Plato’s account of Atlantis in the Timaeus and Critias has been read for a century and a half as describing an island in the Atlantic Ocean. That reading rests on rendering the Greek Atlantikon pelagos as “Atlantic Ocean.” Plato distinguishes pelagos, a partly land-enclosed sea, from okeanos, the encircling world-ocean he uses elsewhere, and he describes the entrance to this pelagos as navigable in the distant past but silted in his own era – a description that cannot apply to the Strait of Gibraltar. G. Sarantitis has argued that the passage instead points to North Africa. This paper summarises the presentation given at IKEK-17, which examined that argument against the palaeoclimatic record of the African Humid Period. During the last Green Sahara, monsoon and Mediterranean rainfall belts converged over a now-hyperarid region, supporting lakes, rivers, grasslands and large-bodied fauna across an area the size of the United States. Marine and continental records indicate that such episodes are not anomalies: over 230 Green Sahara Periods have been identified within the last eight million years, paced by orbital forcing. The presentation drew two conclusions. First, the geography implied by the myth corresponds to a landscape that demonstrably existed, which reframes the question from whether the account was invented to how much of it may preserve memory. Second, and more consequentially for climate discourse, the historical record shows no consistent mapping between the direction of climate change and human outcomes: the drying of the Eastern Sahara dispersed populations and also concentrated them along the Nile, where Pharaonic Egypt arose. Both the Victorian mistranslation and the centurylong delay in interpreting Saharan rock art illustrate a recurring failure – assuming that present conditions are the baseline against which the past must be read.
