5 B rain Changes Underlying the Development of Cognitive Control ...

Y

5 BrainChangesUnderlyingthe

DevelopmentofCognitive

ControlandReasoning

silviaa.bunge,allysonp.mackey,andkirstiej.whitaker

abstract Whatpreciselyischangingovertimeinachild’sbrainleadingtoimprovedcontroloverhisorherthoughtsandbehavior?Thischapterinvestigatesneuralmechanismsthatdevelopthroughchildhood and adolescence and underlie changes in workingmemory,cognitivecontrol,andreasoning.Theeffectsofageandexperienceonspecificcognitivefunctionsarediscussedwithrespecttofunctionalbrainimagingstudies,highlightingtheimportanceofinteractions between prefrontal and parietal cortices in cognitivecontrolandhigh-levelcognition.

What precisely is changing over time in a child’s brain,leading to improved control over his or her thoughts andbehavior? Throughout childhood and adolescence, weimproveatorganizingour thoughts,workingtoward long-term goals, ignoring irrelevant information that could dis-tractusfromthesegoals,andcontrollingourimpulses—inotherwords,weexhibitimprovementsinexecutive function orcognitive control (Diamond,2002;Zelazo,Craik,etal.,2004;Casey, Tottenham, et al., 2005). By the same token, weexhibitincreasedfacilityoverthisagerangeintacklingnovelproblems and reasoning about the world—a capacityreferredtoasfluidreasoning(Cattell,1971).Boththecapac-itytoconsciouslycontrolourthoughtsandactionsandthecapacity toreasoneffectivelyrelyonworking memory,or theability to keep relevant information in mind as needed tocarryoutanimmediategoal.

Neuroscientific research is being conducted to betterunderstandthechangesinbrainstructureandfunctionthatunderlie improved cognitive control and fluid reasoningduringchildandadolescentdevelopment.Morespecifically,researchers seek todeterminehow theneuralmechanismsunderlyingspecificcognitivefunctionschangewithage,howtheydifferamongindividuals,andhowtheyareaffectedbyexperience.

Webeginthischapterwithabriefsummaryofchangesinbrainstructure,focusingprimarilyonprefrontalandparietalcortices,thebrainregionsthathavebeenmostcloselyassoci-ated with goal-directed behavior. We then provide anoverview of functional brain imaging studies focusing onage-relatedchangesinworkingmemory,cognitivecontrol,andfluidreasoningoverchildhoodandadolescence.Becauseworking memory and cognitive control development havebeen discussed extensively elsewhere (Munakata, Casey,etal.,2004;Rubia&Smith,2004;Caseyetal.,2005;Bunge&Wright, 2007), a relatively greater emphasis is placed onrecentstudiesfocusingonthedevelopmentoffluidreasoning.

Structural brain development

Thebrainundergoesmajorstructuralandfunctionalchangesoverchildhoodandadolescence thatmay, inpart,explainchangesinbehaviorandcognition.Asexplainedinchapter2,byKostovicandJudaš,rapidchangesoccurattheneu-ronallevelintheprefrontalcortex(PFC)inthefirstfewyearsoflife,followedbyslower,protractedchangesthroughado-lescence(Petanjek,Judas,etal.,2008).Whilebrainchangesat the cellular level can only be examined in postmortembrain tissue, advances in neuroimaging techniques havemade itpossible tostudygrossanatomicaldevelopment invivo.Structuralmagneticresonanceimaging(MRI)methodsmakeitpossibletoquantifyage-relatedchangesincorticalthickness (Sowell,Peterson, et al., 2007), in the volumeofspecificbrainstructures(Gogtay,Nugent,etal.,2006),andin the thickness and coherence of white matter tractsconnecting distant brain regions to one another (Giedd,Blumenthal,etal.,1999;Klingberg,Vaidya,etal.,1999).

Cortical thickness follows an inverted U-shaped patternoverdevelopment.Up tomiddlechildhood (ages8 to12),increasedthicknessofthegraymatteratthesurfaceofthebrain reflects increased density of neurons and dendrites.Thereafter, decreased gray matter thickness reflects thepruningofexcessdendritesandneurons,aswellasincreasedmyelinationofaxonalprojectionstotheseneurons(Giedd,2004).

bunge,mackey,&whitaker:brainchanges&thedevelopmentofcognitivecontrol 73

silvia a. bunge Helen Wills Neuroscience Institute andDepartment of Psychology, University of California at Berkeley,Berkeley,Californiaallyson p. mackey, and kirstie j. whitaker Helen WillsNeuroscience Institute, University of California at Berkeley,Berkeley,California

2

1

32

Gazzaniga_05_Ch05.indd 73 2/3/2009 8:37:57 PM

Y

74 developmentandevolution

The developmental trajectory of changes in corticalthickness varies across brain structures. In PFC andparietal cortex, gray matter volume peaks around age10–12(Giedd,2004).Thereafter,graymatterlossoccursatdifferent rates indifferent subregionsof thePFC,and it isconsideredoneindexofthetimecourseofmaturationofaregion(Sowell,Peterson,etal.,2003).WithinthePFC,graymatter reduction is completed earliest in the orbitofrontalcortex, followed by the ventrolateral PFC (VLPFC) andthen by the dorsolateral PFC (DLPFC) and rostrolateralPFC (RLPFC) (Gogtay, Giedd, et al., 2004; O’Donnell,Noseworthy, et al., 2005). Differences in maturationaltime course between prefrontal subregions could helpaccountfordifferencesintherateofdevelopmentofdistinctcognitivecontrolprocesses(Bunge&Zelazo,2006;Crone,Wendelken,etal.,2006).

DevelopmentalchangesininterregionalconnectivityhavebeenstudiedwithanMRI-basedmethodknownasdiffusiontensorimaging(DTI).ResearchusingDTIhasshownthatstrengtheningoffrontal-parietalnetworksisassociatedwithimprovedperformanceonworkingmemory tasks (Olesen,Nagy,etal.,2003;Nagy,Westerberg,etal.,2004).Anage-related increase in frontostriatal tract coherence has alsobeenassociatedwithmoreefficientrecruitmentofcognitivecontrol(Liston,Watts,etal.,2006).

Insummary,bothcorticalpruningwithinprefrontalandparietalregionsandincreasedneuronalconnectivitywithinandbetweentheseandotherregionsare likely tounderlieimprovements in cognitive control and fluid reasoningduringdevelopment.Therelationshipsbetweenbehavioralimprovements and changes in brain structure and brainfunction have been explored in recent studies of workingmemory,asdescribedinthenextsection.

Working memory development

Workingmemoryisthebrain’s“mentalblackboard,”allow-inginformation—eithersensoryinputsormemories—tobeheldinmindandmanipulated(Miller,Galanter,etal.,1960;Goldman-Rakic,1992).Consideredacentralcomponentofhuman cognition, the maturation of working memory iscritical for the development of language comprehension,mental calculation, cognitive control, and fluid reasoning.Although children as young as 5 years do not differ fromadultsinsensorimotortasks,performanceontasksthatrelyontheretentionandmanipulationof information, suchasspatialmemoryspanandTowerofLondon,improvesoverchildhood (Luciana & Nelson, 1998). Children’s perfor-manceiscriticallymoderatedbytaskdifficulty:theiraccu-racy rapidly declines as the demands of the task becomemore rigorous and they make more errors. As workingmemory improves, children likewise improve on tests ofcognitivecontrolandfluidreasoning.

Working memory for different types of information ismediated by interactions between domain-specific brainregionsandregionsinPFCandparietalcortex(D’Esposito,2007).Itistheintegrationandrefinementoftheseworkingmemory circuits that underlies age improvements in themaintenance and manipulation of mental representations.Inthenextsubsection,weprovideabriefoverviewoffMRIstudiesexaminingage-relatedchangesinworkingmemory.

Visuospatial Working Memory Most fMRI studiesonworkingmemorydevelopmenthavefocusedontheabilitytokeepinmindaseriesofspatiallocations(Casey,Cohen,etal.,1995;Thomas,King,etal.,1999;Klingberg,Forssberg,et al., 2002; Kwon, Reiss, et al., 2002; Scherf, Sweeney,et al., 2006). The superior frontal sulcus (SFS) and theintraparietalsulcus(IPS),whichhavebeenstronglyimplicatedin adult visuospatial working memory, are increasinglyengagedthroughoutchildhood(Klingberg,Forssberg,etal.,2002;Kwonetal.).Acrosschildren, the levelof fractionalanisotropy in the frontoparietal white matter surroundingthe SFS and IPS is positively correlated with visuospatialworking memory scores (Nagy et al., 2004). Further, thecoherenceofthesewhitemattertractsinthelefthemisphereisgreateramongchildren(age8–18years)whoexhibitthegreatestactivationintheseregions(Olesenetal.,2003).Thusthebrainnetworkunderlyingeffectivevisuospatialworkingmemoryisstrengthenedoverdevelopment.

Atamicroscopiclevel,theincreasedengagementofSFSandIPSduringaBOLDfMRIvisuospatialworkingmemorytaskcouldbedependentononeormorecellularchanges:neuronal pruning, increased myelination, and/or thestrengthening of synaptic connections within or betweenbrainregions.Klingbergandcolleaguesusedcomputationalmethodstodeterminewhichoftheseprocessesarelikelytosupport the development of visuospatial working memory(Edin,Macoveanu,etal.,2007).TheircomputationalmodelofBOLDactivationfoundthatstrengthenedsynapticcon-nectivity within and between brain regions was the mostlikely candidate for increase in activation in these regionsbetweenchildhoodandadulthood.

Justascoreworkingmemorynetworksarestrengthenedover childhoodandadolescence, supportingnetworks thatarenotusedbyadults forworkingmemoryareweakenedoverthisagerange.Usingaspatialworkingmemorypara-digm involving saccadic eye movements, Luna and col-leagues showed that increased recruitment of core regionsinDLPFCin the lefthemisphereandparietal regionswasaccompanied by a weakening and eventual dismissal ofa childhood compensatory circuit involving ventromedialPFC (Scherf et al., 2006).Aqualitative shiftwasobservedin comparing children aged 10–13 and adolescents aged14–17.Comparingadolescentstoadults,thechangesweremore quantitative, evincing refinement of the visuospatial

3

4


Y


workingmemorynetwork.Thismovementaway from thechildhood circuit to the more mature adult network is acommon theme in developmental cognitive neuroscience,anditisfurtherdiscussedinthischapter’ssectiononcogni-tivecontroldevelopment.

NonspatialWorkingMemory InadditiontothefMRIstudies of visuospatial working memory development, wewouldliketohighlightastudyonthedevelopmentofnonspatialworkingmemory,inwhichchildrenaged8–12,adolescentsaged13–17,andyoungadultswereaskedtorememberaseriesofthreenameableobjects(figure5.1A;Croneetal.,2006).Weconsiderfirstthepuremaintenanceconditionofthisstudy,inwhichparticipantswereaskedtoverballyrehearsetheitemsinthe order in which they were presented. All three groupsengagedhighlyoverlappingsetsofbrainregions,indicatingthatthecoreworkingmemorynetworkwasalreadyinplacebymiddlechildhood.However,therewasapositivecorrelationacross participants between task accuracy and level of

activation in left ventrolateral PFC (VLPFC), bilateralDLPFC,andbilateralsuperiorparietalcortex.

Manipulation of Information in WorkingMemory This study of nonspatial working memory(Croneetal.,2006)alsoincludedamanipulationcondition,inwhichparticipantswereaskedtoreversetheorderoftheitems in their head. Children aged 8–12 were dispropor-tionatelyimpairedrelativetoadolescentsandadultsonthismanipulation condition relative to the pure maintenancecondition.Further,children failed toengagerightDLPFCand bilateral superior parietal cortex, regions linked withworkingmemorymanipulation,forthispurpose.Aqualitativeshiftinthecircuitryunderlyingmanipulationwasobservedfrommiddlechildhoodonward, such thatadolescentsandadultsengagedanadditionalmechanismrelativetochildrenaged8–12.Time-seriescorrelationalanalysesshowedthat,for adults, right DLPFC was functionally correlated withbilateralparietalandpremotorcorticesduringmanipulation.

Figure 5.1 Development of nonspatial working memory andworkingmemorymanipulation.(A)Subjectswereaskedtoremem-berthreenameableobjects,presentedfor750mseachandsepa-ratedbya250-msfixationcross.Afterthelastobjecttheinstruction“forward”or“backward”directed theparticipant toeithermen-tallyrehearseorreordertheseobjectsduringthe6,000-msdelay.Finally a probe object was presented and participants indicatedwithabuttonpresswhetheritwasfirst,second,orthirdobjectinthe memorized sequence. Forward trials required pure mainte-

nance,whereasbackwardtrialsrequiredmanipulationinadditiontomaintenance.(B)Group-averagedtimecoursesforactivationinthe right DLPFC during the delay period show that adults andadolescentsrecruitedthisregionmorestronglyduringthehardermanipulationstrials,whereaschildrenshowedthesameactivationinDLPFCforboth“forward”and“backward” tasks. (ReprintedwithpermissionfromCrone,Wendelken,etal.,2006,copyright2006,NationalAcademyofSciences,USA.)


Y


In children, by contrast, right DLPFC activation duringmanipulation was correlated with regions that have notbeen associated previously with this function (unpublishedanalyses).Thusthebrainnetworkunderlyingmanipulationin adults was not yet engaged by children aged 8–12.Importantly,itisnotthecasethatchildrenfailedtoengagethese brain regions during task performance. Indeed,children engaged DLPFC and parietal cortex at encodingandretrievalofitemsinworkingmemory;theysimplyfailedtosufficientlyengagethecircuitrythatsupportsmanipulationat the time when it was required to reverse the order ofobjectsinworkingmemory(figure5.1B).

TheseinvestigationshaveshownhowthedevelopmentofPFCandparietalcortex,aswellastheconnectionsbetweenthem, contributes to increased ability to maintain andmanipulate information online. In turn, an increase inworkingmemorycapacitycontributestoimprovementsinavariety of cognitive functions, including cognitive controlandfluidreasoning.

Cognitive control development

One of the most obvious ways in which children maturebehaviorallyisthattheybecomeincreasinglyabletoignoreirrelevant and distracting information and control theirimpulses while working toward specific goals. The termsexecutive function andcognitive controlrefertomentalprocessesassociatedwiththecontrolofthoughtandaction.Thusfar,developmentalresearchoncognitivecontrolhasbeencon-cernedwithconscious,deliberateformsofcontrol.Putativecognitivecontrolfunctionsarelistedinbox5.1.

Cognitive Control Development: Changes inOne or More Neural Circuits? A key question indevelopmental research has been whether age-relatedchanges in cognitive control are associated with the

developmentofasinglemechanism,suchasthecapacitytostoreorprocessinformation(Case,1992;Dempster,1993),orwitha setofmechanisms (Welsh,Butters, et al., 1991).Behavioral studies suggest that some of these abilitiesmay mature at different rates. For example, the ability toinhibit amotoric responsematures earlier than theabilityto inhibit a response when the task additionally requiresselectiveattention (vandenWildenberg&vanderMolen,2004). Likewise, the ability to switch between task rulesdevelops earlier than the ability to keep a difficult ruleonline (Crone, Wendelken, et al., 2004). Recent advancesusing structural equation modeling indicate that workingmemory, task switching, and response inhibition areseparable latent constructs with distinct developmentaltrajectories (Brocki & Bohlin, 2004; Huizinga, Dolan,et al., 2006).Thusbehavioral researchprovideshints thatdifferent cognitive control functions may have separableneurodevelopmentaltrajectories.

Inarecentstudyofcognitivenetworkdevelopment,Fair,Dosenbach, and colleagues (2007) show that, initially, thestrongest connections between frontoparietal gray matterare anatomically close together. As these regions mature,however, the connections become more functionally rele-vantandreachfurtherafield,presumablytoengagethemosteffectivenetworkforcognitivecontrol(figure5.2).

The protracted myelination of the white matter tractsconnectingtheregionsnecessaryforcognitivecontroloverchildhood and adolescence (Spear, 2007) may explain thecompensatory network required for children to completethesetasks.Iflong-rangeprojectionsarenotsufficientlyinsu-latedbythemyelinsheath,theywillbeunabletocommu-nicatewiththefunctionallyrelevantnetworksutilizedbytheadultbrain.Eventually,duringadolescence,themyelinationissufficienttoallowatransitionfromthelocal,compensa-torymechanismtoamorediffuse,adult,effectivesystem.

Inthefollowingsections,wehighlightafewofthemanybrain-imaging studies that have examined neurodevelop-mentalchangesincognitivecontrol.

Response selection and inhibition Asnotedpreviously,improve-mentsinworkingmemorymanipulation—aformofcogni-tivecontrol—areassociatedwithanincreaseinlateralPFCactivationwithage(Croneetal.,2004).Forsomecognitivecontroltasks,however,maturationisinsteadassociatedwithadecreaseinlateralPFCrecruitment.Forexample,alargestudyinvolvingparticipantsbetween8and27yearsofageby Luna and colleagues (Velanova, Wheeler, et al., 2008)examined functionalmaturational changes associatedwithperformanceonanantisaccadetask.Inthisresponseinhibi-tion task, participants must move their eyes away from avisualstimulusthatappearssuddenlyonthescreen,resistingthe urge to look toward it. The researchers observed anage-related shift away from reliance on DLPFC, toward

Box5.1 CognitiveControlFunctions

1. Selectively attending to relevant information (selective attention) and ignoring distracting stimuli or thoughts(interference suppression/resolution)

2. Selecting between competing response tendencies(response selection) andinhibitinginappropriateresponseten-dencies(response inhibition)

3. Using contextual information to identify currentlyrelevantinformationandappropriateresponses(rule/task-set representation)

4. Reorganizing information currently held in workingmemory(manipulation, updating)

5.Flexibly switchingbetween tasksandperforming twotasksconcurrently(task-switching, dual task performance)

6.Monitoringone’sownactionsandtheconsequencesoftheseactions(performance monitoring, error/feedback processing)

5


Y


posteriorparietalregions(figure5.2).Consistentwithotherwork, this finding indicates a shift away from childhoodcompensatorymechanismstowardthemoreeffectiveadultnetworks.

Inadditiontothisantisaccadestudy,anumberofotherbrain-imagingstudieshavefocusedonage-relatedimprove-ments in the ability to select between competing responsechoices and inhibit inappropriate response tendencies (seeMunakata et al., 2004, and Bunge & Wright, 2007, forreviews).Thesestudieshavemadeuseofavarietyofpara-digms,includingthewell-knownStroop(Adleman,Menon,et al., 2002; Schroeter, Zysset, et al., 2004; Marsh, Zhu,

etal.,2006),go-no-go(Rubia,Smith,etal.,inpress;Rubia,Russell,etal.,2001;Bunge,Dudukovic,etal.,2002;Durston,Thomas,etal.,2002;Tamm,Menon,etal.,2002;Booth,Burman, et al., 2003; Lamm, Zelazo, et al., 2006), andflankertasks(Bungeetal.;Lammetal.;Rubia,Smith,etal.,2006). In task-switching paradigms, the currently relevanttask rule changes without warning, and it is necessary tosuppress the response to the previous rule and also toretrieve the new rule from memory (Crone et al., 2004).ThesestudiesindicatethatoverlappingbutdistinctcircuitsinvolvingregionsofPFC,parietalcortex,andbasalgangliaare involved in various cognitive control tasks (Rubia,Russell,etal.).

Performance monitoring Lunaandcolleaguesusedtheantisac-cadetasktoexaminenotonlyatthedevelopmentofresponseinhibition,butalsothedevelopmentofperformancemoni-toring(Velanovaetal.,2008).Thedorsalanteriorcingulatecortex(dACC),knowntoplayacentralroleinperformancemonitoring (Ford, Goltz, et al., 2005; Polli, Barton, et al.,2005),wasmore stronglyengagedonerror trials inadultsthan inchildrenoradolescents (figure5.3). Increasedper-formancemonitoring,supportedbydACC,islikelytocon-tributetotheobservedimprovementsininhibitorycognitivecontrolwithage.

In summary, cognitivecontrol is considered tocompro-mise a variety of different putative processes, and a fewstudieshavemadeattemptstocomparethedevelopmentaltimecourseofspecificcontrolprocesses (e.g.,Bungeetal.,2002;Croneetal.,2006;Rubiaetal.,2006).However, itremainstobeseenifthesedifferentbehavioralabilitiesarein fact different underlying neural substrates and if theydevelop over separable trajectories. Further work must beundertaken to determine the relative independence andinteractions of the many cognitive control capabilitiesthroughoutdevelopment.

Fluid reasoning development

Fluid reasoning is the capacity to think logically andsolve problems in novel situations (Cattell, 1971). Theconceptoffluidreasoning is integral to theoriesofhumanintelligence(Horn,1967;Cattell,1987;Horn,1988;Carroll,1997; McArdle, 1998; Gray, Chabris, et al., 2003). Com-pared to crystallized, or knowledge-based, abilities it isthought to have a stronger neurobiological and geneticcomponent, leading to the belief that it is less dependenton experience.However, some evidence suggests that it isindeed sensitive to cultural and environmental influences(Flynn,2007).

Thedevelopmentofreasoningabilityiscentraltounder-standingcognitivedevelopmentasawhole,becauseitservesas scaffolding for many other cognitive functions (Cattell,

Figure5.2 Developmentofdistinctcognitivecontrolnetworksthroughchildhoodandadolescence.Regionspreviouslyidentifiedas pertaining to putative task control were analyzed for pairwisetemporalBOLDcorrelationsin(A)childrenand(B)adults.Right-sideROIsaredisplayedon the rightofeachgraphandanteriorROIsat the topofeachgraph.Whereasadultsdemonstrate twoseparate control networks, children show a connection betweenthem. Their networks are connected by a bridge between theanterior PFC and DLPFC, and the dACC and medial superiorfrontal cortexwere incorporated into the frontoparietalnetwork.Inaddition,childrenlackedconnectionsbetweentheDLPFCandthe IPS and inferior parietal lobule. The two separate networksseeninadultsareproposedtointerpretcues,implementtop-downcontrol,andprocessbottom-upfeedback,butusedifferentmecha-nismsandoverdifferenttemporalscales.(Reprintedwithpermis-sionfromFair,Dosenbach,etal.,2007,copyright2007NationalAcademyofSciences,USA.)

7

8

9

6


Y


1987;Blair,2006).Fluidreasoninghasbeenidentifiedasaleadingindicatorofchangesincrystallizedabilities(McArdle,2001). It strongly predicts changes in quantitative ability(Ferrer&McArdle,2004)andreading(Ferrer,2006)amongchildrenaged5to10.Fluidreasoningabilityevenpredictsperformancethroughcollegeandincognitivelydemandingoccupations(Gottfredson,1997).

One form of fluid reasoning is relational reasoning: theability to consider relationships between multiple distinctmentalrepresentations(Gentner,1983;Hummel&Holyoak,1997). Analogical reasoning, more specifically, involvesabstractingarelationshipbetweenfamiliaritemsandapply-ing it to novel representations (Gentner, 1988; Goswami,1989).Inotherwords,forminganalogiesallowsustodeter-minegeneralprinciplesfromspecificexamplesandtoestab-lish connections between previously unrelated pieces ofinformation.Analogical thought isan importantmeansbywhich cognition develops (Goswami, 1989; R. Brown &Marsden, 1990). For example, children use analogies tolearnnewwordsandconceptsbyassociationwithpreviouslylearnedinformation(Gentner,1983).

WhenDoesReasoningAbilityDevelop? Historically,theories of reasoning development focused on children’slimitations.Piagetclaimed that,before the stageof formaloperations around age 11, children are not capable ofmentally representing the relations necessary to solveanalogies(Inhelder,1958).WhenPiagetandhiscolleaguesshowedchildrenpictorialproblemsoftheform“AistoBasC is to...?”andasked themtofind theDtermamongasetofpictures,hefoundthatchildrenoftenchoseitemsthatwere perceptually or semantically related to the C item

(Piaget, 1977). Sternberg and colleagues found similarlimitations in young children’s analogical reasoning,observing an overreliance on lower-order relations duringanalogicalproblem solving (Sternberg, 1980,1982). Ithasbeenarguedthatchildrenasyoungas3yearsoldcansolvesimpleanalogiesaslongastheyarefamiliarwiththeobjectsinvolved and understand the relevant relations (Goswami,1989), but improvements in analogical reasoning areobservedthroughoutchildhoodandadolescence(Sternberg&Rifkin,1979;Richland,Morrison,etal.,2006).

Fluid reasoning ability seems to be a distinct cognitivefunction,risingandfallingatitsownrateacrossthelifespan(Cattell, 1987). It follows a different developmental trajec-torythancrystallizedabilities,supportingtheideathattheseareseparablecognitivefunctions(Horn,1991;Schaie,1996;McGrew, 1997). Fluid reasoning capacity increasesvery rapidly until late adolescence and early adulthood,peakingataroundage22anddecliningthereafter(McArdle,Ferrer-Caja,etal.,2002).

Assessing Reasoning Ability One of the mostcommonly used measures of fluid reasoning ability is theRaven’sProgressiveMatricestest(RPM),aclassicvisuospatialtask that can be administered to both children and adults(Raven,1941).Thistestisconsideredanexcellentmeasureof fluid reasoning ability (Kline, 1993) and of intellectualabilityoverall(Wechsler&Stone,1945).

As illustrated in figure 5.4, the RPM includes zero-relational (REL-0), one-relational (REL-1), and two-relational(REL-2)problems.REL-0problemsrequireonlyperceptual matching. REL-1 problems require subjects toconsidereitherverticalorhorizontalchanges(orspatialrela-

Figure5.3 Developmentofresponseinhibitionandperformancemonitoring. Velanova, Wheeler, and colleagues (2008) demon-stratedthatdACCshowedsignificantlygreatermodulationduringerrorversuscorrecttrialsinanantisaccadetask.Thetaskrequiredsubjectstoinhibittheprepotenttendencytolooktowardthestimu-lusandlookintheoppositedirection.Thetimecourseofactivationwithin thedACC, illustratedhere, not only shows thedifference

betweenactivityduringerrorandcorrecttrialsbutalsothatadultsexhibitgreaterdifferentialactivity thanadolescentsandchildren.Foreachagegroup,blackasterisksmark the timepoint showingmeanmaximalpeakactivityforerrortrials,andgrayasterisksmarkthe timepoint showingmeanmaximaldifferencesbetweenerrorand correct trials. (Reprinted with permission from Velanova,Wheeler,etal.,2008,copyrightOxfordUniversityPress.)

10

31

11


Y


Figure 5.4 Sample problem similar to Raven’s ProgressiveMatrices. (A) Zero-relational problem (REL-0) that requires onlyperceptual matching (Answer: 2). (B) One-relational problem(REL-1)thatinvolvesconsiderationofchangeineithertheverticalor horizontal direction (Answer: 1). (C ) Two-relational problem(REL-2)thatrequiresattentiontochangeinboththeverticalandhorizontaldirections(Answer:3).

Figure 5.5 Sample Visual analogy Problem. On this type ofproblem,subjectsmustconsidertherelationshipbetweenthetoptwo images and choose the image that completes the bottomanalogy(Answer:3;2isthesemanticlure).

tions)acrossfiguresina3×3gridtoinferthemissingpieceinthebottomrightcorner.REL-2problemsrequiresubjectstoprocesschangesinboththehorizontalandverticaldirec-tions simultaneously in order to choose the missing piece.Theseproblemsarethemostdifficultbecausetheyrequiretheintegrationoftwovisuospatialrelations.

Analogical reasoning can be assessed behaviorally andin an MRI scanner with propositional analogy problemsinvolvingeitherwordsorpicturesofnameableobjects.Thevisual analogy task used in a recent fMRI study from ourlaboratory(Wright,2008)requireschildrentoselectwhichoffourpicturescompletesananalogy.Theanswerchoicesfor these problems include perceptual and semantic lures(figure5.5).

Neural Correlates of Fluid Reasoning Brainregions important for fluid reasoning have been identifiedthroughstudiesofpatientswith impairedreasoningabilityand neuroimaging studies of healthy adults. Research onpatients in the early stages of frontotemporal dementia(FTD) has shown that reasoning is differentially affectedbasedonthebrainareasmostcompromisedbythedisease.Patientswithfrontal-variantFTDmakeerrorsonanalogicalreasoningproblemsrelatedtolimitedworkingmemoryandtrouble inhibiting inappropriate responses. In contrast,

patientswithtemporal-variantFTDareprofoundlyimpairedon analogical reasoning problems as a result of semanticmemory loss (Morrison, Krawczyk, et al., 2004). Anotherstudyofpatientswithprefrontaldamagerevealedthatthesepatients have a specific deficit in relational integration ascomparedtopatientswithanteriortemporal lobedamage,who are more impaired on tests of episodic and semanticmemory(Waltz,Knowlton,etal.,1999).

ImagingresearchhasnarroweddowntheregioninPFCresponsible for relational integration to the most anteriorpart of lateral PFC (RLPFC). Functional MRI studies ofreasoning, including the RPM task (Prabhakaran, Smith,et al., 1997; Christoff, Prabhakaran, et al., 2001; Kroger,Sabb,etal.,2002)andaverbalpropositionalanalogytask(Bunge,Wendelken,etal.,2005),have implicatedRLPFCinproblemsthatrequirejointconsiderationofmultiplerela-tions.TheotherlateralPFCregionsplayrolesinreasoningthatarenotspecificallyassociatedwithrelationalcomplex-ity.DLPFCmaysupportreasoningbyorganizingrepresen-tations in working memory, selecting between competingresponsealternatives,andmonitoringperformance(Christ-off et al.). Depending on the nature of the task, differentbrain regions contribute to fluid reasoning. Left VLPFC(Broca’s area) supports reasoning by retrieving semanticrelations on propositional analogy problems (Bunge et al.;Wright, 2008).Likewise, hippocampus andparietal cortexmay play a role in reasoning by representing individual

12


Y


visuospatialrelationsinvisuospatialtasksinvolvingrelationalintegration,includingRaven’sProgressiveMatrices(Crone,in press) and transitive inference problems (Wendelken,underreview).

Parietalcortexisconsistentlyengagedinhigh-levelcogni-tivetasksliketheRPM(Grayetal.,2003)andsharesstrongconnections with PFC (Petrides & Pandya, 1984; Fuster,2002).Astudyofindividualdifferencesinreasoningabilityinadultsshowedthatstrongerprefrontalandparietalrecruit-mentonadifficultworkingmemorytaskisassociatedwithbetter fluid reasoning, as measured by an RPM-type task(Grayetal.).ThelevelofactivationinleftlateralPFCandbilateralparietalcortexaccountedformorethan99.9%ofthe relationship between fluid intelligence and workingmemoryperformanceintheseadults.

Takentogether,theprecedingstudiessuggestthatmatu-rationofRLPFCshouldleadtoimprovementsinrelationalprocessing,whilematurationofBroca’sarea,hippocampus,andparietalcortexshouldleadtobetterreasoningthroughimprovedrepresentationofindividualverbalandvisuospa-tialrelations.

How Does the Brain Change to Allow forImprovements in Reasoning Ability? The neuroi-magingresearchdescribedsofarhasidentifiedbrainregionsthatcontributetoreasoninginadults.However,researchersare justnowbeginningtotrackhowtheseregionsdevelopduring childhood and how this development leads toimproved fluid reasoning ability. As noted earlier in thischapter, in the section on structural brain development,DLPFC, RLPFC, and parietal cortex develop relativelyslowly:corticalgraymatterlosscontinuesthroughtheearlytwenties (Giedd, 2004). A study by Shaw and colleaguesshowedthatthetrajectoryofchangesincorticalthicknessinseveral prefrontal regions differed across children withsuperior,high,andaverageintelligence(Shaw,Greenstein,etal.,2006).Surprisingly,childrenwithsuperiorintelligenceexhibited a delayed peak of cortical thickness in anteriorPFCrelativetotheothergroups,aroundage11asopposedtoage7–8 inchildrenofaverage intelligence.Theprecisesignificance of this intriguing finding is as yet unclear. Inparticular, the role of environmental factors has not beenexplored; in this study sample, IQ was correlated withsocioeconomicfactors.However,thisfindingindicatesthatcognitive ability is related to the particular time course ofcorticalmaturationinfrontalregions,ratherthanthesizeofagivenregionata specificage.Thisfindingspeaks to theuniqueinsightsthatcanbegainedfromlongitudinalstudiesofbraindevelopment.

Whilestructuralimaginghasprovidedcriticalinsightintothe neural changes that underlie reasoning development,functional neuroimaging is essential to understand howchangesinbrainfunctionleadtochangesinbehavior.This

section highlights the first two fMRI studies of reasoningabilityinchildren.

Visual analogies In the first study, our laboratory (Wright,2008) presented children aged 6–13 and young adultswith visual analogy problems (figure 5.5). Children werecapable of identifying analogous relationships betweenpairsofimages,butmadedisproportionatelymoremistakesthanadultsontheanalogyproblemsrelativeto1-relationalproblems that required themto select fromseveral imagestheone thatwasmost semanticallyrelated toacue image(figure5.5A).

InleftVLPFC,aregioninvolvedintheeffortfulretrievalof individual semantic relations between items (see, forexample,Badre&Wagner,2007),noconsistentdifferenceswereobservedbetweenchildrenandadults(Wright,2008).However,olderchildrendidengagethisregionmorestronglythanyoungerchildren,indicatingthatleftVLPFCcontrib-utes increasingly to controlled semantic retrieval betweentheagesof6and13.

In bilateral RLPFC, the time-course analyses providedstrongevidence foran immatureactivationprofile inchil-dren(figure5.6).ThepeakofactivationinRLPFCoccurredatleast4secondslaterforchildrenthanforadults,despiteminimaldifferences inresponse timesbetween thegroups.In fact, for children, RLPFC activation peaked after themotor cortex activation associated with the behavioralresponse. Overall, consistent with a model whereby rela-tively more rostral PFC matures later than caudal PFC(Bunge&Zelazo,2006),largerdifferencesbetweenchildrenand adults were observed in RLPFC than in VLPFC.Changes in the function of RLPFC over childhood andadolescencemaycontribute to improvements in reasoningability, and individual differences in RLPFC functioningmayexplain,atleastinpart,whysomepeoplehaveagreatercapacityforfluidreasoningthanothers.

Raven’s Progressive Matrices Inthesecondstudy,ourlabora-tory (Crone, in press) presented children aged 8–12 andyoungadultswithproblemsadaptedfromtheRaven’sPro-gressiveMatrices(figure5.4).Behaviorally,childrenmadeadisproportionatenumberoferrorsontheREL-2problemsrelativetoREL-1problems,andtheirresponsetimesontheREL-2 problems did not differ from those of adults. Thisfinding suggests that children selected responses for thesedifficultproblemsbeforeadequatelyconsideringbothdimen-sionsofrelationalchange.

Inadults,RLPFCactivationwasgreaterforREL-2prob-lemsthanforREL-1problems.WhilechildrenalsorecruitedRLPFC,theydidnotexhibitsustainedpreferentialrecruit-ment of RLPFC for REL-2 problems as compared withREL-1problems.Togetherwiththeresponsetimedata,thisfindingsuggests that thechildrenweremore likely to treat

13


Y


theREL-2problemssimilarlytoREL-1problems,consider-ingonlyasingledimensionofchange.ActivationofRLPFCassociatedwiththeREL-2problemsincreaseswithage,indi-catingthatdevelopmentofRLPFCintegrationmechanismoccurs,atleastinpart,overtheagerange(8–12years)thatwasstudied.UnlikeRLPFC,theinferiorparietallobulewassensitivetothenumberofrelationsinadultsandshowedanimmaturepatternofactivationinchildren.Insummary,thisstudyprovidesevidencethatthedevelopmentofreasoningisassociatedwithfunctionalchangesinRLPFCinresponsetorelationalintegration.

Insummary,fluidreasoningabilitycomesonlineearlyinchildhood but continues to develop through adolescenceandevenintoadulthood.IntelligenceinadultsisrelatedtoconnectivitybetweenPFCandparietalcortex(Shawetal.,2006).Structuralneuroimagingstudies (Giedd,2004)haveshownthatdevelopmentoftheseregions,PFCandparietalcortex, follows a prolonged developmental time coursethat matches behavioral data on reasoning in childhood(Richland et al., 2006). Initial functional neuroimagingstudieshaveshownthatchildrenrecruitbrainregionssimilartothosethatadultsusetosolveanalogyproblems,butwithdifferentpatternsofactivationsuggestingfunctionalimma-turity(Wright,2008;Crone,inpress).

Conclusions

Agrowingliteratureindicatesthattheincreasedrecruitmentof task-related regions in prefrontal and parietal regionscontributeto improvements ingoal-directedbehaviorovermiddlechildhoodandadolescence.Thepatternofdevelop-mentalchangesinbrainactivationhasbeengenerallychar-acterizedasashiftfromdiffusetofocalactivation(Durston,Davidson,etal.,2006)andfromposteriortoanterioractiva-tion (Rubiaetal., inpress;T.Brown,Lugar,etal.,2005).Differencescanbequantitative,withoneagegroupengag-ingaregionmorestronglyorextensivelythananother,and/or qualitative, with a shift in reliance on one set of brainregionstoanother(T.Brownetal.,2005;T.Brown,Petersen,etal.,2006;Rubiaetal.,inpress;Scherfetal.,2006;Badre&Wagner,2007).Importantly,theprecisepatternofchangeobserveddependsonthetask,theagesbeingexamined,andthebrainregioninquestion.Byfurthercharacterizingneu-rodevelopmental changes in cognitive control processeswithinsubjectsandacrossarangeoftasks,wehopetobetterunderstandthedevelopmentofthehumanmind.

CurrentandFutureDirections Byaroundage12,the ability to hold goal-relevant information in mind and

Figure5.6 RLPFCregionsofinterestandtimecourses.Ontheleftsideofthisimage,therightandleftRLPFCregionsofinterestareshowninasagittalview.Therightsidedisplaystimecoursesfromtheseregionsfrombaselineat2secondsthrough18seconds

aftertrialonsetforbothchildrenandadults.Bilaterally,thepeakofactivationoccursabout4secondslaterinchildrenthaninadults,and these regions even appear deactivated during the first fewsecondsofstimuluspresentation.

14


Y


use it to select appropriate actions is already adequate,although not fully mature. It is of great interest to trackbrain function associated with working memory andcognitive control earlier in childhood, when these abilitiesare first acquired. Optical imaging studies can be con-ducted from infancy onward, although the spatiotem-poral resolution of this method is suboptimal. It is nowpossibletoacquirefMRIdatainchildrenasyoungasfouryearsofage(Cantlon,Brannon,etal.,2006),althoughnotwithoutchallengeslikeheadmotion,lowaccuracy,andpoorattentionspan.

Animportantfuturedirectionistodeterminetheextenttowhichobservedagedifferencesinbrainactivationreflectharddevelopmentalconstraints(e.g.,therequiredanatomi-cal network is simply not yet in place at a given age) asopposed to lackofexperiencewithagiven typeof taskorcognitive strategy. Training studies involving several agegroupswouldallowustoinvestigateeffectsofageandeffectsofpracticeindependentlyandtotestwhetherinherentagedifferences in performance and brain activation are stillpresent after substantial practice (Luna & Sweeney, 2004;Qin,Carter,etal.,2004).

Thus far, all but one (Durston et al., 2006) of the pub-lisheddevelopmentalfMRIstudiesonworkingmemoryorcognitive control have compared groups of individuals atdifferentages.While thesecross-sectional studiesarevalu-able,theyprovideonlyacoarseindicatorofdevelopmentalchange.Itisalsoimportanttoconductlongitudinalstudiesto characterize intraindividual changes in brain functionwithage.

To understand how goal-directed behavior is achieved,it will be necessary to know how PFC and parietalcortices interactwithotherbrainregions.It is thematura-tion of a specific network, rather than a particular brainregion,thatdetermineshoweffectivelyagivenbrainprocessis carried out. Some information about these interactionscan be gleaned from functional connectivity analyses offMRI data. Another approach is to acquire fMRI andEEG data in the same group of participants, either inseparate sessions or simultaneously (Debener, Ullsperger,et al., 2005). An important current and future directionfor developmental neuroimaging studies is to examinedevelopmental changes in interactions between brainregions, furthering the work of Fair and colleagues (2007)demonstratedinfigure5.2.

Theexaminationofthenormaldevelopmentalpathwaysof distinct control functions will be important for under-standing sensitive periods in brain development. Forexample,damagetoPFCinchildhoodhasamuchgreaterimpactthandoesdamageinadulthood, likelybecausethisregionisimportantforacquiringskillsandknowledgeduringchildhood(Eslinger,Flaherty-Craig,etal.,2004).

Finally, a better understanding of neurodevelopmentalchanges in healthy children will lead to insights into thereasonsforimpoverishedgoal-directedbehaviorinanumberofneurodevelopmentaldisorders,suchasattention-deficit/hyperactivitydisorder(Vaidya,Bunge,etal.,2005;Durston,Mulder, et al., 2006) and Tourette syndrome (Peterson,Pine,etal.,2001;Baym,Corbett,etal.,2008).

REFERENCES

Adleman, N. E., Menon, V., et al. (2002). A developmentalfMRI study of the Stroop color-word task. Neuroimage, 16(1),61–75.

Badre,D.,&Wagner,A.D.(2007).Leftventrolateralprefrontalcortex and the cognitive control of memory. Neuropsychologia,45(13),2883–2901.

Baym, C. L., Corbett, B. A., et al. (2008). Neural correlatesof tic severity and cognitive control in children with Tourettesyndrome.Brain,131(1),165.

Blair,C.(2006).Howsimilararefluidcognitionandgeneralintel-ligence?Adevelopmentalneuroscienceperspectiveonfluidcog-nitionasanaspectofhumancognitiveability.Behav. Brain Sci.,29(2),109–125;discussion,125–160.

Booth,J.R.,Burman,D.D.,etal.(2003).Neuraldevelopmentofselectiveattentionandresponse inhibition.Neuroimage,20(2),737–751.

Brocki,K.C.,&Bohlin,G.(2004).Executivefunctionsinchil-drenaged6to13:Adimensionalanddevelopmentalstudy.Dev. Neuropsychol.,26(2),571–593.

Brown,R.G.,&Marsden,C.D.(1990).CognitivefunctioninParkinson’sdisease:Fromdescriptiontotheory.Trends Neurosci.,13(1),21–29.

Brown, T. T., Lugar, H. M., et al. (2005). Developmentalchanges in human cerebral functional organization for wordgeneration.Cerebral Cortex,15(3),275–290.

Brown,T.T.,Petersen,S.E.,etal.(2006).Doeshumanfunc-tionalbrainorganizationshiftfromdiffusetofocalwithdevelop-ment?Dev. Sci.,9(1),9–11.

Bunge,S.A.,Dudukovic,N.M.,etal.(2002).Immaturefrontallobe contributions to cognitive control in children: EvidencefromfMRI.Neuron,33(2),301–311.

Bunge,S.A.,Wendelken,C.,etal.(2005).Analogicalreasoningandprefrontalcortex:Evidenceforseparableretrievalandinte-grationmechanisms.Cerebral Cortex,15(3),239–249.

Bunge, S. A., & Wright, S. B. (2007). Neurodevelopmentalchanges inworkingmemory and cognitive control.Curr. Opin. Neurobiol.,17(2),243–250.

Bunge,S.A.,&Zelazo.P.D.(2006).Abrain-basedaccountofthedevelopmentofruleuseinchildhood.Curr. Dir. Psychol. Sci.,15(3),118–121.

Cantlon,J.F.,Brannon,E.M.,etal.(2006).Functionalimagingofnumericalprocessinginadultsand4-y-oldchildren.PLoS Biol,4(5),e125.

Carroll,J.B.(1997).The three-stratum theory of cognitive abilities.NewYork:Guilford.

Case,R.(1992).Theroleofthefrontallobesintheregulationofcognitivedevelopment.Brain Cogn.,20(1),51–73.

Casey,B.J.,Cohen,J.D.,etal.(1995).ActivationofprefrontalcortexinchildrenduringanonspatialworkingmemorytaskwithfunctionalMRI.2,221–229.

15

16

17


Y


Casey,B.J.,Tottenham,N.,etal.(2005).Imagingthedevelop-ingbrain:Whathavewelearnedaboutcognitivedevelopment?Trends Cogn. Sci.9(3),104–110.

Cattell, R. B. (1971). Abilities: Their structure, growth and action.Boston:Houghton-Mifflin.

Cattell, R. B. (1987). Intelligence: Its structure, growth and action.Amsterdam:North-Holland.

Christoff, K., Prabhakaran, V., et al. (2001). Rostrolateralprefrontal cortex involvement in relational integration duringreasoning.Neuroimage,14(5),1136–1149.

Crone,E.A.,Wendelken,C.,etal. (2004).Contributions of pre-frontal subregions to developmental changes in rule use. San Diego:SocietyforNeuroscience.

Crone, E. A., Wendelken, C., et al. (2006). Neurocognitivedevelopmentoftheabilitytomanipulateinformationinworkingmemory.Proc. Natl. Acad. Sci. USA,103(24),9315–9320.

Crone, E. A., Wendelken, C., van Leijenhorst, L.,Honomichl,R.,Christoff,K.,&Bunge, S.A. (inpress).Neurocognitivedevelopmentofrelationalreasoning.Dev. Sci.

D’Esposito, M. (2007). From cognitive to neural models ofworkingmemory.Philos. Trans. R. Soc. London [Biol.],362(1481),761–772.

Debener, S., Ullsperger, M., et al. (2005). Trial-by-trialcoupling of concurrent electroencephalogram and functionalmagnetic resonance imaging identifies thedynamicsofperfor-mancemonitoring.J. Neurosci.,25(50),11730–11737.

Dempster,F.N.(1993).Resistancetointerference:Developmen-tal changes in a basic processing mechanism. In M. L. H. R.Pasnak(Ed.),Emerging themes in cognitive development: Vol. 1. Founda-tions(pp.3–27).NewYork:Springer-Verlag.

Diamond, A. (2002). Normal development of prefrontal cortexfrom birth to young adulthood: Cognitive functions, anatomyand biochemistry. In S. A. Knight (Ed.), Principles of frontal lobe function (pp. 466–503). Oxford, UK: Oxford UniversityPress.

Durston,S.,Davidson,M.C.,etal.(2006).Ashiftfromdiffusetofocalcorticalactivitywithdevelopment.Dev. Sci.,9(1),1–8.

Durston, S., Mulder, M., et al. (2006). Activation in ventralprefrontal cortex is sensitive to genetic vulnerability for atten-tion-deficit hyperactivity disorder. Biol. Psychiatry, 60(10),1062–1070.

Durston,S.,Thomas,K.M., et al. (2002).Aneuralbasis forthedevelopmentofinhibitorycontrol.Dev. Sci.,5(4):F9-F16.

Edin, F., Macoveanu, J., et al. (2007). Stronger synaptic con-nectivity as a mechanism behind development of workingmemory-relatedbrainactivityduringchildhood.J. Cogn. Neuro-sci.,19(5),750.

Eslinger,P.J.,Flaherty-Craig,C.V.,etal.(2004).Develop-mental outcomes after early prefrontal cortex damage. Brain Cogn.,55(1),84–10320.

Fair, D. A., Dosenbach, N. U., et al. (2007). Development ofdistinct control networks through segregation and integration.Proc. Natl. Acad. Sci. USA,104(33),13507–13512.

Ferrer,E.,&McArdle, J. J. (2004).Anexperimentalanalysisofdynamichypothesesaboutcognitiveabilitiesandachievementfrom childhood to early adulthood. Dev. Psychol., 40(6),935–952.

Ferrer,E.McArdle,J.J.,Shaywitz,B.A.,Holahan,J.N.,Marchione, K., & Shaywitz, S. E. (2006). Longitudinalmodels of developmental dynamics between reading andcognition from childhood to adolescence. Dev. Psychol., 43,1460–1473.

Flynn, J. (2007). What is intelligence? New York: CambridgeUniversityPress.

Ford, K. A., Goltz, H. C., et al. (2005). Neural processesassociatedwithantisaccade taskperformance investigatedwithevent-relatedfMRI.J. Neurophysiol.,94(1),429–440.

Fuster, J. M. (2002). Frontal lobe and cognitive development.J. Neurocytol.,31(3–5),373–385.

Gentner,D.(1983).Structure-mapping:Atheoreticalframeworkforanalogy.Cogn. Sci.,7,155–170.

Gentner,D. (1988).Metaphoras structuremapping:Therela-tionalshift.Child Dev.,59,47–59.

Giedd,J.N.(2004).Structuralmagneticresonanceimagingoftheadolescentbrain.Ann. NY Acad. Sci.,1021,77–85.

Giedd, J.N.,Blumenthal, J.,etal. (1999).Braindevelopmentduring childhoodandadolescence:A longitudinalMRI study.Nature Neurosci.,2(10),861–863.

Gogtay, N., Giedd, J. N., et al. (2004). Dynamic mapping ofhuman cortical development during childhood through earlyadulthood.Proc. Natl. Acad. Sci. USA,101(21),8174–8179.

Gogtay,N.,Nugent,T.F.,3rd,etal.(2006).Dynamicmappingofnormalhumanhippocampaldevelopment.Hippocampus,16(8),664–672.

Goldman-Rakic,P.S.(1992).Workingmemoryandthemind.Sci. Am.,267(3),110–117.

Goswami,U.(1989).Relationalcomplexityandthedevelopmentofanalogicalreasoning.Cogn. Dev.,4,251–268.

Gottfredson,L.S. (1997).Whygmatters:Thecomplexityofeverydaylife.Intelligence,24,79–132.

Gray,J.R.,Chabris,C.F.,etal.(2003).Neuralmechanismsofgeneralfluidintelligence.Nature Neurosci.,6(3),316–322.

Horn,J.L.(1988).Thinkingabouthumanabilities.InJ.R.N.a.R. B. Catell (Eds.), Handbook of multivariate experimental psychology(pp.645–685).NewYork:AcademicPress.

Horn, J. L. (1991). Measurement of intellectual capacities: Areview in theory. In K. S. McGrew, J. K. Werder, & R. W.Woodcock(Eds.),Woodcock-Johnson technical manual(pp.197–246).Allen,TX:DLMTeachingResources.

Horn,J.L.,&Cattell,R.B.(1967).Agedifferencesinfluidandcrystallizedintelligence.Acta Psychol. (Amst.),26,107–129.

Huizinga,M.,Dolan,C.V.,etal.(2006).Age-relatedchangeinexecutivefunction:Developmentaltrendsandalatentvariableanalysis.Neuropsychologia,44(11),2017–2036.

Hummel,J.E.,&Holyoak,K.J.(1997).Distributedrepresenta-tionsof structure:A theoryof analogical access andmapping.Psychol. Rev.,10(3),427–466.

Inhelder,B.P.,J.(1958).The growth of logical thinking from childhood to adolescence.NewYork:BasicBooks.

Kline, P. (1993). The handbook of psychological testing. London:Routledge.

Klingberg,T.,Forssberg,H., et al. (2002a). Increasedbrainactivityinfrontalandparietalcortexunderliesthedevelopmentof visuospatial working memory capacity during childhood.J. Cogn. Neurosci.,14(1),1–10.

Klingberg, T., Forssberg, H., et al. (2002b). Training ofworkingmemoryinchildrenwithADHD.J. Clin. Exp. Neuropsy-chol.,24(6),781–791.

Klingberg, T., Vaidya, C. J., et al. (1999). Myelination andorganizationofthefrontalwhitematterinchildren:AdiffusiontensorMRIstudy.Neuroreport,10(13),2817–2821.

Kroger,J.K.,Sabb,F.W.,etal.(2002).Recruitmentofanteriordorsolateralprefrontalcortexinhumanreasoning:Aparametricstudyofrelationalcomplexity.Cerebral Cortex,12(5),477–485.

18

19

20

21

22


Y


Kwon,H.,Reiss,A.L.,etal. (2002).Neuralbasisofprotracteddevelopmental changes in visuo-spatialworkingmemory.Proc. Natl. Acad. Sci. USA,99(20),13336–13341.

Lamm, C., Zelazo, P. D., et al. (2006). Neural correlates ofcognitive control in childhood and adolescence: Disentanglingthecontributionsofageandexecutivefunction.Neuropsychologia,44(11),2139.

Liston,C.,Watts,R.,etal.(2006).Frontostriatalmicrostructuremodulates efficient recruitment of cognitive control. Cerebral Cortex,16(4),553–560.

Luciana,M.,&Nelson,C.A.(1998).Thefunctionalemergenceofprefrontally-guidedworkingmemorysystemsinfour-toeight-year-oldchildren.Neuropsychologia,36(3),273–293.

Luna,B.,&Sweeney,J.A.(2004).Theemergenceofcollabora-tivebrainfunction:FMRIstudiesofthedevelopmentofresponseinhibition.Ann. NY Acad. Sci.,1021,296–309.

Marsh, R., Zhu, H., et al. (2006). A developmental fMRIstudy of self-regulatory control. Hum. Brain Mapping, 27(11),848–863.

McArdle,J.J.(2001).Alatentdifferencescoreapproachtolon-gitudinaldynamicstructuralanalysis.InS.d.T.R.Cudeck&D. Sörbom (Eds.), Structural equation modeling: Present and future: A Festschrift in honor of Karl Jöreskog (pp.341–380).Lincolnwood,IL:ScientificSoftwareInternational.

McArdle, J. J., Ferrer-Caja, E., et al. (2002). Comparativelongitudinalstructuralanalysisofgrowthanddeclineofmultipleintellectual abilities over the lifespan. Dev. Psychol., 38(1),113–142.

McArdle, J. J., & Woodcock, J. R. (1998). Human cognitive abilities in theory and practice. Mahwah, NJ: Lawrence ErlbaumAssociates.

McGrew,K.S.(1997).Analysisofthemajorintelligencebatteriesaccording to a proposed comprehensive Gf-Gc framework. InJ. L. G. D. P. Flanagan & P. L. Harrison (Eds.), Contemporary intellectual assessment: Theories, tests, and issues (pp.131–150).NewYork:Guilford.

Miller,G.A.,Galanter,E.,etal.(1960).Plans and the structure of behavior.NewYork:Holt,Rinehart&Winston.

Morrison,R.G.,Krawczyk,D.C.,etal.(2004).Aneurocom-putationalmodelofanalogicalreasoninganditsbreakdowninfrontotemporal lobar degeneration. J. Cogn. Neurosci., 16(2),260–271.

Munakata, Y., Casey, B. J., et al. (2004). Developmentalcognitiveneuroscience:Progressandpotential.Trends Cogn. Sci.,8(3),122–128.

Nagy, Z., Westerberg, H., et al. (2004). Maturation of whitematterisassociatedwiththedevelopmentofcognitivefunctionsduringchildhood.J. Cogn. Neurosci.,16(7),1227–1233.

O’Donnell, S., Noseworthy, M. D., et al. (2005). Corticalthicknessofthefrontopolarareaintypicallydevelopingchildrenandadolescents.Neuroimage,24(4),948–954.

Olesen,P.J.,Nagy,Z.,etal.(2003).CombinedanalysisofDTIand fMRI data reveals a joint maturation of white and greymatter ina fronto-parietalnetwork.Brain Res. Cogn. Brain Res.,18(1),48–57.

Petanjek, Z., Judas, M., et al. (2008). Lifespan alterations ofbasal dendritic trees of pyramidal neurons in the humanprefrontalcortex:Alayer-specificpattern.Cerebral Cortex,18(4),915–929.

Peterson,B.S.,Pine,D.S.,etal.(2001).Prospective,longitudi-nal study of tic, obsessive-compulsive, and attention-deficit/hyperactivity disorders in an epidemiological sample. J. Am. Acad. Child Adolesc. Psychiatry,40(6),685–695.

Petrides,M.,&Pandya,D.N.(1984).Projectionstothefrontalcortexfromtheposteriorparietalregionintherhesusmonkey.J. Comp. Neurol.,228(1),105–116.

Piaget, J. M., J., & Billeter, J. (1977). La formation descorrelates.InJ.Piaget(Ed.),Recherches sur l’abstraction reflechissante(pp.115–129).Paris:PressesUniversitairesdeFrance.

Polli, F. E., Barton, J. J. S., et al. (2005). Rostral and dorsalanteriorcingulatecortexmakedissociablecontributionsduringantisaccadeerrorcommission.Proc. Natl. Acad. Sci. USA,102(43),15700–15705.

Prabhakaran,V.,Smith,J.A.L.,etal.(1997).Neuralsubstratesof fluid reasoning: An fMRI study of neocortical activationduringperformanceof theRaven’sProgressiveMatricesTest.Cogn. Psychol.,33,43–63.

Qin, Y., Carter, C. S., et al. (2004). The change of the brainactivation patterns as children learn algebra equation solving.Proc. Natl. Acad. Sci. USA,101(15),5686.

Raven, J. C. (1941). Standardization of progressive matrices.Br. J. Med. Psychol.,19,137–150.

Richland, L. E., Morrison, R. G., et al. (2006). Children’sdevelopmentofanalogicalreasoning:Insightsfromsceneanalogyproblems.J. Exp. Child Psychol.,94(3),249–273.

Rubia,K.,Russell,T.,etal.(2001).Mappingmotorinhibition:Conjunctive brain activations across different versions of go/no-goandstoptasks.Neuroimage,13(2),250–261.

Rubia,K.,&Smith,A.(2004).Theneuralcorrelatesofcognitivetimemanagement:Areview.Acta Neurobiol. Exp.,64,329–340.

Rubia, K., Smith, A. B., et al. (2006). Progressive increase offrontostriatal brain activation from childhood to adulthoodduring event-related tasks of cognitive control. Hum. Brain Mapping,27(12),973–993.

Rubia,K.,Smith,A.B.,etal.(inpress).Developmentofinhibi-tionanderrordetectioninfMRI.Hum. Brain Mapping.

Schaie, K. W. (1996). Intellectual development in adulthood: The Seattle Longitudinal Study. New York: Cambridge UniversityPress.

Scherf, K. S., Sweeney, J. A., et al. (2006). Brain basis ofdevelopmentalchangeinvisuospatialworkingmemory.J. Cogn. Neurosci.,18(7),1045.

Schroeter,M.L.,Zysset,S.,etal.(2004).Prefrontalactivationdue to Stroop interference increases during development—Anevent-relatedfNIRSstudy.Neuroimage,23(4),1317–1325.

Shaw,P.,Greenstein,D.,etal. (2006). Intellectualabilityandcortical development in children and adolescents. Nature,440(7084),676–679.

Sowell,E.R.,Peterson,B.S.,etal. (2003).Mappingcorticalchange across the human life span. Nature Neurosci., 6(3),309–315.

Sowell,E.R.,Peterson,B.S.,etal.(2007).Sexdifferencesincorticalthicknessmappedin176healthyindividualsbetween7and87yearsofage.Cerebral Cortex,17(7),1550–1560.

Spear,L.P.(2007).Thedevelopingbrainandadolescent-typicalbehaviorpatterns:Anevolutionaryapproach.InD.W.Romer(Ed.), Adolescent psychopathology and the developing brain: Integrating brain and prevention science(pp.9–30).NewYork:OxfordUniver-sityPress.

Sternberg, R. J., & Rifkin, B. (1979). The development ofanalogical reasoning processes. J. Exp. Child Psychol., 27(2),195–232.

Sternberg,R.J.D.,C.(1982).Thedevelopmentofhigher-orderreasoninginadolescence.Child Dev.,53,209–221.

Sternberg, R. J. N., G. (1980). Developmental patterns in thesolutionofverbalanalogies.Child Dev.,51,27–38.

23

24

25

26

28

27


Y


Tamm,L.,Menon,V.,etal.(2002).Maturationofbrainfunctionassociated with response inhibition. J. Am. Acad. Child Adolesc. Psychiatry,41(10),1231–1238.

Thomas, K. M., King, S. W., et al. (1999). A developmentalfunctional MRI study of spatial working memory. Neuroimage,10(3,Pt.1),327–338.

Vaidya,C.J.,Bunge,S.A.,etal.(2005).Alteredneuralsubstratesofcognitivecontrol inchildhoodADHD:Evidencefromfunc-tional magnetic resonance imaging. Am. J. Psychiatry. 162(9),1605–1613.

van denWildenberg,W.P.M.,& van derMolen,M.W.(2004).Developmentaltrendsinsimpleandselectiveinhibitionofcompatibleandincompatibleresponses.J. Exp. Child Psychol.,87(3),201–220.

Velanova, K., Wheeler, M. E., et al. (2008). Maturationalchanges in anterior cingulate and frontoparietal recruitmentsupport the development of error processing and inhibitorycontrol.Cerebral Cortex.

Waltz, J. A., Knowlton, B. J., et al. (1999). A system forrelational reasoning in human prefrontal cortex. Psychol. Sci.,10(2),119–125.

Wechsler,D.,&Stone,C.P.(1945).Wechsler Memory Scale.NewYork:PsychologicalCorporation.

Welsh, K., Butters, N., et al. (1991). Detection of abnormalmemory decline in mild cases of Alzheimer’s disease usingCERAD neuropsychological measures. Arch. Neurol., 48,278–281.

Wendelken,C.B.,S.A. (under review).Relationalprocessing:Distinctcontributionsofanteriorprefrontalcortexandthehip-pocampus.Nature Neurosci.

Wright, S. B., Matlen, B. J., Baym, C. L., Ferrer, E., &Bunge, S. A. (2008). Neural correlates of fluid reasoning inchildrenandadults.Front. Hum. Neurosci.,1(8).

Zelazo,P.D.,Craik,F.I.,etal.(2004).Executivefunctionacrossthelifespan.Acta Psychol. (Amst),115(2–3),167–183.

30

29


Y


Y

AUTHOR QUeRy FORMDear Author

During the preparation of your manuscript for publication, the questions listed below have arisen. Please attend to these matters and return this form with your proof.

Many thanks for your assistance.

Query Query RemarksReferences

1. Au:confirmaffiliations.

2. Au:PlsfollowAPAstyleforcitationsandreferences.Forthreetofiveauthors,giveallnamesatfirstmention;thereafter,onename+etal.forsixasmore,useetal.atalltimes.

3. Au:Klingberg...:Isthis’02aor’02b?IsthesecondKlingberg...aorb?

4. Au:DoesBOLDneedtobespelledout?

5. Au:HanVelanova...beenpublished?Ifnot,plsgiveas“inpress”.

6. Au:PlsupdateRubia,Smith...inpreesifithasbeenpublished.

7. Au:HasVelanova...beenpublished?

8. Au:Compromise?Comprise?

9. Au:Wedon’thaveHorn’67orMcArdle’98.ShouldthesebeHorn&Cattell’67andMcArdle&Woodcock’98?

10. Au:Wedon’thaveFerrer’06.IsthisFerrerMcArdle&Shaywitz,Holahan,Marchione,&Shaywitz’06?

11. Au:Wedon’thavePiaget’77.IsthisPiaget&Billeter’77?

12. Au:IsthisWright,Matlen,Baym,Ferrer,&Bunge’08?+p.251&p.249×2(alsoonp.247).

13. Au:Crone,inpress:IsthisGrone,Wendelken,vanLeijenhorstHonomichl,Christoff,&Bunge?i.e.Croneetal.?(alsop.250,p.251).

14. Au:PlsupdateRubia...inpressifithasbeenpublished.(×2).

15. Au:Plsgiveallauthorsnamesiftherearesixonfewer.Ifmorethansix,givesixnamesfollowedby“etal.”

16. Au:R.Brown...’90:Trends Neurosci.OK?That’swhatwe’rebeendoingandalsothewayoursourcegiveit.

17. Au:Casey,Cohen...:Plsgivejournalname.

GAZ_04


Y


18. Au:Crone...inpress:Plsclarify“W.,C.;”(WendelkenOK?).Plssupplydataifithasbeenpublished.

19. Au:Dempster:“M.L.H.R.”OK?Diamond:OKasedited?

20. Au:Eslinger...pp.84–10320?Whatshouldthisbe?

21. Au:Horn...’88:“J.R.N.a.R.B.Catell”?Plsclarify.

22. Au:Inhelder,B.P.,J.:WhatisJ.?Jr.?

23. Au:McArdle’01:‘S.d.T.R.’OK?

24. Au:McGrew:“J.L.G.D.P.”?Isthiscorrect?

25. Au:Rubia,Smith’04:IsthatActa Neurobiol. Exp.[Warsz.]?

26. Au: Pls update Rubia, Smith...in press if it has beenpublished.

27. Au:Spear:Whatis“E.F.”?asecondeditor?

28. Au:Sternberg,R.J.D.,C.Sternberg,R.J.N.,G.aretheseOK?Whatismeant?

29. Au:Velanova...:Plsgivevol,andpages.

30. Au: Wendelken...: “S. A.”? “& Bunge, S. A.”? Wright...:pages?

31. Au:PslupdateVelanova...ifithasbeenpublished.

32. Au:Therunningfootistoolongtofit,weshortenedit.Plscheckandadvise.


5 B rain Changes Underlying the Development of Cognitive Control ...

Documents

Transcript of 5 B rain Changes Underlying the Development of Cognitive Control ...