“DISSECTING” QUARTZITE AND BASALT BIPOLAR FLAKE SHAPE: A MORPHOMETRIC COMPARISON OF EXPERIMENTAL REPLICATIONS FROM OLDUVAI GORGE, TANZANIA

Abstract

In this pilot study we explore the relationship between bipolar flake shape and stone raw material differences. We conducted a morphometric analysis of 40 quartzite and 40 basalt experimentally replicated specimens from Olduvai Gorge, Tanzania. We carried out two sets of statistical analyses to investigate the shapes of quartzite and basalt bipolar flakes. Our first set of analyses focused on statistical comparisons of five morphometric variables recorded on the two samples of bipolar flakes. For the second set of analyses we used principal components analysis (PCA) to investigate shape differences among the quartzite and basalt bipolar flakes. Our results suggested that there are no significant differences amongst the quartzite and basalt samples. These results are consistent with the hypothesis that the close association between quartzite and bipolar reduction during the Lower Paleolithic at Olduvai Gorge is due to a single motivating factor, namely, expediency. However, we emphasize that more experiments altering test parameters and variables are needed, as are direct quantitative comparisons between experimental and archaeological datasets. Future experiments may uncover additional motivating factors for the prominent quartzite-bipolar reduction found at Olduvai Gorge.

Dans cet article on explore le rapport entre la forme de l’écaille faite de la taille sur enclume et les différences des matières premières de pierre. Nous avons fait une analyse morphométrique de 40 spécimens bipolaires de quartzite et de 40 de basalte qui étaient répliqués expérimentalement et qui sont d’origine d’Olduvai Gorge, Tanzanie. Nous avons dirigé deux analyses statistiques pour étudier la forme. La première s’est concentrée sur la comparaison statistique de cinq variables morphométriques trouvées sur les deux exemples des écailles bipolaires. Dans la deuxième analyse nous avons utilisé le style PCA (analyse des éléments principaux) pour étudier les différences de forme dans le quartzite et le basalte. Ces résultats sont en rapport avec l’ hypothèse que le proche rapport entre le quartzite et la réduction bipolaire pendant l’ère paléolithique dans Olduvai Gorge est causé par un seul facteur, l’opportunité. Néanmoins, nous soulignons le fait qu’il faut plus d’expériences qui changeront les paramètres et les variables, aussi bien que des comparaisons quantitatives et directes entre les ensembles des données expérimentales et archéologiques. C’est bien possible que les expériences dans l’avenir découvriront plus de facteurs motivationnelles pour l’association bipolaire-quartzite connue à Olduvai.

Keywords:

• Bipolar reduction
• Quartzite
• Basalt
• Experiment
• “Lithic Morphometric Comparative Anatomy
• ” Olduvai Gorge
• Tanzania
• Lower Paleolithic

Acknowledgments

Financial support for this research was provided by the British Academy Grant “Can Olduvai Gorge Shed Light on the Middle Stone Age”? (SG112454) and a Leverhulme Trust Early Career Fellowship (ECF-2011–567). MIE is currently supported by a University of Missouri College of Arts and Sciences Post-Doctoral Fellowship. We are grateful to Justin Pargeter and Hilary Duke for inviting us to participate in the SAA Symposium from which this paper originated. We are also appreciative to the anonymous reviewers and to Grant McCall.

Notes

1 In a recent article Hiscock (2015) expressed concern that “weak” definitions of bipolar reduction used by us and others (Gurtov and Eren 2014:285; also Eren 2010; Eren et al. 2013; Diez-Martin et al. 2011:692) — which demand minimal procedural similarity for inclusion — have been used over “stronger” definitions. We disagree with Hiscock's attributions and suggest that the “weak” versus “strong” dichotomy is more precisely characterized as a “broader” versus “more specific” dichotomy. We also suggest that the choice to use a broader or more specific definition depends on the question being asked, and does not preclude the use of the alternative (i.e. broader or more specific) in future analyses. Indeed, the juxtaposition of results coming from analyses that use broader versus more specific definitions may itself be interesting or informative depending on the question. In the cases that Hiscock (2015) highlights, the topics of study do not necessarily require a specific definition of “bipolar,” as they are concerned with comparing “bipolar” more generally (knapping with use of an anvil) with “freehand” percussion (knapping without use of an anvil). The choice of a broader or more specific definition also depends on whether researchers have the archaeological resolution to distinguish between different types of bipolar (striking a core directly at 90° versus simply resting a core on an anvil before hitting it). Indeed, it would be irresponsible to use a specific definition of bipolar in archaeological analyses when, to our knowledge, the ability of archaeologists to morphometrically identify and distinguish between different types of bipolar procedures suggested by Hiscock does not currently exist. In terms of the experimental dataset used in this paper and in Gurtov and Eren (2014), bipolar reduction did indeed operationally aim to fit Hiscock's (2015) specific definition, namely “a hammer strikes a rock which is resting on an anvil, and strikes into it at close to 90° and in line with the point at which the rock is in contact with the anvil.” However, the experimental dataset was created in the field, and we did not possess a video recording or some other means with which we could control for, or confirm later, that indeed every blow was close to 90°, despite that being the knapper's goal. Further, Hiscock's definition is vague (i.e., how close to 90° is “close”?) and therefore difficult to operationalize. Given these issues, we felt it was more conservative to use a broader definition of bipolar in this study and in Gurtov and Eren (2014:285), “Bipolar reduction can be defined as a percussion technique in which a stone core is placed on an anvil and struck with a hammer to produce flakes”).

2 It is noteworthy that because the experimental dataset used here was created for another analysis, in some respects its use here acts as a “blind” test. Blind testing has yet to be regularly practiced in stone tool replication experiments. However, it is important to understand that blind testing is merely another type of experimental variable control whose use depends entirely on the question being asked. Sometimes the use of blind and non-blind replication experiments in conjunction may be the most productive experimental strategy. Other times use of “blind” stone tool replicated assemblages (e.g. see discussion of Tactikos (2003) in Eren et al. (2008)) or blind tests may confound results, in some cases it may be the most robust research strategy, while in still other situations it may not matter one way or another.

3 During the revision of this manuscript, an experimental analysis of novice knappers using chert was published by Putt (2015). Her results were consistent with Gurtov and Eren's (2014) results that bipolar reduction is more expedient than other forms of knapping.

Additional information

Notes on contributors

Alia N. Gurtov

Alia N. Gurtov is a Ph.D. candidate at the University of Wisconsin, Madison, WI.

Briggs Buchanan

Briggs Buchanan is an Assistant Professor at the University of Tulsa, Tulsa, OK.

Metin I. Eren

Metin I. Eren is the University of Missouri College of Arts and Sciences Post-Doctoral Fellow, Columbia, MO.