Many years ago when I was in a grade-eight math class, I was sitting looking out the windows at the dinosaurs playing. Ok, despite what my daughter thinks, I am not quite that old. What I was looking at was planes circling around in the distance. It turns out that a plane had crashed. It was a Handley Page HPR-7 Herald 211 operated by Eastern Provincial Airlines and all eight people on board were killed. Now, it is sometimes claimed that science cannot explain the past. It’s even argued that historical sciences like paleontology, archeology, and cosmology, somehow use different methods of discovering the past, than say, determining the reason of a plane crash and that is again different from the method for discovering the laws of nature. In reality, the methods are all the same.
I suppose, in response to the plane crash, people could have sat around and made predictions for future plane crashes but instead they used science to try to discover the past—what had caused the plane to crash. In this case it turned out to not be so difficult. The Aviation Safety Network describes the cause thus: Failure of corroded skin area along the bottom centre line of the aircraft beneath stringer No.32 which resulted in structural failure of the fuselage and aerial disintegration. This was found out by a metallographic examination which provided clear evidence of stress corrosion in the aluminum alloy. The planes of this type that were remaining in service were repaired to prevent them from crashing as well.
The approach to understanding why the Eastern Provincial Airline’s plane had crashed followed a similar approach to any other plane crash: you analyse the debris, gather records from the black box and whatever other information is available, and construct a model for what happened. You test the model by making predictions for future observations; for example, that corrosion will be found on other planes of the same type. This sounds very much like the standard scientific method as proposed originally by Roger Bacon (1220 – 1292) and followed by scientists ever since: observe, hypothesize, test, rehypothesize, and repeat as necessary.
The same technique is used for any reconstruction of the past, be it plane crashes, the cause of Napoleon’s death, archeology, paleontology, evolution, and cosmology. The cause of Napoleon’s death is quite interesting as an exercise in forensic science. The original cause of death was suggested to be gastric cancer. But that is too mundane a cause of death for such an august figure. So the conspiracy advocates went to work and suggested he was poisoned by arsenic. How to test? Easy look for arsenic in samples of hair. Well, that was done and arsenic was found. Case closed? Not quite. Were there other sources of arsenic than deliberate poisoning? Yes, the wall paper in his room had arsenic in it. Also further investigation revealed that he had been exposed to arsenic long before he went to St. Helena. In support of the caner hypothesis his father also died of stomach cancer. The current consensus is that the original diagnosis was correct. He died of stomach cancer. But notice the play of events: hypothesis—arsenic poisoning, testing—look for arsenic in hair samples, refine hypothesis—check for other sources of arsenic, etc. We can see here the classic process of science being played out in reconstructing the past.
We can continue this technique into the more distant past: When did humans evolve? Why did the dinosaurs die out? How did the earth form? How did the solar system form? What if anything preceded the big bang? All of these questions can be tackled using the standard methods of science. Observations of present tell us about the past, counting tree ring tells us when the tree started to grow.
The interplay between what might be called natural history and natural laws is very intricate. We must interpret the past in order to extract the natural laws and use the natural laws to interpret the past. All our models of science have, explicitly or implicitly, both an historical and a law component. In testing a model for how the universe works—ie to develop the laws—we conduct an experiment. Once the experiment is finished, it becomes history and interpreting it is historical science. For example, why did the OPERA experiment claim to see faster than light propagation for neutrinos? Or is the bump seen in searches for the Higgs boson real or an artifact of the detector? Those investigations are as much forensic science as trying to decide why Napoleon died or the dinosaurs went extinct. Thus, all science is historical and sometimes, quite explicitly. Einstein abandoned the cosmological constant based on an alternate model for the history of the universe, namely that it is expanding rather than static.
So, we have science as a unified whole, encompassing the past, present, and future; the natural laws entangled with the natural history. But what about the dinosaurs I did not see out of the math-room windows? We can be quite sure they did not exist at that time and that Fred Flintstone did not have one as a pet (a saber-toothed pussy cat is another story). The study of evolution is much like that for plane crashes. You study the debris, in the case of evolution that “debris” includes fossils and the current distribution of species. Consider the fossil Tiktaalik roseae, a tetrapod-like fish or a fish-like tetrapod, that was found a few years ago. One can engage in futile semantic arguments about whether it is a fish, or a tetrapod, or a missing link, or whether it is the work of the devil. However, the significant point is that a striking prediction has been confirmed by a peer-reviewed observation. Using evolution, a model of fossil formation, and a model of the earth’s geology, a prediction was made that a certain type of fossil would be found in a certain type of rock. Tiktaalik roseae dramatically fulfilled that prediction and provides information on the fish-tetrapod transition.
The cause of plane crashes, Napoleon’s death, evolution, and the extinction of dinosaurs can all be explored by using the same empirically-constrained model-building techniques as the rest of science. There is only one scientific method.
To receive a notice of future posts follow me on Twitter: @musquod.