Scientists (by alaric)
I've been reading a book lately called "The Brain that Changes Itself", which discusses a once-controversial theory that the adult brain can rewire itself in the same manner as the developing brain, with the main difference being that it just requires more effort to focus the attention. It used to be believed that once the brain had finished developing, its structure and function was fixed. However, it turns out that, with the right approach, the victims of strokes and the like can retrain their brain to perform the lost functions with different bits of neural tissue.
But that's not what I'm writing about today.
One of the things that has struck me in the book's account of how the neurbiologists rejected this controversial idea, along with previous impressions I had obtained from other sources, is that the academic community is riddled with idiots who reject evidence that contradicts their beliefs about their field.
To see why this is crazy, look at it like this. Science is, largely, about finding the underlying truths of the Universe. The problem is that these underlying truths can rarely be directly observed (and we've figured most of the ones that can out by now). One cannot directly percieve an electron, but one can deduce its existance by perceiving the effect of an electron gun in an evacuated chamber pointed at a phosphorescent screen. But there are multiple interpretations of that experiment - perhaps there are tiny charged particles being released which stimulate the screen into producing light... or perhaps the electron gun actually causes the metal of the negative electrode to ablate and the resulting ion cloud then condenses into an invisibly thin thread which coils out across the vacuum until it touches the screen, whereupon electricity flows directly down the wire and causes the spot of light. Perhaps the vacuum is required, not because air inhibits the free motion of electrons, but because the air disrupts the formation of the thread.
Yet we can rule out the thread theory in a number of ways, and there are other experiments that show that electrons are discrete charged particles. It's the weight of a whole heap of evidence that all reinforce the correct theory and disprove all the alternative theories. However, one can never be entirely sure that another theory has yet to be discovered, which all the existing experiments fail to disprove - but which leads to the development of an experiment which disproves the electron theory, and reinforces the new theory. Perhaps there are no electrons; but the "electron theory" has provided us with useful predictions, and nobody has yet found fault with it. So we stick with it. Even if it's wrong, it's useful - and if we ever find it's wrong, that will give us the clues required to find a better theory.
But there are levels of deduction involved here. We directly observe the construction of electron guns and the appearance of spots of light with our eyes. We apply previously reinforced beliefs that the electrical power supply we connect the electron gun to will provide a voltage, and that the electron gun will therefore emit electrons. We observe the appearance of a spot of light, and therefore conclude that the electrons flew through the chamber and caused the spot of light. And from that, combined with existing knowledge about the nature of light and matter, we construct a theory that electrons can travel through a vacuum then cause phosphorescent screens to glow. Each level of further deduction is less certain then those that it builds upon, since its truth depends on their truth, plus a further step of deduction - which might be wrong in itself.
So what do we do when new evidence comes and appears to disprove our theory? Say somebody publishes the results of an experiment that show that, if a kitten is within one metre of the chamber, the spot of light on the screen grows into the kanjii for "potato". The electron theory does not predict this. Have we disproved the theory of electrons? Or have we merely discovered that kittens emit complicated high-frequency magnetic fields that disturb the paths of nearby electron beams? Well, I'm sure further experiments would be performed, surrounding kittens with Hall sensors and SQUIDs and the like, but for now, let's imagine we only have that one data point to look at.
Electron theorists would probably question the validity of the experiment at all. For a start, it is a leap of faith that the experiment was set up correctly. Perhaps the electron gun itself is defective and projects the kanjii symbol directly, and the kitten has nothing to do with it. Perhaps there are coils under the bench generating magnetic fields that steer the beam to draw the symbol, either accidentally or as part of a deliberate academic prank. In this case, with seemingly unrelated objects (kittens) having suspiciously unexpected consequences (kanjii characters), that is a distinct possibility, so the kitten theorists would be under additional burden of proof to recreate the experiment - and to ask electron theorists to defend their theory by recreating the experiment themselves to show that it does not occur with "trusted" equipment. For sure, the academic community does need some level of protection from a "denial of service" attack from charlatans assaulting it with fraudulent claims that have to be tediously experimentally dismissed. There is scope to accidentally perform flawed experiments due to overlooking some factor or failing to test all the equipment used for defects, leading to honest results that turn out to be misleading. This gives some credibility to the concept that some data can be rejected out-of-hand for contradicting widely-held theories, but it is all too easy to take conformist censorship at this level too far and reject evidence that actually shows flaws in currently-sacred theories.
But what if the conflicting evidence is less silly, or it is independently and widely confirmed in other experiments, showing there is definitely some effect at work? Perhaps kittens do emit mysterious high-frequency magnetic fields - in which case, our theory of electrons is still valid; it's just our theory of kittens which was wrong. As physicists are often more familiar with electrons than kittens, it's easy for them to defend their electron theory and question the researcher's grasp of kitten theory, thereby making it somebody else's problem. Meanwhile, biologists asked to defend the theory that mammal tissue can't generate intense, high frequency, magnetic fields might point to excellent arguments about the maximum rates of charge movements in various tissues, and tell the physicsts that their electron theory must be all wrong. At least we now have some kind of debate, rather than outright censorship, but - particularly in cross-specialisation problems like this one - it's all too easy for both sides to just ignore the evidence and blame it on the other.
But what makes scientists so defensive? Good scientists realise that the data is all we can be sure about (and, even then, we must be careful of experimental errors, or failing to control for unknown influences). They treat theories as temporary affairs, which suffice until they are found wanting, or something better is found. Where does this academic Nazism emerge, where academics will often jump immediately to questioning the motives and competence of people who hold views that contradict the mainstream, leading to the mainstream remaining mainstream long after the weight of contradicting evidence becomes overpowering?
I think a part of the problem is the fact that scientists with new ideas have to fight so hard to get them heard over the mainstream in the first place - they find it hard to give up the fighting mentality once they've been accepted.
Another part of the problem might be human nature - scientists are taught the existing lore of their field in lectures where they soak it all up, and probably record it in their minds as unassailable truth. I suspect they are much more open to reconsider theories they encounter as "new" after having lived, for a while, in an academic world in which no theory explaining the behaviour in question had yet emerged. Theories considered "complete" when they were learnt are probably rarely questioned.
Fri 29th Jul 2011 7:15 pm 
Subscribe
By Seth, Tue 9th Aug 2011 @ 8:58 am
the trouble with science is people. Although the scientific process is sound, the minds implementing it are likely to have personal gain as a motivating factor. That, and wondering where the next grant will come from.
By @ndy, Mon 22nd Aug 2011 @ 11:51 am
Are you familiar with the experiments to determine to charge on an electron? The first experiment was widely accepted but eventually turned out to be drastically wrong. However, it took a number of intermediate experimenters who gradually changed the value towards the correct one before it could be accepted. The "correct" value was just too far from the current value for it to be immediately accepted by the mainstream.