in this method of reasoning, can hardly have been otherwise than sensible of that intellectual comprehension, or length and breadth of survey, which we have in view; since one demonstration is often connected with another, much in the same way as the subordinate parts of separate demonstrations are connected with each other; and he, therefore, finds it necessary, if he would go on with satisfaction and pleasure, to gather up and retain, in the grasp of his mind, all the general and subordinate propositions of a long treatise.

§. 20. Further considerations on the influence of demonstrative reasoning.

But, on the other hand, there are some results of a very great attention to sciences, which require the exclu sive application of demonstrative reasoning, of a less favourable kind.

(I) It has been thought among other things, that this form of reasoning, when carried to a great length, has a tendency to render the mind mechanical. That is, while it increases its ability of acting in a given way, it diminishes the power of invention, and prevents its striking out into a new path, different from that, which it has been in the habit of going over. And hence it is, that men of the strictest virtue and the most powerful intellect have sometimes discovered an unexpected weakness, and made extraordinary mistakes, when placed in certain new situations. We may illustrate our meaning by a single instance, although perhaps not one of the strongest kind. The celebrated Turgot, who combined the purest moral sentiments with the rarest intellectual endowments, was what may be termed a mathematical politician. History has recorded the result. When the king of France called him to direct the political concerns of the French empire, he decidedly failed, where half the talents and integrity had firmly held the helm amid political tempests. That great and virtuous mind, when called away from the abstractions of science to deal with the realities of life and mankind, which prejudice and passion, weakness and

power, interest and suffering presented before him, found too late, that we cannot estimate the intellect as we can estimate the arc of a circle, and that the calculus, which can measure the flight and eccentricities of the stars, may not succeed in ascertaining the momentum and the obliquities of human nature.*

(II) An exclusive culture of demonstrative reasoning nourishes a spirit of scepticism; or perhaps we may say, diminishes the power of belief. The exclusive mathematician has been accustomed to yield his assent to demonstration only; and it is but natural, that he should find some difficulty in being satisfied with any lower degree of evidence. This disposition to doubt will be, in some measure, experienced, even in the transition from pure to mixed mathematics; at least there will be an absence of that full and delighted satisfaction, which had hitherto been enjoyed. Still more will it be felt, when he is called upon to judge of events, and duties, and actions of common life, which do not admit of the application of demonstration. In a word, it has been supposed to unfit the mind in a considerable degree for accurate discriminations as to moral evidence on all subjects whatever, where that species of evidence is alone admissible; and also for fair and correct judgments in matters of taste.

Such, on the whole, being the result of an exclusive

Remarks to this effect, in respect to Turgot, have been frequently repeated; and that too by writers of different political sentiments. Take the following as an instance from no very friendly writer.

"Turgot was in youth what he was in age, grave, industrious, argumentative, and undecided;-a theorist, who could with difficulty descend to practice, and who passed his time out of office, and lost it when in, in a vain search after perfection, and in the Quixotic folly of attempting to subject human affairs to the precision of mathematical problems. The plausibility of reducing the art of administration to a system raised Turgot to office, and its impracticability drove him from it. He was generally right in his conception, but he did not know how to execute it; and he brought into the cabinet an immense stock of knowledge on every subject, except man." London Quarterly Review, Vol. XXVI, p. 231.

attention to sciences, which admit of demonstration alone, a restricted pursuit of them is all, that can be safely recommended. In making this remark, however, it is not meant, that we would absolutely set limits to the prosecution of them, but would only propose, that other modes of mental discipline should be prosecuted at the same time. Those, who aim at a perfect education, will not "canton out to themselves a little Goshen in the intellectual world," which is to receive all their labours, and leave the rest of the vast field of the mind to neglect, but will bestow a suitable share of culture on every part of it.

CHAPTER NINTH.

MORAL REASONING.

§. 21. Of the subjects and importance of moral reasoning.

MORAL REASONING, which is the second great division or kind of reasoning, concerns opinions, actions, events &c.; embracing in general those subjects, which do not come within the province of demonstrative reasoning. The subjects,to which it relates, are often briefly expressed by saying, that they are matters of fact; nor would this definition, concise as it is, be likely to give an erroneous idea of them.Skill in this kind of reasoning is of great use in the formation of opinions concerning the duties, and the general conduct of life. Some may be apt to think, that those, who have been most practised in demonstrative reasoning, can find no difficulty in adapting their intellectual habits to matters of mere probability. This opinion is not altogether well founded, as we have seen in the preceding chapter. Although that species of reasoning has a favourable result in giving persons a command over the attention, and in some other respects, whenever exclusively employed it has the effect in some degree to disqualify them for a correct judgment on those various subjects, which properly belong to moral reasoning. The last, therefore, which has its distinctive name from the primary signification of the Latin MORES, viz. manners, customs, &c. requires a separate consideration.

§. 22. Of the nature of moral certainty.

Moral reasoning causes in us different degrees of assent, and in this respect differs from demonstrative. In demonstration there is not only an immediate perception of the relation of the propositions compared together; but in consequence of their abstract and determinate nature, there is also a knowledge or absolute certainty of their agreement and disagreement. In moral reasoning the case is somewhat different.- -In both kinds we begin with certain propositions, which are either known or regarded as such. In both there is a series of propositions successively compared. But in moral reasoning, in consequence of the propositions not being abstract and fixed, and, therefore, often uncertain, the agreement or disagreement among them is in general not said to be known, but presumed; and this presumption may be more or less, admitting a great variety of degrees. While, therefore, one mode of reasoning is attended with knowledge; the other can properly be said to produce in most cases only judgment or opinion. But the probability of such judgment or opinion may sometimes arise so high, as to exclude all reasonable doubt. And hence we then speak, as if we possessed certainty in respect to subjects, which admit merely of the application of moral reasoning. Although it is possible, that there may be some difference between the belief attendant on demonstration, and that produced by the highest probability, the effect on our feelings is at any rate essentially the same. A man, who should doubt the existence of the cities of London and Pekin, although he has no other evidence of it than that of testimony, would be considered hardly less singular and unreasonable, than one, who might take it into his head to doubt of the propositions of Euclid. It is this very high degree of probability, which we term moral certainty.

§. 23. Of reasoning from analogy.

MORAL REASONING admits of some subordinate divisions; and of these, the first to be mentioned is reasoning from analogy. The word, analogy, is used with some vagueness,