What follows is not a “professional” take on the scientific method. I am just gathering into one place my own thoughts concerning what makes scientific hypotheses more or less “persuasive” or “real” or “probably correspondent to the real world.” Here, I am using “science” in its limited modern sense – that is – the study of physical realities and the laws that govern their behavior. This sphere of explanation corresponds roughly to Aristotle’s “material” and “efficient” causality, without necessarily making reference to the philosopher’s “formal” or “final” causality. The abandonment of the latter by advocates of “scientism” (the view that science is the chief or only means of explaining any phenomenon at all) is demonstrably fraught with philosophical and practical problems, but I don’t want to get into that here. Nevertheless, I’ve been in several conversations of late with both conservative Christians and with confident atheists that have motivated me to collect my thoughts about what makes scientific claims “persuasive” into one place. And here, I do not mean to imply that any scientific theory will meet all the below criteria, but only that their persuasive power is either “more” or “less” intense depending upon the amount of the below criteria which are met and to the precise extent that each individual criterion is met. And so, in relative order of importance:
1. Scientific hypotheses build upon the sorts of observed events and facts that are very difficult to deny, even by those who are motivated to do so. For instance, if my brother and I got into a fight about whether or not he could lift 300 pounds, and I bet him 10 dollars that he could not lift 300 pounds, I’d be very motivated to observe his failure to lift 300 pounds. However, if he in fact lifted 300 pounds in front of my face, and there was evidence that the weights he lifted were real (believe me, I’d check!), then I would be confronted with the sort of fact that I could not choose to disbelieve even if I wanted to. That is to say, there are certain sorts of observations that, in a fashion, usually tend to compel belief and assent. The closer science sticks to these sorts of observations, the more persuasive it is. A recent example of this might be many discoveries in neuro-science. If someone was tempted to believe that the single direction of causality between the mind and the brain were from the consciousness to the brain (as a sort of tool through which the mind manipulates the body), they would have a very difficult time accounting for the discoveries of neuroscience. Indeed, I daresay that one really cannot be confronted by the findings of neuroscience and really believe that the causal relationship between the mind and brain is uni-directional. Now, likely very few people ever actually believed that (despite certain popular narratives!), but this just illustrates a very firm group of observations by modern scientists that built a cumulative case that is very persuasive. Things that happen to our brain can and do affect our consciousness. Inasmuch as science sticks close to these sorts of observations, its conclusions are at their most persuasive. The further we move into theoretical “models,” this degree of persuasion incrementally ceases. For instance, there are many well-accepted scientific theories (many of which are probably correct) which nevertheless no not “compel” belief in any obvious manner – or rather – they are not built upon observations that compel belief. Often, they are hypothesized as theoretical models to explain disparate phenomena (more on this below). Certain physics theories (String theory?) might be the least controversial examples here. To be fair, it is possible that certain observations would virtually compel belief if they were understood, but are the sorts of observations which take a specialist to even “see.” More controversialy, I suspect that many current culture wars over science are less a function of being unpersuaded than by a refusal to observe – less a function of mind than of will.
2. A good scientific theory or model should demonstrate “fittingness” with the observed phenomenon. I have often seen folks take all explanations which are not absolutely certain as epistemically equal. This is simply not the case. Some explanations are better than others, and obviously so. They are better to the extent to that they have a sort of aesthetic “fittingness” with the phenomena being explained. For instance, the hypothesis “Little magic blue people inside of atoms account for all of their behavior” is not a demonstrably false statement. But it also doesn’t account for any atomic behavior any better than, say, “little magic red people,” or “”little scientific blue people” or “little magic blue kittens.” That is to say, while this hypothesis is consistent with all of the evidence, it doesn’t actually “account” for the evidence in a way that is unique from, well, any other explanation. A good scientific theory or model isn’t just “consistent with” the evidence, but uniquely explanatory. It doesn’t remain content with mere logical consistency but seeks aesthetic harmony. The principle is something like, “If this theory is true, not only would it be consistent with these observed phenomenon, but you would expect the phenomenon to look exactly like, or very close to, what we find them to be in reality.” The recent science of plate tectonics is probably a good example of a theory that meets this criteria. In any case, closely related to this point…
3. Scientific theories are more persuasive the more often they generate predictions which lead to further observations. It is particularly neat when a scientific model can say something like, “If the world is as I expect it to be, then I should observe x in place y” BEFORE they observe x in place y! If a scientist can pull that off, and the more they can pull that off, they are really onto something. Their success in discovery means that the world is yielding its fruit to their hypothesis. And this suggests that their hypothesis is correspondent to the world. There are many examples of this (Einstein’s predictions related to relativity being a famous example), but one of the most exciting in recent times has been the prediction(s) related to the Large Hadron Collider in Switzerland – specifically the prediction and (hopefully!) discovery of the Higgs particle.
4. Scientific theories lack persuasive power to the extent that they are unfalsifiable. The mark of pseudoscience is that the target is always moving. I am sure there are many examples of this in various health studies and paranormal theories. Conspiracy theories are a good foil here. The more you attack the central premise, the more the conspiracy gets complicated and the ground of explanation shifts. “You think everyone wants to kill you? Then why are most people so nice to you and why haven’t they killed you yet?” “Because they are toying with me and fattening me up for the perfect ceremonial kill moment, of course! I mean, who wants to kill someone without some irony?” You get it. There’s no way to falsify the hypothesis. But good science is falsifiable. Do you want to disprove Newton? Throw something in the air that doesn’t fall back to the ground. Do you want to disprove evolution? As one scientist famously put it, produce a fossilized rabbit from the precambrian period.
5. Finally, good science recognizes when it has made a wrong turn. It knows its limits. This is controversial, but it is not for that reason less true. If “science” leads you to a view which says that human reasoning is all just deterministic and a fundamentally irrational response to stimuli, then you have falsified the persuasiveness of your theory. If your hypothesis leads you to conclude something which is utterly counter-intuitive (i.e. “we are likely a computer simulation of another species”), you have likely gone of the rails. It is not that science cannot lead to counter-intuitive conclusions (“wave-particle” duality, anyone?). But there are basic phenomena to be explained – some of which are so basic that they cannot be explained “away” or “reduced” without violence to what they simply “are” (Consciousness is a good one here). Fundamentally, I’m saying that science really cannot do without mind and reason and at least some level of reliable observation. Perhaps Plantinga’s “evironment within which the mind is intended to function” is the basic phenomenological zone which needs explanation here. Whatever the case, certain things (and we could argue about what these are – I’m just going for principle here) are basic and constitute the phenomena to be explained rather than phenomena which are reduced to other things. I’m talking to you, social scientists! Once again, scientific conclusions can be counter-intuitive, but they cannot be counter-being and the basic nature of being. This seems obvious to me. It is not as though, by the way, that metaphysics is any different here. Metaphysics leads to counter-intuitive conclusions as well. Or at least, it points to realities which cannot be fully grasped by the human mind – and in this sense, there is a certain fittingness between science and metaphysics. Indeed, one might even say that the realm of rational science behaves precisely like classic metaphysics would anticipate it to behave. Hmmmm…
Addendum: I should have mentioned (number 6?) that a hypothesis is more persuasive when it has independent lines of confirmation or fittingness. So, for instance, evidence from different fields of biology, chemistry, physics, geology – fittingness with philosophy, reason, etc. This means that even if a particulate line of evidence is negated, the whole is not necessarily negated. Basically, there is something of a “cumulative case” principle here.