"Certainty" by Leonard Peikoff

Good morning. Today we reach the climax, in a way, of this series because we are going to focus on the reward of the art of thinking, that is, the art of thinking correctly, the art of integrating, suppressing incorrect context, thinking in essentials, thinking in principles, following hierarchical structure and the rest that we've discussed. The reward, cognitively speaking, is certainty. And that is the main subject of our talk, of our lecture today. Now, this is not a new topic for this class in the way in which principles, for instance, was largely new because I covered certainty at some length in my book on Objectivism. But there are a lot of aspects of certainty that I would have included in a book on certainty which I had to ruthlessly excise because my goal was to have the total of Objectivism between two covers. And if I had treated certainty as extensively as I had wanted to, it would have ended up with a book on every chapter to give parallel treatment, and probably a book on every sequence. And then still we would not know what is objectivism, because you couldn't hold it all because of the crow. But in any event, there's a lot of important points in regard to certainty in clarifying it that I had to omit from the book. It's in the nature of applications to various areas that trouble people. So for today, I've taken four of the most difficult or complex but common problems, confusions that Objectivists have in regard to certainty, even when they accept the basic ideas in my book. All of these problems center one way or another on the issue that certainty is contextual. And that is such a revolutionary idea that the full implications and applications of it are often difficult for people to grasp. So I've really just chosen four topics. But before I go to those, I want to review three highlights from the book as the main background. From my book as the main background. For that, I want you to keep in mind for this discussion today. Now, the book was simply too heavy to carry all the way from California, so I just xeroxed a few pages. If you have it, it's page 172. I'm going to start there, the bottom paragraph. Man is a being of limited knowledge, and he must therefore identify the cognitive context of his conclusions. And here comes the key sentence from the point of view of our present class. In any situation where there is reason to suspect that a variety of factors is relevant to the truth, only some of which are presently known, he is obliged to acknowledge this fact. The implicit or explicit preamble to his conclusion must be on the basis of the available evidence that is within the context of the factors so far discovered, and the following is the proper conclusion to draw. And then thereafter, the individual must continue to observe and identify. Should new information warrant it, he must qualify his conclusion accordingly. Now, the key sentence. There is, in any situation where there is reason to suspect that a variety of factors is relevant to the truth, only some of which are presently known, then you have to acknowledge it. And that is going to turn into the question right away. Well, how do you know whether there is or isn't reason to suspect that relevant factors exist which you don't know is every principle that we come to in any field in that condition? Because, after all, we're limited in knowledge. How do we know we're not going to learn something more? Or is there a division and what counts as reason to suspect? Now, I worded that, you know, like a legal document that was going up against the lawyers of Perot, so that I had every word there. Reason to suspect, not even reason to think. Reason to suspect. I didn't say how many factors. A variety of factors. I didn't say prove the truth is relevant to the truth. I've got it. As specific as it needs to be to have meaning and as general as it has to be in order to capture everything that's involved. And one of the things we want to do as we discuss these topics is clarify what that means and how it applies. Now, another thing I want to call to your attention is the paragraph on 175 refuting the idea that an absolute is some principle or idea independent of everything else, independent of every other factor, cognition, which is the standard definition of an absolute, since Kant and I have a whole reputation to that. You remember the philosophy professor who says that gravity is not an absolute because its operation depends on circumstances and therefore it's not isolated. It can't be 100% certain, because how do you know that circumstances won't affect it? And then this is my summing up paragraphs. The modern definition of absolute represents the rejection of irrational metaphysics and epistemology. It is the inversion of a crucial truth. And this is in italics. And I don't often put a whole sentence is in italics. I dole that out very sparingly. But this one is relationships are not the enemy of absolutism. They are what make it possible. Relationships are not the enemy of absolutism. They are what make it possible. We prove a conclusion on the basis of facts logically related to it and then integrate it into the sum of our knowledge. That Process is what enables us to say everything points to this conclusion. The total context demands it. Within these conditions, it is unshakeable about an isolated revelation by contest. We could never be secure since we would know nothing that makes it so. We could count on nothing to keep it so. Either which I think is a great line. The fact of context does not weaken human conclusions or make them vulnerable to overthrow. On the contrary, context is precisely what makes a properly specified conclusion invulnerable. Now, there are many cases where it doesn't seem as though that's true. And some of those I want to discuss tonight. To hit that on the head for one tonight, this morning, to hit that on the head definitively. And now one last paragraph tracing the need for the concept of certainty, which I want you to keep in mind before we start. The concept of certainty designates knowledge from a particular perspective. It designates. And now here comes two key words, some complex items of knowledge, not all complex items, and no simple items, but some complex items of knowledge considered in contrast to the transitional evidential states that precede them. And then I say in parentheses that by extension the term may be applied to all knowledge, perceptual and conceptual, to indicate that it is free of doubt. But that is not the essential and important meaning of the term, and it is not the sense we are using it in. Here. A conclusion is certain when the evidence in its favor is conclusive, that is, when it has been logically validated. At this stage, one has gone beyond substantial evidence, rather, which would be probability. Rather, the total of the available evidence points in a single direction, and this evidence fulfills the standard of proof. In such a context, there is nothing to suggest even the possibility of another interpretation, and there are therefore no longer any grounds for doubt. Now, that's basically the bare bones of the material that I am counting on. Now, I just want to say a few words right here before we develop on this issue, that certain applies to some complex items of knowledge. In simple direct matters, there's only two states, nothing and something. No grasp and grasp. You see it or you don't. Print as you see a table or you don't. There's no gradual accumulation of evidence. You remember your wife's face. The image occurs to you or it doesn't, and there's 100,000 more than that matters like that. In cases like that, there is no gradual adducing of evidence. There is no transitional continuum of getting closer and closer to conviction. There is no progression from possible to probable to certain, and there is no possibility of doubt or error. Now, if all knowledge was in that state, the concept of certain would be odious. You know what I mean? Useless. Certain applies really to inductive integrations of data. And not just inductive integrations, but higher level. That's what we mean by complex, because the ostensive, what you can directly perceive, is cut and dried. There's no question of it. You either get it or you don't. And I say complex because even the early runs of the higher level, such as, for instance, on the level of furniture versus table, may be so straightforward. Either you get it or you don't. The reason I say in, that's why I say some complex, because there can be complex things that are absolutely cut and dry, particularly also in mathematics. But I say the issue is inductive, because only in induction are you truly collecting data gradually as an indispensable means of reaching the truth. When you want a form of generalization, for instance, there is no either or of the kind you have in the instances I cited. It's not like seeing a chair or not. It inherently requires a whole number of observations, a process of abstraction, integration to this and an integration to that, until it gradually masters the whole field. And the same is for formulating a scientific theory or a hypothesis. Again, you have to accumulate evidence gradually to show that this explanation integrates with and would account for everything that we experience and nothing else would. And the same applies to a prediction of the future. I can't perceive it yes or no, the way I can the present. It requires to make such a prediction considerable knowledge of the laws governing the world and of the current status of the relevant variables. And that kind of information comes in gradually. These are the kind of cases where certainty applies as a crucial new concept, and what it then signifies is the end of the continuum from possible to probable to certain. And that is the sense in which I use it. So I don't make much of certainty in the sense of. I'll answer all skeptics by saying, well, all of our knowledge is certain. I don't think we would need the concept certain. All we would have to say is, I know. It's because of this category of gradually accumulating inductive knowledge that we need the category. All right, now let me plunge into the first of the common questions I get about certainty. Certainty about the future, now, about the present and the past. Most people have no problem, if they haven't majored in philosophy, with grasping that certainty is attainable because it's clear the outcome has already happened. If I say we are meeting at 10:40am on July 7, 1992, as part of conceptual conferences for this summer, we know that right here and now, as of this minute is happening. And there's no question, is it or isn't it? If I say Bush withdrew too early from the Iraqi Gulf War and consequently has caused a disaster, that's in the past, and so there's no doubt whatever about that. But if I say Bush will lose the election, that's in the future. And can we be sure of a statement like that? Can we even be sure of a statement like we will meet again in two days? Because what if I get struck by an illness, or the building gets struck by a bomb, etc. And so on. Now, the answer to all these questions about certainty in regard to the future lies in the idea of certain within a context. An example I'm going to use, I've used before in courses, but it has a special gripping quality to people because a lot of people are worried about flying, and most of you probably have to fly to get out of here. So this would be a way of talking you into the fact that you have every reason to think your flight will be safe if you do. But we're going to ask now just as a. As a question, will your flight home on the airplane be safe? That's a prediction. And the question is, can you be certain as you step on the plane that this flight will land safely? Now, you remember what we said. You can't have certainty only if you have a standard of proof. What would be the proof that it's safe? Well, the answer is you have to identify all the known conditions on which the outcome depends and then check each against your specific plane. If everything meshes and nothing else is relevant to your knowledge, assuming you have substantial experience, then you are certain. But if some of the data contradict, then you cannot have certainty. You haven't met the standard of proof. The most you can have is probability. Thus, suppose, just to simplify it, we know that the following are the determinants of safety. Now, I just wrote down some airplane experts that obviously put down many more. The record of the pilot and also his current condition of health and sobriety, the maintenance and condition of the aircraft that is loaded with fuel, the history of this particular type of plane, the weather, the size of the load, et cetera. You can add, there's obviously several others, but you enlist whatever within your knowledge. And if you're not sure, you ask an expert. There is nothing wrong with Asking most of what we do in a division of labor society is rely on an expert. And that's fine so long as we choose the expert according to rational qualifications and then don't just follow him blindly, but he has to make sense within his own terms. But in any event, you come up with a list of a plane is safe if A, B, C, D, E. And then you say this flight is A, B, C, D, E, therefore this plane is safe. Now that is actually deductive reasoning. And therefore the problem is to get the major premise, the plain is safe, you have A, B, C, D, E. But the point is this major premise is contextual, and that's where the phrase comes in. Do you have any reason to suspect that another factor would be relevant that you haven't yet considered? If so, if you have reason to suspect it, then you have to check it out. But you have to have reason. You can't say, well, maybe the state of politics in Ecuador would affect the oil deliveries from Venezuela, and maybe as a result they're starting to dilute the gasoline, maybe it's going to crash. You have to have a reason to suspect, not to know, but to at least justifiably say, is this a relevant factor? Now, in the context that we normally operate in, the experts know all the reasons that there are a factor and they check them all out and assume you go to a reputable airline, you check its record, and so on. Within your context, you have to say, I know that this plane is safe within the context. Now contrast that, for instance, with the prediction that a flight on the same airline two years from now, from here to New York or wherever, will be safe now. Two years long time in the future. It's all different in the world between this Sunday, because they got a lot of questions. Will the world still be around in two years? Will Pan Am or whatever your flight is still be in business, etc. And the farther you go into the future, the more unknowns, the more the structures of the plane will be modified and the less you'll know about which ones are safe and which aren't. So as you get farther and farther into the future, you are less and less certain of any given concrete. This does not mean you are less certain of principles, of philosophy, of generalizations, but their application to a specific prediction in the future becomes more and more problematic as you go farther and farther, because there's more and more possibility of something coming up that you hadn't counted on. So it has to be your predictions about the future have to be Within a framework where you have reason to say there is no basis to suspect any emergence of a new factor within this time frame that would affect this prediction. Any predictions about 50 years from now, therefore, are out of the question. Because so much can happen in 50 years that you simply have to make it probable. Unless you do it the way philosophers make predictions, which is they say this country will ultimately become a dictatorship. Now no one can refute that, even though it's a prediction about the future, because if you die, it's not yet ultimately, it's ultimately as whatever happens is finally it's going to be that way. So, so you're safe as a philosopher in making a prediction. But if you're making a concrete prediction about two weeks from now, that's a different proposition. The present future can definitely be certain contextually. Now let's answer this question. Let's say the plane fulfills all the requirements and conditions you outline and you've got an expert, you've gone over the whole thing with aeronautical engineers and et cetera. Couldn't somebody still say, well look, there's still a little time between now and then. Couldn't there be a bomb on it that we just can't discover in our search? Or what if a war breaks out, like for instance the LA riots broke out when I was in New York and the plane could not land at LAX because it was densely covered with smoke. Or what if a meteorite that is presently simply unpredictable came down and crushed it? Now you have to just take my word for it. Any semi inventive person can think of a hundred things like that about any prediction about the future, even including the fact that you're going to turn your eyes in the next second to the right because he'll have some ray striking you before you can do it. Now what is the answer to this question? This is exactly the kind of thing that is covered by contextual specification. If you have diligently examined the case there is no evidence of a bomb, etc, etc, of a meteorite and so on, you are entirely justified in saying within the context of evidence available, it is certain that this plane will be safe. Now let us just assume the worst. Assume that a meteorite falls and hits your plane in midair and you crash. Were you wrong when you said within the context of my knowledge, it is certain that the plane will be safe. And then bang, this meteor comes out of the blue and it hits you and your plane crashes. Were you? Do you then say in your dying breath, I was wrong, the plane is not Safe. No, you were not wrong. You committed no fallacy. There was no lapse in logic. There was no misinterpretation of the available data. It was a bolt from the blue in a literal sense which could not be foreseen. And that is precisely what is covered by. In the present context. That means, judging by all the evidence now available, this is the true conclusion, and that is and remains true when later evidence comes that doesn't contradict the former. The later evidence tells me there is now another piece of evidence which I had no basis to hypothesize at the time. And in this new context, there may or there will be a crash. That is exactly what's covered in saying in accordance with my present context. In other words, you cannot call it an error that you didn't know everything or foresee everything. You were right, you were certain. So now you committed no fallacy and no lapse in logic, and you covered yourself about a new condition precisely by saying within the context. Within the context is tantamount in these cases to saying, barring unforeseeable accidents, there will be a safe flight. And this is absolutely true. And nor is it ever contradicted by given accidents, there won't be a safe flight. Nor can you say, well, because I can't foresee an accident, maybe it's possible. If you're going to hypothesize it, you have to have some grounds to hypothesize it. Now that brings me. All of these questions are interconnected. That brings me right away to the second big question with regard to certainty. I said, so you can put a line now, although this is a continuous discussion, I said, you have to have the real, real evidence to support a future possibility like a crash before you can say it's uncertain what's going to happen. And at this point someone will come up inevitably with the statement, but there is always real evidence for everything. There is always some real evidence for every conceivable possibility. And where does this person come up with the evidence for anything? With a word starting with S, T, A T, statistics. Now, there's statistics, actual or possible, on everything, and there's always one out of some number that happens. So isn't there always some evidence against your prediction? And if that is so, how can you ever say this is 100% proved? You know, there's a statistical chance. And that being so, aren't you out of the picture now, for instance, you can go through all of this, you know, study of the plane and so on, and someone will come up to you and say, well, look, it's a fact that one out of a thousand planes crash. We have kept statistics. Every thousand flights, one crashes. That's it. Now, you don't know you could be one of those thousand. So you have one chance in a thousand. As long as there's one chance in a thousand, you're on a plane, it's uncertain. You don't know that Pigov is going to show up on Thursday because according to, you know, the latest study From Johns Hopkins, 1 out of 75 lecturers does not show up on a Thursday. And you've got great chance that he will. 74 out of 75. But there's still evidence that he won't. And the evidence is that one out of every 75 don't. Now you see that you can do that with everything, depending only on your ingenuity and your ability to find figures. Now let us take this question and ask, well, is it then really possible for the airplane to crash if it's true? Well, let's assume the statistics are true because they make no difference that there's one in a thousand that do crash. The first thing I want to do is remind you of a distinction I made in my lectures but did not put in my book between metaphysical and epistemological possibility. I used that distinction, but I didn't call it that. If you say, is it possible for, for an airplane or this airplane to crash, does it have the capacity, the capability of crashing? That is the metaphysical sense of possibility. The sense of does this entity have a certain potentiality? And if you say yes, that tells you something about the nature of the entity, because what it can do or what can happen to it is a function of what it is. A feather, for instance, can't crash. Or if there was an anti gravity ray that absolutely blanketed the sky, then the plane could not crash. It literally would have no capacity to fall. But as it stands, certainly the plane can crash. It has that capability as against a feather. But this is very different from the epistemological use of possibility. It's very different from advancing a hypothesis about this particular situation. It's very different from saying maybe it will fall today. Metaphysical possibility says only the thing has a certain capacity under certain conditions. But it is your obligation to show that those conditions are met before you can hypothesize that maybe it will fall today. Now, you could say maybe if there was some structural damage to the wings and. Or there was a severe storm that you hadn't anticipated and. Or the pilot comes out of the cockpit reeking of rye, then you have some evidence. You say, you know, maybe this thing is going to fall because you have evidence that some of the conditions of crashing are being met. But in the normal case, all of the data is pro safety and there's nothing whatever in favor of the crash. And, and you cannot hypothesize a maybe out of nothing. Now, if you keep that in mind, let's come back to. But Statistically, there is 1 in 1,000 chance of crashing. If that's true, I just made up that statistic, so isn't that always evident? Now, I want to dictate to you a sentence, but I tried to work exactly, because it states for the first time publicly, the objective is position on statistics. So this is basically out of Ayn Rand's mouth. To me, statistics are applicable only when one, you are unavoidably ignorant about a concrete. Or about a given concrete. And two, some action is necessary. In other words, it can't be deferred. Now let me repeat that. Statistics are applicable only when one, you are unavoidably ignorant of a given concrete, and two, some action is necessary. In other words, it can't be postponed or deferred. You've got to do something now. You have to simply have to act. You can. You're in an emergency situation. Now let's concretize this. We take the same example about the plane, but you have to imagine, just to simplify, you're in a desert. Somehow you got there, we just dropped you off, you come upon your pilot, but you come upon a plane. You're a pilot in the sense that you can drive the plane, but that's it. You come upon a plane of which you know nothing. You're not an engineer, you don't know its structural integrity. All you can see is a plane, it's got two wings. And I bet you if you do such and such and start moving, the thing would get us up in the air. And the question is, should you decide to fly it out or not? You can't just sit there and say, I'll wait till the evidence comes in. Because you could die by the time the evidence comes in. Now you say, do I have any alternative? Well, you say, I'm in the desert. I have no idea whether there's an oasis around the corner or so many miles from here that I'll die of thirst before I get there. The first might crash and kill you, and the second, you might parch to death and you have nothing that you know about your concrete case. No maps, no idea of the topography, no idea of the safety of the plane, no idea. If you take 10 steps, you're going to be so weak, you know that you're going to collapse. But if you, if you take this thing 10ft off the ground, you're going to crash. And that's your only two ways out. Now, if at that point you could make a phone call, obviously you would phone to the plane manufacturer if you could find them, or at least to the local authorities and ask to deliver a bottle of Coke. But supposing that you could only phone a statistician because nobody else knows anything, and you get only two statistics, People who fly airplanes, one out of a thousand crash, taking into account all planes of all kinds. If you don't know where you are. But when people wander around in the desert looking for an oasis, one in 2,000 die of thirst. That's all you know. Then if you say, I gotta do something, I got a better chance of finding the oasis, I got two chances in a thousand, as opposed to getting out by plane, where there's only one chance in a thousand. So you go for the oasis. Now that this kind of case does have real application very often. But the condition of it is you are ignorant. You have no way to get access to the facts of about your different alternatives. So you just try to take the best one on blind statistics. And surgeons often have to do this, isn't that right? I see you nodding, Dr. Laporte. You don't know if a tumor is malignant or not. Now, sometimes you can test for it or take biopsies and so on, but in this case, let us say there's no way to get at it, to inspect it, so there's no way to determine is it malignant or not. And you consult statistics and you find that there's one out of a hundred chance it is malignant. And if it is, action is imperative. So then you take a look at the statistical results of the operations and you find, let us say, in what, 5% of the cases they die and 95 they're cured. And then you have to put the statistics together and decide where is the best chance for a completely unknown situation. Well, I don't know if he's got it, but there's such and such a statistical chance, and I don't know what the complications will be, but there's such and such a statistical chance and see whether he has a better chance of remaining alive from blind statistics, leaving them alone or operating. But while there are cases like that, that is not the normal case. That is not the case in which most people use statistics. First of all, let me point out to you that when you rely on an expert, that is not the same as ignorance in a division of labor society. As I said earlier, much of your evidence is indirect through experts. That is, you count on the plane, you count on the food in the restaurant not being poisoned, et cetera, by having a certain reason to say, I trust this person, this is the background, this is the context, and so on. Therefore, I have no reason to suspect otherwise. But that is, if you're rational in choosing your experts, whether they're cooks or aviators or whatever, that is still an evidential basis for your action. Now, of course, if you just take some expert on blind faith, then it is not rational. If a doctor tells you to amputate your arm because he read it in the Reader's Digest, you say, well, he's a doctor, he should know, then you definitely are not rational. I'm not countenancing that sort of thing. But if you choose your expert according to the record of his experience, his knowledge and so on, and that what he advocates makes sense in your own terms, he is a man who has access to the evidence, so you are not ignorant. And that's of the concrete in question. And that's the situation in ordinary life. The point here now is, therefore you can put it this way. Knowledge of the concrete, whether that knowledge is direct on your own or indirect through experts. Knowledge of the concrete erases the applicability of statistical probabilities. How's that for a mouthful? Knowledge of the concrete, whether direct or indirect through experts, erases the applicability of. Of statistical probabilities. In other words, when you know something about a given concrete, you have no business applying statistics to it. Statistics are an expression of ignorance. They are a quantification of ignorance. They are not a form of cognition. They are not knowledge. They are a calculation to permit action in the face of ignorance. As soon as you have knowledge of a particular, the statistics become insignificant. They are displaced, they are pushed out of the picture by your knowledge. So if I know one in a thousand planes crash, and this plane has been studied thoroughly by a proper expert, he wipes out the one in a thousand. I mean, you can see this on obvious examples. I say one in a hundred of my students are libertarians, masking their identity. So maybe you are one. Now, that's fantastic. Even if it's true that one of a hundred are my Knowledge of him completely erases that possibility. I have no business applying statistics if I know him. Now, if you're a completely unknown audience to me, I'm entitled, and I'm talking to Norton about the question, period. I'm entitled to say there's a good chance there's going to be a couple libertarians in there, because about 1 out of 100 is. But I'm not talking about anybody in particular, just saying somewhere. But as soon as you come to a particular prediction, you've got to take into account what you know. And as soon as knowledge is applicable, statistics are out. So then let's push the question to this. Can you say that it's impossible for your plane, given the proper conditions, to crash? Well, impossible is the negation of possible. So it depends what you mean by impossible. In the metaphysical sense, it's always possible. It has the capacity to do it under certain conditions. But in the epistemological sense, in the case that we've outlined, it is non possible. In other words, you cannot justifiably call it possible without some evidence. And in this case we are assuming there's no such evidence. So in that sense, it is not, quote, possible that it will fall. There is no evidence to sustain a maybe. Now, you still need to say within my present context, or accidents aside, but then remember, you need to have a basis to hypothesize any specific accident. Now, if you try to concretize the alternative, you will see that it has only one result, and that is total psychotic paranoia. If you went by statistics and could hypothesize epistemological possibility just in a vacuum because you don't know anything. You couldn't go out on the street because one in a million people are hit by a meteorite. Every sound you hear at home means maybe it's a burglar. Every quarrel you have with your wife or husband means maybe it's a divorce. Every car you see in the street, maybe it's an accident. Every stranger you meet, maybe it's a child molester. Now all of these, that last one is already current wisdom in the country, but all of these are the result that you would get into if you said you don't need any context, you just take from statistics. One in 100 are child molesters. You're one, therefore maybe you are. That is completely invalid reasoning. And the end of it is simply to be consumed by terror. And that's why I call this use of statistics, although it poses as being dryly mathematical and numerical. I call this A blatant emotionalism. It's fear determining epistemological status. Fear versus evidence. In other words, emotion above reason and reality. Let's sum up this second topic with the following observations. You do not plan for a metaphysical possibility in the absence of specific, not statistical, evidence for it. You do not plan for a metaphysical possibility in the absence of specific, not statistical evidence for it. One in 87 women have twins. I believe this was true before the fertility drug. That's probably less than that now. It's possible for a woman to have twins metaphysically, but once you see this woman and she's had a sonogram and there's quadruplets or there's a single baby, in either case, the one in 87 become completely irrelevant. I could just as well say one in a thousand is a foreign agent. Maybe you are. One in a trillion is a barber. Maybe you are. I mean, you just cannot walk around talking like that. It's completely baseless. It's only when action is necessary, which is completely blind, that you latch onto probability or latch onto statistics, because you have to be blind and yet you have to act. Now, I want to conclude this point by telling you how this originally came up in my discussions with Ms. Rand. Most of the formulations giving you were hers to me and erased to take them down. But I didn't want to put it in the book because I felt it's too detailed. And if I give this level of detail, uncertainty, the whole book will have to be on that level of detail. So there was a lot of things that I was very much in conflict about. They would clarify things, but then they raised questions. Why don't you clarify equally in other chapters? So I had to just keep my hand off with the idea that someday I would give lectures on what was left over, of which these are some. But this, I tell you, is the story of how this topic came up. I was talking one night with Ms. Rand about statistics, which she did not like. I mean, she allowed their use under the appropriate circumstances, but she certainly did not favor that mode of approach. And I said to her, well, let's just work out a particular case to see what it would involve. And she said, okay, what case? I said, we were in her apartment on the sixth floor. I said, well, how many people are in the lobby right now? What is the chances? Let's figure it out statistically. And she said, well, you go ahead. How would you figure it out statistically? I said, well, I don't have the exact figures, but the idea would be it's night now, so there's fewer chances that a person would be in the lobby. They say 1 in 100, whereas during the day maybe it's 1 in 20. On the other hand, it's New York City. So chances maybe one in a thousand that someone will go out at night versus in Boise, they might go one out of three. It's a really warm evening, so the chances are much less. And however the moon is out, so the chances are more. And she looked at me, she said, what are you going to do with all that? I said, well, you try to get a model, you know, put it all together so you end up with a net, you know, one out of whatever under these conditions. And then, you know, and she looked at me and she said, that is the most fantastic thing I ever heard. And I said to her, well, what would you do if you wanted to know how many stages are in the lobby? She said, I would take the elevator and count them. And that's where she made the point to me in emphatic terms. Never use statistics unless ignorance is unavoidable. If you have a way to acquire knowledge directly, it is an aberration, it is a perversion, it is a corruption to use statistics to try to figure out the answer. So I think that should clarify this issue for you about is there always evidence? There's always statistics, but there is not always evidence. I then want to take the break a little early so I don't get caught in the middle of the next point. So let's take a. A nine minute break. I'd just like to say that I've already received seven or eight letters from people, letters fulfilling the assignment for Thursday. So I'm giving them priority if they're what I want, if they're either very good or very bad, I can't use them. But if they're just right, you know, with enough wrong so you can learn something, I have already too much. So don't do all the work of handing it in on the certainty that you're going to be picked because some people are already ahead of you. But if you want to try, there's still a chance that there'll be some new ones left. Okay, now let us turn to the third of the four topics. They're all related one way or the other, as you see to knowledge is contextual and it's applying that point. If knowledge is contextual, is it necessary or proper to say the preface in the present context about everything, Do you say? And this I put on the brochure in the present context of knowledge, the senses are valid. In the present context of knowledge, Adler shrugged is a great novel. In the present context of knowledge, we're having a lecture. Well, obviously, no, you don't say it about everything. So of what knowledge, if any, do you specify in such and such a context? And of what knowledge, if any, do you not? Is all knowledge contextual in this sense or not? Are there categories of knowledge about which it's inappropriate to say it's true? Quote in the present context. Now, the answer is yes, there are. And it comes down to the same issues which make the concept of certainty necessary. The following are not contextual in the sense in which you have to say in the present context, Perceptual cognition. And I'm including in that memory, direct memory, and I'm including all automatized conceptual identifications. So I see this platform that's a perceptual. I look at it, it's a platform automatically. The concept is so integrated and so obvious that there's no process of thought at this stage of the game. And I have a memory. There was one yesterday. All of that just simply comes to me directly. It doesn't mean you couldn't create a context when you didn't learn it. It means that at this stage of the game, it automatizes and appears to you as yes or no. You've got it or you have it. There is no accumulation of evidence to get a better and better case. And of course, for percepts, there's nothing antecedent to serve as a context. As infants, we add sensations, which were the context for percepts. But as adults, we don't have the sensations anymore, only the percepts. And therefore, as far as we're concerned, they come just flat out of reality. We just look and see. Now, I say the automatized use of concepts, because contrast that with looking at a person who's angry and saying he's hostile. Now, that is not like this is a lectern, that's a picture, etc. You're there into a whole character of the person, not just one observation. And you can look at him and say, he's holding a knife, dripping with blood. But to say he's hostile, you have to know more about what are the reasons, and so on. It's just the same way that a doctor can have a completely automatized vocabulary and he can know the meaning of pneumonia perfectly. But when he looks at a given patient, he may not know is this pneumonia or not, because there's so many factors that have to be checked and so many things he has to discriminate it against. And some counts are blurry. Now, obviously, wherever you have a case that the concept is not automatically applicable, in other words, it's too complex to be self evident, then it's not in the category that we're here talking about. That's precisely the type of case where you have to build up a case gradually by the accumulation of evidence, and then that falls in the inductive side. But whenever we're talking about perceptual cognition or such simple conclusions as the automatized conceptual identification of it, and of course this includes direct memory, then it is ridiculous to say in the present context, because that's what you start with. As adults, when we're starting to acquire knowledge, we start with precisely this and then we try to build everything else out of that. We can't start without a vocabulary, we can't start without percepts, we can't start without memory. Now, all those things had a context. You couldn't form concepts without percepts, you couldn't form percepts without sensations. But by the time you reach the adult level, which is where philosophy begins, what you can take for granted as knowledge, as your starting point is what I've just said. Perceptual observations, their counterparts in memory and the automatized conceptual identification of them. That's where we begin. And therefore it would be completely wrong to say about those in the present context of knowledge, this is a, this is a fan. Because that would be as. As though I were to say it's like pneumonia. And this is back to where the concept of certainty comes from. It's only in those cases where certainty is theoretically applicable in its meaningful sense that you can raise the question of context. Therefore, besides what I've just said, all axioms from any science, including all corollaries or deductions from them, you do not say in the present context, an axiom is what a starting point where you begin. So it can't have a context, but that would be something that came before it, that made it possible. So for instance, A is a, is a starting point. Now does it have a context? Obviously, in the sense that you open your eyes, you only see something, and you don't even know it's something till you get to the perceptual level and contrast this thing with some other thing. But nevertheless, once you've done that and you then reach adulthood, you can't start below A is A and give an argument or show where it comes from without using is an axiom. It's your starting point, and therefore starting points. You don't say in the present context because there would be no context without those starting points. They define everything that you're going to learn in that field. Now, this applies to mathematics and philosophy, particularly in mathematics. You never say in the present context. Why? Because you check the validity once and for all. You don't go through possible, probable, uncertain. The theorem either follows or it doesn't follow from the axiom. You do have a context, of course, but it itself reduces eventually entirely to axioms with no other future knowledge relevant. And you can of course, program a computer to perform all these deductions. It needs no judgment, no weighing, none of the inductive necessities. You just have to go through a routine process, and computers can do it a lot faster and better than we can do it. So, in effect, the reason that math escapes the need to say in the present context is that it's a tremendously delimited subject. You say, I'm going to disregard every blessed aspect of the things that I observe, except their quantitative features. I'm going to abstract away from everything variable except how many of them I have, when there's one and another, regardless of what happens to the two or how different they are. And therefore, of course, in a subject so tremendously delimited, you can have eternal, unmodifiable deductions, because it goes right back to the axioms, which, if they're properly chosen, are the starting point and are incontestable. The same applies in a way to philosophy. It is in a certain way, like math, even though philosophy does use induction in a way that mathematics does not. What is true, though, is philosophy is like math in one crucial respect, it is a tremendously delimited subject. It is not a subject like physics that says, I am open to anything about matter, whatever I find, big or small, interesting or boring, it's wide open or like psychology, which says whatever you find, whether it's about some minds or all minds, or this mind or that mind or whatever that's part of this field. Now, those fields have masses of knowledge in a proper society, constantly coming in with no delimitations, and consequently they have to always to specify their context. But in philosophy, we say at the outset we only care about one thing, fundamentals, the eternal universal principles that underlie everything else, and which are the framework for evaluating evidence in every field. Well, obviously that itself can't depend upon an earlier framework. And so on to infinity the principles of epistemology. You cannot say in the present context. You couldn't say, for instance, in the present context, knowledge has to have, has to be logical, because then you're saying, what? In some future context, it could be mystical, or what would you be? How would you know if somebody then comes up to you and says, well, I just got a revelation, the context has changed. And you say, well, there's no logic behind that. He said, yeah, but you said logic was applicable in the present context. I got a revelation. So you obviously, all epistemological principles reaffirm themselves in the attempt to deny them the way as a does. That's the same is true. There's logic and hierarchy, et cetera. So there is an objective test. But we delimit epistemological principles to those that are applicable to all knowledge. Therefore, no future knowledge can possibly amend them because they are the standard for determining whether something is future knowledge. So it's different concretely from math, but it's similar in that the subject is, I almost say artificially delimited, but let us say deliberately delimited. Now, let me say parenthetically, that because epistemology is in this status, so is ethics, because ethics, properly understood, as I hope you get from my book, is a corollary of epistemology. It's epistemology in practice. So it's really an offshoot. And therefore, in essence, all of philosophy is in the same category as mathematics. You do not say in this context of knowledge, although, of course, to stress again, you had to learn the principles of philosophy in a certain order, and they had a preceding context and so on. But you learn all the crucial ones implicitly. And once you reach the adult stage, it's up to a philosopher to make explicit that everything else you know depends on this, and that then becomes the governing context for everything else. So it can't itself have an earlier context. I hope you get this distinction. It would be absurd to say in the present context the senses are invalid. That would mean there's some reason to suspect that they're not. What could the reason be if it isn't based on sensory data, which would wipe out the doubt? And the same for Adler shrugged. To take the example that I gave on the brochure, can I say, Adler shrugged, is the greatest novel ever written or to be written? No, you can't say that even in the present context of knowledge, because you're making a prediction about the unlimited future. And we have already enough knowledge to know that when people build on other people's knowledge, they can do even more surpassing things than the preceding. So Aristotle and Hugo, Ayn Rand would have said made her possible. She could not have written what she did without what she learned from them. I don't know about generations from now someone that will come and build on her and turn out something that we can't even fathom or imagine. At this point, can you say, this is a great novel? Yes, that you can say regardless of context. You don't have to say in this context because there's only two things required to say that. The principles of aesthetics, which are timeless like any philosophic principle, and your knowledge of the book, which is ultimately perceptual. You read it, you know it, and when you relate the book to those principles, you could come up with the assessment, great, regardless of who turns out what in the future. So that is as a philosophic evaluation, that is also, you can't say in this context. The only time context would come up would be if for some reason you were limited in your knowledge of the concrete. That is, in an aesthetic evaluation, you see part of a movie and then you're called out to perform an operation. And it started very intriguingly, you could say something, well, as far as I know, part I saw in this context, it was a good movie, but who knows what came later? You see, but when you have complete knowledge of the concrete and of the timeless aesthetic principles, it becomes ridiculous to say in the present context. Would I say there is no God in the present context? No, because that's inherently a rejection of the axioms of existence, consciousness and identity. And since they're absolutes, his non existence is an absolute. Now, having given you all these cases where you will be happy to hear, you shouldn't say in the present context. I want to remind you most fervently that even if you had to say in the present context, that would not have been the disaster that many of you feel it would have been. Contextual does not mean uncertain. All you do when you say in the present context is in effect give your listener this information. In this case, my conclusion is not self evident or perceptual. It's built on an accumulation of knowledge as follows. That's it. That does not mean the opposite of my conclusion is possible. It does not mean my conclusion is untrustworthy. All it means is you know only so much at any given time. And you have to be able to live with the fact that our knowledge on many issues is limited. You have to be able to acknowledge that it's limited without fearing that this somehow undercuts it or makes it less than certain or opens the door to its overthrow. If the acknowledgement that your knowledge is limited, if that makes you feel your knowledge is shaky, or undercut or open to overthrow, then you are implicitly holding Watt as the cognitive standard of certainty omniscience. And that of course is mysticism and impossible. So when you say context, in the present context it does not amount to well, this is the way it looks so far, but how do I know what's going to happen in the future? It does not amount to that. To state I'm going to repeat this just because one of these formulations will get through. To state in the present context is not to assert that anything else is possible. It is not to say maybe one day I will discover something to upset this. It is only to say the following. Everything known now supports this, but I know there is more to learn and what the more will be, I cannot now even speculate since I don't know it. Therefore I can't even say that maybe it will prove something different. All I can say is there is more to come and I acknowledge this. And I know if my method is right that the more will not contradict what I have so far, the more if you arrived at your contextual knowledge by the right method merely adds a new condition, as in the example I gave in the book about the RH factor. So let me try to hit this one more way. Am I leaving open a chance that I will have to amend my knowledge by saying in the present context, no, I'm not leaving anything open. All I am saying is I don't know everything. Well you say what is the point of making such a statement then if you're not leaving open? The answer is twofold. First of all, it's intellectual honesty. You are acknowledging your non omniscience and that you are using a certain form of validation on this information as against that which is appropriate for axioms and percepts. It's basically you're stating, in other words, this is an inductive conclusion and I'm not omniscient. But that is not the same as saying it's uncertain. And secondly, when you put down the context, that means practically speaking, that if somebody has the proper credentials, you will entertain evidence of an amendment in your conclusion if such evidence exists. This is the big point. If you know A is A, you don't say in the present context, if somebody comes to you and says I have fantastic evidence that nobody discovered before, that A is an A. You just laugh the guy out of court. You have every right to say, I don't even listen to this. I want to hear this guy. I guess you're interested in maniacs or the latest from philosophers, but you know that A is A is the axiom underlying everything, and therefore nobody can argue against it. It. But if somebody comes to you and says, I have a piece of evidence against the theory of relativity, that your initial reaction will not be, this is absurd, because you know that scientific theories are contextual. And then you ask, well, what is, you know, who is this guy, Is he a scientist, etc. That you could think, all right, at least I'll listen to you. That's the practical difference that it makes. It doesn't mean you know less or that you're in any way in doubt. It means you're ident when you say in the present context. It's just you are identifying the status of a certain method of coming to a conclusion and will act accordingly if and when there's reason to do so. Now I can't, I can't keep saying it, but I know that there are people that can't live with this. So I'll try one more last statement. Is all knowledge contextual? Answer briefly like this. All knowledge is relational. All knowledge. Remember Ayn Rand's point that you can't even discriminate a blue sky if you couldn't contrast it with a cloud or some other color? So that's differentiating, which means relating one thing to another. In that sense, there is no item of knowledge on any level, from sensations and axioms on up. It is not relational. But when we say knowledge is contextual, we're using that in a special way as a norm for adult thinking. But to say knowledge is contextual as I use it imposes a responsibility that you have to look at the conditioning background and summarize it and be aware of it in order to proceed. Whereas you don't say A as a, as contextual because that you take for granted as the basis of entering this field. Now, if you get that, you got it, and if you don't, you don't. This point is really either self evident, it's like the percepts. It's self evident or unintelligible. And I just can't. I know there are certain people that you cannot beat into the ground on this, and that's because they will not stifle that other skeptic context that goes back to lecture one. They Keep wanting to equate certain as omniscience. And therefore they want the right to say, I'll never look at another thing about this. This is dogma out of context, which is what a skeptic defines certainty as. And if you know that that's wrong, you must squelch that desire every time it comes up. And the method we used in lecture one, otherwise you will never get this. But once you atrophy that context, once you break the association between certainty and omniscience, this is so self evident that it becomes baffling why you have to go on a half an hour about it. But all of this chewing is necessary because that context of omniscience as the standard is completely automatized in most people's minds. So at least if you don't understand me, I've made you feel guilty about that. Now I have one last and also very important. Topic about certainty, which came as a big surprise to me when I found out that students were routinely drawing a certain conclusion. I mean, really good, bright philosophical students. And I'll tell you what they were doing. They didn't know it entirely, but they were saying, as long as you specify your context, you can never be wrong. In other words, context can work miracles. It eliminates human fallibility. To achieve the state of infallibility about anything, all you have to do is say, in the present context of knowledge, such and such is the case. And then no matter what comes up to review, say, well, in the present context, I didn't have, I didn't know about this, so therefore it was true. Now this is utterly wrong. Utterly. Context is not an easy back door into dogmatism. You can't get away with murder by simply uttering in the present context and then say anything. Well, that's the way it is. Let me write down, I ask you to write down this sentence. Context does not eliminate the possibility of error. Context does not eliminate the possibility of error. No philosophy is going to make you infallible. You can follow the method, the logical, objective method of thought to the utmost that is volitionally possible to you. You can specify your conclusion contextually and still be wrong. The most you could say in such a case is you have no right to say I am wrong or I'm in error, or even maybe I'm in error in that context. In that context you have to say, I have reached the truth. But at a later time, if you come to grasp that you were in error back then, an error you simply didn't catch, it was still an error. So when you look at it, you say I made a mistake there. I was wrong to think such and such, even though at that time I was right to think it. And the epistemological situation in such a case is you were wrong. In fact. Fact. But you were fully rational in holding that you were right. Now that is not a contradiction, it's a common occurrence. A common occurrence to be wrong and yet be fully rational in holding that you are right. Now, let's give you the simplest example. You balance your checkbook, you get the statement from the bank, all the entries you check off against your checkbook seem fine. You go over the addition, even with a calculator, you're focused and conscientious within reason. Let's say you don't spend more than three hours over it. And then you think to yourself, absolutely, justifiably, I'm right. This is how much I've got. Then you get the next statement from the bank and it doesn't reconcile what the hell is going on here. And you go over and you realize you inadvertently copied one figure wrong from one column to another. Now, does that mean you were non objective? Not at all. Objectivity is not infallibility, otherwise there will be no use having such a concept. Objectivity is a guide for a volitional consciousness and therefore you can be completely objective and wrong because you tried to apply the method, but you simply didn't do it correctly. Now here is a case in the checkbook case where you were objective, you did what you could grasp to be possible. You observed, you checked, you used logic, you verified so far as you could grasp a procedure to be necessary. But lapses or momentary inadvertence is possible, however focused and reality oriented you are. So the right analysis is metaphysically possible. So the right analysis is when you first balance the check, you say this is right, which you were justified to do at that point. But then you come to discover the discrepancy of the figures and you see it was mistaken. In other words, even given the data available, you erred in interpreting it owing to your failure in method, even though you were trying your best and no one can blame you. Now you have to clearly distinguish in your mind two different types of situations. Two different types. One is where the issue of context is applicable, where your knowledge is limited, you commit no error. But later, knowledge teaches you a qualification and never contradicts your old that applies whenever your knowledge was really knowledge. In other words, when you have correctly applied the method of objectivity, then you can Say new knowledge will never overthrow this. But if you err in your method, even if you are doing your utmost to be reality focused, the new knowledge will contradict your first conclusion. If I inadvertently neglect some figures in my checkbook not through evasion, but just through inattention when I had no reason to suspect or I mis added the column of figures and I ended up with an improper balance, this is false even in my context of knowledge. It's a misinterpretation of data available to me through the use of an improper method. Now, no amount of context specifying will annul this possibility. You can't say, well, X was the sum in my context at the time, because context is an objective fact. It denotes all the actual evidence available and the evidence was available when you balanced your check to correct. It's not that you had limited data. It's not that there were a lot of things about your deposits or other people's deposits or the safety of the bank that you didn't know. You had every blessed piece of evidence you needed to come to the right conclusion and you made a mistake, well then in that case, the problem is not that you have limited data, but that you used the wrong method to process the data. The wrong method. In other words, there was a lapse of logical rigor, even though that may be entirely not your fault and not blameworthy. Well, in a case like that, an error is an error and you can't salvage it by saying in my context, whereas the context type of case, as in the example from my book, the A type bloods are compatible, there was no false processing, no lapse in logic and no error. A true conclusion was gained, real knowledge was gained, but it depends on conditions not fully known. Here is a case where you used logic reach the truth, but you don't know everything. So you have to say in my present context of knowledge, and that means given the facts now available, this is the proper conclusion to draw the actual logical conclusion from these facts. In other words, to put it another way, your context, when you use that legitimately, cannot include your mistakes, only your knowledge. So this is a very important distinction to make. It's one thing to amplify your knowledge. It's a different thing to correct your mistakes. It's one thing to have the gist of a truth, but not yet all its qualifications. It's another thing to be flat out wrong to make a mistake. The only time we're entitled to say that we're wrong is if the mistake method of processing the available data is Erroneous. But if the method is valid, then you can't call it an error because you did come to what was actually a fact and no further knowledge will upset it. Therefore, the test of whether something is an error or not is the quality or nature of the method used to reach it. Did you commit a logical error even within your context, some identifiable fallacy, non sequitur, contradiction? Or did you follow logic impeccably except that you are non omniscient? That's the distinction. And maybe this way of looking at it will be helpful. There are two so called problems of knowledge here. One is fallibility and one is non omniscience. Those are not the same thing, because animals are non omniscient, but are infallible in the sense that as far as they go, being only on the perceptual level, they're right. A man has two ways, two issues here. He can err even within the material available to him. He can err in interpreting it through lapses of method. And on top of that, he only knows so much. There's a whole world still outside his ken. The first problem is taken care of by logic, the second by specification of context. You get that? All right. Now let me give you one example and you tell me in conclusion, this was a real example given to me in the seminar. Is this a case of error on the part of the person or a case of limited contextual null? Now, these examples can get really excruciating, but I took one that's not really excruciating. This is the example given me by a student. You raise your hand when you decide what it was. The child. Child is raised in an environment which consists entirely of good people. He is told that man is made in God's image, that God is good, and that's why men are. And he believes that all men are good. Later he comes across evil men and revises his conclusion. Was his first conclusion all men are good an error or a contextually delimited truth? Now you think about that for a minute. Was his conclusion that all men are good an error, perhaps understandable error, or a contextually delimited truth which later knowledge does not overthrow? Now let me see by a show of hands, how many think it's a contextually delimited truth? Wow. And how many think it's an error? Well, group two is correct. Now, does somebody want to tell me why? Now, you see, you have in mind a pejorative use of error. You have in mind error, as he did something bad and here's this poor innocent kid. What could he do about it? If he's got, you know, situation like that, he's more or less doomed to come to the wrong conclusion. Okay, even so, that is not relevant to is it wrong? Is it true in any context that all men are good? Is that like under some conditions all a bloods are compatible? No, it's not. A man is volitional, therefore no universal moral assessment is possible. And therefore any universal assessment is necessarily mistaken. Now what? What? So when you say someone committed an error, that is not the same as saying you bad person, What. What fallacies did the child commit? Very understandably. At least two fallacies that are identified in elementary logic classes. Hasty generalizing is one. And second, argumentum ad verecundiam appeal to authority. In other words, parents and people he respect told him this and it fit with, you know, his hasty generalizing. Therefore he couldn't imagine that his parents are deceiving him. So he's not blameworthy. But he did not reach the truth. He committed a logical error even within his context. What would correct him is not new information, because there are people who know all about the whole world and still insist that all men are good for the same reasoning. What would correct him would not be new information or new qualifications, but understanding a method of thought. That means his problem lay in his method, not in the delimitation of his knowledge. I'd go so far as to say his error may even be unavoidable. It may be the case that he had no option in that environment, circumstances, but to come to that error. But then I would hasten to say it's unavoidable to a child. Why? What does he lack that makes him helpless in the face of all these potential errors? Epistemology, methodology, logic, a conscious knowledge of how to come to conclusions and when a conclusion is valid and when it isn't. Until a child or a civilization reaches epistemology, they have no way to distinguish valid from invalid conclusions. And it's luck. And in this situation, errors are inevitable and you are helpless. But there's still errors. Now, is there any condition in which errors are avoidable? I would say yes. Errors are avoidable only to a man with access to the proper methodology. Epistemology, once you become sophisticated, logically speaking, conscious of the issue of method and the proper rules of thought. Once you do that, I say no. Error as such is inevitable, however twisted the evidence, however limited the evidence, however warped a picture you're getting. Because as a being who has free will and is methodologically conscious, you always have the option to refrain from coming to a conclusion to say this is not yet proven. There are other possibilities. So I would summarize it like this. To a methodologically conscious adult, no error as such is inevitable, although many errors are possible, understandable, even honest and moral. But that's not the same as inevitable, unavoidable. Now the reason I stress this particular point is because it has a lot of importance for the issue of the validation of free will. And the failure to grasp this confuses a lot of people on the topic of free will. So that's. I just want to make this final observation. In conclusion, if you were to say, as an advocate of free will, that in some cases error is inevitable to a methodologically conscious adult, you would then be back in the position of the determinist. Namely, how can you rely on your mind? How do you know you haven't been so programmed as to swallow an error without your choice or participation? That's what we tell a determinist. You can't trust your mind because we're not automatically correct. Therefore, inevitably, sometimes you've got to be wrong, and therefore how do you know? Now the only consistent way to carry that out is to say that if you have free will and you use it to apply to a self conscious method, then no error can be built in in such a way that you just couldn't avoid making it. Savages, primitives, children, anybody who doesn't have the essentials of a valid method. And by the essentials of a valid method, I don't mean a great deal more than what we're saying in this course. Context and hierarchy are probably the two most crucial. If you have that you have a chance at every moment, you are never so programmed or warped by an outside confusion that you have to be wrong. You can of course be wrong. Maybe it's very likely you're going to be wrong, but you can't say it's inevitable the way it is with the determinism. So we are not in the position of the determinist. We can be right, and that's why we can assert certainty even though our knowledge is limited. But it's not therefore that we are infallible. We can be right and we can be wrong. And you have to know the conditions under which each. Well, I hope you get the idea that context is real, but not miraculous. And on this point I'll close for today. Thank you. Now I'll take questions. Yes. Regarding standards of proof. First of all, Is a standard of proof an item of knowledge? And if so, how can you be certain of it? Well, that's the whole question of induction, really. The question is, is a standard of proof an item of knowledge? And if so, how can you be certain of it? The whole question of the problem of induction is really, how do you arrive at the standard of proof? How do you know what criteria has to be fulfilled in order for an induction or a scientific theory or generalization to be justified? Now, in the book, I carefully took a situation which was widely acknowledged, namely motive, opportunity and means. Thanks to murder mystery writers, we had covered all the bases there. And you could see in logic, you know, that a person would have to have a motive, he'd have to have the chance, and if there was some technical means, he had to have access to it. And if you had all those three and nobody else did it, it had to be him that became, you know, it was a simple, straightforward case. Now, my theory of induction is this. The problem of induction will be solved when you can say about any scientific theory the equivalent of when you know motive, opportunity and means. It's true. But obviously it's going to have to be much more abstract than that because it's going to have to cover all different complexities of scientific theory. And that's where you're going to need a combination of a scientist and an epistemologist to do it. An epistemologist could give you broad criteria, as I have in this course, when it integrates with everything, you know, when you observe the hierarchical structure, you know, when it's both reducible and horizontally integrable, et cetera, et cetera, when it's defined in terms of essentials, all that. But there are certainly many cases, actual and possible, of different scientific theories which at a given point all obey, as far as we can see, all of these abstract possibilities. And we need to have some procedure to tell us if you were in that field, if you are a scientist, where to look to make the decision, how to choose among them, how to distinguish. And they really don't have that. Now. The only thing they have right now as a method is the hypothetical deductive method, which is that if this theory is true, it predicts such and such in the future, then the prediction comes true. And then you therefore say, that verifies the theory. But of course, that doesn't tell you that only this theory predicts that particular event. So it's always leaves open the question, are there other possible explanations? Now, of course, you have to remember that you can't assert possibility arbitrarily, but by the same reason, you can't assert your theory arbitrarily. So there's cases of perfectly legitimate disagreement in complex scientific fields precisely because no one has yet solved this problem on the broadest level. See, it's the case where what we need desperately is the intermediate knowledge. We know the broad philosophic principles that a scientific theory must live up to. But we. And we know in many daily issues, such as a murderer, what specific things will be proof. But on the intermediate level of breadth, where you're making generalizations about atoms or not everything, and not just, you know, simple, straightforward things in front of your eyes, that is really the essence of the problem of induction, how to formulate the standard of proof. And I don't purport to have done it. And I answer all questions insofar as they impinge on philosophy by saying, any inductive generalization that I use, I apply the tests of taking instances in every possible category that I can think of and showing that it's applicable, integrating it to everything I can about the nature of the entity and all other entities. That is, integrating it reductively and horizontally, and then saying, if it truly is inductive within the present context, and that's it. I will not, you know, make a dogmatic absolute. But that doesn't mean that I am uncertain, for instance, that concepts are identifications with their measurements omitted. Now, Ms. Rance describes this as an inductive generalization. She has obviously not observed every concept, so how does she say that this is certainty? Now, remember here the great thing about philosophy, it applies no matter what your stage of knowledge. That's the whole difference between philosophy and science. Sciences keep evolving because they deal with concrete, so they're continuously adding their knowledge. But philosophy deals only with the mechanism of the mind, no matter what error or no matter how it functions. So she simply has to say, from the time of Aristotle or historical record to the present, let's take concepts of nouns, let's take concepts of verbs, adverbs, every part of speech you want, concepts of measurements, every type that you could conceivably think. And that's exactly what I did with her when I was trying to learn it. Concepts of substances like water had me completely baffled, how it was measurement omission. We spent maybe two weeks, and I wrote a whole long paper on why water as a concept is formed by measurement omission, how that differed from ice, and so on and so on. And then she said, in every single case I see that this is what happened. I integrate this to mathematics and I see that it's the same principle. I see that the essence of this is a type of integration. And that's what all consciousness is. I've connected it to every concrete and every abstraction within the sum of knowledge. I've answered all the traditional problems. I'm not omniscient. Then I said to well, what if something comes up tomorrow that you reach by something other than measurement omission? And she said, what? I said, I haven't the faintest idea. She said, why are you thinking of it then? If it comes up by something else? You know, it might be like a clover that grows out of the ground. It's not a concept. It comes up from the roots, from the soil. But she said, you can't hypothesize what if everything I know is wrong? Because then you have no basis on which even to think. So I am in the enviable position of being able to say at once I would like to see a solution to the problem of induction for scientists, while feeling that for philosophy it is not necessary for us to be able to declare certainty about our inductive principles. And that is a position which would certainly not go over well with any person that has tenured position in this country. But nevertheless, that is my view, yes. Would you have to approach the problem of induction differently in the natural sciences versus the social sciences? I don't want to comment on that because that is really too technical. How would I differentiate induction in the natural and the social sciences? I don't even want to talk about induction anymore because I've already said more than I know about it. I like something about certainty. You still are worried about the airplane, Christina? Yeah, go ahead. No, it's a question about the. Clarify the statistics. Yes. If after you reach a certainty that, for example, in the airplane state somebody comes up and says that the statistics that they have taken refer they've been taken with all those data that you have reached. So these are not statistics like about all the planes, but only of planes that been checked, well, then it's impossible. My question does this then reflect just a metaphysical possibility? Well, no. This is an impossible situation. If we have covered in our knowledge all of the conditions and this plane comes under it, that statistics have to be 100 out of 100 get there safely. It would be impossible to say one in a thousand crash under conditions in which none crash. So I mean, what you're trying to say is informed statistics is a self contradiction. Statistics is Inherently a method of ignorance. As soon as you start knowing something about the concretes, your statistics qua statistics go in the background. It's only qua ignorant that you can apply. So if you take the sum context of your knowledge, there's no room for statistics. Yes, a red short sleeved shirt. Harry, I noticed in your discussion of statistics you said there's about a con three. Yeah. Is this why actuarial tables can be relied upon by insurance? Yes, absolutely. Very good point. He says I said statistics are inapplicable or applicable only when you're ignorant of a concrete, but they are applicable in many other cases where you're not trying to make statements about concretes, but only about broad groups. Just as I said, for instance, I can legitimately say in ignorance one out of 100 is going to be a libertarian because I don't know anything. In that same way Harry points out, actuaries who make predictions can say five out of a thousand are going to die before 60 from having an angiogram taken or whatever the statistics happen to be. And they know that simply because that's happened time and again. And that's. They don't know why. They certainly don't know that it's you. They couldn't even remotely guess. All I can say about you is if you're just like all the rest, but if you've already had a heart transplant or a heart attack or whatever the change, the chances are completely wiped out that you predicted in your empty statistical prediction. For the same reason, I hesitate to say this, statistics are sometimes helpful in the early stages of a science. For instance, if you have a science like psychology, which doesn't exist, as in the stage around Thales that Thales was in, in relation to philosophy, they have a few tentative feelers and some mystical theories and they don't know what the hell is coming on, it's okay to do statistical studies and say 93.8%, the people we interviewed feel better when they think well of themselves than when they hate themselves. Now, you know, you get that all the time and it's obviously not an accident. And if you had a decent basic understanding of human nature, even the kind that philosophy gives you, you would know that there's an underlying causal relationship there and that they better investigate the other 3% to see that they've come to the correct conclusion, because there's a necessary tie between self esteem and happiness. But if you had no idea of that and you saw this vast predominance, you say, well, this is something worth investigating. I wonder if there's something to it. If there's something to it. Now remember, there's lots of statistical correlations which hold for long periods of time which are utterly without meaning. They actually did a study once of the death rate in a certain province in India and the number of semicolons on the New York Times editorial page. And they went up and down together. The more semicolons, the more deaths, the less, the less. There was, within certain limits, a direct concomitant variation. But they just dismiss it as this is ludicrous, it's a pure coincidence, because there's no conceivable logic unless the semicolons are poisoned and the paper is red, you know, by the Indians who died. But leaving aside some mechanism like that is just a coincidence. So you cannot say when you find 100% correlation, this is knowledge. If all you have is statistics, you do not have knowledge. If all you have is statistics, however terrific the correlation, the most you have is a potential avenue to investigate a possibility that's worth running down lead that you may as well follow in the absence of anything else. But your ideal should be to acquire knowledge by some proper methodology and hierarchy so that you're not left at the mercy of juggling statistics. And if your whole science consists of nothing but collecting and reporting statistics, which is what it basically is today, then you have no science. And that's why anybody can believe anything and can prove anything by the latest study, because none of it is connected to anything and nobody can explain why for anything. And everybody's figures conflict with everybody else's figures. And any scientist therefore can prove that the earth is going to turn hot tomorrow or cold tomorrow, you know, or the fish are going to fry or they're going to chill just because they put together two things which correlate this way versus that way. And then they make arbitrary computer models and they say, well, this was the correlation. All that is bs, that none of that is knowledge, let alone science, which is supposed to be a highly systematic, rational form of knowledge. Well, I think I've made my animus against statistics clear, but that does not to say that there aren't uses for it. Yes, you said that certainty is a subcategory of nullity, right? What would be an example of knowledge that was uncertain? No, no, no, I said certainty is a subcategory of knowledge. What would be example of knowledge that is uncertain? I don't say all knowledge is certain or uncertain. I say there's category of knowledge to which that dichotomy doesn't apply like perception. It is. I add nothing to that by saying I'm certain of it because I open my eyes and see it. If that's all there was and everything was like that, you would just simply say, I know it, that's it. You wouldn't have such an idea as I know it for certain as against, for probable. That whole topic only comes up because there are complex inductive conclusions. And so the section of knowledge to which certainty or uncertainty applies is in the complex inductive aspects. The rest is neither certain or uncertain. Now, if you want to, as I said, you can use the term certain in the broadest sense as a polemical term. To counter skeptics say not open to doubt. But in a proper world you would not use it that way because no one would think that it was open to doubt. You know your perceptual evidence and therefore you wouldn't have a term to stand for the absence of something that no one would suspect. So in that broadest sense, the use of the term certain is a cultural, not a philosophical use. The philosophic use applies only as I've indicated, and, Well, give me the choices because I don't know what's unclear. How about differentiating between that and data? Well, data and statistics. Data is much more abstract, everything. Data means the given. That's just Latin, plural for given. And therefore perceptual data are given to them. Nothing to do with statistics. The information previously known and validated is the given for the future inquiry. It's the context. Statistics is a very, very specific kind of data. Numerical correlations with no understanding of. Correl correlations or discorrelations. One in a thousand cases, one in a million, 910, 3.8, etc. It's the attempt to associate events or situations causally simply by their numerical preponderance or by the numerical pattern of their coexistence. And that is much more specific than. I mean, I can't see how you can even relate the two to each other. One is you have to perceive and the other is you have to think in the wrong way. Who else has a question? Yes, I have a question about what is the context of something like a mirage, but it's not actually there. Yes, it is actually there. He's asking the question, what about a mirage? That you see it and it looks real and so on, but it's not actually there. Well, what do you mean by the context? Your intellectual or the physical situation? Intellectually, the only context for a Mirage, since it's a perceptual phenomenon, is that you are there with your senses and when you see it, you are not in any way distorting it's dislike perception of any other kind. But it happens to have one tricky aspect. It's the light waves from some real object or place transferred so far from the place that you see it where it isn't. But what you're still reacting to, if it's a mirage and not a hallucination, is an actual physical fact which you are perceiving. You simply have to learn to interpret it. It's exactly like the figure looking bent in water. So there's a physical basis and all the it has no more context than any other perception. But you need a context to interpret it, because it happens to be a perception which if you're unwary, you'll automatically think is a certain way. Once you come to know this type of case and you're on the desert, you say to yourself, is that a mirage or not? You see? So there's no problem with that? Yes. Is horizontal integration sometimes necessary to establish certainty? And if so, does this necessitate defining the concept of proof broadly enough to include horizontal integration? Yes, and yes, horizontal integration is always necessary to achieve certainty, because in those areas where certainty is involved, integration across time is the crucial element. And if you don't integrate to everything else you know, then you have no basis, you know, to reach the conclusion. And the second one was, what does it necessitate defining the concept of proof broadly enough to include horizontal? Absolutely. Proof absolutely has to include as a broad concept, if you're not using it simply as validation. Now, validation comes in different forms. Validation I use as the broadest possible term to mean any form of relating a content to reality, any form of establishing your correspondence to reality. That might be self evidence, that might be deductive proof, which is where the word proof most often is used. Or it might be inductive. Now you could call that proof if you want, or call it just induction, but whatever your name for that category of reasoning, horizontal integration is an essential of cognition. It's an essential of proof. Yes, I asked the question because in the book you brought up the term. You describe proof as a type of reduction. Yeah, Distinguished reduction from horizontal integration. Right. So I saw reduction as well. I see what you're saying. I think that that's perhaps a legitimate criticism, although I was trying to take into account that both are essential. So I'm making only this distinction. You could justify my wording this way. Proof, in the sense of deriving a conclusion from reality, is the process of reduction. If you truly do reduce it, you have proved it. But the trick is, so you can say proof is only reduction, but the trick is you must validate that you truly have reduced it. And the way to do that is by the horizontal integration. And consequently, the real justification of saying proof is the complete integration. But I was there trying to analyze and then later put it together. But if you want to say, you know, you could rewrite it as saying proof requires both reduction. And I said the same thing, only I called it logic. I said the use of logic requires both. But I wouldn't fight over when we apply the word proof, as long as you understand you have to do both. You see, the reason I did that is because in mathematics, they can have a proof, you see, without the horizontal integration. If they really are, you know, you can check it by computer, you see? Yes. Thought it might lead to concepts that confusions like certain and fully certain people making a distinguish. No, there's no distinction between certain and fully certain, any more than pregnant or fully pregnant, certain as you have it or you don't. Any adjective, any adverb in front of it is already a mistake. I mean, unless you're just talking in common sense, you know, you say I'm 100% certain, you just want to beat the other person with the idea you are certain. But it's easier instead of yelling R to give three or four adjectives. But in actual fact, any modifier of certain that would intend to heighten it is a redundancy, because if there was a certainty that wasn't fully certain, it would be a probability. Well, we have about one minute, Diane, I'll give you the last one. Thank you. This question is about your example of the child and accepting the premise that all men are good and that it was actually there, although it wasn't blameworthy, and the discussion of the you've got to have the method of logic first. Then that made me question an adult who doesn't have a proper method of logic. Is he blameworthy because he doesn't have the proper method? Well, that's a. Is an adult who doesn't have the proper method of knowledge innocent like the child, even though erroneous? Or is he blameworthy because he doesn't have the proper method? Now, that's a moral question. That is not a question of epistemology or certainty. Obviously, the adult in your case, as you stated, has already made an error through a bad method. So there's no question pertaining to our lecture. But you want to just know, out of curiosity, how do you morally appraise adults who do not know the right method and therefore continuously make mistakes? And the only answer to that is it depends on their access to the right method. If they are schooled enough or read enough or are brilliant enough to come up with the idea that some kind of method is necessary and these are its essential steps, then of course they're responsible. But in the vast majority of cases, to cite statistics, in 999 out of a thousand, a person can't do it because it takes a genius and it takes a tremendous amount of work. And therefore most people, for most of their mistakes, especially as they get into the broader, more abstract realms, are not morally blameworthy because they don't know what the hell they're doing. The only morally brainworthy part is that they posture as though they do. No, but if they were to say, in fact, I don't know why my mind comes up with the things it does and therefore I'm not sure, that would be terrific, because then they would need to come up with the next question. Well, let's find out how to use our minds, you see. So the dishonesty is not in lacking the method, but in just turning off the whole subject and going with what you've got, even though, you know it's no good. Now, even that I don't blame most people for. Because they don't really know what they're lacking, and they certainly find the opposite of it if they go to college in courses that are supposed to offer you logic or methodology. So most people are really in a bind and they could be more honest than they are, but they couldn't be substantially ahead of where they are at this stage of philosophic development. That's basically what I would say. I think that's woozy enough to cover all cases, thank you very much.