Ebook Brain and behavior (5E): Part 2

PartIVComplexBehavior
Chapter12.LearningandMemory
Chapter13.IntelligenceandCognitiveFunctioning
Chapter14.PsychologicalDisorders
Chapter15.SleepandConsciousness

iStock/michellegibson


12LearningandMemory
LearningastheStorageofMemories
Amnesia:TheFailureofStorageandRetrieval
MechanismsofConsolidationandRetrieval
APPLICATION|TheLegacyofHm
WhereMemoriesAreStored
TwoKindsofLearning
WorkingMemory
CONCEPTCHECK
BrainChangesinLearning
Long-TermPotentiation
HowLTPHappens
ConsolidationandSleep
ChangingOurMemories
APPLICATION|TotalRecall
INTHENEWS|EnhancingSoldiers’LearningWithNeurostimulation
CONCEPTCHECK
LearningDeficienciesandDisorders
EffectsofAgingonMemory
Alzheimer’sDisease
RESEARCHSPOTLIGHT|Alzheimer’sinaDish
KorsakoffSyndrome

CONCEPTCHECK
InPerspective
ChapterSummary
StudyResources
Afterreadingthischapter,youwillbeableto:
Explainhowthebrainisinvolvedinthedifferenttypesofmemory.
Diagramtheneuralinvolvementinprocessingofinformationthatisstoredinmemory.
Describethechangesthatoccurinthebrainaslearningproceeds.
Examinehowmemorychangesduringaging.
Contrasttheimpactsofnormalaginganddisordersonmemory.

Attheageofseven,HenryMolaison’slifewasforeverchangedbyaseemingly


minorincident:Hewasknockeddownbyabicycleandwasunconsciousforfive
minutes.Threeyearslater,hebegantohaveminorseizures,andhisfirstmajor
seizureoccurredonhis16thbirthday.Still,Henryhadareasonablynormal
adolescence,takenupwithhighschool,scienceclub,hunting,androller-skating,
exceptforatwo-yearfurloughfromschoolbecausetheotherboysteasedhim
abouthisseizures.
Masterthecontent.
edge.sagepub.com/garrett5e
Afterhighschool,hetookajobinafactory,buteventuallytheseizuresmadeit
impossibleforhimtowork.Hewasaveraging10smallseizuresadayand1
majorseizureperweek.Becauseanticonvulsantmedicationswereunableto
controltheseizures,Henryandhisfamilydecidedonanexperimentaloperation
thatheldsomepromise.In1953,whenHenrywas27,asurgeonremovedmuch
ofbothofhistemporallobes,wheretheseizureactivitywasoriginating.The
surgeryworked,forthemostpart:Withthehelpofmedication,thepetitmal
seizuresweremildenoughnottobedisturbing,andmajorseizureswerereduced

toaboutoneayear.Henryreturnedtolivingwithhisparents.Hehelpedwith
householdchores,mowedthelawn,andspenthissparetimedoingdifficult
crosswordpuzzles.Later,heworkedatarehabilitationcenter,doingroutine
taskslikemountingcigarettelightersoncardboardstoredisplays.
Henry’sintelligencewasnotimpairedbytheoperation;hisIQtestperformance
evenwentup,probablybecausehewasfreedfromtheinterferenceofthe
abnormalbrainactivity.However,therewasoneimportantandunexpected
effectofthesurgery.Althoughhecouldrecallpersonalandpubliceventsand
remembersongsfromhisearlierlife,Henryhaddifficultylearningandretaining
newinformation.Hecouldholdnewinformationinmemoryforashortwhile,
butifheweredistractedorifafewminutespassed,hecouldnolongerrecallthe
information.Whenheworkedattherehabilitationcenter,hecouldnotdescribe
theworkhedid.Hedidnotremembermovingintoanursinghomein1980,or
evenwhatheateforhislastmeal.Andalthoughhewatchedtelevisionnews
everynight,hecouldnotremembertheday’snewseventslaterorevenrecallthe
nameofthepresident(Corkin,1984;B.Milner,Corkin,&Teuber,1968).

Discoveringthephysicalbasisoflearninginhumansandothermammalsisamongthe


greatestremainingchallengesfacingtheneurosciences.
—T.H.Brown,Chapman,Kairiss,&Keenan,1988

Henry’sinabilitytoformnewmemorieswasnotabsolute.Althoughhecouldnot
findhiswaybacktothenewhomehisfamilymovedtoafterhissurgeryifhe
wasmorethantwoorthreeblocksaway,hewasabletodrawafloorplanofthe
house,whichhehadnavigatedmanytimesdaily(Corkin,2002).Overtheyears,
hebecameawareofhiscondition,andhewasveryinsightfulaboutit.Inhisown
words,
Everydayisaloneinitself,whateverenjoymentI’vehad,andwhatever

sorrowI’vehad....Rightnow,I’mwondering.HaveIdoneorsaid
anythingamiss?Yousee,atthismomenteverythinglookscleartome,but
whathappenedjustbefore?That’swhatworriesme.It’slikewakingfroma
dream;Ijustdon’tremember.(B.Milner,1970,p.37)
Overaperiodof55years,Henrywouldbethesubjectofahundredscientific
studiesthathecouldnotremember;hewasknowntotheworldaspatientHMto
protecthisprivacy.Inthenextseveralpages,youwillseewhymanyconsider
HM’ssurgerythemostsignificantsingleeventinthestudyoflearning.

LearningastheStorageofMemories
Someone-celledanimals“learn”surprisinglywell,forexample,toavoid
swimmingtowardalightwheretheyhavereceivedanelectricshockbefore.I
haveplacedthetermlearninquotationmarksbecausesuchsimpleorganisms
lackanervoussystem;theirbehaviorchangesbriefly,butifyoutakealunch
breakduringyoursubject’straining,whenyoureturn,youwillhavetostartall
overagain.Suchatemporaryformoflearningmayhelpanorganismavoidan
unsafearealongenoughforthedangertopassorlingerinaplacewherefoodis
moreabundant.Butwithouttheabilitytomakeapermanentrecord,youcould
notlearnaskill,andexperiencewouldnothelpshapewhoyouare.Wewill
introducethetopicoflearningbyexaminingtheproblemofstorage.


Amnesia:TheFailureofStorageandRetrieval
HM’ssymptomsarereferredtoasanterogradeamnesia,animpairmentin
formingnewmemories.(Anterogrademeans“movingforward.”)Thiswasnot
HM’sonlymemorydeficit;thesurgeryalsocausedretrogradeamnesia,the
inabilitytoremembereventspriortoimpairment.Hisretrogradeamnesia
extendedfromthetimeofsurgerybacktoabouttheageof16;hehadafew
memoriesfromthatperiod,buthedidnotremembertheendofWorldWarIIor
hisowngraduation,andwhenhereturnedforhis35thhighschoolreunion,he

recognizednoneofhisclassmates.Bettermemoryforearliereventsthanfor
recentonesmayseemimplausible,butitistypicalofpatientswhohavebrain
damagelikeHM’s.Howfarbacktheretrogradeamnesiaextendsdependson
howmuchdamagethereisandwhichspecificstructuresaredamaged.

Howdoesstudyingamnesiahelpusunderstandmemory?
HM’ssurgerydamagedordestroyedthehippocampus,nearbystructuresthat
alongwiththehippocampusmakeupthehippocampalformation,andthe
amygdala.Figure12.1showsthelocationofthesestructures.Becausetheyare
onorneartheinsidesurfaceofthetemporallobe,theyformpartofwhatis
knownasthemedialtemporallobe(rememberthatmedialmeans“towardthe
middle”).BecauseHM’ssurgerywassoextensive,itisimpossibletotellwhich
structuresareresponsibleforthememoryfunctionsthatwerelost.Studiesof
patientswithvaryingdegreesoftemporallobedamagehavehelpeddetermine
whichstructuresareinvolvedinamnesiaand,therefore,inmemory.Henrydied
in2008attheageof82,buthecontinuestocontribute,astheaccompanying
Applicationexplains.
Thehippocampusconsistsofseveralsubstructureswithdifferentfunctions.The
partknownasCA1providestheprimaryoutputfromthehippocampustoother
brainareas;damageinthatpartofbothhippocampiresultsinmoderate
anterogradeamnesiaandonlyminimalretrogradeamnesia.Ifthedamage
includestherestofthehippocampus,anterogradeamnesiaissevere.Damageto
theentirehippocampalformationresultsinretrogradeamnesiaextendingback
15yearsormore(J.J.Reed&Squire,1998;Rempel-Clower,Zola,Squire,&
Amaral,1996;Zola-Morgan,Squire,&Amaral,1986).Moreextensive


retrogradeimpairmentoccurswithbroaderdamageordeterioration,likethat
seeninAlzheimer’sdisease,Huntington’sdisease,andParkinson’sdisease,
apparentlybecausememorystorageareasinthecortexarecompromised(Squire

&Alvarez,1995).
Figure12.1TemporalLobeStructuresInvolvedinAmnesia.

Sources:(a)From“HM’sMedialTemporalLobeLesion:FindingsFrom
MagneticResonanceImaging,”byS.Corkin,D.G.Amaral,R.G.
González,K.A.Johnson,andB.T.Hyman,1997,JournalofNeurosicence,
17,pp.3964–3979.Copyright©1997bytheSocietyforNeuroscience.
Usedwithpermission.(b)Adaptedwithpermissionfrom“Remembranceof
ThingsPast,”byD.L.SchacterandA.D.Wagner,Science,285,pp.1503–
1504.Illustration:K.Sutliff.©1999AmericanAssociationforthe
AdvancementofScience.ReprintedwithpermissionfromAAAS.

MechanismsofConsolidationandRetrieval
HM’smemoryimpairmentconsistedoftwoproblems:consolidationofnew
memoriesand,toalesserextent,retrievalofoldermemories.Consolidationis
theprocessinwhichthebrainformsamoreorlesspermanentphysical
representationofamemory.Retrievalistheprocessofaccessingstored
memories—inotherwords,theactofremembering.Whenaratpressesa
levertoreceiveafoodpelletorachildisbittenbyadogoryouskimthroughthe


headingsinthischapter,theexperienceisheldinmemoryatleastforabrief
time.Butjustlikethephonenumberthatisforgottenwhenyougetabusysignal
thefirsttimeyoudial,anexperiencedoesnotnecessarilybecomeapermanent
memory;andifitdoes,thetransitiontakestime.Untilthememoryis
consolidated,itisparticularlyfragile.Newmemoriesmaybedisruptedjustby
engaginginanotheractivity,andevenoldermemoriesarevulnerabletointense
experiencessuchasemotionaltraumaorelectroconvulsiveshocktreatment(a
meansofinducingconvulsions,usuallyintreatingdepression).Researchers
dividememoryintotwostages,short-termmemoryandlong-termmemory.

Long-termmemory,atleastforsomekindsoflearning,canbedividedintotwo
stagesthathavedifferentdurationsandoccurindifferentlocations(Figure12.2),
aswewillseelater(McGaugh,2000).

Mostmemories,likehumansandwines,donotmatureinstantly.Insteadtheyaregradually
stabilizedinaprocessreferredtoasconsolidation.
—YadinDudai

Ananimalstudyclearlydemonstratesthatthehippocampusparticipatesin
consolidation.Ratsweretrainedinawatermaze,atankofmurkywaterfrom
whichtheycouldescapequicklybylearningthelocationofaplatform
submergedjustunderthewater’ssurface(Figure12.3;Riedeletal.,1999).
Then,forsevendaystherats’hippocampiweretemporarilydisabledbyadrug
thatblocksreceptorsfortheneurotransmitterglutamate.Elevendayslater—
plentyoftimeforthedrugtocleartherats’systems—theyperformedpoorly
comparedwithcontrolsubjects(Riedeletal.).Researchershavebeenableto
“watch”theconsolidationhappeninginhumans,usingbrainscansandeventrelatedpotentials.Presentingwordsorpicturesactivatedthehippocampusand
adjacentcortex;howwellthematerialwasrememberedlatercouldbepredicted
fromhowmuchactivationoccurredinthoseareasduringstimuluspresentation
(Figure12.4;Alkire,Haier,Fallon,&Cahill,1998;Brewer,Zhao,Desmond,
Glover,&Gabrieli,1998;Fernándezetal.,1999).
Figure12.2StagesofConsolidation.


Source:Reprintedwithpermissionfrom“Memory—ACenturyof
Consolidation,”byJ.L.McGaugh,Science,287,pp.248–251.Copyright
2000AmericanAssociationfortheAdvancementofScience.

Application:TheLegacyofHM

Source:WikimediaCommons.
NotonlydidHenryMolaisondevotemuchofhislifetonumerousscientificinvestigations,but
hisbrainwillcontinuetobethesubjectofstudyformanyyearstocome(Lafee,2009).Soon
afterhisdeath,Molaison’spreservedbrainwasinaplasticcoolerstrappedinaseatonaflight
fromBostontoSanDiego;inthenextseatwasJacopoAnnese,directoroftheBrain
ObservatoryattheUniversityofCaliforniaatSanDiego.Afterseveralmonthsofpreparation,
AnneseandhiscolleaguesdissectedMolaison’sbrainintoslicesasthinasthewidthofahair


(70µm).The53-hour,uninterruptedprocedurewasrecordedandlive-streamedoverthewebto
allowscientificscrutinyandtoincreasepublicawarenessandengagement(Anneseetal.,2014).
EachsliceofHM’sbrainwasmicroscopicallyphotographedwithsuchresolutionthatthedata
fromeachonewouldfill200DVDs.Thedatawerethencombinedintoathree-dimensional
reconstructionofthebrain,whichisavailableonline.Scientistscannavigatethroughittothe
areaoftheirinterestandthenzoomintothelevelofindividualneurons.HM’smemory
problemsmadehimperhapsthemoststudiedsubjectinneuroscience.Ironically,themanwho
couldnotrememberwillneverbeforgotten.

Animalsthatweregiventheglutamate-blockingdrugatthetimeoftesting
insteadofimmediatelyaftertrainingalsohadimpairedrecallinthewatermaze,
indicatingthatthehippocampushasaroleinretrievalaswellasconsolidation.
ResearchershaveusedPETscanstoconfirmthatthehippocampusalsoretrieves
memoriesinhumans(D.L.Schacter,Alpert,Savage,Rauch,&Albert,1996;
Squireetal.,1992).Figure12.5showsincreasedactivityinthehippocampi
whiletheresearchparticipantsrecalledwordslearnedduringanexperiment.The
involvementofthehippocampusinretrievalseemsinconsistentwithHM’s
abilitytorecallearliermemories.Butthememoriesthatpatientswith
hippocampaldamagecanrecallareofeventsthatoccurredatleasttwoyears
beforetheirbraindamage.Manyresearchershaveconcludedthatthe
hippocampalmechanismplaysatime-limitedroleinconsolidationandretrieval,

apointwewillexamineshortly.Thisdiminishingroleofthehippocampus
wouldexplainwhyoldermemoriessufferlessthanrecentmemoriesafter
hippocampaldamage.
Figure12.3AWaterMaze.

Figure12.4HippocampalActivityRelatedtoConsolidation.


Source:From“PETImagingofConsciousandUnconsciousMemory,”by
M.T.Alkire,R.J.Haier,J.H.Fallon,andS.J.Barker,1996,Journalof
ConsciousnessStudies,3,pp.448–462.
Figure12.5HippocampalActivityintheHumanBrainDuringRetrieval.


Source:Reprintedwithpermissionfrom“ConsciousRecollectionandthe
HumanHippocampalFormation:EvidenceFromPositronEmission
Tomography,”byD.L.Schacteretal.,ProceedingsoftheNational
AcademyofSciences,USA,93,pp.321–325.Copyright1996National
AcademyofSciences,USA.
Theprefrontalareaisalsoactiveduringlearningandretrieval,andsome
researchersthinkthatitdirectsthesearchstrategyrequiredforretrieval(Buckner
&Koutstaal,1998).Indeed,theprefrontalareaisactiveduringeffortfulattempts
atretrieval,whereasthehippocampusisactivatedduringsuccessfulretrieval
(seeFigure12.5;D.L.Schacteretal.,1996).Wewilllookattheroleofthe
frontalareaagainwhenweconsiderworkingmemoryandKorsakoffsyndrome.
WesawinChapter5thatrewardssuchasdrugsincreaseactivityindopamine
neurons.Becauserewardplaysacrucialroleinlearning,wemightsuspectthat
dopaminehassomefunctioninlearning,andthatisindeedthecase.Blocking
dopaminereceptorsinguineapigsshortlyafterlearningimpairsconsolidation
andmemory(K.-N.Lee&Chirwa,2015),andincreasingdopaminelevelsby

injectingthedopamineprecursorlevodopaimprovesmemoryinhumans
(Chowdhury,Guitart-Masip,Bunzeck,Dolan,&Düzel,2012).Novel
experience,suchasexploringanunfamiliarenvironment,increasesdopamine
activity;placingratsinanovelenvironmentbeforeorafterlearningimproves
learning(S.-H.Wang,Redondo,&Morris,2010).Inhumans,learningcanbe
increasedbydopamine-enhancingstimulationassimpleasviewingnovel
photographsfromNationalGeographic(Fenker,Frey,Schuetze,&Heipertz,
2008).
Dopamine-enhancingstimulationiseffectivewhetheritoccursbeforeorafter
thelearningexperience;thisisbecausedopaminedirectlyaffectsconsolidation
oflong-termmemory,ratherthanbyimprovingshort-termmemory(Lisman,
Grace,&Duzel,2011).Dopaminereleaseinitiatesthesynthesisofproteinsin
thepostsynapticneuron.Theseplasticity-relatedproteinsarenecessaryfor
consolidationtooccur,aswewillseelater,anddrugsthatblocktheirsynthesis
inhibitlearning(Clopath,2012).
Dopaminedoesnotsignalrewardssomuchasitsignalserrorsinprediction.
Firingincreasesindopamineneuronsonlyiftherewardisunexpected—eitherof
greatervaluethanusualoroccurringwhenithasbeeninfrequent(Schultz,
2016).Iftherewardisexpected,thefiringrateremainsthesameanditdeclines
iftherewardislessthanexpected.Inotherwords,evolutionhastailored
learningspecificallytohelpuscopewithchangesinourenvironmentandinour
circumstances.

WhereMemoriesAreStored
Thehippocampalareaisnotthepermanentstoragesiteformemories.Ifitwere,
patientslikeHMwouldnotrememberanythingthathappenedbeforetheir
damageoccurred.Accordingtomostresearchers,thehippocampusstores
informationtemporarilyinthehippocampalformation;then,overtime,amore

permanentmemoryisconsolidatedelsewhereinthebrain.Astudyofmicethat
hadlearnedaspatialdiscriminationtasksupportedthehypothesis:Over25days
ofretentiontesting,metabolicactivityprogressivelydecreasedinthe
hippocampusandincreasedinthecorticalareas(Bontempi,Laurent-Demir,
Destrade,&Jaffard,1999).
Isthereaplacewherememoriesarestored?


Toexplorefurthertherelationshipbetweenthesetwoareas,Remondesand
Schuman(2004)severedthepathwaythatconnectsCA1ofthehippocampus
withthecortex.Thelesionsdidnotimpairtherats’performanceinawatermaze
duringtrainingor24hours(hr)later,butafter4weekstheratshadlosttheir
memoryforthetask.Theresultssupportedthehypothesisthatshort-term
memorydependsonthehippocampusbutlong-termmemoryrequiresthecortex
andaninteractionovertimebetweenthetwo.Topindownthewindowof
vulnerabilityofthememory,theresearcherslesionedtwoadditionalgroupsof
animalsatdifferenttimesfollowingtraining.Thoselesioned24hraftertraining
wereimpairedinrecallfourweekslater,butthosewhosesurgerywasdelayed
untilthreeweeksaftertrainingperformedaswellasthecontrols.This
progressionapparentlytakeslongerinhumans.ChristineSmithandLarrySquire
(2009)usedfMRItoimagethebrain’sactivitywhilesubjectsrecallednews
eventsfromthepast30years.Activitywasgreatestinthehippocampusand
relatedareasassubjectsrecalledrecentevents,declinedastheyrecalledevents
asfarbackas12years,andstabilizedafterthat.Atthesametime,activity
increasedprogressivelywitholdermemoriesintheprefrontal,temporal,and
parietalcortex.So,yourbrainworksratherlikeyourcomputerwhenittransfers
volatilememoryfromRAMtotheharddrive—itjusttakesalotlonger.
InChapter3,youlearnedthatwhenWilderPenfield(1955)stimulated
associationareasinthetemporallobesofsurgerypatients,heoftenevoked
visualandauditoryexperiencesthatseemedlikememories.Wespeculatedthat

memoriesmightbestoredthere,andmorerecentresearchhassupportedthat
idea,withmemoriesforsoundsactivatingauditoryareasandmemoriesfor
picturesevokingactivityintheoccipitalregion(Figure12.6;M.E.Wheeler,
Petersen,&Buckner,2000).YoualsosawinChapter9thatwhenwelearna
newlanguage,itisstorednearBroca’sarea.Namingcolors(whichrequires
memory)activatestemporallobeareasnearwhereweperceivecolor;identifying
picturesoftoolsactivatesthehandmotorareaandanareainthelefttemporal
lobethatisalsoactivatedbymotionandbyactionwords(A.Martin,Haxby,
Lalonde,Wiggs,&Ungerleider,1995;A.Martinetal.,1996);andspatial
memoriesappeartobestoredintheparietalareaandverbalmemoriesintheleft
frontallobe(F.Rösler,Heil,&Henninghausen,1995).Thus,allmemoriesare
notstoredinasinglearea,noriseachmemorydistributedthroughoutthebrain.
Rather,differentmemoriesareindifferentcorticalareas,apparentlyaccordingto
wheretheinformationtheyarebasedonwasprocessed.
Aninterestingexampleisthecellsinvolvedinplacememory.Placecells,which


increasetheirrateoffiringwhentheindividualisinaspecificlocationin
theenvironment,arefoundinthehippocampus.Eachcellhasaplacefield
(overlappingsomewhatwithothers),andtogetherthesecellsformamapofthe
environment.Thismapdevelopsduringthefirstfewminutesofexploration;the
cells’fieldsarethenremappedonenteringanewenvironment,buttheyare
restoredonreturningtotheoriginallocation(Figure12.7;Guzowski,Knierim,
&Moser,2004;M.A.Wilson&McNaughton,1993).Thefieldsaredependent
onspatialcuesintheenvironment,includingvisual,tactile,andevenolfactory
cues(Shapiro,Tanila,&Eichenbaum,1997).Placecellsdomorethanindicate
anindividual’scurrentlocation.Forexample,theycontributethecontextof
locationthatissoimportantinmemoriesofevents(D.M.Smith&Mizumori,
2006).Theyalsoprovidespatialmemoryrequiredforplanningnavigation;as
ratspausedatchoicepointsinamazewithwhichtheywerewellexperienced,

cellswithplacefieldsinthealternativesectionsfiredinsequence,asiftherats
weresimulatingthetwochoices(Johnson&Redish,2007).FunctionalMRIhas
confirmedthathumanshaveplacecells;theiractivityissoprecisethatthe
investigatorscoulddeterminethesubject’s“location”inacomputer-generated
virtualenvironment(Hassabisetal.,2009).
Figure12.6FunctionalMRIScansofBrainsDuringPerceptionandRecall.


Source:From“Memory’sEcho:VividRememberingReactivatesSensorySpecificCortex,”byM.E.Wheeleretal.,ProceedingsoftheNational
AcademyofSciences,USA,97,pp.11125–11129,fig.1c,d,e,f,p.11127.
©2000NationalAcademyofSciences,USA.

TwoKindsofLearning
LearningresearcherswereinforarevelationwhentheydiscoveredthatHM
couldreadilylearnsomekindsoftasks(Corkin,1984).Onewasmirrordrawing,
inwhichtheindividualusesapenciltotraceapatharoundapattern,relying
solelyonaviewoftheworksurfaceinamirror.HMimprovedinmirrordrawingabilityoverthreedaysoftraining,andhelearnedtosolvetheTowerof


Hanoiproblem(Figure12.8).Buthecouldnotrememberlearningeithertask,
andoneachdayofpracticehedeniedevenhavingseentheTowerpuzzlebefore
(N.J.Cohen,Eichenbaum,Deacedo,&Corkin,1985;Corkin,1984).Whatthis
means,researchersrealized,isthattherearetwocategoriesofmemory
processing.Declarativememoryinvolveslearningthatresultsinmemoriesof
facts,people,andeventsthatapersoncanverbalizeordeclare.Forexample,
youcanrememberbeinginclasstoday,whereyousat,whowasthere,andwhat
wasdiscussed.Declarativememoryincludesavarietyofsubtypes,suchas
episodicmemory(events),semanticmemory(facts),autobiographicalmemory
(informationaboutoneself),andspatialmemory(thelocationoftheindividual
andofobjectsinspace).Nondeclarativememoryinvolvesmemoriesfor

behaviors;thesememoriesresultfromproceduralorskillslearning,
emotionallearning,andstimulus-responseconditioning.Learningmirror
tracingorhowtorideabicycleorsolvetheTowerofHanoiproblemare
examplesofnondeclarativelearningor,morespecifically,proceduralorskills
learning;rememberingpracticingthetasksinvolvesdeclarativelearning.
Anotherwayofputtingit,whichisadmittedlyabitoversimplified,isthat
declarativememoryisinformational,whilenondeclarativememoryismore
concernedwiththecontrolofbehavior.Justaswehavewhatandwhere
pathwaysinvisionandaudition,wehaveawhatandahowinmemory.

Whatarethetwokindsoflearning?
Figure12.7RecordingsFromPlaceCellsinaRatinaCircularRunway.

Source:ReprintedbypermissionfromMacmillanPublishersLtd.From


“NeuralPlasticityintheAgeingBrain,”byS.N.BurkeandC.A.Barnes,
2006,NatureReviewsNeuroscience,7,pp.30–40.NaturePublishing
Group.
Themainreasontodistinguishbetweenthetwotypesoflearningisthatthey
havedifferentoriginsinthebrain;studyingthemcantellussomethingabout
howthebraincarriesoutitstasks.Foryears,itlookedlikewewerelimitedto
studyingthedistinctionintherarehumanwhohadbraindamageinjusttheright
place;hippocampallesionsdidnotseemtoaffectlearninginrats,soresearchers
thoughtthatratsdidnothaveanequivalentofdeclarativememory.Butitjust
tookselectingtherighttasks.R.J.McDonaldandWhite(1993)usedan
apparatuscalledtheradialarmmaze,acentralplatformwithseveralarms
radiatingfromit(Figure12.9).Ratswithdamagetobothhippocampicouldlearn
thesimpleconditioningtaskofgoingintoanylightedarmforfood.Butifevery
armwasbaitedwithfood,theratscouldnotrememberwhicharmstheyhad

visitedandrepeatedlyreturnedtoarmswherethefoodhadalreadybeeneaten.
Figure12.8TheTowerofHanoiProblem.

Conversely,ratswithdamagetothestriatumcouldrememberwhicharmsthey
hadvisitedbutcouldnotlearntoenterlightedarms.BecauseParkinson’s
diseaseandHuntington’sdiseasedamagethebasalganglia(whichincludethe
striatum),peoplewiththesedisordershavetroublelearningproceduraltasks,
suchasmirrortracingortheTowerofHanoiproblem(Gabrieli,1998).
Incidentally,thetermdeclarativeseemsinappropriatewithrats;researchershave
oftenpreferredthetermrelationalmemory,whichimpliesthattheindividual
mustlearnrelationshipsamongcues,anideathatappliesequallywelltohumans
andanimals.


Figure12.9ARadialArmMaze.

Source:MauroFermariello/ScienceSource.
Youalreadyknowthattheamygdalaisimportantinemotionalbehavior,butit
alsohasasignificantroleinnondeclarativeemotionallearning.Becharaandhis
colleagues(1995)studiedapatientwithdamagetobothamygdalaeandanother
withdamagetobothhippocampi.Theresearchersattemptedtoconditionan
emotionalresponseinthepatientsbysoundingaloudboathornwhenablue
slidewaspresentedbutnotwhentheslidewasanothercolor.Thepatientwith
amygdaladamagereactedemotionallytotheloudnoise,indicatedbyincreased
skinconductanceresponses(seeChapter8).Hecouldalsotelltheresearchers
whichslidewasfollowedbytheloudnoise,buttheblueslideneverevokeda
skinconductanceincrease;inotherwords,conditioningwasabsent.Thepatient
withhippocampaldamageshowedanemotionalresponseandconditioning,but
hecouldnottelltheresearcherswhichcolortheloudsoundwaspairedwith.
Thisneuraldistinctionbetweendeclarativelearningandnondeclarative

emotionallearningmaywellexplainhowanemotionalexperiencecanhavea


long-lastingeffectonaperson’sbehavioreventhoughthepersondoesnot
remembertheexperience.
Theamygdalahasanadditionalfunctionthatcutsacrosslearningtypes.Both
positiveandnegativeemotionsenhancethememorabilityofanyevent;the
amygdalastrengthensevendeclarativememoriesaboutemotionalevents,
apparentlybyincreasingactivityinthehippocampus.Electricalstimulationof
theamygdalaactivatesthehippocampus,anditenhanceslearningofanonemotionaltask,suchasachoicemaze(McGaugh,Cahill,&Roozendaal,1996).
Inhumans,memoryforbothpleasantandaversiveemotionalmaterialisrelated
totheamountofactivityinbothamygdalaewhileviewingthematerial(Cahillet
al.,1996;Hamann,Ely,Grafton,&Kilts,1999).

WorkingMemory
Thebrainstoresatremendousamountofinformation,butinformationthatis
merelystoredisuseless.Itmustbeavailable,notjustwhenitisbeingrecalled
intoawarenessbutwhenthebrainneedsitforcarryingoutatask.Working
memoryprovidesatemporary“register”forinformationwhileitisbeing
used.Workingmemoryholdsapasswordyoujustlookeduplongenoughfor
youtotypeitin;italsoholdsinformationretrievedfromlong-termmemory
whileitisintegratedwithotherinformationforuseinproblemsolvingand
decisionmaking.Withoutworkingmemory,wecouldnotdolongdivision,plan
achessmove,orevencarryonaconversation.
Whyisworkingmemoryimportant?
ThinkofworkingmemoryasliketheRAMinyourcomputer.TheRAMholds
informationtemporarilywhileitisbeingprocessedorused,buttheinformation
isstoredelsewhereontheharddrive.Butweshouldnottakeanyanalogytoo
far.Workingmemoryhasaverylimitedcapacity(withnoupgradesavailable),
andinformationinworkingmemoryfadeswithinseconds.So,ifyoumakea

mistakeenteringthepasswordyoujustlookedup,you’llprobablyhavetolookit
upagain.

Thepersonrecallsinalmostphotographicdetailthetotalsituationatthemomentofshock,
theexpressionofface,thewordsuttered,theposition,garments,patternofcarpet,recalls
themyearsafterasthoughtheyweretheexperienceofyesterday.


—G.M.Stratton,1919

Thedelayedmatch-to-sampletaskdescribedinChapter11providesanexcellent
meansofstudyingworkingmemory.Duringthedelayperiod,cellsremainactive
inoneormoreoftheassociationareasinthetemporalandparietallobes,
dependingonthenatureofthestimulus(Constantinidis&Steinmetz,1996;
Fuster&Jervey,1981;Miyashita&Chang,1988).Cellsintheseareas
apparentlyhelpmaintainthememoryofthestimulus,buttheyarenotthe
locationofworkingmemory.Ifadistractingstimulusisintroducedduringthe
delayperiod,thealteredfiringintheselocationsceasesabruptly,buttheanimals
arestillabletomakethecorrectchoice(Constantinidis&Steinmetz,1996;E.K.
Miller,Erickson,&Desimone,1996).Cellsintheprefrontalcortexhaveseveral
attributesthatmakethembettercandidatesasworkingmemoryspecialists.Not
onlydotheyincreasefiringduringadelay,buttheyalsomaintaintheincrease
despiteadistractingstimulus(E.K.Milleretal.,1996).Somerespond
selectivelytothecorrectstimulus(diPellegrino&Wise,1993;E.K.Milleret
al.).Othersrespondtothecorrectstimulus,butonlyifitispresentedina
specificpositioninthevisualfield;theyapparentlyintegrateinformationfrom
cellsthatrespondonlytothestimuluswithinformationfromcellsthatrespond
tothelocation(Rao,Rainer,&Miller,1997).Prefrontaldamageimpairs
humans’abilitytorememberastimulusduringadelay(D’Esposito&Postle,
1999).Allthesefindingssuggestthattheprefrontalareaplaysthemajorrolein

workingmemory.
Althoughtheprefrontalcortexservesasatemporarymemoryregister,its
functionisapparentlymorethanthatofaneuralblackboard.InChapters3and
8,youlearnedthatdamagetothefrontallobesimpairsaperson’sabilityto
governhisorherbehaviorinseveralways.Manyresearchersbelievethatthe
primaryroleoftheprefrontalcortexinlearningisasacentralexecutive.Thatis,
itmanagescertainkindsofbehavioralstrategiesanddecisionmakingand
coordinatesactivityinthebrainareasinvolvedintheperceptionandresponse
functionsofatask,allthewhiledirectingtheneuraltrafficinworkingmemory
(Wickelgren,1997).

ConceptCheck
TakeaMinutetoCheckYourKnowledgeandUnderstanding


Whatdeterminesthesymptomsandtheseverityofsymptomsofamnesia?
Describethetwokindsoflearningandtherelatedbrainstructures.
Workingmemorycontributestolearningandtootherfunctions.How?

BrainChangesinLearning
Learningisaformofneuralplasticitythatchangesbehaviorbyremodeling
neuralconnections.Specializedneuralmechanismshaveevolvedtomakethe
mostofthiscapability.Wewilllookattheminthecontextoflong-term
potentiation.

Howdoneuronschangeduringlearning?

Long-TermPotentiation
Morethan50yearsago,DonaldHebb(1940)statedwhathasbecomeknownas
theHebbrule:Ifanaxonofapresynapticneuronisactivewhilethe

postsynapticneuronisfiring,thesynapsebetweenthemwillbe
strengthened.Wesawthisprincipleinactionduringthedevelopmentofthe
nervoussystem,whensynapticstrengtheninghelpeddeterminewhichneurons
wouldsurvive;someofthatplasticityisretainedinthematureindividual.
Researchershavelongbelievedthattounderstandlearningasaphysiological
process,theywouldhavetofigureoutwhathappensattheleveloftheneuron
and,particularly,atthesynapse.
Long-termpotentiation(LTP)isapersistentstrengtheningofsynapsesthat
resultsfromthesimultaneousactivationofpresynapticneuronsand
postsynapticneurons(Cooke&Bliss,2006).LTPcanbeinducedinthe
laboratorybystimulatingbothneuronsatthesametime,orbystimulatingthe
presynapticneuronadequatelytocausethepostsynapticneurontofire.Asyou
canseeinFigure12.10a,thepostsynapticneuron’sresponsetoateststimulusis
muchstrongerfollowinginductionofLTP.WhatisremarkableaboutLTPisthat
itcanlastforhoursintissueculturesandmonthsinlaboratoryanimals(Cooke
&Bliss).LTPhasbeenstudiedmostlyinthehippocampus,butitalsooccursin
severalotherareas,includingthevisual,auditory,andmotorcortex.SoLTP


appearstobeacharacteristicofmuchofneuraltissue,atleastintheareasmost
likelytobeinvolvedinlearning.
Figure12.10LTPandLTDintheHumanBrain.

Source:From“Long-TermModificationsofSynapticEfficacyinthe
HumanInferiorandMiddleTemporalCortex,”byW.R.Chenetal.,
ProceedingsoftheNationalAcademyofSciences,USA,93,pp.8011–
8015.Copyright1996NationalAcademyofSciences,USA.Usedwith
permission.
Neuralfunctioningrequiresweakeningsynapsesaswellasstrengtheningthem.
Long-termdepression(LTD)isadecreaseinthestrengthofsynapsesthat

occurswhenstimulationofpresynapticneuronsisinsufficienttoactivate
thepostsynapticneurons(Cooke&Bliss,2006).Potentiationcanbedepressed
inapostsynapticneuronbyapplyingalow-frequencypulsetothepresynaptic
neuronforafewminutes,causingthepresynapticneurontofirebutnotthe
postsynapticneuron(Figure12.10b).LTDisbelievedtobethemechanismthe
brainusestomodifymemoriesandtoclearoldmemoriestomakeroomfornew
information(Stickgold,Hobson,Fosse,&Fosse,2001).


Activityinpresynapticneuronsalsoinfluencesthesensitivityofnearby
synapses.Ifaweaksynapseandastrongsynapseonthesamepostsynaptic
neuronareactivesimultaneously,theweaksynapsewillbepotentiated;this
effectiscalledassociativelong-termpotentiation(Figure12.11).Associative
LTPisusuallystudiedinisolatedbraintissuewithartificiallycreatedweakand
strongsynapses,butithasimportantbehavioralimplications,whichiswhyit
interestsus.Electricshockevokesastrongresponseinthelateralamygdala,
wherefearisregistered,whileanauditorystimulusproducesonlyaminimal
responsethere.Rogan,Stäubli,andLeDoux(1997)repeatedlypairedatonewith
shocktothefeetofrats.Becauseofthisprocedure,thetonealonebeganto
evokeasignificantlyincreasedresponseintheamygdala,aswellasan
emotional“freezing”responseintherats.Youmayrecognizethisscenarioasan
exampleofclassicalconditioning;wecouldeasilychangethelabelsinFigure
12.11from“Strongsynapse”to“Electricshock”andfrom“Weaksynapse”to
“Auditorytone.”ResearchersbelievethatassociativeLTPisthebasisof
classicalconditioning,andRoganetal.’sresultssupportthatview.LTP,LTD,
andassociativeLTPcanallbesummedupintheexpression“Cellsthatfire
togetherwiretogether.”

HowLTPHappens
LTPhasbeenstudiedmostoftenintheneuronsconnectingCA1andCA3ofthe

hippocampus,andwewillusethosefindingsasourmodelherewithoutgoing
intothevariationsthatoccurinotherareasofthebrain.LTPisinducedthrougha
cascadeofeventsatthesynapse.InCA1(andinmostlocations)the
neurotransmitterinvolvedisglutamate,whichisdetectedbytwotypesof
receptors:theAMPA(alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic
acid)receptorandtheNMDA(N-methyl-d-asparticacid)receptor.Initially,
glutamateactivatesAMPAreceptorsbutnotNMDAreceptors,becausetheyare
blockedbymagnesiumions(Figure12.12).DuringLTPinduction,activationof
theAMPAreceptorsbythefirstfewpulsesofstimulationpartiallydepolarizes
themembrane,andthisdislodgesthemagnesiumions.Theresultinglargeinflux
ofcalciumionsactivatesahostofproteinkinases,enzymesthatalteroractivate
otherproteins(Lüscher&Malenka,2012).Oneoftheproteinkinases,CaMKII
(calcium/calmodulin-dependentkinaseII)isrequiredforLTP.Micewithtwo
mutant,nonfunctioninggenesforthealphaformofCaMKIIfailtoshowLTP;
thosewithonemutantandonefunctioninggenedoshowLTP,butitisnot
consolidatedintolong-termmemory(Figure12.13;Frankland,O’Brien,Ohno,


Kirkwood,&Silva,2001).Severalplasticity-relatedgenesareactivatedaswell
(Kelleher,Govindarajan,&Tonagawa,2004);theyalongwithnewlyactivated
proteinsbeginproducingstructuralchangesinthesynapse(Lüscher&
Malenka).
Figure12.11AssociativeLTP.

FiguresBroughttoLife
Figure12.12ParticipationofGlutamateReceptorsinLTP.


FiguresBroughttoLife
Figure12.13RetentioninNormalandαCaMKII-DeficientMiceOverTime.

Source:ReprintedbypermissionfromMacmillanPublishersLtd.From
“αCaMKII-DependentPlasticityintheCortexIsRequiredforPermanent
Memory,”byP.W.Frankland,C.O’Brien,M.Ohno,A.Kirkwood,&A.J.
Silva,2001,Nature,411,pp.309–313.Figure1.NaturePublishingGroup.
Within45–60minutesafterLTP,postsynapticneuronsdevelopdramatically
increasednumbersofdendriticspines,outgrowthsfromthedendritesthat
partiallybridgethesynapticcleftandmakethesynapsemoresensitive
(Figure12.14;N.Toni,Buchs,Nikonenko,Bron,&Muller,1999).Existing
spinesalsoenlargeorsplitdownthemiddletoformtwospines(Matsuzaki,
Honkura,Ellis-Davies,&Kasai,2004).Anotherimportantstructuralchangeis
theappearanceofnewAMPAreceptors,whichincreasesynapticstrength
(Lüscher&Malenka,2012).Thesecomefromapoolofsilentreceptorsthatare
transportedintothespinesfromwithinthedendrite;theycanrecyclebetween
thecytoplasmandthemembraneorintheotherdirectionwithinmeretensof
minutes.Afurtherchangethatoccursinsupportoflearningisthegenerationof
newneuronsinthehippocampus;thoughtherateofneurogenesisisrelatively
lowinadults,overthelifespannewneuronsadduptoanestimated10%–20%
ofthepopulation(Jacobs,vanPraag,&Gage,2000).Numerousstudiesshow
thatlearningisimpairedbyblockingneurogenesisandenhancedbyincreasing
newcellbirth.Newneuronsaremoreactivethanmatureones,havealower
thresholdforLTPinduction,andarebetteratmakingfinediscriminations,such