HANDBOOKof

DEVELOPMENTALPSYCHOLOGY

Edited byJaan Valsiner and Kevin J. Connolly

SAGEPublicationsLondon’ ThousandOaks’sNew Delhi

21Adult CognitiveDevelopment:

Dynamicsin theDevelopmentalWeb

KURT FISCHER, ZHENG YAN and JEFFREY STEWART

Adulthood normally spansmore than 60 years,startingfrom aboutage20, andthecognitivechangesduring thoseyearsarevast.Accumulatedevidenceindicatesthat cognitivedevelopmentin adulthoodisrich, complex,anddynamic,perhapsevenmore sothan in infancy andchildhood, with manyfactorsacting together in various contexts to producesystematic,dynamicvariation.For instance,it canbcobserved that adults frequentlyshow regressionperformancesand movedown to lower levels ofcognitive skill and then construct higher levels,instead of always following a simple forwardprogression. This kind of backward transitionphenomenonin adult cognitiveprocessesshowsaninterestingandimportant cognitive advancement,one thatmay seemfrustratingandcounter-intuitiveto manyintelligent adults.

Backwardtransitionis just thetip ofthe largeice-bergofcomplexcognitivedevelopmentin adulthood.In this chapter,wereframeadult cognitivedevelop-mentdynamically.resynthesizingresearchfindingsto revealthecomplexdynamicsbehindthevariabilityin adult cognitive development,andreexaminethelimitationsoftraditionalcognitiveanalyses(Fischer,1980b; Fischer& Bidell, 199$; Valsiner, 1991;vanGeert, 1994).A constructedweb(like thatbuilt innatureby a spider)servesasthe meta-metaphorfordevelopment,andfrom the weh we elaboratethreeImportanttypesofdynamicpattemsin adult cognitivedevelopment:dynamicranges,dynamicstrandsandnetworks,and dynamic constructions.With theseConcepts,we beginto capturetherichnessandcoin-Plexity of adultcognitivedevelopmentandto offeranewstory abotitwhat, how,andwhy adult cognitivedevelopmenttakesplaceovertime.

LADDERS AND WEBS: META-METAPHORSOF ADULT COGNITIVEDEVELOPMENT

The history of scienceshowsthatdiffereirt mets-metaphorsfunctioningascentralmentalmodelshavehadtremendousimpact on scientific thinking (forexample,viewing the earth as the centerof theuniverse,seeingthespiral as thestructureof DNA,consideringthe personasadigital computer).Like-wise, differentmeta-metaphorsdrive fundamentalviewsof adultcognitivedevelopment.Wecategorizetwo major types of meta-metaphorsfor adultdevelopment— laddersandwebs whichengenderdifferentportraitsof adult cognitivedevelopment.

Developmentalladderscharacterizedevelopmentasasimple fixed progression,following monotoniechange,with onestep following anotherin asingledirection. As shown in Figure 21.1, the develop-mental ladder-like trajectory has at least threefeatures:(1) developmentsimply follows a singlestraight line; (2) eachstep is fixed, following thepreviousstepalongtheline; and(3) forwardprogres-sionalongtheline is thesolefonn of development.

Piagefs(1983)cognitive developmentalmodel,as it is us~sallyunderstood,is one of the mostcommon ladder-likemodels of human cognitivedevelopment(althoughPiagethimself hada moredynamicview, asin Piaget,1975).Accordingto thismodel, thinkingprogressesthroughaseriesofstagesand then stops at the level of formal operationsduringadolescence.Many scholarshavebuilt uponthis Piagetianframeworkby extendingthe modelvertically or horizontally in adulthood, addingmore stagesor more unevennessacrossdomains

492 DEVELOPMENTIN ADULTHOOL)

SingleTrendlineof

Developmentin One

Direction

(Alexanderet a!., 1990; Baltes, 1987; Basseches,1984; Berg,2000; Commonset al., 1998; Dawson,1999; Erikson, 1968;Gardner,1983; Gruber, 1981;Kegan. 1982; King & Kitehener,1994; Kohlberg,1969; 1984;Loevinger, 1976; Sinnott, 1998).Thesemodelseitherhave substantiallyexpandedPiaget’smodelalongtheverticaldimensionby addinghighercognitivestagessuchaspost-formaloperationsandadvancedreflective thinking, or have extendedPiagetsmodel alongthe horizontaldimensionbyincluding more eognitivç domainssuchas moralreasoningandself-understanding.

Other models that are groundedprimarily inpsychometricresearch,such as standardizedabilitytesting,oftenhaveacknowledgedphenomenasimilarto Piagetianstages,but have emphasizedcertainupwardanddownwardgeneraldevelopmentaltrendsassociatedwith ageon standardizedtestsof abilities(Baltes,1987;Birren, 1964; 1970;Craik, 1977;Craik& Salthouse,199!; Horn, 1982; Horn & Cartell,1967; Salthouse, 1984; 1992; Sternherg, 1985).Some abilities, such as crystallized intelligence,increasewell into old age,while others,suchasfluidintelligence,begin to decreaseby early or middleadulthood.

Thesevarious developmentalmodelshave sub-stantiallyaddedto knowledgeof cognitivedevelop-mental changesand variations in adults, but allof them, to differing degrees,sharean underlyingladder-likemeta-metaphor.Theytreatadult cognitive

development,like child cognitivedevelopment,ashstaticprogressiveprocessunfolding alongaseries~fixed laddersteps,eitherthroughstagesor through~linearability scales.In short, thismeta-metaphordoh~simplify complex developmentalphenomenaandsketch general developmentaltrends, but at th~expenseof neglecting,downplaying,andevenmis-representingthe variability andrichnessof adultcognitivedevelopment.

In contrast,developmentalwebs portrayadultcognitive developmentas a complex processofdynamic construction within multiple rangesinmultiple directions.As illustratedin Figure21.2, thedevelopmentalweb has at least three importantfeatures:(I) developmentoccursin acomplexmulti-level range;(2) developmentalpathwaysundergodynamictransformationthroughmultiple strandsornetworklinks; and(3) multidireetionalconstructionis the form of development.

Dynamicskill theory (Fischer& Bidell, 1998)analyzesdevelopmentas involving a constructedweb that capturesmuchof the rich variability inhumanbehavior.Central to thevariability, it tumsout, is the fact that activitiestakeplacein specificcontexts. Peopledo not act in a void. Growingadaptivelyin a dynamicworld with varioussocial,emotional,technological,andphysicalchallengesmeansthatbehaviormust fit the immediacyof thesituation.For adescriptionofdevelopmentthat aimsat both rigor andhonesty,thesecontextscannotbeignored. A web capturesthe interconnectedcom-plexity of skills in diversecontexts, asshowninFigure 21.2. Eachweb containsdistinct strandsfordifferentcontextsandactivities,sometimesconverg-ing through coordination, sometimes divergingthroughseparationor differentiation, alwaysbuiltthroughspecificsensorimotorandmentalactivities.Emotional statesalso shapestrands,such as theseparationof positive andnegativeactivities (goodandbad, nice andmean,approachandavoidance).The webmetaphorstressesthatmanycomponentscontributeto ally activity, producingdiverseshapesof development.A personactsinteractively,engagedwith his or hermanyenvironments,andtheactionprocessis dynamic and nonlinear becausetheoutcomeof an actioninvolves more than addingtogether the behavior of the individual and theenvironmentalcomponentsthat contribute to it.Specifically, eachpersonconstructsa uniqueweb,while at the sametime ordering principles helpgeneralizationacrossindividual webs.

Theweb also incorporatesskill variationwithineachstrand.Eachstrandis structuredby acompositeof availablelevels— thedevelopmentalrange— withreferenceto theexperiencesandcontextualsupportsthat contribute to its construction,For any singledomain of action(single strand),aperson’scompe-tenceis not fixed at a particularpoint on the strandbut can vary alonga portion of the strand.Practiceandfamiliarity with adomain,contextualsupportfor

FINISH

FixedSteps

START

Figure 21.1 A developmentalladder

ADULT COGNITJyE DEVETClEMENT:DYNAMICSIN THEDEVELOPMENTALWEB 493

complexactivity, andjoint participationwith othersall affect the level of a person’sactivities alongastrand.Eachsinglestrandshowsthedevelopmentalrangein skill andknowledgeof the individual forthat particular task and domain given varyingamountsofexperienceandcontextualsupport.Laterin the chapterwe will elaboratehow this variabilitycanbeintegratedinto thewebmetaphor.

Conceptually,thedevelopmentalwebdiffers froma developmentalladder in at least six importantways:

I Theweb placesvariation in activity at centerstage,whereastheladderdownplaysvariation,relegatingit to marginalityaserroror individualdifferences,

2 The web is based on individual cognitiveperformance,whereasthe ladder is primarilybasedon averagegroupperformance.

3 The webincludesmultiple cognitive levels ineach person, whereasthe ladder assumesasinglelevel at a time.

4 The web distinguishesmultiple tasks anddomains,whereastheladdertreatsdiversetasksanddomainsin termsof asingleline,

S The webhasinherentlycomplexinterconnec-tions within it, whereasthe ladder doesnotincludenetworkingamongelements,

6 The web shows multiple directions of con-struction, suchas forward consolidationandbackwardtransition,whereastheladderassumesasingledirectionofforwardprogression.

multiplestrands

DYNAMIC RANGES IN TFIE WEB

0a,>‘I,a,a-naE0)a0-c

Ca,Ea0a,>a,~0

Rethinking adult cognitive developmentrequiresestablishingnewmeta-metaphorsto replaceoldmeta-metaphors.Developmentalwebs cancapturemoreof the richnessand complexity of adult cognitivedevelopmentthan ladders. As a powerful meta-metaphor,thewebcanfacilitatebetterunderstandingof what, how, andwhy adults’cognitionchangesincomplexsituationsovertheextremelylongperiodoflife afterchildhood.

Researchshowsthat the complexitylevelsof adultcognitioncontinueto changein two importantways.First, for the samecognitive task, an adult oftenshowsmultiple levelsof cognitionunderdifferentcircumstances.Becauseof the wide rangeof levelsof which adultsarecapable,cognitive performancein adultsvariesmuchmorewidely thanin children,Adults can think more flexibly, dynamically,andcontextually than children, while like childrenthey also continueto makeerrors,evenridiculousmistakes,and to act in simple, primitive ways.Second,the upperlimit of cognitive functioningcontinuesto increasebeyond what Piaget calledformal operations(Inhelder& Piaget,1958; Piaget,1975; 1983). Thus, adults can solvemuch moreabstract and complicated cognitive tasks thanchildren, even while they also can use low-level

Figure21.2 A developmentalweb

494 DEVELOPMENTIN ADULTIJOOD

skills similar to thoseof children. The lengthsofsotnestrandsin the web continueto expandintodevelopment,representinga continuingincreaseinadults’ optimal cognitiveskills andawide rangeofvariationin thelevel of skills thatadultscanusein adomain,

Multiple Levelsof Adult CognitiveDevelopment

Along with the increasein overall complexity ofadults’ cognitive development,both developmentalresearchand everydayobservationsindicatethatadultsshow multiple levelsof cognitivedevelopment,not performanceat onefixed level, Evenverywiseadults usesimple skills when thesituation requiressimpleaction,andfrom timeto time theymaymakeunwise decisionswhendealingwith complextaskswithout sufficient contextualsupportto them. Thedynamicsofadults’multilevelperformancevarywithcontextualsupport,prior experience,andjoint actionwith otherpeople.

OptimalandFunctionalLevels

A central concept in traditional developmentalresearchis thatof ‘upperlimit’: peoplehaveanupperlimit on a given skill beyondwhich they cannotgo. This conceptrequiresmajor revision, becauseevenan adult’supperlimit vanesdynamically with

~j5Ab2 -

>0)

Sm

Abi

AM -

Aba -

Epa

Rp2 -

8 12

contextualsupport.Developmentalresearchdifferen-tiatestwo majortypesof upperlimit on skill perfor-mance,varyingwith contextualsupport:optimallevelandfunctionallevel,Thereis no singlelevel of com-petencein anydomain.instead,in theabsenceoftaskintervention or scaffolding by others, individualsshowgreatvariationin skill levelsin their everydayfunctioning (Fischer& Bidell, 1998;Fischer,Hand,& Russell,1984;vanGeert,2002).Optimal levelsareattainedprimarily in thoseinfrequentcircumstanceswhen environmental conditions provide strongsupport for complexperformance-Suchconditions,including clearlydefinedtasks,familiar materials,practice,and memorypriming of the gist of theactivity, arenot presentin mostsituations-For thisreason,everypersonshowsapersistentgapbetweenthe functional level under typical (low-support)conditionsand theoptimal level affordedby highsupport.

Functional levels tend to be characterizedbyslow, gradual, and continuousgrowth over time,whereasoptimal levels exhibit stage-like spurtsandplateauswithin an upwardtrend,like those inFigure21,3.Thesetwo trendlinesdiverge,becomingmore disparatewith age,becausethey dependondifferent sets of growth processes.The functionallevel resultsfrom thesteadyconstructionofaskill inaparticulardomain overtime, whereasthe optimallevel S-theupperlimit on functioning— is achievedthroughstrongcontextualsupport for a skill com-binedwith organicgrow-lb processesthat reorganizebehavior and brain activity in recurring growth

Age in years

Principles:level Ab4

24 28

Abstractsystems:level AbS

Abstractmappings:level Ab2

singleabstractions:level Abi

-- -- -

- -

F F

16 20

Figure21.3 Depeloj,mentotoptanal andfrenctionallevelsin o domain

ADULT COGNITIVEDEVELOPMENTDYN1MIGS[N TIlE DEVELOPMENTALWEB 495

cycles.Furthermore,thegapbetweenfunctionalandoptimal levelsgro\vswith age.Researchhasfoundafar larger increasewith age in the optimal level fora given skill than in its functional level, andconse-quently the gap increasesfrom early childhoodthroughadulthood(Bullock & Ziegler, 1994;Fischer,Kenny, & Pipp, 1990;Kiteheneretal., 1993:Watson& Fischer,1980).

An exampleof optimal andfunctional levels inabstractionsis thedevelopmentofconceptsofselfinrelationships.In astudy of how Koreanadolescents(grades8 through 13. oradolescentthroughyoungadult) sawthemselvesin relationshipwith others,studentsparticipatedin the Self-in-RelationshipsInterview,which includedbothanopen-endedinter-view abouttheir relationships(low support)andahigh-supportassessment(Fischer& Kennedy,1997;Kennedy, 1994). Support was provided throughtheir creationof adetaileddiagramof the charac-teristicsof specificrelationships.In thehigh-supportassessment,students(a)createddescriptionsoftheircharacteristicswith particularpeople;(b) placedthedescriptionsin oneof threeconcentriccirclesfrommost to least important; and (c) groupedsimilardescriptions,drewconnectinglines to indicaterela-tions,andaddedaplus orminus to indicateemotionalvalence(good,bad, or ambivalent).Then theinter-viewer askedthem a seriesof questionsto elicitexplanationsof their diagramat different develop-mentallevels.In thelow-supportassessmentstudents

Domains

producedonly aslight increaseoverthesix yearsanddidnot achieveeventhelevel of singleabstractions.The same students in the high-supportconditionstartedat a higher level, single abstractions,andmovedup to thelevelofabstractsystems.In addition,their trajectoryshowedspurtsfor theemergenceofabstractmappingsandabstractsystems,similar tothoseshownin Figure213.Muchmoresophisticatedcognitiveskills \verecalledforth with support,whileanabsenceof support ledto low-level skills,

Note that optimal level producesaseriesof spurtsin growth followed by plateausor small drops— adynamicpatternof changethat is commonin devel-opment (Fischer& Bidell, 1998; Fischer,Kenny,& Pipp, 1990); Thatcher,1994; van der Maas&Molenaar, 1992). The fact that the functional levelshowsno suchsystematicvariability underscoresthe potential for missing the telling dynamicsofdevelopmentby examiningperformancein only onecondition andassumingthat it representsthe basicnatureof cognitive development.Growthpatternsdiffer underdifferentconditions,evenfor the‘same’skill in the sameperson, and the dynamics ofthis variability arefundamentalin adult cognitivedevelopment.

How do the spurtsin optimal level relateto thewebof development?Variousstrands/domainsin aweb showa clusterof spurts within a concurrentzone, as illustrated in Figure 21.4, Put anotherway, in the developmentalweb, the optimal level

Mathematics Self in Reflectivejudgmentrelationships

Clustersof diseontinuitiesin emergenceZones

Figure2 1 .4 Clustersol discontinuitiesfor two optimal levelsacrossstrandsanddomains

496 DEVELOPMENTIN ADULTIIOOD

emergeswhen clusters of diseontinuitiesappearacrossmanystrandsin the sametime period. Thisskill phenomenonhasaneurophysiologiealcorrelate,in thatcortical substratesfor the increasein abilityshow developmental changes that mirror thebehavioralones(Fischer& Rose,1994; Thatcher,1994). That is, patternsof cortical activity showspurtsthat areapproximatelyconcurrentwith thespurtsin optimal skill level.

Automatizationand Co-Participation

Optimal andfunctional levelsareonly two of themany skill levels that adults routinely use. Forexample,when peopleact automatically(withoutthinkingor consciouslychoosing),theytypicallyactata low level, as whensomeonestepson thebrakeautomaticallywhenachild runs in front of thecar.Researchershave not directly assessedthedevelopmentallevel of suchautomaticactions,hutthey exist in everydomain, andusually they arerelatively simpleandprimitive.

Ontheotherhand,peoplefrequentlyact togetherwith otherpeople,cooperatingto accomplishatusktogether-. telling a family story, puttingtogetherajigsaw puzzle,playing poker,or building a house.Onepersonscaffoldstheactionsof another,some-times in expertand noviceroles as with teacherand student (Wood, Bruner, & Ross, 1976) andsometimesas more equalcollaborators(Granott,I 993b;Valsiner,1996).In actuality,manysituationsthat psychologistsoften treat as individual arenaturallysocial.Manychildrenpreferto play videogameswith their friends,eitherdirectlysharingthemor talking aboutthem on the phone.Manyscholarswrite paperswith the help of other people,evenwhenonly oneauthoris listed.In co-participationingeneral,peopleen-constructcomplexskills thatoftengo beyond their individual capacity,as Vygotsky(1978)emphasizedwith his conceptof the zoneofproximaldevelopment,andWood,Bmn’er, andRoss(1976)elaboratedwith the conceptof scaffolding.Indeed, the importanceof suchsocialconstructionhasbeenrecognizedfor theentirehistoryofmodernpsychologyandchild development,but it continuesto beneglectedin mostresearchandtheory(Valsiner& vanderVeer, 1988),which is especiallypuzzlingin elaborationsof explicit theoriesof social con-structionsuchasErikson’s(1963). Co-constructiveprocessesareat leastas importantin adultsastheyarein children,

in addition,peoplemoveup anddown in thelevelof their performance,adaptingto thesituation,goal,task,emotionalstate,andtheir co-participants.Real-time analysisof ongoingactivity showshow levelvariesdynamicallywith thesefactors,evenmoreinadults than in children (Bullock & Ziegler, 1994;Fischer & Granolt, 1995; Granott, l993a: 2002;Kuhn et al., 1995; Roberts, 198l; Siegler, 2002;Vaillant. 1977). Asastrandin aperson’swebgrows

— Automaiized

— runctiona]Optima]

- scaffo]decj

Figure 21.5 Developmentalrangein aweb

longer, he or shehasawider rangeof skills to useacrossportionsofthestrand.Figure21.5showshowthefour levelsthatwe havedescribedareevident intheweb. Automatizedskills, markedby thick solidlines,mostlyoccurearly in eachstrand.Functionalskills, perfonnedthoughtfullybut without support,aremarkedby thin solid lines.Optimalskills, whichusuallydependon contextualsupport,occupylaterportionsof the strandandare markedby dashedlines. Seaffoldedskills, in which peoplejointlyperfonnacomplexactivity, aremostcomplexandaremarkedby dottedlines.

Levelsof Optimal CognitiveDevelopment

Adult developmentmustbe understoodin termsofthewhole scopeof developmentfrom infancy,bothbecauselater skills arebuilt on earlierones andbecauseadults routinelyuse skill levels that firstemergein infancy andchildhood (especiallywhenthey move down in a strandof the web to useautomatizedskills, or makebackwardtransitionstobuildnewskills). Dynamicskill theorydescribesthecontext-basedconstructiveprocessofbuilding fromreflexesto actions,from actionsto representations,andfrom representationsto abstractions(Fischer,1980b; Fischer& Bidell, 1998). Cognitive activityundergoesmassiverestructuringduringtheyearsofinfancy andchildhood, graduallybuilding towardconcreteskills andconceptualcategories.In adoles-cenceandearly adulthood,peoplerestructuretheiractivities again, moving from representationstoabstractions,Muchoftherestofadulthoodinvolvesconsolidation,elaboration,integration,synthesis,andextensionof theseabstractskills,

The skill hierarchynot only describescognitivedevelopment,hut providesamler for assessingandstudyingdynamicvariationsin adult activities,Thisrulerallows comparisonof levelsacrossconditionsandtasks,suchasoptimal,functional,andscaffoldedlevels(Figures21.2and21.3),andit makespossibleanalysisof the dynamicsof real-timelearningandproblem-solving,as in backwardstransitionsand

‘F’

F’

— F FF F FF F FF F F F

ADULI’ COGNITIVEDEVELOPMENT:DTNAMICE IN THE DEVELOPMENTALWEB 497

for\vard consolidation,Dynamic analysisof skillrequiressucha scaleto assessvariability and tomodel it, Cognitive developmentresearchhasbeenhamperedby the absenceof suchscalesfor codingactivity acrosstasks,domains,andtrials, exceptinthe arenaof motoractivity; whereCartesiancoor-dinates provide ready-madescales for dynamicanalysis(Rose& Fischer, 1998; Thelen& Smith,1994; vanGeert, 1994).

HierarchyofAdult SkillLevels

Frombirth to 30yearsofage,an individualdevelopsskills through four sequential tiers in a nestedhierarchy.Early reflexesbecomecoordinatedintoactions,actionsarecoordinatedinto representations,andrepresentationsinto abstractions.Eachof thesequalitativelydifferent behavioralrepertoirescyclesthroughasimilar patternofeoordinations— thefourlevelsof eachtier. Movementis from an initial singleexpressionof anability (the first level ofagiventier),to a mappingof two elements(the secondlevel ofa tier), to a sjsFstem that relatesmultiple elements(the third level),andfinally to asystemof systems

(the final level). Eachlevel arisesfrom thegradualcombinationof two or more skills from the priorlevelin aprocessofcoordinationanddifferentiation,Takentogether,the four tiers producea scaleof 13levelsthat increasein complexityandintegration—

a 13-point interval scalefor assessingthedynamicsof developmentandvariation. Reorganizationsofneuralnetworksseemto helpcatalyzedevelopmentof awide rangeof skills at eachnew level (Fischer& Rose, 1994; 1996).

‘I’he levels that characterizethe final tier movethrough single abstractions,abstract mappings,abstractsystems,andabstractsystemsof systems,or principles.We will describethis developmentofincreasedcomplexity of abstractthinking frommiddle childhoodthroughadulthood,asshownin theleft-handpart of Table21.1, andwe will explicatethelevelsthroughdiscussionsofrefteetivejudgment,moraljudgment,identity development,andDarwin’sconstructionof thetheoryof evolution.

The optimal level of representationalsystems(Rp3)usuallyemergesaroundthe ageof 6 yearsinmiddle-classchildrenwith high contextualsupport,and is elaboratedandconsolidatedoverthe next3

Table 21.1 Levelsof’developmentofrepresentationalandabstractskills

Level

Tier

AgeFRepresentations Abstractions

Rpl [Q] IS —24 monthssinglerepresentations

Rp2 [Q— R] 3,5 —4.5 yearsrepresentationalmappings

Rp3: [Q~e—~R~] 6—7 yearsrepresentationalsystems

Rp4/AbI: systemsof representationalsystems,which are

[Q~/’c—sR~’1I IL~2”~T~j

= [Yl 10—12 years

single abstractions

Ab2:abstractmappings [Y—Z] 14—16 years

Ab3:abstractsystems [Yg<—*Z 19—20 years

Ab4:systemsof abstractsystems,whichareprinciples

[Yg~—~Z%~I I[A~—*D~j -

24—25years

Note: Italic lettersdesignaterepresentationalsets, andoutline leuersabstractsets,Subscriptsandsuperscriptsdesignatedifferentiatedsubsets,Long straightlines andarrowsdesignatearelationbetweensetsorsystems.Bracketsmarka singleskill, Notealsothat structuresfrom lower tierscontinueat higherlevels(representationsin abstractskills, etc.),but thefonnulasareomittedbecausetheybecomeso complex.

Agesaremodal for theemergenceof optimallevelsbasedon researchwith middle-classAmericanandEuropeanchildren,Theymaydiffer acrossculturesor socialgroups.

498 DEVELOPMENTIN ADUL THOOD

or 4 years.(The earlierrepresentationallevelsareshownfor completenessandbecausetheyappearinadultproblem-solving,discussedbelow.)This levelis the core of much adult functioning,becauseformany activities peopleneedonly concreteactionsand representations,not sophisticatedabstractthinking. With a skill at this level, a person cancoordinatetwo or threedifferent aspectsof severalrepresentations.For instance,a child Karaand hermotherJanecanplay teacherandstudent,wherethechild/teacherinteracts simultaneouslyand recip-rocally with themother/student:

Play[STUDENT TEACHERIJANE )KARALMOTHER CHILD

Children especiallyenjoy reversingconventionalrolesto assumemorepowerfulandindependentadultrolesin play; andadultsfrequentlycooperatein thispretense.Themotherandthe child both actwith asimilar representationalsystem,asshownin formula1, but the snother’sskill mayinclude an additionalcomponentfor scaffolding her child’s skill, Thecategoryrelations usually remain fully concrete,evenwhenthe story becomescomplex.

Theoptimal levelofsingleabstractions(Rp4/AbI)

emergesat aboutage 10, when youngstersbeginto understandabstractconceptscommonly usedby adults, At this first level of abstractthinking,the ability to relatedifferent explicit instancesofrepresentationsto an intangible conceptbecomescomthonplace.For a I 2-year-oldgirl, travelingwithfriends to a parade, buying her own lunch, andchoosingher own new clothescan all he relatedinthe conceptof independence.Therepresentationalsystemsfor theparade,the lunch, andthe clothes,which eachhavea structuresimilar to that in formula1, are richly coordinatedin a new skill, which isan achievement of representationalcomplexity(Rp4). A diagram for the coordination of twosystemsto form an abstractionis shownin Table21.1.Thecoordinationofparade,lunch,and clothessystemsgives them the power to broadly unifythe threecontentsinto a ssngle abstractconcept,independent:

[INDEPENDENT

At this initial level, abstractionsaresomewhatfuzzybecause they are single: without comparison,abstractionscamsotbeeasilydifferentiatedfrom eachother, The 12-year-old may use the samethreeexamplesfor both independence(formula 2) andindividualist:

When askedhow the two differ, shensuddiesthemtogether,not clearlyarticulatingadifference: ‘They’rethe samething. Both of thesninvolve being free,’Imaginehow complexandconfusingit is whena thirdconceptsuchasliberty is addedto thepot! Thesamekindsof confusioncreatedifficulties in coordinatingone’s own identity with another person’s, oftenleadingto a kind of mergingorglobbing ofidentitieswith a closeftiend orpartner(Erikson. 1968).Adultsaswell asadolescentsshowthis globbingtogetherofdistinctabstractions,andit takesmanydifferentforms(Fischer,Iland, & Russell,1984). At leastadultsarecapableof buildinghigher-levelskills to compareanddifferentiaterelatedabstractions.

The optimal level of abstractmappings (Ab2)(I) appearswhenadolescentsarefirst able to coordinate

two ormoreabstractions,beginningataboutage15.Much sophisticatedadult activity involves this levelof simple relations of abstractions(Colby et al.1983; Commonsetal,, 1998; Cook-Greuter,1999;Dawson,1999; King & Kitchener,1994).Being ableto useoneabstractionin comparisonwith anotherisa greathelpin making thinking moreprecise.Inde-pendenceand individualism arerelatedbut distinct,with independenceinvolving the freedom to dothingson one’sown, andindividualism involving acommitmentto freelychoosingwhoonewantsto be:

1[INDEPENDENT flHFFIV1DtJALIgTj

In tandemwith the increasein cognitiveclarity atthislevel is a jump in social facility becauseof thecapacityto coordinateand differentiateone’s ownabstractionswith someoneelse’s.In identity, apersoncancoordinateanabstractionaboutthemselveswithoneaboutaclose friend or partner,allowing for anew kind of intimate relationship, such as howmy independenceis similar to anddifferent from myfriend’s,especiallyin ourcloserelationship:

[SELF.~~FT:JENID[INDEPENDENT INDFEPENDFENT

(Erikson, 1968; Fischer& .&youb, 1996; Kegan,1982; Loevinger~1976; Noametal,, 1990).

At aroundthe ageof 19 or 20, theoptimal level ofabstract systems (Ab3) emerges,as individuals

2) coordinatemultiple abstractionsandbeginto under-standthesubtletiesandnuancesin abstractrelationsin manydomains, including understandingof selfand others. For instance, the young adult cancompareandrelatethesubtletiesofabstractionslikeconformity and independence.At the prior level,relating different formsof conformity andindepen-dence in different situations is difficult, but withahslraetsystemsit is easierto see:for example,thatboth atschooland with friends I show mixtures ofboth conformity and independence.Similarly withidentity coordination, a mother and father can

I

SELF

ADULT COGNITIVEDEVELOPMENT:D)N4MICSTV THEDEVELOPMENTALWEB 499

uslderstandhow their two identitiesdiffer with theirson anddaughter,anda personcan more readilycoordinatehis or herown conformityandindepen-dencewith aft’iendsor partner’s:

[:T~~~:rFFNirPLMING rDNFFnFLJ.AFfiF4n

IEELF ( >F]F/ifEND[tND’EPENDINT INDEPENDENT

The optimal level of abstractsystemsof systems(Ab4), orprinciples,is thefinal developmentallevelpredictedby skill theory. Emerging underhigh-supportconditionsaroundthe mid-20s,this highestcognitivelevel allowsapersonto coordinateseveralabstractsystemstogether,asdiagrammedin Table21.I. 1-low doesmy own personalidentity relatetomoral dilemmasthatI havefaced,or careerchoicesI havemade,or differentintimaterelationships!havehad(Erikson, 1963; 1968)?By coordinatingtwo orsnoreabstractsystems,a personcanconstructanduseageneralprinciplethatgoesacrosssystems,suchas the Golden Rule in morality and ReflectiveJudgmentin knowledgedilesumas.We will describein some detail how Darwin built his level Ab4principleof evolutionby naturalselection:

[EvaLUTJFDI~, -

SELECTION

Onceconstructed,suchaprinciple canbe extendedto many different abstractsystems,as we willillustrate later.Peopledo notremainat this levelforlong periods,butonly useit asneeded,svithenviron-mentalandsocialsupportsrequiredto sustainit intheday-to-dayactivitiesof living.

Theseskill levelsprovideacomplexityscalewithwhichto assessthe variabilityin people’sactivitiesandto look for patternsof stabilityandorder.Peopledo notact stablyat oneskill level, asin the laddermetaphorfor development.Insteadtheyrangewidelyovermanylevels,sometimeschangingalmostinstan-taneouslyin adaptingto different challenges.Therangeextendsfromlow levelsofactionandrepresen-tation (far belowthoseshownin Table21.1)to thehighest level of abstraction(Brown & Reeve,1987;Bullock & Ziegler, 1994; Fischer& Bidell, 1998;Fiseher&Granott.1995;Granott,1993a;1998;Kuhnet al., 1995; van Geert, 2002). Much of what wedescribein this chapteris therulesfor orderin thispervasivevariationin adultcognitive development.

DevelopmentbeyondAbstractions?

Is thereany evidencefrom thesestudiesthatmightpoint to the developmentof levelsandtiersbeyondthe level of principles (Ab4) — perhapsrelatingprinciplesto eachotheror changingskill capacitiesIn someotherway?Soundandsufficient empiricalevidenceis requiredto answerthis question,andweknowof lirtle thathasbeendecisivebeyondthelevel

of principles for newly emergingoptimal levels,Perhapsadultshaveenoughto do simply general-izing andconsolidatingtheabstractionsrequiredofthem. Hosvever,someinterestingwork by FrancineBenes(1984)onmnyelinationofneuronsin thebrainsuggestsapossiblemajorreorganizationat mid-life,Myelin is the insulationaroundneuralaxonsthatgreatlyimprovesthe speedandefficiency ofneuraltransmission,After yearsof only slow changeinmyelin, adults in their 40s and SOs show myelingrowthspurts for neuronsconnectingtheprefrontalcortexto the limbic system.Onespeculationis thatthischangecreatesmorerefinedcontrolofemotionalimpulses,perhapsin relationwith themasteryof thehighestlevelsof abstractionthatmanyadultsachieveby theseages.With thecapacityto sustaincomplexabstractandprincipledthinking without contextualsupport(at leastin areasofexpertise)comesagreateropportunity to bringwisdom to bearon emotionalequilibrium andself-control.Perhapsthis changeisrelevantto Erikson’s(1963)suggestionthat wisdomis the central issue in his final stageof identitydevelopment.

DevelopmentofReflectiveJudgmentandMoral Judgment

The foundationsof knowledgearea fundamentalissue in cognitivescienceandphilosophy,andJohnDewey (1910)describeda model for the develop-ment of understandingthe basesof knowledge.The goal of educationis what he called reflective/udgment,the‘active,persistent,andcarefulconsid-erationofanybeliefor supposedformofknowledgein the light of the groundsthat support it, andthefurtherconclusionsto which it tends’ (1910, p. 6).Key elementsincludetheuseofevidenceandreason-ing, theframeworksfor knowledgeandbelief, andjustificationsfor conclusions.Developingreflectivethinking is oneof theimportanttasksandintellectualchallengesin adult cognitivegrowth.

Thefoundationsofmoralreasoningareevenmoreimportantthanreflectivejudgmentin humansociety,especially for socially responsibleadults. Moralevaluationandjudgmentareoneof the intellectualchallengesthatadults facein a world with multiple,oftenconflictingmoralstandardsanddecisions.Goodmoralreasoningnot only requiresabstractthinking,butalsocomplexvaluejudgmentsandemotions.TheinfluentialworkofLawrenceKohlberg(1969; 1984)on moralreasoningrevealshow peoplemovein theirthinking from an authoritariannotion of moralitythroughagradualrelativizingoftheirjudgments,andthen to an establishedvalue system (a processgenerallyanalogousto that for reflectivejudgment).Indeed,theresearchonreflectivejudgmentwasbaseddirectly on theresearchandmethodsthat Kohlbergdevisedfor moruljudgment.

A rich researchprogramled by KitchenerandKing hasinvestigatedthe developmentof reflective

(6)

(7)

500 DEVELOPMENTINADULTHOOD

judgmentin adults,aswell asadolescents,includingtests of optimal and functional levels (King &Kitehener,1994;Kitcheneretal., 1993),whereastheresearchon moraljudgmenthasnot assessedthesetwo distinct levels. Kitchener andKing start byaskingpeopleaboutdifficult dilemmasandhow theyknow somethingis eithertme or falsefor suchadilemma.Oneof their standarddilemmasconeemschemicaladditives: arethey good things becausewhen addedto food, they preventsomeillnesses;or arethey bad becausethey may causecancer?Dependingon the responseto this dilemma,peoplecanvaryoversevenstagesof understanding,with theoptimal stagesemergingfrom, roughly,2 yearsofageup to 25 yearsandbeyond(Fischer& Pruyne,2002). The stagesmapexactlyonto the skill levelsoutlinedin Table21.1(Kitcheneretal., 1993).

At the first stage, responsesreflect only anabsolutekind of thinking: a fact or conclusioniseither right or wrong. In moraljudgment,the firstseveralstagesreflectasimilar concreteapproachtomorality: anaction is simply goodor bad. By stage4 of reflectivejudgment(themiddle stage)peoplehavemovedto a relativetypeof thinking: the truthof astatementvarieswith theperspective.Whethersomethingis tme dependson yourbias’.This stageinvolves the construction of single abstractions(Abl) for relativeknowledge,and, as is typical forthis level in general,peoplehavedifficulty movingbeyond the confusion of single, uncoordinatedabstractions.A personknowssimplythat knowledgeis a variablething, andeven though anattempt ismadetojustil5za decisionaboutchemicaladditives,thejustification is neithercoordinatedconceptuallywith thedecisionnordifferentiatedfrom it.

Thenatureofrelativismin moraljudgmentremainsaquestionin theresearch.It is unclearwhetherthereis a distinctstageofrelativism. Two candidatesarestage 3, wheremoral judgment is basedon one’ssocialgroupnorms,or anadditionalstagebetween4and 5, where the relativity of moral judgmenttosociety andculture is recognized(Dawson,2002;Fischer,1-land,& Russell,1984).

In stage5 of reflectivejudgment(the level ofabstractmappings,Ab2) peoplebeginto comparearguments,evidence,andviewpoints,recognizingthatsomeargumentsandconclusionsarebetterthanothers,Argumentsandjustificationsare linkedto acertaincontext or viewpoint, andthereis certainlogic relating to the conclusions,but still peopletakeamostly relativisticstance,At stages6 and7,Dewey’s goal of reflective judgment comesintoplay: the tmth of the propositiondependson thespecificargumentsmadeandthesupportingevidencefor the arguments.With sufficient evidenceandargument,aconclusioncanbe firmly reachedthatgoesbeyonda relativisticdependenceof viesvpoint.Stage6 arguments(abstractsystems,Ab3) recognizethat knowledgeis notalwayscertainbut that strong.justified conclusionscanbe madewith sufficient

evidence.Stage7 arguments(principles,Ab4)moveto fully reflectivejudgment,including formulationof a principle that strong, justified conclusionsreston their evidence,andthat differentkinds ofevidencedependon thesituationandviewpoint fromwhichtheywerecollected.

In moral judgment,Kohlberg’s stages5 and6constitutewhathecalledprincipledmoral reasoning,analogousto the principle of reflectivejudgment.Generalprinciplesareheldto applyacrosscultures,alongwith local variations,andareusedto guidejudgmentsof lawfulandmoralactivities.Empiricalevidencesupportstheexistenceof asocial-contractprinciple,which Kohlbergcharacterizedasstage5,in which peoplearguethat valuesestablishedbynormsin order to promotesocialharmonyfor thegoodofeveryonearesubjectto modificationaccord-ing to thewill of the people.The universalethicalprinciplesthat Kohlberghypothesizedfor stage6,such as the Golden Rule (do unto othersas youwouldhavethemdounto you), remaincontroversialbecauseresearchto datehasfoundfew peoplewhoconsistentlyfunction with suchprinciples(Colbyet al,, 1983).We proposethat researchon dypamicvariation in moraljudgment, such as optimalandfi,mctional levels,will resolvethis dispute.

Onestudyof reflectivejudgmentsuggestsakindof orderin the variationbehindthe emergenceofoptimal levels, Whena level or stagefirst emerges,peopledonot quickly generalizethenewskill acrossall tasksbutwork slowly andpainstakinglyto createthe generalskill. For instance,when the level ofabstractsystemsfirst emergedin thestudentsin thisstudyat aboutage20,theyproducedonly about50%oftheir argumentsatstage6 (Figure21.6).Not untilthe next level emergedat aboutage 25 did theyproducenearly 100%of their argumentsat stage6.In general,eachoptimal level producesa spurt inperfom~ance,as shownin the line for the generalreflectivejudgmentscorein Figure 21.6. But thatlevel doesnot becomepowerfully generalizeduntilsomeyearslater, with theemergenceofthenextlevel(or even a further level beyondthe next). For thefunctional level, the processis much longeryet,reaching50% -only in the late 2Os. In general,functional consolidationof optimal skills requiresmanyyearsof adult development.

An important finding about both reflectivejudgmentandmoraljudgmentis thathighereducatsonplays a central role in their development andconsolidation,All the peoplein the study in Figure21.6werestudents,eithergraduatestudents,collegestudents,or college-boundhigh school students.Researchshowsstronglythat educationplaysa moreimportantrolethanagealone inproducingmovementto sophisticatedjudgmentsabout moral issuesandthenatureof knowledge(Colby et al., 1983; Dawson,2002;King & Kitchener,1994;Restetal, 1999).Theemergenceof anew optimal level is not enoughtoproducethe stable development of a sophisticated

ADULTCOGNITIVEDEVELOPMENT:DIWAMIC5’ IN TI-fE DEVELOPMENTALWEB 501

A~ /

Figure21.6 Developmentofrefieetivefudgnient(datafrom Kitcheneret al,, 1993,)

H 100

90

skill, A stimulatingenvironmentmust catalyzethedevelopmentof the highest stagesof moral andreflectivejudgment,andit maybeessentialfor otherdomainsofadult developmentaswell.

DYNAMIC STRANDSAND NETWORKS IN

THE WEB

Adults developnot only deeplybut alsobroadly.Todeal with complexnatural, social, and spiritualworlds, adults apply, extend, and expand theirsophisticatedcognitive skills in a wide variety ofdistinctivetasksanddomainsthat they encounterinboth academicsettingsandtheir everydaylives,includingjob, profession,health, family relation-ships,child rearing,homepurchaseandmaintenance,self-understanding,emotionregulation,moral rea-soning,religion, andpolitics (Baltes& Staudinger,1993; Erikson, 1978; Fischer, Hand, & Russell,1984; Gardner, 1983; Kegan, 1994; Neugarten,1968; Stemberg,1990). In all phasesof adulthood,people need to update their skill repertoire inmultiple domainsconstantlyin orderto adaptthem-selves to change.Adults must develop multiplespecializedcognitiveskills, suchascritical reading,academicwriting, moraljudgment,householdman-agement,businesspractices,emotionalintelligence,andreligiouspractices,to meetchallengestheyface.The dynamicsandcomplexity of strandsin the webprovide amodel of therichnessandcomplexity ofthis breadthin adult cognitive development.

Thecomplexinterconnectionsamongskill com-ponentsanddomainsin aweb remindus of aneuralnetwork, especiallyof themanydendritesthatcanproliferate from a singleneuronwithin anetwork.With complexnetworking amongmultiple skillsin multiple domains,adults manifestphenomenathat occurnot at all or in muchreducedform inchildren,suchas complexmultiple identities,inter-disciplinaryexpertise,creativity,andwisdom.More-over,dynamicnetworksofstrandsconstantlychangeover time andcontext,andproduceemergentandcomplicated cognitive processesand products.Examplesofhow adultsdevelopin multiple domainsinclude strandsof identity in adulthood,Darwin’sconstructionof the theory of evolution, and theplusesandminusesof cognitiveaging.

Strandsof Adult Identity Understanding

Observing, analyzing, and understandingoneselfis one of the most difficult lifelong intellectualchallengesthat eachadult hasto face. Erik Erikson(1963; 1968), in hisclassicworkon identitydevelop-mentoverthelife course,hadtheinsightthatidentityalwaysdevelopsin relationshipwith otherpeople,especiallyin family, friendship,andwork, Eriksondescribeda developmentallyorderedsequenceofcrises thatreachesits pivotalpointattheendof child-hood with theemergenceof identity in adolescence.Identity is aperson’ssense of who sheor he is andwantsto be, a self-constructedorganizationof emo-tions,beliefs, values,goals,andindividual history. It

- x General optimal RJ (Vi)

——±—— Optimal stage6 tasks 0’2)Functionalstage6taskscY2)~

7

6

5

Stage7

a0)(U0Oa00GO

CO0-acc

4

3-14

/

S

4....

‘V

-80

‘70

-60 ~0,

‘-50

-40 fflr-30

-20 a0.

-10

0I I

16 18 20 - 22 24 26 28

Age in years

502 DEVELOPMENTINADULTHOOD

is not fully achievedandfinished in adolescenceorearly adulthoodbut continuesto be wovenacrossmultiple life strandsgraduallyaswe grow older.

Fromthefirst formationof identity astheclimaxof childhood,adultsextendandcoordinatetheiro~vnidentities with otherpeople’sidentitiei acrosscon-texts andtime periods,progressingthroughthreefurther stages,accordingto Erikson, This conceptof identity haspermeatedmodemsociety,so that itis almost a truism today, althougheverydayuseoftheconceptis often superficial.Most of theempiricalwork on identity developmenthas unfortunatelynot focusedon the full scopeof identity develop-mentduring adulthoodbut hasinsteadconsideredprimarily microdevelopment(substages)within theemergenceofidentity in adolescenceandearlyadult-hood (for example,Marcia, 1980; 1994; Matteson,1977; Phinney,1989; Turkle, 1995).This researchalsoneglectsthe importanceof socialcoordinationofone’sowIa identitywith otherpeople’s.As aresult,considerableconfusionhasreignedaboutthedegreeto which the crises in fact form stages,althoughthis researchhas not actually tested the stagesthemselves.Fortunatelya few studies have gonebeyondthe stageof emergenceof identity (versusrole diffusion) to examinethe full set of stagesEriksondescribed,especiallythroughcaseanalysesandclinical material (Erikson, 1969; 1978;Gilligan,1982; Loevinger, 1976; Neugarten,1968; Noametal., 1990;Vaillant, 1977).

We proposean importantdifferentiationof theidentity frameworkthrough cognitive analysisoftheskills involved in identity formationandcoordi-nationwith others.Articulatingidentity developmentthroughthis skill analysisilluminatesthewaysthatmultiple strandsof identity developsystematicallyin aperson’swebandhowpeopleconstructidentityskills hierarchicallyin a way that correlateswithErileson’s stagecrises,The stagesare shapedbybasic human tasks and issuesthat people shareacrosscultures,suchas learning skills for homeandwork, choosinga romantic partner,making aliving, raisingchildren,andgrowingold. Individualcircumstancesdiffer widely acrosscultures andfamilies, yet the generalpattemof crises (tasksandissues)remainssimilar. In addition, latercrisesbuild up more complexdemandsin life situationsandthe needfor integrationof strandsofone’s lifeweb — an important cognitive challenge(Kegan,1994).

Eachof theidentitystagesbeyondthefirst requiresco-constructionof one’s own abstractidentitywith thoseof other people, and in each casethischallengingtask requiresa minimum skill level,Table 21.2 lists Erikson’s stages.beginningwithidentity, and shows how each onedependson askill structureat a particular level to afford thecoordinationthat thestage requires.Theearlierlevelsbefore the emergence of identity are also shown,becausetheylaythegroundworkfor identitythrough

Table21.2 Developmentofiden4ticationandidentity: relation to Erikson‘s stagesandgeneralizedskill diagt-am.r

Stage6:

Representationaltier:identification

[MEA> YOU8

]

[MEA — YOU8

]

Rp3 [ME~E—* YOU~]

intimacy versusisolation

Stage7:generativltvversusstagnation

Stage8:ego integrityversusdespair

Abstracttier:identification

Rp4 [ME~e.~~5YOUPtflm[SELFw]Abl I

LME~÷—110U27J or (OThER~J

Ah2

Ab3

Ah4

[~fl2w—flit HER~]

[SELiF. &—*

1~’~*--s rJfl’flJ)fifl~>

t’k,te: Peopledevelopspecificskills, not global ones-Theseformulas mustbe filled in with specificcontentto capturearealskill.

Erikson’sstagesof Skill levelidentity: first emergence

‘1’Concreteidentifications

Rpl

Rp2

Stage5:identityversusrolediffusion

,-1 toll are concretepersonalcharacteristics,g to Z areabstractidentitycharacteristics.

ADULTCOGWITIVEOFVELOI’MENT: DYNAMICSIN THE DEVELOPMENTALWEB 503

theformationof concreteidentificationsthatcharac-terizeoneself(ME’) in relation to important others(YOU). Theseidentifications arecoordinatedwitha minimum of single abstractionsto create thebeginningsofidentity in earlyadolescence,‘l’he skillformulas in Table 21.2 are listed with generalcomponents(lettersfor variablesto be filled in) tomakethepointthatasimilar skill structuredevelopsacrossstrands/domainsin theweb. For applicationto real people,notethatthegeneralvariablesneedtobe specified with concretecontent, There arenoinstantly generalandgeneralizableskills,

Thecreationof multiple concreteidentificationsduringchildhoodsetsthestagefor theemergenceofidentity at the endof childhoodarid the beginningof adolescence.For example,Karaplayedateachergamewith hermotherJane,asdescribedin formulaI. With just aminor changein that representationalskill (level Rp3), sheidentifieswith Janeas both ateacherandamother:

Iden41/cationTEACHER TEAc/lENJANE~—* K,4RAMOTI/ER MOTTlER

Shetries to actlike ateacherandmotherherself,notonly in play but in real-life choicesthatshemakes,suchashelping anotherchild with homeworksimilarto thewaythat sheseesJaneteachstudentsandcarefor ayounger sibling andsimilar to the way Janetakes care of Kara’s brother. With many suchconcreteidentifications,achild builds material forthecreationofanabstractidentity.

Thestageofidentityversusrolediffi,ssioninvolvesabstractanswers(notjust one)to thequestion,‘WhoamI?’ A youngpersonbrings togetherat leasttwoconcreterepresentationalsystemslike formula 8.Coordinatingthat identificationwith her identifica-tion with her fatherasphysicianandparent,

Ident(JicationPHYSICIAN PHYSiCIANWALTER KARAFATHER PARENT

asshownfor level Ab I in Table21.2, shecreatesanidentity of herselfascaregiver:

[SELF -- -

CAREGIVER

At the sametime shebuilds up manyotherspecificidentities,suchas selfasindependent(formula2),and sheconstructsher own conceptionsof otherpeople’sidentitiesin a similar way, suchasthatherbest friend Isabelle is independent:

Much of the confusionof earlyidentity formationcomesfrom the multiplicity of strandsof identityformation and the difficulty in relatuig differentabstractidentitycharacteristicsto eachother atthisoptimal leveL To do comparisonsof txvo personalcharacteristicsof her own identity with that oflsabelle,Kara mustdropdown to concretecharac-teristics, using a representationalsystem. Thecoordinationof herown abstractidentity with thatofher friend’sthus remainsout of reach,

Erikson’snextstageof intimacyversusisolationinvolvesthecoordinationofone’s own identity withthatof afriend or partner,andthe cognitivemini-mumis abstractmappings,asshownin Table21.2.WhenKarafocusesonherown independenttenden-cies,thenshe caneasilycoordinateherown identitywith that of Isabelle as sheseesit. in a repeatofformulaS:

[SaF --- —~ FIUENID[JNDEFENI-i-NT .tNDEPENDENTJ (5)

A challengeof intimacy is to haveherown abstrac-tS) tions about Isabelle matching well enoughwith

Isabelle’sactionsandabstractionsto sustaina closerelationship.Contradictionsalsocomeeasilyat thisoptimal level, becausepeople have difficultiesdealing with multiple abstractionsabout selfandother,WhenKara focusesonherown caregiving,forexample,thereniay be a conflict with Isabelle’sindependence.Within herselftoo, her own care-giving canseemto contradicther independence anexampleof the senseof contradictionandconflictthatmanyadolescentsandadultsexperience(Fischer& Kennedy,1997; Harter& Monsour,1992).

Issuesof intimacy, like theissuesof everyoneofErikson’s stages,exist throughoutlife, long beforethe yearsof early adulthoodandlong after them,The reasonsthat theybelong especiallyto earlyadulthoodareprimarily two: (I) peoplecommonlyseekintimacyatthis age,especiallysexualintimacyandlong-termpartnership;and(2) atthis age,cog-nitivecapacitiesmakeit possibleto truly coordinateabstractidentitiesin intimaterelationships.In manyculturesandlife situations,young adults facethechallengeof deepenedrelationships,involving eithersexor work, wherethe coordinationof identities isparamount.Intimaterelationshipsrequireholdingasenseof self,but alsoan opennesstowardtheunique-nessanddepthof another;theyrequirelearningthemajor componentsof anotheridentity,with the twopeoplebecominga dynamicunit, especiallyfor theideal intimatepartnershipthatEriksondescribes,inwhichbothpartnersgrowtogetherto~vardfulfillmentin astablerelationship.

Intimacy of identities can bemucheasierwith- higher levels, becausethen one skill can readilyincorporatemorethan oneor two abstractcharac-teristicsof selfand other,Marriagepartnerswith

(II) children, for example,can sharethe identities of

I I

[ (9)

(10)

FRIEENE:ff4r.!EnENr!~iENT

504 DEVELOPMENTINADUL THOOD

caregivingandindependence,working togethertosupporteachother asparentsand asindependentpersonswith their own separateneedsin anabstractsystem(Ab3):

Justasidentity only beginsto developwith thelevelof single abstractions(AbI), intimacyonly beginswith the level of abstractmappings(Ab2). That iswhy Table 21.2 refers to the first emergenceofErikson’s stagesata spec/ficdevelopmentallevel,

For thelast two stagesEriksondepictsevenmorecomplexlife tasks,The stageof generativityversusstagnationrequiresmeetingthechallengesofproduc-tivity andcreativity,in contrastto feelingsof lackofpurpose,direction, or self-worth, Themost obviousgenerativity is having children, but generativityinvolves muchmorethanprocreation.This processemergeswith abstractsystems,becausewith themonecancoordinatemultiple abstractidentitiesin selfandothers,as in the exampleof parentingandindepen-dence.However, thechallengesof generativityareenormous:peoplemustcoordinatetheiridentitieswiththoseofnot only theirpartners,co-workers,or friends.but alsochildren,agingparents,andotherpeople.Theabstractthinkingat this stagehasto accommodatearich webof interdependence,relating astrong senseofpersonalidentityanditschangesovertimewith theidentities of others,both youngerandolder, whomoneseeksto guidein wayscommensuratewith theirown needsfor identity andchange.

Bern/ce Neugartenemphas/zedhow cogn/t/vedevelopmentcontributedto theprocessof genera-tivity in a groupof successfulmiddle-agedpeople:‘We havebeenimpressedwith the central impor-tanceof whatmightbecalledtheexecutiveprocessesof personalityin middle age[including] the stock-taking,the heightenedintrospection,andaboveall,thestructuringandrestructuringofexperience— thatis, the consciousprocessingof new informationin thelight of whatonehasalreadylearned’ (1968,p. 98).As onewoman from thestudystated,‘It is asif therewere two mirrors beforeme, eachheld atapartial angle.I seepartofmyself in my motherwhois growingolder,andpartof her in me. In theothermirror, Iseepartof myselfin mydaughter.’Kara’smotherJanein sucha situationconsidersher focuson parenting in relationship to her mother anddaughter,aswell asthe independencethat sheseesin differentformsin all threeof them, Thatkind ofcomparison,going beyondtheconcreteparticularsof onesetof actionsto generalidentity analysis,involvesa highly complexabstractsystem:

[CAREGWEp. FA,RENT PARENT -

I fi &J.~Ti5?1ik SELF<H—> M/STIIEi7,[fNDiEFREDERT iNDEPENDENTINDEPENDENT

(13)

Erikson’s final stageinvolves egointegrity versusdespair,with the challengeof putting the greatexpanseofone‘s life into amean/ngfulsynthes/s,andwith thepotentialachievementofwhatmayproperlybecalled wisdom,Understandingthat oneis manyidentities, in interdependencewith many otherpeople, as well as with the social and culturalroles requiredfor themeaningfulparticipationin ahistorical timeandplace— all thesestrandscoalesceinto whatEriksoncalls integrity asoneapproachestheendof life. Failure to accomplishthis synthesismaybringdepressionanddespairatmidline or in oldage.Achievementofsuchagrandsynthesisrequiresnot only the highest level of abstract thinking,systemsof abstractsystems(Ab4) andthe broadintegrativeprinciplesaboutone’s life that theycancreate,but alsoyearsof experiencerelating one’sown identity to thoseof intimatepartners,friends,co-workers,children,aswell asculturalgroupsandhistorical epochs.This is truly a grandcognitiveachievementt

Networksin Darwin’s DevelopmentoftheTheoryofEvolution

We havedescribedidentity developmentglobally,outlining aprocessthat mostpeopledevelopthrough,taking many different pathways with commonthemes(issues,crises).Now we switchto adifferentperspective:analysisof a caseof one person’sconstmctionof amultistrand,networkedweb. Thecaseof CharlesDarwin’s constructionofthetheoryof evolutionportrays thedynamicsof strandsandnetworksin the web. Dynamicanalysis is at itsrichestin analyzingindividual growth in detail(vanGeert, 1994), andDarwin unintentionallyprovideda greatsourceof datafor analyzinghowhe createdthetheoryofevolutionby naturalselection.Darwinkept a seriesof notesbetween1832 and 1839 inwhichherecordedhisobservationsandideasastheydevelopedinto his theoryof evolution by naturalselection.

The way Darwin constructedhis revolutionaryunderstandingis tantamount to a case study ofbuilding complex knowledge networks in adultcognitivedevelopment.At theageof22, in Decemberof 1831,Danvinsetouton afive-yearvoyagearoundtheworld on the ship HMSBeagle,duringwhichherecordedobservationsandthoughtsaboutthenaturalphenomenaheencountered.Towardtheendof thistime, between1837 and 1839, he kept a seriesof specific notebookson his thinking about ‘thetmnsrnutationofspecies’.In 1839at theageo130,hehad constructedwhat becamehis general theory,althoughhewould notdare releaseit to theworld atlargefor another20 years,whenhefinally publishedThe ot-igin ofspeciesin 1859, Becauseof his notesand notebookswe can peerover his shoulderto

I CO-PARENTS]ELF<INDEPENDENT

CD-PARENT 1SPOUSE (12)INDEFENDENTJ

ADULTCOGNITIVE DEVELOPMENT:DYNAMICS[N TIlE DEVELOPMEN7ALWEB 505

see the steps he took in building the theory andcreatingtheprincipleofevolutionby naturalselection(level Ab4). Table21.3 outlinessomeof themajorsteps in Darwin’s web, showing severalseparatestrands(distinctskills for differentdomains)at eachlevel, Detailedexpositioncan be found in severalother sources,especiallyGruber(1973)andFischerandYan(2002).

Before his voyageon theBeagle,Darwinheld aview’ oftheworldinformedby conventionalreligiousbelief, like other scientistsat that time, God hadcreatedtwo separateworlds, thePhysicalWorld ofsubstancesandtheOrganicWorldofplants,animals,and people. The fact that theseworlds hardlyinteractedwasacceptedas God’s law, In termsofskill level, theconceptfor eachof thesetwo worldsrequiredonly a single abstractionfor eachworld,with little needfor a higherlevel becauseofthelackof interactionbetweenthetwo. It was CharlesLyell’sPrinciplesoJ’geology,whichDarwinavidly read onhis voyage,that openedup for him the questionofinteraction,Inspiredby Lyell’s descriptionof gradualchangein the physicalworld, Darwinwasat greatpainsto recordthesupportingevidencehefound.

Darwinbeganto realizethatthephysicalchangeshesawmight relateto thecommonobservationthatcreaturesill-suited to their environmentby somedefecttendedto die, suchas birds with defectivewingsor fish with defectivegills. This phenomenonsuggestedhow thephysicalworld caninfluencetheorganicworld by gettingrid ofill-adaptedorganisms.Darwin’s knowledgeof the practicesof selectivebreedingof animalsalsocontributedto thedevelop-mentofhis insightabouttheactionofphysicalforceson the viability andadaptationof organisms.Histhinking moved beyond single abstractionstoconstructabstractmappings— lawful interactionsbetweentheworldsofthephysicalandthe organic,asillustrated for level Ab2 in Table21.3.

Darwin’s notesportrayhis yearsoffollowing thisinsight in organizingthe countlessobservationshehadgatheredon his voyage.Oneespeciallyimpor-tant exampleis his work on thevariousspeciesofGalapagosfinches: he discoveredthat the differentspecies’feedinghabitswere closely relatedto theshapesof their beaks(a level Ab3 systeminsight).He realizedthat the form of the beak matchedtheway theparticularfinch obtainedits mostcommonkind of food, This adaptivematch pointedto thefinely honedadaptationof theorganismto its envi-ronment,In anotherstrand/domain,Darwin usedhisknowledgeof fossils to analyzehow specieshadchanged(evolved)over long time periods— howcharacteristicsof currentspeciescouldbe relatedto characteristicsofearlierspeciesthroughconceptsofchangeovertime, In this way, hebuilt systemsofabstractionsin severalindependentstrands,whichhesoonwovetogetherto createthetheoryof evolutionby naturalselection.

In attemptingto build his understandinginto acomprehensiveexplanatorynetwork,Darw’in triedout anumberof conceptsbeforediscardingthemasinadequate.Darwin’sreadingof anessayby ThomasMalthus,conceminghow populationscanreproduceat muchhigherratesthan their environmentscansupport,playedacentral role in his formulation ofthefinal theory.Basedonhisnotebooks,it seemsthatDarwin hit uponhis eventualtheoryseveraltimes,but hewasnot ableto generalizeit fully until hehadreconstructedit repeatedly.This is a commonoccurrencein theconstructionof new knowledge,perhapsevenmore so for complexknowledgenet-works.Darwinnotonly hadto coordinateanumberofcomplexrelationships(coralreefs,finches’beaks,specieschangeovereons),butalsohadto generalizethesecoordinationsinto aprinciple— evolutionbynatural selection,Repeatedconstruction is oftenessentialto newunderstanding:indeedit constitutesgeneralization,with componentsbeingworkedintothe new fabricof a generalskill, suchasDarwin’sevolutionaryprinciple. In sum, Darwin’s construc-tion ofthefinal form ofhis famoustheoryillustratesan extremelycomplexprocessof organizationandreorganization of connections across multipledomainsin orderto build the coherent,innovative,andpowerful knowledgenetwork of evolutionarytheory.

OlderAdults’ CognitiveAgeing

Most adults do not create a new principle thatrevolutionizeshumanthinking,butmostdo dealwiththe challengesof cognitive and physical ageing,including the growthof wisdom,at least for somedomains,andthe loss of somespeedandfacility,especiallylatein life, Whenconceptualizedin tennsof dynamicdevelopmentalwebs, ageinginvolvesgrowthcombinedwith decline,wisdom alongwithslowingdown.

The cultural stereotype,at least in manyWesterncountries,is that cognitiveageingmeanscognitivedecline and intellectual deterioration:‘The older,the dumber.’Other commonfalse beliefsare thatpeoplebecomelesshappyandmorelonelywith age.Happily, researchdatapaint a more optimisticportrait. Most adults experiencemore positiveemotionsandmorenumeroussocialconnectionsastheygrow older,with earlyadulthoodbeingoneoftheloneliestandleasthappylife periods,on average(Carstensen,1993; 2000). Likewise for cognition:researchdoesnotsupportthepropositionofan overalldecline in intelligenceduring adulthoodin concertwith the generalphysicalageingprocess(Wechsler,1972). Horn andCattell’s (1967)classicresearchshows the interweaving of gain and loss withcognitive aging. Many kinds of intellectualskillsincreaseslowly but consistentlywith age, eveninresearchlimited to standardizedpsychometrictests.

506 DEVELOPMENTIN ADULTNOOD

Table21.3 DevelopmentofDa,-w/n ‘s theo,’vofevolution(1831-1839):a generaloverview

Level Skill Major eventsDates

AbI: single Lio~i:o or [WORLD 1 From adolescence:musings Before 1831abstmctions [ORGA.NIC [PHYSICAL] aboutcreationandspecies;

separationof organicandphysicalworlds

r ~.devMnt~ . — - -Ab2: abstract I ‘gj-UIRLU —---- -wOiktO Physicalworldeliminates

mappings [PHYStCAL D-RGAN1C deviantorganisms

[wo-~uun-~~~WOR~-.O Peopleselectivelybreed

[HUMAN rORGA.NIC animalsandplants fordesiredcharacteristics

Voyageof Beagle:mastering 1831—1836Lyell’s Principlesofgeologvandcollectingobservations

[‘woRoo ~~~‘WORLO1 Beginningofcoralreeftheory:[D:R-GANIC PHYSICAL] coralsvary with changesin

physicalworld

Ab3: abstract [OROAC H reef PHYSICAL xl Final coral-reeftheory:coral 183 5—1837systems Wa-RIO- ~c—> WORLO reefsgrowascoralsadapt

[ORGANIC Y PHYSiCAL Yj to changingoceandepthsbygrowingupwardto reach light

[G~R0ANICH match PHi- S]1CAL xl Variationsin Galapagosspecies1 WORLD- i—> WORLD- matchspeciescharacteristics[-ORGANIC Y PEYgICAL Yj with physicalniche

10-EG~~NICRI time ORGANIC P.2 1 Many speciesshowsystematicWORLD ‘.) WORLD I changeoverlongtime periods

LO-ROAN1CTi CRGAMC T2 JStrugglingwith ideaofmultiplecreations

monadcreationAb4: systemsof ORGANIC XI PHYSiCAL Xl Inadequateprocessof evolution: 1837

abstractsystems, WORLD WORLD monadtheory,branchingtreewhich areprinciples ORGANIC Yi pHYSICAL ‘ti

treeofUchangeORGANIC X2 PHYSICAL X2‘-:VJOEli) (—3 ‘-WORLDfYp:GANCC Y2 PhYSICAL Y2

Hybridization(insteadofnaturalselection)

evolution

Xi PHYSiCALXI Emergenceoftheoryof 1838—1839W•ORLD ‘. ~:i~JOT/jLD evolutionby naturalORGANiC Yl PHYSICAL Yl selection(heredity,variation,

natural selection naturalselection)- ORGANICX2 PIH’YSICALX2

WORLD <—‘—3 WOI9JLDORGANIC Y2 PHYSI1.AL Y2

Vote: Skill structuresin this tableemphasizerclaiionsbetweenphysicalandorganicworlds in thevariousphasesofDarwin’s work. FischerandVan (2002) describetheactualskills, specifyingthecomponcnisfrom physicalandorganicworlds that Darwinactuallycoordinated.

So,o’ce.r: Table adaptedfrom Fischer& Yan (2002). Sourecsfor historicalinfonuation:Barrett t 1974).t)arwin (1859).Keegan(1989),andespeciallyGrubcr(l 981).

ADULTCOGWITIVEDEVELOPMENT:D1WAMTCSIN THE DElT

ELOP~-tF.NTALWEB 507

Thesereflectxvhatis calledcrv,susllizediuielhgeitce,composedof sk-ills that benefit from accumulatedexperience,suchas vocabulamyandgeneralknow-l-edge.On theotherhand.mattyskillsalsodeclinewithage, especially from middle adulthood,and thesereflect whatis calledfiuid intelligence,composedofskills thatdependonnovelactivitiesaridinfonnation.Most of theactivities that adultsneedto do involveaccumulatedknowledgeandcrystallizedintelligence,andthey getbetterwith age. For example,Schaje’s(1996) longitudinal data indicate that inductivereasoningrisesslightly throughmiddle adulthood,with a gradualdecline begittningonly in lateadult-hood. On the other hand, there are clear, smalldeclinesin speedandphysicalstrengthbeginninginmiddle age(I-Ions, 1982: Salthouse,I-Iarnbrick, &McGuthiy,1998). Illncssis alsoan importantfactor,producingpowerfuldeclinesin skill at anyageandbecomingmorelikely in old age.

Cognitiveageingis clearlymultidimensionalandmultidirectional (Baltes, l987; Baltes & Baltes,I 990; Berg,2000; Bitten, 1970;Craik& Sahhouse,1991; Schaie,1983; Sternbcrg,1985). Researchthatseemsto showonesimplefactorunderlyingageing(or development)is basedon assumptionsandstatis-tical techniquesthat forcecomplexwebsinto single,monolithic dimensionsandprecludeconsiderationof thetexturedrichnessofdevelopmentalwebs(e.g.Gottlieb,Wahlsten,& Lickliter, 1998).

The standardizedtestsusedin mostageingresearchdo not assessthe complexskills that developat thehighest levels of abstraction.or the integrationofemotional andcognitivestrandsthat groundwisdom,Even the reduction in speedwith age sometimescomes from the increasedsophisticationof adultcognitiveskills andtheir thinking processes:morecomplexwebs andnetworkstake longerto processinformation (Atkinson & Shiffrin, 1968; Fischer&Rose, 1994; Gruber, 1973; Schacter,4999). Whatmight becalled cognitivepragmatics,or theculture-basedsoftwareof mindand body, actually improvewith age,asevidencedby numerousstudiesof adultdevelopment(Baltes& Baltes, 1990; Colby et al,,1983;Dawson,2002;Erikson,1969; 1978;Levinson,1978; Loevinger, 1976; Neugarten, 1968; Noamet al,, 1990; Vaillant, 1977; 1986). Oneof the mosttelling findings is the age at which productivityreachesits maximum in creativepeopleworking inhighlycomplexfields,suchashistoriansandnovelists(Denois, 1958; Simonton, 1991; 2000). People inthesefields becomemostproductiveand creative intheir40s,SOs,andeven60s. In contrast,peoplein lesscomplexlytextured fields, suchasmathematiciansandpoets,oftenpeak in their 20sand30s.

How can the complex interconnectionsamongthesefeaturesin cognitive ageinghe analyzedandunderstood?Obviously,theyarenot totally separatefeaturesmoving in differentdirections,With multipleelementsinteractingwith eachother over time, a

dynamicprocessof self-organizationoccursin whichadultsactively organizetheirlimited mentalresourcesinto dynamicskill networksto adaptto their complexlife tteeds.Whilesomecomponents,suchasspeedofactivity and speedof processinginformation,reducethe richnessof the networkpastmiddle age,othercomponents,such its syntheticthinking and inter-personalwisdom,can increasetherichness.Throughthis process,both ageing and aged adults builddynamiccognitivenetworksto meetvariouscomplexlife challenges.Two examplest’rom researchmnvolvethe specific motor skill of typing and the broadcognitive-emotionalskill ofwisdom.

For themotor skill oftyping,oldcr adultsorganizetheir skillsdifferently, anticipatinga wider spanoflettersandcompensatingfor lowerperceptual-motorspeed in simple tasks suchas reactiontime andtapping(Salthouse,1984). In a sampleof 19- to 72-year-olds.older adults maintainedtyping speedbymorepreciselycontrolling the sequencingof key-strokesacrosslargerspansof charactersthan didyounger typists. For many skills besidestyping,dynamic interconnectionand compensationplay asimilar critical role in changesduring adulthood,Dynamiccompensationandadjustmentwithin com-plex cognitive networks are widely observed indomains of memory (Barrett & Watkins, 1986,p. 129),chessplaying(Charness,1981),socialinter-action(Carstensen,l993),emotionalunderstanding(Labouvie-Vief,De-Voe, & Bulka, 1998), andjobchange(Stemberg,1990).

Wisdom is very different from typing, yet it tooinvolves building complexcognitive networks toadaptto needsin life, Wisdomrequiresintegrationofmultipletypesofknowledgeandskill aboutptacticalandethical issuesin humanlife (Baltes& Staudinger,1993; Dawson, 1999; Erikson, 1963; Stemberg,1990). It requiresan implicit, complex, effectiveknowledge network combining multiple domainsoverlongtimeperiodsandextensiveexperience,andit seemsto requirespecialcoordinationofemotionalandcognitiveprocesses.Wisdomseemsto compen-sate for physical slowing in middle and old age,enablingmany adultsto perform syntheticthinkingaboutself and othersin the complexworld, oftenmakingpeopleespeciallyeffective as, for example,political leaders,judges,moral leaders,andscholars,

In summary, adults can develop deeply andbroadly in their skills — building complex identitieswith their family, friends, andcolleagues,creatingnew ideas,practices,or productsthat shapetheirsociety, and building wise ways that go beyondself-interestand immediate response.All theseoutcomes depend on developnsentof networksconnectingmultiple strandsin the web of skill andemotion,

508 DEVELOPMENTINADULTHOOD

DYNAMIC BACKWARD AND FORWARD

CONSTRUCTIONSIN THE WEB

Besidesmultiple levels, strands,and connections,development in adults also mOves in multipledirectionsfor cognitive construction,The completepicture includesnot only complexvariationsin thestrandsandnetworksbutalso dynamicconstructionprocesseswithin theweb. Dynamically,adult cog-nitive developmentmovesforward,backward,andinvariousotherdirections.It formsadynamicweb,andeveneachseparatestrandis dynamic(andfractal),not a linear ladder (Fischer & Granott, 1995).Traditionally developmentis defined as forwardprogression, but cognitive developmentmovesbackwardaswell asforward, ‘Progress’resultsfromacombinationof backwardandforward mnovement,with muchbackwardmovementpreparingthewaytomove forward through the constructionof newadaptiveskills, Throughthus constructingskills inmultiple directions,adultscanhandlecomplextaskseffectively andflexibly andadvancetheir compe-tence, Adults’ backward directions of cognitiveconstructionaresometimestreatedasanindicatoroffailure andmalfunction,especiallyin old age,butinstead,flexible useof simpler andmore complexskills reflectsmaturity andwisdom,Two develop-mentalphenomnenashowsomeof the order in themultidirectionalityof adult development:backwardtransitionandforward consolidation,

Backward Transition

Oneimportantprincipleof dynamicconstructionisbackwardtransition,a movementof activity fromhigher-levelskills downto lower-levelonesfollowedby gradualmovementin fits andstartsbackup tohigher-level,new skills (Duncker, 1945; Fischer,198Gb; Fischer& Granott, 1995; Granott, l993a;l998; 2002). Backward transition or regressionseemsto be a universalstrategythatpeopleusewhentheyaretrying to constructnewskills, asreflectedinexplicit problem-solvingstrategiessuchas‘breakingaproblemdowninto its simplestunits’ and‘startingagainfrom the beginning’. Whenpeopleencounterataskthat theydo nothavetheskills to perform,theyfall down to low-levelactivity — evensensorimotoractionssimilar to thoseofaninfant— so that theycanfigure out thetaskandgraduallybuild towardhigh-level skills, Recentmicrodevelopmentalresearchindicatesthatbackwardtransitionsarepervasiveinadulthood and play such an important role thatsupposedlyinadequate,lowerlevelsof performanceneedto be reexaminedandreevaluated,Backwardtransition leads adults to perform flexibly andtodevise ways of solving complex tasks that areinitially beyondthem,

Nira Granott (l993a; 2002)deviseda method-ology that allows a focus on both realtime skillconstructionandthegeneralizationandconsolidationthataccompanyit, Shehadteamsof adults interactwith a small Lego robot called a wuggle. Thewriggle,whichwasaboutthesizeof a toy truck,wasprogrammedto respondto changesin light/shadow,sound,andtouchby altering its movements,Thedyadswere giventhe taskof figuring out what thewuggle did, while video camerasrecordedinter-actionsanddiscussions,Granott(l993b)reasonedthatbecausemosthumancognition takesplacein thesocialarena,observingdyadswill provide usefulinsight into spontaneouslearning and problem-solving andthe social interactionwill makeovertmanyleamingprocesses.Participantsbeganin asmallgroup in a roomcontainingseveralwrigglesthat movedor satamongcommonobjectssuchastables and boxes, and they formed dyads spon-taneouslyduringtheobservationsession,

In analyzingvideotapesof this session,Granottand her coderseasilyreachedagreementon whatconstitutedan interchange, a single dyadic inter-actionwith the wuggle,andeachinterchangewasscored for complexity using the skill scale, Forinstance,whenadyadunderstoodthat makinga loudnoiseled thewuggle to changedirection, the levelwascodedasanaction mapping— that is, thesecondlevel of theactiontier (Smn2), connectingmakingorhearingasoundwith seeingachangein movement,

In atypical initial encounter,thesubjectswere atfirst confused,thenmoreengagedastheirexperimen-tationbroughtresponsesfrom the wuggle.Granott(1993; 2002)usedthe dyad of Ann andDonaldtoillustratethis process.Encounteringawugglefor thefirst time, they had to learn how it changeditsmovementin responseto asound,In 148 interchangesover27 minutes,theystartedwith mereobservationof thewuggle’smovement(thelowestlevel ofsingleactions,SmI) andgraduallybuilt up throughhigherlevelsas they tried to understandthe wuggle.Thesequencefrom actionto representationunfoldedasfollows, Seeingthewugglemoveis a single action,andhearinga soundis anothersingle action(levelSmI). A mappingof actions(level Sm2) is noticingthat hearinga loud sound goes with seeingthewugglechangemovement,A systemofactions(levelSm 3) is combiningseveralmovementsandsoundsituations.A singlerepresentation—-thewugglereactsto sound— emergesfrom coordinatingseveralactionsystems(level Sm4/RpI). And so forth:mnicrodevel-opmentcontinueswith relating anddifferentiatingrepresentationsofthewuggleandsound,

Among the dyads,however,suchconstructionofarepresentationdidnot proceeddirectly throughthelevels in a ladder-likemanner-As shownin Figure21.7,progresstowarda skill for understandingthewugglewent in fits andstarts— aseriesof backwardstransitionsandreconstructioos,not oneconsistentup’vard construction,Initially Ann andDonald fell

ADULTCOGNITIVE DEVELOPMENT:DYNAMIGSIN THEDEVELOPMENTALWEB 509

I,>0

to

Redogadget

Describe Redogadget gadget

Abi

Rp3

Rp2

Rpl

Sm3

8m2

Smi

Interchange

Figure21.7 Backward transition and microdevelopment in understanding a wuggle:Ann and Donald (Gronott, I993a; 2002)

downto a level far belowtheir capacity,producingseverallevel I actionsandthen building up amorecomplexskill overseveralminutes.Theyinteractedwith the wuggleandmadesoundsandotheractionsto exploreit, in a falteringway graduallybuildingtheir first representationthat the wuggle reactstosound(level4). But thenatinterchange65 somethinginterestinghappened:a wire hadfallen out of thewriggle,andwhentheyplacedit back(ina differenthole,by mistake),the wriggle acteddifferently. Inthe face of this task change their fragile skillcollapsed,droppingbackimmediately to a level Iaction.Overthenext severalminutestheyoncemorerebuiltmorecomplexskills, graduallyretumingto arepresentationthatthe wriggle reactsto sound(level4) and then going further to higher levels still,relatingseveralrepresentationsto eachother.

Thisprocessofbackwardtransitionandreconstruc-tion happenedtwo moretimesin the27 minutesofproblem-solving.At interchange118 Ann andDonaldencounteredanothervariationin thetask:theysetoutto summarizewhat theyknew,andagainthechangein taskledto adrop in their skill — this time level 2mappingsof actions followed by againrebuildingskills to reachrepresentations(levels4 and5). Thenat interchange134 Ann andDonald changedthewiring of thewriggle again,andthey showedback-~vardtransition to low--level actions followed byreconstructionofcomplexactionsandrepresentations.

The repeatedfall andriseof skill levels in theconstructionandgeneralizationofnewknowledgeisa common feature of microdevelopment.Adultleamersalsoshowedthis scallop-shapedgrowth inYan’s (1998; 2000)recentstudyof learningto riseacomputerprogramto dosimplestatisticaloperations.Participantswere graduatestudentswho variedwidely in their expertise,both with computersandwith statisticaloperations.Eachstudentworkedat acomputer,with ateacherat herorhis sideto answerquestionsandto intervenewhenhelp wasneeded,Studentswith intermediatebackgroundexperienceshowedclearscallopingin their leaminggraphs— alow level of skill followedby agradualincreaseandthen anabruptdrop whena new task elementwasintroduced,asillustratedin Figure21.8. Forstudentswith little background,skill level showedwidefluc-tuation initially, andscallopinggraduallyappearedas they became more familiar with the tasks,concepts,andcomputeroperations.Studentswith ahigh degreeof knowledge(experts),on the otherhand,showed,little scalloping,staying generallyattheupper limit of skill requiredby the task, withoccasionaltransientdrops.

Anotherimportantfinding of thesestudiesis thatadults function at a level appropriatefor the taskathand, which may be far below their upper limit(either optimalor functional level). In Figure 21.8,for example, a highly intelligent adult graduate

4’

0 20 40 60 80 100 120 140

510 DEVELOPMENI’IN ADULTHOOD

Abi -

Rp3 -

0

0

~ Rp2 -

Rpl -

Sm3 I I I ] F I I I I I I I I

5 10 15 20 25 30

Activity

Figure21.8 Scalloptng in learning a statistics operation (Tan. 1998,’ 2000)

studentperfonnedwith skills thatatmaximumwereonly reprcsentationalsystems(Rp3). This optimallevelfirst emergesat6 or 7 yearsofage,andtheadultstudent was capableof muchmore sophisticatedactivity, including high levelsof abstraction,whichsheregularlydemonstratedin otherclassactivities.The reasonfor the low level thatsheandall otherstudentsdemonstratedin this studyis that thetaskrequiredonly this level — nothingmore!

Yanfurtheraskedhow interactionwith theteacheraffectedlearning.Hefoundthattheupwardarcof thescallopoften followed the teacher’sresponseto arelevantquestion,especiallyfor the intermediate-level students: the scaffolding provided by theteacher’sresponseallowedthe studentto build upunderstandingofthetask,Thesupportofferedby theinstructor through clues and priming facilitatedthetemporaryrisesin skill levelevidentin scalloping.

Justasateachercanprovidea scaffoldto supporta student’sconstructionof nrmderstanding,peoplecan supporttheir ow-n skill constructionthrougharecently discoveredmechanismn called bridging(Fischer& Bidell, 1998;(iranott, l993a; Granott &Parziale, 2002; Parziale. 1997), A bridging shellalloxvs people to bootstrap themnselvesto newknowledgeby creatinga temporarytargetor open-endedshell for what is asyetunconstructed.Theshellis a framework (an attractor in dynamic systemsterms) for guiding a cun’ent level of perfonnancethroughthesearchspaceto thenexthigherlevel — likean algebraicskill fonnulawith unknownvariablesthatapersonusesto guidediscoveryduringproblem-solving.In thesvuggletask,dyadscontinuallycreatedshells that helpedbridgetheir explorationof thewuggleto higherlevels.For example,thedyadKevin

and Marvin noted that their wuggle showeda‘reaction’ to somethingthat they did, but theycouldnot articulate either the cause of the reaction(unknownvariableXa) or thenatureof thewuggle’schangein activity (unknownvariable~h) (Granort,

Fischer,& Parziale,2002). They used a sketchymappingskill asa shell to bridgetheir constructionofan understandingof thesetwo factors:

[(X) ~ (14)

Explorationof the wuggleguidedby this shell ledthem to aseriesof moreexplicit skills basedon theshell,starting with therealization:‘When it comesover hereandas soonas it getsunderueathpart ofthe shadowthere, it startschangingits behavior,’This statementof acausalrelationshipbeganto fillin the shell:

[SHADOW IT (WUGGLE)(ON WUCCLE) CHANCE(S)

L BEHAVIOR (15)

Aswith all newkno\vledge,the newskill remainsatemporaryone until it canbe reconstructedseveraltimeswith sufficientvariationso thatit stabilizes,Ina similar manner,adults rise bridging fi’equently toguidetheir o\vn learning.This processof bridgingcriesout for researchto unpackhow adultsguidetheirownlearningamid development.(Notethat boldfont indicatessensorimotorskills, which arebasedin action.Thelevelsprior to representationsinvolveactions,whichfonn thebasisfor represenlalions.)

The examples’with wrigglesand computerpro-gramnsdemonstratethatknowledgeis not simply a

Participant07, Session1

‘

‘ I — — — —-

Scallops

ADULTCOGNITIVE DEVELOPMENT:DYNAMICS[N THEDEVELOPMENTALWEB 511

stable accomplishment. In both studies, peoplemovedto high skill levelsin a short time, hut whenthey encountereda small changein thetask, theyinstantly fell backto lower levels.Generalskills mustbebuilt throughthis repetitiveprocessofdoingandredoinga taskto stabilizeandgeneralizeit, When-everatask is changed,tlmere is l,ackward transitionandreworking.graduallyleading to amore stablerepresentation(Bever, 1982;Duneker,1945;Fischer,1980a; l980b; Fischer & Bidell, 1998; Granott,1993a; 1998; Granott, Fischer,& Parziale,2002;Werner, 1948). The pointhereis thatpeopledonotsimply work up to a level of skill andthen keepitavailablefor all similar circumstances.For knowl-edgeto becomereadilyaccessibleacrosstasks anddomains,it hasto bereconstructedmultiple times,probablywith its flexibility detenninedin largepartby therangeof variationsin thetaskswhenapersonhasto reconstructit.

Knowledgedisappearseasilyandhasto berecon-structed.It is unstable,Relativelystableknowledgecomesonly with extensivegeneralizingreconstruc-tion for familiar tasksandsituations,(Knowledgecanbe stablein thecommunitywithoutbeingstablein the individual.) With so little researchon thenaturallydynamicvariationin individual activitiesand knowledgein the real world, scientistsandeducatorshave too easily treatedknowledge asstable,evenfixed — thusperpetuatingamyth of stableindividual knowledge that permeates humanlanguageandculture(Lakoff& Johnson,1980).

ForwardConsolidation

Thephenomenonof forwardconsolidationinvolvesadifferentpatternof movementduringadult develop-ment:Theoptimalperformancethatcomeswith highcontextual support is gradually consolidatedintofunctional performancewithout contextualsupport.Most cognitive-developmentalresearchexaminesonly conventionalforw-ard progressionwithin thesame contextual condition: from lower optimalperformanceto higher, or from lower functionalperformanceto higher,or from lowerperformanceonastandardizedtestto higher.Whenyoung adultsface,for example,a difficult dilemma,suchaswhetherchemicaladditivesto food arehelpful orhannflml, orwhetheran unwed, poor, young woman who ispregnantshould considergiving up the child foradoption,they will showahigherlevel of reasoningwith optimalcontextualsupportthanwithoutit. Theycannotsustaintheoptimal level on their own,buttheycanrememberit vaguelyandbuild a bridgingshellthat eventually leads them to consolidation andmasteryof thehigherlevel skill withoutsupport.Asshownin the developmentalweb in Figure 21.5,forwardconsolidationtakesplacealongthe strandsso that the optimalportion will graduallybe turnedinto the functmonal portion. This kind of forward

consolidationis pervasivein adulthoodandplaysanimportantrole in adults’ cognitive development.

Onedemonstrationof forwardconsolidationis thepatternof emnergenceof perfonnanceof skills at agiven level. ‘Fhe skills thatemergeat oneoptimallevel predominatenot whentheyemerge,but yearslatcm; often upon emergenceoftime next oprimal level,or eventheoneafter that (Fischer,Kenny, & Pipp.1990;Kitcheneret al,, 1993). In other words, theconsolidationof skills at a given level takesplacewith theemergenceof the next level. In reflectivejudgment,skills for stage6 first spurted at age20,the usualage when the optimal level of abstractsystems(Ab3) emerges.llowever, studentsat thatage only producedabouthalf of theirargumentsatstage6, asshownin Figure21.6. Not until five yearslater at age 25 did stage6 performancejump tonearly 100%, (Age 25 was also when stage7performancejumpedto 50% astheoptimal level ofprinciples,Ab4. emerged.)

The numberof yearsthat it takesadultsto movefrom optimal level to consolidationof functionallevel variesgreatlyacrossdomainsandindividuals.Forreflectivejudgnent,Table21,4describestheagerangebetweenemergenceof anoptimal level for astageandtheconsolidationofthat skill atfunctionallevel,Theseagesarebasedmostlyon researchwithAmericanstudentswho haveacollegeeducationorplan to attendcollege,and of coursetheyvaryforpeoplefromotherculturaloreducationalgroups.Forexample,stage5 reasoningemergesin manystudentsasearlyas IS yearsof ageunderhigh support,but inlow support(functional)situations,it is notseenuntilsomewherebetween18 and30 yearsof age.Simi-larly, stage6may appearat 20 yearsunderoptimalsupport,but it is not consolidatedat functionalleveluntil 25 to 40years.Notethat theagesfor functionallevel involveonly adultswhoactuallyshowedthosestagesin research.Many adults neverreach thehigheststagesin anyparticulardomain,asevidencedevenin researchwith college-educatedadults,

It takesyearsfor anindividual adult to movefromemergenceof anoptimalperformanceto consolid-ation of a functional performance.Darwin tookseveralyearsof intensethinking with high self-scaffoldingandlong immersionto movefrom thetheory of coral reefs to the principle of evolutionby natural selection, even though the coral-reeftheorywaslaterseenasan instanceofthe principle(Fischer& Yan, 2002). Theextensionofthat theoryto hundredsof problemsin biology went on forthe restof his life. Forwardconsolidationis both achallengingcognitivejourneyandasignificantintel-lectimaiaccomplishment,whetherforan extraordinarythinkeroranordinaryadult,

Why are there such gaps in the timing andperfonnanceof reflectivejudgmentandotherskills?Catastrophetheory (a kind of dynamics)helpstoexplain thesenonlinearprocesses.When a numberof influencesact together, they can produce a

512 DEVELOPMENTIN ADULTHOOD

Latehigh school, college,andabove16 to 23 yearsNeverfor manypeople anddomainsEarly graduateschool19 to30 yearsor olderNeverfor manypeopleanddomains

Advancedgraduateschool23 to 40 yearsor older

Neverfor manypeopleanddomainsAdvancedgraduateschool

30 to 45 or older.Neverfor manypeopleanddomains.

Note:this tableincludesonly thelast five ofthesevenstages,which aretheonesthat adultsusemost.I Agesfor emergenceoffunctional level varywidely,andso theseestimatesarecoarse.

Sources:reviewsandresearchby King andKitchener(1994),KitchencrandFischer(1990),Kitcheneret al. (1993),as’well asBasseches(1984),Colby et al, (1983),Cook-Greuter(1999),Davison(2002),Fischer,Kenny, andPipp(1990),Peny(1970),Restet al, (1999),andVaillant (1977).

nonlinearpatternwith a complexshapethat includespowerful discontinuitiescalled catastrophes (vander Maas & Molennar, 1992; Zeeman, 1976).Catastrophetheory describeshow a developingpathway can bend back on itself over time as itprogresses,giving a distinctivescallopedshapeto theascendingpathway.This backwardbendingshowsaremarkableparallelto thespikinessandgappinessinthe developmentofoptimal levelin agiven domainor in the confluenceof integratingdomains,asshownin Figures 21.3 and 21.6 (Fischer & Bidell, 1998;Rose & Fischer,1998). In a sense,the cognitivecapacitiespressedinto serviceunderhigh-supportconditionsareunstableuntil thepersonhasconsol-idated them through extensive experienceandpractice.The instability takestwo forms: (I) dcve-lopmnent of optimal performanceshows suddenjumps and drops;and (2) the level appearsanddisappearswith variationin contextualsupport.

Backwardtransitionandforwardconsolidationaswell asothergrowthprocessesform foundationsforthe dynamic phenomenaof adult development.Thesedynamicprocessesoperatewithin thestrandsof thedevelopmentalweb,andtheycreatethe widerangeof levelsof everydayskill. Skills range fromlargedropsto basic levels in backwardtransition tohigh levels of new skill constructedon thesebasicactions.They rangefrom low levelsof automatizedactionsto functional levels of unaidedactionsandfurtherup to optitual levels of supportedactions.Theyevenextendto the highreachesof collaborative

action,Therangeofvariationis especiallybroadandpervasive in adulthood, even more so than inchildhood.

At least three reasonsaccountfor this broadvariation, First, adults have a wider rangeof skillsavailablebecausetheyare capableof going all theway from elementarysensorimotoractionsto com-plex abstractions.Second,thehigh-levelabstractionsofwhichadultsarecapableareespeciallysubjecttothe influencesof culture and education,even morethanthe basicskills of childhood. Third, adultstendto specialize in particulardomains,basedon theirlifechoicesand situations — enteringonejob and notanother,one avocationandnot another,one familyrole andnot another,

CONCLUSION: RICHNF.SSAND COMPLEXITY

OFADULT DEVELOPMENT

Accumulatedevidenceindicatesthat Piaget’sfonnaloperationsis not the endofhumancognitivedevelop-ment. Instead, developmentover the 60 years ofadulthoodis an importantpart of thewhole pictureofhumancognition.Adult cognitivedevelopmentisrich anddynamnic,like a complexwebthat is con-stantly changingwith multiple levels, strands,net-works, anddirections.The wisdom and intelligenceof anadult cannothecapturedby onedevelopmcntal

Table21,4 Appro.titnate ages for optimal and functional levels of reflective judgment

Stageofreflectivejudgment Emergenceof optimal level Emnergenceof functional level

Middle schoolandhigh schoolage.12 to 17 years

Pm-c-reflectivejudgmentStage3 (levelRp3) 6 to 7 years

Quasi-reflective judgnmentStage4 (level AbI) 10 to 12 years

StageS(level Ab2) 14 to 16 years

Reflective judgmentStage6 (level Ab3) 19 to 21 years

Stage7 (level AM) 24 to 26 years

.1.~

ADULT COGNITIVE DEVELOPMENT: DIW4MICS IN THE DEVELOPMENTAL WEB 513

level, one domain, one pathway,or one direction,During adulthood, intelligence commonly movestobecome more sophisticated, flexible, synthetic,constructive,and socially orietsted— morecomplexanddynamic.Cognitivedevelopmentin adulthoodtakes a numberof different shapes,and it occursthroughasetoffine-grainedmechanismsfor buildingandadaptingskills,Specificskills emergeat onelevelbut requirelong periodsofconsolidationbeforetheypredominatein ordinary contexts, They emergeabruptly as new optimal levelsfor a givendomain,hut developsnoreslowly andgraduallyasfunctionallevelsof everydayaction,An importantmechanismfor construction of new knowledge is backwardrecursionto lower levels of actionsand represen-tationsfollowed by repeatedrebuilding of a skilluntil it is consolidatedand stabilizedas availableand generalizable.A major impetus to the wosk ofconstructingnew approachesto the challengesof arapidly changing world is through leaming withothers— colleagues,friends,mentors,parents,rela-tives, andeventheir children.Researchin the futurewill needto unpackfurthertherichness,complexity,and dynamics of adult cognitive development,building knowledgeof thedevelopmentalwebsanddynamicprocessesthatwe havebegunto describe,By openingup thescopeof researchandtheorytoanalysisof thedynamicsof variability and change,we can better understandthe true richness andcomplexityof eachadult’s adaptiveconstructionofknowledge.This new knowledgeof what, how,andwhy adult cognition changescan eventuallyhelpmillions ofadults to meetnew challengesfrom theircomplexnatural, social, andspiritualworlds moresuccessfullyandenjoyably.

REFERENCES

Alexander,C.N..Davies,IL., Dixon, CA., Dillbeck, M.C.,Druker.S.M..Oetzel,R.M., Muehlman,J.M., & Orme-Johnson,D.W. (1990). Growth of higher stagesofconsciousness:Maharishi’sVedic psychologyofhumandevelopment.In C.N. Alexander& E:J. Langer (eds),Higher stages oJ human development: Perspectives onadult growth (pp. 286—341). New York: OxfordUniversityPress.

Atkinson,R.C.,& Shiffrin,R.M. (1968).Humanmemory:A proposed system and its control processes.InK.W. Spence& J.T. Spence(eds),The psychology oflearning and motivation: Advances in reseat’ch andtheory (Vol.2.pp. 89—195).New York: Academic.

Bakes,PB, (1987). Theoreticalpropositionsof life-spandevelopmentalpsychology:On the dynamicsbetweengrowth and decline, Developomental P.s’s’chology. 23,611—626.

Baltes, PB., & Baltes, M.M. (eds) (1990). Successfulaging: Perspectivesfront thebehavioralsciences.NewYork: CambridgeUniversityPress.

Baltes.PB.,& Mayer,KU. (ed.)(1999).The Berlin aging

study: Agingfrom 70 to 100. New York: CambrmdgcUniversityPress.

Baltes,PB., & Staudinger,U.M. (1993). The searchfor apsychologyof wmsdom. Current Directions in Psycho-logical Science, 2, 75—80.

Barrett,G.V., & Watkins,S.K. (1986). Word familiarityandcardiovascularhealthasdetersninantsofage-relatedrecall differences.JournalofGerontology,41,222—224.

Barrett, PH. (1974). Darwin’s early and unpublishednotebooks. New York: Dutton.

Basseches,M. (1984). Dialectical thinking and adultdevelopment. Norwood,NJ: Ablex.

Benes,F. (1994).Developmentofthecorticolirnbicsystem.In 0. Davison& K.W. Fischer(eds),Human behaviorand the developing brain (pp. 176—206).New York:Gtsilford.

Berg,C.A. (2000).Intellectualdevelopmentin adulthood.in R.J. Steinberg (ed.), Handbook of intelligence(pp. 117—140).New York: CambridgeUniversityPress.

Bever, TO. (ed.)(1982).Regressionsin mentaldevelop-oment: Basic phenomena and theories. Hillsdale, NJ:Erlbaum.

Birren,I.E. (1964). The psychology of aging. Englewood,NJ: Prentice-Hall.

Birren, I.E. (1970).Towardanexperimentalpsychologyofaging.American Psychologist, 25, 124—135.

Brown, AL.. & Reeve,R. (1987). Bandwidthsof compe-tence: The role of supportivecontextsin learning anddevelopment.In L.S. Liben (ed.) Development andlearning: Conflict or congruence? (pp. 173—223).Hillsdale,NJ: Erlbaum.

Bullock, M., & Ziegler, A. (1994). Sctent(fic reasoning.

Researchreport. Munich: Max Planck Institute forPsychologicalResearch.

Carstensen,L.L. (1993). Motivation for social contactacrossthe life span: A theory of socioemotionalselectivity. In I.E. Jacobs(ed), Nebraska symposiumon motivation (pp. 209—254). Lincoln: University ofNebraskaPress.

Carstensen,L.L., Pasupathi,My.. Mayr,U., & Nesselmade,I.E. (2000).Emotionalexperiencein everydaylife acrossthe adult life span.JournalofPersonalityand SocialPsychology, 79, 644-655,

Cattell, R.B. (1963). Theory of fluid and crystallizedintelligence:A critical experiment.Journal of Educa-tional Psychology, 54, 1—22.

Chamess,N. (1981). Aging and skilled problemsolving.Journal of Experimental Psychology: General, 110,21—38.

Colby, A., Kohlbeing,L., Gibbs, I., & Liebersnan.M.(1983). A longitudinal study of moral development.Monographs of tIme Society for Research in ChildDevelopment; Vol. 48, no. I, serial 200.

Commons,M.L., Trudeau,El., Stein,S.A., Richards,F.A.,& Krause,SR.(1998).Hierarchicalcomplcxity oftasksshowstheexistenceof developmentalstages.Develop-mental Review, 18, 237—278.

Cook-Greuter,SR. (1999).Postautonomnousegodevelop-ment: A study of its nature and development.

514 DEVELOPMENTINADULTHOOD

Unpublisheddoctoral dissertation,Harvard GraduateSchool ofEducation,HumanDevelopment& Psychol-ogy, Cambridge,MA.

Craik, F.I.M. (1977). Age differencesin humanmemory.In i.E. Birren & K.W. Schaic (eds),Handbookof thepsychology of aging (pp. 384—420).Ndw York: VanNostrandReinhold.

Craik,F.I.M., & Salthosmse,TA. (eds)(1991). The hand-book of aging and cognition. Hmllsdale,NJ: Erlbaum.

Darwin, CR. (1859). The origin of species. New York:Penguin.

Dawson,T.L. (1999). A good educationis. . .‘: A life-spaninvestigationof developmentalandconceptualfeaturesof evaluativereasoningabouteducation.DissertationAbstracts international: Section B. The Sciences &Engineering, 60(3—B),1329.

Dawson,T.L. (2002).New tools, new insights:Kohlberg’smoral reasoningstagesrevisited. international Journalof Behavior Developnsent, 26, 154—166.

Dennis, W. (1958). The age decrementin outstandingscientific contributions: Fact or artifact? AmericanPsychologist, 13, 457—46n,

Dewey,J. (1910).how we thi,mk, Amherst,NY: PrometheusBooks, 1991.

Duncker,K. (1945). On problem solving. PsychologicalMonographs, Vol. 58, serial 270.

Erikson, E.H. (1963). Childhoodandsociety (2nd edn).New York: Norton.

Erikson,E.H. (1968).Identity: Touth and crisis. NewYork:Norton,

Erikson, F. (1969). Gandhi’s truth: On the origins ofomilitont nonviolence. New York: Norton.

Erikson,E. (ed.)(1978).Adulthood. NewYork: Norton.Fischer,K.W. (1980a).Learningandproblemsolvingas

the developmentof organizedbehavior. Journal ofStructural Learning, 6, 253—267.

Fischer,K.W. (l9SOb).A theoryofcognitivedevelopment:l’he control and constructionof hierarchiesof skills.Psychologicol Review, 6,477—531.

Fischer,K.W., & Ayoub, C. (1996).Analyzing develop-tnent of working models of close relationships:Illustration with a caseof vulnerabilityandviolence. InG.G. Noam & K.W. Fischer(eds).Developnsent andvul,merohility in close relationships (pp. 173—199).Hillsdale,NI: Erlbaum.

Fisclser,K.W., & Bidcll, T.R. (1998). Dynamic develop-ment of psychologicalstructuresin action andthought.In F.M. Lerncr(cd),Handbookofchildpsychology: VoL1. Theoretical models qf Ituosan dei.’elopmnent (5th edn,pp.467—

561).New York: Wiley.

Fischer, K.W., & Granott, N. (1995). Beyond one-dimemssional change: Parallel, concurrent, socmallydistributed processestn learning and dcvelopment.i-Iuoton Developnsent, 38, .302—314.

Fischer.K.W., HamscLH.H.,& Rtmssell,S.L.. (1984).Thedevel-opment ofabstraetsonsin adolescenceandadulthood. InM. Commons,F’.A. Riclsards,& C. Annon(eds),Beyondf4rttmol operations (pp. 43”~73). New York: Praeger.

Fischer.K.W.. & Kennedy,B. (1997). ‘Fools for analyzingthemany shapesof development:The caseof self—tn-

relationshipsin Korea. In K.A. Rcnninger& E. Amsel(eds),Processes of developotent (pp. 117—152).Mahwah,NI: Eribaum.

Fischer, K.W., Kenny, S.L., & Pipp, S.L. (1990). Howcognitive processesand environmental conditionsorganizediscontinuitiesin the developmentof abstrac-tions. In C.N. Alexander& E.J. Langer(eds),higherstages of hutnan development: Perspectives on adultgrowth, (pp. 162—IS?).New York: Oxford UniversityPress.

Fischer,K.W., & Pruyne.E. (2002).Reflectivethinkinginadulthood:Development,variation, andconsolidation.In I. Demick(ed.) Handbookofadult developnsent.

Fischer,K.W., & Rose,S.P,(1994).Dynamicdevelopmentofcoordinationofcomponentsin brain andbehavior:Aframeworkfor theory andresearch.In 0. Dawsoa&K.W. Fischer(eds),human behavior and the developingbrain (pp. 3—66) New York: Guilford,

Fischer,K.W., & Rose,S.P,(1996).Dynamicgrowthcyclesof brain and cognitive developsnent.In R. Thatcher,G.R. Lyon,J. Rumsey,& N.Kinasnegor(edt), Develop-mental neuroinmnging: Mopping the development of brainand behavior (pp. 263—279).Ncw York: Academic.

Fischer,K,W., & Yan, Z. (2002).Darwin’sconstnmctionoftlse theoryof evolution: Microdevelopmentof explana-tionsof specmesvariationandchange.In N. Granott&I. Parziale(eds),Mierodevelopmnent: Transition processesin des’elopnmettt and learning (pp. 294-318).Cambridge:CambridgeUniversityPress.

Gardner,H. (1983).Framesofnund: The theoryofmnultipleintelligences. New’ York: Basic.

Gillsgan, C. (1982) In a d~fJeremmt voice: Psychological

theory and women’s development. Cambridge,MA:ilarvardUniversstyPress.

Gottlieb, G., Wahlsten,D., & Licklitcr, R. (1998). Thesignificance of biology for human development:Adevelopmental psychobiological systems view’. InW. Damon(seriesed.) & F.M. Lerner (ed.),handbookofchildpsychology:Vol. 1. Theoreticalnsodelsofhtomtandevelopment (Stlm edn,pp. 233—274).NewYork: Wiley.

Granott,N. (I 993a).Microdevelopmentofcoconstructionof knowledgeduringps’oblemsolving: Puzzledminds,weird creaturesand w’uggles. Unpubltsheddoctoraldissertation, MIT. Cambridge, MA. Disse,’totion.4hstroeis Inter,tatiommal. 54 (lOB), 5409.

Granott, N. (1993b). Patterns of interaction in thecoconstructionof knowledge: Separateminds, jointeffort, and weird creatures.In R.H. Wozniak& K.W.Fsscher(edt), Developnsent in context: Acting antithinking in specific e,tviro,tme,tis (pp. 183—207).Ilillsdale, NI: Erlbatmm.

Granott, N. (1998). We learn, thes’efore ‘ye dcvelop:I.earningversusdevelopment.‘ ordevelopimsgIeamtng?In M. C. Smith & T. Pourchot(eds),.4dult learning omit!developntemtt: Per.c’pectivesfroos edticationolpsvc’hologs’(pp. 15-34).Mahwah,NJ: Erlbatmm.

(iranott,N. (2002).Howmierodevelopmentct’eatesmacro—development:Reiteratedseqttences.backwardtranss-tions,andthe zoneofcurrentdevelopment.In N. (‘iranost& J. Parzsale tedsI, Mies’odevelopntemtt: Transition

ADULTCOGNITIVE DEVELOPMENT:DTNAMICE [N THEDEVELOPMENTALWEB 515

processes in developmmtemt t ond leam-tm ing (Chapter8).Cambridge:C;msnhrmdgcUsssversstyPress.

Granotl,N., Fischem’.K.W,. & Parziale,J. (2002).Bridgingto theunknown:A transitionmechanmsmin leansingandproblem-solving.In N. Granott & 3. Parziale (cds),zt’tierodes”elopmnent: Tm’onsition pt-oc’esses in developsmteot

ond (corning (pp. l31--156). Cambridge:CambridgeUniversityPress,

Granott,N., & Parziale,I. (eds)(2002). Miem-odevelopment:Transition processesitt developnten t and learning.Cambridge:CambrtdgeUniversity Press,

Grtmbcr, HE. (1973). Courage and cognitivegrowth inchildren andscientists.In M. Sehwcbel& I. Raph(eds),Piaget in the classroons, New York: Basic,

Gmber, lIE. (1981).Do,-sm’in ott nton (2ndedn). Chicago:UniversityofChicagoPress.

Hatter,S.. & Monsour.A. (1992).Developmentalanalysisofconflict causedby opposingattributesin theadolescentself-portrait.Developnse,mtol Psychology, 28, 251--260.

Horn, IL. (1982). The aging of human abmlities, InB.B. Wolman ~ed.), Handbook of developnmentnlpsychology. EnglewoodCliffs, NJ: Prentice-Hall,

Horn,J.L., & Cattell,RB. (1967).Age differencesin fluidandcrystallizedintelligence.Acto Psychologico, 26,107—129,

Inhelder.B., & Piaget, 1. (1958). The growth of logieolthinking from childhood to adolescence (trans.A.P.S.Seagrim)Ness’ York: Basic. Originallypublished1955),

Keegan, FT. (1989). Flow’ CharlesDarwin becameapsychologist.In D.B. Wallace & HF. Gruber (eds),Creative people at work: Twelve cognitive case studies(pp. 107—l25). New York: Oxford UssiversityPress,

Kegan,R. (1982). The evolvingself Problemandprocessin human development.Cambridge, MA: HarvardUniversityPress,

Kcgan,F. (1994). In over our heads: The mental demandsof modern life. Cambridge,MA: Harvard UniversityPress,

Kennedy. B.P. (1994). The development of self-understandingin adolescentsin Korea. Unpublisheddoctoraldissertation,HarvardUniversity, Cambridge,MA. Dissertotion Abstracts international: Section B. TheSciences & Emrgineerimtg, 55 (7—8), 3036,

King, P.M., & Kitchener, KS. (1994). Developingreflective judgment: Understanding and promotingintellectual growth and critical thinking in adolescentsand adults, SanFrancisco:lossey-Bass.

KitcheneLKS., & Fischer,K,W, (1990).A skill approachto the developmentof reflective thinking. In D. Kuhn(ed), Developmental perspectives ott teaching andlearning thinking skills. Contributions to HumanDevelopment, 21(4). 48—62.Basel,Switzerland:Karger.

Kmtchener,KS., Lynch,CL., Fischer,K.W., & Wood,P.K.(I 993), Developmentalrangeof retlectis’ejudgtnent:Theeffect of contextualsupportamsd practiceon develop-mentalstage.DeveloposemttalPsychology.29,893—906,

Knhlberg, L. (1969). Stageandsequence:The cognitivedevelopmentalapproachto socialization,In D.A, Goslin(ed),Handbook of.tot’iohzoiioms ilteoty amid resewv’h (pp.347—480).Chicago:Rand.McNally.

Kohlberg,L. (1984).Moral stagesandmoralization:Theeognitive-dcvclopmcntalapproach.In L. Kohlbergfed.),The psn:hologv of mo,vtl developotent: The ‘motto-c andsolid/tv of ,norol stages (pp. 170—205). San Francisco:Hamper& Row,

Kuhn, D., Garcia-Mila, M., Zohar, A.. & Andersen,C.(I 995? Strategies of knems’ledge acquisition. Mono-graphs of the Society ,for Research in ChildDeveloposemtt,Vol. 60, no, 4, serial245).

Lakoff G., & Johnson.M. (1980).Metophot-z we live by.

Chicago:UniversityofChicagoPress.Labouvie-Vief, U., De-Voe, M., & Bulka. D. (1998).

Speakingaboutfeelings:Conceptionsofemotionacrossthelife spams.In L.M. Powell & TA. Salthouse(eds)(1998). Essential papers on the psychology of aging(pp.

48I—S 10). New’ York: New York UniversityPress.

Levmnson.DI. (1978). The seasons of a mon’s ltji. New’York: Knopf.

Loevinger,J. (1976). Ego developntent: Concepts andtheories, SanFrancmsco:Iossey-Bass.

Marcia,I.E. (1980).Identity ip adolescence,In J. Adelson(ed),Handbook of adolescent psychology. New York:Wiley.

Marcia,I.E. (1994).The empiricalstudyofego identity. InHA. Bosma, T.L.G. Graafsma,lID. Grotevant, &DI. de Levita (eds), Identity and development: An

interdisciplinary approach (pp. 67—80). ‘rhousandOaks,CA: Sage.

Matteson,DR. (1977). Explorationandcommitment:Sexdifferencesandmethodologicalproblemsin theuseofmdentmty status categories.Journal of Tooth andAdolescence, 6,353—374,

Nesmgarten,Bk. (ed.) (1968). Mtddle age and egtvg.

Chicago:UniversityofChicagoPress,Noam, G.G., Powers,SI,, Kilkenny, F., & Beedy, J.

(1990).Theinterpersonalselfin life-spandevelopmentalperspective:l’heoty. measurement,and longitudinalcase analyses. In PB, Baltes, DL. Feathennan.&R.M. Lerner(ed),L~fespan development and behavior(Vol. 10, pp. 59—104). Hillsdale,NJ: Erlbaum.

Pav’siale.J. (1997). Microdevelopmentduringanactivitybased sciencelesson. Dissertation Abstracts Inter-notional: Sectio,t B. The Sciences & Engineering, 58(5—B), 2723.

Perry,W.G, Ir (1970).Fornts of intellectual andethicaldevelopment in the college years. New York: ilolt,Rinehart,& Winston.

Phinney,iS. (1989).Stagesof ethnicidentitydevelopmentin minority groupadolescents,Journalof’Earki Adoles-cence, 9,34—49,

Piaget,I. (1975).L’bquilibrationdesstructurescognitives:Problhmecentraldu dbveloppement(Equilibrationofcognitive structures:Central problemof development).Etudes d’Epistdntoiogie Gdmtétique, 33.

Piaget,1. (1983).Piaget’stheory, In P.11,Mussen(seriesed.)& W. Kessen(ed.) blandhook of child pysehology:

Vol. 1. History, theory, amid m,methods (pp. 103—126),NewYork: Wilcy.

Rest,J.. Narvaez,D.. Bebeau.M.J,, & Thoma.5.). (1999).Postconventio,mol mo,’ol thinking. Mahwah,NJ: Erlbaum.

516 DEVELOPMENTIN ADULTHOOD

Roberts, RI. Jr (1981). Errors and the assessmentofcognitivedevelopment.In K.W. Fischer(ed),(ognitive

development: New directions j’or child development(Vol. 12,pp. 69—78). SanFrancisco:Jossey-Bass.

Rose, S.P.,& Fischer,K.W. (l998). Models andrulers itsdynamicaldevelopment.British Jou~nal of Develop-nsental Psychology, 16 (Pt I), 123—131.

Salthouse,TA. (1984). Effectsof ageandskill in typing.Journal of ExperimentalPsychology:General, 113,345—371,

Salthouse,TA. (1992). Mechanisms of age—cognitionrelations in adulthood. Hillsdale,NJ: Erlbaum,

Salthouse,TA., Hambriek, D.Z., & McGuthmy, K.E.(1998). Shared age-relatedinfluenceson cognitiveand noncognitivevariables,Psychology & Aging, 13,486—500,

Schacter,DL. (1999). Thesevensinsofmemory:Insightsfrom psychologyandcngnstmveneuroscience,AmericanPsychologist, 54, l82—203.

Schaie,K,W. (1983).Consistencyand changesin cognitiveftmnctioningoftheyoung-oldandold-old, In M. Schmidt-Scherser(eds),Aging in the eighties and beyond. NewYork: Springer

Schaie, K.W. (1996). Intellectual development inodulthood: The Seattle longitudinal study. New York:CambridgeUniversityPress.

Siegler, R.S. (2002). Microgenetic studies of self-explanation.In N. Granott& I. Parziale(eds),Micro-developtnent: Transition processes in development andlearning (Chapter1). Cambridge:CambridgeUniversityPress.

Simonton,D.K. (1991).Creativeproductivitythroughtheadultyears.Generations, 15, 13-16.

Sinsomiton,O.K. (2000).Ct’eativedevelopnsentasacquiredexpertise: Theoretical issues and an empirical test,Developmental Review, 20, 283—318.

Sinnott, ID. (1998). The development of logic inadulthood: Postformal thought and its applications. NewYork: Plenum,

Smith,MC., & Pourchot,T. (ed.)(1998),Adult learning

and development: Perspectives front educationalpsychology. Mahwah,NJ: Frlbaum,

Stes’nberg,F. (1985). Beyond JQ: A ttiarehie theory ofintellt~’eaee, NewYork: CambridgeUniversityPress.

Sternbcrg,F.J. (ed.)(1990). Wisdom: Its nature, origins,and development. New York: CansbridgeUniversityPress.

Tennant,M. (1997).Psychology and adult learn ing (2ndedn).New York: Foutledge.

Thatcher, F.W. (1994) Cyclic cortical reorganization:Originsoflsumasscognmtivedevelopment,In U.Daw’son& K.W. Fischer (eds), Human behavior and the

developing b,’ain (pp. 232—266).New York: Guilford.Thelen. E,. & Smith. L.B. (1994). A dynamic systent.t

approach to the developnsent of cognition and action,Cambridge,MA: MIT Press.

Turkle, S.(1995).Life on thescreen:Identity in theageofthe intem-net, New York: Simon& Schuster,

Vaillant, G.E. (1977).Adaptation to ltjè. Boston:Little,Brown,

Vaillant, G.E. (ed.) (1986), Enspirieoi studies of egomechanisms of defense. Washington.DC: AmericanPsychiatricPress,

Valsiner,1, (1991). Constructionof thesnental:From the‘cognitive revolution’ to the study of development.Theomy & Psychology, I, 477—494.

Valsiner,J. (1996). Co-constrnctionismand development:A soeio-historictraditton,Anuario de Psicologia, 69,63—82.

Valtiner, J,,& vanderVeer,F. (1988).Onthesocialnatureof human cognition: An analysis of the sharedintellectual roots of George I’Ierbert Mead and LevVygotsky, Journal for the Theory of Social Behaviour,IS, 117—136.

van der Maas,H., & Molenaar,P. (1992).A catastrophe-theoreticalapproachto cognitivedevelopment.Psycho-logical Review, 99, 395—417,

van Geert,P. (1994).Dynamic systems of development:Change hetvr’een complesity’ and chaos, London:HarvesterWheatsheaf,

van Geert,P. (2002).Developmentaldynamics,intentionalaction,andffizzy sets.In N. Granott& J. Parziale(eds).Mierodevelopment: Transition processes in developmmtentand learning (chapter 12). Cambridge: CambridgeUniversityPress,

Vygotsky, L.S. (1978). Mind in society: The developmentof higher psychological processes. Cambridge,MA:HarvardUniversityPress,

Watson,MW., & Fischer,K.W. (1980).Developmentofsocialrolesin elic/tedandspontaneousbehavior duringthepre-schoolyears.Developmental Psychology, 16,484-494,

Wechsler,D. (1972). ‘Hold’ and ‘don’t hold’ test, InSM. Chown (ed.) Human aging. New York: Penguin.

Werner,H, (1948). Comparative psychology of mentaldevelopntent. New York: ScienceEditions,

Wood, D., Brnner,IS.. & Foss,G. (1976). The role oftutoring in problemsolving. Child Psychology’ andPsychiatry, l7, 89—100.

Yan, Z. (1998).Meastiringmicrodcvelopmentof under-standingtheVMS-SAS structum’e:A developmentalscalepilot. Unpublishedqualifying paper.llar’,’ard GraduateSchool of Education,Cambridge,MA.

Yan, Z. (2000).Dynamicanalysisof microdevclopmentinlearning a computerprogram. Unpublisheddoctoraldissertation,Harvard GraduateSchool of Education,Cambridge,MA.

Yan, Z., & Fischer,K.W. (2002).Always underconstruc-tion: Dynamicvariationsin adultcognitivedevelopment.Human Development, 45, 14 I—I 60.

Zeeman, E.C. (1976). Catastrophetheory. ScientificAmerican, 234(4),65—83.