Save in extremely rare cases,—supposing such to occur at all, as perhaps in the instant of surprise, and in states approximating to pure sensation,—save in such extremely rare cases, where the mind, like a mirror, passively reflects external impressions, intellectual activity may always be reduced to one of the two following types: associating, combining, unifying, or dissociating, isolating, and separating. These cardinal operations underlie all forms of cognition, from the lowest to the highest, and constitute its unity of composition.
Abstraction belongs to the second type. It is a normal and necessary process of the mind, dependent on attention, i. e., on the limitation, willed or spontaneous, of the field of consciousness. The act of abstraction implies in its genesis both negative and positive conditions, and is the result of these.
The negative conditions consist essentially in the fact that we cannot apprehend more than one quality or one aspect, varying according to the circumstances, in any complex whole,—because consciousness, like the retina, is restricted to a narrow region of clear perception.
The positive condition is a state which has been appropriately termed a “psychical reinforcement” of that which is being abstracted, and it is naturally accompanied by a weakening of that which is abstracted from. The true characteristic of abstraction is this partial increment of intensity. While involving elimination, it is actually a positive mental process. The elements or qualities of a percept, or a representation, which we omit do not necessarily involve such suppression. We leave them out of account simply because they do not suit our ends for the moment, and are complementary.[2]
Abstraction being, then, in spite of negative appearances, a positive operation, how are we to conceive it? Attention is necessary to it, but it is more than attention. It is an augmentation of intensity, but it is more than an augmentation of intensity. Suppose a group of representations a + b + c = d. To abstract from b and c in favor of a, would ostensibly give a = d - (b + c). If this were so, b and c would be retained unaltered in consciousness; there would be no abstraction. On the other hand, since it is impossible for the whole representation d to be suppressed outright, b and c cannot be totally obliterated. They subsist, accordingly, in a residual state which may be termed x, and the abstract representation is hence not a but a + x or A. Thus the elements of abstract representations are the same as those of concrete representations; only some are strengthened, others weakened: whence arise new groupings. Abstraction, accordingly, consists in the formation of new groups of representations which, while strengthening certain elements of the concrete representations, weaken other elements of the same.[3]
We see from the above that abstraction depends genetically upon the causes which awaken and sustain attention. I have described these causes elsewhere,[4] and cannot here return to their consideration.
It is sufficient to remark that abstraction, like attention, may be instinctive, spontaneous, and natural; or reflective, voluntary, and artificial. In the first category the abstraction of a quality or mode of existence originates in some attraction, or from utility; hence it is a common manifestation of intellectual life and is even met with, as we shall see, among many of the lower animals. In its second form, the rarer and more exalted, it proceeds less from the qualities of the object than from the will of the subject; it presupposes a choice, an elimination, of negligible elements, which is often laborious, as well as the difficult task of maintaining the abstract element clearly in consciousness. In fine, it is always a special application of the attention which, adapted as circumstances require to observation, synthesis, action, etc., here functions as an instrument of analysis.
A deeply-rooted prejudice asserts that abstraction is a mental act of relative infrequency. This fallacy obtains in current parlance, where “abstract” is a synonym of difficult, obscure, inaccessible. This is a psychological error resulting from an incomplete view: all abstraction is illegitimately reduced to its higher forms. The faculty of abstracting, from the lowest to the highest degrees, is constantly the same: its development is dependent on that of (general) intelligence and of language; but it exists in embryo even in those primitive operations which are properly concerned with the concrete, i. e., perception and representation. Several recent authors have emphasised this point.[5]
Perception is par excellence the faculty of cognising the concrete. It strives to embrace all the qualities of its object without completely succeeding, because it is held in check by an internal foe,—the natural tendency of the mind to simplify and to eliminate. The same horse, at a given moment, is not perceived in the same manner by a jockey, a veterinary surgeon, a painter, and a tyro. To each of these, certain qualities, which vary individually, stand in relief, and others recede into the background. Except in cases of methodical and prolonged investigation (where we have observation, and not perception) there is always an unconscious selection of some principal characteristics which, grouped together, become a substitute for totality. It must not be forgotten that perception is pre-eminently a practical operation, that its mainspring is interest or utility, and that in consequence we neglect—i. e., leave in the field of obscure consciousness—whatever at the moment concerns neither our desires nor our purposes. It would be superfluous to review all the forms of perception (visual, auditory, tactile, etc.), and to show that they are governed by this same law of utility; but it should be remarked that the natural mechanism by which the strengthened elements and the weakened elements are separated, is a rude cast of what subsequently becomes abstraction, that the same forces are in play, and are ultimately reducible to some definite direction given to the attention.