Society for the Study of Social Problems A Comment on Disguised Observation in Sociology Author(s): Kai T. Erikson Source: Social Problems, Vol. 14, No. 4 (Spring, 1967), pp. 366-373 Published by: University of California Press on behalf of the Society for the Study of Social Problems Stable URL: http://www.jstor.org/stable/798850 Accessed: 19/08/2010 13:16 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=ucal. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact email@example.com. Society for the Study of Social Problems and University of California Press are collaborating with JSTOR to digitize, preserve and extend access to Social Problems. http://www.jstor.org 366 SOCIAL PROBLEMS cians, etc.; in short, all people as they doing what one can do without prom- carry out jobs for which they are in ising to keep informationconfidential. some sense publicly accountable.One Since publicly accountableindividuals of the functions of our discipline, often recognize the fact of their ac- along with those of political science, countability and the useful purposes history, ecqnomics, journalism, and that might be served by sociologists intellectual pursuits generally, is to studying them, one can often gain a further public accountability a soci- in good deal of cooperationwithout the ety whose complexity makes it easier promise of confidentiality.5 for people to avoid their responsi- We are suggesting that sociologists bilities. in this respect have the same rights We would suggest that, in principle, that journalistshave. Our understand- anyone is publicly accountablefor the ing of the social process may be such actionswhich it is his duty to perform. that we do not use this right in the Most of the time, however, since soci- same way as journalists, because we ologists are not muckrakers, is not it are not interested in momentarysen- sations but in developing an under- necessary or desirable to single out individuals or even clearly identifiable standing of the persisting tendencies small groups. In such situations one of social systems, large or small. may reasonablyuse confidentialityas 5 Clearly the question of methodology is an inducementto cooperation.In other also involved here. It is easier to maintain situations, however, this is clearly the position we have outlined if the meth- ods used are primarily observational (as unwarranted.If one wishes to study was true of Sudnow's work) than it is with the functioning of courts, or of a interviewing methods where the interaction mayor's office, or of General Motors, of the interviewer and respondent may or of unions, it is perhaps better to more easily generate expectations of con- fidentiality which must be dealt with in put up with the difficulties of only some forthright fashion. A COMMENT ON DISGUISED OBSERVATION IN SOCIOLOGY* KAI T. ERIKSON Yale University At the beginning of their excellent The purpose of this paper is to paper on the subject, Howard S. argue that the researchstrategy men- Becker and Blanche Geer define par- tioned in the last few words of that ticipant observation as "that method description represents a significant in which the observer participatesin ethical problem in the field of sociol- the daily life of the people under ogy. In point of sheer volume, of study, either openly in the role of re- course,the problem is relativelysmall, searcheror covertlyin some disguised for disguised participant observation role. . . 1 is probablyone of the rarestresearch * Paper read at the annual meetings of techniques in use among sociologists. the Society for the Study of Social Prob- But in point of general importance, lems, Chicago, 1965. the problem is far more serious- 1 Howard S. Becker and Blanche Geer, partly because the use of disguises "Participant Observation and Interviewing: seems to attract a disproportionate A Comparison," Human Organization, 16 (1957), pp. 28-32. amount of interest both inside and Disguised Observationin Sociology 367 outside the field, and partly because fessional meetings. But a good deal it offers a natural starting point for more is at stake here than the sensi- dealing with other ethical issues in tivities of any particularperson, and the profession. my excuse for dealing with an issue In recentyears,a handful of studies that seems to have so many subjective have been reported in the literature overtones is that the use of disguises based on the work of observerswho in social research affects the profes- deliberatelymisrepresented their iden- sional climate in which all of us work tity in order to enter an otherwise and raisesa numberof methodological inaccessible social situation. Some of questions that should be discussed these studies have alreadyprovoked a more widely. good deal of comment-among them, I am assuming here that "personal for instance, the cases of the anthro- morality" and "professional ethics" pologist who posed as a mentalpatient are not the same thing. Personalmoral- by complaining of symptoms he did ity has something to do with the way not feel,2 the sociologists who joined an individual conducts himself across a gathering of religious mystics by the range of his human contacts; it is professing convictions they did not not local to a particulargroup of per- share,3the Air Force officerwho bor- sons or to a particularset of occupa- rowed a new name, a new birth date, tional interests.Professionalethics, on a new personal history, a new set of the other hand, refer to the way a mannerismsand even a new physical group of associatesdefine their special appearance order to impersonatean in responsibilityto one another and to enlisted man,4and the group of gradu- the rest of the social order in which ate studentswho venturedinto a meet- they work. In this sense, professional ing of Alcoholics Anonymouswearing ethics often deal with issues that are the clothes of men from other social practicalin their applicationand lim, classes than their own and the facial ited in their scope: they are the terms expressionsof men suffering from an of a covenant among people gathered unfortunatedisability.5 together into a given occupational In taking the position that this kind group. For instance,it may or may not of masqueradingis unethical, I am be ethical for an espionage agent or naturally going to say many things a journalist to represent himself as that are only mattersof personalopin- someone he is not in the course of ion; and thus the following remarks gathering information,but it certainly are apt to have a more editorialflavor does not follow that the conduct of a than is usual for papers read at pro- sociologist should be judged in the same terms; for the sociologist has a 2 William C. Caudill et al., "Social Structure and Interaction Processes on a differentrelationshipto the rest of the Psychiatric Ward," American Journal of community,operatesunder a different Orthopsychiatry, 22 (1952), pp. 314-334. warrant,and has a differentset of pro- 3 Leon Festinger, Henry W. Riecken, fessional and scientific interests to and Stanley Schacter, When Prophecy Fails. Minneapolis: University of Minnesota protect. In this sense, the ethics gov- Press, 1956. erning a particular discipline are in 4 Mortimer A. Sullivan, Stuart A. Queen, many ways local to the transactions and Ralph C. Patrick, Jr., "Participant Ob- that discipline has with the larger servation as Employed in the Study of a world. Military Training Program," American So- ciological Review, 23 (1958), pp. 660-667. The argumentto be presentedhere, 5 John F. Lofland and Robert A. Le- then, is that the practice of using jeune, "Initial Interaction of Newcomers in masks in social researchcompromises Alcoholics Anonymous: A Field Experi- ment in Class Symbols and Socialization," both the people who wear them and Social Problems, 8 (1960), pp. 102-111. the people for whom they are worn, 368 SOCIAL PROBLEMS and in doing so, violates the terms of course,is whetherwe have the right of a contract which the sociologist to inflict pain at all when we are aware should be ready to honor in his deal- of these risks and the subjects of the ings with others. There are many re- study are not. The second, however, spects in which this is true, but I will is perhaps more important from the be dealing here in particularwith the narrowpoint of view of the profession relationship between the sociologist itself: so long as we suspect that a and a) the subjects of his research, method we use has at least some po- b) the colleagues with whom he tential for harming others, we are in works, c) the students he agrees to the extremely awkward position of teach, and d) the data he takes as his having to weigh the scientific and so- subject matter. cial benefits of that procedureagainst The first of these points has to do its possible cost in human discomfort, with the responsibilitiesa sociologist and this is a difficult business under should accept towardother institutions the best of circumstances. we hap- If and other people in the social order. pen to harm people who have agreed It may seem a little cranky to insist to act as subjects,we can at least argue that disguised observation constitutes that they knew something of the risks an ugly invasion of privacyand is, on involved and were willing to con- that ground alone, objectionable.But tribute to that vague program called it is a matter of cold calculation to the "advance knowledge".But when of point out that this particularresearch we do so with people who have ex- strategycan injure people in ways we pressed no readinessto participatein can neither anticipate in advance nor our researches (indeed, people who compensate for afterward. For one would presumably have refused if thing, the sheer act of entering a hu- asked directly), we are in very much man transaction the basis of delib- on the same ethicalposition as a physician erate fraud may be painful to the who carries out medical experiments people who are thereby misled; and on human subjects without their con- even if that were not the case, there sent. The only conceivable argument are countlessways in which a stranger in favor of such experimentationis who pretends to be something else that the knowledge derived from it is can disturbothers by failing to under- worth the discomfort it may cause. stand the conditions of intimacy that And the difficultieshere are that we prevail in the group he has tried to do not know how to measurethe value invade. Nor does it mattervery much of the work we do or the methodswe how sympathetic the observer is to- employ in this way, and, moreover, ward the persons whose lives he is that we might be doing an extraordi- studying: the fact of the matter is nary disserviceto the idea of detached that he does not know which of his scholarship if we tried. Sociologists actions are apt to hurt other people, cannot protect their freedom of in- and it is highly presumptuousof him quiry if they owe the rest of the com- to act as if he does-particularly when, munity (not to mention themselves) as is ordinarilythe case, he has elected an accounting for the distress they to wear a disguise exactly becausehe may have inadvertently imposed on is entering a social sphere so far from people who have not volunteered to his own experience. take that risk. So the sheer act of wearing dis- The second problem with disguised guises in someone else's world may observationto be consideredhere has cause discomfort, no matter what we to do with the sociologist's responsi- later write in our reports;and this pos- bilities to his colleagues. It probably sibility raises two questions. The first, goes without saying that researchof Disguised Observationin Sociology 369 this sort is liable to damagethe repu- afford to jeopardize, and we have tation of sociology in the larger society every right to be afraid that such and close off promising areas of re- people may dose their doors to socio- search for future investigators. This logical researchif they learn to become is true in the limited sense that a too suspicious of our methods and particular agency-say, for example, intentions. Alcoholics Anonymous-may decide The third objection to be raised that its integrity and perhaps even its here, if only as a note in passing, con- effectiveness was violated by the ap- cerns the responsibilities the profes- pearanceof sociologists pretending to sion should accept toward its students. be someone else and deny access to The division of labor in contemporary other students who propose to use an sociology is such that a considerable altogether different approach.And it proportionof the data we use in our is also true in the wider sense that any work is gathered by graduatestudents researchtactic which attractsunfavor- or other apprentices,and this propor- able notice may help diminish the tion is even higher for researchpro- general climate of trust toward soci- cedures that require the amount of ology in the community as a whole. energy and time necessaryfor partici- So long as this remainsa serious pos- pant observation. Of the dozen or sibility, the practice of disguised ob- more observerswho took part in the servation becomes a problem for studies I have cited, for example, all everyone in the profession; and to but one was a graduatestudent. Now this extent, it is wholly within the a number of sociologists who have bounds of professional etiquette for one sociologist to challenge the work engaged in disguised observationhave of another on this score. reported that it is apt to pose serious moral problems and a good deal of This objection has been raised sev- personal discomfort, and I think one eral times before, and the answermost might well argue that this is a heavy often given to it is that the people burdento place on any person who is, who are studied in this fashion-alco- by our own explicit standards,not yet holics or spiritualists or mental pa- ready for professional life. I am not tients, for example-are not likely to suggesting here that students are too read what we say about them anyway. immature to make a seasoned choice Now this argumenthas the advantage in the matter. I am suggesting that of being correct a good deal of the they should not be asked to make what time, but this fact does not prevent it one defenderof the method has called from being altogether irrelevant. To "real and excruciating moral deci- begin with, the experienceof the past sions" while they are still students few years should surelyhave informed and presumably protected from the us that the press is more than ready Xarious dilemmas and contentions to translate our technical reports into which occupy us in meetings like this news copy, and this means that we -particularly since they are so likely can no longer provideshelter for other to be academically,economically,and people behind the walls of our own even psychologically dependent upon anonymity.But even if that were not those elders who ask them to choose.6 the case, it is a little absurdfor us to The fourth objection I would like claim that we derive some measureof to raise here about the use of under- protectionfrom the narrownessof our audience when we devote so much 6 To keep the record straight, I might time trying to broaden it. The fact is add that I first becameinterestedin these that we are increasinglyreaching au- matterswhen I was a graduatestudentand diences whose confidence we cannot applied for one of the observerposts men- tioned here. 370 SOCIAL PROBLEMS cover observationis probablythe most scious mind of the participant.It may important-and yet the most remote be possible for someone to mimic the from what is usually meant by the conventional gestures of fear, but it term "ethics."It seems to me that any is impossible for him to reproduce attempt to use masqueradesin social the small postural and chemical researchbetrays an extraordinary dis- changes which go with it. It may be respect for the complexitiesof human possible for a middle-cass speaker to interaction, and for this reason can imitate the broader accents of lower- only lead to bad science. Perhaps the class speech, but his vocal equipment most important responsibilityof any is simply not conditioned to do so sociologist is to appreciatehow little without arousing at least a subliminal he really knows abouthis intricateand suspicion. It may be possible for a elusive subjectmatter.We have at best trained person to rearrangethe slant a poor understandingof the human of his body and re-set his facial mus- mind, of the communicationsignals cles to approximate the bearing of that link one mind to another, or the someone else, but his performance social structures that emerge from will never be anything more than a those linkages-and it is the most ar- rough imposture.Now we know that for rant kind of over-simplification us these various physiological, linguistic, to think that we can assess the effect and kinetic cues play an important part which a clever costumeor a few stud- in the context of human interaction, ied gestureshave on the social setting. but we have no idea how to simulate The pose might "work" in the sense them-and what is probablymore to that the observeris admitted into the the point, we never will. For one situation; but once this passage has thing, we cannot expect to learn in a been accomplished, how is he to judge matterof hours what others have been his own influence on the lives of the practicing throughout a lifetime. For people he is studying? This is a seri- another, to imitate always means to ous problem in every departmentof parody, to caricature, to exaggerate science, of course, and a good deal of certain details of behavior at the ex- time has been devoted to its solution. pense of others, and to that extent any But the only way to cope with the person who selects a disguise will nat- problem in even a preliminaryway is urally emphasize those details which to have as clear a picture as possible he assumesare most importantto the of the social properties that the ob- he In character is portraying. doing so, server is introducing into the situa- of course,he is really only portraying tion, and this is altogether impossible a piece of himself. It is interestingto if we ourselves are not sure who he speculate, for example, why the Air is. We can impersonateother modes Force lieutenant mentioned earlier of behavior with varying degrees of thought he needed to present himself insight and skill, but we cannot repro- as a near-delinquent youth with a visi- duce them; and since this is the case, ble layer of personal problems in or- it seems a little irresponsible for a der to pose as an enlisted man. sociologistto assumethat he can enter Whatever the reasoning behind this social life in any masquerade that particularcharade, it would certainly suits his purpose without seriously be reasonablefor someone to suspect disruptingthe scene he hopes to study. that it tells us more about the investi- When people interact, they relate gators' impression of enlisted men to one anotherat many differentlevels than it does about the men themselves at once, and only a fraction of the -and sincewe have no way of learning messages communicated during that whether this is true or not, we have interchangeare registeredin the con- lost ratherthan gained an edge of con- Disguised Observationin Sociology 371 trol over the situation we are hoping psychologist, the observer was recom- to understand.What the investigators mended for re-classificationon the had introducedinto the situation was groundsthat he appearedquite anxious a creatureof their own invention, and over the death of his father. Now it would be hardly surprising if the these events may indeed suggest that results of their inquiry corresponded the pose was successful, for the ob- to some image they had in advanceof server was trying to look somewhat the enlisted man's condition. (It is delinquentand did have a story mem- perhapsworth noting here that imper- orized about the death of his father sonation always seems easier for peo- in an auto accident. But who would ple looking down rather than up the care to argue that the diagnosis of the status ladder. We find it reasonableto sergeant and the psychologist were assume that officers "know how" to inaccurate? Surely something was portray enlisted men or that sociolo- wrong, and if they perceived an edge gists have the technical capacity to of uneasinesswhich remindedthem of pose as drunks or religious mystics, anxiety or detected a note of furtive- but it is not at all clear that the re- ness which looked to them like delin- verse would be equally true.) quency, they may only have been This, then, is the problem. If we responding to the presence of a real provide observers with special masks conflict between the observer and his and coach them in the "ways" of the mask. We may leave it to the psycho- private world they are hoping to en- analysts to ask whether vague anxie- ter, how can we learn what is happen- ties about "killing" one's father are ing to the people who meet them in an unlikely impressionfor someone to this disguise? What information is leave behind when he is parading registered in the unconscious minds aroundwith a new name, a new back- of the other people who live in that ground, a new history, and, of course, world? How does the social structure a new set of parents. The authors of accommodate this peculiarinvasion? to the article tell us that the observer It is clear, I think, that something "did have something of a problem to happens-something over which we transformhimself from a 27-year-old, have no control. Let me relate two college trained, commissioned officer incidentsdrawn from the studies men- into a 19-year-old, near-delinquent tioned earlier.The first has to do with high school graduate,"and this is cer- the Air Force officerwho posed as an tainly easy to believe.8 What is more enlisted man. In their report of the difficultto believe is that such a trans- study, the investigators used several formationis possible at all-and if it is pages of a short paper to describethe not, we can have very little confidence elaborate masqueradethey had fash- in the information gathered by the ioned for the observerand the coach- observer. Since we do not know to ing he had received in the ways of what kind of creaturethe enlisted men the adolescent sub-culture. "So suc- were responding, we do not know cessful was the tutoring," reads the what sense to make of what they said brief report, "that when the time for and did. 'enlistment' arrived, the recruiting The second example comes from sergeant . . . suggested that the ob- the study of the apocalypticreligious servernot be acceptedby the Air Force group. At one point in the study, two because by all appearanceshe was a observersarrivedat one of the group's juvenile delinquent."7And later, dur- meeting places under instructions to ing an interview with a service 8 StuartA. Queen, "Comment,"Ameri- 7 Sullivan, Queen, and Patrick, op. cit., can Sociological Review, 24 (1959), pp. p. 663. 399-400. 372 SOCIAL PROBLEMS tell quite ordinary stories about their trol over the effects introducedby the experience in spiritualismin order to observer.A companyof recruitswith create as little commotion as possible. a disguised officerin its midst is sim- A few days afterwards,however, the ply a different kind of organization leader of the group was overheard than one without the same ingredient; explaining that the two observershad a group of spiritualists which num- appearedupset, excited, confused, and bers as many as eight observersamong unsure of their errand at the time of its twenty or so membershas a wholly their original visit, all of which helped different character than one which confirm her suspicion that they had does not-and so long as we remain somehow been "sent" from another unable to accountfor such differences, planet. In one sense, of course, this we cannot know the meaning of the incident offered the observers an in- information we collect. triguing view of the belief structure of the cult, but in another sense, the In one of the most sensible pieces leader's assessment of the situation written on the subject, Julius Roth was very shrewd: after all, the ob- has remindedus that all social research servers had been sent from another is disguised in one respect or another world, if not anotherplanet, and she and that the range of ethical questions may have been quite right to sense which bear on the issue must be vi- that they were a bit confused and un- sualized as falling on a continuum.10 sure of their errandduring their early Thus, it is all very well for someone moments in the new job. "In both to argue that deliberate disguises are cases," the report informs us, the vis- improper for sociologists, but it is its of the observers "were given as quite anothermatterfor him to specify illustrations that 'strange things are what varieties of researchactivity fall happening.'"9 Indeed, strange things within the range of that principle. were happening; yet we have no idea Every ethical statement seems to lose how strange they really were. It is al- its crisp authority the moment it is most impossible to evaluate the reac- carried over into marginal situations tion of the group to the appearance where the conditions governing re- of the pair of observers because we search are not so cearly stipulated. do not know whether they were seen For instance, some of the richest ma- as ordinary converts or as extraordi- terial in the social sciences has been narybeings. And it makesa difference, gatheredby sociologistswho were true for in the first instance the investi- participantsin the group under study gators would be observing a response but who did not announce to other which fell within the normal range members that they were employing of the group's experience, while in this opportunity to collect research the second instancethey would be ob- data. Sociologistslive careersin which serving a responsewhich would never they occasionallybecome patients, oc- have taken place had the life of the casionally take jobs as steel workers group been allowed to run its own or taxi drivers, and frequently find course. themselves in social settings where My point in raisingthese two exam- their trained eye begins to look for ples, it should be clear, is not to insist data even though their presencein the on the accuracyof these or any other situation was not engineered for that interpretations,but to point out that purpose. It would be absurd,then, to a wide variety of such interpretations insist as a point of ethics that sociolo- is possible so long as one has no con- 10 Julius A. Roth, "Comments on 'Se- 9 Festinger, Riecken, and Schacter, op. cert Observation,'" Social Problems, 9 cit., p. 241-242. (1962), pp. 283-284. The Observer,the Experimenterand the Group 373 gists should always introduce them- importantquestions about invasion of selves as investigatorseverywherethey privacy as an ethical issue. go and should inform every person In the meantime,the time has prob- who figures in their thinking exactly ably come for us to assume a general what their researchis all about. posture on the question of disguised But I do think we can find a place to begin. If disguised observationsits participantobservationeven if we are not yet ready to state a specific ethic, somewhereon a continuumand is not and a logical first step in this direction easily defined, this only suggests that would be to assess how most members we will have to seek further for a of the profession feel about the mat- relevant ethic and recognize that any ter. I am not suggesting that we poll line we draw on that continuumwill one anotheron the merits of adopting be a little artificial. What I propose, a formal code, but that we take some then, at least as a beginning, is the kind of unofficial reading to learn following: first, that it is unethicalfor what we can about the prevailing cli- a sociologist to deliberatelymisrepre- mate of opinion in the field. If we dis- sent his identity for the purpose of cover that a substantial number of entering a private domain to which sociologists are uncomfortable about he is not otherwise eligible; and sec- the practice, then those who continue ond, that it is unethical for a sociolo- to employ it will at least know where gist to deliberately misrepresentthe characterof the researchin which he they stand in respectto the "collective is engaged. Now these negative sanc- conscience"of their discipline. And if tions leave us a good deal of leeway we discover that only a scatteringof -more, perhaps, than we will even- sociologists are concerned about the tually want. But they have the effect matter,we will at least have the satis- of establishing a stable point of ref- faction of knowing that the profession erence in an otherwise hazy territory, -as a profession-has accepted the and from such an anchored position responsibility of knowing its own as this we can move out into more mind. THE OBSERVER, THE EXPERIMENTER AND THE GROUP* THEODORE M. MILLS Yale University Imagine,if you will, the plight of the move next? We wonder if he will ever centipede who suddenly stops to ask: move again. He suffersfrom the centi- How do I do this? What system do I pede complex.l employ?How am I relatedto my legs? Stephen Potter applies the concept, Which one, or is it several? should I you will recall, in his advice to the * Written for the symposiumon Ethical golfing gamesman: when your oppo- and Methodological Problemsin Social Psy- nent is on the green and aboutto putt, chological Experimentsunder the auspices ask him what muscles he brings into of the Society for the PsychologicalStudy play, and from what part of the body of Social Issues at the AmericanPsycholog- ical Association convention, Chicago, Illi- 1 Brought to my attentionby Robert F. nois. September1965. Bales.
Pages to are hidden for
"Disguised Observation in Sociology by Erikson"Please download to view full document