Some people will look at a historical event and draw definitive conclusions from it. Irrespective of whether they are correct or not in their conclusion, the methodology is broken.
Compare and contrast a scientific experiment with a historical event. If you've ever taken a class with a laboratory component, you know how much effort and precision goes into the design of experiments: they must be carefully controlled so that you can be sure that the effect you are seeing has the cause that you are hypothesizing. Famous examples of mistaken experiments abound, and these are mistakes that are very subtle and yet completely alter the conclusion that can be drawn from the experiment.
Now contrast this with the process of reading about a historical event and attempting to draw a definitive conclusion about cause and effect from it: nothing is controlled, there is only one data point. Could you imagine trying to submit a paper proving Newton's law of gravity not by actually running experiments in which you drop things (carefully!) and measuring them (carefully!), but instead by reading about someone dropping something in a history book??
History is great for two things: for hypothesis creation, and for negative proof via counter-examples. It is terrible for hypothesis *proving*, however.
Statistician Paul Holland (1986) cautions that there can be “no causation without manipulation,” a maxim that would seem to rule out causal inference from nonexperimental data.This is a perfect example: if you don't get to manipulate an experiment to try to tease out the causal factors, you really can't make causal inferences. And yet, this is bread and butter in 99% of common discourse! Every talk radio, newspaper editorial, television advocacy program, politician, and blog lives on this. Just look, for example, at the number of definitive statements that have been made about the recent economic situation. Sure, there is interesting stuff to see here to try to formulate hypotheses that we can then go and analyze and experiment with in much more detail, but we cannot draw conclusions from this set of history, there are way too many uncontrolled variables.
The frustrating result is that 99% of what is said by many reasonably smart people is just wasted oxygen/pen/electrons. It's not that it's wrong, but that because its methodology is wrong, it adds nothing to the information about the subject.
Just one of the many things that frustrates me about the world I live in.
