2002ElsevierScienceLtdAllrightsreservedPrintedinGreatBritainDesignmethodologyandthenatureoftechnicalartefactsPeterKroesDepartmentofPhilosophyFacultyofTechnologyPolicyandManagementDelftUniversityof ID: 314997
Download Pdf The PPT/PDF document "0142-694X/02$-seefrontmatterDesignStudie..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
0142-694X/02$-seefrontmatterDesignStudies(2002)287302PII:S0142-694X(01)00039-4 2002ElsevierScienceLtdAllrightsreservedPrintedinGreatBritainDesignmethodologyandthenatureoftechnicalartefactsPeterKroes,DepartmentofPhilosophy,FacultyofTechnology,PolicyandManagement,DelftUniversityofTechnology,Jaffalaan5,Nl-2628BXDelft,TheNetherlandsDesignmethodologyaimsattheimprovementofdesignprocesses.Incontrasttothemethodologyofscience,itisstronglyprocessorientedandtakesanormativepointofview.Thepaperarguesthat,despiteitsprocessorientation,designmethodologycannotavoidaddressingquestionsconcerningthenatureoftheproductsbeingdesignedfortworeasons.TheÞrstisthatthedesignprocessandthedesignproductaresointimatelyrelatedtoeachotherthatanunderstandingofthenatureofthedesignprocessrequiresinsightintothenatureofthekindofproductdesigned,andviceversa.Second,inordertojustifyitsnormativestancetowardsdesignprocesses,designmethodologywillhavetoconsidertheissueofthequalityoftheproductbeingdesigned.Tosupporttheseclaims,thenatureoftechnicalartefacts,consideredtobetheoutcomeofadesignprocess,willbeexamined.Itwillbearguedthattheyhaveadualnature:theyarephysicalobjectsontheonehand,andintentionalobjectsontheother.Someoftheconsequencesofthisdualnatureoftechnicalartefactsfortheresearchagendaofdesignmethodologywillbeexplored.c2002ElsevierScienceLtd.AllrightsKeywords:designmethodology,philosophyofdesign,designprocess,designstudies,humanfactorsheaimofdesignmethodologyistoimprovedesignprocesses;thismeansthatittakesanormativestancetowardsitsobjectofstudy.Giventhisaimitisnosurprisethatresearchindesignmethodologyhasalwayshadastrongfocusonthenatureofdesignprocesses.Thestudyofthenatureoftechnicalartefacts,consideredtobetheoutcomeofadesignprocess,hasreceivedlittleattention.Forseveralreasons,however,includingitsnormativestandpoint,designmethodologycannotavoidacloseranalysisofthenatureoftechnicalartefacts.Here,aninterpretationoftechnicalartefactsintermsofadualnaturewhichreferstothefact Carnap,RIntellectualautobi-ographyinPASchilppThephilosophyofRudolfCar-,OpenCourt,LaSalle,ILDorst,KDescribingdesign;acomparisonofparadigmsPhDthesis,DelftUniversityofTech-nology,TheNetherlands,1997Love,TPhilosophyofdesign:ameta-theoreticalstruc-turefordesigntheoryVol21(2000)293313288DesignStudiesVol23No.3May2002thattheyarephysicalandintentionalobjectsatthesametimeisofferedanditisarguedthatthisinterpretationhasfar-reachingconsequencesfortheresearchagendaofdesignmethodology.Thepaperstartswithacom-parisonofdesignmethodologywiththemethodologyofscience.Thisisfollowedbyanexpositionofthedualnatureoftechnicalartefacts.Finally,someconsequencesofthisinterpretationoftechnicalartefactsfortheresearchagendaofdesignmethodologyarediscussed.DesignmethodologyversusmethodologyofscienceTherearetwostrikingdifferencesinorientationbetweenmethodologicalstudiesoftechnicaldesignandofscientiÞcresearch.Designmethodologytakesanormativestancetowardsdesignandisverymuchprocessoriented,whereasresearchmethodologyisdescriptiveandstronglyproductoriented.WewillÞrsthaveacloserlookatthesedifferences.Methodologicalstudiesofscience,aspartofthebroaderdisciplineofthephilosophyofscience,typicallyconcentrateontheoutcomesofscientiÞcresearchprocesses,suchasempiricalclaims,laws,theoriesandexpla-nations,andfocusonquestionsabouttheinterpretationoftheseproductsandtheirreliability(ortruth).ThisproductorientationinthemethodologyofscienceisrelatedtotheclassiclogicalpositivistÕsdistinctionbetweenthecontextofdiscovery(howarephenomena,laws,theories,etc.,discovered?)andthecontextofjustiÞcation(howarephenomena,laws,theories,etc.,justiÞed?)andtheirhighlyinßuentialideathatthereisnoÔlogicÕofscientiÞcdiscovery.Insofaraslogicalpositivistswereprocessoriented,theywereinterestedinaveryspecialkindofprocess,namelyaÔrationalreconstructionofscienceÕ,thatisaccordingtoCarnapÔaschem-atiseddescriptionofanimaginaryprocedure,consistingofrationallypre-scribedsteps,whichwouldleadtoessentiallythesameresultsastheactual[É]processÕ(p.16).Therealresearchprocess,asitwasactuallyperfor-med,wasonlyofminorinteresttothemorofnointerestatall.Admittedly,therehasbeenanempiricalturninthephilosophyofsciencesincetheworkofKuhn,butalthoughthisturnhasledtomoreinterestinactualresearchprocesses,particularlyinexperiments,theunderlyingissuesremain,asbefore,issuesabouttheinterpretationandjustiÞcationofscien-tiÞcknowledgeclaims.Moreover,mainstreammethodologyofscienceisnotdrivenbytheaimtoimprovethepracticeofscientiÞcresearch.Itismainlyadescriptiveactivity.Designmethodology,characterisedbyoneofitsleadingÞgures,NigelCross,asÔthestudyofprinciples,practicesandproceduresofdesignÕ(seenote1),aimsatimprovingdesignpracticeandisstronglyprocessoriented. Cross,NAHistoryofdesignMJDeVries,NCrossandDPGrantDesignmethodologyandrelationshipswithscienceKluwerAcademicPublishers,Dordrecht(1993)pp15n,DThereßectiveprac-titioner;howprofessionalsthinkinactionAshgate,AldershotBucciarelli,LLTheMITPress,Cam-bridge,MA(1994)Galle,PDesignrationaliz-ationandthelogicofdesign:acasestudyDesignStudies17No3(1996)253DesignmethodologyFigure1Differencesinfocusbetweenresearchmethodologyanddesign AccordingtoCrossshistoryofdesignmethodology,thefoundingfathersofthisdisciplinehadastrongnormativeattitude:designmethodologyshouldcontributetotheimprovementofdesignpractice,particularlybyexploitingscienticmethods.Crossshistoryalsoillustratesthestrongprocessorientationofdesignmethodology.Hementionsvecategoriesofrecentworkinthiseld,fourofwhichareexplicitlyprocess(activity)oriented(namely,thedevelopmentofdesignmethods,themanagementofdesignprocess,thenatureofdesignactivityandthephilosophyofdesignmethodwhichdealswiththephilosophicalanalysisandreectionondesignactivity);onlytheworkwhichheclassiesundertheheadingnatureofdesignproblemsisnotprocessoriented.Alookatthebroadereldofdesignresearchindeedconrmsthestrongbenttowardsprocessesandactivitiesinthiseld.AccordingtoDorst,twoparadigmswithindesignresearchcanbedistinguished,designasrationalproblemsolvinganddesignasreectivepractice,andbothareprocess.Schostheoryaboutthereectivepractitioner,whichhasattractedmuchattentioninrecentyearswithindesignresearch,approachesdesignasareectiveprocess.Bucciarelliswork,alsowellknownwithineld,analysesdesignasanessentiallysocialprocess.Finally,aquickscanofthecontentsofvolumes16(1995)through22(2001)ofoneoftheleadingjournalsinthisDesignStudies,conrmsthestrongprocessorientation.Thetopicsaddressedtypicallyconcern:(creativityin)designthinking,designeducation,designeffort,conceptualdesignasaprocess,designprogress,communicationofdesignknowledge,managingdesigninformation,theroleofcomputersindesign,designasacognitiveactivity,decisionmakingindesign,etc.Thisjournalexplicitlypresentsitselfasaforumforthediscussionanddevelopmentofthetheoreticalaspectsofdesign,includingitsmethodologyandvalues,butalmostwithoutexceptionthemethodologicalcontributionsconcerntheactualmethodsandtech-niquesusedinsolvingdesigntasks,notthemethodsandtechniquesusedinjustifyingtheoutcomeofadesignprocess(seenote2).So,therearetwodifferencesbetweendesignmethodologyandresearchmethodology:theformertakesanormativestanceandisprocessoriented,thelatterisdescriptiveandproductoriented(seeFigure1).Comparedto VereinDeutscherInge-nieure(VDI)approachtothedesignoftechni-calsystemsandproducts:Guid-elineVDI2221BeuthVerlag,Berlin(1987)KroesPEngineeringdesignandtheempiricalturninthephilosophyoftechnologyKroesandAMeijersempiricalturninthephilosophyoftechnology(CMitcham(seriesResearchinphilosophyandVol20,JAI/Elsevier,Amsterdam(2000)19Simon,HAThesciencesoftheartiednTheMITPress,Cambridge,MA(1996)Kroes,PATechnicalfunc-tionsasdispositions:acriticalVol5No3(2001)16Losonsky,MThenatureofVol65(1990)81290DesignStudiesVol23No.3May2002researchmethodology,designmethodologyisinterestedinaratherdiffer-entkindofrationalreconstruction,namelyinaschematiseddescriptionofreal(notimaginary)designprocedures,consistingofrationallypre-scribedsteps,whichshouldleadtoessentiallybetter(notthesame)resultscomparedwithexistingdesignprocedures.Notsurprisingly,therefore,designmethodologyhasresultedinavarietyofschemesfordividingthedesignprocessintovariousphases,varyingfromtheverysimpleanalysisevaluationschemeto,forinstance,theratherdetailedandelabor-ateschemeproposedbytheVereinDeutscherIngenieure(VDI)Becauseofthesedifferences,designmethodologyandmethodologyofsciencebearlittleresemblancetoeachother.Itisevenconfusingtocallbecausethatsuggeststhattheyaddresssimilarkindsofquestionsfordesignandresearch.Thatisnotinfactthecase.Supposeweweretoconstrueaeldcalleddesignmethodologyanalogoustotheeldcalledmethodologyofscience.Thenitwouldhavetodealwiththefollowingkindsofquestions,someofwhichwillsurfaceagainlateron(seenote3).Whatisadesign?Whatmakesadesignagoodorasuccessfuldesign?Whatarethepropercriteriaforevaluatingproposedsolutionsforagivendesignproblem?Isitpossibletocharacteriseinageneral(logical?)waynotionssuchastheeffectivenessandefciencyofdesignsolutions?Howcanaproposedsolutionforadesignproblemberationallyjusti-Howcandesigndecisionswithregardtotrade-offsbetweencondesignspecicationsberationallyjustiShiftingattentionfromdesignsolutionstodesignmethods,thefollowingquestionscropup.Doesthecorrectapplicationofdesignmethodsguaranteeasuccessfuloutcomeormakeasuccessfuloutcomeprobabletoacertaindegree?Ifso,istherealogicofdesignmethods,thatis,canweunderstandthispropertyofdesignmethodsfromalogical(analytical)pointofview?Furthermore,atechnologicaldesign(ideally)containsanexplanationofhowagivenphysical(chemical,biological)devicerealisesacertainfunc-Howissuchatechnologicalexplanation,i.e.anexplanationofafunc-tionintermsofaphysical(chemical,etc.)structure,possible? DesignmethodologyWhatkindofadequacyconditionsapplytotechnologicalexplanations?ThesequestionsconcerneitherthejusticationoftheoutcomeofthedesignprocessorarationalreconstructionofthedesignprocessintheCarnapiansense(i.e.intermsofimaginarystepsandprocedures).Designmethodology,asithasbeenpractiseduptillnow,haslargelyneg-lectedthesequestions.Becauseitaimsattheimprovementofdesignprac-tice,ithasfocusedmainlyonthedesignprocess.Byanalysingindetailthenatureofthisprocess,ittriestorationallyreconstructitinthesensedescribedearlier.Inmyopinion,however,designmethodologywillhavetoaddresssomeoftheissuesdescribedaboveforatleasttworeasons.Therstisthatthedesignprocessandthedesignproductaresointimatelyrelatedtoeachotherthatanunderstandingofthenatureofthedesignprocessrequiresinsightintothenatureoftheproductdesignedandviceversa.Considerthedesignofvariouskindsofartefacts,e.g.thesteeringwheelofacar,anairbag,acar,acartransportsystem,apolicesystem,alawontrafcregulations.Roughlyspeaking,theseartefactsmaybeorderedonanaxisrangingfromtechnicalobjectsthroughsocio-technicalobjectstosocialobjects.Itisamatteroffactthatthedesignprocesseswhichlieatthebasisofthesevariouskindsofartefactsdifferstrongly.Itseemsimplausiblethatitwillbepossibletoconstructadomain-inde-pendenttheoryaboutdesignprocesses,whichwillcoverallthesecases(seenote4).Ananalysisofthedesignprocessoftechnicalartefactsshouldthereforetakeintoaccountthespecicnatureofthoseobjects(seenote5).Second,thenormativestancetakenbydesignmethodologytowardsthedesignprocessimpliesthatitcannotescapequestionsconcerningthequal-ityoftheoutcomeofthatprocess.Sincethatoutcomeisthedesignofatechnicalartefact,ithastoaddresssomeofthequestionslistedaboveaboutcriteriaforsuccessofadesign.So,letusnowturntoacloseranalysisofthenatureoftechnicalartefacts.ThedualnatureoftechnicalartefactsAccordingtotheviewdefendedbelow,technicalartefactshaveadualnature:ontheonehandtheyarephysicalobjects(man-madeconstructions)thatmaybeusedtoperformacertainfunction,ontheotherhandtheyareintentionalobjectssinceitisthefunctionofatechnicalartefactthatdistinguishesitfromphysical(natural)objectsandthisfunctionhasmean-ingonlywithinacontextofintentionalhumanaction.Beforepresentingthisdualnatureviewoftechnicalartefacts,IwillbrieydiscussHerbertstheoryonarticialthingsasexposedinhisclassicThesciencesofthearti(inthefollowingtext,pagenumbersrefertothisbook).Thistheoryprovestobeausefulsteppingstonetothedualnatureview. 292DesignStudiesVol23No.3May2002ForSimon,thescienceofthearticialwillcloselyresemblethescienceofengineeringbecauseengineeringdealswiththesynthesisofthings.Incontrasttothescientist,theengineerandmoreinparticularthedesignerconcernedwithhowthingstobehowtheyoughttobeinordertoattaingoals,andto(pp.45).Oneofthestrikingfeaturesof(technical)artefactsispreciselythattheycanbecharacterisedintermsoffunctionsandgoals.FunctionsandgoalsareanalysedbySimoninthefollowingway(p.5):Letuslookalittlemorecloselyatthefunctionalorpurposefulaspectofartithings.Fullmentofpurposeoradaptationtoagoalinvolvesarelationamongthreeterms:thepurposeorgoal,thecharacteroftheartifact,andtheenvironmentinwhichtheartifactperforms.Forinstance,thepurposeofaclockistotelltimeandthecharacteroftheclockreferstoitsphysicalmake-up(gears,springs,etc.,foramechanicalclock).Finally,theenvironmentisimportantbecausenoteverykindofclockisusefulineveryenvironment;sundialscanonlyperformtheirfunc-tioninsunnyclimates.SimonsanalysisofartefactsisrepresentedinaschematicwayinFigure2(seenote6).AccordingtoSimon,theenvironmentofanartefactisveryimportantbecauseitmouldstheartefact.Heconsiderstheartefacttobeakindofenvironment,thesubstanceandorganis-ationoftheartifactitself,andanenvironment,thesurroundingsinwhichitoperates(p.6).Theinnerenvironmentoftheartefact,itscharacter,isshapedinsuchawaythatitrealisesthegoalssetintheouterenvironment(p.10).Therefore,thescienceofthearticialhastofocusonthisinterface,sincethecialworldiscenteredpreciselyonthisinterfacebetweentheinnerandouterenvironments;itisconcernedwithattaininggoalsbyadaptingtheformertothelatter(p.113).sdistinctionbetweeninnerandouterenvironmentpointstotwodifferentwaysoflookingattechnicalartefacts.Lookedatfromtheouterenvironment,thetechnicalartefactpresentsitselfprimarilyassomething,whateveritsinnerenvironment,thatfullsacertaingoal,purposeorfunc-tion.Fromthisperspectivetheartefactischaracterisedprimarilyinafunc-tionalway;theinnerenvironmentremainsablackbox.Lookedatfromtheinnerenvironment,theartefactisdescribedassomekindofphysicalsystem;fromthisperspective,thegoalthatitfullsintheenvironmentremainsablackbox(seenote7).AsSimonremarks(p.7)airplane,orabird,wecananalyzethembythemethodsofnaturalsciencewithoutanyparticularattentiontopurposeoradaptation,without DesignmethodologyFigure2SchematicrepresentationofSimonanalysisofartefactswithan referencetotheinterfacebetweenwhatIhavecalledtheinnerandouterThesetwodifferentwaysofcharacterisinganartefact,intermsofitsinnerandouterenvironment,correspondcloselytowhatwecallthedualnatureoftechnicalartefacts.Theviewthattechnicalartefactshaveadualnaturendsitsoriginintheobservationthatweemployinourthinking,speakinganddoingtwobasicconceptualisationsoftheworld,andthatwedonotknowhowtointegratethesetwotogetherintoonecoherentconceptualisation(seenote8).Ontheonehand,weseetheworldasconsistingofphysicalobjectsinteractingthroughcausalconnections.Thiswillbecalledtheconceptualisationwhichisemployedanddevelopedbythephysicalsciences.Ontheotherhand,weseetheworldasconsistingpartlyofagents(primarilyhumanbeings),whointentionallyrepresenttheworldandactintentionallyinit,andwhosebehaviourisexplainedpartlyintermsofreasons(andnotcauses).Thisistheconceptualisationofthe 294DesignStudiesVol23No.3May2002worldwhichunderliesmostofthesocialsciences.Oneaspectofthislatterconceptualisationisthatcertainactivitiesareinterpretedintermsofrealis-ationsofgoalsandthatfunctionsareattributedtocertainobjectsoractivi-ties.Theexistenceofthesetwodifferentconceptualisationsposesaprob-lemincaseswherebothoffercompetingexplanationsforthesamekindofphenomenon,e.g.forraisingahandtovoteinameeting:oneintermsofphysiologicalcauses,theotherintermsofreasons.Thisisthewell-knownmindbodyproblem.Thequestionthatconcernsusishowtechnicalartefactstintothesetwoconceptualisationsoftheworld.Ourstartingpointforexploringthisissuewillbethefollowingcharacterisationoftechnicalartefacts:technicalarte-factsareobjectswithatechnicalfunctionandwithaphysicalstructureconsciouslydesigned,producedandusedbyhumanstorealiseitsfunction(seenote9).Inshort,atechnicalartefactisaphysicalobjectwithatechni-calfunction.Thischaracterisationofatechnicalartefactmakesitahybridkindofobjectwhichdoesnottineitherthephysicalortheintentionalconceptualisation.Lookeduponasmerelyphysicalobjects,technicalarte-tintothephysicalconceptualisationoftheworld;thewaythearte-factworkscanbeexplainedintermsofcausalprocesses.Butasamerephysicalobject,itisnotatechnicalartefact.Withoutitsfunction,theobjectlosesitsstatusasatechnicalartefact.Thismeansthattechnicalartefactscannotbedescribedexhaustivelywithinthephysicalconceptualisation,sinceithasnoplaceforitsfunctionalfeatures.Butneithercanitbedescribedexhaustivelywithintheintentionalconceptualisationsinceitsfunctionalitymustberealisedthroughanappropriatephysicalstructureandtheintentionalconceptualisationhasnoplaceforthephysicalfeaturesofatechnicalartefact(seenote10).Hencetheconclusionthattechnicalarte-factshaveadualnature:ontheonehandtheyarephysical,ontheotherintentionalobjects.Accordingtotheabovelineofthought,thenotionoftechnicalartefactisrelatedtothreekeynotions,namelythenotionofaphysicalstructure,ofa(technical)functionandofacontextofintentionalhumanaction(seeFigure3andnote11).Theinclusionofthecontextofhumanactionintoouranalysisofartefactsneedssomeclarication,sincewehavecharacterisedtechnicalartefactsearlierasphysicalobjectswithtechnicalfunctions.Ihaveincludedthecontextofhumanactionbecauseitmakesnosensetospeakabouttechnicalfunctionswithoutreferencetoacontextofhumanaction.Asremarkedearlier,functionaldiscourseispartoftheintentionalconceptualisationoftheworld;itismeaninglesstospeakabouttechnicalfunctionswithouta Searle,JRTheconstruc-tionofsocialrealityBooks,London(1995)Meijers,AWMrelationalontologyoftechnicalPKroesandAMei-TheempiricalturninthephilosophyoftechnologyMitcham(seriesed)),inphilosophyandtechnology20(2000)81DesignmethodologyFigure3Thedualnatureoftechnicalartefactswithan contextofintentional(human)action.Thiscanbeexpressedinanontologi-calwaybysayingthatsomecontextofhumanactionisconstitutiveforatechnicalfunction.ThisisinlinewithSearlesclaimthattechnicalfunc-tionsareattributed,inorwithregardtosomecontextofhumanaction,toobjects;theyarenotintrinsicpropertiesofthoseobjects.Inhisanalysisoftherelationalontologyoftechnicalartefacts,Meijersalsoclaimsthathumanactionisconstitutiveforfunctions:Acentralpartofmyargumentfocusesonfunctionsandfunctionalproperties.Thesearerealisedbythephysicalstructureoftheartifacttogetherwiththepracticeofitsdesignanduse(p.81).Thus,inFigure3functionandcontextofhumanactionareintimatelyconnected;theybothbelongtothedomainoftheintentional.Technicalartefactshaveadualnaturesincetheyareatthesametimepartofthedomainofthephysicalandoftheintentional.Therearesomenotabledifferencesbetweenouranalysisoftechnicalarte-factsandSimons.Simonsnotionofgoalorpurposehasbeenreplaced 296DesignStudiesVol23No.3May2002bythenotionoffunction.Thismayseemaninsignicantmovebutitisnot,becausewemayattributefunctionstotechnicalartefactsbutnotgoals(inthesenseofanaimoranend(telos)).Thatnotionhasitsplaceinacontextofintentionalhumanaction;withinsuchacontextameansusedtoachieveagoal(end,aim)isattributedafunction.Thus,Simonsanalysisimplicitlyreferstoacontextofhumanactionbyreferringtogoalsandpurposes.Furthermore,thenotionofenvironmenthasbeenreplacedbythenotionofcontextofhumanaction.Itcouldbearguedthatthisisalsoaminorchange,becauseoneformofenvironmentisacontextofhumanaction.Simonsclaimthattheartefacthastoadapttoitsenvironmentthenreducestothe,ratherobvious,claimthattheartefacthastoadapttothecontextofhumanactioninwhichitisused.Nevertheless,thisisanotice-ablechangebecauseitbringsoutthefactthatnotanykindofenvironmentisrelevantfortheanalysisoftechnicalartefacts;onlyreferencestoenvironmentscomprisingacontextofhumanactionareappropriate.Inhisexampleofthesundial,forinstance,Simoninterpretstheenvironmentinaphysicalway(sunnyclimatesaretherequiredenvironmentforsundials).Butthisisproblematic.Itisnotthisphysicalenvironmentthatturnstheobjectinvolved,astickthatcastsashadowonasurface,intoanartefactofthetypesundial.Onlywithinthecontextofhumanaction(e.g.oforderingeventsorcomparingtimeintervals)doesthisphysicalobjectacquireafunctionandbecomeatechnicalartefact(atime-keepingdeviceorclock).ThemaindifferencebetweenSimonsanalysisandoursisthatthelattergivesamuchmoreprominentandexplicitplacetoacontextofhumanactioninanalysingthenatureoftechnicalartefacts.Theadvantageofthisisthatitbringsmuchmoreintotheopenthedualnatureoftechnicalartefacts:wecannotmakesenseoftechnicalartefactswithouttakingintoconsiderationtheirphysicalstructure,butalsonotwithouttheircontextofintentionalhumanaction.WithinSimonsanalysisthisdualnaturestaysmoreimplicitandisrelatedtothetwodifferentperspectivesontechnicalartefacts,namelytheperspectiveoftheinnerenvironment(physicalstructure)andtheperspectiveoftheouterenvironment(contextofhumanaction).Notethattheabovecharacterisationofatechnicalartefactinvolvespro-cessesinanessentialway:withoutsomecontextofhumanaction(activity,processes)thenotionoffunctionlosesitsmeaning,andwhatisleftofatechnicalartefactwithoutitsfunctionisjustsomephysicalobject.Inordertoarriveatabetterunderstandingofhowthedesignprocessisinvolvedincharacterisingatechnicalartefact,wehavetotakeacloserlookatwhatwehavecalledcontextofhumanaction.Thisisaverygeneralandrather Designmethodologyvagueterm.Withregardtotechnicalartefacts,atleasttwosignicantkindsofcontextofhumanactioncanbedistinguished,namelythedesigncontextandtheusercontext(seeFigure4andnote12).Inthesetwocontextsthetechnicalartefactmanifestsitselfindifferentways.Inthedesigncontext,themainemphasisliesonhowtoconstructaphysicalsystem(object)thatrealisesacertainfunction.Thisfunctionisoftendescribedintermsofalistofspecicationswhichtheobjecttobedesignedmustmeet.HereweencounterwhatSimoncallstheofatechnicalartefact.Inthecontextofuse,thepresentsitself.There,thefunctionoftheartefactinrelationtotherealisationofgoals(ends)isofprimeimportanceandthephysicalconstitutionofthetechnicalartefactbecomesofsecondaryimportance.Note,however,thatinthecontextofdesignaswellasinthecontextofuse,thetechnicalartefacthasadualnature:withinthedesigncontextitisnotonlyaphysicalstructure,justasitisnotonlyafunctioninthecontextofuse.Inmanycases,thereisnocontinuitybetweenacontextofdesignandacontextofuseinthesensethatthesamepeoplewhodesignatechnicalartefactalsouseit.Thissituationcreatesproblemswithregardtothecom-municationoffunctionsbetweendesignersandusers.Towhatextentisitpossibletodesignatechnicalartefactsothatitwillcommunicateitsfunction,i.e.thefunctionitwasdesignedfor,toitspotentialusers?Orisitthecasethatthetechnicalartefactitselfplaysnointermediaryroleatallinthecommunicationofitsfunction,whichmeansthatthiscommunicationhastobeestablishedbyothermeans?ItisinterestingtonotethatDiperthasworkedoutatheoryoftechnicalartefactsinwhichit Figure4Technicalartefactsandtheircontextsofdesignandofuse Dipert,RRArtifacts,artworksandagencyTemple,Phil-adelphia(1993)Kroes,PAexplanations:therelationbetweenstructureandfunctionoftechnologicalobjectsVol3No3(1998)http://scholarRosenman,MAandGero,Purposeandfunctionindesign:fromthesocio-culturaltothetechno-physicalVol19(1998)Kroes,PAReßectionsontechnologicaldesignasartÕgratedDesignandProcessVolIDPT-3(1998)298DesignStudiesVol23No.3May2002isadeningfeatureofartefactsthattheyareexplicitlydesignedtocom-municatetheirartefactualityandfunctionalitytotheirusers.Thisviewpresupposesthatinprincipleitispossiblethatanartefactitselfcatesitsfunction.However,sincethefunctionisnotanintrinsicpropertyoftheartefact,itisnotclearwhatitisintheartefactitselfthatisthesourceforthecommunicationofitsfunction.Partofthisproblemmaybesolvedbytakingintoconsiderationtheusermanualofatechnicalartefact.Ausermanualhasatleasttwofunctions:itisameanstocommunicatetheintendedfunctiontotheuserandtomakesthisfunctionaccessibletotheuserbyprescribingwhichactionshavetobeperformedtorealisetheintendedfunction.Ifweassumethatausermanualisanintegralpartofthetechnicalartefact,thenpartofthecommunicationproblemcanbesolvedeasilybywayoftheusermanual(seenote13).Buteveninthatcasethequestionastohowmuchofthefunctionofatechnicalartefactcanbecommunicatedwithoutrecoursetoausermanualisofgreatimpor-tancefordesignpractice.Tostudythismatter,designmethodologywillhavetofocusonthenatureofatechnicalartefact,moreinparticularonthenotionoftheintendedorproperfunctionofanartefact,onthevariouswaystechnicalartefactsmaycommunicatetheirfunction,andwhatkindoftheoryofcommunicationthispresupposes.AtthispointIconcludetheexpositionofthedualnatureoftechnicalartefacts.Inthenextsection,Iwillexploresomeoftheconsequencesofthisinterpretationofthenatureoftechnicalartefactsfortheagendasettingofdesignmethodology,particularlyfromthepointofviewoftherelation-shipbetweenthedesignprocessandthedesignproduct.Discussion:consequencesfortheresearchagendaofdesignmethodologyOuranalysisoftechnicalartefactsashavingadualnatureofitselfleadstoaquestionthatisofcrucialimportanceforunderstandingthenatureofdesignprocessesandthereforedeservesaprominentplaceontheresearchagendaofdesignmethodology.Thisquestionis:Howcanweaccountforthefactthatdesignersareabletobridgethegapbetweenafunctionalandastructuraldescriptionofatechnicalartefact?Thattheyareabletobridgethisgapstandswithoutquestion.Butfromaphilosophicalpointofview,wearedealingherewithtwodifferentconceptualisationsofanartefact.Itisnotclearhowthesetwoarerelatedtoeachotherandhowitispossibletogofromoneconceptualisationtotheother.Schematically(seeFigure5),adesignprocessmaybecharacterisedasstartingwithafunctionaldescriptionofthedesiredartefact;thismaybeconsideredtobetheinputofadesignprocess.Thisfunctionaldescriptionisablackboxdescriptionwithregardtothephysicalstructureofthetechnicalartefact.Itisprecisely DesignmethodologyFigure5Thedesignpro-cessandthegapbetweenfunctionalandstructural thetaskofthedesignertollthisblackboxwithaphysicalstructuresuchthatthisstructurewillrealisetheintendedfunction.Theoutputofadesignprocess,therefore,isadescriptionofaphysicalstructurewhichadequatelyperformsthefunction,thatis,withadesignofthetechnicalartefact(whichmaybetakentoincludetheusermanual).Giventhisinterpretationofthedesignprocess,twoobservationsmaybemade.First,designersmanagetobridgethegapbetweenfunctionalandstructuraldescriptionsofartefactsinasystematicway;theyuseallkindsofdesignmethodstohelpthemsolvetheirdesignproblems.Second,theyareinmostcasesabletoexplainwhyaproposeddesignwilladequatelylitsfunction.Fromthepointofviewofthedualnatureoftechnicalartefacts,theseobservationsraisethefollowingquestions:Whatkindofdesignmethodsareusedbydesignerstobridgethegapbetweenthetwomodesofdescribingtechnicalartefacts?Howarewetointerprettheroleofthesedesignmethodsinbridgingthegapbetweenthetwoconceptualisationsoftheartefact.Inotherwords,canweprovidearationalaccountoftheuseofthesedesignmethods,showingwhytheiruseissuccessful,giventheconceptualgap?HowdodesignersexplainthefunctionofanartefactintermsofitsHowcanafunctionbeexplainedintermsofaphysicalstructure,giventheconceptualgapbetweenthetwokindsofdescriptionsinvolved(seenote14)?Inordertoanswerthesequestionsmuchempiricalandconceptualworkstillremainstobedone.Butgiventheaimofdesignmethodologytoimprovedesignpractice,itcannotavoidaddressingthesequestions:with-outclaricationoftheseissuesanadequateunderstandingofthenatureofthedesignoftechnicalartefactsis,tosaytheleast,problematic(seenote15).naltopicthatIwouldliketodrawattentiontoconcernsthequalityofadesign,inparticularthenotionofasuccessfuldesign.Itisself-evidentthatdesignmethodologyhastoestablishsomecriteriaforthequality,thesuccessandthefailureofdesignprocessesifwearetotakeitsnormative 300DesignStudiesVol23No.3May2002stancetowardsdesignprocessesseriously.Otherwise,thenotionofanimprovementofadesignprocesslosesitsmeaning.Thesecriteriaarealsonecessarytoupholdtheideathatdesigningtechnicalartefactsispartlyarationalactivity(seenote16).Withoutsomecriteriaforimprovementorprogress,thenotionofrationalitybecomesproblematic.So,whatarethecriteriaforqualityonthebasisofwhichdesignprocessesmaybeevaluated?Inlinewiththeirprocessorientation,designmethodolo-gistsseemtohaveapproachedthisproblemprimarilyfromthepointofviewoftheorganisationandmanagementofdesignprocesses.Therearemanyprescriptivephasediagramsofhowtosplituptheoveralldesignprocessintovariousparts.Thesuggestionisimplied,explicitlyorimplicitly,thatfollowingthesediagramswillleadtooratleastcontributetothequality(success)ofthedesignprocess.Thus,implementingadequatelytheprescriptivephasediagrambecomesacriterionforsuccess.Withoutanassumptionofthiskind,therationalebehindthesediagramsbecomesproblematic.Thismaybepartoftheanswer,butitishighlyquestionablewhetheritaddressestherealissuesinvolved.Itisnotdiftoimagine,andprobablyhasoftenactuallybeenthecase,thatadesignprocessfollowspainstakinglyalltherequiredproceduresandneverthelessitsoutcomeisdeemedafailurebythepeopleinvolved.Insuchcases,thedesignprocesshastobeconsideredasuccess,whereasitsoutcomeisafailure(theproverbialsuccessfuloperationwithadeadpatient).Con-versely,abadlyorganisedandpoorlymanageddesignprocessmayleadtoanexcellentdesign.Therelationshipbetweenthequalityofadesignprocessandthequalityofitsoutcomedoesnotseemtobestraightforward.Abidingbytherulesofproceduralrationalityisnotasufcientcriterionforsuccess(neitherdoesitseemtobeanecessarycriterion).Moreisinvolved,namelythecriteriaintermsofwhichtheoutcomeofadesignprocessisevaluated.So,wearriveatthequestion:Whatisagoodorsuccessfuldesign?Thatitselfisacomplicatedissueanditisdoubtfulthatthereisonesetofcriteriathatisuniversallyvalidineverycontext.Inouranalysisoftechnicalarte-factswehavedistinguishedsofartwodifferentcontextsofhumanaction,namelythedesigncontextandtheusecontext.Itisnotatallself-evidentthatthesamecriteriaforqualityapplyinbothcontexts.Withinadesigncontext,ageneralcriterionforsuccessmaybethataproposeddesignmeetsallthespecicationsandconstraints.Aparticulardesignthatsatisesthiscriterionmayneverthelessbeconsideredafailureinthecontextofusebecauseitdoesnotmeettheexpectationsorsatisfytheneedsoftheusers;thelatterwillbetheircriterionofsuccess.Thissituationmaybedueto,forinstance,poorcommunicationbetweendesignersandusersaboutthe Designmethodologydesiredfunctionality.Butevenifweassumethatthecommunicationaboutneedsandfunctionsisawless,thenitisneverthelessdoubtfulwhetherthecommunityofdesignersappliesthesamecriteriaofqualityasthecommunityofusers.Forinstance,theintroductionofan,i.e.achangeinthephysicalmake-upofanartefactthatdoesnotaffectinanywayitstechnicalfunction,mayconsiderablyimprovethequalityofadesignjudgedinthecontextofdesign,whereasinthecontextofuseitsqualityremainsthesame.Apartfromthecontextofdesignandthecontextofuse,technicalartefactsgureinmanyothercontextsofhumanaction,suchasthecontextofproduction,contextofmaintenance,contextofconsumermarkets,etc.Eachofthesecontextshasitsowncriteriaforqualityandsuccesswhichmayberelevanttothewaythequalityofthedesignoftheartefactisevaluated.Aestheticcriteriaposeaproblemoftheirowninevaluatingthequalityofadesignbecauseitisaproblemtondobjectivestandardsforthesecriteria(seenote17).Moreover,theimportanceofthesecriteriavar-iesstronglyoverdifferentengineeringdomains(forinstance,inmanyareasofelectricalandmechanicalengineeringtheyarealmostirrelevant,whereasinarchitecturetheyareimportant).Thereisbutoneconclusiontobedrawnfromtheforegoing,namelythataclearinsightintothenotionofthequality(success)ofadesignoradesignprocessislacking.Giventhisconclusion,itisratherremarkablethat,althoughdesignmethod-ologyprofessestoaimatimprovingdesignprocesses,ithas,tomyknowl-edge,notaddressedtheseissuessystematically.Eitherthesuccessofadesignprocessdependswhollyorinpartonthesuccessoftheoutcomeofthisprocess,inwhichcaseitappearsratherobviousthattheaboveissuesaboutthequalityofadesignshouldrankhighontheresearchagendaofdesignmethodology.Orthesuccessofadesignprocessdoesnotdependatallonthesuccessofitsoutcome,butinthatcasetherationaleforimprovingthedesignprocess,thatis,fortheaimofdesignmethodology,becomesproblematic.Designmethodologyneedsafoundationforitsnor-mativepointofviewondesignprocesses,anditappearsplausiblethatthisfoundationispartlytobefoundincriteriaforasuccessfuldesign.IwouldliketothankthemembersoftheDepartmentofPhilosophyoftheDelftUniversityofTechnologyfortheirvaluablecommentsonanearlierversionofthispaper. 302DesignStudiesVol23No.3May2002QuotedinDorst(p.8).Thereseemstobesomeconfusionaboutwhatdesignmethodologyisaboutandhowitrelatestothewidereldofdesignstudies.Iwillnotgointothismatter;formypuposes,Crossdescriptionisagoodstartingpoint.AnotableexceptionisGalleFormoredetailsabouthowdesignmethodologyinthissensetsintothebroadereldknownasthephilosophyofengineeringdesign,seeKroesThisclaimdoesnotgoundisputed;seeforinstance,Simon,whoremarks:Theintellectualactivitythatproducesmaterialartifactsisnodifferentfundamentallyfromtheonethatprescribesremediesforasickpatientortheonethatdevisesanewsalesplanforacompanyorasocialwelfarepolicyforastate(p.111).Inthefollowinganalysis,Iwillconsidertheoutcomeofadesignprocesstobeatechnicalartefact.Althoughadesignisnotyetitselfsomethingthatjustiablymaybecalleda(full-blooded)technicalartefact,itisanintegralpartofaprocessthatproducesatechnicalartefact.Moreover,theultimatevalidationofadesigninvolvestheactualmakinganduseofthetechnicalartefactdescribedinthedesign.Thus,ifthedesignvalidationphaseistakentobepartofthedesignprocess,thisprocessimplicitlyimpliesthemakingoftheintendedtechnicalartefact.Thearrowsstandforconceptualimplication:thenotionofanartefactconceptuallyimpliesthenotionofacharacter,agoalorpupose,andanenvironment.Formoredetailsabouttheblackboxcharacteroffunctionalandphysicaldescriptionsofobjects,seePartsofthefollowingarebasedontheNWOgrantapplicationThedualnatureoftechnicalartifacts1999,writtenjointlybyAnthonieMeijers,MaartenFranssen,PieterVermaas,WyboHoukesandtheauthor.Forthefulltextofthisapplication,seehttp://www.dualnature.tudelft.nl.Ofcourse,allkindsofdemarcationproblemsariseaboutsoftwareornaturalobjectsusedforpracticalpurposes.Iwillleavethoseproblemsaside.Thischaracterisationseemstobeadequatefortechnicalarte-factswhicharetheresultofengineeringdesignanddevelopment.Thisisrelatedtothefactthatafunctionaldescriptionis,fromaphysicalpointofview,ablackboxdescriptionofanobject;ingeneral,afunctionaldescriptionstatesthatsomething,whateveritmaybefromaphysicalpointofview,maybeusedasameanstorealiseacertainstateofaffairs.Inamoreorlesssimilarway,Losonskyanalysesthenatureofartifactsintermsofthefollowingthreefeatures:internalstructure,purposeandmannerofuseForanactiontheoreticalaccountofthedesignandusecontext,seeW.Houkes,P.Vermaas,K.DorstandM.deVries,Designanduseasplans:anactiontheoreticaccount,inthisissueofDesignStudiesPartofthecommunicationproblem,becausethereisnoguaranteethatauserwillreconstructfromthetechnicalartefact(includingitsusermanual)afunctionthatisidenticaltotheintendedfunction.Anargumentforincludingtheusermanualinthetechnicalartefactisthatitstrengthensthetiesbetweenatechnicalartefactandacontextofaction,sincetheusermanualprescribeshowtheartefactinquestionhastobeusedinordertorealiseitsintendedfunction.Foradiscussionofthisissue,seeKroesThefactthatdesignershavetodealwithtwoconceptualisationsoftheworldhasnotgoneunnoticedindesignmethodology.RosenmanandGero,forinstance,explicitlycharacterisedesignasinvolvingthetransitionofconceptsfromthesocio-culturalenvironmenttothedescriptionoftechnicalobjects(p.161).Thenotionofrationalitymaybeinterpretedinthiscontextinvariousways.Itmaybetakeninthesensethatitispossibletoprovideargumentswhycertaindesigndecisionswillleadtobetterresultsthanotherones.Ifwetakedesigningtobeagoal-orientedactivity,thenthenotionofrationalityasadaptationofmeanstoanendmaybeapplied(meansForadiscussionoftheroleofaestheticcriteriaindesignandhowthisroleaffectsthequestionwhetherdesignisanartorascience,seeKroes