A cross-cultural investigation of young children's spontaneous invention of tool use behaviours

Abstract

Through the mechanisms of observation, imitation and teaching, young children readily pick up the tool using behaviours of their culture. However, little is known about the baseline abilities of children's tool use: what they might be capable of inventing on their own in the absence of socially provided information. It has been shown that children can spontaneously invent 11 of 12 candidate tool using behaviours observed within the foraging behaviours of wild non-human apes (Reindl et al . 2016 Proc. R. Soc. B 283 , 20152402. ( doi:10.1098/rspb.2015.2402 )). However, no investigations to date have examined how tool use invention in children might vary across cultural contexts. The current study investigated the levels of spontaneous tool use invention in 2- to 5-year-old children from San Bushmen communities in South Africa and children in a large city in Australia on the same 12 candidate problem-solving tasks. Children in both cultural contexts correctly invented all 12 candidate tool using behaviours, suggesting that these behaviours are within the general cognitive and physical capacities of human children and can be produced in the absence of direct social learning mechanisms such as teaching or observation. Children in both cultures were more likely to invent those tool behaviours more frequently observed in great ape populations than those less frequently observed, suggesting there is similarity in the level of difficulty of invention across these behaviours for all great ape species. However, children in the Australian sample invented tool behaviours and succeeded on the tasks more often than did the Bushmen children, highlighting that aspects of a child's social or cultural environment may influence the rates of their tool use invention on such task sets, even when direct social information is absent.

The zone of latent solutions and its relevance to understanding ape cultures

“…Yet, there is empirical evidence that this zone also exists for humans: Reindl et al ( 2016 ) provide data on some of the tool-use behavioural forms that are within human children’s ZLS that do not seem to require copying (as they can be reinnovated by task-naïve children (note that the validity of this conclusion was increased by taking tasks from different ape species)). This finding has now been replicated and extended cross-culturally, which further demonstrates that it does not depend on specific cultural—including form—backgrounds (Neldner et al 2020 ). Thus, the ZLS hypothesis completes Vygotsky’s original ZAD and ZPD concepts by providing a missing third zone: the ZLS (as the largely phylogenetically-based and form-copying-independent baseline of the ZAD).…”

“…see the approach used by Reindl et al 2016 ). Validity is again further increased when the specifics of the surrounding cultural (including its contained forms) and environmental settings does not change the results in substantial ways (see the approach used by Neldner et al 2020 ). 35 Finally, for many modern human copying-dependent forms, history provides ample valid natural latent solution experiments.…”

“…donut + croissant = cronut 57 )—we call this latter case indirect copying of forms (or recombination, in Muthukrishna and Henrich’s 2016 terminology). Below we will first explore the case of direct copying, before turning to indirect copying (see also Neldner et al 2020 for a similar differentiation).…”

“…However, there are also examples in humans. Not only can young children spontaneously reinnovate ape behavioural forms of tool use (Reindl et al 2016 ; Neldner et al 2020 ), but various form phenomena reliably occur and stabilise at the end of so-called transmission experiments in humans without ever having been inputted at start (e.g. from simple drawn data patterns to lab-based language structures (compare Kirby et al 2008 )).…”

“…While this may occasionally even be the case also in modern humans (e.g. yawning forms, but also some basic tool use; Reindl et al 2016 ; Neldner et al 2020 ), the combined likelihood of this happening for millions of specific human form phenomena (e.g. all books ever written, all machines ever built, all songs ever sung) can likely be ignored.…”