For the feeling to be cognitive in the specific sense, then, it must be self-transcendent; and we must prevail upon the god to create A reality outside of it to correspond to its intrinsic quality Q. Thus only can it be redeemed from the condition of being a solipsism.  If now the new created reality resemble the feeling’s quality Q I say that the feeling may be held by us to be cognizant of that reality.

This first instalment of my thesis is sure to be attacked.  But one word before defending it ‘Reality’ has become our warrant for calling a feeling cognitive; but what becomes our warrant for calling anything reality?  The only reply is—­the faith of the present critic or inquirer.  At every moment of his life he finds himself subject to a belief in some realities, even though his realities of this year should prove to be his illusions of the next.  Whenever he finds that the feeling he is studying contemplates what he himself regards as a reality, he must of course admit the feeling itself to be truly cognitive.  We are ourselves the critics here; and we shall find our burden much lightened by being allowed to take reality in this relative and provisional way.  Every science must make some assumptions.  Erkenntnisstheoretiker are but fallible mortals.  When they study the function of cognition, they do it by means of the same function in themselves.  And knowing that the fountain cannot go higher than its source, we should promptly confess that our results in this field are affected by our own liability to err.  The most we can claim is, that what we say about cognition may be counted as true as what we say about anything else.  If our hearers agree with us about what are to be held ‘realities,’ they will perhaps also agree to the reality of our doctrine of the way in which they are known.  We cannot ask for more.

Our terminology shall follow the spirit of these remarks.  We will deny the function of knowledge to any feeling whose quality or content we do not ourselves believe to exist outside of that feeling as well as in it.  We may call such a feeling a dream if we like; we shall have to see later whether we can call it a fiction or an error.

To revert now to our thesis.  Some persons will immediately cry out, ‘How can a reality resemble a feeling?’ Here we find how wise we were to name the quality of the feeling by an algebraic letter Q. We flank the whole difficulty of resemblance between an inner state and an outward reality, by leaving it free to any one to postulate as the reality whatever sort of thing he thinks can resemble a feeling,—­if not an outward thing, then another feeling like the first one,—­the mere feeling Q in the critic’s mind for example.  Evading thus this objection, we turn to another which is sure to be urged.