Bechtel.the Endogenously Active Brain

7/28/2019 Bechtel.the Endogenously Active Brain

1/22

TheEndogenouslyActiveBrain:

TheNeedforanAlternativeCognitiveArchitecture

WilliamBechtelDepartmentofPhilosophy,CenterforChronobiology,and

InterdisciplinaryPrograminCognitiveScienceUniversityofCalifornia,SanDiego

Abstract

Mostproposalsofcognitivearchitecturesincognitivescienceandaccountsofbrainprocessesinneuroscienceconstruethemind/brainasreactive:processingisinitiatedbyastimulusandterminatesinaresponsetoit.Butthereisgrowingevidencethatbrainsareendogenouslyactive:oscillationsinelectrochemicalactivityatmultiplefrequenciesareongoinginthebrainevenintheabsenceofstimuliandstimuliservetomodulatetheseoscillationsratherthaninitiateactivity.Moreover,evidenceisgrowingthatthisendogenousactivityisusedinvariousinformationprocessingactivities.Iappealtoevidencefromsingle-cellrecording,EEG,andrestingstatefMRItosupporttheclaimofongoingoscillatorybehaviorinthebrainandidentifyseveralwaysitmaycontributetocognition.Ifcognitivescienceistounderstandhowweperformcognitivetasksitneedstodevelopcognitivearchitecturesthatincorporatethesortofendogenousdynamicactivityexhibitedbythebrain.

1.TheSearchforCognitiveArchitecturesThecognitivetraditionisdistinguishedfromitsbehavioristpredecessorbyfocusingoninformationprocessingmechanismsthatarethoughttoexistwithinthemind-brainand

hypothesizedtoexplainbehavior.Cognitiveresearchershaveoftenassumed,eitherimplicitlyorexplicitly,thatthebraintakesininformationthroughthesenses,representsit,performsoperationsontherepresentations,andrespondsbyeitherchangingitsinternalstateorplanningandexecutingactions.Intheearlydecadesofcognitivescienceitwasdifficulttoidentifytheneuralprocessesthatservedasrepresentationsforhigh-levelcognitiveprocessessuchasmemory,reasoning,andproblemsolvingandtheneuraloperationsthroughwhichtheywereprocessed.Asaresult,mostcognitivetheoriesofinformationprocessinghadtorelyonindirectmeasuressuchasreactiontimesanderrorpatternstoguideandevaluatehypothesesastotherepresentationsemployedandtheoperationsperformedonthem.Forexample,Sternberg(1966)usedthetimerequiredforsubjectstodeterminewhetheratestitemwasonalistofitemstheyhadmemorizedto

determinethathumansperformexhaustiveserialsearch.Onewaycognitivescientists,especiallythosefocusedoncomputationalmodelingofcognition,furtherconstrainedtheirinquirywastodevelopproposalsastothenatureofthecognitivearchitecture.Acognitivearchitecturespecifiestheprimitiveoperationsthemind/brainisthoughttoperform.ProposedarchitecturessuchasNewellsSOAR(Laird,Newell,&Rosenbloom,1987)andAndersonsACT-R(Anderson,1990,2007)weredefendedontheoreticalgroundsthattheypossessedtheappropriateprimitivecapacities

7/28/2019 Bechtel.the Endogenously Active Brain

2/22

forcognitivesystemssuchasfoundinhumans.Forthoseskepticaloftheproposalthatthemindperformsoperationsonsymbolstructures,theparalleldistributedprocessingframeworkorconnectionismprovidedanalternativeclassofarchitectures(Rumelhart&McClelland,1986),emphasizinginparticularfeedforwardnetworkstrainedbybackpropagation.(Foranoverviewandtaxonomyofcognitivearchitectures,seeDuch,

Oentaryo,&Pasquier,2008.)Researchersworkingwithaparticulararchitectureconstrainthemselvestoaccountforvariousaspectsofhumancognitiveactivityusingtheprimitiveoperationsprovidedinthearchitecture.Noneofthesearchitectureswasoriginallydevelopedprimarilywithaneyetocharacterizingtherepresentationsandprocessesusedinthebrain(althoughconnectionistssometimesrefertotheirapproachasbrain-stylecomputing),butastechniqueshavebeendevelopedtorelatecognitiveprocessingtoneuralactivity,advocatesofsomecognitivearchitectureshavetriedtoshowthattheirarchitecturefitswithourunderstandingofhowthebrainfunctions.Thesecognitivearchitecturestypicallyadoptareactiveperspectiveonthemind/brain.Cognitiveactivityisassumedtobeginwiththepresentationofataskorstimulus,whichis

representedandtherepresentationisthentransformedviaoperationsspecifiedbythearchitecture.Thisreactiveconceptionofcognition(itoccursinresponsetoastimulus)hasalsobeensharedastheneurosciencesbegantoprovideinsightintotherepresentationsandoperationsperformed.Thefirstsuccessesinidentifyingneuralprocessesthatrepresentinformationresultedfrominvestigationsofsensoryandmotorprocessinginwhichitwaspossibletolinkbrainactivity(typicallyspikingratesofneurons)withsensorystimuliormotoractivities.Withrespecttovisualprocessing,forexample,researchersbeginningwithKuffler(1953)andHubelandWiesel(1962,1968)employedsuchtechniquesassingle-andmulti-cellrecordingtodeterminewhatfeaturesofvisualstimuliwerecorrelatedwithspecificneuronalactivity.Theactivityoftheseneuronswasthenviewedasrepresentingthecorrelatedfeaturesofthevisualstimulus,andresearchers

hypothesizedoperationsthroughwhichtheserepresentationsweresuccessivelytransformedinahierarchyofprocessingareas(vanEssen&Gallant,1994;forananalysisofthishistory,seeBechtel,2008).Inthe1970sresearchesdevelopedtechniquesforrelatingtheelectricalsignalrecordedatthecorticalsurface(electroencephalographyorEEG)tostimuluspresentation(measuringwhatareknownasEvokedResponsePotentialsorERPs).WhileERPstudiesprovidedlittleinformationaboutthespatiallocusofactivity,theyofferedinformationaboutthetemporalorderofprocessing.Forexample,KutasandHillyard(1980)foundthatwhenthelastwordofasentencewasanomaloustheEEGexhitibedanegativedeflectionpeakingabout400millisecondslater(hence,anN400response).TheintroductionofPositronEmissionTomography(PET)inthelate1980sandfunctionalMagneticResonanceImaging(fMRI)inthe1990sallowedresearcherstobegin

tolocatecognitiveoperationsinbrainregionsinhumans.Petersen,Fox,Posner,Mintun,andRaichle(1988),forexample,adaptedthesubtractivemethod,developedbyDondersinthemid-19thcenturyforidentifyingthetimerequiredforacognitiveoperation,toidentifythebrainregionsresponsibleforsuchoperations.Thus,intheirpioneeringPETstudytheysubtractedthebloodflowmeasuredwhensubjectssimplyreadanounaloudfromthatmeasuredwhentheyfirstgeneratedarelatedverbandreaditaloud.Theyfoundincreasedactivityintheleftprefrontalcortex(alsothecerebellumandtheanteriorcingulate),whichtheyarguedtobethelocusofthesemanticprocessingrequiredforthetask.

7/28/2019 Bechtel.the Endogenously Active Brain

3/22

Researchersadoptingthereactiveperspectivehaveprovidedagreatdealofinformationaboutinformationprocessinginthebrain,especiallyinareasofsensoryandmotorprocessingandincreasinglywithrespecttomemory,attention,andemotionalresponses.Buttherearereasonstobeskepticalabouttheadequacyofthereactiveperspective.Since

thepioneeringstudiesofLorentedeN(1938),researchershaverecognizedthatthereareatleastasmany,andlikelymanymore,backwardsandcollateralprojectsthanforwardones.1Thereactiveperspective,however,hasbeenabletoprovidelittleinsightintowhatcontributiontoinformationprocessingtheseoperationsperformsincewhatevercontributiontheymakehasalreadyaffectedhowbrainareasrespondtostimuliandtypicallycannotbeseparatelyidentified.(Oneexceptionhasbeentheinvestigationofhowattentionmodifiesprocessingofperceptualstimuli.SeeCorbetta&Shulman,2002.)Althoughithasreceivedfarlessattention,thereisanalternativetraditioninbrainresearchthathasemphasizedtheendogenousactivityofthebrain,viewingstimuliasperturbingon-goingbrainactivity,notinitiatingactivity.ThistraditiontracesbackatleasttoThomas

GrahamBrown(1911,1914),whostudiedneuralmechanismsformotorbehaviorindecerebratecatsintheLiverpoollaboratoryofCharlesScottSherrington,achiefproponentofadoptingareactiveperspectiveonthebrain.Bothwereinterestedinmotoractivitysuchaswalkingwhichappeartooriginatefrominsidetheorganism.Sherrington(1923)extendedthereactiveframeworktoexplainsuchbehaviorintermsofasequentialreflexmechanism,bywhichperipheralinput(e.g.,tothecatsfeetwhenplacedonamovingtreadmill)producedasequenceofneuralsignals(tothespine,withinthespine,andouttoflexorandthenextensormuscles).Eachcycleofsteppingresultedinrenewedinput(sensoryfeedback)andhenceongoing,rhythmicsteppingmovements.Browndiscoveredthathecouldobtainsimilarrhythmicsteppingevenafterisolatingthespinalcordfromafferent(peripheral)inputbycuttingthedorsalrootnerves.Heaccountedforitsrhythmic

outputsbyproposingacouplednetworkofspinalneuronsoneforflexionandoneforextensionthateachinhibitedtheothersbehaviortogenerateoscillations.Brownsproposalsanticipatedresearchonwhatcametobeknownasthecentralpatterngenerator(Wilson&Wyman,1965).Morerecently,Brownsperspectiveonendogenousactivityhasbeenadvancedbyresearchersusingtheverytoolsemployedbythereactivetraditionsingle-cellrecording,EGG,andfMRI.InthispaperIwillarguethattheevidencefortheendogenouslyactiveperspectiveonthemind-brainisextremelycompellingandthatinlightofitcognitiveresearchersshouldfundamentallyreconceivetheirconceptionsofcognitionandcognitivearchitecturestoincorporateandrecognizethesignificanceofendogenousactivity.Ibeginbyconsidering

whatitmeanstoclaimthatabiologicalmechanismexhibitsendogenousactivityandfocusonhowindividualneuronsareendogenouslyactive,thenexplorehowEEGandfMRIare1ForthemostpartinthispaperIwillfocusonconnections,forward,recurrent,andcollateralwithinthecerebralcortexandhippocampalformation.Butthesepresentonlyaportionofthestory.Mostsensoryinformationreachescortexviaregionsinthethalamus,andtherearealmosttentimesasmanyrecurrentprojectionsfromcortexbacktothesethalamicregionsasthereareforwardprojection.

7/28/2019 Bechtel.the Endogenously Active Brain

4/22

revealingsynchronousactivityofpopulationsofneuronsandinspiringproposalsastohowtheyfigureinneuralprocessing.Ithenconcludebyemphasizinghowthisbodyofresearchpointstotheneedforfundamentalrevisionsinthecognitivearchitecturesemployedincognitivescience.

2.EndogenouslyActiveMechanismsintheBrainClaimingthatthebrainisendogenouslyactivemaystrikesomeascomparabletoproposingthatitisaperpetualmotionmachine.Thatis,however,farfromwhatisbeingproposed.Alllivingorganisms,andaccordinglythosewithanervoussystemandabrain,areopeninthethermodynamicsensetomatterandenergyfromtheirenvironment.Whatisdistinctiveaboutbiologicalorganismsisthattheyareorganizedsystemshencenotinequilibriumwiththeirenvironmentandthattheymaintainthemselvesinthisnon-equilibriumstatedespitethetendencyexhibitedbyclosedsystemstowardsequilibrium(highentropy).Theequilibriumtendenciesaremanifestinthecontinualdegradationoforganicstructuresthatrequirelivingsystemsregularlytorepairthemselves(Rosen,1991)orelseceasetoexistas

organizedsystems.Moreover,theymustperformtheoperationsneededtoconstructthemselvesincorporatematterintotheorganizedstructurewithwhichtheyareidentified.MaturanaandVarela(1980)havereferredtothisasautopoiesisandvieweditasthefoundationofcognition(seeLyon,2006,forasystematicdiscussion).Crucialtoautopoiesis,butnotemphasizedbyMaturanaandVarela,isthatorganismsmustcaptureandemployfree-energyfromtheirenvironmentsinautopoiesisandrepair.Ruiz-MirazoandMoreno(2004)havemadethisacentralfeatureintheircharacterizationoflivingorganismsasautonomousandBarandiaranandMoreno(2006)haveextendedtheviewtocognition.Organismsmustinitiatetheseactivitiesofautopoiesisandrepairfromwithin,andthis

entailsthattheybeendogenouslyactive,notreactive.AsGnti(2003)emphasizedinhisproposalofthechemotonastheminimalchemicalsystemcapableofexhibitingthecharacteristicsoflife,cyclicprocessesthatcanregularlyreturntothesameconditionarefundamentaltoautonomoussystemsasaresultofcyclicorganization,anorganismcanregularlyrestoreitselftotheconditionswhereitcanperformtheoperationsnecessarytobuildandrepairitself.Aslongastheorganizationofthecyclicallyorganizedsystemisadequatetorecruitfreeenergyfromtheenvironment,itcancontinuetoiteratethestagesinthecycle.Ifonetracksvariablesrepresentingstatesofthesystemthroughtime,theywilloscillate.Iftherearetimedelayswithinthecyclicsystem,andespeciallyiftheoperationswithinthecyclearenon-linear,theseoscillationscanbesustainedindefinitelyprovidedsufficientmatterandenergyareavailable.Goodwin(1963)pioneeredtheanalysisof

sustainedoscillationsinbiologicalsystems,showinginacomputationalmodelhowafeedbackloopinspiredbyJacobandMonods(1961)operonmodelforgeneregulationinbacteriacould,withappropriateparameters,generateasustainedoscillatorysystem.Subsequently,othertheoristssuchasGoldbeter(1996)havepursuedtheapproachandhavedrawnattentiontoasubstantialnumberofendogenousoscillationsinbiologicalsystems.(ForfurtherdiscussionseeBechtel&Abrahamsen,2011.)

7/28/2019 Bechtel.the Endogenously Active Brain

5/22

Whilemechanismsthatareendogenouslyactiveclearlyoccurinnatureandarefundamentaltolivingsystems,manyoftheaccountsofmechanismandmechanisticexplanationadvancedbythenewmechanisticphilosophersofscienceareinadequatetocharacterizethem.Followingwhathasbeenthedominantemphasisinmechanisticresearchinbiologyoverthepastseveralcenturies,theseaccountsemphasizethepartsand

operationsofmechanismsandtheprocessesbywhichresearchersdecomposemechanismsintotheirpartsandoperations(Bechtel&Richardson,1993/2010;Bechtel&Abrahamsen,2005;Glennan,1996,2002;Machamer,Darden,&Craver,2000;Thagard,2003;Wimsatt,2007).Theseaccountsofmechanismalsonotetheimportanceofhowcomponentsareorganizedbut,likemanyscientists,theytendtoprivilegesequentialorganization.Thus,Machamer,Darden,andCraverincludeintheirdefinitionofamechanismthattheyareproductiveofregularchangesfromstartorset-uptofinishorterminationconditions(p.3).Sequentialorganization,however,cannotproduceendogenousactivitythisrequiresnegativefeedbackorothercyclicdesigns.Moreover,notallnegativefeedbacksystemssustainoscillations;onlyoneswithnon-linearcomponentsandappropriateparametervaluescandoso.Determiningwhethera

particularmechanismwillgeneratecontinuedactivity(e.g.,oscillation)orsettleintoastablestaterequiresemployingthetoolsofcomputationalmodelinganddynamicalsystemsanalysis.BechtelandAbrahamsen(2010;seealsoBechtel,2011)designateexplanationsthatinvokecomputationalmodelingtounderstandpatternsofchangeovertimeinthepropertiesofthepartsandoperationsofamechanismasdynamicmechanisticexplanations.Sustainedoscillatorsarethesimplestendogenouslyactivemechanismsaslongastheycanrecruitfreeenergyfromtheirenvironmenttheyarecontinuallyactive.Neuronsareexamplesofsuchsustainedoscillators.Theendogenousnatureofneuronalactivity,however,isoftennotappreciated.Sinceearlyinthe20thcentury,whentechniqueswere

developedforrecordingtheelectricalactivityofindividualneurons,neuroscientistshaveoftentreatedthegenerationofactionpotentialsastheprincipalactivityofneurons.Thecommonpicture,bothintextbooksandinphilosophicaldiscussions,isthatneuronsresideattheirrestingpotentialuntiltheyreceivesufficientinputsontheirdendritestoraisethevoltageabovethreshold,afterwhichtheygenerateanactionpotentialontheiraxonand,overtime,returntotheirrestingpotential.Onthisview,neuronsarereactivecomponents.Althoughmostaccountsofneuralprocessingfocusonactionpotentials,theyrepresentonlyasmallpartoftheelectricalactivityinthebrain.Evenbeforeindividualactionpotentialscouldberecorded,researcherssuchasduBois-Reymond(1848-1884)hadrecordedelectricalpotentialsfrommusclecellsandneurons.Bernstein(1912)hadidentifiedagreaterconcentrationofpotassium(K+)ionsinsidethaninsidethecell,and

proposedthattheelectricalcurrentresultingfromdiffusionoutofthecellexplainedthenegativepotentialofthecellatrest.Hefurtherproposedthat,whenexcited,otherionswoulddiffuseacrossthemembrane,eliminatingthediffusionpotentialoftheK+ions.HodgkinandHuxley(1952)determinedthatinadditiontothepotassiumgradient,thereisasodium(N+)gradient(withN+concentrationsgreateroutsidethecellduringtherestingpotentialphase)andthattheconductanceofeachionvariesindependentlyanddependsonthevoltageacrossthemembrane.Exceptwhentherelativeconcentrationofagivenionon

7/28/2019 Bechtel.the Endogenously Active Brain

6/22

eachsideofthemembrane(itsmembranepotential)equalsitsreversalorequilibriumpotential,ionswillfloweitherintooroutofthecell.HodgkinandHuxleysinterestinthesecurrentswasprincipallytoexplaintheactionpotential,whichmostresearcherscontinuedtoviewfromareactiveperspective.Butsome

researchers,workingasHodgkinandHuxleyhadwithinvertebrateneurons,foundspecializedpacemakerneuronsthatgeneratedtheirownrhythmicactionpotentials(Alving,1968).Othersbegantofocusonthevarietyofvoltage-gatedandothercurrentsthatcouldbeobservedacrossneuronalmembranesandexploredthecomplexpatternsofchangeinthesecurrentsnotonlyinaxonsbutalsointhedendritesandcellbodiesofneurons(reviewedbyKandel,1976).Initiallymammalianresearchers,whofollowedinthetraditionofSherringtonandtreatedneuronsasintegratinginputsandfiringwhentheseinputspushedthemabovetheirthreshold,viewedthesestudiesskeptically.Thisbegantochangeinthe1970sand1980swhenLlinsandhiscollaboratorsfoundavarietyoffunctionallyimportantioncurrentsin

neuronsoftheinferioroliveandcerebelluminmammalsandbirds.Mostwerespatiallydistributedandgatedbyvoltageinadifferentmannerthanthesodiumandpotassiumchannelsintheaxon,equippingthemforfunctionsotherthanthedirectgenerationofactionpotentials.Notably,thedendriteswereendowedwithchannelsprovidinghigh-thresholdconductancetocalcium(Ca2+)ions,enablingdynamicallycomplexdendriticexcitationincontrasttoearlierassumptionsofpassivetransmissionofsignalsfromsynapses.2Moreover,thecellbodiesofsomeneuronsintheinferiorolivehadadifferentkindofcalciumchannelwithaseeminglyparadoxicallow-thresholdconductancethat,ininteractionwithsodiumandhigh-thresholdcalciumconductances,enabledtheseneuronstofunctionassingle-celloscillatorscapableofself-sustainedrhythmicfiringindependentofsynapticinput(Llins,1988,p.1659).3Theysenttheserhythmicactionpotentialsto

targetneuronsinthecerebellumthatwereabletorespondatthesamefrequency,qualifyingthemasresonatorsinthedynamicallexiconchampionedbyLlinsreacting,butinwaysshapedbytheirinternalproperties.Llinsalsoinvestigatedspontaneousoscillationsinelectricalpotentialselsewhereinthebrain.Llinsfindingsrevealedthattheneuronsareoscillatorssomeofwhichmaintainoscillationsontheirownwhileothersresonatetooscillationsinitiatedbyothers.Thishasconsequencesnotjustforhowweconceiveofneuronsbuthowweunderstandtheirinteractions.AsHuygenshadobservedinhis1665letterstodeSluse(lettersno.1333of24February1665,no.1335of26February1665,no.1345of6March1665publishedinHuygens,1888),aslongasthereisameanstoconveyasignal(evenaveryweakone)

betweentwopendulumclocks(apendulumclockisanoscillator)withnearlythesame2ThislinkednerveexcitabilitywiththeCa2+-dependentsecondmessengersystemthatisimportantforregulatinggeneralcellularfunctions.3Forfurtherexposition,seeBuzski(2006,pp.181-183),whocomments:Thesefindings...illustratethatnaturewenttoalotoftroublebringingtogetherthesechannelsattherightdensitiesandlocationjusttoserveonepurpose:oscillation.Forevidenceextendingthefindingstosensoryneuronsinvariousmammalianspecies,seeHuguenard(1996).

7/28/2019 Bechtel.the Endogenously Active Brain

7/22

periods,theytendtosynchronizetheiroscillationsandexhibitacommonphaserelation.Whensignalscanbepassedbetweenmultipleoscillators,somewithdifferentnaturalfrequencies,complexpatternsofactivitycandevelop.Agivencomponentmaysynchronizewithsomebutoscillateoutofphasewithothercomponents.

Emphasizingtheoscillatorydynamicsacrossneuralmembranesisnotinconflictwithfocusingonthemasgeneratingactionpotentialsinresponsetostimuli.Inputsreducetheelectricalpotentialoftheneuronandactionpotentialsarisewhentheycrossathreshold.Iftheelectricalpotentialofneuronsisoscillating,therewillbeperiodswhentheyarelesspolarizedandhenceclosertothresholdandotherperiodswhentheyaremorepolarized.Aninputthatmaysufficetopushneuronsacrossthethresholdwhentheyarehypopolarizedmaybeinsufficienttodosowhentheyarehyperpolarized.Moreover,whentwoneuronsarelinkedinacircuitandtheiroscillationsaresynchronized,theperiodswhenthefirstneuronismostlikelytogenerateanactionpotentialwillcorrespondtoperiodswhenthesecondislikelytogenerateanactionpotentialinresponse.Ontheotherhand,whentheyaredesynchronized,thesecondneuronwillbelesslikelytogeneratean

actionpotentialatthetimewhenthefirstneuronismostlikelyto.Iturnnowtoconsideringhowsuchdynamicalbehaviorofneuronsismanifestinpopulationsofneuronsengagedincognitivetasks.3.OscillationsDetectedwithEEGandLFP

Avarietyoftechniqueshavebeendevelopedthatcandetectsynchronizedelectricalpotentialsofpopulationsofneurons.Oneofthefirstinvolvedelectrodes,initiallyinsertedintothescalpandsubsequentlyplacedonthescalp.InhispioneeringresearchinwhichhecoinedthenameElektrenkephalogramm(electroencephalogramorEEGinEnglish)forthetechnique,Berger(1929,1930)reportedthathecouldrecordelectricaloscillationsfrom

leadfoilelectrodesaffixedtotheheadsofhumanbeings.Inparticular,hedetectedlarge-amplitudeoscillationsofapproximately10Hzwhensubjectswereawakewiththeireyesclosedthathecalledalpharhythmsandsmalleramplitude,fasteroscillations(20-30Hz)thatbecameapparentoncealpharhythmsdeclinedwhensubjectsopenedtheireyes,receivedinputfromanothersense,orperformedanattentiondemandingtask(seeFigure1).

7/28/2019 Bechtel.the Endogenously Active Brain

8/22

Figure1.Inthiseight-secondextractofarecordingmadebyBerger(1930),theupperlineisasubcutaneousEEG.Itshowsthreesecondsofpredominantlyalphawavesthatwere

blocked0.27secondsafterhestrokedthesubjectshandwithaglassrod(indicatedbyarrowBonthe10Hztimingsignalatthebottom).ForatleastthenexttwosecondstheEEGshowslower-amplitude,higher-frequencybetawaves,notalphawaves.Themiddlelineisanelectrocardiogramrecordedsimultaneously.ExtractedfromFigure5inGloorstranslation(1969,p.82)ofBerger(1930).

BergerandothersinitiallyassumedthattheEEGwasgeneratedfromactionpotentials.Eventuallyresearchersrecognizedthattheyresultednotfromactionpotentialsbutfromsynchronizedpost-synapticpotentialsindendritesdiscussedintheprevioussection(Bremer,1958).Moreover,researchsuchasthatofJahnsenandLlins(1984)onthethalamusandthalamocorticalrelayneuronshelpedlinkdynamicbehaviorofindividualneuronstothelarge-scaledynamicsseeninEEG.InthedecadesafterBergerspioneeringwork,researchersidentifiedbothfasteroscillations(greaterthan30Hz)thattheydesignatedgammarhythmsaswellasavarietyofsloweroscillations(delta,0-4Hz,andtheta,4-8Hz,rhythms).TypicallyoscillationsatdifferentfrequenciesareallcombinedintheelectricalactivityresearchersrecordandtechniquessuchasfastFourieranalysisarerequiredtodifferentiatecomponentsintheoverallsignal.Moreover,researchershavealsofoundthatthesameactivitycanbedetectedfromelectrodesimplantedintobraintissueaslongastheelectrodesarenottooclosetoanygivenneuron.Whendetectedinthiswaythecurrentsarereferredtoaslocalfieldpotentials.ThemostcommonapplicationofthevariousrhythmsdetectedwithEEGhasbeentodifferentiateoverallcognitivestatesstatesofactiveawareness(gamma),quietresting(alpha),andvariousstagesofsleep(delta,theta).FormanyyearslittleprogresswasmadeinlinkingEEGoscillationstomorespecificallycharacterizedcognitivefunctions.Instead,theprimaryuseofEEGincognitiveresearchhasbeeninERPstudiesthatreflectthereactiveconceptionofneuralprocessing.Thedifferenceintheoverallelectricalactivityproducedduetothestimulusisquitesmallinrelationtotheelectricalactivitythat

7/28/2019 Bechtel.the Endogenously Active Brain

9/22

originateseitherendogenouslyorinresponsetootherstimuli,soERPresearchersmustaverageovermanytrialsinordertoextractasignal.Ironically,theyaretherebyremovingfromconsiderationtheendogenousbackgroundactivitythatBergerhadfirstidentified.Inrecentdecades,however,someresearchershavebeguntoidentifycorrelatesof

cognitiveactivitydirectlyintheEEGsignalitself.Oneofthefirsthypothesesresultedfromthedetectionofgammaoscillationsinvisualprocessingareasinanesthetizedcatspresentedwithmovingbarswithaparticularorientation(Gray,Konig,Engel,&Singer,1989).Singerandhiscolleaguesdevelopedaninterpretativeframeworkinwhichgammaoscillationsservedtotemporarilybindtogethertheneuralrepresentationsofdifferentfeaturesofaparticularobjectwhicharerepresentedintheactionpotentialsofneuronsindifferentvisualprocessingregions(Singer,1999,characterizedthisasthebinding-by-synchronizatonhypothesis).Subsequentresearchershavedetectedgammaoscillationsinawake,alertcats,monkeys,humans,andotherspecies,inresponsetothepresentationofshapesandsmoothlytransformingshapes(reviewedbyFries,2009).Gammaoscillationshavealsobeenfoundinsomatosensoryandauditoryprocessingareas.Recentstudieshave

shownthatgammasynchronizationrequiresnotjustinputsbuttop-downactivationduetoattention.Thesegammaoscillationsappeartobetheresultofinitialinputstopyramidalcellsthatthenelicitresponsesinbasketcellsthatdistributeaninhibitorysignalthatblocksfurtheractionpotentialsinthepyramidalcellsonwhichtheysynapseuntiltheireffectswearoffinasynchronizedmanner,allowingsynchronizedspikingbehavior.Thepopulationofbasketcellsarethenentrainedtomaintainaregularoscillationwhichleaveswindowsinwhichactionpotentialscanbegeneratedinthepyramidalcells(Hasenstaub,Shu,Haider,Kraushaar,Duque,&McCormick,2005).Theeffectofsynchronizedactionpotentialsallowstheoutputsofmanycellstobereceivedbyadownstreamneuroninacommontemporal

windowsothattheysumsufficientlytoproduceanactionpotentialinit.Binding-by-synchronizationretainsareactiveconceptionofbrainprocessingsinceitrepresentsthefeatureselicitedbyastimulusascomingtobebound.Butotherinvestigationsofgammaoscillationsuggestthattheymayreflectanendogenouscomponentofbrainprocessing.Whittington,Traub,andJefferys(1995)foundthatpharmacologically-isolatedinhibitoryneurons,whenprovidedwithatonicexcitatorydrive(activationofmetabotropicglutamatereceptors),generategammaoscillationswithoutspecificinputs.ResearcherssoonfoundotherconditionsthatwouldgenerategammaoscillationsandWangandBuzski(1996)pioneeredtheprojectofconstructingcomputationalmodelstoexamineconditionsunderwhichnetworksofinhibitory

interneuronswouldgenerategammaoscillations.However,theirmodelrequiredconditionsnotfoundintheexperimentalpreparations.Kawaguchi,Katsumaru,Kosaka,Heizmann,andHamas(1987)discoveryoffast-spikinginhibitoryneuronsthatexpressthecalciumbindingproteinparvalbuminprovidedaclueastohowtheseoscillationsarise.Parvalbuminexpressingneuronsformnetworksofmutuallyconnectedneuronsinwhichgammaoscillationsarecreatedandthentransmittedtopyramidalcellsviatheiroutputconnectionstothesecells(Bartos,Vida,&Jonas,2007).Theparvalbumin-expressingcellsarehighlyactiveduringgammaoscillationsduringwhichtheygenerateactionpotentials

7/28/2019 Bechtel.the Endogenously Active Brain

10/22

oncepercycleinaphase-lockedmanner(Gloveli,2005),supportingthehypothesisthattheyplayafundamentalroleingeneratinggammaoscillations.Moreover,theyaffectthepyramidalcellsinadistinctivewaytheydonotcausethemtohyperpolarize,butusechloridechannelstoshuntelectricalactivity.Theresultisamechanismthatisrobustagainsttheheterogeneityofthetonicdrivewithhighexcitationanearlyconductance-

dominatedphasecausessubsequentactionpotentialstooccurlaterwhereaswithlowexcitationthedepolarization-dominatedphasecausessubsequentactionpotentialstobeadvanced.Theresultistocreateregular,clock-likeepochsinwhichactionpotentialscanarise;informationprocessingisthusstructuredintodiscretetemporalwindows.Fries(2009)proposesthatthesegammaoscillationsmaybelocatedwithinslowerthetaoscillationsinsuchafashionthatthetaoscillationsprovidelargertemporalwindowsinwhichgammasynchronymaydevelopbutistheninterruptedsoastoprocessnewinput.AsevidenceforthisproposalhecitesRollenhagenandOlsons(2005)recordingsfromindividualinferotemporalneuronsafterfirstpresentingapreferredstimulusandthen600mslateraddinganon-preferredtothepreferredstimulus.Presentationofthesecond

stimulusresultedinareductioninfiringrate,followedbyanapproximately5Hz(thetarange)oscillationbetweenperiodsofenhancedandreducedfiring,suggestingathetafrequencyoscillationbetweentheresponsetothepreferredandnon-preferredstimulus.Sincesuchoscillationisalsoobservedinrecordingfromindividualneuronswhenonlythepreferredstimulusispresented,Friesproposedthethetaoscillationsareanendogenousprocessthatallowsaperiodfordevelopingaresponsetoastimulusfollowedbyaperiodthatopenstheprocessingsystemuptoseekingnewinputratherthansimplycontinuingtorespondtothefirststimulus.Itthusinsuresthatanyresponsetoastimulusisonlymetastableandthatprocessingdoesnotstopwithit.Bothgammaandthetaoscillationsarealsoexhibitedinthehippocampuswhere

researchershaveadvancedhypothesesastotheircontributiontocognitiveprocessingofspatialinformation.Thehippocampusisorganizedintoaloopstructure(Figure2)itreceivesinputfromtheentorhinalcortex(EC)thatistransmittedalongapathwaytothedentategyrus,theCA3fields,theCA1fields,andsubiculumbeforesendinginputsbacktotheentorhinalcortex.ThereisalsoapathwaydirectlyfromtheECtotheCA1fields.

7/28/2019 Bechtel.the Endogenously Active Brain

11/22

Figure2.ThelooparchitectureofthehippocampusinwhichinputfromentorhinalcortexiseitherprocessedinthedentategyrusandCA3beforereachingCA1orisdirectlyreceivedbyCA1.

Followingreportsthatratswithhippocampallesionsexhibitdeficitsinthesortsof

navigationtasksthathadledTolman(1948)toproposethattheypossessacognitivemapoftheirenvironment,OKeefebegandetailedfunctionalanalysisoftheresponsecharacteristicsofhippocampalneurons.RecordingfromindividualcellsintheCA1andCA4regions(CA4isasmallregionnotcommonlydiscussed)ofthehippocampus,OKeefeandDostrovsky(1971)identifiedanumberofcellsthatgeneratedseveralburstsofactionpotentialswheneveraratwasinaparticularlocationinitsenclosure.OKeefelatertermedtheseplacecellsandarguedthattheyconstitutedtheratscognitivemapofitsenvironment(O'Keefe&Nadel,1978).DuringthesameperiodRanck(1973)identifiedcellsthatgeneratedburstsofactionpotentialsinthethetarangewhentheratwasmoving.Overthefollowingdecaderesearchfocusedprimarilyonplacecells,especiallyonconditionsinvolvingchangesintheenclosurethatwouldcausedifferentcellstoproduceburstsof

spikesdespitetheratbeinginthesamelocation.OKeefeandRecce(1993)initiatedanewphaseofresearchwhentheyexploredtherelationbetweenplacecellactivityandthetaoscillationsanddeterminedthattheburstsofactionpotentialsfromplacecellsexhibitedasomewhathigherfrequencythantheongoingthetaoscillation.Asshowninfigure3,theinitialburstwouldtypicallyoccurduringthepeakofthethetaoscillationsandsubsequentburstswouldprecessearlierinthethetaoscillationssothatbythetimetheratlefttheplacefield(theregionthatelicitedactivityfromaparticularcell),theburstswouldhaveadvancednearlyafullcycle.Consideringonlytheactivityofasingleplacecell,theamountofprecessionspecifieshowlongtherathasbeenintheplacefieldandthushowfarithasadvancedthroughit.Fromapopulationperspective,however,theinformationisevenricher.Amongcellswhoseplacefieldspartiallyoverlap,thoseplacecellswhoseaction

potentialshaveprecessedthemostrepresentplacestheratenteredearlierwhiletheonesinwhichactionpotentialshaveprecessedtheleastrepresentthosetherathasjustentered.Thus,theactivityacrossthepopulationofplaceinrelationtothetaspecifiestheratstrajectory.

Figure3.Illustrationofthetaprecession.Asratrunsalongthemaze,itcrossestheplacefieldofacell.Theplacecellspikes,showninred,precessagainsttheunderlyingthetaoscillation,firingfirstjustafterthepeakandmovingprogressivelyearlyonsubsequentthetacycles.FromWang2010.

Inadditiontothetaactivity,thehippocampusalsoexhibitsgammaoscillationsattwodifferentfrequencies,oneoriginatinginmedialEC,whichexhibitsfastgammaoscillations(>60Hz),andtheotherintheCA3region,whichgeneratesslowgammaoscillations(

7/28/2019 Bechtel.the Endogenously Active Brain

12/22

fromwhichtheyreceiveinput.FromtheseconsiderationsandthefactthatCA1atdifferenttimesexhibitsfastandslowgammaoscillations,Colgin,Denninger,Fyhn,Hafting,Bonnevie,Jensen,Moser,&Moser(2009)havehypothesizedthatthesedifferentgammaoscillationsregulatetheprocessingofinformationatdifferenttimestheCA1fieldssynchronizewiththefastgammaoftheentorhinalcortexandrespondtoinputfromitand

atothertimessynchronizewiththeslowgammaofCA3andhencerespondtoinputsfromit.TheprocessinginDGandCA3isthoughttofacilitaterecognizingalocationasapreviouslyexperiencedone;whenthatfails,CA1respondsdirectlytotheassumedtobenovelplacerepresentedbytheECactivity.Animportantprocessinthehippocampusislong-termpotentiation(LTP),whichstrengthenssynapticconnectionsonwhichaninputisreceivedwhentherecipientneuronactuallygeneratesanactionpotential.LTPoccursonlywhenaneurongeneratesaspikeduringthethroughofthethetacycle,whichhappenswhenCA1issynchronizedwithentorhinalcortex.ThislimitslearninganewrepresentationtooccasionswhentheDG-CA3loopsfailstorecognizethelocation,asituationinwhichitisappropriatetoacquireanewrepresentation.Whenanoldlocationisrecognized,CA1synchronizeswithCA3andspikesoccuratthepeakofthethetacycle

whenLTPisblocked.AfteraperiodduringwhichtheonlywayresearcherswereabletorelateEEGactivitytocognitionwasthroughERP,whichadoptedareactiveperspective,researchersarebeginningtoformulatehypothesesastohowoscillations,especiallythoseinthethetaandgammabands,thatariseindependentlyofspecificstimulifigureininformationprocessing.TheyhavebeenproposedtocreatetimewindowsinwhichneuralprocessesmaydevelopresponsestoinputsbutthenbreakoutoftheresponsetosampleotherinputsaswellaswaysofgatingtheflowofinformationbetweenbrainregionsandcontrollingprocessessuchasLTP.Theseproposalssuggestthatthebrainisadynamicallyactiveprocessingsysteminwhichtheongoingdynamicsstructuresandregulatesprocessingintime.

4.OscillationsDetectedwithPETandfMRIOneofthelimitationsofEEGforstudyingcognitiveprocessinginthebrainisthatwhileitprovideshigh-resolutioninformationaboutthetimingofelectricalactivity,itisverydifficulttolocalizespatiallythesourceoftheelectricalsignal.ThusalthoughasmallcadreofresearchersemployedEEGandERPtoinvestigatecognitiveprocessinginthebraininthe1970sand1980s,theirinvestigationsfailedtosparkthedevelopmentofcognitiveneuroscienceasaprominentfieldofinquiry.Rather,thishappenedonlyafterthedevelopmentoftechniquessuchaspositronemissiontomography(PET)inthelate1980sandfunctionalmagneticresonanceimaging(fMRI)inthe1990s.Thesetoolsused

tomographicmethodstolocalizethesourceofeitheraradioactiveormagneticsignalthatwasassumedtoberelatedtospecificcognitiveactivities.Asdescribedabove,intheinitialresearchwithPETandfMRIstudiesresearcherssoughttoidentifyregionsinwhichbloodflowwascorrelatedwithcognitiveactivities.ThuslikeERPresearchers,PETandfMRIresearchersinitiallyviewedthecognitivesystemasareactivemechanismandfocusedondetectingincreasedactivitythatcouldbeattributedtothetaskasubjectwasperforming.Accordingly,theypaidlittleattentiontothevaryingbackgroundsignalexceptinsofarasitposedchallengesindetectingthesignalattributedtothestimulusortask.

7/28/2019 Bechtel.the Endogenously Active Brain

13/22

Inthemid-1990ssomeresearchers(e.g.,Ghatan,Hsieh,Wirsn-Meurling,Wredling,Eriksson,Stone-Elander,Levander,&Ingvar,1995;Baker,Rogers,Owen,Frith,Dolan,Frackowiak,&Robbins,1996)drewattentiontothefactthatsomebrainregionsseemedregularlytoexhibitreducedactivity(weredeactivated)intaskconditionsthaninaresting

state(aconditioninwhichthesubjectlayquietlyinthescannerwithnoassignedtask).Todeterminehowwidespreadthiseffectwas,Shulman,Corbetta,Buckner,Fiez,Miezin,Raichle,andPetersen(1997)conductedameta-analysisofPETstudiesinwhichataskconditionwascomparedtoanon-taskconditioninwhichthesamestimuluswaspresented(thisproducedresultsverysimilartotherestingconditionwithnostimulus).Theyidentifiedaspecificsetofbrainareasthatwerereliablylessactiveintasksituations:thejunctionofprecuneusandposteriorcingulatecortex(PCC),theinferiorparietalcortex(IPC),theleftdorsolateralprefrontalcortex(leftDLPFC),amedialfrontalstripthatcontinuedthroughtheinferioranteriorcingulatecortex(inferiorACC),theleftinferiorfrontalcortex,theleftinferiorfrontalgyrus,andtheamygdala.

ShortlythereafterRaichleandhiscollaboratorsshiftedtheirperspectivefromviewingtheseareasasdeactivatedintaskconditionstoviewingthemasbeingmoreactiveinnon-taskconditionsandbegantocharacterizethemasconstitutingadefaultmodenetworkonewhichperformsactualfunctionsbestcarriedoutwhentherearenoexternaltaskdemands.OneclueastotheirfunctionwasprovidedbyAndreasen,O'Leary,Cizadlo,Arndt,Rezai,Watkins,Ponto,andHichwa(1995)whofoundthatepisodicmemorytaskswereonetypeoftaskthatdidnotresultinreducedactivityintheseregions.Someresearchersdrewuponthisfindingtosuggestthatthedefaultmodenetworksupportsundirectedthinkingorwhathadbeenlabeledmind-wandering(Antrobus,Singer,Goldstein,&Fortgang,1970).ThiscoheresthereportsbyAndreasenetal.ssubjectsthat,whenrequiredtolayinthescannerwithnotaskrequirements,theythoughtabouttheirownpastorplannedfuture

activities.Thesearebothactivitiesthatdrawuponepisodicmemory.Adoptingthemind-wanderinghypothesis,Buckner,Andrews-Hanna,&Schacter(2008,p.2)linkmind-wanderingtotheabilitytocarryoutflexibleself-relevantmentalexplorationssimulationsthatprovideameanstoanticipateandevaluateupcomingeventsbeforetheyhappen(p.2).IndefendingthisviewtheycitenotonlyAndreasenetal.sresultsbutalsocorrelationsfoundbyMasonetal.(2007)betweenstimulusindependentthoughtsandactivityinthedefaultnetwork.Gilbert,Dumontheil,Simons,Frith,andBurgess(2007)offeranalternativeviewthatactivityinthedefaultnetworkgenerateslow-levelgeneralizedawarenessorwatchfulness.ThisalternativegainssupportfromHahn,Ross,andSteins(2007)findingsofincreasedactivityinthedefaultnetworkin

atarget-detectingtaskwhenthetargetcouldappearanywhere,butnotwhenitwasexpectedinaspecificlocation.Apotentialproblemforbothofthesetreatmentsofthefunctionofthedefaultmodenetworkisthatithasbeenfoundtobeactivenotjustwhenindividualsareawakebutalsoduringsleep(Fukunaga,Horovitz,vanGelderen,deZwart,Jansma,Ikonomidou,Chu,Deckers,Leopold,&Duyn,2006;Larson-Prior,Zempel,Nolan,Prior,Snyder,&Raichle,2009)andunderanesthesia(Vincent,Patel,Fox,Snyder,Baker,VanEssen,Zempel,Snyder,Corbetta,&Raichle,2007)circumstancesinwhichneitherspontaneousconsciousthoughts(suchasAndreasenetal.ssubjectsreport)nor

7/28/2019 Bechtel.the Endogenously Active Brain

14/22

generalizedawarenessandwatchfulnessoccur.Evenifactivityinthedefaultmodenetworkisfoundintheabsenceofconsciousthought,thedefaultmodenetworkmaynonethelessprovideaneurophysiologicalfoundationformind-wanderingorwatchfulnesswhenthesubjectisengagedinconsciousthought.Mind-wanderingandwatchfulnessarefeaturesofourmentallifethathavebeenneglectedbythereactiveframeworkbutas

Buckneretal.suggestmaybeimportantcognitiveactivitiesthatenableustoanticipateandcopewithourenvironments.Theinitialstudiesofthedefaultmodenetworkdidnotfocusonthedynamicsoftheactivityintheseareas,butastudybyBiswal,Yetkin,Haughton,andHyde(1995)showedhowfMRIcouldbeusedtoidentifydynamicprocessesintaskactivatednetworksthatweresoonextendedtothedefaultmodenetwork.TheseresearchersperformedatimeseriesanalysisofafMRIrecordingtakenevery250msandobservedverylow-frequencyoscillations(

7/28/2019 Bechtel.the Endogenously Active Brain

15/22

Buckner,Andrews-Hanna,andSchacter(2008,pp.4-5)summeduptheperspectiveprovidedbythisresearch:Thedefaultnetworkisabrainsystemmuchlikethemotorsystemorthevisualsystem.Itcontainsasetofinteractingbrainareasthataretightlyfunctionallyconnectedanddistinctfromothersystemswithinthebrain.

TheCordesetal.resultsnotedaboveshowedthatevenintherestingstateonecoulddemonstratecorrelatedactivityacrossnetworksthatwouldthenbejointlyactivatedintaskconditions.Foxetal.(2005)foundcorrelatedactivityintherestingstateinanetworkthatwasespeciallyactiveduringattention-demandingtasks(intraparietalsulcus,frontaleyefield,middletemporalregion,supplementarymotorareas,andtheinsula)andshowedthatactivityinthatnetworkandinthedefaultnetworkwereanticorrelated.Thatis,theareasthatwerepositivelycorrelatedwithineachnetworkwerenegativelycorrelatedwithareasintheothernetworkanoutcomemoreinterestingthanazerocorrelation.Foxetal.emphasizedthat:

anticorrelationsmaybeasimportantascorrelationsinbrainorganization.Littlehasbeensaidpreviouslyintheneuronalsynchronyliteratureregardingtheroleof

anticorrelations.Whilecorrelationsmayserveanintegrativeroleincombiningneuronalactivitysubservingsimilargoalsorrepresentations,anticorrelationsmayserveadifferentiatingrolesegregatingneuronalprocessessubservingoppositegoalsorcompetingrepresentations(p.9677).

TheoscillationsfoundusingfMRIareonetotwoordersofmagnitudeslowerthanthoseobservedwithEEG.Dothey,likethetaandgammaoscillations,playaroleincognition?Sofarnotalotisknownabouttheirfunctionalsignificance,butaveryintriguingfindingbytheRaichlegroupsuggeststhatitmayhavebehavioralconsequences.AsIhavenoted,inreactivestudiesresearchersconstantlyconfrontthefactthatthereisgreatvariabilityintherecordedsignal,whichisoftentreatedasnoise.Fox,Snyder,Zacks,andRaichle(2006)

devisedaninnovativestrategyforshowingthatalargepartofthisvariabilitymaybetheendogenousoscillation.Assubjectsperformedabutton-presstaskwiththeirrighthandwhenevertheyidentifiedaneweventinavideo,amotoractivitythatisdirectedbytheleftsideofthebrain,theresearchersidentifiedtheongoingoscillationinsensory-motorareasonbothsidesofthebrain.TheyfoundthattheoscillationsintherightsideofthebrainaccountedformuchofthevariationfoundinthefMRIsignalontheleftsideofthebrain.Moreover,inasubsequentstudythesamegroupfoundthatthespontaneousfluctuationexhibitedinthecontralateralhemispheretothatcontrollingthebuttonpushaccountedfortheforcewithwhichasubjectpressedthebutton(Fox,Snyder,Vincent,&Raichle,2007).Thisfindingissuggestivethatendogenousactivitymaydirectlyaffectedmentalprocessesandbehavioralactivities.

5.DynamicallyOrganizedNetworksintheBrain

Thepatternofactivitywithincorrelatednetworksandbetweenanticorrelatednetworks,whetherthesenetworksaretaskcharacterizedorthedefaultmodenetwork,duringconditionswheresubjectsarenotrequiredtoperformanytasksrevealsthatthereishighlyorganizedendogenousactivityinthebrain.DuringthesameperiodaswhenfcMRIwasrevealingthesefunctionalnetworks,othertechniques(e.g.,diffusiontensorMRI)were

7/28/2019 Bechtel.the Endogenously Active Brain

16/22

beingdevelopedtoanalyzestructuralconnectivityinthebrain.Thestructurallyandfunctionallycharacterizednetworksappeartobehighlycorrelated.Greicius,Supekar,Menon,&Dougherty(2009)showedthattheregionsinthedefaultmodenetworkareanatomicallyconnected;vandenHeuvel,Mandl,Kahn,&Pol(2009)foundthateightoftheninenetworkstheyidentifiedintherestingstatecorrespondtoonesthatcanbe

characterizedanatomicallyasconnectedbyfibertracts.Moreover,thesenetworkshaveasimilartypeofstructure.Whetheranalyzedstructurallyorfunctionally,theyhavebeenfoundtoexhibitasmall-worldstructure(Watts&Strogratz,1998).Small-worldorganizationisaformoforganizationthatliesbetweenregularlatticesandrandomlyorganizednetworksasinlatticesmostconnectionsarebetweennearbyunitsbutafewconnectbetweendistantlocations.Forinformationprocessingpurposes,latticestructureshavethevirtueofcreatingmodulesofunitsthatfunctiontogethertoperformaparticulartaskwhereasrandomnetworksexhibitashortpathofcommunicationacrossthenetwork.Small-worldnetworkpossessbothpropertiesandmanyreal-worldnetworkshavebeenfoundtoexhibitsmallworldorganization,includingthedefaultmode

networkandtaskdirectednetworks(Sporns,2010).Inthinkingaboutcognitivearchitectures,itiscommontoconstruethemassetdownbeforecognitiveactivityandsimplysettingthepossibilitiesforsuchactivity.Butactivityinthebrainisnowunderstoodascapableofmodifyingthestructuralorganizationofthebrain.ThebrainisplasticandprocessessuchasLTP,discussedabove,canchangetheconnectivityofthebrain.AnintriguingpossibilitythatRaichlehasadvancedisthatthesynchronizedslowoscillationswithinnetworksmayplayaroleinsculptingthebrain.GongandvanLeeuwen(2004)showedinacomputationalmodelthatwhenchangingconnectionstrengthsbetweenunitsoccurredwhenunitsexhibitsynchronizedactivity,small-worldorganizationwoulddevelopinthenetwork.Evenwhenthesenetworksbeginwithrandom

connectivity,theyself-organizeintoclusterslinkedtoeachotherthroughhubs.Rubinov,Sporns,vanLeeuwen,andBreakspear(2009)appealtoGongandvanLeeuwensmodeltosuggestthataspatternsofsynchronizedactivityoccurinthebrain,thebrainmightself-organizeintoasmallworldnetworkwithhubs.Someofthesynchronybetweenneuronsmightbetheresultofspecificstimuliortasksthebrainisrequiredtoperform.Butifthesynchronizationbetweenbrainregionsisongoing,occurringduringtherestingstateaswellasundertasksconditions,thensomeofthesynchronizedactivitythatshapesthewiringofthebrainwillbearesultofendogenousfunctioning,notjuststimulusinducedactivity.6.DevelopinganEndogenouslyActiveCognitiveArchitecture

IntheprecedingsectionsIhavepresentedevidencefromavarietyofsourcesthatbrainsnotonlyrespondtotasksorstimulibutalsoareendogenouslyactiveandthatthisactivitymaybeimportantforunderstandingcognition.Tobeginwith,neuronsexhibitendogenousoscillationsintheelectricalactivityontheirmembranesandtheyaremorelikelytogenerateactionpotentialswhentheyareinahypopolarizedphasethanwhentheyarehyperpolarized.Inaddition,someneuronsarecapableofgeneratingactionpotentialsindependentlyofanyinput.Further,attheleveloflargepopulationsofneuronsbothEEG

7/28/2019 Bechtel.the Endogenously Active Brain

17/22

andfMRIrevealongoingoscillatoryactivitythathastraditionallybeentreatedasnoise.EEGdetectsoscillationsinavarietyoffrequencybandsabove1Hz.Bergeridentifiedactivitydominantlyinthealphabandwhensubjectsaresittingquietlywithoutstimulation,butevenduringthisperiodthereareoscillationsinotherfrequencyranges.Boththetaandgammaoscillationsincortexappeartobeimportantincreatingtemporalwindowswith

gammawindowsappearingtoenableorganizedprocessingofinputsandthetaoscillationsservingtoperiodicallyresettheprocessingsystemtorespondtoalternativeinputs.Inthehippocampusdifferentregionsoscillateatdifferentfrequencies,withsomeregionsabletoentraintoandthenrespondtotheoscillatoryactivityofdifferentregions.ThetaoscillationsprovideatimingwindowinwhichtherelativetimingofactionpotentialscarriesinformationandregulateswhetheractivitysuchasLTPistooccur.Therecentfindingofoscillationsatlessthan.1HzwithfMRIprovidefurtherperspectiveontheroleofendogenousactivityinthebrain,especiallyinrevealingongoingcoordinationofactivityinmultiplebrainnetworksevenduringtherestingstate.Thereissuggestiveevidencethatthisactivityismanifestinbehaviorandunderliesongoingthoughtnottiedtotasks(mind-wandering).Thepossibleroleofthisactivityinsculptingthebrainsuggeststhatitis

importantinestablishingtheprocessingcapacitiesofthebraini.e.,thecognitivearchitecture.Theevidenceofendogenousactivityinthebrainandsuggestionsthatitmaybeimportantforcognitionposesachallengetocognitivesciencethatreliesonproposedcognitivearchitecturesthatarepurelyreactive.Moreover,itraisesthequestionofhowcognitiveresearchersmightdevelopaccountsofthecognitivearchitecturethatincorporateendogenousactivity.Withindividualneurons,endogenousactivityappearsintheongoingoscillationsintheelectricalpotentialsoverthemembranes.Onewayofcreatinganendogenouslyactivesystemistoconstructitoutofoscillators.Populationsofoscillatorshavethepotentialtosynchronize,butdependingonthepatternofconnectionsbetween

them,theymayonlyformtemporarycoalitionsofsynchronizedcomponentsandsoexhibitmeta-stability.Accordingly,oscillatorswithprocessesforsynchronizingthemconstitutesystemscapableofthesortsofendogenousactivityfoundinthebrain.Cognitivearchitecturesareinstantiatedincomputationalmodels.Itisrelativelyeasytodevelopcomputationalmodelsthatexhibitongoingoscillationsinfieldsofbiologydealingwithoscillatoryphenomena,suchascircadianbiology,researchersemploymodelsofoscillatorycomponents.Someoftheserepresentthecomponentsofthemechanism(involvinginthecircadiancase,thetranscriptionandtranslationofgenes)whileothersabstractfromthesedetailsandemploy,forexample,equationsdescribingavanderPoloscillator(Bechtel&Abrahamsen,2010).Populationsaremodeledbyaddingtermstothe

basicequationsforoscillatorstocharacterizethetransmissionofsignalsbetweenthem.Inthecontextofneuroscience,computationalmodelershaveincreasinglydirectedtheirattentiontomodelingmembranepotentialsandusingthesemodeltounderstandtheroleofmembranepotentialsingeneratingpopulation-levelrhythms(Destexhe&Sejnowski,2003).Thesemodelsprovideresourcesforcreatingcomputationalmodelsofcognitivearchitecturesthatexhibitendogenousdynamicalactivity.Developingandusingsuchalternativecognitivearchitectureswouldprovidecognitiveresearchersabasisforinvestigatingthepossiblecontributionsofendogenousactivitytocognition.Ontheother

7/28/2019 Bechtel.the Endogenously Active Brain

18/22

hand,byrelyingonlyoncognitivearchitecturesthatdonotincorporateendogenousactivity,cognitivescientistsmaylacktheresourcesformodelingcognitionthatoccursinbrainsthatdomakeendogenousactivityavailableasacognitiveresource.References

Alving,B.O.(1968).SpontaneousactivityinisolatedsomataofAplysiapacemakerneurons.JournalofGeneralPhysiology,51,29-45.

Anderson,J.R.(1990).Theadaptivecharacterofthought.Hillsdale,NJ:LawrenceErlbaumAssociates.

Anderson,J.R.(2007).Howcanthehumanmindoccurinthephysicaluniverse?Oxford:OxfordUniversityPress.

Andreasen,N.C.,O'Leary,D.S.,Cizadlo,T.,Arndt,S.,Rezai,K.,Watkins,G.L.,etal.(1995).Rememberingthepast:twofacetsofepisodicmemoryexploredwithpositronemissiontomography.AmericanJournalofPsychiatry,152,1576-1585.

Antrobus,J.S.,Singer,J.L.,Goldstein,S.,&Fortgang,M.(1970).Mindwanderingand

cognitivestructure.TransactionsoftheNewYorkAcademyofSciences,32,242-252.Baker,S.C.,Rogers,R.D.,Owen,A.M.,Frith,C.D.,Dolan,R.J.,Frackowiak,R.S.J.,etal.

(1996).Neuralsystemsengagedbyplanning:aPETstudyoftheTowerofLondontask.Neuropsychologia,34,515-526.

Barandiaran,X.,&Moreno,A.(2006).Onwhatmakescertaindynamicalsystemscognitive:Aminimallycognitiveorganizationprogram.AdaptiveBehavior,14,171-185.

Bartos,M.,Vida,I.,&Jonas,P.(2007).Synapticmechanismsofsynchronizedgammaoscillationsininhibitoryinterneuronnetworks.NatureReviewsNeuroscience,8,45-56.

Bechtel,W.(2008).Mentalmechanisms.London:Routledge.Bechtel,W.(2011).Mechanismandbiologicalexplanation.PhilosophyofScience,78,533-

557.Bechtel,W.,&Abrahamsen,A.(2005).Explanation:Amechanistalternative.Studiesin

HistoryandPhilosophyofBiologicalandBiomedicalSciences,36,421-441.Bechtel,W.,&Abrahamsen,A.(2010).Dynamicmechanisticexplanation:Computational

modelingofcircadianrhythmsasanexemplarforcognitivescience.StudiesinHistoryandPhilosophyofSciencePartA,41,321-333.

Bechtel,W.,&Abrahamsen,A.(2011).Complexbiologicalmechanisms:Cyclic,oscillatory,andautonomous.InC.A.Hooker(Ed.),Philosophyofcomplexsystems.Handbookofthephilosophyofscience(Vol.10,pp.257-285).NewYork:Elsevier.

Bechtel,W.,&Richardson,R.C.(1993/2010).Discoveringcomplexity:Decompositionandlocalizationasstrategiesinscientificresearch.Cambridge,MA:MITPress.1993

editionpublishedbyPrincetonUniversityPress.Berger,H.(1929).berdaasElektroenkephalogrammdesMenschen.ArchivfrPsychiatrie

undNervenkrankheiten,87.Berger,H.(1930).berdaasElektroenkephalogrammdesMenschen.ZweiteMitteilung.

JournalfrPsychologieundNeurologie,40.Bernstein,J.(1912).Elektrobiologie.DieLehrevondenelektrischenVorgngenim

OrganismusaufmodernerGrundlagedargestelltBraunschweig:Vieweg&Sohn.

7/28/2019 Bechtel.the Endogenously Active Brain

19/22

Biswal,B.,Yetkin,F.Z.,Haughton,V.M.,&Hyde,J.S.(1995).Functionalconnectivityinthemotorcortexofrestinghumanbrainusingecho-planarMRI.MagneticResonanceinMedicine,34,537-541.

Bremer,F.(1958).Cerebralandcerebellarpotentials.PhysiologicalReviews,38,357-388.Brown,T.G.(1911).Theintrinsicfactorsintheactofprogressioninthemammal.

ProceedingsoftheRoyalSocietyofLondon.SeriesB,ContainingPapersofaBiologicalCharacter,84,308-319.

Brown,T.G.(1914).Onthenatureofthefundamentalactivityofthenervouscentres;togetherwithananalysisoftheconditioningofrhythmicactivityinprogression,andatheoryoftheevolutionoffunctioninthenervoussystem.TheJournalofPhysiology,48,18-46.

Buckner,R.L.,Andrews-Hanna,J.R.,&Schacter,D.L.(2008).Thebrain'sdefaultnetwork.AnnalsoftheNewYorkAcademyofSciences,1124,1-38.

Buzski,G.(2006).Rhythmsofthebrain.Oxford:OxfordUniversityPress.Colgin,L.L.,Denninger,T.,Fyhn,M.,Hafting,T.,Bonnevie,T.,Jensen,O.,etal.(2009).

Frequencyofgammaoscillationsroutesflowofinformationinthehippocampus.

Nature,462,353-357.Corbetta,M.,&Shulman,G.L.(2002).Controlofgoal-directedandstimulus-driven

attentioninthebrain.NatureReviewsNeuroscience,3,201-215.Cordes,D.,Haughton,V.M.,Arfanakis,K.,Wendt,G.J.,Turski,P.A.,Moritz,C.H.,etal.

(2000).MappingfunctionallyrelatedregionsofbrainwithfunctionalconnectivityMRimaging.AmericanJournalofNeuroradiology,21,1636-1644.

Destexhe,A.D.,&Sejnowski,T.J.(2003).Interactionsbetweenmembraneconductancesunderlyingthalamocorticalslow-waveoscillations.PhysiologicalReviews,83,1401-1453.

DuBois-Reymond,E.(1848-1884).UntersuchungenberthierischeElektricitt.Berlin:Reimer.

Duch,W.,Oentaryo,R.J.,&Pasquier,M.(2008).CognitiveArchitectures:Wheredowegofromhere?Proceedingsofthe2008conferenceonArtificialGeneralIntelligence(pp.122-136):IOSPress.

Fox,M.D.,Snyder,A.Z.,Vincent,J.L.,Corbetta,M.,VanEssen,D.C.,&Raichle,M.E.(2005).Thehumanbrainisintrinsicallyorganizedintodynamic,anticorrelatedfunctionalnetworks.ProceedingsoftheNationalAcademyofSciencesoftheUnitedStatesofAmerica,102,9673-9678.

Fox,M.D.,Snyder,A.Z.,Vincent,J.L.,&Raichle,M.E.(2007).Intrinsicfluctuationswithincorticalsystemsaccountforintertrialvariabilityinhumanbehavior.Neuron,56,171-184.

Fox,M.D.,Snyder,A.Z.,Zacks,J.M.,&Raichle,M.E.(2006).Coherentspontaneousactivity

accountsfortrial-to-trialvariabilityinhumanevokedbrainresponses.NatureNeuroscience,9,23-25.

Fries,P.(2009).Neuronalgamma-bandsynchronizationasafundamentalprocessincorticalcomputation.AnnualReviewofNeuroscience,32,209-224.

Fukunaga,M.,Horovitz,S.G.,vanGelderen,P.,deZwart,J.A.,Jansma,J.M.,Ikonomidou,V.N.,etal.(2006).Large-amplitude,spatiallycorrelatedfluctuationsinBOLDfMRIsignalsduringextendedrestandearlysleepstages.MagneticResonanceImaging,24,979-992.

7/28/2019 Bechtel.the Endogenously Active Brain

20/22

Gnti,T.(2003).Theprinciplesoflife.NewYork:Oxford.Ghatan,P.H.,Hsieh,J.C.,Wirsn-Meurling,A.,Wredling,R.,Eriksson,L.,Stone-Elander,S.,et

al.(1995).Brainactivationinducedbytheperceptualmazetest:APETstudyofcognitiveperformance.Neuroimage,2,112-124.

Gilbert,S.J.,Dumontheil,I.,Simons,J.S.,Frith,C.D.,&Burgess,P.W.(2007).Commenton

"Wanderingminds:Thedefaultnetworkandstimulus-independentthought".Science,317,43b-.Glennan,S.(1996).Mechanismsandthenatureofcausation.Erkenntnis,44,50-71.Glennan,S.(2002).Rethinkingmechanisticexplanation.PhilosophyofScience,69,S342-

S353.Gloor,P.(1969).HansBergerontheelectroencephalogramofman.Amsterdam:Elsevier.Gloveli,T.(2005).Differentialinvolvementoforiens/pyramidaleinterneuronesin

hippocampalnetworkoscillationsinvitro.J.Physiol.(Lond.),562,131-147.Goldbeter,A.(1996).Biochemicaloscillationsandcellularrhythms:Themolecularbasesof

periodicandchaoticbehaviour.Cambridge:CambridgeUniversityPress.Gong,P.,&vanLeeuwen,C.(2004).Evolutiontoasmall-worldnetworkwithchaoticunits.

EurophysicsLetters,67,328-333.Goodwin,B.C.(1963).Temporalorganizationincells;adynamictheoryofcellularcontrol

processes.London:Academic.Gray,C.M.,Konig,P.,Engel,A.K.,&Singer,W.(1989).Oscillatoryresponsesincatvisual

cortexexhibitinter-columnarsynchronizationwhichreflectsglobalstimulusproperties.Nature,338,334-337.

Greicius,M.D.,Krasnow,B.,Reiss,A.L.,&Menon,V.(2003).Functionalconnectivityintherestingbrain:Anetworkanalysisofthedefaultmodehypothesis.ProceedingsoftheNationalAcademyofSciencesoftheUnitedStatesofAmerica,100,253-258.

Greicius,M.D.,&Menon,V.(2004).Default-modeactivityduringapassivesensorytask:Uncoupledfromdeactivationbutimpactingactivation.JournalofCognitive

Neuroscience,16,1484-1492.Greicius,M.D.,Supekar,K.,Menon,V.,&Dougherty,R.F.(2009).Resting-StateFunctional

ConnectivityReflectsStructuralConnectivityintheDefaultModeNetwork.CerebralCortex,19,72-78.

Hahn,B.,Ross,T.J.,&Stein,E.A.(2007).Cingulateactivationincreasesdynamicallywithresponsespeedunderstimulusunpredictability.CerebralCortex,17,1664-1671.

Hasenstaub,A.,Shu,Y.,Haider,B.,Kraushaar,U.,Duque,A.,&McCormick,D.A.(2005).InhibitoryPostsynapticPotentialsCarrySynchronizedFrequencyInformationinActiveCorticalNetworks.Neuron,47,423-435.

Hodgkin,A.L.,&Huxley,A.F.(1952).Aquantitativedescriptionofmembranecurrentanditsapplicationtotheconductionandexcitationofnerve.JournalofPhysiology,117,

500-544.Hubel,D.H.,&Wiesel,T.N.(1962).Receptivefields,binocularinteractionandfunctional

architectureinthecat'svisualcortex.JournalofPhysiology,160,106-154.Hubel,D.H.,&Wiesel,T.N.(1968).Receptivefieldsandfunctionalarchitectureofmonkey

striatecortex.JournalofPhysiology,195,215-243.Huguenard,J.R.(1996).Low-thresholdcalciumcurrentsincentralnervoussystem

neurons.AnnualReviewofPhysiology,58,329-348.Huygens,C.(1888).Oeuvrescompltes.LaHaye:M.Nijhoff.

7/28/2019 Bechtel.the Endogenously Active Brain

21/22

Jacob,F.,&Monod,J.(1961).Geneticregulatorysystemsinthesynthesisofproteins.JournalofMolecularBiology,3,318-356.

Jahnsen,H.,&Llins,R.R.(1984).Electrophysiologicalpropertiesofguinea-pigthalamicneurones:aninvitrostudy.TheJournalofPhysiology,349,205-226.

Kandel,E.R.(1976).Cellularbasisofbehavior:Anintroductiontobehavioralneurobiology.

SanFrancisco:W.H.Freeman.Kawaguchi,Y.,Katsumaru,H.,Kosaka,T.,Heizmann,C.W.,&Hama,K.(1987).Fastspikingcellsinrathippocampus(CA1region)containthecalcium-bindingproteinparvalbumin.BrainResearch,416,369-374.

Kuffler,S.W.(1953).Dischargepatternsandfunctionalorganizationofmammalianretina.JournalofNeurophysiology,16,37-68.

Kutas,M.,&Hillyard,S.A.(1980).Readingsenselesssentences:brainpotentialsreflectsemanticincongruity.Science,207,203-205.

Laird,J.E.,Newell,A.,&Rosenbloom,P.S.(1987).SOAR:Anarchitectureforgeneralintelligence.ArtificialIntelligence,33,1-64.

Larson-Prior,L.J.,Zempel,J.M.,Nolan,T.S.,Prior,F.W.,Snyder,A.Z.,&Raichle,M.E.

(2009).Corticalnetworkfunctionalconnectivityinthedescenttosleep.ProceedingsoftheNationalAcademyofSciences,106,4489-4494.

Llins,R.R.(1988).Theintrinsicelectrophysiologicalpropertiesofmammalianneurons:Insightsintocentralnervoussystemfunction.Science,242,1654-1664.

LorentedeN,R.(1938).Analysisoftheactivityofthechainsofinternuncialneurons.JournalofNeurophysiology,1,207-244.

Lyon,P.(2006).Thebiogenicapproachtocognition.CognitiveProcessing,7,11-29.Machamer,P.,Darden,L.,&Craver,C.F.(2000).Thinkingaboutmechanisms.Philosophyof

Science,67,1-25.Mason,M.F.,Norton,M.I.,VanHorn,J.D.,Wegner,D.M.,Grafton,S.T.,&Macrae,C.N.

(2007).Wanderingminds:Thedefaultnetworkandstimulus-independentthought.

Science,315,393-395.Maturana,H.R.,&Varela,F.J.(1980).Autopoiesis:Theorganizationoftheliving.InH.R.

Maturana&F.J.Varela(Eds.),Autopoiesisandcognition:Therealizationoftheliving(pp.59-138).Dordrecht:D.Reidel.

O'Keefe,J.A.,&Dostrovsky,J.(1971).Thehippocampusasaspatialmap.Preliminaryevidencefromunitactivityinthefreelymovingrat.BrainResearch,34,171-175.

O'Keefe,J.A.,&Nadel,L.(1978).Thehippocampusasacognitivemap.Oxford:OxfordUniversityPress.

O'Keefe,J.A.,&Recce,M.L.(1993).PhaserelationshipbetweenhippocampalplaceunitsandtheEEGthetarhythm.Hippocampus,3,317-330.

Petersen,S.E.,Fox,P.T.,Posner,M.I.,Mintun,M.,&Raichle,M.E.(1988).Positronemission

tomographicstudiesofthecorticalanatomyofsingle-wordprocessing.Nature,331,585-588.

Ranck,J.B.(1973).Studiesonsingleneuronsindorsalhippocampalformationandseptuminunrestrainedrats:PartI.Behavioralcorrelatesandfiringrepertoires.ExperimentalNeurology,41,462-531.

Rollenhagen,J.E.,&Olson,C.R.(2005).Low-FrequencyOscillationsArisingFromCompetitiveInteractionsBetweenVisualStimuliinMacaqueInferotemporalCortex.JournalofNeurophysiology,94,3368-3387.

7/28/2019 Bechtel.the Endogenously Active Brain

22/22

Rosen,R.(1991).Lifeitself:Acomprehensiveinquiryintothenature,origin,andfabricationoflife.NewYork:Columbia.

Rubinov,M.,Sporns,O.,vanLeeuwen,C.,&Breakspear,M.(2009).Symbioticrelationshipbetweenbrainstructureanddynamics.BMCNeuroscience,10,55.

Ruiz-Mirazo,K.,&Moreno,A.(2004).Basicautonomyasafundamentalstepinthe

synthesisoflife.ArtificialLife,10,235-259.Rumelhart,D.E.,&McClelland,J.L.(1986).Explorationsinthemicrostructureofcognition.Volume1.Foundations.Cambridge,MA:BradfordBooks,MITPress.

Sherrington,C.S.(1923).Theintegrativeactionofthenervoussystem.NewHaven:YaleUniversityPress.

Shulman,G.L.,Corbetta,M.,Buckner,R.L.,Fiez,J.A.,Miezin,F.M.,Raichle,M.E.,etal.(1997).Commonbloodflowchangesacrossvisualtasks:I.increasesinsubcorticalstructuresandcerebellumbutnotinnonvisualcortex.JournalofCognitiveNeuroscience,9,624-647.

Singer,W.(1999).Neuronalsynchrony:Aversatilecodeforthedefinitionofrelations?Neuron,24,49-65.

Sporns,O.(2010).Networksofthebrain.Cambridge,MA,USA:MITPress.Sternberg,S.(1966).High-speedscanninginhumanmemory.Science,153,652-654.Thagard,P.(2003).Pathwaystobiomedicaldiscovery.PhilosophyofScience,70,235-254.Tolman,E.C.(1948).Cognitivemapsinratsandmen.PsychologicalReview,55,189-208.vandenHeuvel,M.P.,Mandl,R.C.W.,Kahn,R.S.,&Pol,H.E.H.(2009).Functionallylinked

resting-statenetworksreflecttheunderlyingstructuralconnectivityarchitectureofthehumanbrain.HumanBrainMapping,30,3127-3141.

vanEssen,D.C.,&Gallant,J.L.(1994).Neuralmechanismsofformandmotionprocessingintheprimatevisualsystem.Neuron,13,1-10.

Vincent,J.L.,Patel,G.H.,Fox,M.D.,Snyder,A.Z.,Baker,J.T.,VanEssen,D.C.,etal.(2007).Intrinsicfunctionalarchitectureintheanaesthetizedmonkeybrain.Nature,447,83-

86.Vincent,J.L.,Snyder,A.Z.,Fox,M.D.,Shannon,B.J.,Andrews,J.R.,Raichle,M.E.,etal.

(2006).Coherentspontaneousactivityidentifiesahippocampal-parietalmemorynetwork.JournalofNeurophysiology,96,3517-3531.

Wang,X.J.,&Buzsaki,G.(1996).Gammaoscillationbysynapticinhibitioninahippocampalinterneuronalnetworkmodel.JournalofNeuroscience,16,6402-6413.

Watts,D.,&Strogratz,S.(1998).Collectivedynamicsofsmallworlds.Nature,393,440-442.Whittington,M.A.,Traub,R.D.,&Jefferys,J.G.R.(1995).Synchronizedoscillationsin

interneuronnetworksdrivenbymetabotropicglutamatereceptoractivation.Nature,373,612-615.

Wilson,D.M.,&Wyman,R.J.(1965).Motoroutputpatternsduringrandomandrhythmic

stimulationoflocustthoracicganglia.BiophysicalJournal,5,121-143.Wimsatt,W.C.(2007).Re-engineeringphilosophyforlimitedbeings:Piecewise

approximationstoreality.Cambridge,MA:HarvardUniversityPress.