Bronze Age Myths? Volcanic Activity and Human Response in the Mediterranean and North Atlantic Regions

Article excerpt

A search for precision beyond that currently available is a frequent aspect of archaeological interpretation. Tensions exist as a result of the need to resolve events on a human time-scale using techniques often incapable of producing such accuracy or precision. Dendrochronology, ice-core analysis and tephrochronology, where data-resolution can be constrained either by annual to sub-annual banding or precise isochrones, can make important contributions to tackling the persistent chronological problems in archaeology. In these interdisciplinary transfers there is always the danger that the necessary caution about the ways in which the data are used may be lost. This problem is particularly acute when the events being studied are real, or imagined catastrophes (cf. White & Humphreys 1994). Catastrophes - be they the destruction of Bronze Age Thera, the modern island of Santorini in the Aegean, or the apparent collapse of Middle Bronze Age settlement in upland Britain - are headline news; of such things myths and reputations are born and enter the literature as if proven fact. This paper examines some of the available evidence for these two Bronze Age 'catastrophes', the one real and in need of a calendar date, the other hypothesized on archaeological grounds and dated by a tenuous link through tree rings to an Icelandic volcano.


Since Marinatos (1939) connected a major eruption on Santorini, which destroyed the extensive Late Bronze Age town at Akrotiri, with the end of Minoan Crete, the date of this eruption has generated more discussion and controversy than perhaps any other cataclysmic event in prehistory. Initial archaeological considerations favoured a date close to c. 1500 BC (Renfrew 1990a), whilst later, calibrated radiocarbon dates tended towards the 17th century BC (Kuniholm 1990). In 1977, Hammer noted a correlation between acidity, measured by electrical conductivity in the Crete ice core from central Greenland, and the timing of volcanic events on a world scale. It was further suggested that one particular acidity peak lay sufficiently close to the archaeological evidence for the date of the eruption to be that generated by Santorini (Hammer et al. 1980). By the counting of annual layers of ice accumulation in the core, this provided a date of 1390[+ or -]50 BC. On the additional evidence of the Dye 3 core from southern Greenland, this was subsequently revised to 1645[+ or -]7 BC (Hammer et al. 1987). LaMarche & Hirschboeck (1984), working on tree rings from the American Southwest, had noted 'frost rings' (lines of severely retarded growth) which they associated with unseasonally cold conditions and correlated with major eruptions; on this basis, they suggested a date of 1628-1626 BC for the Santorini eruption. This was taken up by researchers at the Queen's University Palaeoecology Centre in Belfast, who sought volcanic impact in the extended oak chronology from Ireland. Baillie & Munro (1988) located a particularly narrow series of rings beginning in 1628 BC; influenced by Lamb's (1970) discussion of the impact of volcanoes on climate, they equated these with a stratospheric dust veil from Santorini. Despite several cautionary comments from both archaeologists (Manning 1988; Warren 1988) and geologists (Pyle 1989; 1990), the 1628 BC date, or one close to it, continues to be accepted (e.g. Michael & Betancourt 1988), without questioning why the effects of the Santorini eruption should be especially recognizable in the ice-core and tree-ring sequences. Large-scale explosive volcanic activity is common on a global scale (Zielinski et al. 1996), and so before accepting the possibility that the Santorini eruption can be recognized by unusual perturbations in the regional records of ice-cores or tree-rings, the case for its distinctive character must be proved. Despite the lack of critical assessment of these basic assumptions, Renfrew (1990b), in his summing up to the Third International Congress on Santorini in 1988, went so far as to suggest that, by the time of the next Congress, the date of the eruption would be unequivocally known to within one year. …