Whose Side Are We On? Author(s): Howard S. Becker Source: Social Problems, Vol. 14, No. 3 (Winter, 1967), pp. 239-247 Published by: University of California Press on behalf of the Society for the Study of Social Problems Stable URL: http://www.jstor.org/stable/799147 . Accessed: 22/06/2014 17:18 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . 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 support@jstor.org. . University of California Press and Society for the Study of Social Problems are collaborating with JSTOR to digitize, preserve and extend access to Social Problems. http://www.jstor.org This content downloaded from 188.72.96.102 on Sun, 22 Jun 2014 17:18:10 PM All use subject to JSTOR Terms and Conditions WHOSE SIDE ARE WE ON?* HOWARD S. BECKER Northwestern University To have values or not to have onlyreadtheresults. Will theresearch, values:thequestionis alwayswithus. we wonder,be distorted by thatsymWhen sociologists to study pathy?Will it be of use in the conundertake of scientific theoryor in the problemsthat have relevanceto the struction worldwe live in, theyfindthemselves applicationof scientific knowledgeto Someurgethem the practicalproblemsof society?Or caughtin a crossfire. notto takesides,to be neutraland do willthebias introduced bytakingsides researchthatis technically correct and spoil it forthoseuses? valuefree.Otherstellthemtheirwork We seldom make the feelingexis shallowand uselessif it does not plicit.Instead,it appearsas a lingering to a value worryfor sociologicalreaders,who expressa deep commitment position. would like to be sure theycan trust area This dilemma, whichseemsso pain- whattheyread,and a troublesome fulto so many,actuallydoesnotexist, of self-doubtfor those who do the forone of its hornsis imaginary. For research,who would like to be sure it to exist,one wouldhave to assume, thatwhatever theyfeelare sympathies as someapparently unseemlyand will do,thatit is indeed not professionally possibleto do researchthatis uncon- not, in any case, seriouslyflawtheir taminatedby personal and political work. That the worryaffectsboth indicatesthat I proposeto arguethatit readersand researchers sympathies. differis notpossibleand,therefore, thatthe it lies deeperthanthesuperficial we shouldtake ences that divide sociologicalschools questionis notwhether and thatits rootsmustbe sides, since we inevitablywill, but of thought, of societythat ratherwhosesidewe are on. soughtin characteristics affect our methodus whatever all, I willbeginbyconsidering theprobpersuasion. lem of takingsidesas it arisesin the ologicalor theoretical If thefeelingweremadeexplicit,it studyof deviance.An inspectionof thiscasewill soonrevealto us features would takethe formof an accusation of the researcher thatappearin sociologicalresearchof thatthe sympathies his have his work and distorted biased all kinds.In thegreatest of subvariety Beforeexploring itsstructural jectmatterareasand in workdone by findings. all the different methodsat our dis- roots,letus considerwhatthemanifest posal, we cannotavoid takingsides, meaningof thechargemightbe. forreasonsfirmly basedin socialstruc- It mightmeanthatwe haveacquired ture. some sympathy with the group we to deterus frompubWe maysometimes feelthatstudies studysufficient of devianceexhibittoo greata sym- lishing those of our resultswhich pathywiththepeople studied,a sym- mightprovedamagingto them.One in the researchcarried can imaginea liberalsociologistwho pathyreflected out. This feeling,I suspect,is enter- set out to disprovesome of the comheld abouta minority tainedoffand on bothby thoseof us monstereotypes hisinvestigation who do suchresearch and by thoseof group.To his dismay, reveals that some the stereotypes of in our work other us who, areas, lying are unfortunately true.In theinterests of justiceand liberalism, he mightwell *Presidentialaddress, delivered at the an- be and mighteven succumb tempted, nual meetingof the Societyfor the Study to suppressthose of Social Problems,Miami Beach, August, to the temptation, 1966. findings,publishingwith scientific This content downloaded from 188.72.96.102 on Sun, 22 Jun 2014 17:18:10 PM All use subject to JSTOR Terms and Conditions 240 SOCIAL PROBLEMS candor the otherresultswhich con- have stillnot provedit false.Recoghis beliefs. firmed nizing the point and promisingto But thisseemsnot reallyto be the addressit eventually, I shall turnto heartof thecharge,becausesociologists the typicalsituationsin which the of bias arises. who studydeviancedo not typically accusation hide things about the people they When do we accuseourselvesand of bias? I think study.Theyaremostly willingto grant ourfellowsociologists thatthereis something ingoingon that an inspectionof representative put the deviantsin the positionthey stanceswould show that the accusaare in, evenif theyare notwillingto tion arises,in one important class of grantthatit is whatthe people they cases,whentheresearch givescredence, studiedwereoriginally accusedof. in anyseriousway,to theperspective A more likely meaning of the of thesubordinate groupin somehierIn the case of charge,I think,is this.In the course archicalrelationship. of ourworkand forwho knowswhat deviance,the hierarchical relationship privatereasons,we fallintodeepsym- is a moral one. The superordinate are those pathywiththepeoplewe arestudying, partiesin the relationship so thatwhile the restof the society who represent the forcesof approved viewsthemas unfitin one or another and official morality;the subordinate respectfor the deferenceordinarily partiesare thosewho, it is alleged, accordeda fellowcitizen,we believe have violatedthatmorality. thattheyareat leastas good as anyone Thoughdevianceis a typicalcase, else,moresinnedagainstthansinning. it is byno meanstheonlyone. Similar Becauseof this,we do notgive a bal- situations, and similarfeelingsthatour ancedpicture.We focustoo muchon workis biased,occurin the studyof and prisons, asylums questionswhoseanswersshowthatthe schools,hospitals, in theright in the studyof physicalas well as supposeddeviantis morally citizenmorally in the mentalillness,in the studyof both and theordinary wrong.We neglectto ask thoseques- "normal" and delinquentyouth.In tionswhoseanswerswouldshowthat thesesituations, thesuperordinate parthe deviant,afterall, has done some- tiesare usuallytheofficial and profesin chargeof some rottenand,indeed,pretty sional authorities thingpretty whilethesubormuchdeserveswhathe gets.In conse- important institution, of the dinatesare thosewhomakeuse of the quence,our overallassessment Thus, the problembeing studiedis one-sided. servicesof thatinstitution. What we produceis a whitewashof policeare thesuperordinates, drugadthe deviantand a condemnation, if dictsare the subordinates; professors of thoserespecta- and administrators, onlyby implication, principals and are the superordinates, ble citizenswho,we think,havemade teachers, while thedeviantwhathe is. studentsand pupils are the subordiIt is to this versionthatI devote nates; physiciansare the superordithe rest of my remarks.I will look nates,theirpatientsthe subordinates. All of thesecases represent one of first,however,not at the truthor of thecharge,butratherat the the typicalsituationsin which refalsity in whichit is typically searchersaccuse themselvesand are circumstances of knowl- accusedof bias. It is a situationin madeand felt.The sociology and tensionexist us to distinguish between which,whileconflict edgecautions and an assess- in the hierarchy, thetruthof a statement the conflict has not underwhich becomeopenlypolitical.The conflictmentof thecircumstances is made; thoughwe ing segmentsor ranksare not orgathat statement no one attempts to its sourcein the nized forconflict; tracean argument to of thepersonwhomadeit,we altertheshapeof thehierarchy. interests While This content downloaded from 188.72.96.102 on Sun, 22 Jun 2014 17:18:10 PM All use subject to JSTOR Terms and Conditions WhoseSide Are WleOn? 241 thatsubordinates subordinates may complainabout the posesof ourresearch, treatment theyreceivefromthoseabove have as much rightto be heard as thattheyare as likely them,theydo notproposeto moveto superordinates, a positionof equalitywiththem,or to to be tellingthetruthas theysee it as that what they say reverse positions in the hierarchy.superordinates, has a rightto be Thus, no one proposesthat addicts aboutthe institution should make and enforcelaws for investigated and have its truthor faleven thoughresponpolicemen,that patientsshould pre- sityestablished, assureus thatit is unscribefor doctors,or thatadolescents sible officials shouldgive ordersto adults.We can necessary becausethechargesare false. call thistheapoliticalcase. We can use the notionof a hierIn thesecondcase,theaccusation of archyof credibility to understand this bias is made in a situationthat is phenomenon. In anysystem of ranked franklypolitical.The partiesto the groups,participants take it as given hierarchicalrelationshipengage in that membersof the highestgroup eitherto havetherightto definethewaythings organizedconflict, attempting maintainor changeexistingrelations really are. In any organization,no of power and authority. Whereasin matterwhatthe restof the organizathefirstcase subordinates are typically tionchartshows,thearrowsindicating and thushave,as we shall theflowof information unorganized pointup, thus see, littleto fear froma researcher,demonstrating (at leastformally)that subordinate partiesin a politicalsitua- thoseat thetop have accessto a more tion may have much to lose. When completepictureof whatis goingon thesituation is political,theresearcherthan anyoneelse. Membersof lower mayaccusehimselfor be accusedof groupswill have incomplete informabias by someoneelse when he gives tion,and theirview of realitywill be of either partialand distortedin consequence. credenceto the perspective I leave Therefore, partyto the politicalconflict. fromthepointof viewof a thepoliticalforlaterand turnnow to well socializedparticipant in the systheproblemof bias in apoliticalsitua- tem,anytale told by thoseat the top tions.1 deservesto be regarded intrinsically We provokethe suspicionthatwe as themostcredibleaccountobtainable are biasedin favorof thesubordinateof the organizations' workings.And partiesin an apoliticalarrangementsince,as Sumnerpointedout,matters when we tell the storyfrom their of rankand statusarecontained in the pointof view.We may,forinstance, mores,2 thisbeliefhas a moralquality. investigatetheir complaints,even We are,if we are propermembers of about thegroup,morally thoughtheyare subordinates, boundto acceptthe the way thingsare runjust as though definitionimposed on realityby a one oughtto give theircomplaints as superordinate to group in preference much credenceas the statements of the definitions espousedby subordinWe provokethe ates. (By analogy,the same argument responsibleofficials. chargewhenwe assume,for the pur- holds forthe social classesof a community.) Thus, credibilityand the 1 No situationis necessarilypolitical or disapolitical. An apolitical situation can be rightto be heardare differentialIy transformedinto a political one by the tributedthroughthe ranks of the open rebellion of subordinateranks,and a system. political situation can subside into one in As sociologists,we provoke the which an accommodationhas been reached and a new hierarchybeen accepted by the 2 William Graham Sumner, "Status in participants.The categories,while analytically useful, do not representa fixed divi- the Folkways,"Folkways,New York: New sion existingin real life. AmericanLibrary,1960, pp. 72-73. This content downloaded from 188.72.96.102 on Sun, 22 Jun 2014 17:18:10 PM All use subject to JSTOR Terms and Conditions 242 SOCIAL PROBLEMS and others, to their accountof how the adult chargeof bias,in ourselves to credence and defer- worldtreatsthem.But whydo we not byrefusing give ence to an established statusorder,in accuse other sociologistswho study which knowledgeof truthand the youthof being biased in favor of rightto be heardare not equallydis- adults?Most researchon youth,after tributed."Everyoneknows" that re- all, is clearlydesignedto findoutwhy for adults, sponsible professionalsknow more youthare so troublesome aboutthingsthanlaymen,thatpolice ratherthan askingthe equallyinterand theirwords estingsociologicalquestion:"Whydo are morerespectable than adults make so much trouble for oughtto be takenmoreseriously those of the deviantsand criminals youth ?" Similarly, we accusethosewho withwhomtheydeal. By refusingto takethecomplaints of mentalpatients of credibility, we seriouslyof bias; what about those acceptthehierarchy forthe entireestab- sociologistswho only take seriously expressdisrespect lishedorder. the complaints of physicians, families ? We compoundour sin and further and othersaboutmentalpatients in thedirecWhythisdisproportion provokechargesof bias by not giving of bias? Why do attention and "equal time" tion of accusations immediate of we moreoftenaccusethosewho are to the apologiesand explanations thanthose officialauthority. If, for instance,we on theside of subordinates ? the way whoareon thesideof superordinates are concernedwith studying of life inmatesin a mentalhospital Because,when we make the former we will natu- accusation,we have, like the well buildup forthemselves, withtheconstraintssocialized membersof our society rallybe concerned and conditionscreatedby the actions mostof us are,acceptedthehierarchy of the administrators and physicians of credibilityand taken over the madebyresponsible officials. who run the hospital.But, unlesswe accusation and also make the administrators The reasonresponsible officials make so frequently the objectof our study(a theaccusation is precisely physicians I will considerlater), we becausetheyareresponsible. Theyhave possibility withthecareand operawill not inquireinto whythosecon- beenentrusted of ourimportant are present. tionof oneor another ditions and constraints officials institutions: law enWe will not give responsible schools,hospitals, or whatever. a chanceto explain themselvesand forcement, They are the give theirreasonsfor actingas they ones who, by virtueof theirofficial that goes do, a chanceto show whythe com- positionand the authority withit,are in a positionto "do somearenotjustified. plaintsof inmates It is odd that,when we perceive thing"whenthingsare notwhatthey are the ones bias,we usuallysee it in thesecircum- shouldbe and, similarly, stances.It is odd becauseit is easily who will be held to accountif they ascertainedthat a great manymore failto "do something" or if whatthey studiesare biased in the directionof do is, forwhatever reason,inadequate. in this the interestsof responsibleofficials Becausetheyare responsible thanthe otherway around.We may way,officials usuallyhaveto lie. That accusean occasionalstudent of medical is a grossway of puttingit, but not sociologyof havinggiven too much inaccurate.Officialsmustlie because of patients. thingsare seldomas theyoughtto be. emphasisto thecomplaints Butit is notobviousthatmostmedical For a greatvarietyof reasons,wellinstitutions are sociologistslook at thingsfromthe knownto sociologists, They do not performas pointof viewof the doctors?A few refractory. biased societywould like themto. Hospitals maybe sufficiently sociologists in favorof youthto grantcredibilitydo notcurepeople; prisonsdo notre- This content downloaded from 188.72.96.102 on Sun, 22 Jun 2014 17:18:10 PM All use subject to JSTOR Terms and Conditions WhoseSide Are We On? 243 habilitateprisoners;schools do not someonemalignsit. The sociologist will be spared educatestudents.Since theyare sup- who favorsofficialdom of bias. posed to, officials developwaysboth the accusation of denyingthe failureof the institu- And thus we see why we accuse tion to performas it should and ex- ourselvesof bias only whenwe take It is bewhichcannotbe the side of the subordinate. plainingthosefailures hidden.An accountof an institution'scause,in a situation thatis notopenly operationfromthe pointof view of political,withthemajorissuesdefined therefore castsdoubton as arguable,we join responsible subordinates offithe official line and maypossiblyex- cials and the man in the streetin an of thehierarchy unthinking acceptance pose it as a lie.3 We assumewith them For reasonsthatare a mirror image of credibility. in an thatthe man at the top knowsbest. of thoseof officials, subordinates have We do not realizethatthereare sides hierarchical relationship apolitical no reasonto complainof the bias of to be takenand thatwe are taking sociologicalresearchorientedtoward one of them. Subordi- The same reasoningallows us to theinterests ofsuperordinates. has the whytheresearcher nates typicallyare not organizedin understand for same worryabout the effectof his sucha fashionas to be responsible on his work as his unintheoveralloperationof an institution.sympathies What happensin a schoolis credited volved colleague. The hierarchyof whose of society is a feature and adminis- credibility or debitedto the faculty we cannotdeny,even if we and held existence trators; theycan be identified to believe to account.Even thoughthefailureof disagreewithits injunction a schoolmaybe thefaultof thepupils, themanat thetop. When we acquire to withsubordinates sympathy theyare notso organizedthatanyone sufficient we foranyfailure see thingsfromtheirperspective, of themis responsible in thefaceof but his own. If he does well, while knowthatwe are flying cheat what "everyoneknows."The knowlothersall aroundhim flounder, and steal,thatis none of his affair, edge givesus pause and causesus to the doubt of of honorcodes to share,howeverbriefly, despitethe attempt makeit so. As long as thesociological our colleagues. When a situationhas been defined reporton his school says that every the secondtypeof case I studenttherebut one is a liar and a politically, are quitedifwillfeelcompla- wantto discuss,matters cheat,all thestudents somedegree have ferent. Subordinates the one are cent,knowingthey excepand,withthat,spokestion.More likely,theywill neverhear of organization of thereportat all or,if theydo, will men, theirequivalentof responsible whiletheycannot Spokesmen, reasonthattheywill be gone before officials. forwhat be held responsible actually does it make? long,so whatdifference of theirgroupdo,makeasserThe lackoforganization amongsubor-- members of an institutionalizedtionson theirbehalfand are held redinatemembers meansthat,havingno re- sponsibleforthe truthof thoseasserrelationship for the group's welfare, tions.The groupengagesin political sponsibility theylikewisehave no complaintsif activitydesignedto change existing andthecredihierarchical relationships affects directly 3 I have stated a portion of this argu- bilityof its spokesmen is not Credibility mentmore brieflyin "Problemsof Publica- itspoliticalfortunes. tion of Field Studies," in ArthurVidich, the onlyinfluence, but the groupcan JosephBensman,and Maurice Stein (Eds.), of realill-afford havingthedefinition Reflectionson CommunityStudies, New its spokesmendisYork: JohnWiley and Sons, 1964, pp. 267- ity proposedby credited,for the immediateconse284. This content downloaded from 188.72.96.102 on Sun, 22 Jun 2014 17:18:10 PM All use subject to JSTOR Terms and Conditions 244 SOCIAL PROBLEMS thatwe mustgrasp quence will be some loss of political know,forinstance, the perspectives of both the resident power. Superordinategroups have their of Watts and of the Los Angeles too,andtheyareconfrontedpolicemanif we are to understand spokesmen withthesameproblem:to makestate- whatwenton in thatoutbreak. mentsaboutrealitythatare politically Second,it is no secretthat most effectivewithout being easily dis- sociologistsare politicallyliberal to of the one degreeor another.Our political credited.The politicalfortunes dictatetheside we will be superordinategroup-its ability to preferences hold the statuschangesdemandedby on and, since those preferences are lowergroupsto a minimum-donot sharedbymostof our colleagues,few forthe are readyto throwthe firststoneor dependas muchon credibility, is grouphas otherkindsof poweravail- are even awarethatstone-throwing able as well. a possibility. We usuallytakethe side When we do researchin a political of theunderdog;we are forNegroes situationwe are in double jeopardy, and againstFascists.We do not think for the spokesmenof both involved anyonebiasedwho does researchdeto theimplica- signedto prove that the formerare groupswill be sensitive tionsof our work.Sincetheypropose not as bad as peoplethinkor thatthe of reality, latterare worse.In fact,in thesecirdefinitions openlyconflicting our statement of our problemis in it- cumstanceswe are quite willing to self likelyto call into questionand regardthequestionof bias as a matter makeproblematic, at leastforthepur- to be dealtwithbytheuse of technical one or theother safeguards. posesof ourresearch, definition. And our resultswill do the We are thusapt to takesides with same. equal innocenceand lack of thought, The hierarchy of credibility reasons,in both operates thoughfor different in a different way in the political apoliticaland politicalsituations.In situationthanit does in the apolitical the first,we adopt the commonsense one. In the politicalsituation,it is view which awards unquestioned to the responsibleofficial. preciselyone of the thingsat issue. credibility Since the politicalstrugglecalls into (This is notto denythata fewof us, of theexisting becausesomethingin our experience questionthelegitimacy rank system,it necessarilycalls into has alertedthemto thepossibility, may of hierarchy questionat the same timethe legiti- questiontheconventional in the special area of our macyof the associatedjudgmentsof credibility of who has a expertise.)In thesecondcase,we take Judgments credibility. rightto definethe natureof reality ourpoliticsso forgrantedthatit supin dictatingwhose thataretakenforgrantedin an apoli- plantsconvention tical situation become mattersof side we will be on. (I do not deny, either,thatsomefew sociologists may argument. Oddlyenough,we are,I think,less deviatepoliticallyfromtheirliberal likely to accuse ourselvesand one colleagues,eitherto the rightor the of biasin a politicalthanin an left,and thusbe moreliable to quesanother apoliticalsituation,for at least two tionthatconvention.) In anyevent,evenif our colleagues of reasons.First,becausethehierarchy in has been openlycalledinto do notaccuseus of bias in research credibility question,we are awarethatthereare a political situation,the interested at leasttwo sidesto the storyand so partieswill. Whethertheyare foreign to investigatepoliticianswho object to studiesof do notthinkit unseemly the situationfromone or anotherof how the stability of theirgovernment in the interest of the contendingpointsof view. We maybe maintained This content downloaded from 188.72.96.102 on Sun, 22 Jun 2014 17:18:10 PM All use subject to JSTOR Terms and Conditions WhoseSide Are We On? 245 the UnitedStates(as in the Camelot or anotherside to a relationship and civilrightsleaders willthusbe affected, at least,byhaving affair)4or domestic who object to an analysisof race suggestedto him what the relevant and issuesare.A student of problemsthat centerson the alleged arguments deficiencies of the Negro family(as medicalsociologymaydecidethathe in thereception theperspective of the givento theMoynihan will takeneither Report),5 interested partiesare quick patientnor the perspectiveof the to make accusationsof bias and dis- physician, but he will necessarily take tortion.They base the accusationnot a perspectivethat impingeson the on failuresof techniqueor method, many questionsthat arise between buton conceptual defects. Theyaccuse physiciansand patients; no matter thesociologist notof gettingfalsedata what perspective he takes,his work butof notgettingall thedatarelevant eitherwill take into accountthe attior it will not.If to the problem.They accusehim,in tudeof subordinates, otherwords,of seeingthingsfromthe he failsto considerthequestionsthey of only one partyto the raise,he will be workingon the side perspective If he does raisethose conflict. But theaccusationis likelyto of the officials. be made by interested and doesfind,as he partiesand not questionsseriously themselves. by sociologists may,thatthereis somemeritin them, WhatI havesaid so faris all sociol- he will then expose himselfto the and of all those ogy of knowledge,suggestingby outrageof theofficials who award themthe top and forwhat sociologists whom,in whatsituations of credibility. reasonssociologists will be accusedof spot in the hierarchy bias and distortion. I have notyetad- Almostall the topicsthatsociologists dressedthe questionof the truthof study,at least thosethathave some the accusations, of whetherour find- relationto the real worldaroundus, for are seen by societyas morality plays ingsare distorted byour sympathy thosewe study.I haveimplieda partial and we shallfindourselves, willy-nilly, answer,namely,thatthereis no posi- takingpartin thoseplayson one side tion fromwhichsociologicalresearch or the other. can be done thatis not biasedin one There is anotherpossibility.We or anotherway. may,in somecases,takethe pointof We mustalwayslook at thematter view of somethirdpartynot directly fromsomeone'spoint of view. The implicatedin the hierarchywe are Thus, a Marxistmight scientistwho proposesto understand investigating. must,as Mead longago pointed feelthatit is notworthdistinguishing society out, get into the situationenoughto betweenDemocratsand Republicans, have a perspective on it. And it is or betweenbig businessand big labor, that his likely perspectivewill be in eachcase bothgroupsbeingequally of theworkers. greatlyaffected by whateverpositions inimicalto theinterests are takenby any or all of the other This would indeed make us neutral in that varied situation. with respectto the two groups at participants Even if his participation is limitedto hand, but would only mean thatwe readingin thefield,he willnecessarilyhad enlargedthescopeof thepolitical readthearguments of partisans of one conflictto includea partynot ordinarily broughtin whose view the was taking. 4 See Irving Louis Horowitz, "The Life sociologist and Death of Project Camelot," TransacWe can neveravoid takingsides. tion, 3 (Nov./Dec., 1965), pp. 3-7, 44-47. So we are left with the questionof 5 See Lee Rainwater and William L. whether takingsidesmeansthatsome Yancey, "Black Families and the White distortion is introduced intoour work 3 House," ibid., (July/August,1966, pp. so greatas to makeit useless.Or, less 6-11, 48-53). This content downloaded from 188.72.96.102 on Sun, 22 Jun 2014 17:18:10 PM All use subject to JSTOR Terms and Conditions 246 SOCIAL PROBLEMS whethersome distortion is methodology are no help here.They drastically, introducedthat must be taken into tellus how to guardagainsterror,but accountbeforetheresultsof ourwork theydo not tell us how to makesure can be used. I do not referhere to that we will use all the safeguards feelingthatthe picturegivenby the availableto us. We can, for a start, researchis not "balanced,"the indig- tryto avoid sentimentality. We are nationarousedby havinga conven- sentimental whenwe refuse,forwhatdefinition of real- everreason,to investigate somematter tionallydiscredited ity given priorityor equalitywith that should properlybe regardedas what"everyone We are sentimental, esknows,"forit is clear problematic. thatwe cannotavoidthat.That is the pecially,when our reasonis thatwe problemof officials, spokesmenand would prefernot to know what is interested parties,not ours.Our prob- going on, if to know would be to lem is to make sure that,whatever violatesomesympathy whoseexistence point of view we take,our research we maynot evenbe awareof. Whatmeetsthe standardsof good scientificeverside we are on, we mustuse our work,thatourunavoidablesympathiestechniquesimpartially enoughthat a do notrenderourresults invalid. beliefto whichwe are especiallysymWe mightdistortour findings, be- patheticcould be proveduntrue.We causeof oursympathy withone of the mustalwaysinspectourworkcarefully parties in the relationshipwe are enough to know whetherour techstudying, by misusingthe tools and niques and theoriesare open enough of ourdiscipline. We might to allow thatpossibility. techniques introduceloaded questions into a Let us consider, whatmight finally, or act in somewayin a seema simplesolutionto theproblems questionnaire, fieldsituationsuchthatpeoplewould posed.If thedifficulty is thatwe gain be constrained to tellus onlythekind sympathy withunderdogsby studying of thingwe are alreadyin sympathythem,is it notalso truethatthesuperwith. All of our researchtechniques ordinates in a hierarchical relationship are hedged about with precautionaryusuallyhave theirown superordinates measuresdesignedto guard against withwhomtheymustcontend?Is it these errors.Similarly,thoughmore not true that we mightstudythose or subordinates, abstractly, everyone of our theories superordinates precontainsa setof directives sentingtheirpoint of view on their presumably whichexhaustively coversthefieldwe relations withtheirsuperiors and thus are to study,specifying all the things gaininga deepersympathy withthem we areto lookat and takeintoaccount and avoidingthe bias of one-sided withthosebelow them? in our research. By usingour theories identification and techniquesimpartially, we ought This is appealing,but deceptively so. to be able to studyall thethingsthat For it onlymeansthatwe willgetinto needto be studiedin sucha wayas to the same troublewith a new set of get all the facts we require,even officials. It is true,forinstance,thattheadthoughsomeof thequestionsthatwill be raisedand some of the factsthat ministrators of a prisonarenotfreeto will be producedrun counterto our do as theywish,not freeto be rebiases. for sponsiveof thedesiresof inmates, If onetalksto suchan official, But the questionmay be precisely instance. this. Given all our techniquesof he will commonly tell us, in private, in the theoretical and technicalcontrol,how thatof coursethesubordinates can we be surethatwe willapplythem relationship have some righton their and acrosstheboardas they side, but thattheyfail to understand impartially need to be applied?Our textbooks in thathis desireto do betteris frustrated This content downloaded from 188.72.96.102 on Sun, 22 Jun 2014 17:18:10 PM All use subject to JSTOR Terms and Conditions WhoseSide Are We On? 247 by his superiorsor by the regulations conditionsare the same elsewhere.I Thus,if a prison referto a moresociologicaldisclaimer theyhaveestablished. is angeredbecause we in whichwe say,forinstance, administrator thatwe take the complaintsof his inmates have studiedthe prisonthroughthe we mayfeelthatwe can get eyesof the inmatesand not through seriously, aroundthatand get a morebalanced the eyes of the guardsor otherinhim and his volvedparties.We warnpeople,thus, pictureby interviewing If we do, we maythenwrite thatourstudytellsus onlyhow things associates. a reportwhichhis superiorswill re- look fromthat vantagepoint-what spond to with criesof "bias." They, kinds of objects guards are in the in theirturn,will saythatwe havenot prisoners'world-and does not atpresenteda balancedpicture,because temptto explainwhyguardsdo what we havenotlookedat theirside of it. theydo or to absolvethe guardsof And we mayworrythatwhattheysay what may seem, fromthe prisoners' is true. side, morallyunacceptablebehavior. The point is obvious.By pursuing This will not protectus fromaccusathis seeminglysimple solution,we tionsof bias,however,fortheguards arriveat a problemof infinite bytheunbalanced regress. willstillbe outraged For everyonehas someonestanding picture.If we implicitlyaccept the of credibility, above him who preventshim from conventional hierarchy doing thingsjust as he likes. If we we willfeelthestingin thataccusation. It is something of a solutionto say questionthe superiorsof the prison a state department of thatover the yearseach "one-sided" administrator, or prisons,theywill com- study will provoke furtherstudies corrections plain of thegovernorand thelegisla- thatgradually enlargeour graspof all ture.And if we go to the governor the relevantfacetsof an institution's and thelegislature, theywill complain operation. But thatis a long-term soluof lobbyists, thepublic tion, and not muchhelp to the inpartymachines, and the newspapers. Thereis no end dividualresearcher whohas to contend to it and we can neverhave a "bal- withtheangerof officials who feelhe anced picture"untilwe have studied has donethemwrong,thecriticism of all of societysimultaneously. I do not thoseof his colleagueswhothinkhe is a one-sidedview,and his proposeto hold mybreathuntilthat presenting own worries. happyday. thedemands What do we do in the meantime? We can,I think, satisfy of our scienceby alwaysmakingclear I theanswersaremoreor less the limitsof what we have studied, suppose obvious.We takesidesas ourpersonal markingtheboundaries beyondwhich and dictate,use politicalcommitments our findings cannotbe safelyapplied. our theoretical and technicalresources Not just the conventional disclaimer, that might in whichwe warnthatwe have only to avoid the distortions intoourwork,limitourconstudieda prisonin New Yorkor Cali- introduce carefully, recognizethe hierforniaand the findings maynot hold dclusions forwhatis is, and in the otherforty-nine states-which archyof credibility since field as best we can the accusations is nota usefulprocedure anyway, thefindings mayverywell hold if the and doubtsthatwill surelybe ourfate. This content downloaded from 188.72.96.102 on Sun, 22 Jun 2014 17:18:10 PM All use subject to JSTOR Terms and Conditions