Latest / Philosophy / 025 - XXV The Validity of Inference
Transcript
- 0:00Speaker 1: Chapter twenty five of Philosophy by Bertrand Russell. The Sleeper
- 0:04Speaker 1: Fox recording is in the public domain. The validity of inference.
- 0:11Speaker 1: It is customary in science to regard certain facts as
- 0:14Speaker 1: data from which laws and also other facts are inferred.
- 0:19Speaker 1: We saw in chapter seven that the practice of inference
- 0:23Speaker 1: is much wider than the theories of any logician would justify,
- 0:27Speaker 1: and that it is nothing other than the law of
- 0:29Speaker 1: association or of learned reactions. In the present chapter, I
- 0:34Speaker 1: wish to consider what the logicians have evolved from this
- 0:37Speaker 1: primitive form of inference, and what grounds we have as
- 0:40Speaker 1: rational beings for continuing to infer. But let us first
- 0:45Speaker 1: get as close a notion as we can of what
- 0:48Speaker 1: should be meant by a datum. The conception of a
- 0:51Speaker 1: datum cannot be made absolute. Theoretically, it should mean something
- 0:56Speaker 1: that we know without inference. But before this has any
- 0:59Speaker 1: definite meaning, we must define both knowledge and inference. Both
- 1:04Speaker 1: these terms have been considered in earlier chapters. For a
- 1:08Speaker 1: present purpose, it will simplify matters to take account only
- 1:12Speaker 1: of such knowledge as is expressed in words. We considered
- 1:16Speaker 1: in chapter twenty four the conditions required in order that
- 1:20Speaker 1: a form of words may be true for present purposes. Therefore,
- 1:24Speaker 1: we may say that knowledge means the assertion of a
- 1:27Speaker 1: true form of words. This definition is not quite adequate,
- 1:32Speaker 1: since a man may be right by chance, but we
- 1:35Speaker 1: may ignore this complication. We may then define a datum
- 1:39Speaker 1: as follows. A datum is a form of words which
- 1:43Speaker 1: a man utters as the result of a stimulus, with
- 1:46Speaker 1: no intermediary of any learned reaction beyond what is involved
- 1:51Speaker 1: in knowing how to speak. We must, however, permit such
- 1:55Speaker 1: learned reactions as consist in adjustments of the sense organs
- 1:59Speaker 1: or in mere ins increase of sensitivity. These merely improve
- 2:03Speaker 1: the receptivity to data, and do not involve anything that
- 2:06Speaker 1: can be called inference. If the above definition is accepted,
- 2:11Speaker 1: all our data for knowledge of the external world must
- 2:15Speaker 1: be of the nature of percepts. The belief in external
- 2:19Speaker 1: objects is a learned reaction acquired in the first months
- 2:22Speaker 1: of life, and it is the duty of the philosopher
- 2:25Speaker 1: to treat it as an inference whose validity must be tested.
- 2:30Speaker 1: A very little consideration shows that logically the inference cannot
- 2:34Speaker 1: be demonstrative, but must be at best probable. It is
- 2:39Speaker 1: not logically impossible that my life may be one long
- 2:43Speaker 1: dream in which I merely imagine all the objects that
- 2:46Speaker 1: I believe to be external to me. If we are
- 2:50Speaker 1: to reject this view, we must do so on the
- 2:52Speaker 1: basis of an inductive or analogical argument, which cannot give
- 2:57Speaker 1: complete certainty. We proceed other people behaving in a manner
- 3:02Speaker 1: analogous to that in which we behave, and we assume
- 3:05Speaker 1: that they have had similar stimuli. We may hear a
- 3:09Speaker 1: whole crowd say oh at the moment when we see
- 3:12Speaker 1: a rocket burst, and it is natural to suppose that
- 3:15Speaker 1: the crowd saw it too. Nor are such arguments confined
- 3:19Speaker 1: to living organisms. We can talk to a dictaphone and
- 3:23Speaker 1: have it afterwards repeat what we said. This is most
- 3:27Speaker 1: easily explained by the hypothesis that at the surface of
- 3:30Speaker 1: the dictaphone events happened while I was speaking which were
- 3:34Speaker 1: closely analogous to those that were happening just outside my ears.
- 3:39Speaker 1: It remains possible that there is no dictaphone and I
- 3:43Speaker 1: have no ears, and there is no crowd watching the rocket.
- 3:46Speaker 1: My percepts may be all that is happening in such cases,
- 3:50Speaker 1: but if so, it is difficult to arrive at any
- 3:54Speaker 1: causal laws and arguments from analogy are more misleading than
- 3:58Speaker 1: we are inclined to think them. As a matter of fact,
- 4:01Speaker 1: the whole structure of science, as well as the world
- 4:04Speaker 1: of common sense, demands the use of induction and analogy
- 4:09Speaker 1: if it is to be believed. These forms of inference, therefore,
- 4:13Speaker 1: rather than deduction, are those that must be examined if
- 4:17Speaker 1: we are to accept the world of science, or any
- 4:19Speaker 1: world outside of our own dreams. Let us take a
- 4:23Speaker 1: simple example of an induction, which we have all performed
- 4:26Speaker 1: in practice. If we are hungry, we eat certain things
- 4:30Speaker 1: we see and not others. It may be said that
- 4:33Speaker 1: we infer edibility inductively from a certain visual and old
- 4:38Speaker 1: factory appearance. The history of this process is that children
- 4:42Speaker 1: a few months old put everything into their mouths unless
- 4:45Speaker 1: they are stopped. Sometimes the result is pleasant, sometimes unpleasant.
- 4:51Speaker 1: They repeat the former rather than the latter. That is
- 4:54Speaker 1: to say, given that an object having a certain visual
- 4:58Speaker 1: and old factory appearance has been found pleasant to eat,
- 5:02Speaker 1: an object having a very similar appearance will be eaten.
- 5:06Speaker 1: But when a certain appearance has been found connected with
- 5:08Speaker 1: unpleasant consequences when eaten, a similar appearance does not lead
- 5:13Speaker 1: to eating next time. The question is what logical justification
- 5:18Speaker 1: is there for our behavior, given all our past experience.
- 5:22Speaker 1: Are we more likely to be nourished by bread than
- 5:25Speaker 1: by a stone? It is easy to see why we
- 5:28Speaker 1: think so, but can we, as philosophers, justify this way
- 5:32Speaker 1: of thinking. It is of course obvious that unless one
- 5:36Speaker 1: thing can be a sign of another, both science and
- 5:39Speaker 1: daily life would be impossible. More particularly, reading involves this principle.
- 5:46Speaker 1: One accepts printed words as signs. But this is only
- 5:49Speaker 1: justifiable by means of induction. I do not mean that
- 5:53Speaker 1: induction is necessary to establish the existence of other people,
- 5:58Speaker 1: though that, also, as we have seen, is true. I
- 6:01Speaker 1: mean something simpler. Suppose you want your haircut, and as
- 6:06Speaker 1: you walk along the street you see a notice hair
- 6:08Speaker 1: cutting first floor. It is only by means of induction
- 6:12Speaker 1: that you can establish that this notice makes it in
- 6:15Speaker 1: some degree probable that there is a haircutter's establishment on
- 6:19Speaker 1: the first floor. I do not mean that you employ
- 6:22Speaker 1: the principle of induction. I mean that you act in
- 6:25Speaker 1: accordance with it, and that you would have to appeal
- 6:28Speaker 1: to it if you were accompanied by a long haired
- 6:30Speaker 1: skeptical philosopher who refused to go upstairs till he was
- 6:34Speaker 1: persuaded there was some point in doing so. The principle
- 6:39Speaker 1: of induction prima facie is as follows. Let there be
- 6:43Speaker 1: two kinds of events A and B, for example, lightning
- 6:47Speaker 1: and thunder, and let many instances be known in which
- 6:51Speaker 1: an event of the kind A has been quickly followed
- 6:54Speaker 1: by one of the kind B, and no instances of
- 6:57Speaker 1: the contrary. Then either a sufficient number of instances of
- 7:02Speaker 1: this sequence or instances of suitable kinds, will make it
- 7:06Speaker 1: increasingly probable that A is always followed by B, and
- 7:10Speaker 1: in time the probability can be made to approach certainty
- 7:14Speaker 1: without limit, provided the right kind and number of instances
- 7:18Speaker 1: can be found. This is the principle we have to examine.
- 7:23Speaker 1: Scientific theories of induction generally try to substitute well chosen
- 7:28Speaker 1: instances for numerous instances, and represent number of instances as
- 7:32Speaker 1: belonging to crude popular induction. But in fact popular induction
- 7:38Speaker 1: depends upon the emotional interest of the instances, not upon
- 7:42Speaker 1: their number. A child who has burnt it's hand once
- 7:45Speaker 1: in a candle flame establishes an induction, but words take
- 7:49Speaker 1: longer because at first they are not emotionally interesting. The
- 7:54Speaker 1: principle used in primitive practice is whatever on a given
- 7:59Speaker 1: occasion immediately precedes something very painful or pleasant is a
- 8:04Speaker 1: sign of that interesting event. Number plays a secondary part
- 8:08Speaker 1: as compared with emotional interest. That is one reason why
- 8:13Speaker 1: rational thought is so difficult. The logical problem of induction
- 8:18Speaker 1: is to show that the proposition A is always accompanied
- 8:22Speaker 1: or followed by B can be rendered probable by knowledge
- 8:26Speaker 1: of instances in which this happened, provided the instances are
- 8:29Speaker 1: suitably chosen or very numerous. Far the best examination of
- 8:35Speaker 1: induction is contained in mister J. M. Keyne's Treatise on Probability.
- 8:40Speaker 1: There is a valuable doctor's thesis by the late Jean
- 8:43Speaker 1: NiCoT Les problem Lugique Lin duc Chon, which is very
- 8:47Speaker 1: ably reviewed by R. B. Braithwaite in Mind, October nineteen
- 8:51Speaker 1: twenty five. A man who reads these three will know
- 8:56Speaker 1: most of what is known about induction. The subject is
- 9:00Speaker 1: technical and difficult, involving a good deal of mathematics. But
- 9:04Speaker 1: I will attempt to give the gist of the results.
- 9:07Speaker 1: We will begin with the condition in which the problem
- 9:10Speaker 1: had been left by J. S. Mill. He had four
- 9:13Speaker 1: canons of induction by means of which, given suitable examples,
- 9:17Speaker 1: it could be demonstrated that A and B were causally
- 9:21Speaker 1: connected if the law of causation could be assumed. That
- 9:25Speaker 1: is to say, given the law of causation, the scientific
- 9:29Speaker 1: use of induction could be reduced to deduction. Roughly, the
- 9:33Speaker 1: method is this, we know that B must have a cause.
- 9:37Speaker 1: The cause cannot be C or D or E et cetera,
- 9:41Speaker 1: because we find by experimental observation that these may be
- 9:45Speaker 1: present without producing B. On the other hand, we never
- 9:49Speaker 1: succeed in finding A without its being accompanied or followed
- 9:53Speaker 1: by B. If A and B are both capable of quantity,
- 9:58Speaker 1: we may find further that the more there is of A,
- 10:01Speaker 1: the more there is of B. By such methods we
- 10:04Speaker 1: eliminate all possible causes except A. Therefore, since B must
- 10:10Speaker 1: have a cause, that cause must be A. All this
- 10:14Speaker 1: is not really induction at all. True induction only comes
- 10:19Speaker 1: in proving the law of causation. This law Mill regards
- 10:23Speaker 1: as proved by mere enumeration of instances. We know vast
- 10:28Speaker 1: numbers of events which have causes, and no events which
- 10:31Speaker 1: can be shown to be uncaused. Therefore, it is highly
- 10:35Speaker 1: probable that all events have causes. Leaving out of account
- 10:40Speaker 1: the fact that the law of causality cannot have quite
- 10:43Speaker 1: the form that Mill supposed, we are left with the
- 10:45Speaker 1: problem does mere number of instances afford a basis for induction?
- 10:51Speaker 1: If not, is there any other basis? This is the
- 10:54Speaker 1: problem to which mister Keynes addresses himself. Mister Keynes holds
- 10:59Speaker 1: that a deduction may be rendered more probable by number
- 11:02Speaker 1: of instances, not because of their mere number, but because
- 11:06Speaker 1: of the probability if the instances are very numerous, that
- 11:10Speaker 1: they will have nothing in common except the characteristics in
- 11:14Speaker 1: question we want. Let us suppose to find out whether
- 11:18Speaker 1: some quality A is always associated with some quality B.
- 11:23Speaker 1: We find instances in which this is the case. But
- 11:26Speaker 1: it may happen that in all our instances some quality
- 11:30Speaker 1: C is also present, and that it is C that
- 11:33Speaker 1: is associated with B. If we can so choose our
- 11:37Speaker 1: instances that they have nothing in common except the qualities
- 11:40Speaker 1: A and B, then we have better grounds for holding
- 11:43Speaker 1: that A is always associated with B. If our instances
- 11:48Speaker 1: are very numerous, then even if we do not know
- 11:51Speaker 1: that they have no other common quality, it may become
- 11:54Speaker 1: quite likely that this is the case. This, according to
- 11:58Speaker 1: mister Keynes, is the sole value of many instances. A
- 12:03Speaker 1: few technical terms are useful. Suppose we want to establish
- 12:07Speaker 1: inductively that there is some probability in favor of the
- 12:11Speaker 1: generalization everything that has the property capital F also has
- 12:17Speaker 1: the property lowercase F. We will call this generalization G.
- 12:24Speaker 1: Suppose we have observed a number of instances in which
- 12:27Speaker 1: uppercase F and lowercase F go together, and no instances
- 12:32Speaker 1: to the contrary. These instances may have other common properties
- 12:36Speaker 1: as well. The sum total of their common properties is
- 12:39Speaker 1: called the total positive analogy, and the sum total of
- 12:43Speaker 1: their known common qualities is called the known positive analogy.
- 12:48Speaker 1: The properties belonging to some, but not to all, of
- 12:51Speaker 1: the instances in question are called the negative analogy. All
- 12:55Speaker 1: of them constitute the total negative analogy. All those that
- 12:59Speaker 1: are known constitute the known negative analogy. To strengthen an induction,
- 13:06Speaker 1: we want to diminish the positive analogy to the utmost
- 13:09Speaker 1: possible extent. This, according to mister Keynes, is why numerous
- 13:14Speaker 1: instances are useful on pure induction, where we rely solely
- 13:19Speaker 1: upon number of instances without knowing how they affect the analogy.
- 13:24Speaker 1: Mister Keenes concludes Treatise in Probability, page two thirty six.
- 13:30Speaker 1: We have shown that if each of the instances necessarily
- 13:33Speaker 1: follows from the generalization, then each additional instance increases the
- 13:38Speaker 1: probability of the generalization, so long as the new instance
- 13:42Speaker 1: could not have been predicted with certainty from a knowledge
- 13:45Speaker 1: of the former instances. The common notion that each successive
- 13:50Speaker 1: verification of a doubtful principle strengthens it is formally proved therefore,
- 13:56Speaker 1: without any appeal to conceptions of law or of causality.
- 14:00Speaker 1: But we have not proved that this probability approaches certainty
- 14:03Speaker 1: as a limit, or even that our conclusion becomes more
- 14:07Speaker 1: likely than not. As the number of verifications or instances
- 14:11Speaker 1: is indefinitely increased. It is obvious that induction is not
- 14:16Speaker 1: much use unless, with suitable care, its conclusion can be
- 14:20Speaker 1: rendered more likely to be true than false. This problem
- 14:24Speaker 1: therefore necessarily occupies mister Keynes. It is found that an
- 14:29Speaker 1: induction will approach certainty as a limit if two conditions
- 14:32Speaker 1: are fulfilled. One if the generalization is false. The probability
- 14:38Speaker 1: of its being true in a new instance when it
- 14:40Speaker 1: has been found to be true in a certain number
- 14:43Speaker 1: of instances however great, that number may be falls short
- 14:47Speaker 1: of certainty by a finite amount. Two, there is a
- 14:51Speaker 1: finite a priori probability in favor of our generalization. Mister
- 14:57Speaker 1: Keynes uses finite here in a special sense. He holds
- 15:01Speaker 1: that not all probabilities are numerically measurable. A finite probability
- 15:06Speaker 1: is one which exceeds some numerically measurable probability, however small.
- 15:12Speaker 1: For example, our generalization has a finite a priori probability
- 15:17Speaker 1: if it is less unlikely than throwing heads a billion
- 15:20Speaker 1: times running. The difficulty is, however, that there is no
- 15:25Speaker 1: easily discoverable way of estimating the a priori probability of
- 15:29Speaker 1: a generalization. In examining this question, mister Keynes has led
- 15:34Speaker 1: to a very interesting postulate, which, if true, will he thinks,
- 15:39Speaker 1: give the required finite a priori probability. His postulate, as
- 15:45Speaker 1: he gives it, is not quite correct, but I shall
- 15:47Speaker 1: give his form first, and then the necessary modification. Mister
- 15:52Speaker 1: Keanes supposes that the qualities of objects cohere in groups,
- 15:57Speaker 1: so that the number of independent qualities is much less
- 16:00Speaker 1: than the total number of qualities. We may conceive this
- 16:04Speaker 1: after the analogy of biological species. A cat has a
- 16:08Speaker 1: number of distinctive qualities which are found in all cats.
- 16:12Speaker 1: A dog has a number of other distinctive qualities which
- 16:15Speaker 1: are found in all dogs. The method of induction can,
- 16:19Speaker 1: he says, be justified if we assume that the objects
- 16:23Speaker 1: in the field over which our generalizations extend do not
- 16:27Speaker 1: have an infinite number of independent qualities, that, in other words,
- 16:32Speaker 1: their characteristics, however, numerous cohere together in groups of invariable connection,
- 16:38Speaker 1: which are finite in number, page two fifty six, again
- 16:43Speaker 1: page two fifty eight. As a biological foundation for analogy, therefore,
- 16:48Speaker 1: we seem to need some assumption as that the amount
- 16:51Speaker 1: of variety in the universe is limited in such a
- 16:54Speaker 1: way that there is no one object so complex that
- 16:58Speaker 1: its qualities fall into an infinite number of independent groups,
- 17:03Speaker 1: or rather that none of the objects about which we
- 17:06Speaker 1: generalize are as complex as this, or at least that
- 17:10Speaker 1: though some objects may be infinitely complex, we sometimes have
- 17:15Speaker 1: a finite probability that an object about which we seek
- 17:18Speaker 1: to generalize is not infinitely complex. This postulate is called
- 17:24Speaker 1: the principle of limitation of variety. Mister Keynes again finds
- 17:29Speaker 1: that it is needed in attempts to establish laws by statistics.
- 17:33Speaker 1: If he is right, it is needed for all our
- 17:36Speaker 1: scientific knowledge outside pure mathematics. John NiCoT pointed out that
- 17:41Speaker 1: it is not quite sufficiently stringent. We need, according to
- 17:46Speaker 1: mister Keynes, a finite probability that the object in question
- 17:50Speaker 1: has only a finite number of independent qualities. But what
- 17:54Speaker 1: we really need is a finite probability that the number
- 17:57Speaker 1: of its independent qualities is less than some assigned finite number.
- 18:03Speaker 1: This is a very different thing, as may be seen
- 18:06Speaker 1: by the following illustration. Suppose there is some number of
- 18:10Speaker 1: which we know only that it is finite. It is
- 18:13Speaker 1: infinitely improbable that it will be less than a million,
- 18:17Speaker 1: or a billion, or any other assigned finite number, because
- 18:21Speaker 1: whatever such number we take, the number of smaller numbers
- 18:25Speaker 1: is finite and the number of greater numbers is infinite.
- 18:29Speaker 1: It is the stronger assumption which is needed to justify induction.
- 18:34Speaker 1: This result is very interesting and very important. It is
- 18:38Speaker 1: remarkable that it is in line with the trend of
- 18:40Speaker 1: modern science. Eddington has pointed out that there is a
- 18:44Speaker 1: certain finite number which is fundamental in the universe, namely
- 18:48Speaker 1: the number of electrons. According to the quantum theory, It
- 18:52Speaker 1: would seem that the number of possible arrangements of electrons
- 18:56Speaker 1: may well also be finite, since they cannot move in
- 18:59Speaker 1: all possible orbits, but only in such as make the
- 19:03Speaker 1: action in one complete revolution conform to the quantum principle.
- 19:08Speaker 1: If all this is true, the principle of limitation of
- 19:11Speaker 1: variety may well also be true. We cannot, however, arrive
- 19:15Speaker 1: at a proof of our principle in this way, because
- 19:18Speaker 1: physics uses induction and is therefore presumably invalid unless the
- 19:23Speaker 1: principle is true. What we can say in a general
- 19:27Speaker 1: way is that the principle does not refute itself, but
- 19:31Speaker 1: on the contrary, leads to results which confirm it. To
- 19:35Speaker 1: this extent, the trend of modern science may be regarded
- 19:38Speaker 1: as increasing the plausibility of the principle. It is important
- 19:43Speaker 1: to realize the fundamental position of probability in science. At
- 19:48Speaker 1: the very best, induction and analogy only give probability. Every
- 19:53Speaker 1: inference worthy of the name is inductive. Therefore, all inferred
- 19:58Speaker 1: knowledge is at best. Probably As to what is meant
- 20:01Speaker 1: by probability, opinions differ. Mister Keynes takes it as a
- 20:06Speaker 1: fundamental logical category certain premises may make a conclusion more
- 20:11Speaker 1: or less probable without making it certain for him. Probability
- 20:16Speaker 1: is a relation between a premise and a conclusion. A
- 20:20Speaker 1: proposition does not have a definite probability on its own account.
- 20:24Speaker 1: In itself, it is merely true or false, but it
- 20:28Speaker 1: has probabilities of different amounts in regard to different premises.
- 20:33Speaker 1: When we speak elliptically of the probability of a proposition,
- 20:37Speaker 1: we mean its probability in relation to all our relevant knowledge.
- 20:42Speaker 1: A proposition in probability cannot be refuted by mere observation.
- 20:47Speaker 1: Improbable things may happen, and probable things may fail to happen.
- 20:52Speaker 1: Nor is an estimate of probability relevant to given evidence
- 20:56Speaker 1: proved wrong when further evidence alters the problem probability. For
- 21:01Speaker 1: this reason, the inductive principle cannot be proved or disproved
- 21:05Speaker 1: by experience. We might prove validly that such and such
- 21:09Speaker 1: a conclusion was enormously probable, and yet it might not happen.
- 21:15Speaker 1: We might prove invalidly that it was probable, and yet
- 21:18Speaker 1: it might happen. What happens affects the probability of a proposition,
- 21:23Speaker 1: since it is relevant evidence, but it never alters the
- 21:27Speaker 1: probability relative to the previously available evidence. The whole subject
- 21:32Speaker 1: of probability, therefore, on mister Keane's theory is strictly a
- 21:36Speaker 1: priori an independent of experience. There is, however, another theory,
- 21:42Speaker 1: called the frequency theory, which would make probability not indefinable
- 21:47Speaker 1: and would allow empirical evidence to affect our estimates of
- 21:51Speaker 1: probability relative to given premises. According to this theory, in
- 21:56Speaker 1: its crude form, the probability that an object having the
- 22:00Speaker 1: property uppercase F will have the property lower case F
- 22:05Speaker 1: is simply the proportion of the objects having both properties
- 22:09Speaker 1: to all those having the property uppercase F. For example,
- 22:14Speaker 1: in a monogamous country, the probability of a married person
- 22:17Speaker 1: being male is exactly a half. Mister Keynes advances strong
- 22:23Speaker 1: arguments against all forms of this theory that existed when
- 22:26Speaker 1: his book was written. There is, however, an article by R. H.
- 22:30Speaker 1: Nisbet on the Foundations of Probability in Mind for January
- 22:35Speaker 1: nineteen twenty six, which undertakes to rehabilitate the frequency theory.
- 22:41Speaker 1: His arguments are interesting and suffice to show that the
- 22:44Speaker 1: controversy is still an open one, but they do not,
- 22:47Speaker 1: in my opinion, amount to decisive proof. It is to
- 22:51Speaker 1: be observed, however, that the frequency theory, if it could
- 22:55Speaker 1: be maintained, would be preferable to mister Keynes's, because it
- 22:59Speaker 1: would get rid of the necessity for treating probability is indefinable,
- 23:03Speaker 1: and would bring probability into much closer touch with what
- 23:07Speaker 1: actually occurs. Mister Keynes leaves an uncomfortable gap between probability
- 23:14Speaker 1: and fact, so that it is far from clear why
- 23:17Speaker 1: a rational man will act upon a probability. Nevertheless, the
- 23:22Speaker 1: difficulties of the frequency theory are so considerable that I
- 23:26Speaker 1: cannot venture to advocate it definitely. Meanwhile, the details of
- 23:31Speaker 1: the discussion are unaffected by the view we may take
- 23:34Speaker 1: on this fundamental philosophical question, and on either view, the
- 23:39Speaker 1: principle of limitation of variety will be equally necessary to
- 23:43Speaker 1: give validity to the inferences by induction and analogy upon
- 23:48Speaker 1: which science and daily life depend. End of Chapter twenty five.