190 PSYCHOLOGY. ing of things in general we are notoriously fallible, why not also here? Comte is quite right in laying stress on the fact that a feeling, to be named, judged, or perceived, must be already past. No subjective state, whilst present, is its own object; its object is always something else. There are, it is true, cases in which we appear to be naming our present feeling, and so to be experiencing and observing the same inner fact at a single stroke, as when we say ' I feel tired,' ' I am angry,' etc. But these are illusory, and a little attention unmasks the illusion. The present con- scious state, when I say * I feel tired,' is not the direct state of tire ; when I say ' I feel angry,' it is not the direct state of anger. It is the state of saying-I-feel-tired, of say ing-I-f eel-angry, — entirely different matters, so different that the fatigue and anger apparently included in them are considerable modifications of the fatigue and anger directly felt the previous instant. The act of naming them has momentarily detracted from their force.* The only sound grounds on which the infallible veracity of the introspective judgment might be maintained are empirical. If we had reason to think it has never yet deceived us, we might continue to trust it. This is the ground actually maintained by Herr Mohr. " The illusions of our senses," says this author, "have undermined our belief in the reality of the outer world; but in the sphere of inner observation our confidence is intact, for we have never found ourselves to be in error about the reality of an act of thought or feeling. We Second, internal observation is better fitted to grasp clearly conscious states, especially voluntary mental acts: such inner processes as are ob- scurely conscious and involuntary will almost entirely elude it, because the effort to observe interferes with them, and because they seldom abide in memory." (Logik, ii. 432.)
- In cases like this, where the state outlasts the act of naming it, exists
before it, and recurs when it is past, we probably run little practical risk of error when we talk as if the state knew itself. The state of feeling and the state of naming the feeling are continuous, and the infallibility of such prompt introspective judgments is probably great. But even here the certainty of our knowledge ought not to be argued on the a priori ground that percipi a.nd esse are in psychology the same. The states are really two; the naming state and the named state are apart; 'percipi isesse' is not the principle that applies.
