WE NOW TURN to what you can call the quality of argument, your approach to argument as such. In that context, I want to discuss the topic of rationalism. This is the most widespread and harmful error of today’s intellectuals. It is virtually inescapable by college students, certainly by graduate students in the humanities. It also has the effect of enormously alienating nonintellectuals or nonprofessionals in the field. Rationalism is the arch-disease of today’s intellectuals. If you have not got it yourself, I congratulate you. But even if you have not, you still need to understand it. Otherwise, you will find today’s intellectual world unintelligible, including your professors and their publications, and you will simply end up, in effect, deciding that people are hopeless, and withdrawing. If you understand this error, however, you can inoculate yourself against it and explain the countless forms of it that you see around you.
In philosophy, the term “rationalism” is used to designate a broad school contrasted with “empiricism.” These are two technical terms, standing for two schools of philosophy that have been at war with each other from ancient Greece to the present day. Rationalists, roughly, are those philosophers who stress the role of reason or logic in the acquisition of knowledge, while disparaging or denying the senses. They include thinkers like Plato, Augustine, and Descartes. Empiricists, on the other hand, include people like John Locke, David Hume, and the modern pragmatists, all of whom stress the role of experience, or sensory evidence, while denying or minimizing logic. As pure philosophy, both of these schools lead to disaster, and Objectivism rejects both camps. Objectivist epistemology entirely repudiates the idea of reason versus observation, because it says knowledge is acquired by the use of reason applied to observation. Knowledge is thus the integration of logic and experience, not an either-or situation.
In this book, I want to focus on rationalism in a somewhat different sense from the technical definition that I have just given, and that is rationalism as a profoundly mistaken method of thought. In this sense, it is used not only by technical rationalists, but also by most of the people who call themselves empiricists, as well as by many others who have never heard of either school. The essence of rationalism, as I am using the term, is: deduction without reference to reality. In other words, this method consists of ignoring reality, ignoring observational evidence, seizing in a vacuum on some idea, some premise or theory that sounds plausible to the writer, and then proceeding to deduce from it a set of consequences. The result is like a floating construct. It is completely arbitrary, even though it may appear imposing. There may be a whole chain of “If this, then this; if this, then this; therefore, therefore, QED,” and so on, with all the apparatus of deduction—but it all rests on nothing, like a castle in the air. Very often it culminates in a conclusion that is blatantly the opposite of observed facts, but the person says, “What can I do? I have proved my conclusion, so much the worse for the facts.” In other words, the rationalist starts with an arbitrary preconception, unthinkingly taken as self-evident, the practical equivalent of Plato’s innate ideas. Then there is a chain of deductions, which usually follow quite rigorously, leading to a conclusion that completely defies reality. It is thinking without looking at reality; thinking as a kind of game of manipulating words without reference to what is out there, to existence; thinking in which the primary concern is the relation of one idea to another idea, not the relation of ideas to reality.
Ayn Rand once invented this eloquent example: A rationalist would come out with some argument such as, “Man has only two eyes; therefore, he should be able to see only two things, one with each eye.” The rationalist does not ask why; it sounds neat and symmetrical—two things, two eyes, it all fits. At this point, two schools of philosophy would arise. One would say, “We have to accept the conclusion. Men do see only two things; everything else, they do not actually see—it is all an illusion.” To that, the opposing school would reply, “Men do see countless things, not just two, but that is because of all the hidden eyes that they have.” If you then say to such a rationalist, “Look, it is not true; if you look at reality, you will see that men do, in fact, have only two eyes, and they do see many things,” he will find that irrelevant. Reality does not have any status in his thinking. He has decided that his conclusion about eyes is good, and his idea supersedes all facts.
What rationalism amounts to is obliviousness to reality on the part of a thinker. It is the placement of ideas above reality. Remember that the first crucial point of communication is that it is conceptual, and that you must always bear in mind that a concept is shorthand for concretes. Rationalism represents a total failure on this particular point. A rationalist’s concepts are cut loose from reality altogether, and he creates his own world of concepts. It has a definite structure (rationalists are very big on logic); there is a beginning, a middle, and an end; you always know where he is and where he is going; and there is always, “Therefore, therefore, QED,” just as in geometry, which rationalists love as a subject matter. But it all floats, cut off from reality. The conceptual shorthand, which is supposed to be our way of holding a mass of concretes in a reduced number of units, becomes a world of its own, just like Plato’s world of Forms. As a result, if you ask the rationalist about his imposing structure, pointing to his starting point and asking, “But why that?” he generally looks at you, baffled, and says, “What do you mean, ‘why’? That is where I start.” His whole structure, in other words, collapses very easily.
You can be a rationalist positively, using this method allegedly to prove something of your own—or you can be a rationalist negatively, using this method to refute an opponent. The latter leads to what we can call rationalist polemics. In this case, you are trying to refute an opponent, but not by pointing out facts of reality that he overlooks; you forget about facts. Instead, you show him that his ideas contain an internal contradiction. In other words, in this type of polemics, you accept the basic premise of your opponent. You say to him, “I will concede your fundamental point, but I am going to show you that somewhere along the line, you contradict your own beginning.” This is a very inadvisable method of arguing. There is obviously nothing wrong with showing a man that he has contradicted himself; that can be a valuable thing to do. But if you engage in this type of approach as a characteristic method of arguing, you tend, after a while, to become oblivious to reality. You train your mental eye to focus exclusively on your opponent, to see how you can trap him while accepting his premises. You train yourself to focus not on the relation of what he says to reality, but on a world of ideas apart from reality. This is very harmful to your intellectual development, to your ability to argue or present ideas. It trains you to widen a breach between your mind and reality. It also characteristically fails even as polemics, because once you accept your opponent’s basic premises, he sets the terms thereafter, and you are at his mercy.
This method of rationalism is characteristically used by theologians, to take another example. They have decided in advance of thinking—by means of revelation, they say—what conclusion their reasoning is supposed to lead them to. Then they simply ignore facts. They are oblivious to reality, but they want to “prove” their conclusion because they claim to accept reason, or at least many of them do. Therefore, they cook up an argument that will lead them to the conclusion they know in advance they want to establish. They juggle ideas around until they get them into the pattern in which the ideas allegedly justify the conclusion that was sent by revelation. This is reasoning as rationalization. It is not following reality, or following the facts where they lead, but going through a pretense, making up arbitrary premises and connections in order to come out at a prearranged spot. Unfortunately, this is something done more widely than merely by theologians. It is invalid, and I urge you not to do it.
Obviously, if people engaging in rationalism said openly what they were doing, there would be no problem in detecting it. Usually, however, rationalists do not exist in a really pure form. What you get in a typical real-life case is an individual who ignores reality, but not so consistently that you are dazzled by his ignoring it. He smuggles in facts and observations as necessary, because it is very hard, even in today’s schools, to be completely out of touch with reality. The result is the usual paper, which is half floating free of reality, half tied to some facts. Typically, though, since the facts are smuggled in, they are not introduced in the proper order, or with the right context, and consequently they are unconvincing and fall to pieces the moment you so much as breathe on them. Therefore, you have an allegedly convincing, deductive, impregnable structure that just dissolves into nothing as soon as you ask a few “whys.”
I want to illustrate this by reference to a paper on certainty that was submitted by a graduate student in philosophy a number of years ago (Appendix B). This paper is a very good example of rationalist polemics. Some of it, if you are a beginner in philosophy, may seem technical, but do not worry if any particular point or paragraph eludes you; it is not that important that you grasp every point of it.
Let us begin by observing that this paper on certainty is detached from reality in two crucial respects. It is from a graduate seminar course in philosophy, and the professor has asked the question, “Can you be certain? Is knowledge with certainty possible?” The student is arguing yes. Observe his method: He says right away that he will attempt to demonstrate his viewpoint “by pointing out that any other position involves one in a contradiction.” So it is perfectly well stated from the beginning that this is going to be internal polemics: “I am going to take the opposing viewpoint and deduce from it contradictions.” In what ways, therefore, is the paper by its very nature detached from reality? In the first place, there is no consideration in this paper of any argument in favor of skepticism, the belief that there is no certainty. There is no discussion of any reason anybody would be a skeptic, of what considerations, arguments, facts, or observations might lead to the conclusion that certainty is impossible. Reality is simply irrelevant. The author just begins, “My opponent has a viewpoint. Where it came from, I do not know; it is not up to me to ask. He has an idea, ‘There is no certainty,’ and I use that as my starting point.”
Observe the difference if you were to start by referring to reality. Suppose you said to the skeptic, “Why would anybody think that you cannot be certain?” He might say, just to take one of the many arguments that skeptics actually offer: “People hallucinate; they see things that are not really there—so how do you know, at any given time, that you are not hallucinating? Maybe you are always hallucinating, and therefore you never can be sure of anything.” That is not a good argument, but it is an argument that attempts to start with some actual observation—there is such a thing as hallucination, after all—and then to build a structure on it. If you began like that, then immediately you would have a line of response that would tie you back to reality, because—assuming you knew some philosophy—you would then say to the skeptic: “If you are talking about hallucination, we have distinguished hallucinations from times that we are not hallucinating; otherwise, how would we ever get the concept? As for how you know when you are not hallucinating, it is very simple—the difference between a pink elephant and a real elephant is that when you climb up on the back of a real one, it hurts when you fall to the ground, and you can ride the real one, and the pink one is only visual but you cannot integrate the other sense data, and so on”—and you can show the person that he is making a conclusion unwarranted by the facts. You remain in touch with reality, because you made the skeptic give an argument that started from somewhere. Not in this paper. This is rationalism, and there are no restrictions and no question where the theory came from. There is merely: “He said it, and I am going to deny it.”
The second way in which this paper is, by its very nature, oblivious to reality is that it offers no positive argument in favor of certainty. At no point does the author give a definition or a characterization of certainty, even though that is the central topic. It is certainly controversial, but there is no reference anywhere to what it is. The author is not concerned. His implicit premise is, “Whatever it means, I am going to refute its opposite anyway. Whatever you say against it, I am going to show that you are led to a contradiction, so what do I care what it means? I am not out to show that we can be certain by pointing to any kind of facts of reality. I am out to show that your statement that we cannot be is wrong. That is all. Therefore, it is up to you to worry about what it means.” Notice the difference if the student had begun simply by observing that we have minds, we are capable of perceiving reality, and sometimes the evidence is conclusive and sometimes it is less than conclusive, in which case we make a distinction between the certain and the probable; and, obviously, probable evidence presupposes certain evidence, because “probable” means less than certain, so first we have to grasp the certain; and each of the items of the probable has to be certain itself, so that the sum of it can be said to be probable; and so on. In such case, he would be able very easily to say, “By the nature of consciousness, we start out in contact with reality; we are aware of reality; that is certain,” and there would be no problem. This paper does not do that. It does not try to establish certainty positively, or even ask the other side to refute it negatively. It just starts, with the author ready to deduce contradictions.
This is unconvincing even on its own terms, because suppose the author does deduce contradictions from the original statement. The immediate question is left, and nowhere answered: Why does this thesis lead to contradictions? The student who wrote this cannot say. He does not know what in reality raises the topic. Therefore, all he can say, at best, is: “I can jiggle your formulation around in such a way as to come out with a contradiction”—but why, and what it all means, is not explained.
How, in fact, does the author go about establishing the contradictions? At each crucial point, he makes a passing reference to some certainty or other, some absolute, some framework that is serving as his foundation. Naturally, he does not say, “I am here taking this for granted as certain,” because he is not supposed to be taking anything for granted; he is supposed to be just refuting. Consequently, instead of forthrightly stating his viewpoint and his base, he smuggles it in. When he feels that he is sinking, he brings in something else as further ammunition. But it is not convincing, and, having conceded to the skeptic in advance the right to make the original statement that you cannot be certain, he will never convince anybody by that method.
Here are just a few points to illustrate the futility of this method. At the very beginning, the student says, “I want to point out that any position other than ‘certainty is possible’ involves one in a contradiction.” Any skeptic worth his salt would immediately say, “So what? Everybody’s viewpoint on everything is contradictory; that is my whole reason for being a skeptic. No one can ever attain consistency. So what does that prove?” Or the skeptic will say, “Just because it is a contradiction, how do you know it is wrong? We do not know anything—that is my whole viewpoint.” This student is trying to say that if you cannot know anything, then so-and-so follows. But think of that idea: If you cannot know anything, what follows from that? Nothing. If you cannot know anything, close up shop and go back to the cradle. You cannot say, “Let us assume we know nothing; now here are all the things that follow from total ignorance.” You would not get into that position if you said at the outset, “You are denying the existence of human consciousness, because certainty is inherent in consciousness as such.” The author, though, has left out reality. He is not saying what certainty is, or what the facts are, and therefore he is doomed from the outset.
To take another example, consider the infinite regress argument as elaborated in the second paragraph. It amounts to this: A skeptic says, “You cannot be certain,” and then his opponent says, “Are you certain of that?” and the skeptic says, “No, I am not certain of that,” and then his opponent says, “Are you certain of that?” and then he says, “No, I am uncertain that I am uncertain that I am uncertain,” and so on. That is something like party polemics. But by itself, without anything further, that is not a valid argument. Observe what would happen if you had a knowledgeable skeptic. A knowledgeable skeptic would immediately come back and say, “I can do exactly the same thing with ‘certain.’ I can force you into an infinite regress, because you say, ‘I am certain of so-and-so’—well, are you certain of that?” And you say, “Yes, I am certain that I can be certain.” “Are you certain of that?” You say, “Well, I am certain that I can be certain that I can be certain.” As many times as you can ask him is he certain of uncertainty, he can ask you whether you are certain of certainty. So you, too, are in an infinite regress. Immediately, of course, you feel in your mind that there is some funny difference here. But this paper does not say what the difference is—namely, that both sides are claiming knowledge, and knowledge implies certainty. The advocate of certainty is happy to say, “Yes, I imply a whole infinite chain of certainties,” but the skeptic cannot be happy, because he is denying certainty and therefore systematically contradicting himself. What is it, though, that we have to establish to make that distinction? Why does knowledge require certainty; why does every claim commit you to certainty? The fact of infinite regress does not establish that. What does? The fact that certainty is inherent in being conscious as such, and skepticism is the invalidation of consciousness. Since that is so, any skeptic who opens his mouth is forced into a fatal infinite regress, because he is trying to say something while saying that he is incapable of it—but only because of the essential fact here, to which the author of this paper is oblivious.
In the third paragraph, the author goes on to the next step. He says that this version of skepticism is itself not an intellectual position, that nothing is being said, and therefore, nothing is being asserted that you could consider or refute. That is the neatest trick of the week. Not only has he just refuted the skeptic’s position; he went on at great length to do so; but if nothing has been said, how could he be doing this? If nothing has been said, that would be as though the advocate of skepticism were saying, “Ish de triddle da gloop de tweedle.” There, nothing really is being said, so you cannot refute it. Here, though, the argument is, “You said nothing, and here is what it leads to.” Of course, what the author means is, “You are not saying anything with certainty.” But the skeptic comes back and says, “Of course I am not saying anything with certainty. My whole point is that I cannot; you cannot; there is no such thing as certainty. Why should there be?” There is nothing in all of this that answers that point. The author concedes the opponent’s premise. He is allegedly going to show what follows from it, but nothing follows from it, and in conceding it, he has wiped himself out.
Observe also that in the same paragraph he says, “The truth of this derives from the Law of Excluded Middle, which states that anything either exists or does not exist.” Notice that we are leaving open the question of whether you can be certain of anything; we do not know that yet. Right at this point, though, the author brings in a basic law of logic. He has not said, “My foundation is that certain axioms are absolutes, and logic is one of them,” because then the question is, “Can you be certain?” and he is omitting that. But if you cannot be certain, how do you suddenly bring in the Law of Excluded Middle, as if no one could possibly contest that? If the skeptic is going to contest all consciousness, he is certainly going to be prepared to contest the certainty of logic.
I am passing over many other points, but I want you to see how what seems to be at least partly plausible just simply disintegrates. Take a look, for another example, at the fourth paragraph, in which the author is trying to imagine that the original position has been amended to take account of his allegedly destructive criticism. The skeptic’s amended view is now, “You cannot be certain of anything except one thing, and that is uncertainty.” And now the author is going to say, “Let us accept what the skeptic says as the given; I will not bother with the fact that it is a contradiction. Now I am going to show all the other things he says that he would also have to be certain of.” Let us see if any of this stands up. Imagine what a skeptic, standing over the author’s shoulder, would reply to each of these arguments. The author says that the skeptic must be certain that something exists about which he is uncertain. The author thinks that he has established existence, so he is happy. But what would the skeptic say on the basis of what has been stated here? He would say, “Yes, something exists about which I am uncertain, namely uncertainty.” You might say, “What are you uncertain of? How can you have uncertainty without recognition of reality?” But that is the positive base that has been left out. If you just go by the formulation, “Something must exist about which the skeptic is uncertain,” the skeptic would say, “Yes, uncertainty exists, and that is what I am certain of.” To the next phrase, “He who is uncertain exists,” the skeptic would say, “Indeed, and all I am is a state of uncertainty.” Or the author says that the skeptic, “who is certain of uncertainty, is capable of being certain of (at least part of) that which exists.” To which the skeptic might reply, “Yes, the part I am uncertain of is certainty.” Or the author says that the skeptic must be certain that you can communicate to others; the skeptic would say, “I do not know if there are others; maybe there are, maybe there are not.”
As another example of rationalist polemics, look at the next paragraph. The author is starting a new tack now. He is still not satisfied that he has really wiped out the skeptic, so he is going to try to show that all of the arguments for skepticism must have something in common. He does not give the arguments for skepticism, because he is here concerned not with facts, but with deduction. So he is going to show what the skeptic argument would have to be if the skeptic were to offer it, which he has not. This is very typical of rationalist polemics. The author figures, “How could my opponent defend this case, even if it is not, in fact, what he says?” Then he refutes this complex argument that he has made up. Almost invariably, his opponent then hears it all and says, “But that is not the argument I would have offered.” And the rationalist is very disappointed, because he has analyzed the whole thing, only it is not the argument.
There are a number of standard arguments used to justify skepticism, such as the existence of illusions, hallucination, or error. But the author of this paper has made up his own: “The skeptic would have to show that the very concept of certainty is itself contradictory.” Why? He says, “That is the only thing you could know could not exist—a contradiction. So if you know that certainty does not exist, you have to know that it is a contradiction.” How does the author establish that that is the only way you could demonstrate the impossibility of certainty? He says nothing about that. He has in his mind, “We know that ‘A is A’ is an absolute; and if something violates ‘A is A,’ it could not possibly exist”—neither of which points, of course, he has made in this paper. He is prepared to stop there; if there is any other way of something not existing, he cannot think of it. To a rationalist, this sounds good enough. He thinks that gives him a nice firm basis to go on, and then he is going to show that skepticism cannot hold this position. But the whole thing is completely arbitrary. All the skeptic has to say is, “Where do you get this idea that the only way to prove that something does not exist is by means of a contradiction?” There is no argument offered in this paper for that idea, which is certainly much more controversial than the question of skepticism.
The same thing is true in the next paragraph. The author is now trying to show that the skeptic’s position is meaningless because contradictions are meaningless. Observe that he has left open the question, “Can you be certain of anything?”—in other words, “Is man conscious?” He is prepared to leave that debate open. But he does know for certain that in polemics, he can use the idea, “Anything contradictory is meaningless.” He has a theory and a definition of meaning, but he does not know whether man is conscious. How does he know that his theory of meaning is true? There are all kinds of questions here that the author is simply oblivious to. If it is true that everything that does not refer to reality is meaningless, what about the term “Santa Claus”? It does not refer to reality, but is it meaningless? What about the term “God”? It does not refer to reality if you are an atheist, but is it meaningless? If it is, there is an awful lot of discussion that is completely meaningless.
This author simply decided as follows: “I have got a good argument here: If something is contradictory, it does not refer to reality, and if it does not refer to reality, it is meaningless; therefore, if it is contradictory, then it is meaningless. The skeptic would have to show that certainty is contradictory, but then certainty would be meaningless—and then his argument is wiped out, because his very position is meaningless. That is what I am going to do.” He is in his own rationalist world; he has got his perfect little deduction, which ends up wiping out the opponent’s position, and he is simply oblivious to such an obvious objection as, “How do you know what is meaningless, and what about these words that do not refer to reality and are yet meaningful?”
That much should be ample simply to give you an illustration of rationalist polemics. This paper is well structured and clear, and it did receive a grade of A, but it is thoroughly rationalistic. Even if you do not remember the particular examples, or if you find any particular point abstruse, you should try to just hold in mind such parts as you can get, because it is such a terrific example of this problem that I could not resist using it.
To conclude, let me tell you that the author of this paper was me. This was one of the first papers I submitted in graduate school, and I did not do it as a literary exercise. I thought this was perfectly good at the time that I submitted it. I was quite convinced; I thought I really had refuted the skeptics. This was my first year in graduate school, and this was entirely how I had been trained. It took many years to overcome this kind of rationalism, but I think I have shown you that across time you can do it.