this same self-love. "Moderation is a fear of falling under the envy and contempt which those deserve who are intoxicated by their good fortune." 1 "Pride plays a larger part than good will in the remonstrances we address to those who have committed faults." 2 "There is no passion in which selflove reigns so powerfully as in sexual love." 3

...

No one has defined with equal subtlety and profundity the self-regarding sentiment. "L'amour-propre est l'amour de soi-même et de toutes choses pour soi; il rend les hommes idolâtres d'eux-mêmes, et les rendroit les tyrans des autres, si la fortune leur en donnoit les moyens il ne se repose jamais hors de soi, et ne s'arrête dans les sujets étrangers que comme les abeilles sur les fleurs, pour en tirer ce qui lui est propre. . . . On ne peut sonder la profondeur ni percer les ténèbres de ses abîmes. . . . Là, il est souvent invisible à lui-même . . . mais cette obscurité épaisse qui le cache à lui-même, n'empêche pas qu'il ne voie parfaitement ce qui est hors de lui; en quoi il est semblable à nos yeux, qui découvrent tout, et sont aveugles seulement pour euxmêmes. . . . Il est tous les contraires, il est impérieux et obéissant, sincère et dissimulé, miséricordieux et cruel, timide et audacieux: il a de différentes inclinations, selon la diversité des tempéraments qui le tournent et le dévouent tantôt à la gloire, tantôt aux richesses, et tantôt aux plaisirs. Il en change selon le changement de nos âges, de nos fortunes et de nos expériences. . . . Il ne faut donc pas s'étonner s'il se joint quelquefois à la plus rude austérité, et s'il entre si hardiment en société avec elle pour se détruire, parce que, dans le même temps qu'il se ruine en un endroit, il se rétablit en un autre. ... Voilà la peinture de l'amour propre, dont toute la vie n'est qu'une grande agitation." +

That, however, everything cannot be reduced to self-love, wide as is its diffusion, that it is in constant interaction with other and opposite systems in our character, is evident from its own self-deceptions, self-justifications, and frequent 1 'Reflexions, Sentences et Maximes Morales de La Rochefoucauld,' xviii.

2 Op. cit. xxxvii.

Op. cit. i. (de l'édition de 1665).

3

Op. cit. cclxii.

assumptions of virtue, and from this disinterested curiosity But the laws of this interaction, if implied, are not elicited. In some cases, however, La Rochefoucauld enunciates genuine, empirical laws of character:-" Weak characters cannot be sincere"; 1 "Envy is destroyed by true friendship, and coquetry by true love."2 Again he observes profoundly: "It is only those who have firmness (of will) who can possess true sweetness (of character); those who seem sweet are generally only weak, which is easily turned to sourness."3

The result of an examination of the other great writers of reflections and maxims would bring us to a conclusion not substantially different from that which we have already reached. While we may find in the works of Bacon, Pascal, La Bruyère, Vauvenargues, Joubert, Goethe, an occasional law of character expressed, and many more implied, the bulk of their wisdom is not concerned to formulate such laws.

We must then judge, with regard to the inductive base of the new science, that Mill did not form an accurate opinion of the nature of the material that we could extract from literature, and with no more particular direction than to collect those "empirical laws" that "have been formed in abundance by every successive age of humanity," we shall make little progress, and that the reward is likely to be so disproportioned to our pains that we shall abandon the search in disgust. On the other side, as we have seen, in common with other psychologists of his time, Mill was mistaken as to the functions that could be assigned to the laws of association; and therefore was without any of those "laws of mind" to serve as deductive and explanatory principles, which were indispensable to the working of his method.

Notwithstanding the fact that the generalised observations of the great writers who have understood human nature most profoundly are not couched in a form which we can at once assimilate, and include in a collection of empirical laws of character, they are not without a special value and significance for us. The material which they supply is indisOp. cit. ccclxxvi.

1

Op. cit. cccxvi.

3 Op. cit. cccclxxix.

2

pensable to the inductive basis of the science, and may both help us in discerning laws that are implied in them, and in testing or verifying the laws that we have already discovered.

We have now to inquire in what other directions we have to modify and supplement Mill's method, in order to fit it for our purpose.

G

CHAPTER VIII

OF THE METHOD OF A SCIENCE OF CHARACTER (II)

4. Of certain Qualifications which the Method of a Science of Character should possess

THE chief deficiency of Mill's method is the omission to make any sufficient analysis of character itself, the most difficult and complicated fact that we have anywhere to deal with. He does not appear to have appreciated how abstract and inadequate are the ordinary conceptions of it, or how this abstractness and inadequacy must infect the problems of the new science. Yet the attainment of all difficult ends, setting aside those of our instincts, is dependent on some preconception of them, and the clearness and adequacy of this pre-conception conditions the choice of the means, and, therefore, the chances of success. A confused conception has its inevitable consequence in confused actions, drawn continually from their course because we do not know what that course is. It is for this reason that we have judged it essential to inquire, in the first place, into the constitution of human character, in order that we may form a clear conception of it, and through this clear conception, be able to discern the problems and the goal of the new science.

The conception of character common to those who succeeded Mill in attempting to treat it scientifically, was formed under the influence of analytical psychology, and this influence fastened attention on the three abstract elements-feeling, conation, cognition-assumed to be united in every state of consciousness, so that the conception of character reached along these lines was much the same as that of mind in

general. The problem of a science of character suggested by this conception, will be, therefore, to understand how these three elements,' called also 'aspects,' and sometimes functions,' are related to one another in the different characters of men. Thus arises the conception of a predominance of one or other of them as furnishing a key to the classification of characters. If to this conception we add the account of the four temperaments that has come down to us from Hypocrates and Galen, a second problem is suggested of the innate character as constituted by the predominance of one or other of these temperaments, the nervous, the phlegmatic, the bilious, and the sanguine. And thus we become at length involved in a third and more intricate problem: how the predominance of one or other of these types of temperament is related to the predominance of one or other of the ultimate elements or functions in the same character. Several works 1 have been written along these lines; and as long as they are followed, and the methods of analytical psychology are adhered to, it does not seem possible greatly to vary the conception of character and its problems. We are inevitably committed to quantitative estimates of character-to conceptions of more and less, which cannot be treated scientifically, so long as we cannot measure them, and which leave us with indefinite conclusions. We are taken away from concrete and fruitful problems to follow others which are abstract and even artificial, such as is that of the 'predominance' of one of the fundamental aspects of mind over another from which it is inseparable; and, meanwhile, pre-occupied with them, we lose sight of the concrete facts, and the power of handling them.

2

If, on the other hand, we start from a concrete and synthetic conception of character, we are in harmony with the point of view of all dramatists, historians, biographers, and novelists,

1 'Study of Character,' by Alexander Bain; A. Fouillée, 'Tempérament et Caractère.' P. Malapert, 'Les Eléments du Caractère.' A. Lévy, 'Psychologie du Caractère.' F. Queyrat, 'Les Caractères.'

2 I have considered the meaning which must be attached to such a 'predominance' in a review of M. Fouillée's work, 'Tempérament et Caractère.' See 'Mind,' N. S. vol. v. p. 125 (1896).