On the Philosophy of Discovery, Chapters Historical and Critical

CHAPTER IV.

Aristotle

The views of Aristotle with regard to the foundations of human knowledge are very different from those of his tutor Plato, and are even by himself put in opposition to them. He dissents altogether from the Platonic doctrine that Ideas are the true materials of our knowledge; and after giving, respecting the origin of this doctrine, the account which we quoted in the last chapter, he goes on to reason against it. "Thus," he says11, "they devised Ideas of all things which are spoken of as universals: much as if any one having to count a number of objects, should think that he could not do it while they were few, and should expect to count them by making them more numerous. For the kinds of things are almost more numerous than the special sensible objects, by seeking the causes of which they were led to their Ideas." He then goes on to urge several other reasons against the assumption of Ideas and the use of them in philosophical researches.

Aristotle himself establishes his doctrines by trains of reasoning. But reasoning must proceed from certain First Principles; and the question then arises, Whence are these First Principles obtained? To this he replies, that they are the result of Experience, and he even employs the same technical expression by which we at this day describe the process of collecting these principles from observed facts;—that they are obtained by Induction. I have already quoted passages in which this statement is made12. "The way of reasoning," he says13, "is the same in philosophy, and in any art or science: we must collect the facts (τὰ ὑπὰρχοντα), and the things to which the facts happen, and must have as large a supply of these as possible, and then we must examine them according to the terms of our syllogisms." … "There are peculiar principles in each science; and in each case these principles must be obtained from experience. Thus astronomical observation supplies the principles of astronomical science. For the phenomena being rightly taken, the demonstrations of astronomy were discovered; and the same is the case with any other Art or Science. So that if the facts in each case be taken, it is our business to construct the demonstrations. For if in our natural history (κατὰ τὰν ἱστορί αν) we have omitted none of the facts and properties which belong to the subject, we shall learn what we can demonstrate and what we cannot." And again14, "It is manifest that if any sensation be wanting, there must be some knowledge wanting, which we are thus prevented from having. For we acquire knowledge either by Induction (ἐπαγωγῆ) or by Demonstration: and Demonstration is from universals, but Induction from particulars. It is impossible to have universal theoretical propositions except by Induction: and we cannot make inductions without having sensation; for sensation has to do with particulars."

It is easy to show that Aristotle uses the term Induction, as we use it, to express the process of collecting a general proposition from particular cases in which it is exemplified. Thus in a passage which we have already quoted15, he says, "Induction, and Syllogism from Induction, is when we attribute one extreme term to the middle by means of the other." The import of this technical phraseology will further appear by the example which he gives: "We find that several animals which are deficient in bile are long-lived, as man, the horse, the mule; hence we infer that all animals which are deficient in bile are long-lived."

We may observe, however, that both Aristotle's notion of induction, and many other parts of his philosophy, are obscure and imperfect, in consequence of his refusing to contemplate ideas as something distinct from sensation. It thus happens that he always assumes the ideas which enter into his proposition as given; and considers it as the philosopher's business to determine whether such propositions are true or not: whereas the most important feature in induction is, as we have said, the introduction of a new idea, and not its employment when once introduced. That the mind in this manner gives unity to that which is manifold,—that we are thus led to speculative principles which have an evidence higher than any others,—and that a peculiar sagacity in some men seizes upon the conceptions by which the facts may be bound into true propositions,—are doctrines which form no essential part of the philosophy of the Stagirite, although such views are sometimes recognized, more or less clearly, in his expressions. Thus he says16, "There can be no knowledge when the sensation does not continue in the mind. For this purpose, it is necessary both to perceive, and to have some unity in the mind (αἰσθανομένοις εχειν ἔν τι17 ἐν τῇ ψυχῇ); and many such perceptions having taken place, some difference is then perceived: and from the remembrance of these arises Reason. Thus from Sensation comes Memory, and from Memory of the same thing often repeated comes Experience: for many acts of Memory make up one Experience. And from Experience, or from any Universal Notion which takes a permanent place in the mind,—from the unity in the manifold, the same some one thing being found in many facts,—springs the first principle of Art and of Science; of Art, if it be employed about production; of Science, if about existence."

I will add to this, Aristotle's notice of Sagacity; since, although little or no further reference is made to this quality in his philosophy, the passage fixes our attention upon an important step in the formation of knowledge. "Sagacity" (ἀγχίνοια), he says18, "is a hitting by guess (εὐστοχία τις) upon the middle term (the conception common to two cases) in an inappreciable time. As for example, if any one seeing that the bright side of the moon is always towards the sun, suddenly perceives why this is; namely, because the moon shines by the light of the sun:—or if he sees a person talking with a rich man, he guesses that he is borrowing money;—or conjectures that two persons are friends, because they are enemies of the same person."—To consider only the first of these examples;—the conception here introduced, that of a body shining by the light which another casts upon it, is not contained in the observed facts, but introduced by the mind. It is, in short, that conception which, in the act of induction, the mind superadds to the phenomena as they are presented by the senses: and to invent such appropriate conceptions, such "eustochies," is, indeed, the precise office of inductive sagacity.

At the end of this work (the Later Analytics) Aristotle ascribes our knowledge of principles to Intellect (νοῦς), or, as it appears necessary to translate the word, Intuition19. "Since, of our intellectual habits by which we aim at truth, some are always true, but some admit of being false, as Opinion and Reasoning, but Science and Intuition are always true; and since there is nothing which is more certain than Science except Intuition; and since Principles are better known to us than the Deductions from them; and since all Science is connected by reasoning, we cannot have Science respecting Principles. Considering this then, and that the beginning of Demonstration cannot be Demonstration, nor the beginning of Science, Science; and since, as we have said, there is no other kind of truth, Intuition must be the beginning of Science."

What is here said, is, no doubt, in accordance with the doctrines which we have endeavoured to establish respecting the nature of Science, if by this Intuition we understand that contemplation of certain Fundamental Ideas, which is the basis of all rigorous knowledge. But notwithstanding this apparent approximation, Aristotle was far from having an habitual and practical possession of the principles which he thus touches upon. He did not, in reality, construct his philosophy by giving Unity to that which was manifold, or by seeking in Intuition principles which might be the basis of Demonstration; nor did he collect, in each subject, fundamental propositions by an induction of particulars. He rather endeavoured to divide than to unite; he employed himself, not in combining facts, but in analysing notions; and the criterion to which he referred his analysis was, not the facts of our experience, but our habits of language. Thus his opinions rested, not upon sound inductions, gathered in each case from the phenomena by means of appropriate Ideas; but upon the loose and vague generalizations which are implied in the common use of speech.

Yet Aristotle was so far consistent with his own doctrine of the derivation of knowledge from experience, that he made in almost every province of human knowledge, a vast collection of such special facts as the experience of his time supplied. These collections are almost unrivalled, even to the present day, especially in Natural History; in other departments, when to the facts we must add the right Inductive Idea, in order to obtain truth, we find little of value in the Aristotelic works. But in those parts which refer to Natural History, we find not only an immense and varied collection of facts and observations, but a sagacity and acuteness in classification which it is impossible not to admire. This indeed appears to have been the most eminent faculty in Aristotle's mind.

The influence of Aristotle in succeeding ages will come under our notice shortly.

CHAPTER V.

Additional Remarks on Aristotle

1. ONE of the most conspicuous points in Aristotle's doctrines as bearing upon the philosophy of Science is his account of that mode of attaining truth which is called Induction; for we are accustomed to consider Induction as the process by which our Sciences have been formed; and we call them collectively the Inductive Sciences. Aristotle often speaks of Induction, as for instance, when he says that Socrates introduced the frequent use of it. But the cardinal passage on this subject is in his Analytics, in which he compares Syllogism and Induction as two modes of drawing conclusions20. He there says that all belief arises either from Syllogism or from Induction: and adds that Induction is, when by means of one extreme term we infer the other extreme to be true of the middle term. The example which he gives is this: knowing that particular animals are long-lived, as elephant, horse, mule; and finding that these animals agree in having no gall-bladder; we infer, by Induction, that all animals which have no gall-bladder are long-lived. This may be done, he says, if the middle and the second extreme are convertible: as the following formal statement may show.

Elephant, horse, mule, &c. are long-lived.

Elephant, horse, mule, &c. are all gall-less.

If we might convert this proposition, and say

All gall-less animals are as elephant, horse, mule, &c.:

we might infer syllogistically that

All gall-less animals are long-lived.

And though we cannot infer this syllogistically, we infer it by Induction, when we have a sufficient amount of instances21.

I have already elsewhere given this account of Induction, as a process employed in the formation of our knowledge22. What I have now to remark concerning Aristotle is, that it does not appear to have occurred to him, that in establishing such a proposition as that which he gives as his instance, the main difficulty is the discovery of a middle term which will allow us to frame such a proposition as we need. The zoologist who wanted to know what kind of animals are long-lived, might guess long before he guessed that the absence of the gall-bladder supplied the requisite middle term; (if the proposition were true; which it is not.) And in like manner in other cases, it is difficult to find a middle term, which enables us to collect a proposition by Induction. And herein consists the imperfection of his view of the subject; which considers the main point to be the proof of the proposition when the conceptions are given, whereas the main point really is, the discovery of conceptions which will make a true proposition possible.

2. Since the main characteristic of the steps which have occurred in the formation of the physical sciences, is not merely that they are propositions collected by Induction, but by the introduction of a new conception; it has been suggested that it is not a characteristic designation of these Sciences to call them Inductive Sciences. Almost every discovery involves in it the introduction of a new conception, as the element of a new proposition; and the novelty of the conception is more characteristic of the stages of discovery than the inductive application of it. Hence as bearing upon the Philosophy of Discovery, the statements of Aristotle concerning Induction, though acute and valuable, are not so valuable as they might seem. Even Francis Bacon, it has been asserted, erred in the same way (and of course with less excuse) in asserting Induction, of a certain kind, to be the great instrument for the promotion of knowledge, and in overlooking the necessity of the Invention which gives Induction its value.

3. The invention or discovery of a conception by which many facts of observation are conjoined so as to make them the materials of a proposition, is called in Plato, as we have seen, finding the One in the Many.

In the passage quoted from the Later Analytics, Aristotle uses the same expression, and speaks very justly respecting the formation of knowledge. Indeed the Titles of the chapters of this and many parts of Aristotle's works would lead us to expect just such a Philosophy of Discovery as is the object of our study at present. Thus we have, Anal. Post. B. II. chap. 13: "How we are to hunt (θηρεύειν) the predications of a Definition." Chap. 14: "Precepts for the invention of Problems and of a Middle Term:" and the like. But when we come to read these chapters, they contain little that is of value, and resolve themselves mostly into permutations of Aristotle's logical phraseology.

4. The part of the Aristotelian philosophy which has most permanently retained its place in modern Sciences is a part of which a use has been made quite different from that which was originally contemplated. The "Five words" which are explained in the Introduction to Aristotle's Categories: namely, the words Genus, Species, Difference, Property, Accident, were introduced mainly that they might be used in the propositions of which Syllogisms consist, and might thus be the elements of reasoning. But it has so happened that these words are rarely used in Sciences of Reasoning, but are abundantly and commonly used in the Sciences of Classification, as I have explained in speaking of the Classificatory Sciences23.

5. Of Aristotle's actual contributions to the Physical Sciences I have spoken in the History of those Sciences24. I have25 stated that he conceived the globular form of the earth so clearly and gave so forcibly the arguments for that doctrine, that we may look upon him as the most effective teacher of it. Also in the Appendix to that History, published in the third edition, I have given Aristotle's account of the Rainbow, as a further example of his industrious accumulation of facts, and of his liability to error in his facts.

6. We do not find Aristotle so much impressed as we might have expected by that great monument of Grecian ingenuity, the theory of epicycles and excentrics which his predecessor Plato urged so strongly upon the attention of his contemporaries. Aristotle proves, as I have said, the globular form of the earth by good and sufficient arguments. He also proves by arguments which seem to him quite conclusive26, that the earth is in the center of the universe, and immoveable. As to the motions of the rest of the planets, he says little. The questions of their order, and their distances, and the like, belong, he says, to Astrology27. He remarks only that the revolution of the heaven itself, the outermost revolution, is simple and the quickest of all: that the revolutions of the others are slower, each moving in a direction opposite to the heaven in its own circle: and that it is reasonable that those which are nearest to the first revolution should take the longest time in describing their own circle, and those that are furthest off, the least time, and the intermediate ones in the order of their distances, "as also the mathematicians show."

In the Metaphysics28 he enumerates the circular movements which had been introduced by the astronomers Eudoxus and Calippus for the explanation of the phenomena presented by the sun, moon and planets. These, he says, amount to fifty-five; and this, he says, must be the number of essences and principles which exist in the universe.

7. In the Sciences of Classification, and especially in the classification of animals, higher claims have been made for Aristotle, which I have discussed in the History29. I have there attempted to show that Aristotle's classification, inasmuch as it enumerates all the parts of animals, may be said to contain the materials of every subsequent classification: but that it cannot be said to anticipate any modern system, because the different grades of classification are not made subordinate to one another as a system of classification requires. I have the satisfaction of finding Mr. Owen agreeing with me in these views30.

8. Francis Bacon's criticism on Aristotle which I have quoted in the Appendix to the History31, is severe, and I think evidently the result of prejudice. He disparages Aristotle in comparison with the other philosophers of Greece. 'Their systems,' he says, 'had some savour of experience, and nature, and bodily things; while the Physics of Aristotle, in general, sound only of Logical Terms.

'Nor let anyone be moved by this: that in his books Of Animals, and in his Problems, and in others of his tracts, there is often a quoting of experiments. For he had made up his mind beforehand; and did not consult experience in order to make right propositions and axioms, but when he had settled his system to his will, he twisted experience round and made her bend to his system.'

I do not think that this can be said with any truth. I know no instances in which Aristotle has twisted experience round, and made her bend to his system. In his Problems, he is so far from giving dogmatical solutions of the questions proposed, that in most cases, he propounds two or three solutions as mere suggestions and conjectures. And both in his History of Animals, as I have said, and in others of his works, the want of system gives them an incoherent and tumultuary character, which even a false system would have advantageously removed; for, as I have said elsewhere, it is easier to translate a false system into a true one, than to introduce system into a mass of confusion.