What is most alarming is that we have no idea—none—what natural phenomena could so swiftly rattle Earth's thermometer. As Elizabeth Kolbert, writing in the New Yorker, has observed: "No known external force, or even any that has been hypothesized, seems capable of yanking the temperature back and forth as violently, and as often, as these cores have shown to be the case." There seems to be, she adds, "some vast and terrible feedback loop," probably involving the oceans and disruptions of the normal patterns of ocean circulation, but all this is a long way from being understood.

One theory is that the heavy inflow of meltwater to the seas at the beginning of the Younger Dryas reduced the saltiness (and thus density) of northern oceans, causing the Gulf Stream to swerve to the south, like a driver trying to avoid a collision. Deprived of the Gulf Stream's warmth, the northern latitudes returned to chilly conditions. But this doesn't begin to explain why a thousand years later when the Earth warmed once again the Gulf Stream didn't veer as before. Instead, we were given the period of unusual tranquility known as the Holocene, the time in which we live now.

There is no reason to suppose that this stretch of climatic stability should last much longer. In fact, some authorities believe that we are in for even worse than what went before. It is natural to suppose that global warming would act as a useful counterweight to the Earth's tendency to plunge back into glacial conditions. However, as Kolbert has pointed out, when you are confronted with a fluctuating and unpredictable climate "the last thing you'd want to do is conduct a vast unsupervised experiment on it." It has even been suggested, with more plausibility than would at first seem evident, that an ice age might actually be induced by a rise in temperatures. The idea is that a slight warming would enhance evaporation rates and increase cloud cover, leading in the higher latitudes to more persistent accumulations of snow. In fact, global warming could plausibly, if paradoxically, lead to powerful localized cooling in North America and northern Europe.