symbolic self-completion, attempted influence, and self-deprecation

8/12/2019 Symbolic Self-Completion, Attempted Influence, and Self-Deprecation

1/26

BASIC AND APPLIED SOCIAL PSYCHOLOGY 1981 ,2 2) , 89-114

Symbolic Self-Completion, Attempted Influence, andSelf-DeprecationRobert A. WicklundPeter M. Gpllwitzer

University of Texas at ustin

concept of symbolic self-completion states that people define themselves asmu sicians, athletes, etc. by use of indicators of attainment in those activity realms ,such as possessing a prestige job, having extensive education, or whatever isrecog nized by others as indicating progress toward com pleting the self-definition.The self-completion idea postulates that when important symbolsindicators ofself-definitionare lack ing, the person will strive after further, alternative symb olsof the self-definition. In the present research two modes of symbolizing complete-ness were studied: (1) attempting to influence others directly within the self definitional area; or (2) simple self-descriptions of one s own performance, such thatothers would be exposed to those self-descriptions. Two correlational studiesshow ed that the less education (Study 1) or on-the-job experience (Study 2) subjectshad, the more they desired to influence others. Experimental subjects (Study 3)w ho w ere interrupted wh ile writing a positive, self-descriptive statement man ifestedstronger influence attempts than did subjects who were allowed to finish. Finally,Study 4 exam ined self-descriptions as a self-symbolizing effort. Incom pletenesswas varied by means of a salience-of-past-teacher manipulation. When pressurewas placed on subjects to characterize themselves negatively within their respectiveareas, those for whom a positive teacher was salient were the most willing to benegative. The interpretation of these findings assumes that influencing others, aswell as positive self-descriptions, can further the ind ividu al s sense of having acomplete self-definition.

W e are greatly indebted to Dr. Melvin L. Sn yder, whose understanding of theory and experim entaldesign went a long way toward influencing the approaches taken here. We would also like toacknowledge the work of the experimenters Steven Lewis, Cindy Marshall, and Elysabeth Stewart.We are indebted to William Michael Lynn for assisting with the data analysis. Drs. Richard L.Archer, Otto M. Ewert, Dieter Frey, and Peter Schonbach made a number of enlightening commentson an earlier draft of the manuscript. Portions of this paper were presented at the Annual Meetingofthe Eastern Psychological Association in Hartford, 1980.

Requ ests for reprints should be sent to Robert A. W icklund, D epartment of Psychology , Universityof Texas at Austin, Austin, Texas 78712.

89


2/26

9 WICKLUND AND GOLLWITZER

When people refertothemselvesas ph ysicia ns, rally drivers, or guitar-i s t s , it is commonly thought that they possess qualifications correspondingtothese titles. When someone introduces himself asamedical doctorandproceedsto describe the extent of his education, we are normally prepared to infer acertain medical co mp etence. And should the guitarist report that he has influencedothers' guitar-playing styles,we areprone to conclude thathe is a competentor even gifted mu sician. M ore generally, when a person pursuingagiven activitycan talk abouttheself positively,orotherwise attemptstogain more recognitionfor performing that activity,theaudienceforthese wordsandgesturesislikelyto conclude thatthepersonis indeed well preparedandcom petent.Butsuchaninference is often erroneous.A central observation to be made about the human, within thecontext of anotion of self-completion, is that very central flawsin theperso n's trainingorperformance are covered ov er bywhatweshall call self symbolizing behav-iors.Italso follows fromthe idea, described more fully later, thatapersonwhocurrently possesses numerous, durable indicators of competence isunlikelytoengage in self-symbolizing actions.

The person with a great dealofexperience in an activity, for instance, willnot endlessly bring others' attentiontovarious indicatorsorself-descriptionsofcompetence; such a person will carryout the activity withinan atmosphere ofmodestyandu npretentiousness.Onthe other hand,thefrequent useofopenselfaggrandizement, status symbols, know ing m or e than theother, and effortsto influence others may be taken as signsof the individual's insecurity (i.e., incompleteness in thedomainofactivity).

A core assumption isthat indicators (symbols)ofcompletenessaresubstitut-ableforone ano ther.Theperson w hocanpointtosymbols that supportthe selfdefinition aspiredto (e.g., physician) will tendtoneglectthepursuitof furthersymbols. Thus self-symbolizing efforts willbe undertaken when thepersonislackinginsymbolic indicators ofthe statusof phys ic ian, gui tar is t , orotherself-definition.An investigationbyRoss , B ierbrauer,andPolly (1974) illustrates this line ofreasoning. When subjectswho hadservedasteacherto ayoungboyfound thattheyhadfailed toteachhim thecorrect spellingof a fewwords, they tendedtoattribute blame to himespecially if they were inexperienced college studentsrather than veteran teachers. The results suggest that the veteran teachers, arriving

on the scene with a stockpile of symbolsof their teacher status, werenot dependenton oneparticular instanceof successto further their self-definitions.Historical BackgroundThe idea thatthepotential effectiveness of onesymbolic indicatorcan be substitutedfor by analternative symbolic indicatorwasimplicitin thethinkingofLew in (1926) and several of his students. Their analysis of goal-oriented behavior


3/26

SYMBOLIC SELF COMPLETION 9 1

and interrupted activities is the main conceptual background oft ispaper. Lewinargues that when an organism establislies a particular goal, a tension systemcomes into play and remains until the goal is reached or until the organism leav es the fie ld . Further, should the task be interrupted due to outside forces,the tension system will remain intact, and the psychological effects of that tensionsystem will be detectable even if the person cannot resume the activity imme-diate ly. T his rather simple principle surfaces in a highly testab le, operationalizedform in the often-cited work on memory (Zeigamik, 1927), as well as in theanalysis of interrupted-task resumption by three other Lewin students (Lissner,1933; M ahler, 1933; Ovsiankina, 1928).

We have concentrated especially on Mahler's (1933) research and concep-tualizing of the dynamics of substitute activities. In Mahler's research paradigmsubjects typically were assigned 4- 6 fairly simple task s, such as piecing a mo saictogether, or constructing a tower from building blocks. Described in Lewin'slanguage, the psychological process associated with each task was a specifictension system, which abated only when the task was completed. More impor-tantly, if a task were to be interrupted, the tension system was said to persevere,resulting in subsequent voluntary resumption of the same task.Mahler interrupted each subject on a number of the tasks, using a variety ofcreative devices, and following each interruption the subjects were eventuallyallowed to resume the original task. Approximately 90% of the original taskswere resumed, but there is a more significant feature of the research: Followingsome o fth e interruptions M ahler gave her subjects substitute tasks; after w orkingon these they could resume work on the original. For instance, the original taskmight have been to write out a slogan by piercing the letters of the words intoa piece of paper. When this task was interrupted, the experimenter introducedthe substitute task, which consisted of completing the slogan by writing it out

with a pencil. This kind of substitute activity had a decided impact on thesubsequ ent tendency to resume the original task; the resumption rate was sharplycurtailed. In Lewin's terms, the activity was carried to a tension-reducing con-clusion via a task that was not identical with the original but served the sametension system.Thus Mahler points to the concept of substitutability: Given a goal with somefinite, objective quality (such as constructing a tower), a whole class of goalswould seem to reduce the tension system corresponding to that goal. But afurther

observation needs to be mad e, this one stemm ing from Henle (1944 ). Henle wasof the view that the seemingly concrete tasks found in this Lewinean researchwere related psychologically to certain sup ero rdin ate goalssuch as creativ-ity, industriousness, or intelligence. It is argued that the substitute value ofMahler's alternative tasks was so substantial because subjects were orientedlargely toward these self-defining goalsand not exclusively toward the objec-tive goal of to w er com pletion or hav ing written out a slogan by piercingho les .


4/26

9 2 WICKLUND AND GOLLWITZER

Symbolic Self CompletionA theory of symbolic self-completion can now be spelled ou t, using the conceptsof comm itment to go als , symbols of completeness, and social reality.

Commitment to Self Defining Goals. From Lew in's perspective a goal-spe-cific tension remains active only as long as the person is involved psycho logicallyin the pursuit of the goal. Ovsiankina (1928) found that when the task wasimportant personally to subjects, the resumption effects were strongest. Forsubjects who found the tasks to be personally unimportant (i.e., trivial or ar-bitrary ), the resumption rate was so low that one would be inclined to doubt thepresence of a goal-directed tension system.The foregoing leads us to invoke a boundary condition for the self-completionprocess. In all the research reported here we have tried to insure an ongoing,goal-oriented process directed toward a self-definition. Forone,w e have typicallyasked subjects to name one activity area in which they have a central interest(such as participating in a sport, playing a musical instrument, or studying asubject). Second, we have also established that subjects are continuously activein that area.

How does one recognize a self-defining goal, in contrast to other kinds ofgoals? Self-d efm ing means that people predicate of themselves some qualitythat corresponds to a sense of control and capability, such as p ar en t, ath -l e t e , or a rt is t. Further, and central here, is that these goals are defined byinteracting with others (Mead, 1934) and that the sense of progress toward thegoal requires social recognition.Symbols of Completeness. These may be defined generally as indicators of

on e's standing with respectt a self-defining goal that are poten tially recogn izableby others. At a very rudimentary but important level are the simplest self-de-scriptions (e.g., a person teaching at a university introduces himself to an au-dience as sc ien tis t ). Of course, the human is not wholly dependent on thesekinds of open self-characterizations. There are numerous abbreviations for im-mediate social acknowledgment, many of these describable as status sy m bo ls.Th at is , having a diploma from a graduate school is a broadly recognized symbolof the person 's self-definition, and it will propel the person toward a sense ofcompleteness. Similarlytitles, official occupational positions, and membershipin select interest groups are all socially evolved mechanisms for providing theindividual w ith indicators, or markers of possessing an aspired-to self-definition.For the sake of conven ience it is possible to view such symbols as falling intothree groups, being based on: (a) the background experience or training requisiteto the activity; (b) occupying a position or status that furthers the relevantactivities; or (c) performance of the actitself For each of these three categoriesof symbols the person stands to be recognized as possessing the self-definitionthat is sought after. In the case of a the specific symbol could be education,


5/26

SYMBOLIC SELF COMPLETION 9

intheinstanceof b itmightbe theposition vice-p residen t, and for coneisrecognized asbeingam usician, or lawyer, simplyby remaining highlyactive inmusicianlikeor lawyerlike actions.

Social Reality Mahler (1933) found thatthe tension-reduction potentialofa substitute task depended largelyonwhetheror notsolvingthe substitute taskcarried a social reality. The substitute taskhad tension-reducing propertiesonly when thesolution wasannounced to the experimenter. Once othersac-knowledgethepersonforhaving solvedtheproblem, having solveditbecomesa social factandthuscanserveas a self-defining symbol. This lineofthinkingcanbecarried b acktoCooley (1902),whostates that self-definitions cancomeinto being and remain stable onlybyvirtueof theacknowledgment ofothers.Thus the sense of progress toward a self-defining goal is dependent on theacknowledgment of others.We will call this thesocial reality factor We canalso talk aboutaconceptof broad ening social rea lity . Onceapersonhas anindicatorathand, thesenseof completeness shouldbeenhancedto thedegreethatone can inform more people aboutit, ormore generally, enlargethe scopeof individuals who would potentially recognize the completeness of the self-definition.

Research ParadigmsThe core of the self-completion idea is that symbolsofcompleteness a repotcm ialsubstitutes for oneanother. The central implication is thata lack in symbolicsupport will leadtosymbolizing theselfascompletewithin aparticularself-definition. How canthis notionbebroughtto anempirical test? Firstof all, adefinite symbolic lack, over whichthepersonhas nocontrol,has to beestab-lished. Thiscan beaccomplishedbyexperimentally varyingtheextentofcom-pleteness with respectto agiven symbolor by employing existing lacks (i.e.,comparing individuals who differaprioriinrelative completeness). Second,theperson m ust have an accessible meansofself-symbolizing, such as the possibilityof acquiring new symbolic indicators (e.g ., status symbols; prestige affiliations),orbybroaden ing imm ediate social realitythrough social infiuenee,or via self-descriptions promoting theaspired-to self-definition.

In the research to be presented we have provided subjects with a highlyaccessible way to symbolize the aspired-to self-definition. In the first threestudies, subjects were askedto infiuence others withintheareaof the pertinentself-definition. Tothe extent that the aspiring mathematician can conv ince othersof some point regarding mathematics, those others then become part of the socialreality that recognizestheperso n's statusas amathematician. Thus the subjectswere givenapossibilityofbroadeningthesocial realityof their self-definitionthrough infiuencing others.Asimilar effect has been notedbyFulbright (1966 ),Geyer (1977),andIzraeli (1977),all of whom view attempted infiuencingand


6/26


wielding of power as attempts to restore lacking competence or security. InStudy 4 it is shown that self-descriptions, to be acknowledged potentially byothers, can also reflect the striving for self-completion.

STUDYOverviewSubjects committed to a variety of different activity areas (football, mathematics,guitar, etc.) were given a questionnaire that dealt with the following: (1) It wasimportant to know whether the person's commitment, hence motivation to becom plete in the area, was ong oing. This was operationalized in terms of recencyof performance of the activity. (2) Second, relative prior existence of a symbolof completeness was defined operationally through years of relevant formaleducation. Once these background data had been taken, all subjects w rote a shortessay, ostensibly for instructional purposes. They then were asked to indicate:(1) how many different groups of available target audiences they wanted theiressay to be presented to; and (2) how many people they thought should readtheir essay. These two desire-to-infiuence items are the dependent variables. Itwas hypothesized that the less education the person has in the relevant area, themo re desired infiuence should be manifested. How ever, this relationship shouldhold only for individuals whose commitment to the activity is ongoing. Thetension associated with nonattainment should be relatively absent among thosewho are not actively pursuing the activity (Adler Ko unin , 1939; Ke lley, 1951;Ovsiankina, 1928).

The re e x i s t s no r e a dy ope ra t iona l i z a t ion fo r d r a wing th i scom mitted-noncom mitted distinction, as comm itment traditionally has been de-fined through experimentally induced decisions (Brehm Coh en, 1962; Kies ler,1971). Ho wev er, one straightforward definition, based on Le win's idea of leavingthe field, would be recency of the last engagement in the activity. The symboliccompleteness idea, not being a parametric model, does not specify numericallyhow long an activity m ust be ceased before it no longer holds a tension state,but as it turns out, locating a temporal cutting point is not a difficult problem(reported in the following in more detail).

METHODSubjectsThe subjects were undergarduates from introductory psychology classes, testedin one session. Ofthe 219 subjects tested, 66 were eliminated due to acriterionthat each activity be chosen by at least three subjects. Th is criterion w as imposed


7/26


because it was desired to analyze the data of individual activity areas. Thus 74males and 79 females are reported in the analysis.ProcedureThe first questionnaire concerned itself with subjects' background and containedthe following critical items: (1) A request for subjects to indicate their centralactivity (Subjects w ere asked to write down an area of special com petence, suchas a sport, musical talent, or special area of knowledge); (2) Number of yearsand months of their training in the activity;' (3) How long ago they lastperformedthe activity. This item served to define commitment.The second form, attached to the first, consisted of printed instructions towrite a two-page essay on how to get started and interested in the relevantactivity, for the purpose of teaching someone. Subjects were then instructed tostart writing, and to take up as much of the two pages as they desired.

Once this material was collected they received the second que stionnaire, wh ichasked them to indicate explicitly who should be influenced by what they hadwritten, and how many people should be influenced. The following two itemswere included: (1) First came a description of a sample of different target groups( ave rage high school stu de nts , foreign exchange stu de nts , etc.) availableto the psychology department. It was said that the department intended to givethe subjec t's essay to some of these grou ps, but that it would be difficult to handout the essay to every one of the grou ps. In line with this rationa le, subjectswere then asked to check just those groups they thought should read the essaythey had written; (2) Second, they were requested to indicate the number ofpeople who should read their essay within each category checked. For eachcategory the subject could check any of the following four alternatives: 1-5,6-20, 21-60, over-60.

RESULTSThe prediction may be examined in terms of two kinds of correlations: (1) therelationship between am ount of formal education in the activity area and num berof target groups indicated as infiuence targets; and (2) the relationship betweenamount of formal education in the activity area and gross number of peopleindicated as targets of influence. Fo r thenumber of target groupsmeasure, sub-jec ts could receive a score of between 0 and 12. Thequantity of people measure

'Subjects were asked to distinguish between formal education in the school, private instruction,and university edu cation. As the latter two categories contributed very little to total years of educ ation,and as man y of the activities named by subjects were never associated with private instnic tior 'h eanalyses reported here are solely in terms of formal education in the school.


8/26


was analyzed by assigning a 7 to the 1-5 category, a 2 to the 6-20 category,a i to the 21 -6 0 catego ry, and a 4 to the over-60 category.In order to test the hyp othesis, which predicts an inverse relationship betweenthe preexisting symbolic indicator (amount of education) and tendency to infiu-ence for committed subjects only, it is first necessary to distinguish betweencom mitted and noncomm itted subjects. Table 1 divides subjects into comm ittedversus noncommitted, using three different cutting points: 7 days, 14 days, and30 days. The correlations reported are between: (1) years of education andnumber of target groups; and (2) years of education and quantity of people. Aninspection reveals that the exact cutting point seems to make little difference,

in that the correlations for committed subjects are consistently more negativethan are those for noncommitted subjects.^ Note that independently of the cri-terion for commitment, the correlations for committed subjects are consistentlynegative and significant for both indices. The differences between correlationsfound for committed versus noncommitted subjects are significant for the num-ber-of-target-groups index no matter which criterion is used: z = 2.3 6,p


9/26


It m ight also be added that the length of subjec ts essays played no role in thesefindings. When essay length is partialled out of the correlations, none of the r sis changed by a factor of greater than .02 .One also may examine the hypothesis in terms of means. Figure 1 illustratesthe results of splitting education at the median (4 years) and then evaluating theimpact of education separately for committed and noncommitted subjects. Forthe dependent variable of number-of-target grou ps, the interaction between com -mitment and education is significant, F(l,149) = 6.08,p


10/26


Analysis of the Separate Ac tivitiesThe effect of education may also be analyzed within each category of activity,hence the following analysis based on individual activities was performed, usingonly cornmitted subjects. Table 2 lists the 21 categories of activity together withthe Pearson r s computed for each category. One additional category (auto me-chanic5) included in the general analysis is not listed here because it show ed novariation in education level. These r s were combined after the manner of Mos-teller and Bush (1954), also suggested by Rosenthal (1978). Each r is convertedto a Z, and the resulting mean Z is tested against 0 using a conventionalf test.By this method the combined r s are significantly different from 0 for both thenumber-of-target-groups measure, f(20) = 2.76, p .02, and the quantitym easure, f(20) = 2.7 8,p .02.

TABLECorrelations Between A mount of Education and Tendency to InfluenceListed by individual Groups

N(4)(7 )(3)(5)(3)(4)(6)(5)(5)(4)(5)(6)(5)(4)(8)(6 )(7 )(3)(5)(6)(7)

TopicBusinessDanceDramaEnghsh WritingFrenchGuitarMathematicsNatural SciencesOrchestraPhotographyPianoSingingSpanishBaseballBasketballFootballGolfGymnasticsSoccerSwimmingTennis

Number oftarget groups

- . 8 9- . 2 7- . 7 2

.61- 1 . 0 0- . 9 8

- . 8 9.21- . 3 2- . 8 5- . 3 8- . 2 2- . 8 0,31

,21- . 3 6- . 0 1- . 9 8

.59,27- . 4 3

Quantityof people- . 9 1- . 1 6- . 6 9

.64- . 9 9- . 3 9- . 8 2- . 2 1- . 2 4- . 3 9- . 5 2- . 3 3- . 6 0- . 2 5

.35- . 5 7.39- . 7 7

,46- . 1 5- . 3 9


11/26


DISCUSSIONAn A lternative Viewpoint on the Results and a Com menton Self SymbolizingAn alternative way of looking at these effects is from the standpoint that the lesseducated individuals tend either to be naive about the limits of their abilities orare more excited about their activities owing to novelty. Thus it would not bethe lack of support for on e s self-definition per se, but an overzealousness coupledwith lack of self-knowledge that might lead to attempted infiuence. Althoughthe existence of this kind of process seems plausible, there are a few facets ofthe data that do not lend it full support. For instance, the fact of a consistentdifference forthecomm itment variable implies that itisnot just naivety, curiosity,or enthusiasm but, rather, a certain motivation or sense of goal orientation thatis necessary for the effects. Of course, it would have been possible that thecommitted subjects were also the least learned ones; however, the two com-mitm ent groups were practically identical (p > .50) and relatively high (over4 years) in level of education.

A S ubsidiary S tudy with Noninvolved SubjectsW e have proposed that a self-definition is composed of mutually substitutablesymbols, and the data suggest that if a person is short on one kind of symbolicevid enc e, striving for an alternative is to be expected. B ut it would also be usefulto know whether people typically view such symbolic indicators, as we havecalled them, as hanging together in a unit. From the standpoint of our involvedsubjects, it would be ideal if onlookers (i.e., audiences) assume that there is acompleteness underlying the infiuence effort. Otherwise stated, it is ideal forthe involved, incomplete individual if observers would impute high competenceas a function of the actively involved per son s efforts to persuade.

Hilton, GoU witzer, and W icklund (Note1 gave 50 observer-subjects infiuencedata from Study 1. The observers knew that an essay had been written, that theoriginal subject was a journalist (or whatever specific area), and, most impor-tantly, that the original subject wanted to infiuence a certain number of targetgroups and a certain nu mber of peop le. Given this information the observers hadto try to estimate the number of years of education the involved person hadaccumulated.

The data are again correlations between education and the two infiuencevariables, computed within individual activity areas, and this time the patternis opposite to that in Study 1: The average correlation between the number oftarget groups to be infiuenced and estimated years of education is r = .27, andthe correspond ing figure for quantity of people and estimated years of education


12/26


is r = .30 . In short, it would appear as if the inverse relationship betweenlacking in one symbol and compensating by another is to be found only amongactively committed individuals. If the person is uninvolved, whether as an ob-server or by virtue of not being committed, the inverse quality disappears. Thenoninvolved observers go so far as to assume that these symbols fit together ina positively correlated pattern. T his is fortunate for the self-symbolizing person ,in that to be seen as trying to infiuence results in being acknowledged forposse ssing other parts of the whole that make up the self-definition.

STUDY 2The results of the initial correlational study are consistent with the reasoningregarding the inverse relationship among symbols of completeness and are alsoin accord with the notion of commitment that comes from Lewin (1926). Thesecond study was run as a conceptual replication, but with a format that departedappreciably from that of the first. Rather than gathering subjects with diverseinterest areas , subjects w ith one quasi-occupational interest were collected. Th esewere dormitory advisors at a state university. Again, an aspect of their back-g ro un d was se lec ted a s the op e ra t io na l i z a t ion o f r e l a t iv ecompleteness/incompleteness, and their subsequent desire to infiuence was thenassessed. The procedure had the following features that distinguished it fromthe first study: (1) relative lack of a durable symbol w as operationalized in termsof years of experience at a specific job^that of dormitory advisor; (2) com-mitment was defined operationally through subjects' intent to continue in thejob; (3) rather than a diversity of interests, all subjects belonged to the activityarea dorm itory a dv isor ; (4) the dependent measure of self-symbolizing wasthe number of people the subject deemed appropriate to infiuence. In this case,the objects of infiuence were constituted by a more concrete, tangible group( i .e. , undergraduate advisees).

METHODSubjectsThe subjects were 42 undergraduate dormitory advisors, all from the largestdorm itory at the University of Texas at Au stin. Their participation was voluntaryand anonymous.ProcedureAn interviewer contacted subjects individually and left them a questionnaire,which they filled out anonymously. The following critical items were included:


13/26

SYMBOLIC SELF COMPLETION 1 0 1

(1) to operationalize the presence/absence of an existing symbolic indicator,subjects were asked for their number of years of experience as advisor; (2)commitment was assessed by asking subjects whether they intended to continuewith the j o b : ' 'Wo uld you consider being a Senior Resident Ass ista nt? ' '; subjectscould answer either in terms of y e s , po ss ib ly , or n o ; (3) finally, thedesire-to-infiuence item was highly similar to the quantity-of-people index ofStudy 1. Subjects were asked simply to write down the ideal number of peoplethat should be under the direction of one dormitory advisor.

RESULTSSubjects were divided into committed versus noncommitted groups based onwhether they intended to continue with an advising job. Those who checked n o were placed into the noncommitted group, and those who checked either po ss ib ly or y e s were designated as committed. The resulting correlationsbetween years of experience as advisor and quantity of people are as follows.The relationship between the two variables for the 26 committed subjects wasnegative, r = .36,p .05 , as predicted. There was no appreciable cor-relation for the 16 noncom mitted subjects, r = .01.The significant co rrelationfor the committed group approximates in its magnitude the correlations of StudyI .

DISCUSSIONThe data replicate the pattern of results ofthe first correlational study. Subjectswho were lacking with respect to what may be called back grou nd sec uri ty ,in this case lack of experience, evidenced the strongest desire for a broad realmof infiuence. The effect was again found only among subjects whose com mitm entwas on going, consistent with the Lewinean contention (Adler & Ko unin, 1939;Ovsiankina, 1928) that incomplete activities generate a tension state and impelthe person toward completion only if the personal investment has not faded.

STUDY 3In the following laboratory experiment we used the approach of the Lewinianschool (Lissner, 1933; Mahler, 1933; Ovsiankina, 1928) to create an experi-mentally defined incompleteness. In that earlier research, subjects were inter-rupted partway through a task, and their subsequent resumption of that sameactivity was then assessed. These investigators presumed that an interpolatedsubstitute task could lower the tension state associated with the original task,resulting in decrements in its resumption.


14/26


The present experiment is in much the same vein as that early work, although orig ina l ta sk and sub stitute ta sk have a broader meaning within our the-oretical context. Whereas Lissner and Mahler used such activities as puttingtogether cut-up postcards, or building a mosaic out of stones, the activity hereis more definitely ego-involving. The or ig in al task in this experiment, as wellas the su bs tit ut e task, are both conceptualized as potential symbolic indicatorsof completeness. It is assumed, after the manner of Mahler, that the two kindsof task have a dynamic relation to one another. To the extent that one accom-plishes the first task, the second is less necessary. Or stated in terms of thepresent language, to the extent that a person possesses one kind of symbolicevidence, other indicators of that same talent ace less necessary.Subjects in a variety of activity areas (as ih Study 1) were first given theopportunity to create a strong symbol, in the sense of writing a description oftheir involvement in the area, with the expectation that the description wouldreceive an audien ce. In the midst of their creating that sym bol, half of them we reinterrup ted, thu s generating an incom pleteness with regard to that symbol. It wa sexpec ted that these interrupted subjects would be the most likely to take advantageof a subsequent opportunity to infiuence others. This opportunity came in theform of exerting influence on the training of school children, w ithin the sub ject'sown activity area.

The commitment variable was operationalized from a new standpoint. Allsubjects had a chance to infiuence others on a topic identical to their owninterests, and they had an equal opportunity to infiuence people on a topic(painting) well-removed from their own interest areas. As they were not com-mitted to painting, the process central here should operate primarily for thecritical area. In addition, it was possible to operationalize commitment in thesame way as in Study (i.e., recency of last performance), although the numberof noncommitted subjects {N 5) did not allow a separate analysis.

M THOSubjectsA total of32male undergraduates, enrolled in an introductory psycho logy cou rse,participated as part of a course requirement. Subjects were randomly assignedto the two conditions. One subject from the Interruption condition was dismissedbecause his activity area happened to be the same as the co n tr o l activity area(i .e. ,painting; to be described later). Five of the subjects, tw o in the Interruptioncondition and three in the Control condition, were noncommitted by the 2-weekcriterion, and their data were set aside.


15/26


ProcedureSubjects were run individually by three experim enters. Experimenter describedthe experiment as a study of representative college students' interests. The studywas said to take only about half the time allotted for the subject's experimentalparticipa tion, thus he was to assist a graduate student (Experimenter 3) fromthe Educational Psychology D epartm ent during the last part of the hour.

The first study was then described in more detail. It was said that a numberof essays were to be collected, written by college students, and that each essayshould address the history of the subject's interest in his activity area. Theexperim enter handed the subject a two-page booklet with a formal-looking cove rsheet. On the second page the subject found a short statement of consent, andin this context the experimenter explained that a later publication of the ongo ingresearch required some of these student essays for purposes of illustration. Thelength of the actual essay was to be 200 words, which would take up the secondpage of the booklet. Experimenter 1 asked the subject to proceed to write theessay. At this same time, in another location. Experimenter 3, who had beeninformed of th e su bject's activity area, prepared some instructional ma teria lson the subject's chosen topic, to be used in the second phase ofth e experiment.

Interruption Manipulation When the subject had written approximately one -third of a page. Experimenter 2 entered, and the ensuing conversation made itclear that he was engaged in the same study. In the Control condition he leftafter 2 minutes, and the subject continued his essay. Once the essay was com-plete. Experimenter signaled Experimenter3.B ut in the Interruption conditionthe conversation took a different twist: Experimenter 2 asked the first experi-menter what area of interest the present subject had selected, then pretended tobe su rprised and replied that he had already collected three essays on that top ic,which was enough from one area. At that point Experimenter 1 told the subjectthat his essay could not be used and that she would have to discard it. Just asin the Control condition, she then called Experimenter 3, who took the subjectto her room.

The Editing Measure Experimenter 3 explained that she and some fellowgraduate students in educational psychology were trying to create vocational andrecreational interests in young adolescents from a ghetto neighbo rhood . T he ideawas to infiuence these children to pursue some activity independen tly by teachingthem the basic steps for undertaking it. All kinds of activities were ostensiblybeing considered (e.g., sports, music, art, and even academic areas). The ex-perimenter said that a number of students had already been asked to writeinstructions that would assist the adolescents in getting a start in the variousactivities. The experimenter then explained that it had been helpful when other


16/26


17/26


An analysis of variance was performed on the summed words crossed out andinserted. The hypothesis primarily addressed the interruption effect for thecritical essay ( i.e ., the essay relevant to the sub ject s own activity ar ea), andhere the difference is highly significant, r(24) = 4.14, p


18/26


involving phenomenonwewishtotalk abo ut.By thepresent co nception,com-pletenessin an important activity areacan besymbolized in avarietyofways.If a self-symbolizing effort is interrupted, that tension should then carry overinto whatever alternative symbolic routeisoffered. Inthis case that alternative,or substitute symbolic activity, was the desire to infiuence others. And thedifferences inpatternsofdatafor thecriticaland painting essay suggest againthat commitment is a vital issue. The interruption manipulation tends to havea stronger effect on editingof the critical essay, thusit would appear thatthevariableofcomm itmentcan beoperationalizedinalternative ways.Inthis ca se,of course,theconclusion mustnot beoverdrawn,as therelevant interactionwasslightly shortof an acceptable significance level.One m ightbetemptedtodescribetheaforementioned chainofevents in termsof frustratio n-ag gres sion (DoUard, Do ob, Miller, M owrer, Sears, 1939).The interruption manipulation was an ideal empirical definition of frustration,inthe sense of breaking off goal-oriented activity. To complete this analysis,however, wewould haveto assume that rewriting someone else's compositionconstitutesanaggressive act. The editing,bythis reason ing, wouldbeconstruedasan act ofdisplaced aggression againstaninnocent, nonprovoking bystander.Atthesame time, another possible interpretation of the editing measureisthatit refelctsthesubjec t's desire toh lpthe person w ho so clumsily wrote the essay ,thereby simultaneously helping the adolescents from the ghetto^the targe t groupof the essayas well as assisting Experimenter3. But assuming thatour de-pendent variable is a valid index of aggression, the present results wouldbesurprising in lightof the accumulated literature on attempts to find a frustra-tion-agg ression relation. C iting numerous research programsandcom mentariesontheDoUardet al. hypothesis (Bandura, 1969; Buss,1963,1966; K regarman& W orchel, 1961; Zillmann Cantor, 1976), Zillman (1979) concludes thatablocked goal-oriented reaction,by itself, isunlikelytoproduce aggression.Ata minimum, some elementofattack wouldbenecessary (B uss, 1961;Zillmann,1979).Inshort, the present paradigm is not one that should have been appropriatefor inducing aggression.

STUDY

This experiment deals with another sort of gaining social reality for a self-definition (i.e., a positive self-description). To the degree that a person laysclaim to being proficient orgifted in anarea,and isrecognized for that self-description, that instanceof broadened social reality should contributeto self-completion. Butthereis a complementary phenomenonthe reluctanceto berecognizedasincapable, untrained,ornot prepared.The self-symbolizinginthiscase goes in a direction opposite to that we have thus far discussed. Once asocial realityisimparted toanegative self-description, the person loses symbolic


19/26


completeness,and of course there should generally be nodesire todescribe theself negatively. But more importantgiven that people are put under pressureto manifest modesty, failure, or incompetence, those whohave a strong back-ground of existing symbols should be more ready to allow a negative self-description. Their backgroundofsymbols lends asecurity that allows som e self-descriptive liberties.

The present study was designed around this thinking. Them anipulation ofcompleteness differed markedly from theearlier one and consisted simply of asalience variation. Some subje cts attention was brought to bear on the bestteacherin their activity area, whereas other subjects were made tofocus on theworst teacher. F ollowing this manipulation,allsubjects w ere requested to writea self-description, to beshown toothers, indicating that they had performed verypoorly on atest intheir relevant activity area. Their level ofnegativity wasthentakenas the main dependent variable.

There is a fiirther way to examine the incompleteness process within thiscontext. Basedon thestudy byRoss et al. (1974), wehave reason tothink thatteachers with inadequate experience in their professions are more self-servingwhen asked why their students fail.Itwas found that novice teachers were mo relikely to blame failure on a student than were veteran, classroom teachers. Wecould then surmise that experience itself constitutes an indicator of self-com-pletion,andthat for anexperienced person, thepossibility of describing oneselfpositively in every instance is not a compelling need. Accordingly, in Study 4we subdivided subjects according to whether they had previously taught theactivity. This enabledus to test the hypothesis that thepeople who hadtaughtwould more readily describe themselves negatively. Finally,itwas also possibleto examine theeffect of commitment again, defined after themanner of Study1.

METHODSubjectsThirty female and 29 male undergraduates participated as subjects. Seven ofthese were excluded from the final data analysis. They either failed toname aformer teacherorrefused to describe the only teacher they hadencoun tered intheir areaof interest in the requested negative terms.ProcedureWhen subjects arrivedat the designated classroom, ingroups of approximately15, they were met by a male experimenter. Hehanded out a questionnaire onwhich subjects indicateda special activity area. Italso asked subjects howmanydaysago theactivity waslast engaged in whether they ever taughtinthe ir area


20/26


of interest, and how many teachers they had studied under in this area. Whenall the subjects had finished this questionnaire, the experimenter asked them towork on the Teach er Description Fo rm .The Teacher Salience Manipulation. The instructions on the first page ofthis form explained that teachers differ in their effectiveness and likability. Halfof the subjects w ere then asked to think of the person w ho qualified as the bestteacher they had encountered in their activity area. The other half was asked tothink of their worst teacher. Then they began writing an essay on that person,that was supposed to answer questions related to whether the teacher was inter-

esting, enth usiastic, well qualified, and so forth. Subjects w ere told to fill aboutone pa ge . The purpose of asking them to elaborate on the traits of their teacherswas simply that of maximizing the positive (or negative) teacher's salience.Measurement of Self Deprecation. When subjects had finished they wereasked to make a self-descriptive statement characterizing their performance ona testone they did not actually take^measuring competence in their activityarea. The experimenter handed out a ' 'Performance Feedback Sheet for the Basic

Abilities T e s t. Subjects recorded their area of interest, name, sex, age, and ,f ina l ly , they f i l led in the b lanks of the fo l lowing s ta tement :'' (name) performed worse than % ofth undergraduate sample thathad taken this te st . The experimenter explained that he would need thesefabricated test results to create failure expectancies for forthcoming subjects whowould actually be tested in their areas of competence. The experimenter madeit understood that negative statements (i.e., high percentages), would serve theproclaim ed pu rpose the best. T he experimenter accepted responsibility for show-ing the test results to the forthcoming subjects.Manipulation Checks. Prior to debriefing, subjects responded to the follow-ing questionnaire items (nine-point answer scales): H ow secure do you feel inyou r area of interest when you think of the teacher you just des cr ibed ?, and How good is the teacher you described in the teacher description form?

RESULTSEquivalence of GroupsThere were no significant differences between Best-Teacher-Salient and Worst-Teacher-Salient conditions on any ofthe following m easures: recency of engagingin the activity; teaching ex perience; and number of teachers encountered (all f's< 1.0). The proportions of subjects indicating sp o rt s and n on sp or ts didnot differ between conditions: 46% sports in Best-Teacher-Salient cond ition, and45 % sports in Worst-Teacher-Salient condition.


21/26


Effectiveness of Manipulationr-tests were performed on subjects' reported security in their activity areas andon the perceived quality of the described teacher. The teacher subjects wroteabout was rated more positively in the Best-Teacher-Salient condition than inthe Worst-Teacher-Salient condition, t 39) = 17.80, p


22/26

1 1 0 WICKLUND AND GOLLWITZER

100

i 75

I 50S 25

0Best Worstteacter teoctier

(N=20) iU-ZiExperimentafvoriot ion

Hove Have nottought fought(N=I6) (N 251Correloftonolworiotion

FIG . 3 . Impac t of experim ental and correlational variations of comp leteness on self-deprecation(Study 4).

DISCUSSIONThe p resent study is especially poignant in illustrating a sim ple, direct derivationfrom the theory. A self-description of oneself as inept in a particular instanceis rendered possible because of a preexisting, salient symbol of completeness.Peop le wh o have been asked to dwell on a high caliber teacher are able to bringthemselves to forego a further symbol of completeness (positive self-description).This effect, whether referred to as modesty, magnanimity, or simply self-de-precation, is interesting in that the self-description does not directly reflect theunderlying cond ition of the individual. Instead, the preexisting evidence fo rone 'scompetence is inv rs ly correlated with the subsequent self-description. Theeffectiveness of this salience manipulation would suggest that it might generallybe possible to lull someone into a feeling of completeness, or alternatively, toshake the person's security of self-definition, just by making salient select por-tions of the individual's background. This should not come as a surprise, giventhe accumulated evidence on the inordinate impact of one recent or salientselfrelevant event (Duval & Wicklund, 1972; Taylor & Fiske, 1978).The impact of teaching experience is interesting because of the implication


23/26


that the influence variable, which was the dependent measure for Studies 1-3 ,can just as readily be used as a predictor variable. If one has taught, thus hada formal impact on othe rs, it is possible to forego to some extent a positiveselfdescription.

GENERAL DISCUSSIONIn ordertoplace these results into proper perspective , it simp ortanttodistinguishthis framework from other kinds of psychological states that might possiblycharacterize the present paradigms. The following two ideas are discussed herebecause components of them resemble facets of self-completion research. Thesetwo are: (1) the ubiquitous idea of self-presentation; and (2) cognitive dissonancetheory.

Divergent Antecedents of Public Self DescriptionsAs the concept of self-presentation has been treated in social psychology orsociology (Goffman, 1959; Jones & Pittman , in press; Schlenker, 1975), it isgenerally thought that the immediate social milieu creates a pressure on peopleto present themselves in a way that will gain approval, respect, or generosityfrom those present. For instance, Schlenker (1975) dwells on the notion thatpeople try to appear positive, but also consistent in the eyes of others. Moregenerally, Jones and Pittman (in press) delineate numerous forms of self-pres-entation, ranging from projecting an image of integrity, to efforts to gain powerover the other through strategic self-presentation.

The concept of self-symbolizing rings a bell of presen ting to ot he rs . How -ever, a central distinction should be drawn between the present endeavors andwh at is otherwise referred to as self-pre sen tatio n. In the present framew orkself-descriptions are viewed as originating in the dynamics of the individual.The y are said to result from patterns of personal history , or from experimen tallyinduced symbolic incompleteness. The governing variables within the previousself-presentation work have been viewed largely as constraints of the imm ediatesituation. Thus it is predicted that the person will try to look positive in others'eyes, or modest, or consistent, or powerful, depending on the strategy set offby the situation. The crucial distinction is that over and above such socialconstraints, one may characterize the self-symbolizing individual as governedby dynam ic processes that originate prior to the self-presentational conte xt.Indeed, it is even possible to demonstrate that relatively incomplete individualswill present themselves in ways contrary to self-presentational cues, given thatthe cues call for self-descriptions that are not conducive to generating com plete-ness (GoUwitzer & Wicklund, Note 2).


24/26

1 1 2 WICKLUND AND GOLLWITZER

Post Decisional Processes Versus Self SymbolizingCognitive dissonance theory (Festinger, 1957) is, just as the present notion ofself-completion, a concept about compensation or overcoming of obstacles. Inthe case of dissonanc e theory the process focuses strictly on a conflict betweena behavioral direction (recent decision) and a disparate (dissonant) set of cog-nitions. The context is one of having just decided, and the consequences of thedecision should be foreseeable and freely chosen (Brehm Cohen, 1962; W ick-lund Breh m , 1976). In contrast, the self-completion concept talks about con-structing a self-definition via relevant self-symbols; the domain is not limitedto that of a single, post-decisional period (cf. Studies 1 and 2 ). Further, asymbolic lack does not require volition in order to trigger self-symbolizing, asis illustrated in Studies 3 and 4 . The m anipulations of these studies have the no-choice quality that has been shown to suppress dissonance effects (CooperBreh m, 197 1; Lepper, Zanna, Abelson, 1970; U nd er, Cooper, W icklund,1968). Further, Study 4 employs self-descriptions as a dependent variable; thismeasure has never been a part of dissonance-theory derivations.

The central difference, then, is one of scope: Self-completion deals with along-range endeavor, a process that involves the extended accumulation of sym-bols and that gains a motivational momentum when the individual falls short onsymbolic dimensions. Dissonance theory, quite in contrast, focuses on the spe-cific, one-time decision and holds its analysis to the short run.

ConclusionThe present research points to a special conception of the human being. Thestarting point is a com mitment to a self-definition, and once people fall shorton a symbolic dimension they move toward substituting an alternative symbolof completeness for the lacking one. According to this conception the humanis impatient in regard to defining theself unwilling to tolerate insufficiencieson important dimensions of the self-definition. Thus given an acute lack ofsym bolic support, individuals reach out, on dimensions of interpersonal influenceor self-description, oriented toward gaining a broader acknowledgem ent for theirstatus as journalists, guitarists, Spanish speakers, and so forth.

Accompanying this phenomenon is an apparent willingness on the part ofnoninvolved bystanders to assume that such self-symbolizing efforts stem froma basic competence. The observer study (Note1 that we reported in conjunctionwith Study 1 implies that involved self-symbolizers are indeed gaining broade rrecognition by turning their efforts to influencingothers;the g reatertheattemptedinfluence, the more relevant training is imputed to the person by observers.The directions to be taken in new research have to do with questions ofsalience touched upon here in Study 4 and with the broad issue oftherelativeimportance and longevity of various classes of symbols. The present research


25/26


is viewed as a starting point, as illustrating a conception that may be a funda-mental human process.

REFEREN E NOTES1. Hilton, J. L., GoUwitzer, P. M ., & Wicklund, R. A. W ienthe observer goes astray: Inferring

competence from self-symbolizing efforts. Unpublished manuscript. University of Texas, 1981.2 . GoUwitzer, P. M ., & Wicklund, R. A. Symbolic self completion and self presentation. Un-

published manuscript. University of Texas, 1980.

REFEREN ESAdler, D. L ., & K ounin, J. S . Some factors operating at the moment of resumption of interrupted

tasks. The Journal of Psychology, 1939, 7, 255-267.Bandura, A. Principles of behavior modification. New York; Holt, Rinehart, & Winston, 1969.Brehm, J. W., & Cohen, A. R. Explorations in cognitive dissonance. New York: Wiley, 1962.Buss, A. H. The psychology o f aggression. New York: Wiley, 1961.Buss,A. H . Physical aggression in relation to different frustrations.Journal of Abnormal and Social

Psychology, 1963,67, 1-7.Buss, A. H. Instrumentality of aggression, feedback, and frustration as determinants of physicalaggression. Journal of Personality and Social Psychology, 1966,3, 153-162.

Cooley, C. H. Human nature and the social order. New York: Scribner, 1902.Coop er, J., & Bre hm , J. W. Prechoice awareness of relative deprivation s determinant of cognitive

dissonance. Journal of Experimental Social Psychology, 1971,7, 571-581.DoUard, J., Doob, L. W., Miller, N., Mowrer, O. H., & Sears, R. R. Frustration and aggression.

New Haven, Conn.: Yale, 1939.Duval, S., & Wicklund, R. A. A theory of objective self-awareness. New York: Academic Press,

1972.Festinger, L. A theory of cognitive dissonance. Stanford, Calif : Stanford University Press, 1957.Fulbright, J. W. The arrogance of power. New York: Vintage, 1966.Geyer, D. Der russische Imperialismus. Gottingen, Federal Republic of Germany: Vandenhoeck

& Ruprecht, 1977.Goffman, E. The presentation of selfineveryday life. Garden City, N.Y.: Anchor Books, 1959.He nle, M . The influence of valence on substitution. The Journal of Psychology, 19 4 4 ,17, 11-19.Izraeli, D . Se ttlin g-in : An interactionist perspective on the entry of the new mana ger. Pacific

Sociological Review, 1977,20, 135-160.Jones, E. E., & Pittman, T. S. Toward a general theory of strategic self presentation. In J. Suls

(Ed.) , Psychological perspectives on theself Hillsdale, N.J.: Lawrence Erlbaum Associates,in press.Kelley, H. H. Communication in experimentally cheated hierarchies. Human Relations, 1951, 4,39-56 .

Kiesler, C. A. The psychology of comm itment. New York: Academic Press, 1971.Kre garm an, J. J. , & W orchel, P . Arbitrariness of frustration and aggression. Journal of Abnormal

and Social Psychology, 1961,63, 183-187.Leppe r, M . P. , Zanna, M . P. , & Abelson, R. P. Cognitive irreversibility in a dissonance reduction

situation. Journal ^Personality and Social Psychology, 1970, 16, 191198.Linder, D . E. , Cooper, J., & Wicklund, R. A. Pre-exposure persuasion as a result of comm itment

to pre-exposure effort. Journal of Experimental Social Psychology, 1968,4 , 470-482.


26/26


Lissn er, K. Die Entspannu ng von Bedurfnissen durch Ersatzhandlungen . PsychologiseheForschung,1933, 18, 218-250.

Mah ler, W. Ersatzhandlungen verschiedenen Realitatsgrades.Psychologisehe Forschung, 1 9 3 3 ,18 ,2 7 - 8 9 .Mead, G. H. Mind self and society. Chicago: University of Chicago Press, 1934.Mosteller, F. M ., Busch , R. R. Selected quantitative techniques. In G. Lindzey (E d.), Handbook

of social psychology. Vol. 1:Theory andmethod. Reading, Mass.: Addison-Wesley, 1954.Ovsiankina, M. Die Wiederaufnahme unterbrochener Handlungen. Psychologisehe Forschung,

1928, 11, 302-379.Rosen thal, R. Combining results of independent studies.Psycho logical Bulletin, 1 9 7 8 ,85 ,18 5 -193 .Ross, L ., Bierb rauer, G ., Polly, S. Attribution of educational outcom es by professional and

nonprofessional instructors.Journal of Personality ruiSocial Psychology, 1974,29, 609-618 .Sch lenke r, B . R. Self-presentation: Manag ing the impression of consistency when reality interfereswith self-enhancement. Journal of Personality and Social Psychology, 1975,32, 1030-1037.Tay lor, S. E ., Fiske, S. T. Salience, attention, and attribution: Top of th e head phenom ena. In

L.Berkowitzihd. ,Advancesinexperimentalsocialpsychologyiyol. 11). New York: Academ icPress, 1978.

Wicklund, R. A ., Brehm, J. W . Perspectives on cognitive dissonance. Hillsdale, N.J.: LawrenceErlbaum Associates, 1976.

Zeigamik, B. Das Behalten erledigter und unerledigter Handlungen. Psychologisehe Forschung,1927, 9, 1-85.

Zillmann, D. Hostility and aggression. Hillsdale, N.J.: Lawrence Eribaum Associates, 1979.Zillm ann , D ., Can tor, J. R. Effect of timing of information about mitigating circum stances onemotional responses to provocation and retaliatory behavior. Journal of Experimental Social

Psychology, 1976, 12, 38-55 .

symbolic self-completion, attempted influence, and self-deprecation

Documents