Parker's book on the topic is a good one-volume cover of the history from a global perspective. The climate data are pretty clear on how weird things were, with

Parker's book on the topic is a good one-volume cover of the history from a global perspective. The climate data are pretty clear on how weird things were, with no shortage of "most extreme ever" events from China to England: http://www.amazon.com/Global-Crisis-Climate-Catastrophe-Seventeenth/dp/0300208634/ref=sr_1_1?ie=UTF8&qid=1434393062&sr=8-1&keywords=global+crisis http://www.amazon.com/Global-Crisis-Climate-Catastrophe-Seve... Concepts--all concepts--are made things, and the edges we draw around them are the edges of our attention. People draw those edges in different places, or not at all, depending on their priors. As such, a certain type of person will argue that there wasn't "really" a "general criss" in the 17th century because the panoply of events aren't enough to trigger their edge-detectors, and they feel that their edge-detectors are arbiters of reality. Arguments of this nature are rarely productive. The events happened regardless of what you call them. That the global climate underwent a statistically significant excursion in the 17th century is hard to argue against. That that excursion was also economically significant is likewise difficult to argue against. Unprecedented multi-year runs of poor harvests aren't really something that can be wished away with ideology. Finally, that the economic excursion helped drive political decision making should be relatively uncontroversial, unless one were to argue that politics happens in an economic vacuum. Parker does a good job of emphasizing the contingency of outcomes. People made decisions for reasons that seemed good to them--often of the form "God loves my country, so if I do X, Y will necessarily and certainly happen!"--and results followed for unrelated reasons (often of the form, "Luck.") He also argues that something like 1/3 of the human species died in the crisis, which is fairly astonishing, but given the death tolls for the places where we have data, not entirely unsupportable.