Chapter 10

The method

Well, there we are, finally. Now what are we going to do with all this?

We already have a starting-point, if we can find some use for it: our absolutely certain knowledge that there is something.

We also have a criterion for the security of our starting-point that is better than Descartes' "clear and distinct idea" (which was so clear as to be incapable of being doubted by him, but not others), or Spinoza's "adequate idea" (which was adequate for him but not others), and so on. Our criterion, if you remember, was that something is absolutely certain if to deny it is to assert it (i.e. if the denial only "works" if what is denied is true).

But that still leaves us at the starting-point, because the absolute certainty that there is something is self-awareness of the consciousness as something; the "refutation" of the universal skeptic was that he knew that his doubt was not nothing.

But can we go beyond this? We are, true, physically certain that we are not now dreaming and that we are confronted with a "real world out there." But to deny this is not, on the face of it at least, based on the assertion of it; it is not easy to see how it is impossible for there to be nothing but my consciousness and for me still to be conscious in the way I am. In other words, "How do we know that not everything is a dream?" is not a captious question like, "How do you know that there is something?" How do we know, when we dream sometimes and think we're not dreaming?(1)

It sounds as if we have an effect.

And, of course, the philosophical method we are going to be using is based on noticing effects about our starting-point, our self-conscious consciousness. If it can be shown that something about our consciousness is "impossible-unless" then the "unless" must be the case. And this "unless" may very well be a real world "out there."

What, then, do we do?

First, be very clear what exactly the effect is; and that involves two things: (1) Be sure that there is real evidence on both sides of a contradiction, and the "contradiction" is not based on misreading what is there; and (2) be as precise as possible in picking out just how the facts known contradict each other.

There is the possibility of error here. It is frightfully easy to misread the evidence and find a "contradiction" because you have formulated the problem wrongly. And because, no matter how careful you are, it is still conceivable that you could be misled, we have at this point left the world of absolute certainty.

I might point out, however, that it is if anything easier to dismiss a real effect as simply a "bad formulation" because you don't see any difficulty about what is happening. Anyone who wants to circumvent an effect with the "bad formulation" approach will have to show just exactly where the bad formulation occurs, and how his view of its being "bad" doesn't entail the denial of things that he wants to hold onto.

If anything, there has been as much of this error in the history of philosophy as there has been of the preceding one. And interestingly, the people who dismiss certain recurrent problems in philosophy as "bad formulation" have themselves done so on the basis of a bad formulation of the grounds for dismissal, as the history of the "verifiability" criterion of meaningfulness so abundantly testifies. The various formulations of the criterion neatly showed how the inconvenient problems were non-problems; but at the same time they made "meaningless" a whole series of things that the formulators wanted to keep as not only meaningful but true.(2)

All this says is that at this stage, you have to be very careful, both to make sure that you've got a real effect, and that you don't ignore a real effect because it doesn't strike you as peculiar. And either way, you can be mistaken.

Well, so be it. The Cartesian ideal of an absolutely certain view of the universe, deduced mathematically from an absolutely certain premise, is gone, if you're going to proceed by way of effects and their causes--because there's no way I at least know of to show that the denial of a given effect as an effect is the affirmation of it.

Then is it all "just a theory" and we're back to "Well, that's your opinion"? Not at all. I just got through saying that if you're going to deny that such-and-such is a real effect, you had better be ready to show (a) which of the facts alleged to be in conflict isn't a fact, or (b) just how the conflict isn't really a conflict, though it would seem so to an intelligent person. In other words, you have to take the deception of the philosopher (who presumably is sincere and not a fool) as an effect and give a cause for it, or you are simply "disagreeing" with no rational grounds for disagreeing. And in that case, since his position is rational and yours isn't, you lose.

What I am saying is that to say, "I don't agree, and after all, you could be wrong" is an irrational position to take; and what reason could you have for taking an irrational position? So a position that doesn't immediately "grab" you or which you aren't "comfortable with" is nevertheless to be accepted under penalty of sticking with what is irrational in preference to what is rational. Sorry, but there it is.

Second, we will use the abstractness of "cause" here. It is known that the true explanation for the effect discovered will have all of the properties necessary to remove the contradictoriness of the effect. But since we presumably can't get "outside" our consciousness to observe the causer "as it is in itself," then there is no hope, really, to be able to know any other properties of the true explanation than the ones that are necessary. Hence, all we can know is the cause of the effect (what all explanations would have to have as a common minimum).

At this point, we will make what P. W. Bridgman called an "operational definition": we will take a term and define it as "The cause of this particular effect," meaning "Whatever it is that actually accounts for this effect." We know that there is such a thing, and so we use this definition to refer to it.

Bridgman's "operational definitions" are valid, not because (as he thought) the supposed "unobserved entity" (such as an electron) didn't really exist but was a mask to cover a set of actual observations, but because the electron does and must exist or the observations are contradictory; and therefore, it can be known as "Whatever makes these observations not contradictory." Let's face it: blank photographic plates don't get marks on them by theoretical constructs. So his "operational definition" much more properly should be called a "causal definition," where the cause is defined in terms of its effect, precisely as "whatever accounts for the effect."

Note that this step is absolutely certain. Given an effect, it is absolutely certain that there is a cause; and since all we have done by our causal definition is point to whatever is the cause (without alleging any more about it than that we know there is one), our "definition" cannot be mistaken, because whatever in fact turns out to be the explanation of the effect, it is ex hypothesi what we meant by our term.

This causal definition is just a gimmick, in other words. You don't know any more after you have picked out your term and defined it than you did when you spotted your effect. The difficulty comes in spelling out exactly what fact the cause is, or what properties it actually has. Still, this step is useful because it avoids the clutter of constantly saying, "Whatever it is that accounts for this particular effect" as you try to find out what has to be the case no matter what it is that accounts for the effect. It is like what mathematicians do when they have a long expression that keeps cropping up in the course of their proof; in order to avoid writing it over and over again, they say, "Let Q (or whatever) stand for this expression," and from then on write Q wherever it would appear, until the end, when they "substitute back."

Similarly here. Our causal definition gives us no insight into the meaning of the term we have defined in this causal way; but it does leave us a handy word instead of a cumbersome expression to use in our investigation, provided we keep in mind that as we use it it means exactly--nothing more and nothing less than--"Whatever it is that is the cause of the effect in question," and we remember that the fact that we can use it doesn't mean we know what it is that we are talking about, but merely that there is something referred to by this term.

Third, knowing that there is a cause, and having a term now to refer precisely to it, we are now aware that the cause contains all that is necessary to account for the effect.

Two things can occur at this point: First, if it can be shown that the effect remains a contradiction unless something about the cause is true, we know that this is a fact about the cause.

This is a very dangerous procedure, and most philosophical theories come a cropper on it. In order for it to work, it must not simply seem to you that without the truth of X about the cause, then the effect is still self-contradictory; you must be able to give a positive reason for excluding anything else but X as another explanation. For example, it is quite easy to say that the finiteness of the world makes sense on the supposition that there is an infinite being (God); and it is "obvious" to many that it makes sense on no other supposition. But how do you show that anything but God (e.g. an infinite set of finite causes) leaves the world self-contradictory?

Once again, then, we are confronted with the possibility of error; because we may think that a given fact explains a given effect when it might not, or (far more likely) we may think that we have ruled out all other explanations, and there may be one lurking there so bizarre that it never occurred to us to consider it--and lo and behold, it is the right one.(3)

Secondly, if the effect in question is similar to some other effect we have already studied, then the cause in question necessarily is analogous to the cause of the other effect. We may not know what the cause is in either case, but we know that they are similar. And if we can point to the respects in which the effects are identical, we can then "separate" the causes into aspects and make up a term referring to the (unknown-in-itself) identical aspect of each cause by which it is identical with the other cause.

Once again, picking out these points of identity will not (as we will see) be absolutely straightforward; and so the specter of possible error sticks in its ugly head once again. I don't want to give the impression by pointing out all these difficulties that things are hopeless; but after all, the most brilliant people in the world have been at this enterprise for thousands of years now, and while we have made progress (we really have), we have made plenty of mistakes along the way; and it's as well to know where the weak points are in the various world-views that explain our experience.

Unfortunately, checking procedures by analogy with the experiment and verification stages of scientific method is very difficult to do here, because you can't get at the causer and manipulate it to see what happens. But we are not exactly stuck, and at least sometimes we can pass on to the following:

Fourth, we take the properties we have found belong to the cause and examine what else must be true if they are indeed "there." If these "predictions" turn out not to be facts, then obviously the properties of the cause can't be what we thought they were, since the "predictions" are logically entailed by the property.

This means that the philosophical method I am proposing is actually empirically falsifiable, in the same sense as scientific theories are falsifiable. You can't prove a theory true from its predictions, because the logic of the situation (if Theory, then Prediction) cannot prove the "if" part true by knowing the truth of the "then" part; but the falseness of the "then" necessarily means the falseness of the "if."

We already saw this at least twice. First, when we were looking at Hume's "explanation" of why we think in terms of cause-and-effect, where he said that the juxtaposition of two events led us to expect the second on seeing the first, and so we called the first the "cause" of the second. What I did was make a prediction from this that therefore any constant juxtaposition would induce us to call the prior event the "cause" of the subsequent one, and so we would necessarily (if his theory is true) call the dawn the cause of the sunrise. But we don't. Therefore, his view of why we think in terms of cause is false.

Secondly, we established that the conscious act is conscious of itself by making a prediction from the opposite hypothesis, showing that if it were the case, then we should be uncertain that there is something even though it is still absolutely certain that there is something and we could know that fact.

Well that's some progress, isn't it? We know one view of causal thinking that doesn't work, and we know that it makes sense to say that we know what we know when we know what we know.

One final remark before we actually begin using this method; and this deserves to be stated as another conclusion:

Conclusion 7: we cannot know all about the world using effects and causes.

It is exceedingly unlikely (to say the least) that absolutely every fact would be necessary to explain some other set of facts; and causes are only what is necessary. In all probability there are other aspects of things which are there, but not necessary to explain some otherwise inexplicable set of facts.

Those of a rationalist turn of mind say, "Well of course they have to be connected to other facts; things don't "just happen." True, effects don't "just happen"; but in order to show that every fact is somehow knowable, you would have to show how it is contradictory that there could be a fact which is (a) not directly known, and is (b) not necessary for the knowledge of some other pair of facts. It may be that there are a lot of facts that are parts of causers, for instance, and so are "involved" in the cause-effect relationship; but if they are non-necessary properties of the causer, we can't know what they are unless we directly observe them. What the thief is thinking as he steals your keys may be something that it integrated into his whole being; but there may be no set of observable facts that would allow you to say, "He couldn't have been thinking anything else but this."

So I think we have to give up at least one half of Hegel's premise that "the real is rational and the rational is real." The real is not irrational, which is another way of saying that effects do have causes; but effect and cause do not exhaust the facts of the universe--or at least I don't think that it can be proved that they do. In that case,

Conclusion 8: The real is both rational and non-rational.

I think that by far the more likely view is that, there are facts that are gratuitous: "just there," so to speak. There's no reason for there not being what they are; but there's no reason why they can't not be what they are. They just are; they're facts, not effects. Why was it that the particular two children I have got born rather than the ones who miscarried or the millions of others who could have been formed from the genes of me and my wife? "Because," you say, "these particular genes got together." But that's the point. There's no reason why these genes had to be the ones that got together; they just did. And I rejoice in my children and accept them; they were not "necessary" to fit into some "scheme of things." Aspects of each of them are necessary, and causes of certain effects; but these aspects do not exhaust them; and if we think they do, then we have a very impoverished view of them.

But this is compatible, of course, with some facts being necessarily related to others as being the causes of effects we can observe; and it's time now to take all this apparatus and apply it to our consciousness of our consciousness.



1. It's not enough here to say, "We just know," because in dreams we don't know and don't know we are dreaming; and in waking life, there are hallucinations, where we seem to see or hear what is "out there" and real when in fact it isn't. Generally speaking, we can distinguish perceptions of what is real from dreams and hallucinations; but (a) this is not enough for absolute certainty, and (b) the question here is not whether we can usually do this but how we manage. So I'm not trying here to cast doubt on your conviction that you are awake and reading these lines; all I'm saying is that it involves an effect, which we will explore.

2. For instance, if you say that the only thing that is real is what is observable, then how (now) do you observe what happened yesterday? Yet you know that it really happened. But there is no way now to verify this by observation. And are photons and electrons merely "mental constructs" that don't have a real "component" (i.e. are they the equivalent of saying that little men carrying lanterns is what makes the light bulb light when you plug the lamp in)?

3. For instance, who would have thought that falling bodies were not accounted for by some force pulling them down, but by a warping of space-time?