“‘1— h “7‘— _t___ 7"?“ '5- \H“~.. ‘r‘h-‘p - ‘ U . |"'L ._. uro.‘ D .l‘,; "I‘I ‘ ' In ... pvt. l\.u'- 'Jt '! .‘Olll ‘I‘\‘1‘| I‘xl‘f'fl: w', -o fl .-4-. . l‘.’ 0.03. . v.0 ut- 'vu: |_‘Ca.ln . “;€(‘ 'r ‘fkyu‘. 7.11 SOME EFFECTS OF TABOO COMMUNMRON ON EE‘fiERPERSONAL RELATIONSHIPS Thesis for the Begree of M. A. ECHIGAN STATE, UNEVERSITY iGS-E REBEN JARA ' I 1973‘ -~ ‘ .‘ . D - ‘ ‘ . . ‘ I . . - . . . r . ,‘ ' . . , - , . . , . . - - . - . . . . - . .. - u . . < . O I O . . > I . .‘r . . . . . O V n I d | . . ‘ . - . _ . . . - - u b . ' ‘ I ' v I I ’ ‘ ' ’ r . ‘ - . . v .. I a . : o - . . . . I g \ ' . ’ ‘ ‘ o ' — . . - 1 . ' . _ . —' . p . _ . . ' .. I 1’ .' . . t‘ . . . r - l I . I. . C — v . ‘ I ,, . . . , I . ' ' ' ‘ ' ' . . . - a . . a ' . l ' . , _ . . "' I " v ' .. . a . - - - n a . , v7 '9. - ' ' ; . . . . 0 I O I l . v o , ‘ - ' '4 a. . v ’ . v ' . ' I ‘ ' - .0 - . o . . r c . 0 ,r ' - . I ' -- 0 r c. . a I - , , , . - . .I - . ‘ . r ' . .'-n- -f..'na' ' - ,1 . - . . - r , . - . I ’0 r ' I I . l - O r ' J ' O 1 ' U - . ‘ r u - . ' I V . t ',a' ’ L I ’ J . - . _ c r ' I . 1 . 4 _ . . . 4 ,-, . . . . . 0 pic I r n. ' p o 1- '.. I’ I I ‘ ’1 I .1 'I b . ‘ ‘ 1 ‘ . ‘ l ' ‘ I . i I I . a ' 'n‘... ’ § ' l ‘ 0‘. - . - . _ a 4 I - ’ , - -- . I ’ . ‘ l t .1 . “ . ' . J _ ‘ ‘ f o — o . - -. O - -0 O . .. - . , . o . . . . - - . , v ‘I -.. . . .— I .’ . .-. - . .- O’ -l v .. ‘. o... .‘ .u a‘ O. . — . o .,..-. , _ ‘ v z u - v . n . .I. r o -"l' . . I ‘. .«‘ '0 4. I. a "a -_ f ’ o - . . ' o , . . - - .‘, . . . -- . . . - . ‘, . - . . . .. . - o - o o - . .- . u .- . - . , . r . . ...o. . .. . .. . .'- . . .-- . .. . -9 ‘ C ,‘J,, n _,. .C , - . - . . .., 4 .- .- . . -— . _"o . . '-o ._ .- .- '1‘. a; . on - 'l-‘. .,'-, -.'J “ ‘lul. '7' A ' ' l —. .l'n .. o ' v If/IW//II/l/l/////I////I/I/////I//I///l/I/////I////LW 3 1293 10421 7512 >~ ray-+1": 4' 812'. wwwwfik» S *W. ..-.. ”EMA..- : r ‘. vn y E ' WY"; F -.‘, 3 Y' I '~ ‘ fl} L"W hat. , = «:31 . v. in {y (Trig 1 I: 7i" ' ‘._. *‘L'EL v I - ‘ J's. . (‘4! q J T ‘ A a w 3 , '5 s". *‘ 'xiL a ’7 ‘ A! gig-5W Mien? .. '- 4":- z; .. «EL-w“, LJ '4'; . _ L __—n<_79-\ .‘ W- M 8095_B_w.3.8'1..'gg- ‘ MSU LIBRARIES as. \— RETURNING MATERIALS: P1ace in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. NOV 1 3 2000 ABSTRACT SOME EFFECTS OF TABOO COMMUNICATION ON INTERPERSONAL RELATIONSHIPS By Jose Ruben Jara This thesis is the report of a research undertaking which had the follow- ing objectives: (1) To study the effects of taboo conuunication on interpersonal rela- tionships, along four main dimensions: Safety-credibility, competence-credi- bility, perceived hounphily , and interpersonal attraction. (2) To explore the nature of possible relationships between taboo com- nunicaticn and (a) nonverbal patterns of response, and (b) reciprocated dis- closure. Taboo camunication was defined as that category of message transfer in which the messages are perceived as extremely private and personal in nature because they deal with socially proscribed behavior or because the trans fer of such messages is socially inappropriate. Because there are social sanc- ticns normally associated with taboo conmunication, we can assume that ante- cedents of taboo conuunication are either: 1. Ignorance of the taboo nature of the conuunication; lack of social graces; intent to "shock" or offend; or 2. A belief by the conmmicator that his message will not be construed as any of the above, but will be accepted as honest information about himself. Jose Ruben Jara Since the latter form of taboo commnication may be seen as typically disclosed mly to friends , it indicates to the receiver that he is liked and trusted by the source. Hence, such comnmicatim my be assured to be a re- ward for the receiver. Further, Homan's notion of "distributive justice" (1961) and Jourard's "dyadic effect" (1971) lend support to the hypothesis that such rewards are likely to be returned. In Homan's words, ". . .outcones exchanged will tend to be of conparable value. " In the present study, it was hypothesized that sources disclosing taboo infomatim would be perceived as acre trustworthy, less competent, mare hamphilom , and better liked. Further, considering the high saliency of taboo nessages, these effects were expected to be maintained over time. In order to test these hypotheses , the following experinent was conducted: Each of 29 female subjects interacted with a nale confederate for a series of five meetings over a 2-1/ 2 week period. Each of these meetings lasted appro— ximately 12 minutes. Nineteen of the subjects were randomly assigned to the experimental treatment , in which during the fourth session , the confederate disclosed that he was an ex-convict. The remaining ten subjects were assigned to the control group which did not receive any taboo information in any of the sessions. The dependent variables were measured with questionnaires filled out by the subjects at the end of each session. Statistically significant results were found concerning the predicted immediate effects of taboo communication on safety-credibility and perceived hotmphily. However, it is likely that these significant relationships were not due to exposure to taboo communication, but to methodological limitations . 'Ihe hypotheses concerning competence-credibility and interpersonal attraction were not supported. Jose hiben Jara These results were discussed in terms of possible shortcomings of the present study regarding (a) sample size, (b) lack of control in the experi- mental sessions, and (c) the possibility of biases introduced by the con- federate such as his liking for the subjects. Finally, it was reconuended that fixture research undertakings should be aimed at the development of a scale of degree of "tabooness" for dif- ferent topics, and the exploration of possible effects of taboo communi- cation cm comnmication networks , nonverbal patterns of response , inter- personal trust, and a nurber of other areas of human interaction. d //? 727* SOME EFFECTS OF TABOO COMMUNICATION ON INTERFERSONAL RELATIONSHIPS By Jose Ruben Jara. A THESIS submitted to Midhigan State University in partial fUlfillment of the requirements for the degree of MAS'I'EROFARI‘S Department of Cbnnunication 1973 To my wife, Maye, who helped me to carry out the basic functions of the graduate student life: "Think, wait and fast." ACKNW I wish to express my gratitude to Dr. Everett M. Rogers for his advise and encouragement far beyond academic matters . Dr. Gerald R. Miller and Dr. Charles K. Atkin, also neubers of my Guidance Conmittee, were nost helpful when their assistance was needed. I also want to acknowledge my appreciation of the help received from my colleagues David Bender, Akiba Cohen, Wayne Cromh, James Innowski, Paul Gitschlag, Steve Lodle, and Mark Steinberg. II III IV TABLE OF CONTENTS THEORETICAL FRAMEWORK AND HYPOTHESES Introduction . . . . . . . ............ 'Dflxx>(kmnumication; A.Conceptualization . . . . The Present Study: Rationale and Theoretical Hypotheses . . . . . . . . . . . . . . . . . . . IMETHODODOGY Design . . . . . . . . . . ............ Pretest . . . ........ . ......... The Sample . . . . . . . . . . . . . . . . . . . . The Cbnfederate . . . . . . ......... . . $fimgmdfimwww ....... ....... The pric . . . . . . . . . . . . . . . . . . . . The Sessions . . . . . . . . . . . . . . . . . . . Operationalization of the variables . . . . . . . Factor.Analysis of the Dependent Measures . . . . RESULTS Hypothesis 1 . . . . . . ...... . . ..... Hypothesis 2 . . . . . . . . ........ . . . Hypothesis 3 . . . . . . . . . . . . . . . . . . . Hypothesis H . . . . . . . . . . ........ . The EXploratory Phase . . . . . . . . . . . . . . SUMMARY AND DISCUSSION SW C O O O O O O O O O O l O O O O O O O O 0 Dis ms 81011 O O O O O O O O O O O O O O O O O O O 0 Suggestions for Future Research . . . . . . . . Implications fbr the Practice of Communication . . iii 28 33 3H 35 35 38 no an '46 u7 TABLE 2-1 3-1 LIST OF TABLES Four Factor Rotation Solution . . . ............ Sumnary of Findings for All the Dependent Variables when Comparing the Means of the EbcperinentalGroupatTu-T3................ Smmary of Findings for Safety-(kedibility whenConparinnganchatT3andTu . . . . . . . . . . . Summary of Findings for Conpetence-Credibility menOonparingfixandUcatT3andTu.. ...... Sammy of Findings for Interpersonal Attraction when ConparingfixanchatT3andTu . . . . ....... Summary of Findings for Perceived Honophily whenConparinnganchathandTu . . . . . . ..... Summary of Findings for All Dependent Variables . . . . . . iv 27 33 3a 31% 35 no LIST OF FIGURES 26148.3. 351%- 2-A The Experimental Room for the Present Study . ....... 15 3-A Mean Scores on Safety-Credibility .......... . . . 29 3-B Mean Scores on Competence-Crebility ............ 30 3-C Mean Scores on Interperscnal Attraction . . . . . ..... 31 3—D Mean Scores on Perceived Honophily ............ 32 APPENDIX LIST OF APPENDICES Text of the taperecorded Instructions Given to theSubjects........ . ......... Questionnaire........... ...... Five Factor Rotation Solution: Rotated Factor Millgs O O O O O O O O O O O O O ......... Means and Standard Deviations ......... Interoorrelation Matrix . . ..... y ........ vi #9 SO 55 56 57 Chapter 1 THEORETICAL FRAMEWDRK AND HYPO‘IHESES Introduction An open and honest commmication anong human beings is a deep concern of our age. This concern is clearly reflected in the expectations of stu- dents entering introductory courses in conmunication , in the proliferation --as well as the goals-~of thousands of sensitivity and human relations geups , in the writings of scholars from many different disciplines , in the political campaigns of top-ranked politicians, and in the advertising efforts of many commercial organizations. The professional concerns of counselors, psychologists and psychiatrists, physicians and priests , and the interests of interviewers , lovers , spouses , and bosses , also indicate the need for a better understanding of honesty and openness* as processes of hmnan interaction. Therefore, the need to know the deep and hcnest side of our fellow conmunicators is important in many spheres of our culture . Taboo commi- nication is an avenue to enhance openness in a commicaticn situation. The purpose of the present paper is to formalize and systematize the notion of taboo commication. The paper has the following objectives: (1) To study the effects of taboo communication on interpersonal relationships , along four main dimensions: Safety-credibility, *By "0 e334“— we mean an individual's readiness to disclose honest and truthful inforuation, regardless of the nature of such information. competence-credibility, perceived honophily, and interpersonal attraction. (2) To explore the nature of possible relationships between taboo conmmication and (a) nonverbal patterns of response, and (b) reciprocated disclosure . * Taboo Ccmmnicat ion: A Conceptual ization Taboo commication is that category of massage transfer in which the messages are perceived by both sources and receivers as extremely private and personal in nature because they deal with proscribed behavior or be— cause the transfer of such messages is socially inappropriate . Farberow (1963) notes that the word "taboo" has two different sets of meanings. The first is "sacred" or "consecrated." The second is "uncanny," "dangerous," "forbidden," or "unclean. " The second set of meanings should be used as referents for the word taboo in this thesis. Farberow (1963) cites some other definitions of taboo as something "forbidden by tradition or social usage or other authority; strongly disapproved as conflicting with conventions or settled beliefs, often anong a particular class. . ." It is "a(n).. .interdiction laid upon the usage of certain things or words or the performance of certain act ions . " This conceptualization of taboo comnunication has the following im- plicaticns: (l) The tabooness of a message is two fold: the massage can be taboo be- cause the behavior stated in its content is a taboo (e.g. , "I have *This exploratory part of the study is not the main focus of this work; therefore, only a partial report and discussion is included in Chapter III. been engaged in incestuous practices"), or because the transfer of such message in a taboo (e.g., "My wife is a great sex partner"). Obviously there is a close relationship between the two aspects of taboo commmication. For example, it is difficult to conceive of a message with a taboo behavior stated in its content that is non-taboo in its trans- fer. chever, it is possible to have messages that although dealing with usually acceptable behavior, are harmed from open and free commnication. For instance , sexual behavior among married couples is hardly a proscribed behavior, however, conmunicating about such behavior is not socially appro- priate. Thus, all messages dealing with taboo behaviors are also taboo in their transfer. However, the reverse is not necessarily true. A taboo topic fosters tabooness about its transmission, resulting in decreased probability of its transmission. This decreased probability of transmission will tend to reinforce the taboo nature of the topic itself, which in turn increases the tabooness of its transmission. Of course, this circle can be broken at two levels. First, one can attenpt to reduce the tabooness of the topic; this is very difficult since societal norms change only slowly. * Second, one can attempt to reduce the tabooness of the transmission, and in doing so, a counter—process starts: The more people conmunicating about the message, the lesser the degree of tabooness of that message, which in turn breeds higier levels of commication, and so on. Rogers (19720) mentions an example of this strategy of breaking the taboo- ness of transferring a massage: *Among certain subpopulations (e.g. , students) a change in norms may occur more rapidly. Under usual conditions in India, the decision for vasectomy is very private. Individuals may fear the ridicule of their friends and neighbors , because the individual perceives that he is taking an act not fully sanctioned by the norms of his comm- nity. The taboo on vasectomy was overcome in a highly successful campaign in Ernakulam District in South India in 1971, in which over 63,000 sterilizations were performed in one month. The taboo was overcome, in part, by creation of a festival spirit. The local mass media were employed to convey the message that 'everybody's doing it'.‘ Door—to-door personal contacting of eligible men was concentrated in one comnunity at a time to help create the inpression that the vasectomy adoption decision was a popular one. Adopters were marshalled to travel in groups to the Operating theaters at the Ernakulam City Hall . They marched together, singing songs about family planning and dis- playing signs. The gay festival spirit on the part of the many thousands of adopters helped break down the perception of vasec- tomy as a taboo, and thus contributed to the surprising success of the Ernakulam campaign, which has since been replicated in 25 additional districts in India. (2) Messages vary in their degee of tabooness along a continuum of "very taboo" to "slightly taboo." For instance, incest is a much more taboo topic than abortion in the U.S. (3) The degree of tabooness of a message is determined by the perceptions of the source and receiver. These perceptions generally arise through cultural descriptions and gescril>tions (" 'Fuck' is a bad word-do not use it") which are taught during socialization processes. The socializing agents--parents, teachers, etc.--may provide category descriptions of taboo words ("words relating to sex and defecation are taboo") or they may specify particular words which are taboo ("Shit," "fuck," etc. ). Proscriptions tend to be explicit and unqualified; therefore, the message that one should not say certain things is clear to everyone. I There may be some variance in the degree to which different people will perceive the "tabooness" in a given word. This variance may result from different people's experiences in regard to their relevant person's react- ions to taboo conmunication. For instance, a child's estimation of the degree of tabooness of a given word may depend on the degree to which his parents seemed disturbed at the mention of the word, and/or the degree to which they punished him for using it. (1+) Because there are social sanctions normally associated with taboo com- munication we can assume that antecedents of taboo conmmication are either: i l. Ignorance of the taboo nature of the communication; lack of social graces; intent to "shock" or offend. 2. A belief by the canmmicator that his message will not be construed as any of the above , but will be accepted as honest information about himself. The following scheme identifies some of the variables that can be con- sidered as having a significant influence on the degree of tabooness of messages: * 1. Cultural norms. _I\I_o_ru_§_ are defined as the most frequently occur- ring pattern of overt behavior among menbers of a particular culture. The norms of a culture determine , to a great extent, whether a certain message is taboo or not for that particular culture. For instance , while virginity is not a taboo topic in the U.S. , in Mexico (or many other latin American countries), it would be considered highly taboo. “*Although no claim is made in regard to the exhaustiveness of this list, the mast inportant categories of variables are believed to be included. 2. Societal norms. m a given society, a hierarchy of tabooness can be observed. Further, messages can be categorized in such a hierarchy in terms of their reference to behaviors that are illegal, immoral, or con- trary to strongly held norms . Societal norms emphasize where a certain message is placed along the tabooness continuum. 3. Circumstantial variables such as time, place, and participants. For instance, the message "I have V.D." is not a taboo topic between a physician and his patient. However, such a message would be more taboo when discussed between father and son, and especially taboo when other mem- bers of the family are present. Moreover, such communication is likely to be definitely harmed from the usual dinner table . Further, an issue generally perceived as taboo at time may be much 1 less so at timez. For instance, divorce, which was considered a highly taboo behavior 20 years ago, is much less so in 1973 in the U.S. 1+. Referent of the message. One can distinguish at least two levels: (a) Self referent. The conmmication source discloses information re- garding his personal involvement with the taboo topic; for instance , a woman who discloses that she has had an abortion. (b) General referent. The content of the message only discloses an interest in, or some degree of knowledge about , the general nature of the taboo topic; for instance, an individual who states a position regarding his support of abort ion reform legislation. Generally, the more personal the level of reference to the topic , the more taboo the communication. 5. Individual differences. The way taboo communication has been con- ceptualized implies two dimensions. First, taboo communication has a "socio - lOgical" dimension since it is the society that marks a certain topic or behavior as taboo. Secondly, there is a "psychological" dimension since these taboo topics or behaviors are tempered by the individual ' s percep— tions and personality factors. For instance , while some women feel free to publicize that they have had an abortion, others will tell only their most intimate female friend. By the same token, while some individuals feel highly embarrased when confronted with the disclosure that the other person has had an abortion; sane others might react in a very casual manner. One of the most important variables in this respect might be the individual's general tendency toward oppenness or self-disclosure. * 6. Semantic variables: The connotations of the words employed in messages are sometimes reinforoers of the tabooness of those messages, thus preventing social change. An example of such inhibition of a change is provided by Rogers (1972c): In India, the word 'French Ieather' (condom) had a negative connotation because it was perceived as an object used by soldiers with prostitutes to prevent venereal disease, thus inhibiting the use of condoms as a means of family planning. As soon as the word was changed to "Nirodh," which had a neutral connotation because it was not previously included in the common lexicon, the semantic change facilitated communication and millions of condoms were sold. In smmary, it can be concluded that the degree of tabooness of a mes- sage is a perceived variable affected by cultural, individual (idiosyncratic), and situation-specific variables. The Present Study: Rationale and Theoretical Hypotheses TWO of the most influential theories of social interaction for present purposes are those offered by Homans (1950, 1961) and Thibaut and Kelley *Jourard (1971) defined self-disclosure as "the process of making the self known." (1959). Both theories seek to explain social interaction in terms of its payoffs or outcomes . It is our contention that taboo communication is one such outcome. Most taboo communication is higmly private in nature, confidential in tone , and anxiety—producing. To communicate about a taboo topic is to take a perceived risk: That the receiver may disclose such information to others , with consequent dangers to the original source . Since such information may be seen as typically disclosed only to friends , it indicates to the recipient that he is liked and trusted by the source. Hence, such a communication may be assumed to be a reward or positive outcome for the receiver. "The posi- tive outcomes which one may gain from a social interation are varied and often very subtle" (Worthy _et. al_. , 1969). Another basic principle of social exchange is that among peers, the outcomes exchanged will tend to be of comparable value. This tendency toward fair exchange. is referred to by Homans as "distributive justice," by Gould— ner (1960) as the "norm of reciprocity," and by Jourard (1971) as the "Dyadic effect. " Hypothesis One: In summary, we propose that when an individual becomes the receiver of a taboo communication, he will feel trusted by the source. Fur- ther, the receiver will feel the need to reciprocate such trust in kind, thus , trusting the source. Formally stated , H1: Individuals receiving taboo communication will tend to evaluate the source of the information as having greater safety-credibi- lity than do individuals who have not received such information from the same source. credibility is defined as the degree to which a communication source or channel is perceived by a receiver as trustworthy and competent. This de— finition, derived from Hovland and others (1953) , was somewhat modified by Berlo gt. _ai. , (1970). Although Berlo gt. _ai. proposed a third "meaningful and statistically independent dimension for evaluating message sources: dynamism," this dimension was not included in the present analysis because (1) there were no theoretical reasons to expect any change, (2) as warned by Berlo it: _a_1_. , (1970), there is the strong possibility that this dimen- sion "would not be psychologically independent of the evaluations on the other two dimensions. " Following both Hovland and Berlo, safety—credibiliy is defined as the degree to which the source is perceived as objective, honest, and unselfish. Watzlawick g. a1 . , (1967) advanced the axiom that "Every communication has a content and a relationship aspect such that the latter classifies the former and is therefore a metacommmication. " What has been proposed pre- viously is that in taboo messages, the relationship aspect of the commmi- cation is both great in value and rewarding. The higher the degree of tabooness of the message-content, the higher its perceived metacomrmmica- ticnal value--and therefore the more rewarding. However, this content/ relationship dichotomy suggests the possibility of an interaction between the content and the relationship aspects of a message . Such interaction is particularly relevant in taboo communication since the content of taboo topics by definition deals with proscribed be- havior. Therefore , one confronts a situation where the receiver may react negatively toward the content and positively toward the disclosure of such content . lO Hypothesis wao: This leads to the hypothesis that individuals, react- ing to the relational aspect of the message, will tend to evaluate sources as being higher in safety-credibility (e.g. , I trust you because you trust me by being open and honest with me). However, when reacting to the con- tent aspect of the message, H2: Individuals receiving taboo communication will tend to evaluate the source of the information as having lower competence-credi- bility than do individuals who have not received such informa— tion from the same source. Following Hovland 23' g” (1953) and Berlo gt. E” (1970), competence- credibility is defined as the degree to which the source is perceived as knowledgeable , intelligent , and experienced. "If the reception of a taboo communication is indeed a social reward, the receiver of such communication should be expected to react in a manner consonant with those principles that have been found to govern other types of social exchange" (Worthy 312' all. , 1969). One expectation about social ex- change is that greater rewards are associated with greater interpersonal at- traction. Thibaut and Kelley (1959) pointed out the close relationship be- tween sociometric choice and the ability and willingness to provide positive outcomes. One who provides positive outcomes tends to be liked; likewise , one tends to extend more positive outcomes to those to whom one likes. If, as suggested here, reception of a taboo communication serves as a positive outcome, it is to be expected that liking and taboo communication will be positively related. * Hypothesis three: H3: Individuals receiving taboo communication will tend to be more interpersonally attracted toward the source of the *It should be noted that interpersonal attraction also might facilitate taboo communication , so that the relationship among the two variables would be mutually reinforcing. 11 information than individuals who have not received such informa- tion from the same source. Integpersonal attraction is defined as the degree of favorableness of attitude toward another person . Jourard (1971) noted the importance of perceived interpersonal similarity to the area of self disclosure. On the basis of the kinds of persons sub- jects indicated they disclosed themselves to, Jourard concluded that ". . .dis- closure of the self is a byproduct, among other things, of the perception or belief that the other, the target person is similar to the self." Rogers (Forthcoming) also pointed out the usual occurrence of this "sim— ilarity" phenomenon in the field of family planning comnication. He states: "Taboo communication is more likely to occur among homophilous individuals." These statements suggest a certain degree of directionality and time- order: First homophily, then taboo communication. In the present research, the expected order of the variables has been reversed in order to explore the possibility of creating a certain degree of perceived homophily through taboo communication. H”: Individuals receiving taboo communication will perceive the source as being more homophilous than do individuals who have not received such information from the same source. Perceived homophily is the degree to which one individual perceives a- nother individual as being like himself. Chapter II momLOGY To study the effects of taboo communication and therefore test the hypotheses presented in the previous chapter, the following study was de- signed. Each of 29 female subjects interacted with a male confederate for a series of five meetings, over a 2 1/2 week period. The first four meetings lasted twelve minutes , and the fifth seven. The meetings were scheduled twice a week in either Mondays and Wednesdays, or Tuesdays and Thursdays. The subjects were told that they were interacting with another randomly- assigned subject of the opposite sex. Upon arriving at the experimental room, both subject and confederate were informed of the nature of the experiment and instructed about their task (see Appendix A). The task consisted of the twelve minutes interaction , and, at the end, filling out questionnaires containing the measurement in- struments for the main dependent variables (Appendix B). By plotting the group mean on each of those variables for each of the five sessions, a graphic representation of change over time was obtained (See Figures 3-A, 3-B, 3-C, 3-D). Ten of the subjects were randomly assigned to the control group, while the remaining nineteen were exposed to the experimental stimulus, consist— l2 13 ing of the disclosure by the confederate that he was an ex—convict on parole (who went to prison because of embezzlement). The disclosure in the experi- mental group occurred about midway during the fourth session. The control group did not receive this taboo information. Pretest All the procedures previously outlined, as well as the detailed des- criptions presented in this chapter, were pre-tested with three subjects. All of the subjects were exposed to the experimental stimulus. The Sample In recruiting the subjects, a delicate problem was confronted: They had to be reliable people who would commit themselves to come five times fer approximately thirty minute periods. They also had to be prompt so that a tight sChedule of experimental sessions could be followed. To fUrther complicate the prOblemu neither academic nor economic rewards were available to the subjects. Subjects were 30 unmarried female undergraduates enrolled in introduc— tory communication classes at MiChigan State University. Of the initial 30 subjects, five did not come for the first session and had to be replaced. These subjects as well as those who could not arrive at a specific time on certain dates were asked to return on Fridays to make up fer the lost ses- sion. A.positive outcome of this procedure was the negligible drop-out rate; only one subject did not want to continue in the study after two ses- sions because of other personal committments. The final sample consisted of 28 white and one black college females with a.mean.age of 19.2 years. 114 When recruiting the subjects, they were asked to participate in an "interpersonal communication study in which we are trying to explore how relationships among people develop over time." No mention was made about self disclosure or taboo communication. The Confederate The confederate was a 29 year—old senior in Communication studying at Michigan State University for one term. For control reasons , it was necessary to use only one confederate. Of course, this posed a heavy burden on him since he had to deal with 15 sub- jects four days a week starting at 8:30 a.m. and finishing at 1:30 p.m. This schedule only allowed for twenty minutes per subject: Three for instructions , twelve for the interaction, and five for the confederate to relax and prepare for the next subject. Setting and Procedures The interaction took place in an office located on the fifth floor of South Kedzie Hall on the Michigan State University campus . The subjects were asked to report initially to an office located in the second floor of the same building, where the confederate reported also. Then , both were brougit to the experimental room where trey heard a tape of the instructions (Appendix A). After the interaction had taken place , they were taken to different offices: The confederate to an office in the fifth floor, and the subject to an office in the fourth floor. In this office the subject was asked to fill out the questionnaire corresponding to that session. * *All this maneuvering was intended to prevent the encounter of two or more subjects and the confederate. 15 The experimental room contained a desk with a Chair on each side, a couple of bookcases full of'boxes with stationary, and a stack Of boxes set against the outside wall directly in.front Of the subject (see Figure 2-A). One of these boxes contained a video tape camera. The box had a hole smaller than a dime but large enough to let the lens record the upper half Of the subjects' bodies. This video equipment was used tO record fOur'minutes Of interaction Of session IV during the confederate's disclosure Of the taboo infcrmation. The wires and cables that went out through the windows were not visible to the subjects , and neither was the video tape recorder operated from another office. The video tape equipment used for this experiment was: (1) One Ampex video tape recorder model VR5100 (1" tape), (2) one Sony video camera.model CVC—2100A with a 50mm lens, (3) one Fanon intercome system model classic lN—2, and (it) one Contel T.V. monitor (18" diagcnal). At the end Of the fifth session the subjects were thoroughly debriefed about the nature and the Objectives Of the experiment. NOne Of the subjects had any serious complaints, and some Of them expressed their interest in the results of the present study. fitm1an.ethical point Of view, the most serious issue involved in the experiment was the videotaping Of the subjects. They were infcrmed.of the existance Of their tape and asked if such tape could be analysed. NOne Of them were Opposed. Moreover, some Of them who showed interest in seeing their films were invited for individual presentations during the following weeks . l6 ‘r—---. P 3 h" windows I boxes I D C re .,._. : 5*: l I I L.__‘ I éonfederate g: E D A m 7 ' V‘ _fisubject : door 0——.—- Figure 2-A. The Experimental Room for the Present Study 17 The Topic In choosing the topic Of disclosure for the experimental group, two criteria were followed: In the first place the topic had to be strongly disapproved and conflicting with conventions or settled beliefs . Secondly, although high in its degree Of tabooness, the topic had to be credible enough so that the subjects would not be suspicious Of its legitimacy. Following these guidelines, the topic Of imprisionment due to embezzlement and the subsequent parole Of the confederate , was chosen. Immediately after session four (as well as after all the other sessions), the subjects responded to the open-ended question: "What dO you consider to be the most valuable information your partner gave you today?" A con- tent analysis Of the responses from the experimental group shows that 16 out Of the 19 subjects remembered and wrote down the taboo message. This finding gives some indication Of the high saliency Of the taboo message and the lasting impression it produced. In order to get some idea about the different feelings elicited by the taboo message, a specially designed questionnaire was administered tO the subjects at the end Of session 5, after they had responded the usual questionnaire containing the measurement instruments for the dependent variables. The first question asked "How intimate or personal do you think is the following informat ion that your partner gave you: That he is an ex- convict."* The subjects answered this question by rating the information *"That he is an ex-convict" was handwritten in order to give the sub- jects the impression that the production Of this statement was not pre- arranged. 18 on a 7 interval scale: Extremelyprivate 7:6:5:H:3:2:l Notintimateor andpersonal "’ " " - “ - — personalatall The mean rating Obtained from the 19 subjects was i = 5A7, which yields some indication Of the moderate tabooness of the message. In question 2, the subjects were asked to express their feelings at the moment Of receiving the taboo information. The question was Open-ended and a content analysis Of the responses indicated that the main reaction elicited by the message was surprise. All the feelings expressed by the subjects in answering this question are summarized below: Feeling Number Of people who mentioned this feeling Surprise 1 Curiosity Uncomfortableness Flattery Suspicion Closeness to the other person NO reaction l—‘l—‘www-lrm The next question was: DO you think the disclosure Of that information changed your image of him? TO this question, 17 subjects answered "ro" and 2 answered "yes." When asked how the taboo information changed their image Of the confederate, one Of those two subjects responded that the information made the confederate more interesting as a person, while the other subject said she trought the confederate was a "better person for saying it." Question 1+ was: After your partner's disclosure, do you think you changed your own disclosure toward him (Openness toward him)? Seventeen subjects answered "ro," and two answered "yes," a little. In question 5, subjects were asked if they felt trusted because Of the confederate's disclosure. The answer was given on a seven-interval scale 19 with anchor points Of "NOT TRUSTED AT ALIr-VERY MUCH TRUSTED." The group mean (N=19) was 5.2. Finally, a group mean (N=19) Of 5.6 was Obtained for the question: Did you trust your partner after his disclosure? The semantic differential where the question was tO be answered was "VFW MUCH SO--NOT AT ALL." The Sessions The reasons for including five sessions in the experimental design were: 1. 2. The study would approach "real life" as mch as possible. The confederate could develop a good rapport with the subjects . The taboo communication would seem to be in place and normal, as Opposed to a two-session study in which subjects would probably be suspicion of such an intimate information being released so soon. A rough measurement Of the initial processes Of interpersonal evaluation (regarding the dependent variables) could be Observed Lnder somewhat controlled conditions , such as keeping constant the target Of evaluation 1. e. , the confederate , as well as the infor- mation released by him. The initial fluctuations Of the dependent variables dLe tO inter- persoial evaluation-re-evaluation processes mentioned before would be somewhat more stabilized after three previous sessions, and, therefore, the effects Of taboo communication could be more reliably Observed . Session I. The goal during the first session was to acquaint subject and confed- erate at a superficial level. As expected, the subjects did not follow an 20 identical pattern Of interaction. Nevertheless , the interactions did not differ alomg critical dimensions such as the degree Of intimacy or tabooness Of the information exchanged. It should be stressed that the information provided by the confederate was identical across subjects. Every attempt was made to standardize the mode in which the confederate presented the information about himself through— out the experiment . The information provided by the confederate during the first session was: I am 25 years Old, a senior in Communication at the University Of Colorado. I am here this quarter to study under the gui- dance Of a faculty member and plan to return next year to do graduate work. I am interested in the field Of diffusioi Of innovations . I am a veteran Of the Navy where I served for four years. My favorite sport is cross—country skiing. This Term I am taking a class in Chinese history, a class in Romance film where we study famous directors such as Fel- lini , and an independent study. I live in Denver, Colorado with my family, who are from Italy. I own a window—cleaning business. Obviously this information was scattered during the session , following the flow Of the cmversatim. Session II During this session the subject and the confederate elaborated upon two main topics: Hobbies and vacation plans . The particular information given out by the confederate was: I intend to spend the coming Christmas break in Aspen and Denver , Colorado. I enjoy racing bicycles and play- ing squash. My main hobbie is to read all kinds Of liter— ature but especially non-fiction. The author I like the most is Albert Camus. 21 Session III The main fOcus Of this session was the different attitudes toward the issues brought up by the presidential campaigns such as presidential pre- ferences, legalization Of abortion, the Vietnam1war5 the Congressional race , etc . Session IV The session started with the confederate taking the initiative and asking the subject to state her desires and goals fer a jOb in the present or fUture. He probed the subject Of what she considered would be the near: strains and satisfactions Of such a jOb. Following the subject's response, the confederate told all the subjects about his plans Of'becoming a univer— sity faculty member, and then, fer the control group, he elaborated on how difficult it was to reach that goal because Of the rigorous qualifications one has tt>meet to become a faculty member. Fer the experimental group, he stated the difficulty Of reaching such a goal because Of the handicap Of being;an expconvict. Since this session was being videotaped, it was possible to Observe the reactions Of the subjects toward the disclosure. For the subjects who probed the confederate on the taboo issue, the response was that he had been in prison fOr’two years because Of embezzlement and that at the present time he was on parole. For the subjects that did.not probe on this issue, the session finished.along the same lines Of conversation about’workerelated matters. Session V The purpose Of this session.was twofOld: (l) to Obtain a delayed mea- surement on the dependent variables, (2) to have some time to debrief 22 the subjects. Accordingly, the time was limited to only seven minutes Of interaction. Although the session was left basically Open, for the most part, the subjects talked about it being the last session and their percep- tions Of the study. The confederate ' 5 comments were neutral , and he only elaborated on information received in prior sessions . Operat ionali zat ion Of the Variables The independent variable , taboo communication, was basically Operation- alized as the message: "I would like to become a miversity professor but I will probably have some problems in getting a job because I am an ex—con- vict." As for the dependent measures Of safety— and competence-credibility , the following adjective pairs" were drawn from the Berlo, _e_t_. _a_1_. (1970) study on source credibility: l. SAFETY : safe-unsafe; just-unjust; kind-cruel; friendly-unfriendly; honest-dishonest 2 . COPPEI'ENCEI: trained-Ln'trained; experienced-unexperienced; skilled- unskilled; qualified-unqualified; informed-uninformed These terms served as the anchor points on a seven-interval scale in which the subjects were asked to rate the confederate by marking the interval closest to their feelings. The scales were scored so that a total score Of 35 for any Of the dimensions would be the most extreme positive rating, and a total score Of 5 would be the most extreme negative rating. In the actual questionnaire given to the subjects (see Appendix B), the safety and corpetence scales were alternated and randomly reversed. 7"Berlo 31;. Q. (1970) report these adjective pairs as being "the most representative Of each dimension Of source credibility. " 23 Interpersonal attraction was operationalized as the score Obtained from indexing the following four items: 1 . I was attracted by the other person . 2 . I disliked the other person. 3 . I wouldn't want to interact with the other person again. 1+. I would like to have the other person as a friend. These items were evaluated by the subjects in terms Of the following scale : SA -— Strongly agree with the statement . A -— gage with the statement. N -- Neither agree nor disagree with the statement. D -- DISEEE with the statement. SD -- Strongly disagree with the statement. The positive or negative phrasing Of the statements was randomly deter- mined. The values assigned tO this response scale ranged from 1 to 5. The 1 representing a strongly unfavorable attitude , and 5 a strongly favorable at— titude. A value Of 3 was given tO the neutral statement. Using a similar procedure for evaluation and scoring, perceived homgphily was operationalized as the score Obtained from the rating made by respondents to the item: "I felt the other person was like me." Factor Analysis of the Dependent Measures In. order to test the extent to which the measurement instruments for the dependent variables were indeed measuring different variables , a factor analysis was performed on the data gathered at time”, right after the taboo communication had taken place . The analysis involved a principal axis solu- tion followed by two varimax rotation analyses with Kiel-Wrigley criterions 21¢ set at four and five.* The means and standard deviations for each item, as well as an intercorrelation matrix for all the items , are presented in Ap— pendices D and E, respectively. The four factor solution (Table 2-1) showed that the prime loadings Of the items were basically clustered around the expected dimensions: Cotpe- tence-credibility (F1) , interpersonal attraction (F2) , safety-credibility (F3), and perceived horophily (Fa). However, the competence and safety di- mensions Of credibility were not well structured. For instance, the items Experienced-Unexperienced , and Qualified-Unqualified , that should have loaded on F1, loaded on F3. In the four factor solution, F1 (competence-credibility) accounted for 19 . 8 percent Of the variance, F2 (interpersonal attraction) explained 21 . 6 percent, F3 (safety-credibility) explained 33.3 percent, and F” (perceived horophily) accomted for 10 percent Of the variance,“ yielding an 81+.9 total percent Of variance explained. In the five factor solution (Appendix C) the prime loadings for the item Qualified-Unqualified shifted from F3 to F1. However, the higher load- ing for the item Experienced-Unexperienced remained in F3 . In this solution , it was also Observed that two interpersonal attraction items ("I was attracted by the other person" and "I would like to date the other person") shifted their prime loadings constituting the additional factor. This shift can be ”The five factor solution was performed in order to explore the possi- bility Of increasing the "purity" Of the factors Obtained in the four factor solution, and, at the same time, Observe what would be the composition Of the new factor (which turned out to be physical attractiveness). “This small percentage Of variance explained might be due to the fact perceived homophily was measured with only one item: "I felt the other per- son was like me," and hence was an unreliable measure. 25 Table 2-1. Rotated Factor» Loadingsa (N=29). Fbur Factor Rotation Solution: Factor Loading ItemP __ r1 F2 F3 Pu h2 lam-won: (398) -313 135} 133 808 2JUST-UNJUST 191 -038 LEE" (270) 798 31CIND-CRUEL 291 -278) 3338* 252 897 uFRIENDLY—UNFRIENDLY (389) -163 239* 273 881 SHONEST-DISHONES‘I' (335) -212 912* 198 887 STRAINED-UNTRAINED 139}: -175 (977) 130 907 7EXPERIENCED-UNB(PERIB\ICED (9_19_) -372 705* 069 809 BSICILlED-UNSICILLED 1333* -165 (502) 197 911 9QUALIFIED—UNQUALIFIED (£01) -208 899* 079 897 Norman-0111mm 3111* -231 (597) 179 897 11I was attracted by the other person -098 {21* (995) 278 819 121 liked the other person 328 -§§g_* 392 (991) 892 131 would want to interact again 298 -58_l_&_* 313 (951) 732 1'*1 would like his friendship (520) -_6_ll* 293 398 831 151 liked my partnerc (929) -16_2_* 092 281 899 151 would like to date himc (090) -gt_t_0_* 088 -093 902 171 felt the other person was like me 138 -252 (322) 800* 828 variance Explained by the Factor 19.39%, 21-55% 33-27%310-18% the underlined values represent Where the prime loadings were expected. *—_' Prime loadings ( ) Second.highest loading : Decimal point omitted Items 1-5 correspond to safety-credibility; 6-10 to competence-credibility; 11-16 tO interpersonal attraction; and 17, to perceived homophily. It Should be noted that although these two items were factor-analyzed, they were not included in the index Of interpersonal attraction; ratheru they were separately analyzed in the exploratory part of the study. 0 26 interpreted as a dimension Of interpersonal attraction more directly related to physical attractiveness . The fact that several items did not have very pure loadings , led tO a further analysis Of the intercorrelation matrix (Appendix E) , which showed rather high zero—order correlation coefficients between the vast majority of the items. Nevertheless, careful analysis Of this matrix indicates that the items corresponding to each factor did generally yield higher correla— tion coefficients among themselves than with the items in the other factors . Therefore , they were indexed so that each item was equally weighted, accord- ing to the initial operationalization Of the variables . All these results taken together , point toward the unidimens ionality Of the four dependent variables that can be considered as cosymptoms Of an intimacy-friendship syndrome in interpersonal relations . This interpreta- tion Of the data is consistent with the findings Of Newcomb (1956) who re— ports that perceived similarity is related to attraction , and both Of them to trust. Chapter III RESULTS In the analysis Of the data, there were two types Of change that could be Observed. First, the difference between the mean scores Of the experi- mental group on a given dependent variable at two subsequent times (e.g. , D321.” - 2T3). Second, a conparison between the mean difference scores for the experimental (Dx) and the control (DC) groups. The hypotheses advanced in this thesis seemed to call most closely for the latter type Of test. However, for computational purposes, both tests were performed.it Table 3-1 shows that no statistically significant t values are found for any Of the dependent variables when comparing change on the mean scores for the experimental group at time,4 (T9) and time, (T3). Table 3-1. Summary Of the Findings for All the Dependent Variables When Comparing the Means of the Experimental Group at Tu-T3. Dependent Variables Safety- Corpetence- Interpersonal Peroe ived Credibility Credibility Attraction Homophily REM(N:19)3 29.18 29.05 19.83 2.88 253(N:19)b 29.18 29.28 19.89 2.57 s_.c 0.82 0.89 0.92 0.28 Dx Sfio 2.19 1.89 0.92 0.10 UPI -0.00 -0.21 -0.26 0.10 t value (--)0.00ns (-)0.32nS (-)0.33nS 0.39n8 28 Table 3—1. Summary Of the Findings for All the Dependent Variables When Comparing the Means Of the Experimental Group at Tn-T3' (con't) Mean scores for the experimental group at time . Mean scores for the enqoerimental group at timeg. Standard deviations for experimental and control groups . Mean difference scores. .QOU‘DJ Figures 3-A, 3-B, 3-C, and 3-D, depict the mean scores for the experi- mental and control groups on each Of the dependent variables at each measure- ment time. Note that higher scores indicate more positive evaluations. Empirical Hypothesis One: Mean difference scores on safety-credibility between T9 and T3 will be significantly higher for the experimental group than for the con- trol group. This hypothesis is tested by a t-test for difference between two sample means, Dx and be at Tu-Ts (Table 3-2). The t value is 2.19, which is greater than the 2.05 required for significance at the 5 percent level. Thus, Hypothesis one is not rejected. Note, however, that as Table 3-2 shows, while no change occurs in the means for the experimental group , the mean for the control group drops by -l.2 points. Thus, it is concluded that the significant difference found is not due to the predicted effects Of taboo communication. *(Footnote From previous page) In order to compute the t-test for dif- ference behween two sample means , we had to calculate the t-test between two correlated sample means in the first place. Safety-Credibility Scores 31 29 27 25 23 29 30.1 p 29.5 29,2 29.2 +- or 28.3 28.7 [A P 27.7 __E7\;8 ’/ ‘,’*".----' \\ I 25 9 ” \\ I ‘ .”z \\ I, x’ “~ , 28.5 P b :i n 1 EL _p_ 0| T1 T2 T3 Tb, flf'l'fz, ' " -Control group Figure 3-A. Mean Scores on Safety-Credibility Experimental group Competence—Credibility Scores 31 29 27 25 23 3O r P 29.7 29.0 29.2 29.0 429.2 e ll, 28.7 I I / L 27.1 I, 28 5 ’7’A\‘ I, o” ‘ a’ \ ’ /’ \\\"’ e ,’ 28.3 I p I I I I I I t ,’ I 23.9! I ,ar I A J I _‘g I 0! T1 T2 T3 T9 T5 Figure 3-B. Mean Scores on Competence-Credibility Experimental Group ----COntrol Group Interpersonal Attraction Scores 31 15.5 15.9 L 15.0 _ 19.8 _ 19.8 19.2 _ 13.8 L \ ‘~13.7 13.9 _ ’ 1’ n . . 1 . 0| T1 T2 T3 Tl... T5 Figure 3—C. Mean Scores on Interpersonal Attraction Experimental Group ---—-Cbntrol Group Perceived Homophily Scores 32 L 2.5 2.68 b I 2.97 e \2.3 r ) . i I 1 1 O I T1 T2 T3 Tu IIQS Figure 3-D. Mean Scores on Perceived Hemophily Experimental Group ----- Control Group 33 Table 3-2. Summary Of Findings for Safety-Credibility When Comparing Dx and Do at T3 and Tuna Standard Mean Difference Deviation Group Mean Score gig 25 19 ii t Value Experimental Group - X(N=l9) 0.62 29.16 29.16 0.00 Control Group - C(N=10) 2.19 27.80 26.60 -l.20 1.20 2.19* *P= .05, df=27 Empirical Hypotheses T\wo: Mean difference scores Of corpetence-credibility between T and T will be significantly lower for the eiiperimeital group than those for the control group when comparing Tu'TS' Following the same statistical procedures mentioned before, t is 1.21, which is smaller than the 2.05 required for significance at p=.05, df=27. Thus, hypothesis two is rejected. As shown in Table 3-3, the predicted de- crement in the mean score Of the experimental group does occur and results in the negative Dx, at Tu-T3. However, such difference is very small. More- over, Dc decreases more sharply than Dx, thus, yielding a positive t value Opposite to the predicted direction. aThe sane format will be need in presenting the tables in the remainder Of this Chapter. 39 Table 3-3. Summary Of Findings_for Competence—Credibility When Comparing Dx and Do at T3 and T”. _S_D: T3 In E t value X(N=19) 0.69 29.26 29.05 —0.21 C(N=10) 1.89 27.10 26.30 -0.80 +0.59 1.21 ns Fflpirical typotheses Three: Mean difference scores Of interpersonal at— traction between Tu and T3 will be signifi- cantly higher for the experimental group than those for the control group. Table 3—9 shows that t is 1.52 which is smaller than the 2.05 required for significance at p=.05, df=27. Thus, hypothesis three is rejected. It is interesting tO Observe that the relative strength Of the t value is mostly due to the acute drop in the mean score for the control group as compared with the slighter drop Observed in the experimental group. Table 3-9. Summary Of Findings fgr Interpersonal Attraction When Comparing Dx and Do at T3 and T '4' _S___D_ T3 In E t value X(N=19) 0.92 19.89 114.79 -0.10 C(N=10) 0.92 19.50 114.10 -0.50 +0.90 1.52 ns 35 Epirical Hypotheses Four: Mean difference scores Of perceived homophily between T9 and T3 will be significantly higher for the experimental group than those for the control group. Table 3-5 shows that t is 2.29 which is greater than the 2.05 required for significance at p=.05, df=27. Thus, hypothesis four is not rejected. This result, although significant in the predicted direction, should be cautiously accepted since perceived homophily was measured only with one item. Thus, the measurement is probably unreliable. Table 3—5. Summary Of Findings fog Perceived Homophily When Comparing 5x and Do at T3 and T9“ EEE 13 I9 I? I_XE£EE. X(N=l9) 0.28 2.58 2.68 0.10 C(N=10) 0.10 2.50 2.90 -0.10 +0.20 2.29* *p= .05, df=27 The Exploratory Phase Besides the systematic analysis presented previously , and given the very recent and tentative formalization Of the field Of taboo commnication, it was ttought that an effort should be made to explore the effect Of taboo commmication on a nnmber Of variables. 1. Nonverbal Patterns Of Response: The possibility Of the existance Of certain nonverbal patterns of behavior especially designed tO regulate intimacy levels in interpersonal relationships is intriguing. Since there 36 are no theoretical bases to formulate any kind Of predictions, it was decided to unobstrusively videotape the subject-confederate interaction when the taboo communication occurred, and then, on the basis Of the video tapes, de— velop a coding scheme that would provide the first rough approximation in the search fOr nonverbal patterns Of response tO taboo communication. UnfOrtunately, the lack Of control on the timing fer the delivery Of the stimulus (taboo message) as well as some other problems Of control (e.g., some Of the subjects were chewing gum) made it very difficult to reliably code the interaction. Nevertheless, preliminary analysis Of the tapes show some traces Of recognizable nonverbal pattern fOllcwing taboo communication. This pattern consists in the momentary freezing Of facial gestures and in- creased eye contact, fOllcwed by a greater usage Of adaptors* such as scratch— ing different parts Of the body, wiping the eyes, caressing the hair, etc. , and finally, changes in body posture. 2. Reciprocated Disclosure: The finding that individuals tend to fOl- low patterns Of behavior set by the person with whomnthey are interacting, is consistent throughout the interpersonal attraction literature. Jourard (1971) notes that "disclosure begets disclosure," Hemans (1950) speaks Of "distributive justice," Gouldner (1960) reports that individuals reciprocate liking, etc. Following this line Of thinking, it was thought that taboo communication might also foster or enhance reciprocated taboo communication. Accordingly, the fOllcwing two questions were asked to the subjects in the questionnaire administered after each session: Waiters are nonverbal markers which originated in the satisfaction Of self needs. Over time, however, they become part Of the individual '8 habit repertoire. In social situations, adaptors are triggered Lnder circumstances Of stress or anxiety (c.f. , Harrison, Forthcoming). 37 l. "I felt the other person made himself known tO me." 2. "I felt I made myself known to the other person." The mean scores for question 1 were Zr” = 3.79 and 2T” = 3.52, yielding a positive difference score Of D = 0.26. Further, the results for question 2 showed a decrease in the .mean values from 2% = 3.89 to XI); = 3.79, thus yielding a negative difference Of D = -0.10. In other words, while the subjects did perceive the confederate as being more disclosing after the taboo information was released, they did not per- ceive themselves as being more disclosing toward him. Thus, we conclude that the expected reciprocity Of disclosure did not occur. Chapter IV SUMMARY AND DISCUSSION The Objectives Of this thesis were: (1) To study the effects Of taboo communication on interpersonal rela- tionships , along four main dimensions: Safety-credibility, competence- credibility, perceived homophily , and interpersonal attraction. (2) TO explore the nature Of possible relationships between taboo com- munication and (a) nonverbal patterns Of response , (b) reciprocated disclo- sure. Taboo communication is that category Of message transfer in which the mes sages are perceived as extremely private and personal in nature because they deal with socially proscribed behavior or because the transfer Of such messages is socially inappropriate. Because there are social sanctions normally associated with taboo comnnication, we assume that antecedents of taboo cormunication are either: 1. Ignorance Of the taboo nature Of the communication; lack Of social graces; intent to "shock" or Offend. 2. A belief by the communicator that his message will not be construed as any Of the above, but will be accepted as honest information about himself. Since the latter form Of taboo communication may be seen as typically disclosed only to friends, it indicates to the receiver that he is liked and trusted by the source. Hence, such communication may be assured to be a 38 39 reward for the receiver. Further, Homan's notion Of "distributive justice" (1961) and Jourard's "dyadic effect" (1971) lend support to the hypothesis that such rewards are likely to be returned. In HOman's words, "...Outcomes exchanged will tend to be Of comparable value." In the present study, it was hypothesized that sources disclosing taboo infOrmation would be perceived as more trustworthy, less competent, meme.homophilous, and better liked. Further, considering the high saliency Of taboo messages, these effects were expected to be maintained over time. In order to test these hypotheses, the following experiment was con- ducted: Each Of 29 female subjects interacted with a male confederate for a series Of five meetings over a 2-1/2 period. EaCh of these meetings lasted approximately twelve minutes. Nineteen Of the subjects were randomly assigned to the experimental treatment, in.Which during the fOurth session, the con- federate disclosed that he was an execonvict. The remaining ten subjects were assigned to the control group whiCh did.not receive any taboo infOrma- tion in any Of the sessions. The dependent variables were measured with questionnaires filled out by the subjects at the end Of each session. Statistically significant (p=.05) results (see Table 9-1) were fOund concerning the predicted immediate effects Of taboo communication Of safety- credibility and perceived homophily. HOwever, it is likely that these sig- nificant relationShips were not due tO exposure to taboo communication but to methodological artifacts. The hypotheses concerning competence-credibi- lity and interpersonal attraction were not supported. '40 Table 11-1. Summary Of Findings for All Dependent Variables . Dependent Variables Safety- Competence- Interpersonal Perceived credibility Credibility Attraction Homophily Mean Difference Scores for experimental group (mu-7T3) Dx(N=l9) -0.00 -0.21 -0.10 0.10 Mean Difference Scores for control group (Wu-n3) Dc(N=lO) -l. 20 -0. 80 -0 . 50 -0 . 10 t values 2.19“” 1.21 ns 1.52 ns 2.29* *p= (.05, df=27. As for the exploratory part Of the study, some traces Of a recognizable nonverbal pattern Of response to taboo communication was Observed . This pat- tern consisted in the momentary freezing Of facial gestures and increased eye contact, followed by an increase in adaptors and, finally, changes in body posture. In regard to the question Of reciprocity in taboo communica- tion, no positive results were Obtained, at least under the circumstances Of the present study. Discussion The data presented in Chapter III imply that only one Of the predicted increases in the dependent variables occurred (Perceived homophily) , and even in this case, the increase was so small that it only reached signifi- cance when coupled to a simultaneous decrease in the control group. In 1+1 regard to the other variables, interpersonal attraction as well as competenci— credibility decreased, while safety—credibility stayed constant. These relative decrements of all dependent variables could be interpre- ted in the light of Newcomb' s findings on interpersonal attraction. Newcomb (1961) explains a similar decrement in roommate attractions by suggesting that initial estimates of attraction are autistic. That is, because of a need or desire to be compatible in a roommate situation, individuals tend initially to over-estimate the favorableness of the relationship. These overestimations are later modified by reality, causing a decrease in esti- mates of attraction. This same argument could be extended to the present study where the desire to please the experimenter, the initial strain of being in an experimental situation, the outgoing personality of the confed- erate , etc. are possible factors contributing to an initial overestimation of the confederate' s credibility and attraction. Some other factors are also believed to have contributed to the failure to find more significant results. First, there were procedural limitations such as the discovery by some of the subjects that the confederate was in- teracting with some other person besides them. The effects of such dis- covery upon the subjects ' perceptions of the confederate are uncertain. Though it might have been advisable to discard the subjects who discovered the presence of the confederate , such deletion was not very feasible taking into account fine small number of subjects participating in the experiment. This discussion leads us into the next shortcoming of the present re- search, namely the sample size. The reason for the small size of the pre- sent study was the impossibility of the confederate to interact with a larger number of people . To introduce a second confederate not only repre- 1&2 sented a serious threat to experimental controls but also--in very pragmatic terms--it represented the difficulty of finding another individtal willing to invest at least five hours a day, to obtain a limited academic reward. A more viable alternative to solve the problem of sample size was the use of films or videotapes as stimulus material. However, the usage of mechanical devices would increase the artificiality of the experiment and would take away the value of having an m relationship being developed. The small sample size is also likely to account to a greater extent for the instability of tte means yielded by the control group. The cost of having "real" interactions in the experiment was high. Besides the small sample size, the research also suffered from a lack of control on the information exchanged during the interactions . Although, as indicated in Chapter II, such information exchange was kept as identical as possible across subjects, differences in patterns of interaction indeed occurred and might have introduced considerable amounts of uncontrolled variance. * Further, the small sample size not only made it more difficult to reach statistically significant findings, but also did not permit control for per- sonality variables such as tendency toward openness. And it is likely that such variables might substantially determine the individual's reaction to- ward taboo communication (see Chapter I). A factor in connection with the strength of the taboo information is perhaps in order. Although subjects and confederate interacted for a total time of almost one hour over the five sessions, the taboo information de- fin fact, even the introduction of other taboo messages than the ex- perimental treatment might have cccured. 1+3 livered to the subjects in the experimental group was usually discussed less than 10 seconds (except for a couple of subjects who discussed it for about two minutes). Further, it should be remembered that the degree of tabooness of the message "I am an ex—convict" was necessarily moderate (see Chapter II) . To further complicate the problem, there is the possibility that sub- jects might have been reacting not only toward the taboo information but also toward the confederate's stated desire of becoming a Pln.D. and a faculty professor. The interaction between the two topics might have caused a state of imbalance in the subjects , thus lowering the rewarding value of taboo communication theoretically assumed . Partial support for this contention is shown in the responses to the exploratory questionnaire, presented in Chap- ter II. Another problem which can not be easily dismissed involves the circum- stances under which the messages were given. The videotape, made possible to observe how the variability of moods of the confederate, the receptive- ness of the subject, the flow of the conversation, etc. , influenced the way in which the taboo message was released. Disclosures across subjects varied in intensity, degree of erotion exhibited, speed, tone of voice, etc. , pos- sibly introducing unexpected amounts of variance . This discussion leads to the conclusion that biases introduced by the confederate' s behavior might also be another source of uncontrolled variance . For instance, it was found that the confederate's liking toward the subjects probably affected the evaluations of those subjects towards him. This re- lationship was checked as follows: At the end of the fifth session, the confederate was asked to fill out a questionnaire with the same format as nu those administered to the subjects . Then, an interpersonal attraction index was computed. This index was then correlated with the subjects' index cf interpersonal attraction at time 5 . The correlation coefficient obtained was p=0.l+7. This moderately high correlation which accounts for 22 percent of the variance, may indicate that the confederate 's liking for the subjects influenced, to a certain extent, the interpersonal attraction results ob- tained from the subjects. Still another shortcoming that hirndered the reliability of our results was the multiple administration of the same measurement instruments thus creating problems of sensitization that mnight be responsible for the small amounts of change of the dependent variables . The rationale here is that the subjects may have tried to be consistent with previous responses. For instance , the results in all dependent variables show a narrow range of responses. Finally , it should be made explicit that the situational factors in the present study are important. The findings are limited to situations in which people have a very limited knowledge of one another, and to male-female interactions. Variables such as length of acquaintance, status , or attrac- tiveness of the source, to name a few, should be important in determining reactions to taboo comunication. Suggestions for Future Research In the introduction for his book Taboo Topics , Farberow (1963) states: "When taboos continue or deve10p without useful society-enriching functions or facilitate self-defeating or self-destructive activities, questions should be raised about them." Taboos exist in every culture or society and it is H5 unfortunate that so little work* has been done to understand the communica- tion processes underlying such phenomenon. The first step in fhture study Should be the development of a valid and reliable scale tOTmeasure degrees of tabooness for a.number of topics. Of course, the development of such a scale represents a Challenge, given the basically cultural nature of taboos. TherefOre, extensive testing should be done not only for different populations within a given culture but also cross-cultural validation should be conducted. The lack of suCh a scaling insunument will hamper the development of operationalizations and stimulus material in future researCh. Another area of needed research is the study of the effects of taboo communication upon the morphological and interactional characteristics of communication networks, such as their stability, size, endurability, inte— gration, and openness. This research is particularly important since net- works are the basic structure upon WhiCh social change takes place. A third intriguing area of'researCh is that concerning possible rela- tionShips between taboo communication and nonverbal behavioru It'mfight be interesting to systematically explore (l) the nonverbal patterns of'respgnse to taboo communication, (2) the nonverbal patterns followed by people in order to "£391" some form of taboo communication. For instance, Rogers (1972c) suggests that "receiverS‘who are perceived as (a) trustworthy, (b) competent (to understand), (0) neutral (that is, who do not sanction or disapprove), and (d) confidential, will be more likely to 'drew' taboo CORP *PoSsible reasons for suCh lack of knowledge are: (l) the difficulty involved in the gathering of the data about taboo areas (Rogers, 1972c); (2) the fear that an interest in taboo topics would blight the professional careers of the investigators (Farbercw, 1963). H6 munication messages . " Finally, other areas of inquiry are the effects of taboo communication on: (1) frequency of communication, (2) diffusion of innovations, ( 3) per- suasion, and (1+) usage of interpersonal and mass media channels. Implications for the Practice of Communication Despite the generally negative empirical results reported in this thesis , on the basis of the theoretical discussion about taboo communication, sev— eral implications for the practice of communication follow. At an individual level , it appears that concealment about taboo topics such as sex, body, etc. , is negatively related to mental health (Jourard, 1971; Eagan, 1970). Although this statement is also supported when refer- ing to relatively non-taboo topics, it seems that the higher the degree of tabooness of the concealed information, the higher the degree of tension, anxiety, and maladjustment of the individual. Some forms of taboo communication such as self-disclosure may be re- garded as one means of establishing a bond between two people in which hi gner levels of intimacy are likely to be more rewarding and functional than low levels (Newcomb, 1956). 0 At a group level , the sharing of taboo information among the members of a group can be a good strategy to increase the group's cohesion (Siemmel, 1950), and elicit increased feelings of pertenence and identification with the group (Hagan, 1970). Finally, in a society, increased levels of communication about taboo topics can help to reduce the "tabooness" of such topics, thus speeding the processes of certain kinds of social change. BIBLIOGRAPHY BTBLJOCRAPHY Altman, I., and Haythorn, W. W. Interpersonal Exchange in Isolation. Socio- metry, 1965, 28, L111-1126. Berlo, D. , Lemert, J. , and Mertz, R. Dimensions for Evaluating the Accep— tability of Message Souroes. Public Opinion Quarterly, 1969-70, 33, 563-576. Eagan, G. , Encounter: Group Processes for Interpersonal Growth. Belmont, California: Wadsworth Publishfig Conpany, 1970. Farberow, H. L. (ed.), Taboo Topics. New York: Prentice-Hall, Inc., 1963. Gouldner, A. W. , The Norm of Reciprocity: A Preliminary Statement. American SociOIOgical Review, 1960, 25, 161-178. Harrison , R. An Introduction to Nonverbal Communication (Forthcoming). Homans, A. C. The Human Group. New York: Harcourt, Brace, 1950. , Social Behavior: Its Elementary Forms . New York: Harcourt , Brace and World ,1961 . Hovland, C. I., I. L. Janis, and H. H. Kelley (1953). Communication and Per— suasion. New Haven: Yale University Press. Jourard, S. M. Self-Disclosure. An Enperimental Analysis of the Transparent Self. New YoFk: Wiley, 1971. The Transparent Self. (Revised edition), New York: Van Nostrand Reinhold Co. , I971. , The Transparent Self. Princeton, N. J .: Van Nostrand, 1969. Kelley, H. H. Interpersonal Accomodation. American Psychologist, 1968, 23, 399-”10. McNemnar, Q., fisydnoloiical Statistics. (Fourth Ed. ,). New York: Wiley, 1969. " Newcomb , T. M. The Prediction of Interpersonal Attraction. American Psy- Chologist, 1956, 11, 575-586. , The Acquaintance Process. New York: Holt, Rinehart and Winston, 1961. u7 1+8 ROgers, E. M. Communication Strategies for Family Planning. (Forthcoming), New York, Free Press. , Taboo Communication and Social Change: Family Planning in Asia, and some Suggested Modifications in the Classical Diffusion Model. Paper presented to the Department of Human Communication, Rutgers-The State University, May 5, 1972. Simmel, G. "The Secret and the Secret Society." In K. Wolff (ed.), 'I_'h_e_ Sociology of George Simmel. New York: Free Press, 1960.. Taylor, D. A. The Development of Interpersonal Relationships: Social Pene- tration Processes. Journal of Social Psycholoy, 1968, 75, 79-90. Taylor, D. A., Altman, I. and Sorrentino, R. Interpersonal Exchange As a Function of Reward, and Costs and Situation Factors: Expectancy Confirmation-Disconfirmation. Journal of Experimental Social Psychology, 1969, 5, 325-339. Thibaut, J. W., and Kelley. H. H. The Social Psychologt of Groups. New York: Wiley, 1959. Watzlawick, P., Bevin, J ., and Jackson, D. Pragnatics of Human Communication. New York: W. W. Norton 8 Co., Inc., 196T Worthy, M., Gary, A. L., and Kahn, G. M. Self-disclosure as an Exchange Process. Journal of Personality and Social Psychology, 1969, 13, 59- 63. APPENDIX APPENDIX A TEXT OF THE TAPE RECORDED INSTRUCTIONS GIVEN TO THE SUBJECTS First of all I would like to remind you that you are participating in an interpersonal communication study, in which we are trying to explore how relationships among people develop over time. Your partner has been randomly selected to interact with you during five sessions. Your task today will be divided into two parts. First, you will be talking to each other for approximately 12 minutes. After that , you will be taken to another room where you will be asked to fill out a questionnaire. We encourage you to interact with your partner as NATURALLY AND HONESTLY as you can, so the results obtained from you will be truthful and valid. A final comment. Please DO NOT TALK about this study to anybody else. Your cooperation is this regard will be highly appreciated. Now, before the experimenter leaves the room, do you have any ques- tions? H9 APPENDIX B STUDY ON memo-w. commoner: QUESTIONNAIRE. Introduction: You have just finished the first part of today's session. For the second and last part, we would like you to answer to the questions pre- sented in the following pages. This questionnaire is divided into 3 parts. Each one of the parts contains its om set of instructions. Please READ CAREFULLY the instruc- tions for each part before you answer it, so you krow how to answer the questions for tlat part. If you lnave any questions concerning those in- structions, ask the experimenter to clarify them. Since ALL of the questions we are asking you deal with your percep- tions, attitudes, beliefs, etc. , there are no "right" or "wrong" answers. Therefore, answer them as honestly as you can. RD’IEMBER: your honesty is the only guarantee we have to insure the validity of your results . Answer all the questions. DO NOT leave blank spaces. If you are not positive about how you want to answer a certain question, (because the categories we provide do not fit your perceptions , or for any other reason) mark down the BEST APPROXIMATION to your actual feelings . Firally, remember that your answers are s1rictly confidential and will not be misused in any way. Now, turn the page and begin working. 50 51 PARl‘l Instructions: We are interested in the way people judge other people as sources of messages. Below, you will be presented with a series of descriptive scales. Guided by the general impression that you have of your partner, rate him along those scales. Here is low to use the scales: GOOD X 3 2 1 U— -l 7? -3 If you felt that your partrner was in general an _myextrenel ocd source of information, you would place a check mark in space 3 . quite ocd (but not extremely good), in 2; if 31' t1 cod in 1; if neither _rgrbgd,in0;ifsli t1 bad,in-; quite__cl,in-2,Er'fiif exnrerely bad, in -3. e or neutral space on the scale may also be used for "I don't know" or "I don't think this scale applies" answers. Be sure to put a check mark somewhere along each scale. Put your check within the spaces (as shown above). Put one and only fie check on each scale. I PERCEIVE MY PARTNER AS BEING: UNSAFE SAFE UNIRAINED TRAINED UNJUST JUST EXPERIENCED INI'XPERIENCED CRUEL KIND UNSKILLEID SKILLED UNFRIENDLY FRIENDLY QUALIFIED UNQUALIFIED HONEST DISHONEST UNINFORMED INFORMED 52 PART 2 In this part of the questionnaire, you will be presented with a series of statements about your experiences with the person you just met. For eaCh sentence, respond by writing in front of it the appropriate category of the key: e.g. I felt at ease N key: SAr-if you strongly agree with the statement A-if you 9533;.With the statement N--if you nert er agree nor disagree D—-if you.disag£§e with the statement SD—-if you strongly disagree with the statement 1. I was attracted by the other person 2. I disliked the other person 3. I wouldn't want to interact with the other person again A. I would like to have the other person as a friend 5. I felt the other person was like me 6. I felt the other person made himself known to me 7. I felt I made myself known to the other person Now, in the fellowing line, please assign a number from 1 to 100 to indi- cate how much you liked your partner: . Assuming that you are avai- lable to date your partner, assign a number from 1 to 100 to signify your willingness to accept a date: Open ended question: What do you consider to be the most valuable infOr- mation your partner gave you today: 53 FINAL PART L. How intimate or personal do you thirnk it is the following information that your partner gave you: EXTREMELY PRIVATE NOT INTIMATE OR AND PERSONAL PERSONAL.AT.ALL 2 . Now, please think back very carefully to the moment when he gave you that information; how did you feel? 3 . Do you think the disclosure of that information changed yorr image of him? NO D ms [:1 if yes, HOW: Ll. After your partner's disclosure, do you think you changed your on disclosure (openness toward him). NO D was [:1 ifyes, VERYMUanJ SOMEWHATD LITI'LED NOTATALLD 51+ Did you feel trusted because of your partner's disclosure? NOT TRUSTED VERY MUCH AT ALL TRUSTED Did you trust your partner after his disclosure? VERY MUCH SO NOT AT ALL 555 [U9PTJVIIIXI(3 FIVE FACTOR ROTATION SOLUTION: ROTATED FACTOR LOADINGSa (N=29). Itemb Factor Loading F1 F2 F3 F, F5 h2 lSAFE-UNSAFE (539) -373 §9§f 239 089 859 2JUST-UNJUST (225) 059 £22? 133 208 817 3KIND~CRUEL (329) -203 gigf 186 261 909 ”FRIENDLY-UNERIENDLY (969) —083 259? 209 238 885 5HONEST—DISHONEST (996) -179 zggf 170 191 867 6TRAINED—UNTRAINED 929* -098 (315) 198 239 995 7EXPERIENCED—UNEXPERIENCED (591) -302 626* 038 277 809 8SKILLED-UNSKILLED gggf -061 (369) 186 271 932 gQUALIFIED-UNQUALIFIED gggr -135 (593) 038 237 897 lolNFORMED—UNINFORMED zggé -109 (992) 132 319 898 111 was attracted by the other person 090 -719* 360 (379) ggl_ 867 121 liked the other person 287 -307 (385) 362 gggfi 896_ 131 would want to interact again 195 —190 (398) 186 13;} 850 1”I would like his friendship (387) -292 259 131 zggf 876 lhluenymmmf 2% egg on In E? we 16I would like to date himC 131 agggfi -009 005 (958) 925 17I felt the other person‘was like me 218 -162 299 853*(275) 990 Variance Explained by the Factor 23.33% 11.95% 27.38% 7.78% 18.28% the underlined values represent where the prime loadings were expected. Prime loadings Second highest loadings Decimal point omitted Items 1-5 correspond to safety-credibility; 6-10 to competence—credibility; 11-16 to interpersonal attraction; and 17, to perceived homophily. It should be noted that although these two items were factor-analyzed, they were not included in the index of interpersonal attraction; rather, they were separately analyzed in the exploratory part of the study. on: :: 1 O APPENDIX D PENIS AND STANDARD DEVIATIONS (N=29). Item Mean _S__._D__ lSAFE—UNSAFE 5.68 1.25 2JUST-UNJUST 5 .91 1. 61 3KIND—CRUEL 5.62 1.29 ”FRIENDLY-UNFRIENDLY 5 .96 1. 20 5HONEST-DISHONEST 5 .6 5 1. 39 6TRAJTIEln-UNTRAINED 5.55 1.35 7EXPERIENCED—UNEXPERIENCED 5 .62 1. 32 8SKILLED—UNSICILIED 5. 62 1. 32 9QUALIFIED—UNQUALIFIED 5 . 51 1. 29 loINIORMED-UNINTORMED 5. 79 1. 99 11I was attracted by the other person 3. 03 . 86 121: liked the other person 3.96 .77 13I would want to interact again 3.75 .78 1L‘I would like his friendship 3.79 .97 151 liked my partner 67.37 22.71 151 would like to date him 38.93 28.30 17I felt the other person was like me 3.72 .89 56 APPENDIX E INTERCORRELATION MATRIX* (N=29) Itcm** 1 2 3 4 5 6 7 8 9 10 1 1000 2 663 1000 3 760 744 1000 4 838 720 904 1000 5 822 792 846 849 1000 6 756 546 634 711 737 1000 7 722 595 874 841 748 681 1000 8 744 578 686 751 708 941 733 1000 9 780 677 822 807 822 848 826 847 1000 10 767 724 738 790 787 881 779 891 856 1000 11 601 398 586 444 502 380 511 418 428 433 12 645 494 696 681 671 595 647 647 654 658 13 536 531 609 591 560 533 562 562 546 581 14 556 463 642 597 569 629 684 656 621 674 15 496 306 475 484 472 495 525 489 506 563 16 401 128 315 229 308 302 441 264 294 362 17 524 429 557 552 513 452 481 481 429 508 "Decimal point omnitted “These item numbers correspond to those presented for Appendix C, and D. 57 INTERCORRELATION MATRIX (continued) Item. 11 12 13 14 15 16 17 11 1000 12 691 1000 13 642 802 1000 14 557 741 769 1000 15 544 756 689 819 1000 16 676 552 519 593 728 1000 17 553 639 489 536 484 335 1000 58 ICHIGQN STATE UNIV. LIBRRRIES IIHI llllll llllllHlllllllfllHl 312931042175 2