Take “The moon is made of green cheese.” That’s just plain false, and we know it’s false because (among other reasons) we have taken parts of the moon and looked at them under microsopes.
Some statements are worse than that: some statements aren’t even false. Think of a rambling newspaper horoscope that’s supposed to tell you how your day will turn out, except it’s written in such amorphous terms that you couldn’t pin it down as true or false no matter what happened. Or listen to someone describing a “life force” or “higher power” in a way that makes no contact with anything you could ever see or hear or touch or taste or smell. Stuff like that, stuff that’s beyond assent or dissent because it is so far adrift from any possible observations, I shall call “guff”.
Now we need to be careful. Science, even hard science like physics, has a lot to say about things that we can’t see, hear, taste etc., because they’re too small, or too far away, or they’re inaccessible for some other reason. But physics is not guff. That’s because the things physicists say about electrons etc. are connected to observation in the right kind of way. And what kind of way is that? In other words: what is the line that separates guff from non-guff? For my money, that’s the most important question in philosophy.
At this point I expect a protest from the Scientists’ Union, that the question is theirs, not philosophers’. Well I’m more than happy for scientists to join in on it, but when they do so they will be talking about science at such a high level that I count it as philosophy. Still, I don’t want to quibble over job descriptions. The point is: all welcome.
Another sort of protest comes from those who have suffered a poor education in philosophy, and think that philosophy is all guff. Okay: I’ll admit that philosophy has had more than its share of guff merchants, but that only means that philosophers have had to put up with it more than most, so they’ve hardened against it more than most. If you don’t believe me, here‘s the eighteenth-century philosopher David Hume:
When we run over libraries, persuaded of these principles, what havoc must we make? If we take in our hand any volume; of divinity or school metaphysics, for instance; let us ask, Does it contain any abstract reasoning concerning quantity or number? No. Does it contain any experimental reasoning concerning matter of fact and existence? No. Commit it then to the flames: for it can contain nothing but sophistry and illusion.
Anti-guff warriors like Hume (they called themselves empiricists, by the way, and they called guff metaphysics) had a few different things on their minds. Often they wondered how to justify our knowledge of the world, based on our observations. That’s a perfectly good question, but it isn’t mine, at least not today. I only want to know where to draw the line between guff and non-guff. I don’t even need to prove that I’ve drawn the line in the right place; I’ll be happy as long as it is not obviously in the wrong place. For example, if I were to say “Guff is anything you can’t see”, that would put electrons on the guff side of the line (in fact, never mind the electrons; it would put the back of my sofa on the guff side of the line) — so that would be a fail. I have to do better than that.
Empiricists have had many shots at this: it’s been going on for centuries now. There were early hard-liners who pretty much did say “Guff is is anything you can’t see”. (It sounds better in Latin —Nihil est in mente quod non prius in sensu— but it’s still a fail.) Then some said guff is anything you can’t prove, based on a full statement of all your sensory inputs. Some modified that by allowing types of reasoning of a lower standard than proof. Others thought it was more about what concepts you could define, in a big dictionary where all the definitions eventually came down to terms that describe your sensory inputs. That didn’t quite work, so there was some ingenious fiddling with the concept of “define”, taking it beyond dictionary-style definitions in various ways.
Suffice it to say that none of these attempts worked out. Either they set a standard that even the best science could never meet, or it was a standard that science could only meet if it performed some monumental construction, like compiling the most complicated dictionary the world has ever seen.
Then up stood Karl Popper, and gave us the concept of refutability. A statement is refutable if there is some possible observation that, if it occurred, would disprove the statement. According to Popper, scientific statements are always refutable, whereas guff never is. The amorphous horoscope, for example, is guff because nothing you could ever see, hear, etc. would disprove it. Statements of science, by contrast, always stick their necks out; always take a risk. Of course, the science we currently accept has not actually been refuted, but the point is that it could be. All science is provisional.
Compared to what came before, Popper’s refutablity principle was a breath of fresh air. There was no longer any need to construct arcane definitions or proofs; just say what observation would, if made, lead you to reject a given statement. And it’s especially well suited to general scientific laws, such as “All swans are white”, which are impossible to prove, but can be disproved by even one observation.
I’m not sure, but I get the sense that, outside of philosophy departments, Popper’s analysis is usually seen as the last word in philosophy of science. It isn’t. What it is is another failed attempt, in the long line of failed attempts.
The problem is that many statements in science are what you might call team players. They aren’t refutable one by one, but only in packs. As W. V. Quine put it, “our statements about the external world face the tribunal of sense experience not individually but only as a corporate body.”
Quine, like Hume, had a few different things on his mind, and it isn’t easy to separate his team player theory (it’s called epistemological holism) from his other concerns; but I’ll do my best, with a pretty contrived example.
I put you in a room with a pen and a pencil, and no other equipment, and I ask you to consider three statements:
A. The pen has positive electrical charge.
B. The pencil has positive electrical charge.
C. The pen has positive electrical charge and the pencil has positive electrical charge.
You could move the pen and pencil close together, and if C is true then they should push apart, because like charges repel. Or to put that another way: if you move the pen and pencil close together and they don’t push apart, then C is refuted. Good: so C is refutable in just the way Popper likes. But what about A, and what about B? If you move the two implements together and they don’t push apart, you’ll know that either A is false or B is false (or both are), but you won’t know which. In fact, with your limited apparatus, there is no test available to you that could disprove A, and there is no test available to you that could disprove B. So, by Popper’s criterion, A is guff, B is guff, and C is good science. Yet C is just A and B bundled up; so something must be wrong with Popper’s criterion.
(Objection: “Refutability isn’t a matter of what tests are available to me in a particular small room; it’s about what tests are possible, and it certainly is possible to test A, by taking the pen to a laboratory with more equipment etc.” I concede: the example is oversimple. But when you take a real-world case, where an observation supposedly refutes one statement, careful inspection will reveal something like the structure I’ve sketched: there are several statements (including some that you might have overlooked: is your electromagnet switched on at the wall?) that work together to predict an observation. And if the predicted observation doesn’t occur then, strictly speaking, no one of those statements is refuted; any one of them could still hold, if others are wrong.)
In case that example wasn’t convincing, here’s another, of a different sort. Take “There are some atoms in the universe.” Clearly this statement is science, not guff; yet if you ask a scientist to tell you what possible observation would lead him to reject it, I don’t think he’ll have an answer handy. Also he’ll probably let you know that you’re making an unreasonable demand: the statement is good science regardless of whether he can dream up some answer to your far-fetched question. In other words, the refutability criterion doesn’t work here.
Time, then, for another criterion.
I’ll call this one the contribution criterion: a statement is guff if it makes no contribution to observational predictions. Or to spell that out: Ask what would happen if you dropped the statement. Would you lose the ability to predict some observation that you could have predicted when you had it? If so, then the statement is making a contribution. If not, then it’s guff.
In the pen-and-pencil example, A makes a contribution, because if you dropped it from your overall theory, then logic would compel you to drop C, and then you’d lose your prediction of the pen and pencil pushing apart. So A is not guff. Ditto for B.
In the atoms example, if the scientist dropped the statement “There are some atoms in the universe”, then logic would compel him to drop every law and every specific statement that refers to atoms. Ask what predictions would go missing as a result, and he can give you a flood of answers; nothing difficult or far-fetched about that. So “There are some atoms in the universe” makes a contribution; it is not guff.
On the other hand, the amorphous newspaper horoscope makes no contribution. If, in a moment of weakness, you had let it into your theory of the world, you could drop it now and not lose any observational predictions. So it’s guff.
The contribution criterion is working out pretty well, and it has another pleasing property also: it gives mathematics the thumbs up, because there’s a whole lot of observations that you couldn’t predict if you dropped mathematics. Likewise with logic. I haven’t mentioned this yet, but all the other criteria we’ve discussed were going to need special clauses to keep maths and logic on the non-guff side of the line (and, on the whole, their authors knew that). But now we don’t need any special clauses: maths and logic belong to science for exactly the same reason that physics and chemistry do.
I’ve noticed that this feature disappoints some people. They seem to want maths and logic to be justified in some special way, unrelated to observation. The term they use is “a priori“. They have a hard time explaining how a priori justification works, but they have a strong intuition that it should work somehow. I say that their “intuition” is just their internalization of pre-Quinean philosophy, and its stickiness is a measure of how deep a shift Quine has asked us all to make; but that is no reason not to make it.
So, does the contribution criterion, as I’ve stated it, draw the line between guff and non-guff in exactly the right place? Are we there yet? For now, I’ll only say that we’re on the right track. The criterion needs some spelling out and breaking down, and its surprising consequences need to be explored. Note to self: blog about this again one day.