The discussion of plateaus seems at times to reduce itself to a definition of the term. According to some, only those delays which can be measured by weeks are plateaus. Yet an arrest of visible progress during a number of days demands an explanation quite as much as retardation extending through a month. If the causé cannot be found in the physical condition of the beginner, it must be sought in the complexity of the activity, in the aptness of the learner for the work, or in both. Now, periods of retardation of longer or shorter duration are usually observable in most learning processes. They do not occur when the act of skill is so simple as to be at once mastered, and their location depends in large measure upon the nature of the work in which the learner is engaged. In beginning the study of psychology, for example, a long plateau quite commonly occurs about the end of the third or fourth week. In embryology, on the other hand, it is usually observed at the beginning of the course. The difference between these two subjects of study suggests a possible 1 These two curves may be seen in the author's Learning by Doing, pp. 106, 108.

explanation of these plateaus. In introductory psychology the first week or two is generally devoted to discussing a few fundamental facts. The new terms are not numerous. The effort of the teacher is devoted rather to laying the foundation by establishing the simpler principles. Consequently, there is little chance for confusion. In beginning embryology, however, a large number of new words and terms must be mastered at once. This naturally perplexes the student. His progress is therefore slower at the outset than in psychology. He drops almost immediately to a low level of efficiency which reveals itself in his curve of learning as a plateau. What, then, is the cause of this retardation?

The present writer, from his study of the learning process in typewriting and in beginning the Russian language,1 concluded that these periods of arrest in progress are caused by the need of time for making associations automatic. In typewriting, for example, these associations consist in connections between the letters of words and the location of the corresponding keys of the typewriter. The learner, starting at the zero stage, advances rapidly at first. This initial rapid advance has been found true of all acts of skill and of subjects of study, in which a large mass of new material is not crowded upon the learner at the outset. The writer's investigation of the early stages in learning the Russian language, the curve of which is given below, shows the same characteristic. This curve of learning also indicates the effect of uninstructed learning. It is much more irregular than would have been the case had the learner been aided by a teacher.2

Returning to typewriting, it is clear that after a certain very moderate degree of proficiency has been attained

1 Psychological Bulletin, vol. 1, p. 295, vol. 7, p. 149; Studies in Philosophy and Psychology (Garman Commemorative Volume), p. 297.

"Those who wish more detailed information of the psychology of beginning a language may consult the author's Mind in the Making, p. 199.

progress will be slower and gains will be roughly proportional to the strength of the connections between letters of words and their corresponding keys. Establishing these connections requires practice, and for this practice time is

needed. There are, of course, different degrees of proficiency, depending upon the firmness of the connection between letters and keys, and upon the rapidity with which the associations work. Quite soon the hand goes directly to the vicinity of the desired key, but a few seconds will be required to find the right one. Next the finger, directed by the eye, may strike the key at once, and finally the eye is unnecessary, the sense of location being a sufficient guide.

This is a very sketchy description of the stages of learning to use a typewriter, but it answers the purpose. There are different levels of efficiency, each of which must, so to speak, be consolidated before it is possible to rise to the next level. This consolidation requires practice, the amount needed depending upon the number of details to be mastered and upon their difficulty. In the case of certain individuals especially adapted to a given piece of work, progress may be continuous to a high level of efficiency. There are children, for example, to whom the multiplication tables offer no difficulties. They learn them easily and accurately. Young salesmen may also be found, occasionally, who do not need to learn the art of selling goods. They know how to deal with men and can meet each emergency as it arises. But such children and such men are rare. The majority learn with effort, and from time to time they must stop to organize and reorganize what they have learned.

The periods of arrest in progress-the so-called plateaus of the curve of learning-are, then, intervals for consolidating the information or skill won by the learner during his advance, but not well organized because he was too busy making gains. This cessation of progress, however, is only apparent. It is not a real arrest, for during this time the facts and information are being automatized for ready use. Progress seems to be delayed, because it cannot be measured and recorded so as to be visible to the eye. Loosely accumulated information is only partially

usable. It comes to mind slowly and some of it does not come at all until it has been relearned. More than this, the meaning of facts grows as their connection with other facts is observed, and these new relationships give added significance to what one has learned, and make it more serviceable in one's thinking. This requires time-time for the nervous processes underlying the learning to become set and time for new nervous connections which are the basis of recalling ideas to be established and fixed.

In the early stages of learning the real advance is therefore made during the periods of seeming arrest of progress, because it is then that the consolidation of details is going forward. In acts of skill these details are the finer movements, together with the judgment of the amount, and direction of muscular effort, and in subjects of study they include facts and ideas which must be both learned and organized. Progress during these plateau-periods assumes, of course, that the learner is a zealous worker. This explanation of these periods of apparent retardation is illustrated by the experience of those who have spent several years in a foreign country trying to learn the language. The first rapid, though slight, advance is attended with a feeling of elation. This, however, is followed by a long period of discouragement. The best endeavors seem to bring no progress. The length of this plateau-period varies with different persons-with their aptitude and preparation for the language-but all feel its oppressiveness. Now, an interesting feature of this plateau is the suddenness with which it often disappears. At times it seems to vanish in a night. The word-associations and peculiarities of thought-sequence had been automatized during the long period when no visible progress was being made. Beginners in golf, again, are only too familiar with the long period of discouragement when they seem to be making no advance. Perseverance finally brings results, but for a long time improvement is intermittent.