HF. £351....“ . "s I“! .r! u: ‘ {‘1- ‘ Q 9 .33 I lfofi a l. : ¥ ‘0 Q n 1", o I.— 'Dfiha 4 58; . ’-,c‘!9 333's) -.‘.- . D an...“ #1- 9" A..—u ‘ .‘O u... . z. f 1 en K , v u; § 1‘. szu H; ‘03: " ’1 :1!ch U "a y I A 5 ,b *ué f. a a. H? n? . Y hm A R ”B #1 K State Micki-"inn L ., Uth‘Si‘cy A THEORETICAL AND EMPIRICAL ANALXSIS OF IOYALTY TO AND DEFECTION FROM A BELIEF-SYSTEM By John J. Laffey AN ABSTRAJT submitted to the College of Science and Arts of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF.ARTS Department of Psychology 1957 John J. Laffey A theoretical distinction was drawn between two kinds of defect- ion from.and loyalty to a belief-system. The distinction between dogmatic and intellectual defection, and between dogmatic and intel- lectual loyalty, was said to lie in the relative extent to which the individual beliefs of the belief-system were integrated. This distinction was then related to the concept of dogmatic thought pro- cesses, and to previous findings of the relatively poor integrative thinking of high scorers on a dogmatism scale. It was hypothesized that low dogmatic subjects would tend to defect intellectually from, and high dogmatic subjects tend to defect dogmatically from, or re- main loyal to, an "impossible" experimental belief-system. The Den- ny Doodlebug problem, a complex task of problempsolving requiring the integration of beliefs for a solution, was employed, with log- ical contradictions making a solution impossible. Each subject ad- ministered the problem to another person, who was actually a sec- ond experimenter in disguise, and who sought to stimulate the pro— cess of defection by challenging the reality of a problem solution. The following hypotheses were made and results obtained: hypothesis Ia. There will be a greater incidence 0d dogmatic defect- ion among high dogmatic subjects than among low dogmatic subjects. None of the 26 subjects defected dogmatically. Hypothesis Ib. There will be a greater incidence of intellectual defection among low dogmatic subjects than among high dogmatic sub- jects. The results were not significant, but in the predicted dir- action. Hypothesis III.Low dogmatic subjects will tend to defect earlier in ‘1. O APF’I‘oved [’7 WV [3/ 13.4,” '4! —2— John J. Laffey the experiment or entertain the possibility of defection sooner than will high dogmatic subjects. No significant differences were obtained. Hypothesis III. There will be a greater incidence of dogmatic loy— alty among high dogmatic subjects than among low dogmatic subjects. The difference was significant at the .12 level. Hypothesis IV. Irrespective of dogmatism scores, there was a pos- itive relation, as predicted, between integrative behavior on the study of "party—line" change, reported by Cram (1) and integrative behavior in the present study, significant at the .01 level. It was felt that the primary reason for failure to obtain dogmat- ic defection lay in the lack of authority accompanying the "pseudo- subject's" challenging of the validity of the problem, and the rel- atively greater authority, represented by the original experiment— er, supporting the maintenance of a belief in a problem solution. As a result, subjects tended to remain non—committal during the experi ment. It was concluded that a definite relation between integra— tive thinking in one situation and intellectual defection in another, and between non-integrative thinking in one situation and dogmat— ic loyalty in another, were establiShed. It was suggested that the methodology employed offered nymerous possibilities for studying the effects of manipulating certain aspects of the problempsolving sit- uation so far relatively unexploited. (l) Oram, Alfred, A theoretical and empirical analysis of party line thinking. M.A. Thesis, Mich. State U. Library, 1957. A Theoretical and Empirical Analysis of Loyalty to and Defection from a Belief-System. by John J. Laffey A THESIS Submitted to the School of Graduate Studies of Michigan State University of Agriculture and applied Science in partial fulfillment of the requirements for the degree of Master of Arts. Department of Psychology 1957 ? «a: J~ s” ‘2 ACKNOWLEDGEMENTS Acknowledgement is hereby made to Prof. M. Ray Denny and Dr. Charles Hanley for their valuable suggestions, to Alfred Oram, for his assistance in the gathering of the data, and especially to Prof. Milton Rokeach for his patient supervision and useful criticisms, and finally, to my parents for their many years of encouragement. TABLE OF CONTENTS Page Introduction.............................................l Method...................................................7 The conditions 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 07 The Situation. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 08 The PrOblem0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 08 The hints00000000.00000000000000000000000000000000008 The conditions 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 10 The Situation. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 011 The prOblem0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 11 The hints00000000000000000000000000000000000000000011 Hypothese8000000 00000000000000000000000000000 00000016 Hypothesis I (a) and (b)......................16 IIyPOtheSis II. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 016 HypOtheSis IIIO 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 016 WPOtheSis Iv0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 016 Resu1t80000000000000000000000000000000000000000000017 Hypothesis I (a) and (b)......................17 HyPOtheSis IIO 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 018 HyPOtheSiS IIIO 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 020 HYPOtheSiS 1V0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 020 DiscuSSj-on0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 024 summary0000000000000000000.0000000000000000000000000000050 References.00000000000000000000000000000000000000000000058 LIST OF TABLES Page Table I.................................................34 Table II................................................35 Table III...............................................36 Table IV................................................56 Table v000000000000000000000000000.00000000000000000000057 -iii- INTRODUCTION In the research to be reported here, the problems of defection from and loyalty to a belief-system are being studied. To our knowledge, neither of these phenomena have been investigated experimentally before, nor has the method— ology herein reported been used elsewhere (1). By defection is meant the relinquishing of a previously held belief-system. It is not the giving up of a single belief (e.g., the denial of the necessity of reinforcement for learning by a behaviorist) but of a belief-system (e.g., the rejection of behaviorism, and adoption of a gestalt position, by a former behavioristic psychologist.) By loyalty is meant the adhering to a belief-system in the face of a challenge to the validity of that belief- system. An obvious example is the belief in capitalism maintained by most Americans in Korean prisons, despite the claims of their guards in favor of another socio-economic system of society. It is in the areas of religious and political belief- systems that we can most clearly see defection and loyalty, and in which marked differences in the processes of defection and loyalty may be noted. For example, we readily welcome defectors from Communism, yet we frequently find that not all defect for the same reasons. The defector who escapes from behind the Iron Curtain because his Black Marketeering has attracted the attention of the police, leaving a sick wife and several children, is surely motivated in a different way than the editor of a Soviet journal, who defects because he finds the changes from Bolshevism to Stalinism to anti- Stalinism, always with trials, purges, and murders, to be a corruption of Marxist ideology. As an illustration of defection from a religious belief— system, let us take two former members of a Fundamentalist sect. Defector A, we may find, has left his church because he has continually broken its rules, violated its precepts, and is unable to repent of his sins before the congregation because of fear of punishment. Defector B, on the other hand, may not have had difficulty in living in the manner prescribed by the church, but has found the contradictions between a literal interpretation of the Bible and the changing state of society, between the Adam-Eve myth and the evidence of evolution, and between the church's pessimistic view of man's state and the cultural optimism engendered by modern science to be irreconcilable, and so leaves the church. The distinction which the above examples seem to illus- trate, with respect to the process of defection, is between dogmatic and intellectual defection. The dogmatic defector is one who, due to his failure to utilize the beliefs (rules, doctrines, sets, etc.,) of a belief-system in the solving of problems arising in relation to that system, gives up the belief-system. An intellectual defector, on the other hand, is one who, by virtue of his utilization of the beliefs of the belief-system, discovers the contradictions between beliefs, and comesto reject the belief-system as inadequate to solve problems arising in relation to it. Thus, in our example of defection from a political belief-system, we should consider our Black Marketeer ex-Communist a dogmatic defector, and our ex-Communist journalist an intellectual defector. In our example of defection from a religious belief-system, the first mentioned, defector A, would be considered a dogmatic defector, while the latter, defector B, would be considered an intellectual defector. A similar distinction may be made with respect to loyalty. In the area of esthetic belief-systems, it is well known that Toscanini revered Verdi as his favorite composer, but surely was not blind to the merits of other composers, (even to the extent of being satisfied with his own conducting of Beethoven's Ninth.Symphony only near the very end of his long career). The loyalty to the same belief- system (the music of verdi) expressed by a musical enthusiast who will praise no compositions other than those of Verdi is certainly a different sort of loyalty. The "patriotic" author of the statement, "My country, right or wrong", seems to be expressing a sort of loyalty that we feel not all Americans would share. The distinction in both examples above, seems to be between loyalty to an authority (composer, institution, etc.,) and lpyalty to the ideals which the authority represents, i.e., a distinction between dogmatic and intellectual loyalty. In making the distinction between the two kinds of defection and loyalty, it would seem that it is in the in- tegration of the various beliefs which compose the belief- system that the primary distinction lies. Relatively poor integration of beliefs necessarily results in relatively poor ability to solve problems requiring such integration. Relatively good integration of beliefs results in relatively good ability to solve such problems. Moreover, psychological resistance to integrating beliefs should result in an inability to see contradictions between beliefs. Conversely, contra- dictions between beliefs will best be seen by those who show good integration of the beliefs. In other words, we are hypothesizing that highly dogmatic people-— people with relatively closed belief—systems -- will, by virtue of their resistance to integrating new beliefs-into a new system, be unable to see logical incongruities between beliefs, and will thus remain either dogmatically "loyal" to an "impossible" belief-system, or will defect dogmatically. We should expect, correspondingly, that highly dogmatic people who defect dog- matically will, through lack of integration, be unable to specify the logical reasons why they defected, but that low dogmatic people who defect intellectually, will, by virtue of greater integration, be able to specify the objective grounds that require defection. So far we have more or less assumed the inverse relation- ship between dogmatism and integrative thinking. We have done so drawing upon the definition of dogmatism as "(a) a relatively closed cognitive system of beliefs and disbe- liefs about reality (b) organized around a central set of beliefs about absolute authority, which, in turn, (c) pro— vides a framework for patterns of intolerance and qualified tolerance toward others" (2). Subsequently, Rokeach, McGovney and Denny (5) have experimentally differentiated dogmatism from rigidity (which concerns the overcoming of single sets or beliefs) showing the markedly greater time required by high dogmatic subjects as opposed to low dogmatic subjects, to integrate new beliefs into a problem solution. More recently, Vidulich has shown that the "inability to integrate seems to be a function of the fact that the new beliefs to be integrated into a new system are not all present in the phenomenal field", due to the reluctance of the high dogmatic to replace the old with the new beliefs (4). Since it is at present impossible to subject loyalty and defection in the religious, political, scientific, and other spheres in which it is naturally found, to laboratory study, we will endeavor to construct an experimental analogue to the processes of defection and loyalty that occur in these spheres. We will present subjects with a new belief-system, an artificial world, in which the integration of new beliefs is necessary for the solution of a specific problem. We will, furthermore, manipulate these beliefs so that, in one instance, a second, different solution must be found, and finally, by changing the problem so that the beliefs come into contradic- tion with one another, a solution is logically impossible. The discovery of the nature of this impossibility will be considered an operational definition of intellectual defec- tion. On the other hand, an insistence that the problem is impossible to solve without being able to state why will be considered an operational definition of dogmatic defection. Thus, we are trying to find out whether integrative thinking will occur, in the face of an impossible problem, so that our subjects will see the contradictions among be- liefs, and defect intellectually. Failure to defect, or dogmatic defection (i.e., defection, with inability to specify the reasons that make the problem impossible) will be evidence of non-integrative thinking. METHOD The artificial world to which we introduce our subjects 1 whose is that of a mythical creature called Joe Doodlebug, basic problem is to get to a pile of food in exactly four jumps according to very specific rules to be described shortly. The Doodlebug problem and its variations have been used in the studies previously mentioned (3), (4), and in the studies of Mikol (5) and Swanson (6) for the investigation of integrative thinking in problem-solving. The situation in which a single belief is changed, and a different one substituted in this artificial world, is analagous to the phenomenon of "party-line change", more completely reported by Alfred Cram (7). Inasmuch as the present study was done in connection with Oram's study of "genuine" vs. "party-line" change, in the solving of prob- lems, it will be well to present the original Doodlebug problem, then its modification in the "party-line change" study,and finally its form in the present study. The Conditions Joe Doodlebug is a strange sort of imaginary bug. He can and cannot do the following things: 1. He can jump in only four different directions, north, south, east, and west. He cannot jump diagonally 1) Joe Doodlebug was launched on the psychological world in 1945, thanks to Dr. M. Ray Denny of Michigan State University. —8- (e.g., southeast, northwest, etc.). 2. Once he starts in any direction, that is, north, south, east, or west, he must jump four times in that same direction before he can switch to another direction. 5. He can only jump, not crawl, fly, or walk. 4. He can jump very large distances or very small distances, but not less than one inch per jump. 5. Joe cannot turn around. The Situation Joe has been jumping all over the place getting some exercise when his master places a pile of food three feet directly west of him. Joe notices that the pile of food is a little larger than he. As soon as Joe seesall this food, he stops dead in his tracks facing north. After all his exercise, Joe is very tired, and wants to get to the food as quickly as he possibly can. Joe examines the situation, then says, "Darn it, I'll have to jump four times to get to the food." The Problem Joe Doodlebug was a smart bug, and he was dead right in his conclusion. Why do you suppose Joe Doodlebug had to take four jumps, no more and no less, to reach the food? The Hints (Given to S at specified time intervals.) A.) Joe does not have to face the food in order to eat it. B.) Joe can jump sideways and backwards as well as forwards. C.) Joe is not necessarily at the beginning or end of a series of jumps. He may have been somewhere in a series of jumps. The answer is, of course, that Joe had already taken one jump to the east when the food was placed down, therefore must take three more jumps east, completing the series in that direction, then change his direction, making one jump to the west, landing on top of the food, where he can eat. As mentioned previously, we have tampered with the original problem, so as to make a new solution necessary. Whether subjects will be able to integrate the original beliefs of the problem with the new belief, so as to find the new solution (i.e., make a "genuine" or integrative change) or whether they will persist in trying to apply the solution of the original problem ("party-line" or non- integrative change) is essentially the object of the study reported by Oram. The only difference between the original problem and the variation employed by Oram is that in the latter there is a canopy over the food. This results also in a change in one of the hints, i.e., now Joe must face the food in order to eat it. Thus, for the sentence in The Situation reading, "Joe examines the situation, then says, 'Darn it, I'll have to jump four times to get to the food,'" there is substituted the sentence, "Joe examines the situation, noticing that there is a low canopy over the food, then says, 'Darn it, I'll have to jump four times to get to the food.'" For the Hint reading, "Joe does not -10- have to face the food in order to eat it," there is the substituted hint, "Joe must face the food in order to eat it." Otherwise, the "canopy" problem is identical to the original, and Joe must still make four jumps to get to the food. The correct solution to the "canopy" problem is that Joe had already taken one jump to the west, consequently must take three more jumps west, landing on top of the canopy with his third jump, then make one jump backwards to the south, where he will be facing the food and able to eat. For the present study, we have further modified the "canopy" problem so that it is now impossible to solve. For purposes of clarity, the problem used in this study is presented in full as follows. The Conditions Joe Doodlebug is a strange sort of imaginary bug. He can and cannot do the following things: 1. He can jump in only four different directions, north, south, east, and west. He cannot jump diagonally (e.g., southeast, northwest, etc.). 2. Once he starts in any direction, that is, north, south, east, or west, he must jump four times in that same direction before he can switch to another direction. 3. He can only jump, not crawl, fly, or walk. 4. He can jump very large distances or very small distances, but not less than one inch per jump. 5. Joe cannot turn around. -11- The Situation Joe has been jumping all over the place getting some exercise when his master places a pile of food three feet directly west of him. Joe notices that the pile of food is a little larger than he. As soon as Joe sees all this food he stops dead in his tracks, facing north. After all his exercise, Joe is very tired and wants to get to the food as quickly as he possibly can. Joe examines the situation, noticing that there is a low canopy over the food, then says, "Darn it, since I have already made one jump to the east, 1'11 have to jump four times to get to the food."1 The Problem Joe Doodlebug was a smart bug and he was dead right in his conclusion. Why do you suppose Joe Doodlebug had to take four jumps, no more and no less, to reach the food? arms. A.) Joe must face the food in order to eat it. B.) Joe can jump sideways and backwards as well as forwards. C.) Joe is not necessarily at the beginning or end of a series of jumps. He may have been somewhere in a series of jumps. It may be observed that the only additional change is l) The underlining was, of course, not present in the instructions to the subjects. It is the added condition that Joe had been travelling east which makes the problem impossible to solve. -12- the inclusion of the statement that Joe had been travelling east at the time the food was placed down. Thus, the prob— lem is impossible, for Joe cannot get to the food under the conditions stated, in less than five jumps. Subjects in the present study were 26 of the 40 who originally participated in Oram's study. The latter were the 20 highest and 20 lowest scorers on a 40-item Dogmatism Scale (8) (a high score indicating a greater degree of dog— matism) administered to a class of some 600 college students enrolled in an introductory course in psychology in the spring of 1956. In the "party-line change" experiment, one half the subjects (10 high dogmatics and 10 low dogmatics) were in- terviewed by one experimenter (A.O.) and one half by the other (J.L.). At the conclusion of the first two Denny Doodlebug problems (the original problem plus the canopy variation involving the substitution of a single new belief, each subject was asked to return for the third problem approximately one week later. Inasmuch as it had been found in a prior, unreported study by the same experimenters that subjects given the "impossible" problem to solve in general could not solve it (would not defect) in rather generous amounts of time, the danger of this study becoming one of frustration toler- ance had to be met. Accordingly, a rather ingenious pro- cedure was adopted, modelled after America's annually most profitable criminal technique, the "Confidence Game."1 1) Thanks are gratefully extended to Dr. Charles Hanley for his ability to combine an interest in the nefarious with scientific acumen. -15- Upon returning for the third problem, each subject was met by his experimenter, and told substantially the follow- ing: "You did pretty well on those problems last time. They are so difficult many people never solve them at all. Do you do much work with problems like that? "You don't say. Say, we have a project going on here that ought to interest you. You know, whenever we give problems to people to solve, there's always the chance we get biased results through the experimenter giving uncon- scious clues to the subject. In other words, the experimenter unconsciously tips off the subject when he gets close to the solution. It's probably something like ESP, you know, extra- sensory perception. We are trying something that's never been done before in psychology. 'We are testing subjects, who of course don't know the answer to the problem, but the unusual thing is that we are also using experimenters who don't know the answer either. "The trouble is, that for each subject we need a new experimenter, and there just aren't that many people in the Psychology Department to do the job. So we've had our eye out for intelligent undergraduates who seem to know their way around these problems, to act as an experimenter for us just once. "For work like this you really ought to be paid pretty well, but our research budget is limited, and we can afford k -14- to pay you only one dollar. Your job is simply to give the third problem to another person to solve, and we'll pay you a dollar for doing so. We have it worked out so that there is very little for you to do. Since you haven't been trained in taking notes, everything will be recorded automatically on this tape recorder, which I'll put on, so you won't have to worry about that at all. "You just read the instructions to the subject, and put the problem and hints on the table for him to look at, and then start the stopwatch. He's already had the first two problems, so he shouldn't have any questions. "The rest is very simple. If the subject gives an answer, and you agree with it, just tell him so. On the other hand, if he gives an answer which you think is wrong, tell him so also. In other words, act natural. "Let the subject work for thirty minutes, or until he solves the problem, whichever comes first. If you should notice anything interesting that the subject says or does, you might tell me about it when you finish. "Are there any questions? "We have a subject scheduled to come in about now. You could be the experimenter for him." Proceeding in this way, it was possible to enlist the cooperation of each subject who appeared for the third problem. At this point, the experimenter brought to the subject-experimenter (S—E) the third person, who was to be -15- the alleged subject. However, unknown to the S-E, the new "subject" was in reality the alternate experimenter, dressed in typical campus apparel, and appearing friendly, coopera- tive, but naive. Thus, the two roles were experimentally reversed: each S-E, acting as "experimenter", was actually an unwitting subject for his "subject", the other experi- menter (E-S). When the experimenter had left the room, with the re- corder on, the E-S was allowed to read over the problem, then the 8—H read the problem aloud, seeing it himself for the first time, and following the same procedure he had witnessed when he himself was a subject in the "canopy experiment" conducted by Cram At the end of five minutes, E-S states: "Are you sure there's a solution to this problem?” At five-minute intervals thereafter, E-S states: "This problem is impossible. Could that be the answer, that Joe can't make it in four jumps?" At no time does the E-S give reasons for his statements, but each time attempts to elicit a response of agreement or disagreement from the S—E as to his conclusion of the impossibility of the problem. At the end of thirty minutes, or sooner, if defection has occurred, (i.e., S-E agrees with E-S) the watch is stopped, and the S-E contacts the original experimenter. The latter then thanks his "accomplice," the E-S, for his cooperation, pledges him to secrecy, and dismisses him. —16- The S-E is then asked for his impressions of the experiment, and, if defection has occurred, the reasons why the problem is impossible. If defection has not yet occurred, S-E is asked what he thinks the answer is, and if he now states that the problem is impossible, is asked his reasons. The S-E is then thanked, paid his dollar, and pledged to secrecy. There was no evidence of communication about the experi— ment among the subjects during the period in which the study was carried on. Hypotheses Hypothesis I. (a) There will be a greater incidence of dogmatic defection among high dogmatic subjects than among low dogmatic subjects. (b) Conversely, there will be a greater incidence of intellectual defection among low dog- matic subjects than among high dogmatic subjects. Hypothesis II. Low dogmatic subjects will tend to defect intellectually earlier in the experiment or entertain the possibility of defection sooner than will high dogmatic subjects. Hypothesis III. There will be a greater incidence of dog- matic loyalty among high dogmatic subjects than among low dogmatic subjects. Hypothesis IV. Subjects who exhibit non-integrative think- ing in one situation should likewise exhibit non-integrative thinking in the present study, i.e., defect dogmatically or remain "loyal." Conversely, subjects who exhibit integra- tive thinking in the "party-line change" study should likewise -17- exhibit integrative thinking in the present study, i.e., defect intellectually. Results Hypothesis I. (a) There will be a greater incidence of dogmatic defection among high dogmatic subjects than among low dogmatic subjects. All of the subjects who agreed that the problem was impossible were able to state the reasons why, consequently all were "intellectual defectors." Thus, we have no infor- mation relevant to dogmatic defection. Hypothesis I (a) is untested. (b) There will be a greater incidence of intellectual defection among low dogmatic subjects than among high dog- matic subjects. The results relevant to Hypothesis I (b) are shown in Table I. Originally there were six intellectual defectors in our group of 12 high dogmatic subjects, and seven intellec- tual defectors in our group of 14 low dogmatic subjects. However, what we did not expect (but what previous studies have indicated as not improbable) was that some of our sub- jects would solve the problem which supposedly had been made impossible, by an ingenious yet logical solution. The answer is, of course, that if Joe Doodlebug were to jump three times to the east, circumventing the earth, landing with the third jump on top of the Canopy, he could then make one jump back- ward to the south, and be in position to eat the food. -18- In other studies, this answer has occurred approximately once in 60 times (3). On these occasions, it has been dis- allowed on the grounds that a) Joe cannot jump.pH§§ far, and b) anyway, the earth is flat. In the present situation it was not possible to disallow such a solution as it arose, for it was impossible to know when it actually did arise in the situation, and because here our real experimenter (E-S) could not, by virtue of the role he was playing, reject such a solution. Since this new solution is a manifestation of integrative thinking, we shall reserve the four instances of it (all from low dogmatic subjects) for analysis of Hypothesis III (Table IV) rather than include them here. Thus, Table I reveals six intellectual defectors out of 12 high dogmatic subjects, and seven intellectual defectors out of 10 low dogmatic subjects. The chi-square computed according to Fisher's exact method yields a probability of .29 (9). While this result is in the predicted direction, the probability of its outcome by chance does not warrant rejection of the null hypothesis. Hypothesis I (b) cannot be accepted with the customary degree of confidence. Hypothesis II. Low dogmatic subjects will tend to defect intellectually earlier in the experiment or entertain the possibility of defection sooner than will high dogmatic subjects. In determining time to defect, the actual time at which the subject left the experimental room was obtained from the stopwatch which the subject was to stop himself. In instances -19- where the subject failed to stop the watch, the "pseudo- subject" stopped the watch as soon as the real subject had left the room. Table II reveals the time in whole minutes for each subject, and the subject's status as either high (H) or low (L) in dogmatism. The rank test described by White was employed to deter- mine whether a significant difference in time taken to de- fect was present between the two groups (10). The T value of 125 and the T' value of 105 do not permit the rejection of the null hypothesis. Consequently, there was no signi- ficant difference between high and low dogmatic subjects in time to defect intellectually. From all subjects, excluding those four who found a new solution, the responses to the question of the solva- bility of the problem, given at five minute intervals, were recorded on separate sheets of paper (with a code number) and given in random order to three judges, with no indica- tion of the subject's name, dogmatism score, status as to defection, or of the time during the experiment at which the statement was made. Of a total of 110 such statements, only those 65 on which all three judges agreed were tabu- lated. Thirty-six of these statements were from high dog- matic subjects; 29 from low dogmatic subjects. Table III shows the number of statements of high and low dogmatic subjects which all three judges agreed indicated either "loyalty" to the belief that there was a problem solution, or that the subject was ”entertaining the possibility -20... of defection" from this belief, or that it was uncertain which of the above was indicated. While there is a very slight tendency in reverse of what was predicted, the low magnitude of the difference does not warrant rejection of the null hypothesis. Hypothesis II, therefore, is uncon- firmed. Hypothesis III. There will be a greater incidence of dog- matic loyalty among high dogmatic subjects than among low dogmatic subjects. In testing this hypothesis, we are pitting dogmatic loyalty on the one hand (a manifestation of non-integrative thinking) against intellectual defection on the other (a manifestation of integrative thinking). However, as men- tioned previously, a new solution was found by four subjects, a solution which involves integrative thinking. Conse- quently, we have grouped both instances of integrative thinking -- intellectual defection and the new solution -- and compared them with all instances of dogmatic loyalty. Table IV shows six dogmatically loyal out of 12 high dog- matic subjects, against three dogmatically loyal out of 14 low dogmatic subjects. A chi-square analysis according to Fisher's exact method yields a probability of .12 (9). Since this is not within the customary significance limits, Hypothesis III cannot be considered proven by our results. Hypothesis IV. Subjects who exhibit non-integrative thinking in one situation should likewise exhibit non-integrative thinking in the present study, i.e., defect dogmatically -21- or remain "loyal". Conversely, subjects who exhibit inte~ grative behavior in the "party-line change" study (7) should likewise exhibit integrative behavior in the present study, i.e., defect intellectually. In the "party-line change" study, scores on the Dogma- tism Scale as an index of integrative thinking were related to performance on problems requiring integrative thinking. In the present study, we have used the same scale as an index of integrative thinking, and related this to intellec- tual defection (a manifestation of integrative thinking) or failure to defect (a manifestation of non-integrative think- ing). However, since any scale for measuring a psychological attribute has its shortcomings, it may be that an experien- tial manifeStation of this attribute may have more accurate predictive value. Accordingly, we have related integrative behavior in one problem-solving situation (the "canopy" problem used by Cram) to integrative behavior in the present experiment (the "impossible" problem). Thus, regardless of dogmatism scores, we tried to predict the results of the present study from the kind of thinking manifested by sub- jects in the Wparty-line change" experiment (7). In the latter study, subjects were given the two prob- lems in either of two orders: The so-called "no-canopy" problem (problem A) followed by the so—called "canopy" problem (problem B), or vice versa. Groups which had prob— lem A first will provide the average time required for solution of problem A when there is no opportunity for a savings due to having solved problem B. Similarly, groups having problem B first will demonstrate the "absolute" time required to solve problem B. In the A-B order, problem A was solved in a mean time of 8.96 minutes by low dogmatic - subjects, and in 20.11 minutes by high dogmatic subjects. In the B—A order, low dogmatic subjects solved problem B in a mean time of 15.54 minutes, high dogmatic subjects in a mean time of 20.96 minutes. These times then, represent the "average difficulty" of each of these problems for high and low dogmatic subjects, when each problem is first in the series. If then a subject solves a problem (B) given in the second position (B—A) in less time than his (similarly high or low dogmatic) group solved it when in the first position (A—B) we may conclude that having worked on problem.A first has helped his solution to problem B. Some amount of inte- gration of the beliefs of the problem must have occurred. However, if the time taken to solve a problem in the second position is equal to or greater than the mean time required by other subjects (with similar dogmatism scores) to solve it when it appears first, there has been no saving, and we may conclude that integrative thinking has not occurred to any measurable extent. In analyzing our results, the high and low dogmatic subjects were divided into two groups, according to the order in.which they were presented the two problems in the earlier study (A-B, or B—A). The time required for each -25- subject to solve the second problem (be this A or B) was compared with the mean time required by other high or low dogmatic subjects to solve the same problem, when that problem is given gipgp. When the time taken to solve the problem in the second position is gpeater than the mean time taken to solve the same problem by other high or low dogmatic subjects when it is given.§ip§§, we have assumed this to be an indication of non—integrative thinking. When the time for the problem in the second position is lggg than the mean time taken to solve the same problem in the first position, we have assumed this to be an indication of integrative thinking. On the basis of the two kinds of thinking, (in- tegrative and non-integrative) found in the "party-line change" study, we have made predictions as to the outcome of the present study, with respect to integrative and non-integra- tive thinking. We have predicted integrative thinking in the present study (intellectual defection or new solution) when the subject was judged to have exhibited integrative thinking in the former experiment, and non-integrative think- ing in the present study (dogmatic loyalty) when the subject was judged to have exhibited non-integrative thinking in the former study. Table V compares integrative and non- integrative thinking in the present situation, with integra- tive and non-integrative thinking in the "party-line change" situation. On the basis of the data of the "party-line change" study, it was possible to predict accurately the results of the present study in 19 of the 26 cases. A chi-square -24- analysis yields a value of 6.58, which is significant at the .01 level. Consequently, Hypothesis IV, predicting a positive relationship between integrative behavior in the "party—line change" experiment and in the present ex- periment was borne out. Discussion Since one of the primary aims of this study was to exa- mine the processes of dogmatic and intellectual defection, it may be well to address ourselves to the question of why there was no dogmatic defection. To answer this question we must return to the point at which we set the stage, so to speak, for the observation of the process of defection. In connection with the "party-line change" study (7) as previously described, each subject attempted to solve two of the Denny Doodlebug problems. Actually, each subject felt that he had solved the problems, for extra help was given at the end of the experimental time limit (45 minutes) in order to maintain in the subject a feeling of accomplish- ment and a sense of closure. We may refer to this as the "indoctrination" period, i.e., the period in which the sub- ject is repeatedly exposed to certain new beliefs (or sets, etc.,) by an authority figure (the experimenter) with success in the situation dependent upon accepting these new beliefs. Upon returning for the final problem (the "impossible" problem) the subject is met by the same authority figure, who reinstates to a large extent the situational cues, and who now invests the subject with the authority to act in -25- his (the experimenter's) stead. We may say that, were our belief-system a "cause" in the real world, our subjects are first "indoctrinated" into the "cause," then put to the test in the face of an "infidel." The "infidel" is, of course, our "pseudo-subject," the other experimenter, who in effect challenges the belief-system's validity (or the validity of the "cause") and in the process, challenges the authority for this belief-system as well. However, this "unbeliever," disguised as a fellow under—graduate, carries no authority for his statements. He is probably perceived in most instances as a peer at best. Furthermore, our subjects know that they must stay in the situation 50 minutes at most. Consequently, should certain anxieties be aroused by answering the challenger's statements of disbelief, the subject, if he is clever and verbally skilled, need only manage to avoid commitment at these five-minute intervals, and persevere in this manner to the end of the experiment. Perhaps this explains the fact that our judges were unable to agree completely on 45 of the 110 statements made by subjects during the experi- ment. The subjects may well have been trying merely to avoid the outright commitment our "pseudo-subject" sought to obtain. Even in the ranks of defection, six of the total 15 defectors acknowledged their defection.gply at the con- clusion of the experiment. Apparently, then, even when intellectual defection was a mental reality, it could not always be confessed to our "pseudo—subject,” given the -26.. circumstances as they existed. One may liken this situa— tion to one in which a graduate student, assisting in a course, finds the seeds of doubt sown by a stubborn under- graduate. While the latter does not offer adequate refuta- tion of the professor's theory, the graduate student, with the germination of this doubt aided by critical appraisal, concludes that the theory is, in fact, erroneous. He does not, however, admit this to the undergraduate, but remains noncommittal, and presents his objections privately to the professor. We believe that the primary cause of our failure to obtain dogmatic defection lies in the fact that the figure who calls the belief-system into question carries with him no authority, whereas the belief-system is reinforced by an authority figure found by the subject in the past to be always reliable and correct. This conclusion seems logical in the light of the definition of dogmatism previously quoted, with reference to the cognitive system of beliefs and dis- beliefs about reality which are "...organized around a central set of beliefs about absolute authoripy..." (2). In brief then, our subjects are eXposed to a challenger of their beliefs about Joe Doodlebug and his world, and of the authority behind those beliefs, but a challenger with no authority, hence no strong claim to be heard. The authori- tative speaker on the rostrum often feels no need to reply to the anonymous heckler, no matter how cogent the latter's criticisms. -27- We would expect that by increasing the authoritative aspect of the pseudo-subject, we would greatly raise the frequency of dogmatic defection. What would be the result if, for instance, instead of a "nice, friendly undergraduate," our subjects were faced with an "imperious and awesome faculty-member," who sat glumly, and stated emphatically (if not dogmatically) that the problem is impossible, and that this was obvious to anyone, and demanded an answer? Granted that this might precipitate the equivalent of an anxiety-attack, we would predict that dogmatic defection would necessarily occur, and more frequently in high dog— matic subjects than in low dogmatic subjects. Further research on the problem of defection should involve the introduction of a new authority to challenge the old. One of our results which deserves comment is the finding of a significant relationship between the kind of thinking demonstrated by subjects in the "party-line change" study by Oram (7) and that demonstrated in the present study. As has been suggested above, the rationale underlying the comparison of the two experiments lies in the fact that whereas in the former study, behavioral manifestations of integrative thinking were predicted from scores on a per- sonality scale, the present study employs those same instances of integrative behavior as predictors of behavioral manifesta- tions of a similar thinking process in the loyalty-defection study. We assume that while the paper and pencil test has a certain predictive value, the direct, observable manifestations -28- of the kinds of thinking processes under investigation should be even more accurate prognosticators of the recurrence of manifestations of similar kinds of thought processes. Our results, shown in Table V, bear out this assumption. It is further interesting to note that of the six high dogmatic subjects for whom defection was predicted on the basis of their times taken to solve the second problem in the "party-line change" study, these times were, in four cases below, and in one case identical to, the mean times for the same problem for low dogmatic subjects. Of these six subjects, five did, in fact, defect. The non-defector was the subject whose time was greater than the mean of the low dogmatic subjects. In other words, those high dogmatic subjects who defected tended to have manifested integrative thinking at least comparable to that of the low dogmatic subjects. In short, they had behaved more like low dogmatic subjects than like high dogmatic subjects. What these reversals imply other than that a scale may have its limitations cannot be inferred from this number of exceptions. What seems fairly conclusive is that integra— tive thinkers will tend to defect intellectually from, and -non-integrative thinkers tend to remain loyal to, the belief in a problem solution when there is no solution to the prob- lem. While it is hypothesized that the latter will also tend to defect dogmatically, the present study cannot offer empirical evidence for such a conclusion, although there are at this writing tentative indications that under stressful -29- conditions, integrative thinking tends to be severely in- hibited (11). The suggestion offered, of increasing the authoritative aspect of the "pseudo-subject," would result in increasing the stress on the subject in the experiment. A final word may be said about the experimental metho— dology, which is rather unique. Granted that the task of arranging for an "indoctrination" period and the apparent shifting of roles by "subject" and "experimenter" make for greater difficulties in the scheduling of subjects (the chief cause of the lowering of our N from an original 40) certain avenues of investigation are suggested. In addition to studying the behavior of subjects surreptitious— ly, such a procedure seems to allow for considerable mani— pulation of the experimental situation not otherwise pos- sible. It also makes possible the study of problem-solving behavior when the subject has no recourse to an experimenter (authority figure, ets.,) for guidance. Finally, we feel that social processes of a fairly broad nature may be studied by experimental analogy, as is the case with loyalty to and defection from a belief-system. SUMMARY A theoretical distinction was drawn between two kinds of defection from and loyalty to a belief-system. Dogmatic defection was said to refer to defection due to the indivi- dual's relative failure to integrate the separate beliefs of the belief system into adequate problem solutions. In- tellectual defection was said to refer to defection in which, by virtue to relatively good integration, the individual per- ceives contradictions between beliefs, and comes to reject the belief-system as invalid. Dogmatic loyalty was said to refer to loyalty to an authority, with relatively poor inte- gration of the beliefs (ideals, etc.) represented by the authority, whereas intellectual loyalty was said to involve relatively good integration of the beliefs with less depen- dence on the authority for solutions to problems. This dis- tinction was then related to the concept of dogmatic thought processes, and to previous findings of the relatively poor integrative thinking of high scorers on a dogmatism scale. It was hypothesized that low dogmatic subjects would tend to defect intellectually from, and high dogmatic subjects tend to defect dogmatically from, or remain loyal to, an "impossible" experimental belief-system. The Denny Doodlebug problem, a complex task of problem-solving requiring the integration of beliefs for a solution, was employed, with logical contradictions making a solution impossible. An -51- unusual procedure was employed, in which each subject ad- ministered the problem to another person, who was actually a second experimenter in disguise, and who sought to stimu- late the process of defection by challenging the reality of a problem solution. The following results were obtained with respect to specific hypotheses: l. (a) None of our 26 subjects defected dogmatically. (b) Six of the 12 high dogmatic subjects defected intellectually. Since four of the low dogmatic subjects found another, ingenious solution, they were included in the analysis of Hypothesis V, leaving 10 low dogmatic sub- jects, of whom seven defected intellectually. This outcome has a probability of occurring by chance of .29. Although this result is in the predicted direction, it does not meet the customary level of significance for acceptance of the hypothesis relating a low score in dogmatism to intellectual defection. 2. There was no significant difference between high and low dogmatic subjects in time to defect intellectually. There was no significant difference between high and low dogmatic subjects in number of statements made during the experiment indicating either "loyalty" to the belief in a problem solution, or the entertaining of the possibility of defection. 5. Six of the 12 high dogmatic subjects remained "loyal" to a belief in a problem solution. Since 11 of the low -52- dogmatic subjects either defected intellectually (seven) or found a new solution (four) only three of the 14 re- mained dogmatically "loyal." This result is significant at the .12 level of confidence, which does not meet the customary criterion for rejection of the null hypothesis, but is clearly in the predicted direction. 4. Irrespective of dogmatism scores, there was a positive relation between integrative behavior in the study of "party— line change" reported by Cram (7) and integrative behavior in the present study. This difference is significant at the .01 level. It was felt that the primary reason for failure to obtain dogmatic defection lay in the lack of authority ac— companying the "pseudo-subject's" challenging of the vali- dity of the problem, and the relatively greater authority, represented by the original experimenter, supporting the maintenance of a belief in a problem solution. As a result, subjects tended to remain noncommittal during the experiment. This was felt to account also in part for the inability of three judges to agree on the implications of 45 of the 110 statements obtained from subjects during the experiment relative to a belief in a problem solution. It was concluded that a definite relation between in- tegrative thinking in one situation and intellectual defec- tion in another, and between non-integrative thinking in one situation and dogmatic loyalty in another, were es- tablished. -55- It was suggested that the methodology employed offered numerous possibilities for studying the effects of mani- pulating certain aspects of the problem-solving situation which thus far have been relatively unexploited. -54- Table.I Intellectual defection and dogmatism score. Intell. Def. No Def. Total High Dog. 6 6 12 Low Dog. 7 5 10 p equals .29 one—tailed test. -35- Table II Time to defect and Dogmatism score. Dog. (high or low) Time (minutes) Dog. (high or low) Time L 10 L Non-Def. I! ll 15 15 20 I! I! N u 21 25 50 50 50 30 A 50 50 II I! mmmtflmtflbt“ mmHHHHmtflHmr-flt“ Total: 22 subjects, excluding four "new solution" subjects. T equals 125. T' equals 105. -56- Table III Implications of verbalizations and dogmatism score. "Loyal" "Def." Uncert. Total High Dog. 24 10 2 56 Low Dog. 21 6 2 29 No significant difference. Table IV Dogmatic Loyalty and dogmatism score. Non-integrative thinking Integrative Thinking Total (Dog. Loyalty) (Int. Def.) (New Sol.) High Dog. 6 6 O 12 Low Dog. 5 7 4 14 p equals .12 one-tailed test. -57- Table V Integrative thinking ("Party-line change" study) and intellectual defection. Integr. Non-integr. ~Integr. (Def. or new solution) 15 4 Non-integr. (No def.) 5 6 Chi-square equals 6.58. Significant at .01 level. Two-tailed test. Total 17 9 l. 2. 5. 6. 7. 10. ll. -58- References Taylor, Don. W., and McNemar, 0.W., Problem Solving and Thinking. Annual Review of Psychology, 1955. 6, 455-482. . Rokeach, M., The nature and meaning of dogmatism. Psychol. Review, 61, 5, 1951, 205. Rokeach, M., McGovney, W., and Denny, M.R., A distinc- tion between dogmatic and rigid thinking. J. abnorm. soc. psychol., 51, 1, 1955, 87-95. Vidulich, R., The integration of multiple sets into a new belief system. M.A. Thesis, Michigan State Univer- sity Library, 1956. Mikol, Bernard, Personal Communication on ongoing research. Swanson, Th., Personal Communication on ongoing research. Oram., A., A Theoretical and empirical analysis of party line thinking. M.A. Thesis, Michigan State Univer- sity Library, 1957. Rokeach, M., Political and religious dogmatism: An alternative to the authoritarian personality. Psychol. Monographs, 70, 18, 1956 (whole number 425). Fisher, R.A., Statistical Methods for Research Workers. (6th ed.) 1956, Edinburgh: Oliver and Boyd. White, 0., The use of ranks in a test of significance for comparing two treatments. Biometrics, 1952, 8, 55-41. Koons, P.B., Personal Communication on ongoing research. Date Due Demco-293 "IT'WLTJ‘EWL'WWW