and consistency not to be found in any other writer of the century. From Hume to J. S. Mill, the doctrine received no substantial alteration. It was Hume's aim to state the principles of morality in such a way as to bring it entirely within the domain of science. Granting the truth of his theories, he succeeded admirably. The only object of reasoning,' he says (that is, of ethical reasoning), 'is to discover the circumstances on both sides which are common to these' (the estimable or blamable) 'qualities, to observe that particular in which the estimable qualities agree on the one hand, and the blamable on the other; and thence to reach the foundation of ethics, and find those universal principles from which all censure or approbation is ultimately derived. As this is a question of fact, not of abstract science, we can only expect success by following the experimental method, and deducing general maxims from a comparison of particular instances.' 1 The science of morality, then, is to be based on experience. Hume succeeded so far as he definitely and systematically admitted this appeal. He failed in so far as, from his standing-point, it was impossible to form an adequate conception of the method by which the appeal should be made.

1

93. This method of approaching the problem implies the dismissal of all ontological and teleological speculation. Clarke's method of deducing morality from the intuitions of pure reason must be abandoned along with Butler's method. of discovering morality by divining the purposes of the Creator. Hume's objections to the first method are radical.2 Reason by itself cannot prompt us to act. It can make us aware that an object which excites our passions does or does not exist, or it can show that the means by which we would gratify our passions are or are not adequate. But it is not by itself a motive. "Tis not contrary to reason,' he says, 'to prefer the destruction of the whole world to the scratching of my finger. 'Tis not contrary to reason for me to choose my total ruin to prevent the least uneasiness of an Indian or person totally unknown to me.'3 Strictly speaking, there is no such thing as a combat between reason and passion.

1 Hume's Works, iv. 174.

* Treatise of Human Nature,' book ii. part iii. sec. 3; book iii. part iii. sec. I; appendix i. to ‘Enquiry.' 3 Works, ii. 195.

'Reason is and ought only to be the slave of the passions, and can never pretend to any other office than to serve and obey them.'' The phraseology is wantonly paradoxical in sound, because in his early treatise Hume aimed at being paradoxical. But it expresses the view which would be taken in sober seriousness by all scientific reasoners. The reason is the faculty which enables us to frame a mental picture of the world corresponding to the external reality. It would show that the total suffering caused by the destruction of the world was greater than the suffering caused by scratching my finger. But unless I were benevolent enough to feel for others, the bare fact would not impel me to scratch my finger to save the world, any more than the knowledge that a guinea was worth one-and-twenty shillings would make me prefer a guinea to a shilling if I had no love of money. If I was malevolent instead of benevolent, it might have the contrary effect. Hence all the reasonings of Clarke's school about the eternal and inherent essences of things are thrown away. If sound, they might reveal to us certain truths, but the mode in which those truths affected us would still be a question of experience. These moralists fill the gap in their system, as Hume points out, by suddenly substituting for the copula 'is' or 'is not' the copula 'ought' or 'ought not.' The reason may regulate and guide the passions by enabling us to compare their objects. It cannot supply the place of the passions.

94. The distinction thus drawn between the reason and the passions raises the most difficult of psychological problems. The connection between the emotions and the intellect is indefinitely intricate. Every mental process has its emotional and its intellectual side. It is impossible, therefore, to describe the fully developed structure of the mind without taking into account a whole series of complex actions and reactions between the two factors. And, for this reason, Hume's psychology, set forth in the second book of the Treatise, is the least satisfactory part of his work, as it was that which was most ruthlessly cut down in the Essays. Only a mangled remnant reappears as the brief Treatise on the Passions,' and ends.

1 Hume's Works, ii. 195.

2 Ib. 245.

3 Hume partly recognises this truth in the section 'Malice and Envy' of the Treatise, ii. 159.

abruptly with a half apology. Hume's attempt, indeed, was hopeless. The older philosophy had resolved feelings into beliefs. The passion, pride, for example, was identified with the conviction I am better than my neighbours.' Hume at once accepts a classification founded on this conception, and tries to get rid of the intellectual element implied. The attempt is contradictory. Pride, if pride be an elementary passion, must imply, at least, the intellectual processes necessary to frame some consciousness of myself and neighbours. Hume's effort to evade this conclusion is, at best, a display of wasted ingenuity. Human nature is compounded of too many elements, too intricately blended, for any offhand guesses of the cleverest philosopher to be of much value.

95. But the fact that Hume was not, and could not be, a scientific psychologist, does not destroy the value of his critical assault upon the ontologists. To confute the school of Clarke, little more was required than to show that ethics was not a branch of pure mathematics; for the truth and reality which they ascribed to morality were, on their showing, to be found in the mathematical world alone. Hume's criticism has a wider application. Morality, he says, in substance, cannot be deduced from absolute a priori truths, for it includes an empirical element. This follows from the fact that, if two men (or two races) shared the same intellectual convictions, the actions which resulted would vary according to their emotional compositions. The same truth which to the angelical nature would supply a motive for doing good, would supply to the diabolical nature a motive for doing evil. Hume, for this reason, compares the moral aspects of an action to the 'secondary' qualities. An action is seen as coloured by our emotions as the external world is known, and can only be known as it affects our senses. From the point of view of the earlier philosophy, this was to admit the unreality of vice and virtue, or, in a different phraseology, it would prove vice and virtue to be subjective.'

96. Hume's view of the passions as entirely independent of the intellect, and associated with certain objects by a tie in sense arbitrary, as indeed every causal tie is with

some

Hume's Works, ii. 245.

Hume arbitrary, might seem to sanction this conclusion. If our likes and dislikes might be indefinitely altered or inverted, there could be no science of human conduct. In fact, however, Hume's aim is precisely to discover such a science, but to prove simultaneously that it must be a science of observation. The passions, he says, form a 'regular mechanism,' which is as susceptible of scientific investigation as any branch of natural philosophy. Thus his argument virtually comes to the statement that a scientific morality would imply a psychology, and that psychology must be based upon experience alone. The relation is the same as that between sanitary and physiological science. The laws of moral as of physical health depend upon the structure of the organism, and the nature of that structure is only discoverable through the ordinary methods of scientific investigation. In this sense morality must include an empirical element, unless it be maintained that an a priori deduction of psychology is possible. The assumption of the possibility, to say nothing of the actual performance of such a deduction, depends upon the resolution of the passions into intellectual perceptions. If the passions are in some sense reason, there is some plausibility in attempting to frame an a priori scheme of psychological truths parallel to the so-called a priori scheme of mathematical truths. In that case, again, and in that case alone, morality would be in a sense capable of a priori deduction. We could not, indeed, even in that case, justify the identification of virtue and vice with truth and falsehood, or reason and error, implied in Clarke's substitution of ought' for 'is,' for that would be to show that båd actions were impossible as well as unusual, or to identify moral with scientific laws. But we might show that certain actions had always certain qualities or tendencies, which justified the moral distinction. That is to say, we might find an a priori justification for the utilitarian or 'moral sense' theories.

97. Meanwhile Hume is justified in declaring that morality must be based on experience if psychology be based on experience. We should amend his statement by adding that a complete science of morality would imply a science of sociology as well as of psychology, and requires a wider

1 Last sentence of Treatise on the Passions.'

and more systematic interrogation of experience than he had fully contemplated. There must be not only an empirical, but a variable, element in morality; and this is enough to condemn the hypothesis of Clarke. A scheme of morality deduced from self-evident and necessary truths must produce a code as rigid as its fundamental axioms, and, therefore, incapable of varying with the development of the race. Morality, on the other hand, includes in its primary data an element which varies, though, of course, varies according to definite laws. It must, therefore, give rules varying as the subject-matter varies; just as sanitary science gives one set of rules for men and another for beasts, and prescribes different conduct to a negro and a European. Hume did not fully appreciate this view, because, accepting from the ontologists the doctrine that human nature is always the same, he contemplated only a variation of external circumstances. As he, like all his contemporaries, failed to make allowance for the slow evolution of new social and intellectual conditions, the observed inconsistencies of the ethical code seemed to imply an almost indefinite variability of the moral sense.

98. If this be the true view of the relation between ethics on the one hand, and the sciences of psychology and sociology on the other, and if again, as is perfectly clear, no scientific psychology or sociology existed (even if they now exist) till long after the foundation of morality, one of two results must follow. Either the moral law is revealed by an instinct or inspired faculty, which can act independently of reason, or morality must be an empirical science; that is to say, it must have been discovered like other truths-by a series of experiments. As sanitary rules preceded physiology, ethical rules have preceded psychology. Was the moral law known by revelation, or by a special faculty, or was it explicable by some admitted and normal faculties of human nature? Hume's object is to answer this question by showing the possibility of the last alternative. The ground was already prepared. Innumerable moralists had proved that virtue produced happiness. Hutcheson, with whom Hume corresponded, had agreed, as we have seen, that the test of the morality was its tendency to produce happiness. The one necessary step was to get rid of the teleological view, and