This faculty or habit consists essentially in combining ideas (particular or general), or objects and ideas, so as to form systems different from those occurring in actual experience. The whole has never been perceived, though all its elements have been perceived.

Any association of ideas may be called imaginary if it occurs in an order different from the order of experience. But the term Imagination is properly confined to novel combinations deliberately and consciously formed to serve some utility. It is thus distinguished from Reverie, in which no choice or control enters into the recollection.

We control our ideal associations by means of comparison, which is therefore what distinguishes imagination from reverie. For instance, if I see a vase from which the handle has been broken off, I can imagine the handle restored, but to do this I must be able to compare the broken vase with a similar whole vase, or with the general idea 'whole vase.' The combination I form is novel, for I have never seen this particular vase in a whole state; if I had I should not be imagining it but recollecting it.

There are two principal distinctions to be noticed in imagination; one relates to the mode of forming the imaginary idea, the other to its use.

In the above case we form the whole by mechanical extension or addition. The process is as simple as nailing one piece of wood to another. But suppose the broken vase is of porcelain and the whole one of bronze, the restoration can still be made, but it is no mechanical junction of two previous ideas. It is a fusion of the material supplied by one idea with a form supplied by another. On the same principle a vase may be wholly designed from hints supplied by a score or more of vases, differing in material, in size, colour, decoration, and so forth. In these cases the new idea may be said to be totally different throughout its length from any other and from any object. Yet it is a combination of previous ideas. We do not create any absolutely new idea. This may be called imagination by transfusion. The elements may be so well mixed that it is impossible to trace each back to its origin.

Transfusion may be further complicated by recompounding ideas already compound. This occurs, as we shall see, in forming the 'external world' of materialists and realists.

The two uses to which imaginary ideas are put are the Artistic and the Rational.

We have seen (x) that emotions may be excited by objects or ideas. Hence, agreeable emotions may be excited by suggesting the objects associated with the original agreeable feelings; and novel emotions may be excited by novel combinations of the ideas of experienced objects that have been signs of feelings. From this possibility has arisen that extensive province of activity called Art, which consists in imagining novel combinations of things capable of exciting novel and pleasurable emotions (not feelings), and in finding means of suggesting such ideas to others. Some of these combinations are so subtle, and the emotions they excite so exquisite, that we value the artistic work at a great price, and rank the man who imagined it among the benefactors of his species.

Reason, or the Rational Imagination, does not appeal directly to the emotions. It serves the uses of life by enabling us to imagine what we have not yet experienced but may have to experience, and the quality aimed at is accuracy of intellectual ideation, not emotional pleasure. It is found by experience that an intellect well furnished with ideas may learn to combine them into pictures or preconceptions of the future, and the indirect utility of this accomplishment is very great. If it does not, like art, give immediate sentimental pleasure, it often enables us so to control events that we are brought into conditions affording more lasting satisfaction than many expensive works of art. Reason, then, is the imaginative faculty applied to the purpose of acquiring ideas of experience that has not yet taken place, and it is good in proportion to the similarity of the idea to the anticipated or unknown experience.

Although imagination is more important than generalisation, it has received little attention from metaphysicians. Their treatment of it is not uniform, but it generally exhibits two fundamental defects. They consider it an independent or ultimate faculty, that is, one incapable of resolution into anything more simple. We have seen that it is an application of comparison, and comparison depends on the coincidence of particular ideas.

Then they regard imagination only in its artistic uses, not perceiving that it is also the basis of reason. Reason they treat as generalisation—a vice that pervades all their systems. They put reason and art in essential opposition, whereas the difference between them is only specific—a difference of use.

Some metaphysicians confound imagination with mere recollection. 'It is,' says one of them, 'the faculty representative of the phenomena both of the external and internal worlds.' But there is a great difference between the representation of what we have experienced actually, and the representation of a future and perhaps impossible event: the latter only is imaginative. 'There is no train of ideas,' says another, 'to which the term imagination may not be applied.' If a man at the end of the day calls to mind all the events of the day in a train of ideas, that is recollection, and would be very inappropriately termed imagination. According to a third, imagination has for its object the concrete as opposed to abstractions and generalities. This also is inexact. A traveller may describe in general the qualities of a foreign country or tribe of men, and we shall imagine that generality without a concrete picture. The power of imagining generalities and abstractions necessarily follows from the power of forming them in the first instance.