2 January 2025

Know-How Copying Is Fundamental to Human Culture

Sterelny argues that the role of copying (and especially single model copying) in human culture is overstated. Consequently, he thinks it misguided to characterize cultural evolution as Darwinian, since there are no clear parent–offspring lineages at the cultural level. He also makes a positive case for the role of niche construction and individual learning in cumulative culture. We endorse his positive claim but he overstates his negative case. Showing that cultural development can occur without Darwinian cultural evolution or know-how copying does not mean that know-how copying is superfluous; or that parent–offspring cultural lineages are rare.

Sterelny acknowledges that imitation (we prefer ‘know-how copying’¹Buskell, A. and Tennie, C. [2025]: ‘Mere Recurrence and Cumulative Culture at the Margins’, British Journal for the Philosophy of Science, 76, available at <doi.org/10.1086/717776>.; henceforth: copying) may be necessary for peripheral social rituals like dancing, and that copying may create parent–offspring lineages in these cases. He mostly downplays the roles of copying in the cases of skill mastery central to his argument. However, copying is not only necessary in cases where traits evolve in Darwinian ways. Many human traits depend on copying more indirectly. For example, while Sterelny acknowledges that language contributes to the construction of the social niches that facilitate craftsmanship, he glosses over how languages must be learned. Copying is plausibly necessary for natural language acquisition,²Tennie, C., Call, J. and Tomasello, M. [2009]: ‘Ratcheting up the Ratchet: On the Evolution of Cumulative Culture’, Philosophical Transactions of the Royal Society B, 364, pp. 2405–15.
Moore, R. [2013]: ‘Imitation and Conventional Communication’, Biology and Philosophy, 28, pp. 481–500. and by extension for every trait that depends on it. In spoken languages, if speakers cannot accurately copy words, they won’t have a usable language. For this they must reproduce the sounds used by others. These are often arbitrary, and so unlikely to be produced without copying.³Moore, R. [2013]: ‘Imitation and Conventional Communication’, Biology and Philosophy, 28, pp. 481–500.

Copying is often thought central to human cultural origins precisely because it can be non-verbal, is present in prelinguistic infants, and is largely absent in great apes.⁴Tomasello, M. [2008]: Origins of Human Communication, Cambridge, MA: MIT Press. Of course, copying alone is not sufficient for natural language acquisition. Learners must also understand the goals towards which others use words to understand their communicative functions. Other factors (for example, knowledge of syntax) may play necessary or facilitating roles, too. However, while not sufficient, copying is likely necessary for language—and so for all things dependent on it. Additionally, while languages are unlikely to be learned from single models, words may be, and in one or few shots.⁵Bloom, P. [2000]: How Children Learn the Meanings of Words, Cambridge, MA: MIT Press.

Sterelny’s prime examples concern not language but skill mastery. Again, he downplays the roles of copying, we think prematurely. Empirically it is established that some kinds of cultural learning occur unreliably for agents who are poor at copying, yet reliably for agents who are good at it.⁶Tennie, C., Call, J. and Tomasello, M. [2009]: ‘Ratcheting up the Ratchet: On the Evolution of Cumulative Culture’, Philosophical Transactions of the Royal Society B, 364, pp. 2405–15. Where copying is present, it need not entail that skill acquisition is rapid, or the presence of single-model student–teacher apprenticeships. Still, even learning from multiple models can involve copying (for example, learning a fan chant from a sea of singing fans). Where cultural traits differ across multiple models, copying cannot be excluded per se, since learners may specifically copy only parts of the whole trait from different individuals.⁷Godfrey-Smith, P. [2009]: Darwinian Populations and Natural Selection, Oxford: Oxford University Press. So there may be many cases where copying plays a crucial role, even if this is not salient. Sterelny concludes that copying is largely unnecessary for skill mastery, but he shows only that it is insufficient and not always salient. Sufficiency was never the issue, however, and salience is not a requirement.

Sterelny extends his denying a major role for copying deep into human evolution:

Productive, robust response to the unpredictable and idiosyncratic properties of stone has a deep history, as deep as the Acheulean. For Acheulean handaxes cannot be made by learning and applying a fixed procedure. Raw material is too variable; artisans have to adjust their procedures to the shape and composition of the cobble on which they are working.

Since the mastery of technical and social skills can depend on copying even in non-obvious ways, the roles of copying must be investigated empirically. It is often a matter of controversy.⁸Koops, K., Biro, D., Matsuzawa, T., McGrew, W. C. and Carvalho, S. [2023]: ‘Appropriate Knowledge of Wild Chimpanzee Behavior (“Know-What”) and Field Experimental Protocols (“Know-How”) Are Essential Prerequisites for Testing the Origins and Spread of Technological Behavior’, Animal Behavior and Cognition, 10, pp. 163–68. Such investigations indeed point away from a major role for copying in early Acheulean handaxe production.⁹Tennie, C., Braun, D. R., Premo, L. S. and McPherron, S. P. [2016]: ‘The Island Test for Cumulative Culture in Paleolithic Cultures: The Nature of Culture’, in M. N. Haidle, N. J. Conard and M. Bolus (eds), Vertebrate Paleobiology and Paleoanthropology, Dordrecht: Springer, pp. 121–33. Copying also seems to play little role either in earlier hominin or great ape tool use too, with ever fewer ape cases debated.¹⁰Gunasekaram, C., Battiston, F., Sadekar, O., Padilla-Iglesias, C., van Noordwijk, M. A., Furrer, R., Manica, A. et al.  [2024]: ‘Population Connectivity Shapes the Distribution and Complexity of Chimpanzee Cumulative Culture’, Science, 386, pp. 920–25. This is consistent with data showing unenculturated, untrained apes are generally poor at copying.¹¹Tennie, C., Call, J. and Tomasello, M. [2009]: ‘Ratcheting up the Ratchet: On the Evolution of Cumulative Culture’, Philosophical Transactions of the Royal Society B, 364, pp. 2405–15.
Berio, L. and Moore, R. [2023]: ‘Enculturation and Cognition: Learning from Great Apes’, Biology and Philosophy, 38, available at <doi.org/10.1007/s10539-023-09908-y>.

In stark contrast, late handaxes are more likely to depend upon copying. These cases (currently) mark the beginnings of copying in the archaeological record, around 800 kya.¹²Tennie, C. [2023]: ‘The Earliest Tools and Cultures of Hominins’, in J. Tehrani, J. Kendal and R. Kendal (eds), Oxford Handbook of Cultural Evolution, Oxford: Oxford University Press, available at <doi.org/10.1093/oxfordhb/9780198869252.013.33>. Again, copying was not the only mechanism at play, and the identifiable contribution of copying varies even across Late Acheulean handaxes. Late Acheulean tools at Boxgrove (500 kya) stand out,¹³Tennie, C. [2023]: ‘The Earliest Tools and Cultures of Hominins’, in J. Tehrani, J. Kendal and R. Kendal (eds), Oxford Handbook of Cultural Evolution, Oxford: Oxford University Press, available at <doi.org/10.1093/oxfordhb/9780198869252.013.33>. particularly given the specifics of the tools that were used to make them.¹⁴Putt, S. S., Anwarzai, Z., Holden, C., Ruck, L. and Schoenemann, P. T. [2024]: ‘The Evolution of Combinatoriality and Compositionality in Hominid Tool Use: A Comparative Perspective’, International Journal of Primatology, 45, pp. 589–634. Still, whether copying played a major role is an empirical question not yet settled. The modern skills Sterelny describes may have required copying to get off the ground, in the current generations of learners and/or in their forebears.

Implicitly, Sterelny allies himself with the ‘Paris school’ of cultural evolutionists who,¹⁵Sterelny, K. [2017]: ‘Cultural Evolution in California and Paris’, Studies in History and Philosophy of Biological and Biomedical Sciences, 62, pp. 42–50. unlike ‘California’ theorists, downplay the role of copying in human culture. Paris theorist Olivier Morin writes: ‘Humans […] have no particular aptitude for cultural transmission […] We do not need to speculate that we developed one adaptation to culture, or one imitative faculty, which set cultural history in motion’.¹⁶Morin, O. [2015]: How Traditions Live and Die, Oxford: Oxford University Press, p. 214. For the reasons described above, we are sceptical. Morin asks us to imagine that human culture could be the same without copying, thus, without languages (among other things).¹⁷Morin, O. [2015]: How Traditions Live and Die, Oxford: Oxford University Press, p. 11. Sterenly’s position is more nuanced. Nonetheless, both overlook the subtle, yet pervasive, roles of copying in human culture (and niche construction). Even if not always salient, its contribution to historically significant human skill sets may be necessary. If Sterelny understates the contributions of copying to human culture, he may also underestimate the presence of cultural replicators in our learning environment—and of Darwinian lineages in cultural domains.

Richard Moore & Claudio Tennie