As memorizing is one sort of learning, what we have found in the preceding chapter regarding the learning process should throw light on our present problem. We found animals to learn by doing, and man by doing and also by observation or observation combined with doing. Observation is itself a form of doing, a mental reaction as distinguished from a purely passive or receptive state; so that learning is always active. Observation we found to be of great assistance, both by way of hastening the learning process, and by way of making what is learned more available for future use. Our previous studies of learning thus lead us to inquire whether committing to memory may not consist partly in rehearsing what we wish to learn, and partly in observing it. Learning by rote, or by merely repeating a performance over and over again, is, indeed, a fact; and observant study is also a fact. Let us see how learning is actually done, as indicated by laboratory experiments. The psychologist experiments a great deal with the _memorizing of nonsense material_, because the process can be better observed here, from the beginning, than when sensible material is learned. Suppose a list of twenty one-place numbers is to be studied till it can be recited straight through. The learner may go at it simply by "doing", which means here by reading the list again and again, in the hope that it will finally stick. This pure rote learning will perhaps do the job, but it is slow and inefficient. Usually the learner goes to work in quite a different way. He observes various facts about the list. He notices what numbers occur at the beginning and end, and perhaps in other definite positions. He may group the digits into two-place or three-place numbers, and notice the characteristics of these. Any familiar combinations that {334} may occur, such as 1492, he is likely to spy and remember. Lacking these, he can at least find similar and contrasting number-groups. For example, the list 5 7 4 0 6 2 7 3 5 1 4 0 9 2 8 6 3 8 0 1, which at first sight seemed rather bare of anything characteristic, was analyzed in a way partly indicated by the commas and semicolons, 5, 74, 0; 62, 73; 5140; 9, 286; 380, 1, and memorized easily. These observed facts transformed the list from a shapeless mass into something having definite characteristics, and the observed characteristics stuck in mind and held the rest together. Lists of nonsense syllables, such as wok pam zut bip seg ron taz vis lab mer koj yad are apt to be learned largely by observation of similarities and contrasts, by reading meanings into the syllables, and by grouping into pairs and reading rhythmically. Grouping reduces the twelve syllables to six two-syllabled nonsense words, some of which may suggest meaningful words or at least have a swing that makes them easy to remember. Perhaps the first syllable of every pair is accented, and a pause introduced after each pair; such devices assist memorizing.