The Island Test for Cumulative Culture in Paleolithic Cultures

Research output: Chapter in Book/Report/Conference proceedingChapter (peer-reviewed)


Colleges, School and Institutes

External organisations

  • University of Cape Town


Early Stone Age artifacts have long been assumed to reflect the material record of communities whose members possessed the ability to transmit ideas, behaviors, and technologies from individual to individual through high-fidelity transmission (i.e., involving teaching and/or imitation) , much like humans do today. Recent experimental work has highlighted marked differences between great apes and modern humans in the capacity and/or motivation for some forms of cultural transmission . In particular, high-fidelity mechanisms of social learning , which are thought to underlie the capacity for cumulative culture , appear to be enhanced in – if not unique to – humans. Taken as a group , these experiments suggest it is plausible that a combination of genetic, environmental, and social factors that do not include high-fidelity social learning mediate the “cultures” described for great ape populations to date. It may be that, while the distribution of great ape behavioral variation in time and space is likely affected by low-fidelity social learning (which is widespread in the animal kingdom), the observed variants were invented (i.e., learned) independently by each individual rather than copied from other individuals. Behaviors that do not require high-fidelity transmission between individuals in order to increase in frequency in a population lie within the so-called “zone of latent solutions .” Here, we begin to grapple with the hypothesis that much of the Early Stone Age archaeological record may reflect deeply “canalized” behaviors of hominin toolmakers – those that reside in each individual’s zone of latent solutions – rather than behaviors that necessarily require high-fidelity transmission between individuals. We explore this possibility while eschewing the simplistic notion that variation in stone tool shape, for example, is entirely determined by the genetic variation found in the toolmakers. Instead, we suggest that the variation observed in Early Stone Age artifacts may simply reflect a heavier reliance on behaviors that reside within the zone of latent solutions than on behaviors that make use of high-fidelity social learning . We discuss a thought experiment, called the Island Test , which may be useful for distinguishing hominin behaviors that require high-fidelity transmission from behaviors that do not. We conclude that the Early Stone Age archaeological record is consistent with the possibility that latent solutions explain the behavioral variation inferred from available material culture. Furthermore, we explore reasons why the assumption of high-fidelity transmission associated with Paleolithic industries is difficult to support.


Original languageEnglish
Title of host publicationThe Nature of Culture
Subtitle of host publicationBased on an Interdisciplinary Symposium ‘The Nature of Culture’, Tübingen, Germany
EditorsMiriam N Haidle, Nicholas J Conard, Michael Bolus
Publication statusPublished - 20 Jan 2016

Publication series

NameVertebrate Paleobiology and Paleoanthropology
ISSN (Print)1877-9077


  • Behavior, Apes, Early Stone Age, Oldowan, Acheulean, Tools, Social learning, High-fidelity transmission