Why Tornados Running Backward Do Not Violate the Second Law of Thermodynamics
In the 11 years since I wrote a letter to the editor of the Mathematical Intelligencer titled "Can ANYTHING Happen in an Open System?" I'm sure I have heard at least 100 different reasons why the spontaneous rearrangement of atoms on a barren planet into intelligent brains, libraries full of science texts and encyclopedias, jet airplanes, and computers connected to keyboards, LCDs, laser printers and the Internet does not violate the second law of thermodynamics. For a more technical response, see here and here.
I have recently noted that if we took a video of a tornado sweeping through a town, and showed the video backward, every one of these arguments could equally be used to argue that there was no conflict between the second law and what we were witnessing on the video. So here is a list of the top 5 reasons, taken from my notes of the last 11 years, why a tornado running backward does not violate the second law:
(1) The Earth is an open system, tornados derive their energy from the sun, and while turning rubble into houses and cars represents a decrease in entropy, the increase in entropy outside the Earth far exceeds the decrease seen in this video, so there is no conflict with the second law. This is the traditional argument used by Asimov, Dawkins and many others, and it is the one I have been primarily criticizing, particularly in my AML paper, "A Second Look at the Second Law." My arguments seem to have been effective, because I rarely hear this silly argument any more. Critics seem to have been forced more and more to fall back on secondary objections now. Since I have dealt with this "compensation" argument thoroughly in the above links, I will deal mainly with the secondary objections here.
(2) The second law only applies to thermal entropy, and what is happening in this video does not result in a net decrease in thermal entropy, so there's no conflict with the second law. While the first formulations were all about heat, it is now universally recognized that the second law of thermodynamics can be used, in a quantifiable way, in many other applications. For example, thermal entropy is just a measure of disorder in the distribution of heat as it diffuses, you can define an "X-entropy" to measure disorder in the distribution of any other diffusing component X, and, as I pointed out in my AML paper, these X-entropies are defined by the same equations as thermal entropy, and are equally quantifiable. The second law is all about probability; "carbon entropy" increases in an isolated system for the same reason thermal entropy increases: because that is what the laws of probability predict.
(3) What is happening in this video is too ill defined and too difficult to quantify for the second law to apply. Once you point out that the second law applies, in a quantifiable way, not only to thermal entropy, but to other types as well, the next line of defense is usually to argue that applying it to increases in less quantifiable types of "entropy," such as books burning or wine glasses breaking, is not scientific, even though this is routinely done in physics textbooks. While there are situations where things are so difficult to quantify that it is hard to say what the second law predicts, in other situations -- a tornado running backward, for example -- it is easy. Some things are obvious even if they are difficult to quantify!
(4) Sure, tornados turning rubble into houses and cars is extremely improbable, but natural forces do extremely improbable things all the time; every time we flip a billion coins, the exact outcome is extremely improbable. What the second law prevents is not extremely improbable things, but macroscopically describable things that are extremely improbable from the microscopic point of view. If we flip a billion coins and get an extremely improbable, simply describable result, such as "all heads," then we have reason to be astonished.
(5) There is no conflict with the second law in this video, because the second law only applies to isolated systems, period. It is true that even the later, more general, formulations of the second law, stated in terms of "order" and "disorder," all begin with "in an isolated system," so it can be argued that, technically, a tornado running backward would not violate even these formulations of the second law. But if Isaac Newton had never generalized his law of gravity beyond "the Earth attracts apples," would we say that, technically, the law of gravity does not apply to oranges? There is an obvious generalization of the second law to open systems, which I stated in my AML paper as a tautology: "if an increase in order is extremely improbable when a system is isolated, it is still extremely improbable when the system is open, unless something is entering which makes it not extremely improbable," and illustrated quantitatively using the "X-entropies" mentioned above. And surely tornados turning rubble into houses and cars is still extremely improbable, even if the Earth does receive energy from the sun.
So, if we saw a video of a tornado, running backward, would we conclude that the second law was being violated by what was happening or not? According to many physics textbooks, such as the Ford text quoted in my video "Evolution is a Natural Process Running Backward" (above), the answer is yes. In any case, if we actually watched a video of a tornado, running backward, it would certainly not occur to us to make any of the above arguments to claim that what we were seeing did not technically violate the second law, as formulated in physics textbooks. We would immediately recognize that what we were seeing violated a fundamental law of Nature, whether it violated the manmade formulations of this law or not. And even if we were told that what actually happened took a long time and the video had been speeded up, we would still not be interested in anyone's "scientific" explanation for what we were seeing in the video; we would immediately recognize that the video must be running backward, because what we were seeing was completely unnatural.
So, how does the spontaneous rearrangement of matter on a rocky, barren, planet into human brains and spaceships and jet airplanes and nuclear power plants and libraries full of science texts and novels, and super computers running partial differential equation solving software, represent a less obvious or less spectacular violation of the second law -- or at least of the fundamental natural principle behind this law -- than tornados turning rubble into houses and cars? Here is a thought experiment for you: try to imagine a more spectacular violation than what has happened on our planet.