We may dismiss the theory that association is the sole cause of reproduction (the second defini tion) by pointing out that the previous joining of ideas is not the only cause- for the reproduc tion of one by the other. This may be due—to mention only two exceptions—to mood (we are likely to recall melancholy events when we are in a melancholy mood); or to a similar relation which two ideas ehanee to have to a third las when 'dog' reminds us of 'cat,' both standing in the class of domestic animals). Rut neither is association satisfactorily considered as repro duction or recollection itself. It bevoincs in this case only a subheading. under memory, find tells us nothing of the relation in of process to process. This leaves us with only our final which says no more than that when and ideational elements group themselves under curtain conditions they may be said to associate. Association then ',peonies simply the form of grouping in which the ele ments stand united. Let us look more closely into this group. First, we must distinguish be tween the processes which do the associating and the materials united in association. In the simultaneous. association, it is always the 'sensa tional elements which start the association. These it is in the ease of the stone which make time perception a stone-perception instead of a table or a wall-pereeption. It may happen, however, that a tendeney toward reproduction of certain ideas wholly diverts the course of the association or the assimilation, as it May be called. Think, for example, of the apprehen sive traveler who comes upon the stone in the dusk and takes it for a ghastly la-ad. Or it may be mood, or sentiment. or temperament, or a violent emotion, which determines the course of the perception. Take, as an illustration. the transformation of the stone by poetic fancy into a treasure-casket or a plaything of the gods. When we look again, at the level of the suc cessive associations, for the critical portion of the perception of idea. the part which does the assoeiating, we find that it may he a single ele ment or the whole perception or idea which is the determining factor. The red (sensation) in a sunset may bring the idea of blood; or a geometrical figure (perception) may suggest the arrangement of petals in a flower: or a melody (perception) an opera in which it ha.; been sung.
The consideration of association as a form of mental connection removes mete more. difficulty which the older association ism eneountered. It makes a more intimate relation within a percep tion or a train of ideas than the latter could There is a unity in the associated mem bers. We mar even speak of a whole perception . or train of ideas as being 'bright,' 'vivid.' `pleasing.' as if it were a single thing. This would hardly he appropriate if the connection were entirely external and simply brought the elements into mere juxtaposition.
If now, we are to speak of 'laws of associa tion,' we can mean only the conditions under which a given association is made. These con ditions will stand in place of the regularity with which the environment brings together certain groups of stimuli and presents them to the organism. Conceiving our laws thus. we find that the old category of contiguity comes near est to expressing the truth of the matter. Simi
larity is a very doubtful rubric.. In the first place, the term is too indefinite to he of much service. Almost anything may be more or less similar to anything else, and it is very hard to see how simple likeness should de termine an association. At least, greater like ness does not seem to mean greater assoeiability. It seems to be possible, in every case, to reduce an apparent 'association by similarity' to some other form; there may be partial identity. in which case the identical member of two groups is first experienced with one and then with the other, as the red of the sunset passing over into the red of blood ; or there may be simple con tiguity, the similar objects having stood together in experience; or there may be an indirect asso ciation, such as we found above, where two timings stand in the same relation to a third, as two objects to the name of the class to which they belong.
On the other hand, 'contiguity' can be re garded as an explanation only when we think of it in terms of neural activity. It is true that we can state it in general psychological terms, saying that conscious connections once set. up tend to persist, so that when a single item (sensation, fusion. perception) conies to consciousness, it tends to bring with it other items with which it has appeared before in consciousness. This for mula, which may be regarded either as a law of habit—if we think of the tendency to persist— or as a law of redintegration—if we think of its results--receives. however, its full signifi cance only when we translate it into physiologi cal terms and state it as a tendency in the ne•v ous system to do that which has been (lone be fore. Neural tracts which have once been excit ed together, tenth to funetionate together again; hence. if an excitation finds its way to one part of the brain, there is a tendency for the parts previously involved to lie rei;xcited. Now, just what path an excitation will take in any particu lar case we cannot foretell; but we are able to name with a good deal of certainty the factors which contribute to association, i.e. to state the conditions under which the tendency to form old connections is realized. At this point, ex periment comes to our aid. It enables us to invite association under first one set of eonditions and then another, and to watch the results. If a given process, a. has formed several connec tions, at various times, say, with s, and z, the one of these connections which is actually real ized when a is brought to consciousness will depend upon the frequency of connection of each with a, the recency of connection, the relative vividness of d, s, and z. their relative positions in a series of processes, and their relative power to attract a wandering attention. The compara tive values of these factors are determined by presenting to the observer simple stimuli (as colors, letters, or sounds), arranged in definite groups of two, and then asking him, when one member of a group is given again, to recall the other member which wes previously given with it. The series are so arranged as to bring into prominence, one after another, the various fac tors.