Hobson makes some difficult points well in Chapter 1, section 8. “Pseudoscience” is a professional concern in science with all science societies putting out fires around the world and trying to better explain what science is and what science isn’t. I wanted to underscore some of them. One of my favorites: one repeatedly hears about evolution is that it’s “just a theory.” One also hears it about the Big Bang and other scientific programs like in climate science. This would be amusing if it were not deadly serious. There is a disconnect, I think, between our everyday use of words like “law” and “theory” and what scientists mean by them. More complicated is when words like “fact” and “proven” are used in a context of science.
Let me try to start at the pre-beginning! What I think all believe – note that we don’t sit around talking about this stuff much, so this is my distillation of what we do as scientists. We subscribe to the following commitments, without which scientific explanation, and hence science itself, would be impossible.
- Objects in the universe behave in consistent, predictable ways, through all of time and at all locations.
- A natural phenomenon has to behave the same way over there as it does over here. Likewise, a natural phenomenon has to behave the same way yesterday as it will tomorrow.
- If this were not the case, then any disagreement about the measurement of a phenomenon could be immediately ascribed to the fact that the measurements needn’t be consistent!
- This is related to:
- Natural phenomena are not capricious. Things don’t happen by magic, but for reasons.
- If there are some limits on what we can know about natural phenomena, then if a result is puzzling or a problem seems hard, then we could just say that “Well, we’re not supposed to be able to understand this” and quit. There cannot be any limits to what we can expect to know.
- Frankly, nobody knows why this seems to be the case. Every time mathematics appeared to not somehow work to understand a natural phenomenon, new mathematics was invented. Google “unreasonable effectiveness of mathematics” and before you can even get to “of” and beyond, the search engine has already filled in the rest and presented you with more than a half-million hits on an article written by Eugene Wigner in 1960. It’s still the gold-standard of representing our bafflement that the universe appears to be mathematical!
- This is one of those cases where overthinking isn’t practical. Physics is defined now as the branch of science that seeks to model the objects in the universe using mathematics and test those models using experiments.
With those commitments, anything goes in science, within the bounds of people-related best customs and ethics.
The models that we create act as guides to experiment and experimental results act as inputs to the creation of models. I like to think of this modeling process as metaphorical. That is we adopt a metaphor to describe some feature of the natural universe and then we model the metaphor, not actual reality itself. For example, early understanding of gases was gained by modeling the metaphor that a gas is a set of little hard spheres without size that have velocities and bang on the walls of their containers according to the mechanics of Newton. This worked very well in the sense that it is a model that’s descriptive and predictive and relates the gas pressure and volume to the average motion-energy of the molecules. This in turn was then interpreted as a working definition of temperature. There were areas where it didn’t work so well, and so the metaphor was changed to include a finite size for the molecules, some defined interactions among the molecules, and so on.
This leads to the notion of confirmation and proof, major areas of confusion. There are no proofs in science and waiting for one is nonsense and judging a science negatively because it’s “not been proven” is wrong-headed. There are proofs in mathematics, since it works within a deductive system of rules. But science works also within the inductive arena where you measure and measure and measure and compare your measurements with your model, but you’re always aware that the next measurement may cause you to have to modify or even reject your model. I had only ever seen brown squirrels in my life, before I moved to Michigan. Every squirrel I ever saw was brown. I could indeed have created a Model of Squirrels, the first “Law” of which would have been: All Squirrels Are Brown. Then I moved to Michigan and met black squirrels. They’re all over my yard. When did my Law fail? Why upon observation of my first black squirrel. Had I not moved here, I’d still subscribe to my original theory.
This is the most important idea of all: Scientific models (notice I keep using the word model, not Law) are subject to challenge – by definition. To say that a scientific idea is “just a theory” is ignorant. All scientific models are all “just” theories. Some theories are very well confirmed (every measurement keeps agreeing) and are therefore highly trusted. So trusted, that we tend to refer to them as “laws.” I’ll make a serious point about that with Newton’s “Law” of gravitation. Some theories are less-well confirmed because their tests have been few, or they have not been very strenuous.
This also leads to the most important criterion for deciding whether a purported statement about the world is scientific: If a theory or model cannot be disproven in principle, it is not scientific. Suppose I have a theory that says: The Heights of All Men Are Either Above 6 Feet Tall, Or Less Than 6 Feet Tall. Is that a scientific statement? There is no way to disconfirm that statement and so no, it’s not. How about a theory that says: The Egyptians Built The Pyramids With The Help Of Aliens. Is that scientific? No. There is no experiment that I could perform that could disconfirm that assertion. Lack of evidence of aliens now doesn’t work. They came, they helped, they cleaned up their footprints, and they left. It’s a (wacky) belief, not a scientific statement.
There are disconfirmation tests for all scientific theories and every scientist would recognize it when he/she saw it. This is true of physics, and we’ll talk about the severe tests that the Big Bang model has faced and survived and how a competing theory failed. It’s also true of evolution. This disconfirmation-idea in evolution is sometimes euphemistically called the “precambrian rabbit.” Finding a rabbit dated to precambrian times would cause Evolution serious problems. Theories which are well-trusted usually don’t crumble at a single disconfirming instance, it would take a lot of evidence. But many times well-trusted theories have been significantly modified or indeed abandoned. The point is that the principle of falsification remains. I thought that Hobson did a pretty good job on this point.
A fussy point just for all of you who may have taken a Philosophy of Science course!: “Falsifiability” as an important criterion to distinguish science from pseudoscience is attributed to the Philosopher of Science, Karl Popper and he has been criticized for it ever since. I intentionally didn’t use that word and would not hold to a strict Popper formalism. And, yes, I’m aware of the Duhem–Quine Thesis as a serious critique. But my gentle criterion stands: in-principle disconfirmation is required of scientific statements.