Affection (Lat. ad, and facere, to do something to, sc. a person), literally, a mental state resulting generally from an external influence. It is popularly used of a relation between persons amounting to more than goodwill or friendship. By ethical writers the word has been used generally of distinct states of feeling, both lasting and spasmodic; some contrast it with passion as being free from the distinctively sensual element. More specifically the word has been restricted to emotional states which are in relation to persons. In the former sense, it is the Gr. pathos, and as such it appears in Descartes and most of the early British ethical writers. On various grounds, however---e.g. that it does not involve anxiety or excitement, that it is comparatively inert and compatible with the entire absence of the sensuous element--At is generally and usefully distinguished from passion. In this narrower sense the word has played a great part in ethical systems, which have spoken of the social or parental affections as in some sense a part of moral obligation. For a consideration of these and similar problems, which depend ultimately on the degree in which the affections are regarded as voluntary, see H. Sidgwick, Methods of Ethics, pp. 345-349.
In psychology the terms affection and affective are of great importance. As all intellectual phenomena have by experimentalists been reduced to sensation, so all emotion has been and is regarded as reducible to simple mental affection, the element of which all emotional manifestations are ultimately composed. The nature of this element is a problem which has been provisionally, but not conclusively, solved by many psychologists; the method is necessarily experimental, and all experiments on feeling are peculiarly difficult. The solutions proposed are two. In the first, all affection phenomena are primarily divisible into those which are pleasurable and those which are the reverse. The main objections to this are that it does not explain the infinite variety of phenomena, and that it disregards the distinction which most philosophers admit between higher and lower pleasures. The second solution is that every sensation has its specific affective quality, though by reason of the poverty of language many of these have no name. W. Wundt, Outlines of Psychology (trans. C. H. Judd, Leipzig, 1897), maintains that we may group under three main affective directions, each with its negative, all the infinite varieties in question; these are (a) pleasure, or rather pleasantness, and the reverse, (b) tension and relaxation, (c) excitement and depression. These two views are antithetic and no solution has been discovered.
Two obvious methods of experiment have been tried. The first, introduced by A. Mosso, the Italian psychologist, consists in recording the physical phenomena which are observed to accompany modifications of the affective consciousness. Thus it is found that the action of the heart is accelerated by pleasant, and retarded by unpleasant, stimuli; again, changes of weight and volume are found to accompany modifications of affection--and so on. Apart altogether from the facts that this investigation is still in its infancy and that the conditions of experiment are insufficiently understood, its ultimate success is rendered highly problematical by the essential fact that real scientific results can be achieved only by data recorded in connexion with a perfectly normal subject; a conscious or interested subject introduces variable factors which are probably incalculable.
The second is Fechner's method; it consists of recording the changes in feeling-tone produced in a subject by bringing him in contact with a series of conditions, objects or stimuli graduated according to a scientific plan and presented singly in pairs or in groups. The result is a comparative table of likes and dislikes.
Mention should also be made of a third method which has hardly yet been tried, namely, that of endeavouring to isolate one of the three directions by the method of suggestion or even hypnotic trance observations.
For the subject of emotion in general see modern text-books of psychology, e.g. those of J. Sully, W. James, G. T. Fechner, O. Kulpe; Angelo Mosso, La Paura (Milan, 1884, 1900; Eng. trans. E. Lough and F. Kiesow, Lond. 1896); E. B. Titchener, Experimental Psychology (1905); art. "Psychology" and works there quoted.
See also Affectional orientation -- affective filter