return — a belief that is supported by the other members of the movement. By way of preparation for the predicted event, he has engaged in many activities that are entirely consistent with his belief. In other words, most of the relations among relevant cognitions are, at this point, consonant.
Now what is the effect of the disconfirmation, of the unequivocal fact that the prediction was wrong, upon the believer? The disconfirmation introduces an important and painful dissonance. The fact that the predicted events did not occur is dissonant with continuing to believe both the prediction and the remainder of the ideology of which the prediction was the central item. The failure of the prediction is also dissonant with all the actions that the believer took in preparation for its fulfillment. The magnitude of the dissonance will, of course, depend on the importance of the belief to the individual and on the magnitude of his preparatory activity.
In the type of movement we have discussed, the central belief and its accompanying ideology are usually of crucial importance in the believers’ lives and hence the dissonance is very strong — and very painful to tolerate. Accordingly we should expect to observe believers making determined efforts to eliminate the dissonance or, at least, to reduce its magnitude. How may they accomplish this end? The dissonance would be largely eliminated if they discarded the belief that had been disconfirmed, ceased the behavior which had been initiated in preparation for the fulfillment of the prediction, and returned to a more usual existence. Indeed, this pattern sometimes occurs and we have seen that it did happen to the Millerites after the last disconfirmation and to the Sabbataians after Zevi himself was converted to Islam. But frequently the behavioral commitment to the belief system is so strong that almost any other course of action is preferable. It may even be less painful to tolerate the dissonance than to discard the belief and admit one had been wrong. When that is the case, the dissonance cannot be eliminated by giving up the belief.
Alternatively, the dissonance would be reduced or eliminated if the members of a movement effectively blind themselves to the fact that the prediction has not been fulfilled. But most people, including members of such movements, are in touch with reality and cannot simply blot out of their cognition such an unequivocal and undeniable fact. They can try to ignore it, however, and they usually do try. They may convince themselves that the date was wrong but that the prediction will, after all, be shortly confirmed; or they may even set another date as the Millerites did. Some Millerites, after the last disconfirmation, even ventured the opinion that the Second Coming had occurred, but that it had occurred in heaven and not on the earth itself. Or believers may try to find reasonable explanations and very often they find ingenious ones. The Sabbataians, for example, convinced themselves when Zevi was jailed that the very fact that he was still alive proved he was the Messiah. Even after his conversion some stanch adherents claimed this, too, was part of the plan. Rationalization can reduce dissonance somewhat. For rationalization to be fully effective, support from others is needed to make the explanation or the revision seem correct. Fortunately, the disappointed believer can usually turn to the others in the same movement, who have the same dissonance and the same pressures to reduce it. Support for the new explanation is, hence, forthcoming and the members of the movement can recover somewhat from the shock of the disconfirmation.
But whatever explanation is made it is still by itself not sufficient. The dissonance is too important and though they may try to hide it, even from themselves, the believers still know that the prediction was false and all their preparations were in vain. The dissonance cannot be eliminated completely by denying or rationalizing the disconfirmation. But there is a way in which the remaining dissonance can be reduced. If more and more people can be persuaded that the system of belief is correct, then clearly it must, after all, be correct. Consider the extreme case: if everyone in the whole world believed something there would be no question at all as to the validity of this belief. It is for this reason that we observe the increase in proselyting following disconfirmation. If the proselyting proves successful, then by gathering more adherents and effectively surrounding himself with supporters, the believer reduces dissonance to the point where he can live with it.
In the light of this explanation of the phenomenon that proselyting increases as a result of a disconfirmation, let us take another, and more critical, look at the historical examples we have offered in evidence. There are a number of grounds for feeling unsatisfied with them as proof.
In the first place there is a scarcity of data of the sort required by our analysis. It is an understandable lack, for the people collecting historical records were not concerned with our particular problem, but it is a lack nonetheless. Even our best documented example, the Millerites, contains little evidence on actual proselyting behavior, especially among the mass members. Statements about proselyting must be inferred largely from evidence about the number of adherents and the size and frequency of meetings. But such signs as these are dependent not only on the effort made to proselyte — the desire to convince others —but also on the effectiveness of the efforts and on the state of mind of prospective converts.
Even where there is direct evidence about proselyting attempts, such as the number of speeches made, the fact that Miller and Himes traveled widely, or that the Millerite presses worked twenty-four hours a day, these are activities of the leaders. There is very little concrete evidence of the proselyting activities of the ordinary members, whose behavior is most significant for our purposes. Leaders of a social movement may, after all, have motives other than simply their conviction that they have the truth. Should the movement disintegrate, they would lose prestige or other rewards.
And if the Millerite case is inadequately documented for our purposes, our other examples are even more poorly supported. On the Sabbataian movement we have virtually no data concerning the initial disconfirmation in 1648, for the very good reason that the movement attracted little attention (and, hence, there were few records of it) until it became very large and important.
A second reason for considering historical data alone as inadequate is the small likelihood that this kind of data could challenge our explanation. Suppose we could find record of a mass movement that had apparently collapsed immediately after disconfirmation. In the absence of adequate measurement, we might well conjecture that the members’ commitment to the belief was small — so small that the dissonance introduced by disconfirmation was enough to force the discarding of the belief. Alternatively, if the commitment could be demonstrated to have been heavy, it is still possible that there were attempts to proselyte following disconfirmation, but that these attempts had been unsuccessful. This would be a tenable contention since it is the results of proselyting efforts that generally find their way into historical records rather than the efforts themselves.
There is a type of occurrence that would indeed disprove our explanation — namely, a movement whose members simply maintained the same conviction after disconfirmation as they had before and neither fell away from the movement nor increased their proselyting. But it is precisely such an occurrence that might very well go unnoticed by its contemporaries or by historians and never find its way into their annals.
Since the likelihood of disproof through historical data is small, we cannot place much confidence in the supporting evidence from the same sources. The reader can then imagine the enthusiasm with which we seized the opportunity to collect direct observational data about a group who appeared to believe in a prediction of catastrophe to occur in the near future. Direct observations made before, during, and after the disconfirmation would produce at least one case that was fully documented by trustworthy data directly relevant to our purpose.
One day in late September the Lake City Herald carried a two-column story, on a back page, headlined: prophecy from planet. CLARION CALL TO CITY: FLEE THAT FLOOD. IT’LL SWAMP US ON DEC. 21, OUTER SPACE TELLS SUBURBANITE. The body of the story expanded somewhat on these bare facts:
Lake City will be destroyed by a flood from Great Lake just before dawn, Dec. 21, according to a suburban housewife. Mrs. Marian Keech, of 847 West School street, says the prophecy is not her own. It is the purport of many messages she has received by automatic writing,
