productive communicative mechanism within peer cultures. Rather than undermining cohesion, disagreements invite children to articulate reasons, defend positions, and co-construct new understandings. We add that these sorts of interactions are likely even more useful in peer settings where adults are not actively present and unable to intervene to resolve disagreements by providing correct answers. Langenhoff et al. further point out that research has overwhelmingly focused on domains of knowledge that are “already known” and that are typically taught by adults. Open-ended domains, where children can co-construct new knowledge with peers, remain understudied, even though they likely offer rich ground for peer cultural innovation.

Tomasello extends this point, arguing that social interactions among peers differ fundamentally from those between children and adults. Within peer settings, authority is flattened, and children must negotiate meaning, coordinate perspectives, and reach consensus without relying on deference to expertise. These dynamics, which rely on argumentation (and often disagreement, as Langenhoff et al. point out), are critical social experiences that are formative in children’s emerging understanding of morality and of collaboration. One could even make the stronger claim, à la Piaget (Reference Piaget1948), that peer interactions are not only helpful for internalizing and understanding the moral world but are actually necessary. Further, through their peer negotiations, children can develop novel moral norms and reasoning skills that do not exist in adult cultures. The ways in which children collaboratively engage in moral reasoning in peer interactions and the mutual development of culture and cognition in early life are a fruitful avenue for future research.

Similarly, Burdett and Fong argue that peer interactions provide unique social learning opportunities. They cite experimental work showing that through peer interactions, 3-4-year-olds can innovate complex tools that were not achieved when working alone or when working dyadically with an adult (McGuigan et al., Reference McGuigan, Burdett, Burgess, Dean, Lucas, Vale and Whiten2017). The authors suggest that a greater emphasis on the bidirectionality of peer interaction and cognitive development would add an important dimension to the study of peer cultures. We agree that the relationship between peer interaction and cognitive development is reciprocal: peer interactions not only depend on social-cognitive capacities, such as theory of mind and perspective-taking, but also serve as the very context in which these capacities are refined. The ability to interpret others’ intentions, manage disagreement, and coordinate shared goals is both a prerequisite and a product of peer cultures.

Gabora and Beckage’s commentary offers a novel lens on how peer interaction shapes cognition, emphasizing mental rather than social networks. They argue that cognitive development can be modeled as an autocatalytic process in which children’s ideas – once linked and exchanged within peer groups – mutually reinforce one another, producing self-sustaining cycles of conceptual growth. This framework elegantly formalizes what many developmental theorists have described qualitatively: collaboration among children allows them to reconfigure shared understandings and, in doing so, reshape their own epistemic architecture, enabling “each new generation to escape the mental ruts of the previous.” Peer cultures, then, function as collective engines of conceptual reorganization, where children’s interactions not only generate new cultural forms but also catalyze the emergence of increasingly integrated cognitive systems.

R2.3. Mechanisms of transmission

A rich set of commentaries focused on modes of transmission within peer groups, exploring how factors such as peer teaching and social status help maintain and regulate peer cultures. Together, these commentaries reveal that the transmission of peer cultures is shaped by a complex interplay of hierarchy, teaching, and motivation. They collectively push the study of cultural evolution beyond simple models of imitation and toward a richer account that recognizes children as active teachers, selective learners, and self-motivated innovators.

Vasquez brings attention to the social hierarchy of peer groups, suggesting that social rank plays an important role in which cultural products are retained and transmitted. That is, just as children are attentive to social factors such as prestige when deciding which adults to learn from, they are almost certainly applying these same judgments to their own peer groups, with implications for how peer cultures change and spread. We endorse Vasquez’s suggestion that social learning biases (e.g., as reviewed in Kendal et al., Reference Kendal, Boogert, Rendell, Laland, Webster and Jones2018) are likely highly influential in peer cultures and second the call for more work on this topic. Importantly, we would like to add that peer cultures may have their own prestige valences, which are distinct from those of adults. For example, “class clowns” – students known for using disruptive humor or comical behavior in school settings – have been shown to score highly in leadership as well as peer-conferred social status (Damico & Purkey, Reference Damico and Purkey1978; Ruch et al., Reference Ruch, Platt and Hofmann2014; Wagner, Reference Wagner2019). Class clowns are also considered rule breakers who challenge adult-dominated power structures overtly and covertly, signaling allegiance to the peer group rather than to institutional authority, and their comic talent is directly related to perceived leadership (Ruch et al., Reference Ruch, Platt and Hofmann2014; Wagner, Reference Wagner2019). Taken together, this evidence suggests that different social characteristics can be leveraged into prestige within peer cultures and that these may differ from dimensions of prestige valued by adults in the same communities. We believe there are rich research questions to pursue in this arena, using, for instance, longitudinal or network-based methods to trace how innovations move through peer groups and how social factors like prestige amplify or suppress them.

Corriveau et al. also underscore the importance of social status, highlighting not just social rank but also group membership as a key determinant of how and among whom cultural transmission takes place. But for Corriveau et al., these dynamics affect not just peer learning but also peer teaching. They draw from a rich literature documenting the sophisticated ways in which children (spontaneously) teach one another, facultatively adjusting their teaching based on the learner’s knowledge states. This point is reinforced by Bira and Hewlett’s observational data among children in small-scale societies – which suggest that child-to-child teaching occurs in nearly half of all learning events – and by Qiu and Moll’s commentary, which highlights the sophisticated nature of children’s teaching capacities. Corriveau et al. further suggest that “children may strategically withhold information when teaching outgroup members to maintain ingroup loyalty.” This is a compelling example of the in-group psychology we discuss in the target article (Section 7.3), which begins to emerge around age three and is further expressed through selective teaching – deciding who to teach and who to ignore. We fully agree with the authors that peer teaching is an overlooked but critical component of children’s cultures and that naturalistic studies of peer teaching are lacking (but see Boyette & Hewlett, Reference Boyette and Hewlett2017; Lew-Levy et al., Reference Lew-Levy, Kissler, Boyette, Crittenden, Mabulla and Hewlett2020; Maynard, Reference Maynard2002). Studies that focus on the nuances of how culture is taught and negotiated between children can help us better understand the active role that children play in cultural evolution.

In addition to questions of who peer cultures are learned from, an open question concerns which cultural products are most likely to spread. Perez et al. argue that, as children are cared for and provisioned by adults, the necessity of engaging in extrinsically motivated activities is reduced, allowing individuals to rely more on intrinsic motivational systems – those that are inherently satisfying, as opposed to those driven by external rewards or pressures. This motivational landscape shapes which peer cultural products emerge and endure: practices that are intrinsically engaging, for example, that are playful, challenging, or socially rewarding, are more likely to crop up. This perspective offers a useful complement to models of cultural transmission that emphasize efficiency or utility, suggesting instead that the persistence of peer-created culture may depend as much (or even more) on its motivational appeal as on its functional value.

Further, we found great value in Perez et al.’s argument that as children age, their intrinsic motivations also mature, leading to downstream changes in peer cultural products. Specifically, they recontextualized the Matses case study from the target article, arguing that fishing proficiency may have been a byproduct of intrinsic motivation to explore the river. This interpretation underscores how intrinsic motivations in childhood can seed behaviors that later acquire extrinsic utility. It further highlights a developmental continuity between play and production and suggests that peer cultures may serve as incubators of future adaptive skills, where exploration today becomes expertise tomorrow. In other words, practices that originate as intrinsically motivated play, experimentation, or collective problem-solving may later crystallize into enduring adult traditions, skills, or institutions.

Lastly, Bira and Hewlett extend our discussion of transmission mechanisms to highlight two forms of group transmission: concerted transmission, in which multiple group members communicate the same skill or knowledge to an individual, and cumulative transmission, in which children acquire skills or knowledge through repeated, multimodal interactions with several group members over time, each encounter building on prior experience and progressively scaffolding mastery. They argue that these forms of transmission are key for maintaining the stability of peer culture over time, drawing from rich observational data of Koygu, Nyangatom, and Aka children’s lives to illustrate these differing modes. These understudied transmission pathways represent new conceptual tools for understanding how learning is socially distributed and how peer cultures sustain coherence across generations.

R3.1. Formal modeling of peer cultures

Several commentators extended our framework into the realm of formal modeling, outlining new approaches that could help specify how peer cultures emerge, stabilize, and evolve. Across these contributions, there is a consensus that models of cultural evolution should not treat children merely as noise or passive transmitters in cultural systems but as dynamic agents whose age-structured interactions shape the tempo and trajectory of cultural change.

Perez Velilla and Smaldino argue that the influence of children’s cultural experimentation is not constant across settings but modulated by the surrounding “risk landscape.” In low-risk, buffered environments – where the cost of failure is minimal – exploratory peer-driven learning strategies can flourish, generating and spreading novel solutions. Conversely, in high-risk contexts where mistakes are costly, conservative social learning and vertical transmission tend to dominate, constraining the space for peer innovation. This framework provides an elegant way to formalize the ecological conditions under which peer cultures are most likely to act as engines of adaptation, offering a bridge between developmental plasticity and population-level cultural dynamics (Frankenhuis & Amir, Reference Frankenhuis and Amir2022). Their commentary also converges with Greenfield’s argument that shifting socioecological conditions, such as transitions from subsistence to market economies, systematically alter the cultural value of innovation itself. Both perspectives emphasize that the expression and transmission of peer cultures are context-dependent, varying with the risks, affordances, and incentives present in children’s social and ecological environments. We’d further like to note – as Perez Vellila, Smaldino, and colleagues also recently posited (Reference Pérez Velilla, Beheim and Smaldino2025) – that the risk landscape itself changes across development, as children are provisioned by adults and alloparents, the latter of whose care is increased in harsher environments (Martin et al., Reference Martin, Ringen, Duda and Jaeggi2020). Similarly, in our recent work, we evaluated the role of climatic variability in shaping children’s activities across twelve forager and mixed-subsistence populations (Lew-Levy et al., Reference Lew-Levy, Reckin, Kissler, Pretelli, Boyette, Crittenden, Hagen, Haas, Kramer and Koster2022). We found that localized risks, such as water availability and predation, did relate to how children participated in activities such as domestic work and childcare, but we found no relationship between localized risks and food production. This may be because children’s social groups – including foraging with peers – offer a way to buffer food production risks through collective risk pooling. These findings support the notion that the risk environment itself varies across the lifespan and that social mechanisms can help buffer those risks.

Deffner et al. argue that bridging theoretical models and real-world data is essential for evaluating how peer cultures influence cultural change. They propose the rise of plant-based diets as a case study for modelers to explore, noting that youth have driven these diets as a form of climate activism. In Germany, for instance, adolescents are 2.5 times more likely than adults to be vegan or vegetarian. This case study highlights an important mechanism through which children and adolescents may shape the trajectory of cultural evolution: not necessarily by inventing new domains of culture, but by reorganizing existing ones around new values or practices that can influence broader society. Deffner et al. also emphasize that studying culture change requires attention to factors such as population size and age structure. To this latter point, we would like to offer another relevant case study: In the 1950s, Canadian children born in the post-WWII baby boom were raised in relative affluence in a political climate that emphasized freedom in opposition to the Cold War (Owram, Reference Owram1997). Under these conditions, peer cultures thrived. By the 1960s, the generational identity such peer cultures cultivated “was sufficiently forceful […] to challenge the credibility of the adult world” (Owram, Reference Owram1997, p. 217), leading to the radical politics of the era (e.g., nuclear disarmament, anti-war, feminism).

Deffner et al. further argue that, while accounting for population size and age structure necessarily increases model complexity, it also allows for a more accurate representation of how cultures evolve. The authors outline clear steps for modeling the transmission of norms (including diffusion from young innovators through homophilic networks), estimating parameters from longitudinal or historical data, and simulating interventions to promote climate-friendly behaviors. We support this approach and hope that the case studies in our target article, and those offered by commentators, provide a foundation for such generative models.

Morin offers a conceptual complement to these modeling efforts through the notion of selection for proliferation. He argues that peer cultural forms often spread not because they are necessarily efficient or adaptive, but because they are easily reproduced – through repetition, visibility, or emotional resonance. Combined with the high turnover of peer groups, this creates a distinct cultural milieu: many traditions are short-lived and disappear quickly, but some are so good at spreading, they are retained for very long periods of time. In other words, high turnover increases the variance in how long traditions last. The overall result is a small set of highly “proliferative” children’s traditions (like certain playground games, songs, or jokes) that can endure just as long as some adult cultural traditions, even though the population of children is constantly changing. This mechanism offers a powerful way to reconcile the apparent instability of peer cultures with their surprising continuity over time – an open question also raised by Takada. Rather than viewing rapid turnover as a liability, Morin’s account reframes it as a selective environment that privileges transmissibility over function, producing a form of cultural evolution optimized for engagement rather than efficiency. Echoing Perez et al. and Takada’s argument that intrinsic motivations are key for peer cultural engagement, Morin’s framework invites cultural evolutionary models to move beyond utility-based assumptions and to consider affect and play as central selective forces in children’s cultural systems.

R3.2. Peer cultures among non-human animals

Several commenters extend the presence and function of peer cultures to non-human animals. This perspective reinforces our view that the developmental features of juvenility uniquely foster cultural experimentation and that these dynamics are not limited to humans.

Sueur et al. present a particularly comprehensive and vivid account of non-human animal peer cultures, drawing on a wide range of animal studies to argue that juvenile animals are not merely passive learners but active agents of cultural innovation. They document numerous cases where young individuals generate behavioral novelties – from object play to tool use – that subsequently diffuse through their groups. For example, juvenile macaques have been observed initiating practices such as stone handling and sweet potato washing, behaviors later adopted by older members (Hirata et al., Reference Hirata, Watanabe, Masao and Matsuzawa2008). Altogether, their review makes a compelling case that peer cultures are not uniquely human.

Gray similarly underscores the role of juvenile play as a crucible for cultural innovation. He highlights another case study among macaques (Tan, Reference Tan2017), in which young macaques playfully began to smash rocks together, perhaps just to enjoy the sound (led by intrinsic motivations, as Perez et al. point out). At some point, the rock-smashing among juveniles – and seemingly only juveniles – gained instrumentality as a way to crack open shellfish, allowing access to a nutrient-rich resource. When these juveniles matured into adults, their behavioral repertoires matured with them and were subsequently learned by new generations through imitation. We agree with Gray that there may be rich avenues for research in this domain, examining the role of play in downstream cultural innovations.

McAuliffe and Thornton extend this comparative argument by focusing on the mechanisms through which juvenile-led innovations are transmitted within animal societies. They highlight three possible routes: (1) juveniles may retain their learned behaviors as they age, allowing those innovations to persist and eventually diffuse through the group once the innovators reach adulthood; (2) other individuals, regardless of age, may selectively adopt successful juvenile practices, indicating that competence rather than dominance can guide social learning; and (3) some species can flexibly adjust their information-use strategies, with adults temporarily relaxing age-based learning biases to acquire knowledge from younger individuals when circumstances favor it. Together, these pathways demonstrate that juvenile innovations need not terminate within the peer domain but can, under the right social conditions, permeate the broader cultural system of the group, even in non-human species.

McAuliffe and Thornton also point out an interesting feature of juvenility that may contribute to innovation: juveniles are often of low rank and may therefore need to find creative ways to secure resources that higher-ranking members need not seek. This social positioning may in fact create ideal conditions for experimentation; with less to lose and fewer expectations to conform to, juveniles may exploit niches of opportunity overlooked by more dominant individuals. Such innovation from the margins mirrors patterns seen in human peer cultures, where reduced oversight and marginal status can likewise foster creativity and variation. For example, in many hunter-gatherer societies, children develop specialized foraging tools adapted to their smaller size and lesser strength, allowing them to target food resources usually ignored by adults (see Pretelli et al., Reference Pretelli, Crittenden, Dounias, Friant, Koster, Kramer, Mangola, Saez and Lew-Levy2024 for review). Framing juvenility in this way – as both a socially and developmentally privileged context for exploration – reinforces the idea that peer cultures, across species, function as engines of adaptive novelty within broader cultural systems.

R3.3. Identifying peer cultures in the archaeological record

Several commentators turned to archaeology to examine how children’s cultural contributions might appear – or fail to appear – in the deep past. They noted that while identifying peer cultures and children’s innovations in the archaeological record is difficult, the study of contemporary children and their material culture can provide new insights and help clarify how children contributed to cultural evolution in human history.

Sugiyama argues that Homo tool industries show little evidence of innovation, challenging the idea that children serve as agents of cultural adaptation. She points to the strong, apparent conservatism of Acheulean tool industries produced by other members of our genus, including H. erectus and H. heidelbergensis. Sugiyama also cites research showing a gradual technological transition from the Middle Stone Age to the Late Stone Age in the African archaeological record. She concludes that, until the European Upper Paleolithic, the cultural products of our ancestors reflected continuity rather than innovation. Only in the Upper Paleolithic, Sugiyama contends, do we see a sustained shift toward novel developments in technology and symbolic expression – including art, music, and personal ornamentation. She further argues that this period coincides with longer lifespans and larger social groups, which enabled cumulative knowledge transfer across generations, particularly from grandparents.

The discontinuous view of behavioral modernity suggested by Sugiyama’s commentary – emerging relatively abruptly in H. sapiens rather than gradually across Homo – is challenged by recent archaeological discoveries. Among Neanderthals, recent findings suggest comparability of behavioral and cognitive capacities (Maida, Reference Maida2025) as reflected in evidence of technological diversity and complexity in the lithic and organic record (Florindi et al., Reference Florindi, Santaniello, Aranguren, Grimaldi, Melandri, Puzio and Revedin2024; Leder et al., Reference Leder, Lehmann, Milks, Koddenberg, Sietz, Vogel, Böhner and Terberger2024), creation of large structures (Jaubert et al., Reference Jaubert, Verheyden, Genty, Soulier, Cheng, Blamart, Burlet, Camus, Delaby, Deldicque, Edwards, Ferrier, Lacrampe-Cuyaubère, Lévêque, Maksud, Mora, Muth, Régnier, Rouzaud and Santos2016), and symbolic expression including engravings (Leder et al., Reference Leder, Hermann, Hüls, Russo, Hoelzmann, Nielbock, Böhner, Lehmann, Meier, Schwalb, Tröller-Reimer, Koddenberg and Terberger2021). With regard to the Acheulean record and the contention that there is little evidence of innovation among H. erectus and H. heidelbergensis populations, Hopkinson et al. (Reference Hopkinson, Nowell and White2013, p. 63) argue that this view “does not withstand scrutiny.” The authors review considerable evidence for variability in the Acheulean tool industry, including in handaxe morphology, regional and temporal variation in early appearances of prepared core reduction technology prior to the Middle Paleolithic, and variability in the primary reduction of large cutting tools. Hopkinson et al. further argue that local short-term variability paired with long-term global stasis in Acheulian populations may be explained by (1) relatively short childhoods in comparison to later Homo, limiting opportunities for innovation on local scales, and (2) small group sizes and limited network interconnectivity, reducing the likelihood that innovations would spread between groups. While the exact timing for the evolution of childhood remains uncertain, through greater network connectivity and longer pre-reproductive periods, innovations produced by H. sapiens children could have been disseminated into the wider regional social ecology of the Upper Paleolithic.

Further, our contention was not that peer cultures are the only mechanisms by which communities adjust to change, and there is undoubtedly an important role for grandparents in the human story. However, dissociating the emergence of these two life history stages is complicated, as the evolution of extended childhood and postreproductive longevity likely co-occurred (Kaplan et al., Reference Kaplan, Hill, Lancaster and Hurtado2000), mutually reinforcing each other and creating new social and cognitive conditions for cumulative culture. Where there were grandparents, there were also children. A longer juvenile period potentially afforded more time for exploration and horizontal learning among peers, while longer post-reproductive lifespans expanded opportunities for vertical and oblique transmission from elders. Together, these life history shifts may have forged a developmental bridge between innovation and retention – with children introducing variation through play and experimentation, and grandparents stabilizing and transmitting those successful innovations across generations.

Sugiyama also draws on ethnographic evidence to argue that children in diverse societies display respect and deference toward adult authority. However, we note that such observations are typically made by adult ethnographers usually focused on adult interlocutors. As Barker and colleagues observe, outside the purview of adults, children can be strikingly irreverent toward adult norms and cultural forms. Similarly, Coffey and Scaff demonstrate that the presence of adults can obscure the full complexity of children’s cultural production. We should not assume that children always defer to adults or adult-generated culture when beyond adult supervision.

Finally, we note that the absence of clear evidence for children’s technological innovations does not necessarily imply their absence from the process of cultural evolution. Assuming that children were indeed not contributing much to technological innovation, one explanation may lie in, as Mesoudi points out, higher acquisition costs; it’s possible that children’s contributions to complex technologies may have been hampered and that their contributions are more likely in other domains (a point echoed by Nielsen). However, it’s also possible that children did contribute to technological innovations, but that these contributions are difficult to recognize. For example, among the Papuan Gidra, children were given finished bows, from which they would learn the manufacturing process through reverse engineering (Nishiaki, Reference Nishiaki, Akazawa, Nishiaki and Aoki2013). Rather than produce innovative products, children may have generated novel manufacturing processes, something unlikely to be identified as such in the archaeological record.

Furthering this point, Snyder and Bandini argue that archaeological traces of children’s activity may be present but under-recognized. For instance, the small, irregular flakes, or “practice pieces,” once dismissed as production debris may, in fact, index moments of learning, imitation, and experimentation within early peer groups. They outline concrete methodological steps for identifying children in the archaeological record, such as experimental knapping studies with contemporary children and fine-grained analyses of artifact variability that might reflect learning or play. Encouragingly, a number of archaeologists have already begun to explore this space (Kilgore et al., Reference Kilgore, Liu and Stout2025). In a recent flintknapping study, for instance, children aged 8–11 took turns learning to knap (Milks and Stileman, personal communication). They were seated side by side, with every other child participating in the first session, then switching places so their neighbors could learn during the second session. The experiment leads, Annemieke Milks and Finn Stileman, observed that during the second session, the children who had already practiced began offering advice and instructions – often unsolicited – to their neighbors, supporting their learning through shared experience. While an adult instructor was sometimes present to provide demonstration and individual guidance, in his absence, children often stepped in to take on that teaching role themselves. These observations vividly illustrate the natural drive for peer teaching and learning, even in apprenticeship contexts traditionally assumed to involve adult-only instruction, such as those described by Fischer (Reference Fischer1989) and Takakura (Reference Takakura, Akazawa, Nishiaki and Aoki2013).

Lastly, Miu et al. extend these insights into the domain of formal modeling, offering tools that may eventually allow us to test hypotheses about children’s roles in the archaeological record. They argue that incorporating developmental parameters into models of cultural evolution can help generate expectations about what children’s cultural contributions might look like in the material record, thus providing a theoretical framework for interpreting the material traces of peer learning and innovation (Miu et al., Reference Miu, Andersen, Lew-Levy and Riede2025). We agree that material culture is an ideal focus point for understanding children’s contributions to culture change across contemporary cultures, both in the deep past and among our non-human kin. Further, material culture can provide insights into individual variation in behavior: how tools are used, how toys and small tools facilitate learning (Riede et al., Reference Riede, Johannsen, Högberg, Nowell and Lombard2018), and the role of horizontal and vertical transmission in learning to manufacture tools (Imamura, Reference Imamura, Hewlett and Terashima2016; Jordan, Reference Jordan2015; Meyer et al., Reference Meyer, Haidle and Riede2025), along with population-level effects on network connectivity (Kline & Boyd, Reference Kline and Boyd2010) and environmental variability (Collard et al., Reference Collard, Buchanan, O’Brien and Scholnick2013; Meyer & Riede, Reference Meyer and Riede2025). Current databases, such as the Play-Object-Play database (https://play-object-play.au.dk/), which surveys tools and toys made for and by children in the ethnographic record, offer one avenue for understanding developmental and cross-cultural patterns in children’s object manufacturing and use, including within peer cultures. Morin offers another corpus of material culture: he suggests that online language data may now provide a novel and enduring domain for tracking and assessing peer cultures. The digital material record, from forum exchanges to collaborative gaming and social media interactions, can open new empirical terrains for studying how children learn, create, and circulate cultural innovations. This suggestion underscores a potential shift in developmental research – from inferring cognitive processes in the past to directly observing cultural change in real time and at scales previously inaccessible.

R4.1. Meta-theoretical considerations: Whose voice is heard, and why does it matter?

Dutra argues that scholarship from the Global South – and scholars based there – must be meaningfully integrated into the study of peer culture evolution. We fully agree. The impetus for our target article was shaped by such cross-contextual encounters: the serendipitous discovery of Morin’s article on children’s traditions (Reference Morin2010), for instance, recognized by Lew-Levy because of her fluency in French; our reading of Bame Nsamenang’s work from Cameroon (Reference Nsamenang1992); and our sustained engagement with the communities we have worked with for many years. Inclusion must extend far beyond linguistic or geographic boundaries, and we are optimistic that technological advances such as those that facilitate long-distance engagement and translation will greatly support these efforts.

Dutra also suggests our focus on autonomy in peer cultures may be representative of hegemonic western thought. To clarify, our usage of “autonomy” is actually most closely related to what hunter-gatherer researchers term cooperative autonomy, often defined as “a combination of obligations to the group and protections for individuals against coercion by others” (Endicott, Reference Endicott, Gibson and Sillander2011, p. 81). This framing underlies our depiction of overlapping circles in Figure 1 – reflecting our view that adult and peer cultures are not wholly independent – and our argument that children’s contributions to domains such as food security illustrate cooperation rather than separation. In upcoming empirical work on peer cultures, we plan to collaborate closely with in-country researchers as co-designers and co-authors, and we encourage others to similarly engage with local knowledge systems regarding the value – or limits – of children’s peer groups in their societies.

Takada draws on Malinowski’s notion of the “small republic of children” to argue that peer cultures function as political bodies within which children collectively negotiate norms, resolve conflicts, and make decisions that balance individual autonomy with group cohesion. What insights can we gain from children’s social regulation mechanisms that could be applied to adult societies today? A few key features of peer cultures may be worth highlighting in this regard. First, peer cultures tend to be less authoritarian, and while some age-based hierarchy does exist, it is much looser than the hierarchy that structures most adult communities. Power is often negotiated through consensus and mutual respect rather than fixed authority. Second, children tend to retain a high degree of individual autonomy, exercising choice within collective frameworks that still value equality and inclusion. This balance between independence and interdependence may offer a useful model for how adult societies can foster cooperation without reproducing rigid hierarchies. Third, children are buffered from many of the risks and time demands encountered by adults, giving them space for exploration and creativity. Economic initiatives that offer social support, such as a universal basic income, can similarly provide a protected environment necessary for adults to explore widely, expand their creativity, and innovate solutions to broader societal issues (RSA, Reference RSA2016).

R4.2. Are peer cultures distinct from adult cultures?

A number of commentators raised questions about the boundary between peer and adult cultures, challenging how clearly the two can be separated.

Barker and colleagues, writing from a folklorist perspective, offer their views on the distinctiveness of peer and adult culture, suggesting that the lines between these cultural spheres are often blurred. Citing several (delightful) examples from children’s folklore studies, they argue that children often directly acquire adult cultural forms but transform them into something that belongs distinctly to their own cultures. Barker et al. suggest that developmental needs play an important role in such hybridization. Further, the authors bring terms and histories from folklore studies, which we believe can enrich evolutionary research on this topic. For instance, we find great utility in the term “triviality barrier,” referring to a tendency of scholars to dismiss topics seen as frivolous or trivial, as opposed to treating them as serious subjects of study. We also found it quite stimulating to consider the protective role of the triviality barrier as a way to keep meddling adults out of peer cultures. Whether children explicitly recognize this barrier and whether they strategically utilize it to prevent adult incursion is an important avenue for future work.

Bennis similarly argues that children have not been dismissed but rather ignored in cultural evolution. That is, while adults are unmarked – treated as “natural, normal, or representative examples of a broader category (people)” – children are marked as a deviation from that category. Bennis further emphasizes that peer cultures are not limited to children. That is, people can form peer groups at different life stages, such as in elderhood, or in different social niches, such as members of a tech start-up. While this is true, we would argue that there is something evolutionarily distinctive about childhood that sets it apart from other life stages. The long, protected period associated with this life stage in humans affords children extended time for exploration, experimentation, and horizontal learning – conditions rarely replicated in adulthood. Further, children’s cognition differs markedly from that of adults in ways that make their peer interactions uniquely generative. Compared to adults, children exhibit greater neural plasticity (Fandakova & Hartley, Reference Fandakova and Hartley2020), cognitive flexibility, and exploratory learning tendencies (Gopnik, Reference Gopnik2020; Lucas et al., Reference Lucas, Bridgers, Griffiths and Gopnik2014). Their reasoning is often less constrained by prior knowledge or social convention, enabling them to entertain multiple possibilities and learn through experimentation (Gopnik, Reference Gopnik2020; Gopnik et al., Reference Gopnik, O’Grady, Lucas, Griffiths, Wente, Bridgers, Aboody, Fung and Dahl2017; Lucas et al., Reference Lucas, Bridgers, Griffiths and Gopnik2014). Gray echoes these points in his commentary, highlighting four dispositional traits that are especially pronounced in childhood and together predispose children toward exploration and innovation. Curiosity motivates children to seek out novelty, playfulness encourages creative engagement with both ideas and materials, sociability facilitates rapid diffusion of new behaviors within peer groups, and future orientation directs attention toward emerging or anticipated challenges. These features make peer contexts especially fertile grounds for innovation: in interaction with similarly flexible partners, children can collaboratively generate novel ideas, behaviors, and social practices that adults might overlook or dismiss. Peer interaction thus functions not only as a medium for cultural transmission but also as a potential driver of cultural adaptation and change.

In a related commentary, Dutra argues that the “boundary between childhood and adulthood is permeable, overlapping, and culturally contingent,” advising against an “age-based essentialism.” Wertz & Russell-Wilks echo this point, suggesting that the dichotomization between child and adult culture risks oversimplifying the story. As stated above, we agree that there is not always a sharp boundary between adult and child cultures, in much the same way that there is no sharp boundary between childhood and adulthood. Further, we agree with Dutra that an outsized focus on age as a discrete category risks obscuring the gradual nature of development. However, the existence of gradual change in development does not preclude us from considering the distinctiveness of childhood as a stage with its own social and cognitive affordances. In much the same way, the existence of overlap and influence between child and adult culture does not undermine the value of treating them as analytically distinct. Rather than drawing sharp boundaries between children’s and adults’ cultures and risking essentializing by age, our aim is to provide a framework that lends structure to complex variation.

Wertz & Russell-Wilks raise an additional and important point about whether the underlying psychologies that produce peer culture in childhood and adult culture later in life are markedly different, which would help determine whether peer cultures comprise a qualitatively different form of culture than any other type of culture. They bring attention to a number of candidate domains – such as artifact reasoning and language – which may evidence distinct signatures as a function of an individual’s age. We find Wertz and Russell-Wilks’ question thought-provoking. In their view, if underlying psychological mechanisms are distinct enough, peer cultures would constitute a different type of culture altogether. However, would the importance of peer culture be threatened if these differences were not found? We’re not so sure. All human cultures – adult or child – likely draw on a shared suite of cognitive capacities but can produce distinct products reflecting specific social and physical needs specific to each life stage. The distinctiveness of peer culture, then, may arise from these contextual differences rather than from fundamentally different psychologies. Still, examining whether the underlying mechanisms differ would offer valuable insight into how and why peer cultures develop their unique forms.

Lastly, Qiu and Moll challenge the idea that children’s cultural production arises primarily from peer-generated knowledge. They argue that children’s cultural transmission often focuses on content that originates from adults, which children then adapt, refine, and pass on to others – including both peers and adults. In this view, children’s contribution to cumulative culture lies less in generating novel content and more in their role as pedagogical agents and cultural brokers, transmitting and transforming adult-derived knowledge across social contexts. The authors use as examples children’s translation for parents and grandparents in immigrant contexts, the introduction of technologies, and civic education at home. We believe that while these are compelling points, a focus on adult-derived content need not obscure children’s own innovative contributions. As outlined in our article, peer cultures also generate distinct cultural forms – novel games, rituals, linguistic practices, and social norms – that arise from children’s collective activity rather than direct adult transmission. Further, innovation can occur through the process of cultural brokering. For example, children may transform adult-derived cultural knowledge through copying errors, leading to change in these traits. And children may transmit some adult-derived traits and not others, leading to cultural selection. The invention of peer-derived knowledge and the transformation of adult-derived knowledge represent two potential sources of cultural innovation that require further study.

R4.3. How significant are children’s contributions to culture change?

While many commenters agreed that children likely have their own distinctive cultures, several commentaries raised important questions about the scope and impact of peer cultures on broader trends in cultural evolution.

Mesoudi points out that two types of innovations exist: those that optimize existing knowledge and those that open a new niche or domain by exploiting a new phenomenon (Derex, Reference Derex2021). More optimized cultural products tend to be more complex, carry high cultural acquisition costs, and thus involve long learning periods, limiting children’s abilities to meaningfully innovate. In post-industrial settings, cultural products generally, and technology specifically, have become increasingly complex over time. The associated acquisition costs may hamper children’s abilities to meaningfully innovate. Yet, children may uniquely contribute to the opening of new knowledge niches. Using the example of the computer game industry in the 1980s UK, Mesoudi shows that adolescents mastered programming independently from adults. As computer games advanced, the industry shifted towards trained programmers, eventually excluding adolescents due to the high acquisition costs of game production optimization. But, as the costs of adopting new technology decrease, children and adolescents may again have the opportunity to contribute meaningfully in this space. We agree with Mesoudi that accounting for acquisition costs associated with optimization vs. niche expansion “might help to elucidate exactly when, for what traits, and in what kind of societies, children are likely to contribute to society-wide cultural adaptation.”

Similarly, Nielsen argues that children’s cultural influence is almost certainly domain-sensitive. He notes that while it may be true that children maintain distinctive peer cultures, the claim that these cultures have lasting effects on broader cultural evolution is less certain. Nielsen specifically asks, “If childhood cultures are meaningful, why are children not less inclined to worship the God(s) of those around them (indeed, why don’t they develop, and keep as real, their own)?” We agree that some domains – such as religion, ritual, or formal institutions – are likely less open to peer-driven change. Yet, peer-driven gods are not completely absent from the ethnographic record. Among the Mbendjele BaYaka in the Republic of the Congo, for instance, children are the guardians of a forest spirit named Bolu, who is respected by children and adults alike (Knight & Lewis, Reference Knight and LewisIn Press; Lewis, Reference Lewis2002). Like forest spirits guarded by adults, Bolu has his own ndjanga, or secret path, and specific songs. Unlike forest spirits guarded by adults, however, there is no formal initiation for Bolu. Bolu is an ancient forest spirit and has likely been maintained and transmitted across many generations of children (Lewis, personal communication). However, we agree that other domains are likely far more permeable. In areas characterized by rapid transformation or weak adult expertise – such as new technologies, social media, and digital communication – Nielsen argues that children may act as early adopters and innovators, guiding how these tools are used and understood. Their fluency in navigating emerging ecological and technological landscapes may give them disproportionate influence precisely where cultural adaptation is happening most quickly.

Kempe and colleagues identify language – particularly phonology, morphosyntax, and the lexicon – as a domain in which young children are unlikely to contribute to innovation. They argue that some of the linguistic innovations we highlighted among young children tend to disappear in later ages, as children conform to adult linguistic conventions with higher fidelity. Here we offer several clarifications. First and foremost, Kempe et al. define children as 10 and under, whereas in our target article childhood was defined more broadly, as between 3 and 17 years. By this definition, much of what Kempe et al. describe as adolescent-driven change would still fall within the scope of our argument. Second, the authors challenge the idea that, although children may not shape established languages, they could nonetheless play a key role in the emergence of new ones. Interestingly, the proposition that children contribute to language emergence aligns with Mesoudi’s distinction in the technological domain between niche expansion and optimization: children may drive expansion into new linguistic niches (the creation of novel languages), whereas adults are more likely responsible for optimizing established linguistic systems. Kempe et al. also question whether, in cases such as Nicaraguan Sign Language, innovations arise from children themselves or from their position within an iterated learning chain. They express skepticism toward the former, arguing that children’s limited cognitive capacities constrain their ability to “scan the semiotic affordances of referential domains and monitor alignment with an interlocutor.” Yet by around age five, children’s mentalizing abilities are well developed (Callaghan et al., Reference Callaghan, Rochat, Lillard, Claux, Odden, Itakura, Tapanya and Singh2005), and as Wertz and Russel-Wilks note, their language learning capacities far surpass those of adults. It therefore seems plausible that, when interacting without a shared language, children could infer one another’s intentions and converge on shared communicative acts – processes that might underlie the emergence of new languages. Importantly, we do not claim that children are the only ones to contribute to language emergence, but that, like adults, they can.

Kempe et al. also discuss prestige and frequency biases, which would favor learning from adults. However, this assertion may reflect an outsized emphasis on adult language to the exclusion of peer group language and communication more broadly. Indeed, there is ample evidence suggesting that children’s learning biases are context sensitive. For example, in two naturalistic studies, children between the ages of 2 and 5 years imitated adults in disciplinary-related domains but peers in affective and play-related domains (Grusec & Abramovitch, Reference Grusec and Abramovitch1982; Kuczynski et al., Reference Kuczynski, Zahn-Waxler and Radke-Yarrow1987). In experimental contexts, children aged between 3 and 5 years preferred adults as sources of information about nutritious foods but peers for information related to toys (Shutts et al., Reference Shutts, Banaji and Spelke2010; VanderBorght & Jaswal, Reference VanderBorght and Jaswal2009). We also note that several studies from children outside of the post-industrialized West demonstrate that children receive more language input from other children than adults (e.g., Cristia et al., Reference Cristia, Dupoux, Gurven and Stieglitz2019, Reference Cristia, Gautheron and Colleran2023). Therefore, one possibility is that children contribute to language evolution at the level of the peer group, even if these innovations rarely propagate beyond the peer group. Kempe et al. further argue that “young children require adult scaffolding to reach communicative efficiency.” However, the paper by Matthews et al. (Reference Matthews, Lieven and Tomasello2007) cited to support this claim only included adult scaffolding, with no peer condition for comparison. While we agree that scaffolding is necessary for language development, whether peers or adults scaffold more effectively in different contexts and use the same strategies to do so is still an open and interesting empirical question.

To this latter question, Coffey and Scaff provide evidence that, in fact, adult scaffolding is not necessary for communicative efficiency. Specifically, as we noted above, they argue that linguistic input from peers has been systematically overlooked. As the authors point out, researchers themselves are adults, and their very presence can distort children’s behaviors – prompting them to speak more performatively or deferentially than they otherwise might. To address this, they advocate for long-form audio-based recording methods that capture children’s interactions in real time with minimal adult interference. Long-form recordings show that even in urban western classroom settings, “children talk more amongst each other than they do their teachers and their speech production is better predicted by peer input than adult input.” Morin offers another source of naturalistic linguistic data – text exchanges, forum posts, or collaborative platforms – suggesting that these may offer an unprecedented window into how peer-driven linguistic variation propagates and stabilizes. We believe studies adopting such naturalistic methods have the potential to revolutionize our understanding of how children contribute to each other’s linguistic, cognitive, and social development within peer group settings.