1995-Sutton-&-Staw-ASQ-What-theory-is-not

WhatTheoryisNotAuthor(s):RobertI.SuttonandBarryM.StawSource:AdministrativeScienceQuarterly,Vol.40,No.3(Sep.,1995),pp.371-384Publishedby:JohnsonGraduateSchoolofManagement,CornellUniversityStableURL::07/04/201020:24YouruseoftheJSTORarchiveindicatesyouracceptanceofJSTOR'sTermsandConditionsofUse,availableat://=cjohn.EachcopyofanypartofaJSTORtransmissionmustcontainthesamecopyrightnoticethatappearsonthescreenorprintedpageofsuchtransmission.JSTORisanot-for-profitservicethathelpsscholars,researchers,andstudentsdiscover,use,andbuilduponawiderangeofcontentinatrusteddigitalarchive.Weuseinformationtechnologyandtoolstoincreaseproductivityandfacilitatenewformsofscholarship.FormoreinformationaboutJSTOR,[e-mail]anagement,CornellUniversityiscollaboratingwithJSTORtodigitize,preserveandextendaccesstoAdministrativeScienceQuarterly.(SBR-9022192).Thisessaydescribesdifferencesbetweenpapersthatcontainsometheoryratherthannotheory.The~reislittleagreementaboutwhatconstitutesstrongversusweaktheoryinthesocialsciences,butthereismoreconsensusthatreferences,data,variables,diagrams,andhypothesesarenottheory.Despitethisconsensus,however,authorsroutinelyusethesefiveelementsinlieuoftheory.Weexplainhoweachofthesefiveelementscanbeconfusedwiththeoryandhowtoavoidsuchconfusion.Bymakingthisconsensusexplicit,wehopetohelpauthorsavoidsomeofthemostcommonandeasilyavertedproblemsthatleadreaderstoviewpapersashavinginadequatetheory.Wethendiscusshowjournalsmightfacilitatethepublicationofstrongertheory.Wesuggestthatifthefieldisseriousaboutproducingstrongertheory,journalsneedtoreconsidertheirempiricalrequirements.Wearguethatjournalsoughttobemorereceptivetopapersthattestpartratherthanallofatheoryanduseillustrativeratherthandefinitivedata.Theauthors,reviewers,readers,andeditorswhoshapewhatispublishedinASQinsist,perhapsaboveallelse,thatarticlescontainstrongorganizationaltheory.ASQ'sNoticetoContributorsstates,Ifmanuscriptscontainnotheory,theirvalueissuspect.Aprimaryreason,sometimestheprimaryreason,thatreviewersandeditorsdecidenottopublishasubmittedpaperisthatitcontainsinadequatetheory.ThispaperdrawsonoureditorialexperiencesatASQandResearchinOrganizationalBehavior(ROB)toidentifysomecommonreasonswhypapersareviewedashavingweaktheory.Authorswhowishtowritestrongtheorymightstartbyreadingthediverseliteraturethatseekstodefinetheoryanddistinguishweakfromstrongtheory.TheAcademyofManagementReviewpublishedaforumontheorybuildinginOctober1989.Detaileddescriptionsofwhattheoryisandthedistinctionsbetweenstrongandweaktheoryinthesocialsciencescanbefound,forexample,inDubin's(1976)analysisoftheorybuildinginappliedareas,Freese's(1980)reviewofformaltheorizing,Kaplan's(1964)philosophicalinquiryintothebehavioralsciences,Merton's(1967)writingsontheoreticalsociology,andWeick's(1989)ideasabouttheoryconstructionasdisciplinedimagination.Unfortunately,theliteratureontheorybuildingcanleaveareadermoreratherthanlessconfusedabouthowtowriteapaperthatcontainsstrongtheory(Freese,1980).Thereislackofagreementaboutwhetheramodelandatheorycanbedistinguished,whetheratypologyisproperlylabeledatheoryornot,whetherthestrengthofatheorydependsonhowinterestingitis,andwhetherfalsifiabilityisaprerequisitefortheveryexistenceofatheory.AsMerton(1967:39)putit:Likesomanywordsthatarebandiedabout,thewordtheorythreatenstobecomemeaningless.Becauseitsreferentsaresodiverse-includingeverythingfromminorworkinghypotheses,throughcomprehensivebutvagueandunorderedspeculations,toaxiomaticsystemsofthought-useofthewordoftenobscuresratherthancreatesunderstanding.371/AdministrativeScienceQuarterly,40(1995):371-384Lackofconsensusonexactlywhattheoryismayexplainwhyitissodifficulttodevelopstrongtheoryinthebehavioralsciences.Reviewers,editors,andotheraudiencesmayholdinconsistentbeliefsaboutwhatconstitutestheoryandwhatconstitutesstrongversusweaktheory.Aspiringorganizationaltheoristsfacefurtherobstaclesbecausethereislittleconsensusaboutwhichtheoreticalperspectives(andassociat