tion and fitness are expressive of ideas entirely simple, and do not consequently admit of definition. It might be supposed, that, if every action productive of happiness is fit to be done, it would be a correct definition of a fit action, to call it an action productive of happiness; but this is no more the case than it would be a correct definition of an equilateral triangle, to call it a triangle with three equal angles. That it has three equal angles is true of an equilateral triangle, but not what expresses our very notion of it - to which the office of a definition is confined. So, the fitness of a beneficent action is a truth perceivable in regard to this action; but can by no means be considered as merely expressing what we mean by its possessing such tendency. Obvious as the truth of this remark must appear, it is the neglect of it, the confounding of the benevolent quality of an action with the moral goodness of that quality, as one idea, that has led to one of the most material fallacies connected with the present subject.*

These observations, defining the particular kind of moral truths which, in the dispute regarding the nature of the moral faculty, reason is maintained to discover, will, it is hoped, still farther vindicate the instrumentality of that faculty, by confining such instrumentality within those precise limits beyond which its employment need not be contended for.

187

SECT. II.

Parallel between Moral and Mathematical Truths.

The views which I entertain as to the source of moral distinctions, naturally lead me to offer an opinion on the question, - whether morality is capable of demonstration, after the manner of mathematics.

If this question meant, whether there are any truths in morals that are necessary and immutable, as opposed to uncertain, contingent, or merely probable, I could of course have no hesitation in answering in the affirmative.

Or, if the question meant, whether there are any of the necessary truths in morals that are the subjects of deduction, as opposed to intuition; or, which comes to the same purpose, whether we can reason from necessary first principles of morals to other necessary truths, (which is the strict notion of demonstration,)-it seems to me that instances of such demonstration may very easily be shewn. But, in these, the conclusions are so little removed from their premises, that the statement of such demonstrations, like that of some of the elementary theorems of Euclid, seems little better than ingenious trifling.

And therefore, if the question means (as I imagine it ought to be understood) whether we can, by direct reasoning from necessary principles,

solve practical difficulties in morals; whether we are indebted to demonstration - to reasoning by progressive steps - for any conclusions in morals which lay hid until such progressive reasoning was employed; whether, in short, morals is a demonstrative science, in the sense that demonstration (in the strict import of the term) ever is, or will be an instrument in ordinary use for ascertaining its truths, - then I can have as little hesitation in saying that morals is not a demonstrative science; for all ordinary and practical difficulties in morals have reference to matter of fact-which, of course, is not subject of demonstration.

But still, the difference between the necessary truths of mathematics, and those of morals, in regard to the application of each to cases of fact, differ only in degree, not in kind.

The moralist tells us that an action which produces fit effects is obligatory. But how, it is asked, are we profited by knowing this, unless we are informed whether any particular action, about which a question may exist, does produce fit effects or not?

Well, but what more does the mathematician teach us, in regard to any particular point connected with his science? His information, so far as it is demonstrative, is really as hypothetical as that of the moralist. - Suppose there is a triangular field, on each of the sides of which is a four-sided field. The two smaller four-sided fields belong to me, the largest to a neighbour. I wish to exchange with him. Will a mathematician demonstrate that they are equal? Certainly not, any more than the moralist can demonstrate that the effects of any action are fit. He will tell us, - if the angle opposite to the largest field is a right angle, and if the three four-sided fields are each perfectly square, the field of the one proprietor will be equal to the two fields of the other : and we are left to ascertain, whether, in point of fact, these circumstances exist as supposed or not, just as the moralist leaves us to find out whether the effects of a given mode of conduct will be beneficial or not.

The moralist, then, does not demonstrate that a given action is right or wrong; but so neither does the mathematician demonstrate that certain fields are equal to one another. The one tells us if the action has such and such qualities it is right; the other, if the figures of the fields have such and such qualities, they are equal. In what then does the mathematical differ from the moral truth? It is solely in the degree in which the general hypothetical condition, supporting the particular truth, may be narrowed into a condition of a simpler kind.

The mathematician having found one condition on which a consequence may be established, straightway shews that condition to hold good wherever another exists; that other, where a third; this third, where a fourth; and this fourth, perhaps, where a fifth exists: and the actual existence of this fifth condition, in any given practical instance, may be very easily ascertained, or much more easily than the first, or even the second, third, or fourth.*

Now the moralist just leaves us at a fartheroff stage, whence we have to grope our way with less ease and less certainty. He carries us, so far as he goes, as securely and as unerringly as the mathematician; but he leaves us sooner to our

selves.

That the mathematician can push his premises to so much remoter and more specific conclusions than the moralist, arises from the infinite number

* Thus, if, in answer to the inquiry about the equality of the fields, the mathematician were merely to tell us, that if my fields were equal to two other fields (which two were equal to my neighbour's,) our possessions would be equal, his information would be of little use to us; but we can suppose that this point might be more easily ascertained than that which we ultimately aim at knowing. If he were to tell us, next, that this condition would hold good, if the space FBC (see Euclid's 47 Prop. 1st Book) were equal to DBA, this might be still more easily found. But if he proceeded, farther, to inform us that this circumstance would depend on the parallelism of the lines BD, LA, FB, GC, and the equality of FB, B D, our task would still be easier. Lastly, if he said that this equality, and this parallelism, would certainly exist, if angle A were a right angle, and each of the three fields a square; this would be a point which we should probably be able to ascertain with great ease and certainty.