One obvious effect of the grossness and vagueness in the perceptions of the inexperienced observer, must necessarily be to indentify under the same common appellations, immense multitudes of individuals which the philosopher will afterwards find reason to distinguish

Language carefully from each other; and as language, by its unavoida ble reaction on thought, never fails to restore to it whatever imper& thought fections it has once received, all the indistinctness which, in the mutually case of individual observers, originated in an ill-informed judgment, act on each or in a capricious fancy, comes afterwards, in succeeding ages, to be other. entailed on the infant understanding, in consequence of its incorporation with vernacular speech. These confused apprehensions produced by language, must, it is easy to see, operate exactly in the same way as the undistinguishing perceptions of children or savages; the familiar use of a generic word, insensibly and irresistibly leading the mind to extend its conclusions from the individual to the genus, and thus laying the foundation of conclusions and anticipations which we suppose to rest on experience, when, in truth, experience. has never been consulted.

In all such instances, it is worthy of observation, we proceed ultimately on the common principle, that, in similar circumstances, the same cause will produce the same effects; and, when we err, the source of our errour lies merely in identifying different cases which ought to be distinguished from each other. Great as may be the occasional inconveniences, arising from this general principle thus misapplied, they bear no proportion to the essential advantages resulting from the disposition in which they originate, to arrange and to classify; a disposition on which (as I have elsewhere shown,) the intellectual improvement of the species a great manner hinges. That the constitution of our nature in this respect is, on the whole, wisely ordered, as well as perfectly conformable to the general economy of our frame, will appear from a slight survey of some other principles, nearly allied to those which are at present under our consideration.

reasoning powers of brutes, from the universal conclusions which they appear to found on the observation of particulars. "Les bestes des singuliers concluent les universels, du regard d'un homme seul cognoissent tous hommes," &c. &c. De la Sagesse, Lib. 1. Chap 8.

Instead of saying, that brutes generalize things which are similar, would it not be nearer the truth to say, that they confound things which are different ?

Many years after these observations were written, I had the satisfaction to meet with the following experimental confirmation of them, in the Abbé Sicard's Course of Instruction for the Deaf and Dumb: “J'avois remarqué que Massieu donnoit plus volontiers le même nom, un nom commun, à plusieurs individus dans lesquels il trouvoit des traits de ressemblance; les noms individuels supposoient des differences qu'il n'étoit pas encore temps de lui faire observer." (Sicard. pp. 30, 31.) The whole of the passage is well worth consulting.

(I remarked that Massieu more readily gave the same name, a common name, to several individuals in whom he found any marks of resemblance. Individual names sup. posed differences, which as yet he had not had time to observe.]

It has been remarked by some eminent writers in this part of the island,* that our expectation of the continuance of the laws of na

Belief in y ture has a very close affinity to our faith in human testimony. The uniformity parallel might perhaps be carried, without any over-refinement, a little farther than these writers have attempted; in as much as, in of y laws of both cases, the instinctive principle is in the first instance unlimited, nature has and requires, for its correction and regulation, the lessons of subse- an affinity quent experience. As the credulity of children is originally with-to out bounds, and is afterwards gradually checked by the exampleso forth in which they occasionally meet with of human falsehood, so, in the in- testimony. fancy of our knowledge, whatever objects or events present to our senses a strong resemblance to each other, dispose us, without any very accurate examination of the minute details by which they may be really discriminated, to conclude with eagerness, that the experiments and observations which we make with respect to one individual, may be safely extended to the whole class. It is experience alone that teaches us caution in such inferences, and subjects the natural principle to the discipline prescribed by the rules of induction.

+.

It must not, however, be imagined, that, in instances of this sort, This instin the instinctive principle always leads us astray; for the analogical anticipations which it disposes us to form, although they may not tive princi stand the test of a rigorous examination, may yet be sufficiently just ple salutary for all the common purposes of life. It is natural, for example, that a man who has been educated in Europe should expect, when he changes his residence to any of the other quarters of the globe, to see heavy bodies fall downwards, and smoke to ascend, agreeably to the general laws to which he has been accustomed; and that he should take for granted, in providing the means of his subsistence, that the animals and vegetables, which he has found to be salutary and nutritious in his native regions, possess the same qualities wherever they exhibit the same appearances. Nor are such expectations less useful than natural; for they are completely realized, as far as they minister to the gratification of our more urgent wants. It is only when we begin to indulge our curiosity with respect to those nicer details which derive their interest from great refinement in the arts, or from a very advanced state of physical knowledge, that we discover our first conclusions, however just in the main, not to be mathematically exact; and are led by those habits which scientific pursuits communicate, to investigate the difference of circumstances to which the variety in the result is owing. After having found that heavy bodies fall downwards at the equator as they do in this island, the most obvious, and perhaps, on a superficial view of the ques tion, the most reasonable inference would be, that the same pendulum which swings seconds at London, will vibrate at the same rate

See Reid's Inquiry into the Human Mind, Chap. VI. sect. 24. Campbell's Disserta tion on Miracles, Part I. sect. I. Smith's Theory of Moral Sentiments, Vol. II. p. 206, &c. Boston edition.

under the line. In this instance, however, the theoretical inference is contradicted by the fact ;-but the contradiction is attended with no practical inconvenience to the multitude, while, in the mind of the philosopher, it only serves to awaken his attention to the different circumstances of the two cases, and, in the last result, throws a new lustre on the simplicity and uniformity of that law, from which it seemed, at first sight, an anomalous deviation.

To this uniformity in the laws which regulate the order of physiAnalogy cal events, there is something extremely similar in the systematical bet: uni regularity (subject indeed to many exceptions) which, in every language, however imperfect, runs through the different classes of its formity of words, in respect to their inflections, forms of derivation, and other physical e-verbal filiations or affinities. How much this regularity or analogy centivy (as it is called by grammarians,) contributes to facilitate the acquisiぞ sition of dead and foreign languages, every person, who has receivryulavity ed a liberal education, knows from his own experience. Nor is it wh prevail, less manifest, that the same circumstance must contribute powerfulshutumly to aid the memories of children in learning to speak their mother tongue. It is not my present business to trace the principles in the

in

of languages, human mind by which it is produced. All that I would remark is,

the very early period at which it is seized by children; as is strongly evinced by their disposition to push it a great deal too far, in their first attempts towards speech. This disposition seems to be closely connected with that which leads them to repose faith in testimony; and it also bears a striking resemblance to that which prompts them to extend their past experience to those objects and events of which they have not hitherto had any means of acquiring a direct knowledge. It is probable, indeed, that our expectation, in all these cases, has its origin in the same common principles of our nature; and it is certain that, in all of them, it is subservient to the important purpose of facilitating the progress of the mind. Of this nobody can doubt, who considers for a moment, that the great end to be first accomplished, was manifestly the communication of the general rule; the acquisition of the exceptions (a knowledge of which is but of secondary importance,) being safely entrusted to the growing diligence and capacity of the learner.

Conclusion The considerations now stated, may help us to conceive in what manner conclusions derived from experience come to be insensibly from the individual to the species; partly in consequence jrom expe-extender oss and undistinguishing nature of our first perceptions, and partly in consequence of the magical influence of a common

mines how

extended to name. They seem also to show, that this natural process of geneval ring thought, though not always justified by a sound logic, is not without its use in the infancy of human knowledge.

In the various cases which have been hitherto under our review, our conclusions are said in popular, and even in philosophical language, to be founded on experience. And yet the truth unquestionably is, (as was formerly observed) that the evidence of experience

Exaut extent of 1⁄2 evidence of experience

reaches no farther than to an anticipation of the future from the past, in instances where the same cause continues to operate in circumstances exactly similar. How much this vagueness of expression must contribute to mislead us in many of our judgments will afterwards appear.

The observations which I have to offer upon analogy, considered as a ground of scientific conjecture and reasoning, will be introduced with more propriety in a future chapter.

IV.

Continuation of the Subject.-Evidence of Testimony tacitly recognised as a Ground of Belief, in our most certain conclusions concerning contingent Truths.-Difference be tween the Logical and the Popular Meaning of the word Probability.

In some of the conclusions which have been already under our consideration with respect to contingent truths, a species of evi- Evidence dence is admitted, of which no mention has hitherto been made; I

mean the evidence of testimony. In astronomical calculations, for of testimon example, how few are the instances in which the data rest on the evidence of our own senses; and yet our confidence in the result is not, on that account, in the smallest degree weakened. On the contrary, what certainty can be more complete, than that with which we look forward to an eclipse of the sun or the moon, on the faith of elements and of computations which we have never verified, and for the accuracy of which we have no ground of assurance whatever, but the scientific reputation of the writers from whom we have borrowed them? An astronomer who should affect any scepticism with respect to an event so predicted, would render himself no less an object of ridicule, than if he were disposed to cavil about the certainty of the sun's rising to-morrow.

Even in pure mathematics, a similar regard to testimony, accompanied with a similar faith in the faculties of others, is by no means uncommon. Who would scruple, in a geometrical investigation, to adopt, as a link in the chain, a theorem of Apollonius or of Archimedes, although he might not have leisure at the moment, to satisfy himself, by an actual examination of their demonstrations, that they had been guilty of no paralogism, either from accident or design, in the course of their reasonings?

Philosophi

of probab

In our anticipations of astronomical phenomena, as well as in those which we form concerning the result of any familiar experiment in physics, philosophers are accustomed to speak of the event al sins. as only probable; although our confidence in its happening is not less of complete, than if it rested on the basis of mathematical demonstration. The word probable, therefore, when thus used, does not imply any deficiency in the proof, but only marks the particular nature of that proof, as contradistinguished from another species of evidence. It is opposed, not to what is certain, but to what admits of being demonstrated after the manner of mathematicians.

VOL. II.

18

Specifi dif:

This differs widely from the meaning annexed to the same word in popular discourse; according to which, whatever event is said to be probable, is understood to be expected with some degree of doubt. As certain as death-as certain as the rising of the sun-are proverbial modes of expression in all countries: and they are, both of them, borrowed from events which, in philosophical language, are only probable or contingent. In like manner, the existence of the city of Pekin, and the reality of Caesar's assassination, which the philosopher classes with probabilities, because they rest solely upon the evidence of testimony, are universally classed with certainties by the rest of mankind; and in any case but the statement of a logical theory, the application to such truths of the word probable, would be justly regarded as an impropriety of speech. This difference between the technical meaning of the word probability, as employed by logicians, and the notion usually attached to it in the business of life; together with the erroneous theories concerning the nature of demonstration, which I have already endeavoured to refute,―have led many authors of the highest name, in some of the most important arguments which can employ human reason, to overlook that irresistible evidence which was placed before their eyes, in search of another mode of proof altogether unattainable in moral inquiries, and which, if it could be attained, would not be less liable to the cavils of sceptics.

But although, in philosophical language, the epithet probable be applied to events which are acknowledged to be certain, it is also applied to those events which are called probable by the vulgar. The philosophical meaning of the word, therefore, is more comprehensive than the popular; the former denoting that particular species Bet: 4 vulgar of evidence of what contingent truths admit; the latter being confined to such degrees of this evidence as fall short of the highest. These different degrees of probability the philosopher considers as a probable. scries, beginning with bare possibility, and terminating in that apprehended infallibility, with which the phrase moral certainty is synonymous. To this last term of the series, the word probable is, in its ordinary acceptation, plainly inapplicable.

7.

+ phit.

of

use

The satisfaction which the astronomer derives from the exact coincidence, in point of time, between his theoretical predictions concerning the phenomena of the heavens, and the corresponding Souver of the events when they actually occur, does not imply the smallest doubt, on his part, of the constancy of the laws of nature. It resolves satisfaction partly into the pleasure of arriving at the knowledge of the same felt by y as truth or of the same fact by different media; but, chiefly, into the tronomer at gratifying assurance which he thus receives, of the correctness of his principles, and of the competency of the human faculties to these success of sublime investigations. What exquisite delight must La Place have his calcula-felt, when, by deducing from the theory of gravitation, the cause of tions

the acceleration of the moon's mean motion,-an acceleration which proceeds at the rate of little more than 11' in a century,—he accounted, with such mathematical precision, for all the recorded ob