Reflections on Pleistocene island occupation

Nellie Phoca-Cosmetatou & Ryan J. Rabett Journal of Mediterranean Archaeology 27.2 (2014), 255-259

From the introduction

We are extremely pleased that the topic of hominin island occupation has attracted particular attention in this issue of JMA in the two fascinating and contrasting papers by Runnels and Leppard. These two papers set out to address similar questions, namely: ‘was there an early Palaeolithic human presence on the Mediterranean islands?’ and ‘why has it been so elusive?’ Even so, they come to diametrically opposed conclusions: e.g. ‘the Mediterranean-and by implication other seas-were at times open roads rather than barriers to hominin dispersals’ vs ‘A Mediterranean awash with seagoing archaic hominins is, then, inherently unlikely’.

Why is this? We do not believe the reason to be the different datasets analysed by each author. Rather, such discrepancies are the consequence of a topic still in its infancy. There is great scarcity of reliable evidence of early hominin presence on the Mediterranean islands. As a result, the debate on Pleistocene island occupation often gets caught in a priori expectations that different researchers might have about past hominin behaviour and seagoing abilities. As research into Pleistocene island occupation is currently gathering momentum, it is expected that targeted fieldwork coupled with sophisticated theoretical models will lead to big leaps in our knowledge over the next few years-and quite possibly a subsidence of contrasting views.

In this commentary, we focus on three issues that permeate the Pleistocene island occupation debate as highlighted in the Runnels and Leppard articles. The first of these issues relates to the substantiation of the stone tool evidence, and thus the reliability of the archaeological signature uncovered so far. The second centres around the frequency, and thus the significance, of hominin island presence. The third refers to future directions of research and the application of appropriate models.