The rate of reading, for example, is a measure of the rate of association. Letters become associated together in certain combinations making words, words into word groups and sentences. Recognition is for the most part an associative process. Rapid and accurate association will mean ready recognition of the printed form. Since language units (whether letters, words, or word groups) have more or less preferred associations according to their habitual arrangement into larger units, it comes about that in the normal mind under normal conditions these preferred sequences arouse the apperceptive complex necessary to make a running recognition rapid and easy. It is reasonable to suppose that in the subnormal mind the habitual common associations are less firmly fixed, thus diminishing the effectiveness of the ever-changing apperceptive expectancy. Reading is, therefore, largely dependent on what James calls the "fringe of consciousness" and the "consciousness of meaning." In reading connected matter, every unit is big with a mass of tendencies. The smaller and more isolated the unit, the greater is the number of possibilities. Every added unit acts as a modifier limiting the number of tendencies, until we have finally, in case of a large mental unit, a fairly manageable whole. When the most logical and suitable of these associations arise easily from subconsciousness to consciousness, recognition is made easy, and their doing so will depend on whether the habitual relations of the elements have left permanent traces in the mind. The reading of the subnormal subject bears a close analogy to the reading of nonsense matter by the normal person. It has been ascertained by experiment that such reading requires about twice as much time as the reading of connected matter. This is true for the reason that out of thousands of associations possible with each word, no particular association is favored. The apperceptive expectancy, practically _nil_ in the reading of nonsense material, must be decidedly deficient in all poor reading. Furthermore, in the case of the ordinary reader there is a feeling of rightness or wrongness about the thought sequences. That less intelligent subjects have this sense of fitness to a much less degree is evidenced by their passing over words so mutilated in pronunciation as to deprive them of all meaning. The transposition of letters and words, and the failure to observe marks of punctuation, point to the same thing. In other words, all the reading of the stupid subject is with material which to him is more or less nonsensical.[66] [66] See "Genius and Stupidity," by Lewis M. Terman, in _Pedagogical Seminary_, September, 1906, p. 340 _ff._