American Bar Foundation

Critical Empiricism: [Comment] Author(s): William C. Whitford Source: Law & Social Inquiry, Vol. 14, No. 1 (Winter, 1989), pp. 61-67 Published by: Blackwell Publishing on behalf of the American Bar Foundation Stable URL: Accessed: 08/12/2010 20:23
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at 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 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

Blackwell Publishing and American Bar Foundation are collaborating with JSTOR to digitize, preserve and extend access to Law & Social Inquiry.


Critical Empiricism
William C. Whitford Learning and Values for Thereis a fundamental that deparadox research is traditionally is as The of scribed "empirical." ostensible purpose this research to learn and observaYet, as Trubek Esseremphasize, something observation. by tion cannotbe objective valuefree. Thereare two relatedproblems or valueladen. that renderall observation inherently of that pheDescription the socialworldrequires we groupdiscrete or nomenainto categories we believe, aretrained believe, that to describe In socialhappenings. ourlanguage describe we thesecategories significant with a singleword,and cometo thinkof themas a singlephenomenon, of ratherthan the grouping discretephenomena that could have been in someotherway. Thus,we talkof disputes, we aretrained and grouped to thinkof a marital and resistance an enterprise an environto spat by mentalprotection orderas related socialphenomena. we do But agency's not think of the questionwhetherIvanLendlor MatsWilander the is world's tennisplayer related, best as we because arenot trained describe to this as a "dispute." the Consistently, process whichthe latter by question we ratherthan "disputes The gets resolved call a "game" processing." we usetodaywerecreated ourforebears, theycommonly and categories by reflect abouthowsociety shouldbe organized werewidely that preferences sharedin the culture whichtheywerecreated.In Trubek Esser's in and the of in terminology, categories socialbehavior use todayin empirical research at leastpartly product ideologies have been domiare a of that nant in relevant population subgroups.'
William C. Whitford is GeorgeH. Young-Bascom of Professor Lawat the University of Wisconsin-Madison.LL.B. 1964, Yale University. 1. For a fullerdescriptionof what I have elsewherecalled the problemof conceptualResearchin a Post-CLSWorld," ism, see Whitford, "LoweredHorizons: Implementation 1986 Wis.L Rev.755, 767-72. The argumentI developlaterto justifythe utilityof observation is a refinementof an argumentfirst stated in that earlierarticle. Id at 769. © 1989 AmericanBar Foundation 61



A related problem has come to be called the problem of agency, which is much emphasizedin the Trubek and Esser essay, where it becomes the basis of their principalcriticismof the work of the Amherst group. Even though our thought is to a significantextent imprisonedby the set of social categorieswe have inherited, we retain the capacityto imaginenew groupingsof social phenomena. Hence, we retain the capacity to reject traditionalunderstandingsabout who benefits from our existing socialorder and about the potentialitiesand meansof changingthat order.2 But these new understandings, acquiredthrough imaginationand are in part a function of the valuesand desiresof the through perception, person imaginingand/or perceiving(i.e., the agent). Existingsocial practices reflect an acceptanceof the idea that individualvalues and desires that person acquiresthrough imcan partlydeterminethe understandings and perception. Thus, in sporting events, as in legal trials, we agination seek officials who do not identify emotionallywith either contestant, in part becausewe fear consciouslybiasedjudgment,but also becausewe understandthat perceptionis affectedby emotional attachment.3 In sum, all observationis value laden for two very differentreasons. One reason concerns the value-ladencharacterof the concepts, and ultimatelyall language,that we use to describethe social order, and is much emphasizedby the structuralist philosophicaltradition. The other reason stressesthe impossibilityof fully separatingdescriptionfrom evaluation, and ultimatelycomes from our understandingof the self. Together, as are Trubekand Esseremphasize,these sourcesof nonobjectivity enough to invalidatewhat they call "uniformscientism"-the idea so prevalentin recentWesternculturethat throughuse of the "scientificmethod"we can learn truths about the nature of human society. Trubekand Esser'sessaydoes not take what might seem Interestingly, to many readersto be the next logical step: advocatingabandonmentof observationas a researchstrategyin the face of its inherentlyvalue-laden character. What reason do we have to believe that through observation we can learn anythingthat cannot be learnedfrom analysisof the values containedin the languageand conceptswe have inheritedand of our own personalgoals and desires?4It seems clear that Trubek and Esserbelieve
2. These capacitiesare much emphasized Coombe in an essayin this issue and are by the sourcesof what she there calls subjectivity. 3. A graphic,and tragic,recent exampleof this common understanding the official is explanationgiven for the shooting down of an Iranianpassengerairlinerby an American warshipin the PersianGulf. That explanationattributedthe tragedyto a misreadingof imageson a radarscreenby Americanservicemenin battle for the first time and biasedby the emotions of that circumstance interpretthe imagesas suggestingan attackon their to ship that was in fact not occurring. 4. Though never with quite the specificitystated in the text, there are critiquesof the Law & Society movementby other CriticalLegalScholarsthat come close to questioning the abilityof empiricalresearchto teach anythingnot learnablein other ways. See Kelman,

Critical Empiricism


that, in spite of the problems, something can be learned from observation.5 But their essaydoes not makea positivecase for the utilityof observation, despite its flaws. I attempt that next. The foundationof my case for observationis a plea that we acknowledge the experienceof learning. Learning,as I use the term, is the experience of believingthat one has acquiredan enhancedabilityto describethe causalrelationshipbetweendifferentevents. Becauselearninginvolves inof creasedunderstanding causalrelationships,it implies an abilityto predict the futurewith greateraccuracy. Predictionis a criticalskill if we are to be able to exert greatercontrol over our social environment. The phenomenon of learning, as I have defined it, seems to me to account for experienceswe have all had. In familylife, we learn behaviors that will avoid or alternatelystimulate disagreement,depending on our wants at the moment. Basketball teams learn what offenses seem to work best against different zone and person-to-persondefenses, respectively. Lawyerslearn what kinds of statementsto particularappellatejudges are most likelyto be persuasive.Note that none of these examplesof learning implies an abilityto predictthe futurewith absolutecertainty,just with a of than previouslywas possible. Furthermore, greaterprobability accuracy this belief in an enhanced predictive ability is fully compatiblewith an appreciationthat one's new understandingis only tentative, to be displacedin the futureby other new insights,perhapsfurtherenhancingpredictivecapacityor perhapsmore appropriate the changedcircumstances to of a new age. The second step in my justificationof observationis to link observation to learning. Learningis acquiredin differentways, not all based on observation. We feel as though we are learningwhen through exerciseof deductive mental processeswe discover new implicationsof propositions alreadyaccepted. Use of this analytictechnique appearsto account for manyof the law reviewarticlesthat professthat if the goal is maximization of resources,then the legalrule should be x, whereasif the goal is equality in the distributionof resources,then the rule should be y.6 Other types of learning experiences can be considered "transformative."Learning is as about transformative, I use the term, when it leads to understandings the natureof the world that could not be deduced solely from preexisting understandings.7
36 "Trashing," Stan.L Rev.293 (1984);White, "FromRealismto CriticalStudies: A Truncated IntellectualHistory,"40 Sw. LJ. 819 (1986). 5. Trubekand Esser,at note 77, very specifically deny any rejectionof what they call which I presumeincludesobservation.Their definitionof "empir"investigatory practices," icism,"text at note 84, seemsto contemplatethat what they call empirical researchneed not include observation,however. 6. E.g., Goetz & Scott, "Principlesof Relational Contracts," 67 Va. L Rev. 1089 (1981). 7. Sometimeslearningis said to be transformative only if the learningresultsin dis-



My case for observationrestson its abilityto aid and abet transformative learning. There is no doubt that imaginationand speculativethought are importantand perhapsindispensable tools in acquiringtransformative Archimedesdiscoveredhis famous principlewhile contemplatlearning. ing.8 Observationalone would not have sufficed. But observationdid precede the contemplation, and lent confidence to the product of the coachesobservethe playof their own basketball contemplation. Similarly, and other teams in learningwhat offenseswill work best againstdifferent defenses. Another example,drawnfrom socialresearchon lawthat will be famousrefamiliarto most Americanlegal scholars,is StewartMacaulay's search on the contractual behavior of businessmen. Macaulay'sdirect observationof contractualbehavior, as well as his interviewswith businessmenengagedin contractualtransactions,led him, and laterthe rest of us, to new understandingsabout the relationshipbetween contract law and contractualrelations.9 The case I have made for observation rests on acknowledgmentof mental experiencesI have had, and believe others have had, that I believe has been useful in enabling me to predict the future more accurately. There is no way I can prove, accordingto the standardsof validity normallyprofessedby positivistsocial sciences,that observationhas this utility. The value-ladencharacterof observation,as describedearlier,would prevent any proof of my case, just as it preventsproof, in that positivistic sense, of any other propositionabout the social order. But in asking for of acknowledgment the experienceof learningthrough observation,I believe I am not askingfor a leap of faith that is differentin kind from what is requiredto acceptthe propositionswith which this commentbeganconcerning the value-ladencharacterof observation. How do we know that the goals and desiresof the observerinevitablyaffectthe productof observation? I believe we accept this proposition because after reflection it seems an accurateaccount of experience we have all had. My case for learningand for observationas a source of learningis similarlybased. Observationis the essential ingredientin what I consider empirical research. For one who acknowledges utility of observation,what imthe are there for the practiceof empiricalresearchif one also acplications to cepts that all observationis value laden? It is possiblefor the researcher
cardingwhole world views, or ideologies,in favorof others. But I have in mind as "transformative"even little insights as long they are not logicallycompelled from preexisting knowledge. The second partof this commentdiscusseswhetherlittle insightscan qualifyas "criticalempiricism." 8. Archimedes'principleis that an object submergedin waterdisplacesits volume regardlessof the object'sweight. I was told in my high school physicsclass that Archimedes arrivedat the insight while in the bathtub,whereuponhe jumpedout, unclothed,and ran into the street shouting "Eureka, have it." I 9. Macaulay,"Non-contractual Relationsin Business: A Preliminary Study," 28 Am. Soc. Rev. 55 (1963).

Critical Empiricism


be self-consciously awareof the possibilityof the value-ladencharacterof and perhapseven of some of the valuesthemselves. Awareher/his work, ness can beget becomingmodesty,and it may help limit misinterpretation of results by remindingboth the researcherherself and any consumerof nature. publishedresultsof the researchof its contingentand probabilistic In an earlierarticle'°I arguedthat an awareness the value-ladencharacof ter of observationshould cause the researcherto prefer studies that are aimed at more particularpropositions (i.e., more emphasison studies of particularlocales at particulartimes, less emphasis on studying propositions about common featuresof a wide set of social practices)and that make more use of "softer"sources of information(more in-depth interof views, less statistics,less emphasison reproducibility results)than has been in fashion in social science researchof the past few decades. The Politics of Empirical Research There has been a lengthydebate about what should be meant by the term "criticalempiricism."It may be that the term should be restrictedto awareof its value-ladencharacempiricalresearchthat is self-consciously ter, a view I associatewith Trubekand Esser. The view has been expressed by some membersof the Amherstgroup that the subjectmatterof empirical researchshould bear on whether the term "critical"is appropriately
used in describing the research." To be "critical," according to this view, research must be directed at discrediting the assumptions underlying the existing legal order or at expressing the point of view and advancing the interests of underrepresented groups. Research designed to discover better ways to achieve some specific policy objective does not qualify. It is not my intention to take a position on a purely definitional issue, but I do want to disassociate myself from any implication that research not defined as "critical" by these authors is less valid. All research, even nonempirical research, is value laden for the reasons discussed above. The "turn to interpretation"'2 is not an escape from the values imbedded in the categories we have inherited for describing behavior. The attempt to give voice to excluded interests, advocated by some as a true form of critical empiricism, is just a form of interpretation. Absent the transformative experience, these received categories limit our ability to understand the goals and objectives desired and sought by some social group. And it should be self-evident that interpretive work does not avoid the problem of agency either. Thus, it is no more possible to state objectively the
10. Whitford, 1986 Wis. L Rev. at 776-79 (cited in note 1). 11. See Sarat& Silbey, "The Pull of the PolicyAudience," 10 Law& Policy97 (1988). 12. See Kennedy,58 S. Cal. L. Rev. 251 (1985).



"true"viewpointof some excludedgroupthan it is to state objectivelythe "true"cause of some accident or environmentaldegradation.13 The argumentagainst policy-specific research,if one is to be made, must be based on an assessmentof politicaltactics. Some membersof the Amherst group have suggestedthat policy-specific researchis most likely to advance the interests of the powerfulratherthan the powerless.'4By this view researchers have limited abilityto shape the questions asked. If the researchers want their work to be used by policy-makers, they must addressquestions the policy-makers want answered. Nor can researchers control the interpretationof their results. To avoid cooptation by the powerful, accordingto this view, the only sound political strategyis to avoid researchpointing towardspecific policy goals, reservingone's energies for projectsthat can help mobilize political constituenciesthat will supportfundamentalpoliticalchange. Frequentlysuch researchwill focus on delegitimizing legal order as it presentlyfunctions, demonstrating the how it favors powerful interests and fails to recognize the interests of others. There is much sage adviceto those on the politicalleft in these warnabout the uses made of much policy-specific research. But it would be ings a mistaketo understandsuch adviceto representuniversalpoliticaltruth. In what we call Westerndemocracies, believe it makessense to use such I terms as ruling classes and disadvantagedgroups, but I also think that authorityin these societies is not so hegemonicas to make impossiblereform benefitingthe constituenciesthat the left desiresto serve. The welfare state reforms of the 20th century, though far from perfect, are to preferable what precededthem.15 And becausedesirablereformis possible in these societies,policy-specific researchcan play a politicallyacceptin structuringsuch reformto be modestly more effective. It is able role possible for the researcherto be the coopter rather than the coopted.'6
13. This point is discussedextensivelyby Trubekand Esserand is the basis on which they direct their strongestcriticismat the work of the Amherstgroup. On the subjectivity of causation,see Kelman,"The NecessaryMyth of ObjectiveCausationJudgmentsin Liberal PoliticalTheory," 63 Chi-Kent L Rev. 579 (1987). 14. Sarat & Silbey, 10 Law & Policy. 15. I rejectthe view, associatedwith some strainsof the Marxistleft, that by legitimating the existingorder,the welfarestate reformshave delayedthe revolutionthat represents the only true hope for progressive change. I agreethat the welfarestate reformshave tended to legitimatethe existing order, but I am not so confident that revolution,ratherthan a successfulrepressionand furthersubordinationof disadvantaged groups,would have been the consequenceof a failureto adopt them. Nor is revolutionalwaysa more desirablealternative than incremental reform. Witness(insertwhateverrevolutionary society particularly appallsyou-e.g., Iran). 16. Joel Handler'sresearchon the Americanwelfaresystemand HermanGoldstein's work on the American police seem to me examplesof politicallycorrect implementation research. Both scholarsare intensely empirical,engagingin extensive observation. They have concentratedon studyingand promotingsuccessfulreformexperiments,and I suspect their work has helped these experimentsto surviveand perhapseven to expand. E.g., J.

Critical Empiricism


Not all circumstances be amenableto progressivechange, of course. will For the criticalresearcher, there is no substitutefor close attention to the of the moment. political possibilities

Decision(1986);H. Goldstein, Reshaping PoliceFunction: the The Handler,The Discretionary Problem-oriented Approach (forthcoming,1989).