To this, however, it has been answered, that in estimating pleasure there is no conceivable appeal from the immediate decision of consciousness: that here the Phenomenal is the Real--there is no other real that we can distinguish from it. And this seems to me true, in so far as we are concerned only with the present state. But then--apart from the difficulty just noticed of observing a pleasure while it is felt without thereby diminishing it--it is obvious that in any estimate of its intensity we are necessarily comparing it with some other state. And this latter must generally be a representation, not an actual feeling: for though we can sometimes experience two or perhaps more pleasures at once, we are rarely in such cases able to compare them satisfactorily: for either the causes of the two mutually interfere, so that neither reaches its normal degree of intensity; or, more often, the two blend into one state of pleasant consciousness the elements of which we cannot estimate separately. But if it is therefore inevitable that one term at least in our comparison should be an imagined pleasure, we see that there is a possibility of error in any such comparison; for the imagined feeling may not adequately represent the pleasantness of the corresponding actual feeling. And in the egoistic comparison, the validity of which we are now discussing, the objects primarily to be compared are all represented elements of consciousness: for we are desiring to choose between two or more possible courses of conduct, and therefore to forecast future feelings.
Let us then examine more closely the manner in which this comparison is ordinarily performed, that we may see what positive grounds we have for mistrusting it.
In estimating for practical purposes the value of different pleasures open to us, we commonly trust most to our prospective imagination: we project ourselves into the future, and imagine what such and such a pleasure will amount to under hypothetical conditions. This imagination, so far as it involves conscious inference, seems to be chiefly determined by our own experience of past pleasures, which are usually recalled generically, or in large aggregates, though sometimes particular instances of important single pleasures occur to us as definitely remembered: but partly, too, we are influenced by the experience of others sympathetically appropriated: and here again we sometimes definitely refer to particular experiences which have been communicated to us by individuals, and sometimes to the traditional generalisations which are thought to represent the common experience of mankind.