Here’s the principle. For any proposition, P: If

1. Taking a certain cognitive stance toward P (for example, believing it, rejecting it, or withholding judgment) would require rejecting or doubting a vast number of your current beliefs.
2. You have no independent positive reason to reject or doubt all those other beliefs, and
3. You have no compelling reason to take up that cognitive stance towards P,

then it is more rational for you not to take that cognitive stance toward P.

“Your current beliefs,” Morris goes on, “are like a raft or boat on which you are floating, sailing across the seas of life. You need to make repairs and additions during your voyage. But it can never be rational to destroy the boat totally while out on an open sea, hoping somehow to be able to rebuild it from scratch, or else to swim without it.” (72ff)

The principle passes its own test and is elevated into a basic belief.

I think this opinion is similar to what Victor Reppert has proposed, namely that one should keep believing what one already believes, unless one encounters a good reason to believe otherwise.

Just as it is obvious that pain one feels from being too close to the fire is not actually in the fire, neither is warmth, which is merely a sensation generated by a certain power of the fire. A warm hand put in lukewarm water feels cold, while a cool hand put into the same water, warm. This implies that if heat as it is experienced were really in the water, the water would have to be both cold and warm at the same time, which is impossible. Heat then is relative to the perceiving mind; the same thing can seem warm to one person and cold to another. So, the secondary quality of temperature has to do with the quickness of movement of the molecules in a substance, and it is capable of producing different sensations in different observers.

The operation of the physical causes of secondary qualities is mysterious, at least to the science of Locke’s day; these operations “are hid from us, in some things by being too remote, and in others by being too minute.” (IV, iii, 24) Locke thinks that we can never know the underlying arrangement of corpuscles that cause our ideas of secondary qualities. This is because all of our ideas come from either experience and reflection, and the insensible parts are simply too tiny to see. (Reflection does not apply here at all.) We cannot predict what effects the minute parts of an object will have on us; nor can we, having seen the effect, know how it was produced. (IV, iii, 25) Finally, we cannot know how the structure of our immaterial minds permits us to experience ideas caused in us by material corpuscles. (IV, iii, 28)

It follows that only the ideas of primary qualities really exist. (II, viii, 17) They are “real” in the sense that they persist in the object even if there is no one around to observe it. But the ideas of secondary qualities vanish as soon as the observer, so to speak, leaves the room. These qualities are reduced back to their causes: an array of corpuscles making up the object with the power somehow to produce ideas in us yet which do not in any way resemble these ideas.

Now Mises writes: “Acting requires and presupposes the category of causality. Only a man who sees the world in the light of causality is fitted to act. In this sense we may say that causality is a category of action. The category means and ends presupposes the category cause and effect. In a world without causality and regularity of phenomena there would be no field for human reasoning and human action. Such a world would be a chaos in which man would be at a loss to find any orientation and guidance. Man is not even capable of imagining the conditions of such a chaotic universe.” (Human Action, 22) Thus, knowledge of causes and effects is required for successful improving of one’s lot.

So, animals act as if they knew causes and effects. My cat jumps down to the floor from the refrigerator most carefully, clearly aware of the fact that not every way of landing will be painless. As such, he connects means and ends; he succeeds in attaining his ends through employment of appropriate means. But this is a far cry from saying that cats know things; rather their instincts and conditioned reflexes provide for their needs. To put it another way, action does not require knowledge, only “animal faith.” Though the cat is aware of the causal relationship between food and happiness or between my scooping the dry food with the dish and the food’s appearance in its regular place, that awareness is never expressed nor reflected on nor is the truth used enjoyed, showing that the intellect (which I argued is the conjunction of the three intellectual virtues, one of which is called “knowledge” or “science”) is, indeed, a uniquely human characteristic.

Now animals still interpret or recognize the meaning of things. Thus, smelling food causes my cat to interpret it instinctively as something “good,” something to be loved, something which brings enjoyment. I have written before that interpretation ends with one’s happiness as the ultimate given. But interpreting things is not the same as knowing truths. I say that animals have the former ability though not the latter.

(3) If p were not true, S would not believe that p.
(4) If p were true, then S would believe that p. (Philosophical Explanations, 172ff)

(3) implies that S’s evidence must not lead him to believe that p regardless of whether p is true or false. In the unhappy case in which it does so lead him, since one can’t in principle have good evidence for false beliefs, S’s evidence for p is not, in fact, good, even if the belief happens to be true. Hence S’s belief is unjustified. In other words, the belief is not “sensitive” to truth; the truth of p does not cause S to believe that p, through whatever means or intermediaries.

(4) is relative to a method of knowing. For example, I have a JTB that there is a pen on the desk because I see it. But not in all cases where the pen is on the desk would I know this, because I might not be looking at it or might overlook it, etc. Yet we would still like to say that I know that the pen is on the desk. So, we must fix the method of knowing, e.g., if p were true, then if learned that p via method M (such as seeing the pen), I would infallibly come to believe that p. At the same time Nozick says that for fundamental beliefs like “the world has existed for many years already,” no particular method needs to be specified: “[s]o nested are these statements in our other beliefs and activities, and so do they nest them, that our belief or acceptance of them is… best represented apart from any particular methods.” (185)

Example 1. There are two methods of knowing whether my friend is alive: (M₁) knowing he was alive yesterday and continuing to hold this (highly probable) belief; (M₂) seeing him alive right now. If my friend were dead, then I could not know anything from M₂ and would believe a falsehood of his being alive through M₁. But M₂ outweighs M₁, such that if I were to see him dead, I’d believe that he was dead, despite knowing that he was alive yesterday. And, of course, if I were to see him alive, I’d believe that he was alive. So, condition (3) is satisfied through M₂: the evidence of seeing is more convincing than the evidence of inference from a trend (it’s a good bet under normal circumstances that a person will stay alive tomorrow if he is alive today), and (3) includes reference to this more powerful method of knowing.

Example 2. There are two methods by which a father believes his son to be innocent of a crime: (a) he has faith in him and his righteousness; (b) his trial has acquitted him. Suppose that the trial has instead correctly proved his son to be guilty, would the father still cling to his now false belief? If so, then he does not have knowledge that his son is innocent, despite the actual fact of (a) and (b), because condition (3) is unsatisfied: p is false in some close possible world, yet the father nonetheless believes that p. (a) outweighs (b) for the father, yet we presume incorrectly. Certainly, the outweighing may be correct, such that the father’s faith is true despite the trial’s verdict, in which case the father would have the knowledge of his son’s merits or demerits.

Example 3. Again, we have two pieces of evidence that a certain building is a concert hall: (1) S has been there and seen the plays; (2) S has read government reports that it is a concert hall despite the fact that it really is a nuclear weapon building factory. (No one can approach the alleged hall because of the radiation which affects a person such that he loses any interest in buying a ticket.) Condition (3) seems to be violated, because in the second case the building is not a concert hall, yet S holds the false belief that it is. Must we deny the relevant knowledge to S? Nozick writes that not all the methods of coming to know something need satisfy conditions 1-4, only the weightiest one. It seems that determining which method outweighs all others must be done on a case by case basis with the help of practical wisdom.

Thus, (3) says that S must possess truth-sensitive evidence for p, such that the falsehood of p would invalidate the evidence. The evidence must “track” p‘s truth value. In other words, the evidence must be seen to immediately cease to be good if p is imagined to be false (because, for example, it would be different due to a different method of gathering it). (4) says that good evidence must be available in any situation in which p is true. Nozick’s criteria for knowledge appear to mirror my own (independent) ideas, namely, that there can exist no justificatory evidence for a false belief and that there must exist (though obviously not such as to be always discoverable) evidence for any true belief.

For example, the psychologists must know in great detail how to build and operate their Matrix-like machines. In other words, they must know the unvarnished genuine causes-and-effects if their wards are to feel comfortable floating in their tanks. And if their technological knowledge is, too, arbitrarily simulated by their own, still more clever, deceivers, then those deceivers must ultimately know what is and what is not real. The buck has to stop somewhere.

Thus, the skeptical hypothesis that because we don’t even know that we are not brains-in-vats, we have no knowledge at all, fails on its own terms. There is no reason why “we” must exclude the psychologists or Matrix robots experimenting on us. “They” then are part of “us,” creatures with intelligence and will, and so they have knowledge even if we don’t. An illusion must take place within some reality; there cannot be only illusion without some known reality on which it builds.

Update. The question still remains whether you personally are or are not deceived. We can pursue this problem in two ways. As per my previous idea, there is no such thing as good evidence for false beliefs. It’s time to strengthen that claim by saying that there must be very good evidence for every truth; hence there must be such evidence that you are not a brain-in-a-vat. If it has not been found (although solutions have been proposed), then just keep looking. Second, the hypothesis that you are a BIV can be evaluated on its intrinsic merits, such as simplicity and scope. And surely, it is much simpler to suppose that you know the realities of your daily life than that you are fed false information on purpose by a team of godlike in their powers psychologists. The scope of the BIV hypothesis is enormous compared to the non-BIV hypothesis. The latter simply states that you know what is happening around you. The former, on the other hand, postulates an entire world created for the purpose of tricking you.

Update 2. There remains the possibility that the evil psychologists put you in a tank and then died, leaving the system work on its own. In that case apparently no one will know what’s going on. Fortunately, this is at worst a temporary predicament, because entropy will eventually cause the machines sustaining you in the tank to break down. Only living beings can overcome entropy. Eventually you will either die or wake up from the Matrix-like illusion.

You have no means other than your justification… of determining whether the belief is true; and because the justification is fallible, it gives you no guarantee of the belief’s being true (and thereby of being knowledge). Moreover, if fallibilism is true, then any justification which you might have, no matter how extensive or detailed it is, would not save you from that plight. Thus (given fallibilism), you are trapped in the situation of being able to reach, at best, the following conclusion: “Because my evidence provides fallible justification for my belief, the belief is fallible knowledge if it is true.” At which point, most probably, you will wonder, “Is it true? That’s what I still don’t know. (I have no other way of knowing it to be true.)”

Does it make sense to ask, “Do we know that X is true?”? Let’s go back to the definition. Let X be “Socrates is sitting.” To know that X is true, (1) you must believe that X is true. But to believe that X is true is simply to believe that X, in our case, to believe that Socrates is sitting. (2) “X is true” must be true. But this is so if and only if X is true or X. So, “‘Socrates is sitting’ is true” iff Socrates is sitting. So, Socrates must in fact be sitting if the belief is to be true. (3) You must be justified in believing that X is true. But the only way to be justified that “Socrates is sitting” is true is to gather up evidence that Socrates is indeed sitting. It is clear that to know that X is true is fully equivalent to knowing whatever is asserted by X. So, the question “Do we know that X is true?” is the same question as “Do we know that X?” We cannot know that X in true any more than we can know that X. Thus, because of this equivalence the author of the article is correct in adding that “And so — right there and then — you are denying that your belief is knowledge, because you are denying that you know it to be true.”

On the other hand, we can suppose that a belief is true and ask whether then fallible justification is enough to make your (true) belief knowledge:

To put that observation more simply, this epistemological question asks whether a belief which is fallibly justified, and which is true, is (fallible) knowledge. This is the question of whether your belief is knowledge, given (even if only for argument’s sake) that it is true.

In the first case the truth value of a belief is not independent of its justification; in the second case it is an independent variable. But then why not call knowledge simply a justified belief if the truth of a belief is judged according to the strength of the justification for it? At least in theory we can never “know that a belief is true.” All we can do is be well-justified in believing something and trust that things will turn out OK.

In the previous post I suggested that no false belief can be well-justified. But true beliefs can be poorly justified. In both cases our claim to knowledge will be invalid. In other words, in theory, no matter how confident you feel about some fact, whether the fact holds or not is entirely accidental to your justification for it.

Can you “know that you know X”? I think this phrase asks whether you can be confident that you know X. This is exactly like asking whether you are confident that X, because if you have a JTB that X, then it’s easy enough to have a JTB that you have a JTB that X. All you have to do is scan your own mind and determine your attitude toward X. That knowledge seems to be incorrigible.

But if the truth of any belief, at least as far as we can tell, is a function of the quality of our justification for this belief, is truth forever out of our grasp? Must the only criterion of truth be human happiness? If things work as expected, then we have a reason to claim knowledge. If they don’t, then our belief was false, in spite of how carefully we worked on justifying it. This pragmatic theory of knowledge seems the only one available to salvage the concept of truth. But then what of things like metaphysics? Does it “work”? Or is some knowledge justified infallibly?

Suppose that both Nogot and Havun are in the room with you.

(1) You have a well-justified false belief that Nogot owns a Ford.
(2) Hence you have a well-justified belief that someone in the room owns a Ford.
(3) That belief happens to be true, because Havun owns a Ford.

So, do you know that someone in the room owns a Ford? The answer would seem to be “no.”

Now clearly, if we are to salvage knowledge as JTB, then we must block the inference from (1) to (2). One way to do it is to say that there is no such thing as good evidence for false propositions or well-justified false beliefs. If you believe something false, then somewhere along the way you screwed up. Then the argument, while valid, will no longer be sound, because (1) is a contradiction and is false.

It follows that all (genuinely) well-justified beliefs are true. (It does not fortunately follow that all true beliefs are well-justified.) The obvious question is, how much justification is enough? Who decides when a proposition is finally well-justified? And how do I know where I ought to look for more evidence? I think, however, that these are pragmatic concerns about the best procedures for seeking truth about various matters. Certain standards can be established, such as the scientific method, or reliabilism, or which sciences are empirical and which are a priori, etc.

Look, getting back to our case, the evidence I have does make it reasonable to think that someone in the room owns a Ford. Why? Precisely because of Nogot. It’s his presence that causes me to feel good about my belief. But were I to discover the truth (by accumulating still better evidence), I’d immediately stop suspecting Nogot and therefore anyone in the room of having a Ford. I’d cease to be well-justified in this. (2) could then no longer be inferred. Can we not say that my evidence in favor of Nogot’s owning a Ford was lacking?

It may be objected that if we are actually fundamentally deceived brains-in-vats or victims of an evil demon, then we do seem to have well-justified false beliefs. I mean, what standard of justification am I possibly proposing if after an exhaustive investigation of the world I falsely conclude that I am not a brain-in-a-vat? Well, I’d simply deny that we are brains-in-vats. It’s a possibility not worth taking seriously. In practice we indeed judge truth by the strength of justification. But the theoretical point still stands: if we believe a lie, we must not have done our job or “epistemic duty” well enough. Here perhaps the Cartesian argument against our being systematically deceived from the existence and goodness of God might be employed. The world is complex and its secrets must be teased out of it, but it is not one big delusion.

Does all this lead us to anti-realism of any kind? Are only “mature sciences” productive of truths? How do we determine what’s a mature science? Do we know X if X or technologies based on X “work”? As one of my professors said, clearly coming down on the side of pragmatism, “We should do the best we can to justify our beliefs and hope that the world cooperates.”