Rondam Ramblings: The Scientific Method, part 4: Eating elephants and The Big News Principle
This is the fourth in a series about the scientific method and how it can be applied to everyday life. In this installment I'm going to suggest a way to approach all the science-y stuff without getting overwhelmed. There is an old joke that goes, "How do you eat an elephant? One bite at a time." That answer might be good for a laugh, but it wouldn't actually work, either for a real elephant (if you were foolish enough to actually attempt to eat a whole elephant by yourself) nor for the metaphorical science elephant. Modern science has been a thing for over 300 years now, with many millions of people involved in its pursuit as a profession, and many millions more in supporting roles or just doing it as a hobby. Nowadays, over 100,000 scientific papers are published world-wide every single day. It is not possible for anyone, not even professional scientists, to keep up with it all. Fortunately, you don't have to consume even a tiny fraction of the available scientific knowledge to get a lot of mental nutrition out of it. But there are a few basics that everyone ought to be familiar with. For the most part this is the stuff that you learned in high school science class if you were paying attention. I'm going to do a lightning-quick review here, a little science-elephant amuse bouche. What I am about to tell you may all be old hat to you, but later when I get to the more interesting philosophical stuff I'll be referring back to some of this so I want to make sure everyone is on the same page. It may be tempting to skip this, especially if you grew up hating science class. I sympathize. Science education can be notoriously bad. It may also be tempting to just leave the elephant lying in the field and let the hyenas and vultures take care of it. The problem with that approach is that the hyenas and vultures may come for you next. In this world it really pays to be armed with at least a little basic knowledge. I'm going to make a bold claim here: what I am about to tell you, the current-best-explanations provided by science, are enough to account for all observed data for phenomena that happen here on earth. There are some extant Problems -- observations that can't be explained with current science -- but to find them you have to go far outside our solar system. In many cases you have to go outside of our galaxy. How can I be so confident about this after telling you that there is so much scientific knowledge that one person cannot possibly know it all? The source of my confidence is something I call the Big News principle. To explain it, I need to clarify what I mean by "all the observed data." By this I do not mean all of the data collected in science labs, I mean everything that you personally observe. If you are like most people in today's world, part of what you observe is that science is a thing. There are people called "scientists". There is a government agency called NASA and another one called the National Science Foundation. There are science classes taught in high schools and universities. There are science journals and books and magazines. The best explanation for all this is the obvious one: there really are scientists and they really are doing experiments and collecting data and trying to come up with good explanations for that data. This is not to say that scientists always get it right; obviously scientists are fallible humans who sometimes make mistakes. But the whole point of science is to find those mistakes and correct them so that over time our best explanations keep getting better and better and explain more and more observations and make better and better predictions. To see that this works you need look no further (if you are reading this before the apocalypse) than all the technology that surrounds you. You are probably reading this on some kind of computer. How did that get made? You probably have a cell phone with a GPS. How does that work? It's not hard to find answers to questions like "how does a computer work" and "how does GPS work" and even "how does a search engine work." Like everything else, these explanations are data which requires an explanation, and the best explanation is again the obvious one: that these explanations are actually the result of a lot of people putting in a lot of effort and collecting a lot of data and reporting the results in good faith. This is not to say that there aren't exceptions. Mistakes happen. Deliberate scientific misconduct happens. A conspiracy is always a possibility. But if scientific misconduct were widespread, if falsified data were rampant, why does your GPS work? If there is a conspiracy, why has no one come forward to blow the whistle? This is the Big News Principle: if any explanation other than the obvious one were true, then sooner or later someone would present some evidence for this and it would be Big News. Everyone would know. The absence of Big News is therefore evidence that no one has found any credible evidence against the obvious explanation, i.e. that there are in fact no major Problems with the current best theories. The name "Big News Principle" is my invention (as far as I know) but the idea is not new. The usual way of expressing it is with the slogan "extraordinary claims require extraordinary evidence." I think this slogan is misleading because it gets the causality backwards. It is not so much that extraordinary claims require extraordinary evidence, it's that if an extraordinary claim were true, that would necessarily produce extraordinary evidence, and so the absence of extraordinary evidence, the absence of Big News, is evidence that the extraordinary claim, i.e. the claim that goes against current best scientific theories, is false. The other important thing to know is that not all scientific theories are the same with respect to producing Big News if those theories turn out to be wrong. Some theories are very tentative, and evidence that they are wrong barely makes the news at all. Other theories are so well established -- they have been tested so much and have so much supporting evidence behind them -- that showing that they are wrong would be some of the Biggest News that the world has ever seen. The canonical example of such a theory is the first and second laws of thermodynamics, which basically say that it's impossible to build a perpetual motion machine. This is so well established that, within the scientific community, anyone who professes to give serious consideration to the possibility that it might be wrong will be immediately dismissed as a crackpot. And yet, all anyone would have to do to prove the naysayers wrong is exhibit a working perpetual motion machine, which would, of course, be Big News. It's not impossible, but to say that the odds are against you would be quite the understatement. By way of very stark contrast, our understanding of human psychology and sociology is still very tentative and incomplete. Finding false predictions made by some of those theories at the present time would not be surprising at all. So our current scientific theories range from extremely well-established ones for which finding contrary evidence would be Big News, to more tentative ones for which contrary evidence would barely merit notice. But there is more to it than just that. The space of current theories has some extra and very important structure to it. The less-well-established theories all deal with very complex systems, mainly living things, and particularly human brains, which are the most complicated thing in the universe (as far as we know). The more well-established theories all deal with simpler things, mainly non-living systems like planets and stars and computers and internal combustion engines. This structure is itself an observation that requires explanation. There are at least two possibilities: 1. The limits on our ability to make accurate predictions for complex phenomena is simply a reflection of the fact that they are complex. If we had unlimited resources -- arbitrarily powerful computers, arbitrarily accurate sensors -- we could based on current knowledge make arbitrarily accurate predictions for arbitrarily complicated systems. The limits on our ability are purely a reflection of the limits of our ability to apply our current theories, not a limit of the theories themselves. 2. The limits of our ability to make accurate predictions for complex phenomena is because there is something fundamentally different about complex phenomena than simple phenomena. There is something fundamentally different about living systems that allow them to somehow transcend the laws that govern non-living ones. There is something fundamentally different about human minds and consciousness that allows them to transcend the laws that govern other entities. Which of these is more likely to be correct? We don't know for sure, and we will not know for sure until we have a complete theory of the brain and consciousness, which we currently don't. But there are some clues nonetheless. To wit: there are complex non-living systems for which we cannot make very good predictions. The canonical example of this is weather. We can predict the movements of planets with exquisite accuracy many, many years in advance. We can't predict the weather very accurately beyond a few days, and sometimes not even that. It was once believed that the weather was capricious for the same reason that people can be: because the weather was controlled by the gods, who were very much like people but with super-powers. Nowadays we know this isn't true. The reason the weather is unpredictable is not because it is controlled by the gods, but because of a phenomenon called chaos, which is pretty well understood. I'll have a lot more to say about chaos theory later in this series, but for now I'll just tell you that we know why we can't predict the weather. It's not because there are gods operating behind the scenes, it is that there are certain kinds of systems that are just inherently impossible to make accurate predictions about even with unlimited resources. Nature itself places limits on our ability to predict things. It is unfortunate, but that's just the Way It Is. So our inability to make accurate predictions about living systems and human consciousness is not necessarily an indication that these phenomena are somehow fundamentally different from non-living systems. It might simply be due to their complexity. We don't have proof of that, of course, but so far no one has found any evidence to the contrary: no one has found anything that happens in a living system or in a human brain that can't be explained by our current best theories of non-living systems. How can I know that? Because if anyone found any such evidence it would be Big News, and there hasn't been any such Big News, at least not that I've found, and I've looked pretty diligently. Because of the fact that, as far as we can tell, our current-best theories of simple non-living systems can, at least in principle, explain everything that happens in more complex systems, we can arrange our current-best theories in a sort of hierarchy, with theories of non-living systems at the bottom, and theories of living systems built on top of those. It goes like this: at the bottom of the hierarchy are two theories of fundamental physics: general relativity (GR) and something called the Standard Model, which is built on top of something called Quantum Field Theory (QFT), which is a generalization of Quantum Mechanics (QM) which includes (parts of) relativity. The details don't really matter. What matters is that, as far as we can tell, the Standard Model accurately predicts the behavior of all matter, at least in our solar system. (There is evidence of something called "dark matter" out there in the universe which we don't yet fully understand, but no evidence that it has any effect on any experiment we can conduct here on earth.) The Standard Model describes, among other things, how atoms are formed. Atoms, you may have learned in high school, are what all matter is made of, at least here on earth. To quote Richard Feynman, atoms are "little particles that move around in perpetual motion, attracting each other when they are a little distance apart, but repelling upon being squeezed into one another." Atoms come in exactly 92 varieties that occur in nature, and a handful of others that can be made in nuclear reactors. (Exercise for the reader: how can it be that atoms "move around in perpetual motion" when I told you earlier that it is impossible to build a perpetual motion machine?) The details of how atoms repel and attract each other is the subject of an entire field of study called chemistry. Then there is a branch of chemistry called organic chemistry, and a sub-branch of organic chemistry called biochemistry which concerns itself exclusively with the chemical reactions that take place inside living systems. Proceeding from there, biochemistry is a branch of biology, which is the study of living systems in general. The foundation of biology is the observation that the defining characteristic of living systems is that they make copies of themselves, but that these copies are not always identical to the original. Because of this variation, some copies will be better at making copies than others, and so you will end up with more of the former and less of the latter. It turns out that there is no one best strategy for making copies. Different strategies work better in different environments, and so you end up with a huge variety of different self-replicating systems, each specialized for a different environment. This is Darwin's theory of evolution, and it is the foundation of modern biology. Here I need to point out one extant Problem in modern science, something that has not yet been adequately explained. There is no doubt that once this process of replication and variation gets started that it is adequate to account for all life on earth. But that leaves a very important unanswered question: how did this process start? The honest answer at the moment is that we don't yet know. It's possible that we will never know. But people are working on it, and making (what seems to me like) pretty good progress towards an answer. One thing is certain, though: if it turns out that the answer involves something other than chemistry, something beyond the ways in which atoms are already known to interact with each other, that will be Big News. Beyond biology we have psychology and sociology, which are the study of the behavior of a particular biological system: human brains. Studying them is very challenging for a whole host of reasons beyond the fact that they are the most complex things known to exist in our universe. But even here progress is being made at a pretty significant pace. Just over the last 100 years or so our understanding of how brains work has grown dramatically. Again, there is no evidence that there is anything going on inside a human brain that cannot be accounted for by the known ways in which atoms interact with each other. Note that when I say "the known ways in which atoms interact with each other" I am including the predictions of quantum field theory. It is an open question whether quantum theory is needed to explain what brains do, or if they can be fully understood in purely classical terms. Personally, I am on Team Classical, but Roger Penrose, who is no intellectual slouch, is the quarterback of Team Quantum and I would not bet my life savings against him. I will say, however, that if Penrose turns out to be right, it will be (and you can probably anticipate this by now) Big News. It is also important to note that no non-crackpot believes that there is any evidence of anything going on inside human brains that is contrary to the predictions of the Standard Model. Speaking of the Standard Model, there is another branch of science called nuclear physics that concerns itself with what happens in atomic nuclei. For our purposes here we can mostly ignore this, except to note that it's a thing. There is one and only one fact about nuclear physics that will ever matter to you unless you make a career out of it: some atoms are radioactive. Some are more radioactive than others. If you have a collection of radioactive atoms then after a certain period of time the level of radioactivity will drop by half, and this time is determined entirely by the kind of atoms you are dealing with. This time is called the "half-life" and there is no known way to change it. In general, the shorter the half life, the more radioactive that particular flavor of atom is. Half lives of different kinds of atoms range from tiny fractions of a second to billions of years. The most common radioactive atom, Uranium 238, has a half life of just under four and a half billion years, which just happens by sheer coincidence to be almost exactly the same as the age of the earth. There is another foundational theory that doesn't quite fit neatly into this hierarchy, and that is classical mechanics. This is a broad term that covers all of the theories that were considered the current-best-explanations before about 1900. It includes things like Newton's laws (sometimes referred to as Newtonian Mechanics), thermodynamics, and electromagnetism. The reason classical mechanics doesn't fit neatly into the hierarchy is because it is known to be wrong: some of the predictions it makes are at odds with observation. So why don't we just get rid of it? Three reasons: first, classical mechanics makes correct predictions under a broad range of circumstances that commonly pertain here on earth. Second, the math is a lot easier. And third and most important, we know the exact circumstances under which classical mechanics works: it works when you have a large number of atoms, they are moving slowly (relative to the speed of light), and their temperature is not too cold. If things get too fast or too small or too cold, you start to see the effects of relativity and quantum mechanics. But as long as you are dealing with most situations in everyday life you can safely ignore those and use the simpler approximations. This, by the way, is the reason for including Step 2 in the Scientific Method. As long as you are explicit about the simplifying assumptions you are making, and you are sure that those simplifying assumptions actually hold, then you can confidently use a simplified theory and still get accurate predictions out of it. This happens all the time. You will often hear people speak of "first order approximations" or "second-order approximations". These are technical terms having to do with some mathematical details that I'm not going to get into here. The point is: it is very common practice to produce predictions that are "good enough" for some purpose and call it a day. Classical mechanics -- Newton's laws, electromagnetism, and thermodynamics -- turn out to be "good enough" for about 99% of practical purposes here on earth. The remaining 1% includes things like explaining exactly how semiconductors and superconductors work, why GPS satellites need relativistic corrections to their clocks, and what goes on inside a nuclear reactor. Unless you are planning to make a career out of these things, you can safely ignore quantum mechanics and relativity. And here is more good news: classical mechanics is actually pretty easy to understand, at least conceptually. It's the stuff that is commonly taught in high school science classes, except that there it is usually taught as a fait accompli, without any mention of the centuries of painstaking effort that went into figuring it all out, nor the ongoing work to fill in the remaining gaps in our knowledge. The reason this matters is that it leaves people with the false impression that science is gospel handed down from on high. You hear slogans like "trust the science." You should not "trust the science." You should apply the scientific method to everything, including the question of what (and who) is and is not trustworthy. And the most important question you can ask of anyone making any claim is: is this consistent with what I already know about the world? Or, if this were true, would it be Big News? And if so, have you seen any other evidence for it elsewhere? It is important to note that the converse is not true. If someone makes a claim that would be Big News if it were true but it doesn't seem to have made a splash, the best explanation for that it usually that the claim is simply not true. But just because a claim does end up being Big News doesn't necessarily mean that it's true! Cold fusion was Big News when it was first announced, but it ended up being (almost certainly) false nonetheless. Big News should not be interpreted as "true" but something more like "possibly worthy of further investigation."
