of DEVELOPMENTAL PSYCHOLOGY

1 ThousandOaks sNewDelhi21 AdultCognitiveDevelopment:DynamicsintheD evelopmentalWebKURTFISCHER,ZHENGYANandJE FFREYSTEWARTA dulthoodnormallyspansmorethan60years,sta rtingfromaboutage20, ,complex,anddynamic,perhapsevenmoresotha nininfancyandchildhood,withmanyfactorsac tingtogetherinvariouscontextstoproducesy stematic, ,itcanbcobservedthatadultsfrequentlyshow regressionperformancesandmovedowntolower levelsofcognitiveskillandthenconstructhi gherlevels,insteadofalwaysfollowinga , , ,andreexaminethelimitationsoftraditional cognitiveanalyses(Fischer,1980b;Fischer& Bidell,199$;Valsiner,1991;vanGeert,1994) .Aconstructedweb(likethatbuiltinnatureby a spider)servesasthemeta-metaphorfordevelo pment,andfromthewehweelaboratethreeImpor tanttypesofdynamicpattemsinadultcognitiv edevelopment:dynamicranges,dynamicstrand sandnetworks, ,webegintocapturetherichnessandcoin-Plex ityofadultcognitivedevelopmentandtooffer anewstoryabotitwhat,how.

2 META-METAPHORSOFADULTCOGNITIVEDEVELOPMEN TT hehistoryofscienceshowsthatdiffereirtmet s-metaphorsfunctioningas centralmentalmodelshavehadtremendousimpa ctonscientificthinking(forexample,viewin gtheearthasthecenteroftheuniverse,seeing thespiralasthestructureofDNA,considering thepersonasadigitalcomputer).Like-wise, simplefixedprogression,followingmonotoni echange,withonestepfollowinganotherina ,thedevelop-mentalladder-liketrajectoryh asatleastthreefeatures:(1)developmentsim plyfollowsa singlestraightline;(2)eachstepisfixed,fo llowingthepreviousstepalongtheline;and(3 ) (1983)cognitivedevelopmentalmodel,asitis us~sallyunderstood,isoneofthemostcommonl adder-likemodelsofhumancognitivedevelopm ent(althoughPiagethimselfhada moredynamicview,asinPiaget,1975).

3 Accordingtothismodel,thinkingprogressest hrougha ,addingmorestagesormoreunevennessacrossd omains492 DEVELOPMENTINADULTHOOL)SingleTrendlineof DevelopmentinOneDirection(Alexandereta!. ,1990;Baltes,1987;Basseches,1984;Berg,20 00;Commonsetal.,1998;Dawson,1999;Erikson ,1968;Gardner,1983;Gruber,1981; ;King&Kitehener,1994;Kohlberg,1969;1984; Loevinger,1976;Sinnott,1998).Thesemodels eitherhavesubstantiallyexpandedPiaget smodelalongtheverticaldimensionbyaddingh ighercognitivestagessuchaspost-formalope rationsandadvancedreflectivethinking,orh aveextendedPiagetsmodelalongthehorizonta ldimensionbyincludingmoreeognitiv ,suchasstandardizedabilitytesting,oftenh aveacknowledgedphenomenasimilartoPiageti anstages,buthaveemphasizedcertainupwarda nddownwardgeneraldevelopmentaltrendsasso ciatedwithageonstandardizedtestsofabilit ies(Baltes,1987;Birren,1964;1970;Craik,1 977;Craik&Salthouse,199!

4 ;Horn,1982;Horn&Cartell,1967;Salthouse,1 984;1992;Sternherg,1985).Someabilities,s uchascrystallizedintelligence,increasewe llintooldage,whileothers,suchasfluidinte lligence, ,butallofthem,todifferingdegrees, ,likechildcognitivedevelopment,ashstatic progressiveprocessunfoldingalonga series~fixedladdersteps,eitherthroughsta gesor through~ short,thismeta-metaphordoh~simplifycompl exdevelopmentalphenomenaandsketchgeneral developmentaltrends,butatth~expenseofneg lecting,downplaying, ,developmentalwebsportrayadultcognitived evelopmentasa ,thedevelopmentalwebhasatleastthreeimpor tantfeatures:(I)developmentoccursina complexmulti-levelrange.

5 (2)developmentalpathwaysundergodynamictr ansformationthroughmultiplestrandsornetw orklinks;and(3) (Fischer&Bidell,1998)analyzesdevelopment asinvolvinga ,ittumsout, dynamicworldwithvarioussocial,emotional, technological, descriptionofdevelopmentthataimsatbothri gorandhonesty, webcapturestheinterconnectedcom-plexityo fskillsindiversecontexts, ,sometimesconverg-ingthroughcoordination ,sometimesdivergingthroughseparationordi fferentiation, ,suchastheseparationofpositiveandnegativ eactivities(goodandbad,niceandmean,appro achandavoidance).Thewebmetaphorstressest hatmanycomponentscontributetoallyactivit y, personactsinteractively,engagedwithhisor hermanyenvironments, ,eachpersonconstructsa uniqueweb, compositeofavailablelevels thedevelopmentalrange withreferencetotheexperiencesandcontextu alsupportsthatcontributetoitsconstructio n,Foranysingledomainofaction(singlestran d),a person scompe-tenceisnotfixedata particularpointonthestrandbutcanvaryalon ga domain.

6 DYNAMICSINTHEDEVELOPMENTALWEB493complexa ctivity,andjointparticipationwithothersa llaffectthelevelofa person ,thedevelopmentalwebdiffersfroma developmentalladderinatleastsiximportant ways:IThewebplacesvariationinactivityatc enterstage,whereastheladderdownplaysvari ation,relegatingit tomarginalityaserroror individualdifferences,2 Thewebisbasedonindividualcognitiveperfor mance, ,whereastheladderassumesasinglelevelata ,whereastheladdertreatsdiversetasksanddo mainsintermsofa singleline,SThewebhasinherentlycomplexin terconnec-tionswithinit,whereastheladder doesnotincludenetworkingamongelements,6 Thewebshowsmultipledirectionsofcon-struc tion,suchasforwardconsolidationandbackwa rdtransition,whereastheladderassumesa ,> I,a,a-naE0)

7 A0-cCa,Ea0a,>a,~ powerfulmeta-metaphor,thewebcanfacilitat ebetterunderstandingofwhat,how,andwhyadu lts ,forthesamecognitivetask, ,cognitiveperformanceinadultsvariesmuchm orewidelythaninchildren,Adultscanthinkmo reflexibly,dynamically,andcontextuallyth anchildren,whilelikechildrentheyalsocont inuetomakeerrors,evenridiculousmistakes, andtoactinsimple, ,theupperlimitofcognitivefunctioningcont inuestoincreasebeyondwhatPiagetcalledfor maloperations(Inhelder&Piaget,1958;Piage t,1975;1983).Thus,adultscansolvemuchmore abstractandcomplicatedcognitivetasksthan children, ,representinga continuingincreaseinadults optimalcognitiveskillsanda widerangeofvariationinthelevelofskillsth atadultscanuseinadomain,MultipleLevelsof AdultCognitiveDevelopmentAlongwiththeinc reaseinoverallcomplexityofadults cognitivedevelopment,bothdevelopmentalre searchandeverydayobservationsindicatetha tadultsshowmultiplelevelsof cognitivedevelopment,notperformanceat onefixedlevel,Evenverywiseadultsusesimpl eskillswhenthesituationrequiressimpleact ion, multilevelperformancevarywithcontextuals upport,priorexperience.

8 Upperlimit :peoplehaveanupperlimitona ,becauseevenanadult supperlimitvanesdynamicallywith~j5Ab2->0 ) ,varyingwithcontextualsupport:optimallev elandfunctionallevel,Thereisnosingleleve lof ,intheabsenceoftaskinterventionorscaffol dingbyothers,individualsshowgreatvariati oninskilllevelsintheireverydayfunctionin g(Fischer&Bidell,1998;Fischer,Hand,&Russ ell,1984;vanGeert,2002).Optimallevelsare attainedprimarilyinthoseinfrequentcircum stanceswhenenvironmentalconditionsprovid estrongsupportforcomplexperformance-Such conditions,includingclearlydefinedtasks, familiarmaterials,practice,andmemoryprim ingofthegistoftheactivity,arenotpresenti nmostsituations-Forthisreason,everyperso nshowsa persistentgapbetweenthefunctionallevelun dertypical(low-support)

9 ,gradual,andcontinuousgrowthovertime,whe reasoptimallevelsexhibitstage-likespurts andplateauswithinanupwardtrend,likethose inFigure21, ,becomingmoredisparatewithage, skillina particulardomainovertime,whereastheoptim allevelS-theupperlimitonfunctioning isachievedthroughstrongcontextualsupport fora skillcom-binedwithorganicgrow-lbprocesse sthatreorganizebehaviorandbrainactivityi nrecurringgrowthAgeinyearsPrinciples:lev elAb42428 Abstractsystems:levelAbSAbstractmappings :levelAb2 ,mentotoptanalandfrenctionallevelsinodom ainADULTCOGNITIVEDEVELOPMENTDYN1 MIGS[ ,thegapbetweenfunctionalandoptimallevels gro\ givenskillthaninitsfunctionallevel,andco nse-quentlythegapincreasesfromearlychild hoodthroughadulthood(Bullock&Ziegler,199 4;Fischer,Kenny,&Pipp,1990;Kiteheneret al.)]

10 ,1993:Watson&Fischer,1980). studyofhowKoreanadolescents( adolescentthroughyoungadult)sawthemselve sinrelationshipwithothers,studentspartic ipatedintheSelf-in-RelationshipsIntervie w,whichincludedbothanopen-endedinter-vie wabouttheirrelationships(lowsupport)anda high-supportassessment(Fischer&Kennedy,1 997;Kennedy,1994).Supportwasprovidedthro ughtheircreationofa ,students(a)createddescriptionsoftheirch aracteristicswithparticularpeople;(b)pla cedthedescriptionsinoneofthreeconcentric circlesfrommosttoleastimportant;and(c)gr oupedsimilardescriptions,drewconnectingl inestoindicaterela-tions,andaddeda plusor minustoindicateemotionalvalence(good,bad ,orambivalent).