Lew-Levy & Amir convincingly argue for the importance of peer cultures for the cultural niche and cumulative cultural evolution in recent Homo sapiens. They extend this argumentation by suggesting a relevance of peer cultures to hunter-gatherer prehistory and human adaptiveness for overcoming environmental change. These ideas, however, require additional development if we are to integrate the peer culture concept into the wider constellation of evolving human cognitive and cultural traits.

Because peer cultures (with the full suite of components sensu Lew-Levy & Amir) seem to be unique to humans, investigation into the evolutionary context of peer cultures should be targeted at the preserved evidence in the archaeological record, i.e., material culture like stone tools (Bandini, Harrison, & Motes-Rodrigo, Reference Bandini, Harrison and Motes-Rodrigo2022; Schick & Toth, Reference Schick and Toth1994; Toth & Schick, Reference Toth and Schick2018). Debates on the cultural and linguistic abilities of extinct hominins notwithstanding (Morgan et al., Reference Morgan, Uomini, Rendell, Chouinard-Thuly, Street, Lewis, Cross, Evans, Kearney, de la Torre, Whiten and Laland2015; Putt, Woods, & Franciscus, Reference Putt, Woods and Franciscus2014; Snyder, Reeves, & Tennie, Reference Snyder, Reeves and Tennie2022; Stout et al., Reference Stout, Toth, Schick and Chaminade2008, Reference Stout, Rogers, Jaeggi and Semaw2019; Tennie et al., Reference Tennie, Premo, Braun and McPherron2017; Toth & Schick, Reference Toth and Schick2018), there is no non-material culture (e.g., the H. erectus equivalent of nursery rhymes) that we can analyze directly as can be done with living cultures.

The archaeological record can generally be rather sparse and incomplete (Dibble et al., Reference Dibble, Holdaway, Lin, Braun, Douglass, Iovita, McPherron, Olszewski and Sandgathe2017; Perreault, Reference Perreault2019). First, taphonomic processes can simply “erase” much of what was originally produced and left behind by hominins. There are also theoretical and methodological constraints limiting our ability to properly identify cognitive processes and cultural phenomena from the past (Dibble et al., Reference Dibble, Holdaway, Lin, Braun, Douglass, Iovita, McPherron, Olszewski and Sandgathe2017). Depending on what contexts we look at, peer cultures and/or their underlying prerequisites (like extended childhoods) could also have simply been absent to start with. Which prompts the question: can we find peer cultures sensu stricto in the archaeological record (see also Meneganzin & Currie, Reference Meneganzin and Currie2025)?

Theoretically, we could identify prehistoric peer cultures based on the features outlined by Lew-Levy & Amir, including both processes (i.e., high-fidelity transmission including biased horizontal transmission between children and variation-generation via innovation) and observable outcomes (i.e., segregation of child and adult spaces; parallel, distinct peer and adult cultures; and the preservation of abandoned adult knowledge).

However, behavioral and cognitive processes do not fossilize and can be difficult to reconstruct (Dibble et al., Reference Dibble, Holdaway, Lin, Braun, Douglass, Iovita, McPherron, Olszewski and Sandgathe2017; Schick & Toth, Reference Schick and Toth1994). High-fidelity transmission (or, “know-how copying”: Bandini et al., Reference Bandini, Motes-Rodrigo, Steele, Rutz and Tennie2020; Tennie et al., Reference Tennie, Bandini, van Schaik and Hopper2020) alone is a matter of intense debate in archaeology involving many distinct accounts (Morgan et al., Reference Morgan, Uomini, Rendell, Chouinard-Thuly, Street, Lewis, Cross, Evans, Kearney, de la Torre, Whiten and Laland2015; Snyder et al., Reference Snyder, Reeves and Tennie2022; Stout et al., Reference Stout, Rogers, Jaeggi and Semaw2019; Tennie et al., Reference Tennie, Premo, Braun and McPherron2017; one account is shown in Figure 1). Furthermore, learning biases and (play-driven) innovation are practically impossible to track down in (earlier) periods with our current methodologies. Pre-existing child knapping anecdotes and experiments illustrate child knapping innovation potential (Ferguson, Reference Ferguson2003; Sternke & Sørensen, Reference Sternke, Sørensen, McCartan, Schulting, Warren and Woodman2009), but these cases involved vertical transmission from adults to children. It is also unclear if and how material outcomes of children’s innovation processes differ from outcomes resulting from human adults’ innovation processes (Pargeter et al., Reference Pargeter, Liu, Kilgore, Majoe and Stout2023; Snyder et al., Reference Snyder, Reeves and Tennie2022; or even adult apes; Toth et al., Reference Toth, Schick, Savage-Rumbaugh, Sevcik and Rumbaugh1993).

Figure 1.

A simplified evolutionary timeline of traits prerequisite for the emergence of peer cultures (Caspari & Wolpoff, Reference Caspari, Wolpoff, Smith and Ahern2013; Gunz et al., Reference Gunz, Neubauer, Maureille and Hublin2010, Reference Gunz, Neubauer, Falk, Tafforeau, Le Cabec, Smith, Kimbel, Spoor and Alemseged2020; Meneganzin & Currie, Reference Meneganzin and Currie2025; Paige & Perreault, Reference Paige and Perreault2024; Snyder et al., Reference Snyder, Reeves and Tennie2022; van Schaik, Pradhan, & Tennie, Reference van Schaik, Pradhan and Tennie2019).

Figure 1 Long description

A diagram illustrating the evolutionary timeline of traits prerequisite for the emergence of peer cultures. The diagram is divided into four main sections: Stone tool technology, Cultural transmission, Life history, and Demographic structure. The timeline spans from 2.58 million years ago to the present, covering the Pliocene Epoch and the Pleistocene Epoch, which is further divided into Early, Middle, and Late periods. Stone tool technology evolves from Oldowan tools to Early Acheulean tools, Late Acheulean tools, and Levallois tools. Cultural transmission progresses from no know-how copying, teaching, or language to cumulative culture with know-how copying, teaching, and language. Life history shows a transition from protracted development without a childhood stage to postnatal brain development suggestive of a childhood stage. Demographic structure indicates increasing survival of older individuals and an increased proportion of older individuals with multiple cohabiting generations.

If we look at outcomes in the archaeological record rather than at processes, we still run into issues. But, archaeological sites, especially in earlier prehistory, tend to be time-averaged and spatially adjacent materials are not always from temporally related events (Perreault, Reference Perreault2019). Consequently, peer culture outcomes (segregation of child and adult spaces and parallel peer and adult cultures, if they were present) may be undetectable due to the poor spatiotemporal resolution at many evolutionarily relevant sites. Finding the “preservation of abandoned adult knowledge” could also be difficult. “Simpler” technological modes re-appear and co-occur with more “advanced” modes throughout prehistory, but with current methods, determining whether maintenance of older technological modes is due to children versus adults is impossible. For example,

Late Neandertals may have used both “simpler” Mode 1 and “advanced” Mode 3 tools due to convenience or utility (Shipton et al., Reference Shipton, Clarkson, Bernal, Boivin, Finlayson, Finlayson, Fa, Pacheco and Petraglia2013) and not as part of parallel peer cultures, although such a hypothesis would be worth considering.

Overall, it would be difficult to find unequivocal evidence in Pleistocene contexts for peer cultures sensu Lew-Levy & Amir. But what if we first take a step back: can we find more general evidence of hominin children’s activities, i.e., by differentiating children’s cultural products from adult’s cultural products? There has been concerted attention given to children’s activities in later prehistory (H. sapiens from the Late Paleolithic onwards; e.g., Finlay, Reference Finlay2015; Riede et al., Reference Riede, Walsh, Nowell, Langley and Johannsen2021). Empirical investigation of children even earlier – necessary for evaluating peer cultures as a driver of human evolution – has been more limited (see work of Nowell, Reference Nowell2021).

Testing this requires observable outcomes like artifact attributes and assemblage-level patterns, which might vary according to the toolmaking agent’s age. Here, we especially need markers that can distinguish child novice toolmakers from adult novices (for early stone toolmaking and not, e.g., Mesolithic toolmaking; Sternke & Sørensen, Reference Sternke, Sørensen, McCartan, Schulting, Warren and Woodman2009). For example, child knappers make smaller flakes compared to adult experts, but so do adult novices (e.g., Pargeter, Khreisheh, & Stout, Reference Pargeter, Khreisheh and Stout2019, Reference Pargeter, Liu, Kilgore, Majoe and Stout2023). Smaller flakes can also relate to economic or technological decisions by knappers (e.g., Gallotti & Mussi, Reference Gallotti and Mussi2015). Thus, more precise methods of discerning between different sources of artifact and assemblage variation need to be developed.

To make inferences about the activities of children in early prehistory – from our perspective – we require experimentally produced reference datasets. Knapping experiments (on early stone toolmaking in the Pliocene to Middle Pleistocene) with human children can elucidate the ontogenetic and cultural processes relevant to toolmaking, while also giving better indications of what children’s cultural products would look like. Ethical concerns (related to consent and safety) have been a major obstacle to such studies, but recent advances in artificial knapping materials (e.g., Snyder et al., Reference Snyder, Boysen, Orellana-Figueroa, Tennoe and Reeves2025), 3D-printing, and virtual reality should allow safer and easier experimental data collection.