that are the basis of the recognition? The answer is: Very distinctly, recognition does affect the impressions. The activity involved in recognition alters the data of sense, and that in almost every case. Two of the ways in which this alteration occurs are these: (1.) In recognizing, we complete present data by remembered past data, and so seem to experience more than is actually given to our senses. Thus, then, in reading, we read over misprints (even against our own will), thinking that we see words when we do not see them, or when we see only parts of them. Again: in listening to an indistinct speaker we often supply what is lacking in the sounds he makes, and seem to hear whole words when we really hear but fragments of words. Or, merely whistling a few notes, we recall to ourselves, and seem to have present, the complex instrumental harmony of some music that we have heard played. Or, in dim twilight, we imagine the form of a man, and seem to see it plainly in detail, when, in fact, a mass of shrubbery, or a coat on a chair, is the one source of our impressions. In all these cases, the activity of recognition alters the data of sense, by adding to them, by fiUing out the sketch made by them. (2.) However, even the qualities of sense-impressions are altered according to the way in which we recognize their objects. The colors of a landscape are dimmer, and less significant as colors, so long as we recognize the objects in the landscape. Look under your arm, with head inverted, and the colors flash out with unwonted brilliancy. For when you so look, you lose sight of the objects as such, and
