unseenreadnovel

ORIGINALPAPERBrainRoutesforReadinginAdultswithandwithoutAutism:EMEGEvidenceRachelL.
MoseleyFriedemannPulvermu¨llerBettinaMohrMichaelV.
LombardoSimonBaron-CohenYuryShtyrovPublishedonline:9June2013TheAuthor(s)2013.
ThisarticleispublishedwithopenaccessatSpringerlink.
comAbstractReadingutilisesatleasttwoneuralpathways.
Thetemporallexicalroutevisuallymapswholewordstotheirlexicalentries,whilstthenonlexicalroutedecodeswordsphonologicallyviaparietalcortex.
Readerstypicallyemploythelexicalrouteforfamiliarwords,butpoorcomprehensionplusprecocityatmechanically'soundingout'wordssuggeststhatdifferencesmightexistinautism.
CombinedMEG/EEGrecordingsofadultswithautisticspectrumconditions(ASC)andcontrolswhilereadingrevealedpreferentialrecruitmentoftemporalareasincontrolsandadditionalparietalrecruitmentinASC.
Furthermore,alackofdifferencesbetweensemanticwordcategorieswasconsistentwithprevioussuggestionthatpeoplewithASCmaylacka'default'lexical-semanticprocessingmode.
Theseresultsarediscussedwithrefer-encetodual-routemodelsofreading.
KeywordsReadingDual-routemodelHyperlexiaSemanticsEEGMEGIntroductionDespitemultipleconceptualreformationssinceKanner's(1943)classicautismdescription,language/communicationabnormalitiesandimpairmentshaveremainedacornerstoneofthediagnosisofautismspectrumconditions(ASC).
Withintheauditorydomain,childrenandadultswithASClacktheneuralpreferenceandbehaviouralinclinationtowardsspeechsoundstypicallypresentfromaveryearlyage(Klin1991;Kuhletal.
2005;Muller2007;Groenetal.
2008;Laietal.
2012).
Thoughthismightbesecondarytoabroaderfailureinsocialorientation(Rapin1997;Dawsonetal.
1998;Swet-tenhametal.
1998;Schultzetal.
2000),itwouldappeartobeindependentofsensorydecitsandhasbeenarguedtobespecictohumanspeech(Cˇeponieneetal.
2003).
Thesestudiessuggest,therefore,thatlinguisticstimulimaybetreatedinaqualitativelydifferentwaywithintheautisticbrain.
Thisisequallytrueinthevisualdomain,thoughtheprocessingofwrittenwordsinASChasreceivedlessattention.
Gaffreyetal.
(2007),inanfMRItaskofsemanticdecision,discoveredunusuallyelevatedrecruitmentofvisualcortex(striateandextrastriateareas,BA17,18,19).
SimilarstrongrecruitmentofextrastriatecortexduringasentenceprocessingtaskwasreportedbyKanaetal.
(2006).
ElectronicsupplementarymaterialTheonlineversionofthisarticle(doi:10.
1007/s10803-013-1858-z)containssupplementarymaterial,whichisavailabletoauthorizedusers.
R.
L.
Moseley(&)Y.
ShtyrovMRCCognitionandBrainSciencesUnit,15ChaucerRoad,CambridgeCB27EF,UKe-mail:rachel.
moseley@mrc-cbu.
cam.
ac.
ukF.
Pulvermu¨llerBrainLanguageLaboratory,DepartmentofPhilosophyandHumanities,FreieUniversita¨t,Berlin,GermanyB.
MohrAngliaRuskinUniversity,Cambridge,UKM.
V.
LombardoS.
Baron-CohenDepartmentofPsychiatry,AutismResearchCentre,UniversityofCambridge,Cambridge,UKY.
ShtyrovCentreforFunctionallyIntegrativeNeuroscience,AarhusUniversity,Aarhus,Denmarke-mail:yury.
shtyrov@cn.
au.
dkY.
ShtyrovCentreforLanguagesandLiterature,LundUniversity,Lund,Sweden123JAutismDevDisord(2014)44:137–153DOI10.
1007/s10803-013-1858-zSinceloweractivityinBA45duringsemanticprocessinghasalsobeenreported(Harrisetal.
2006),severalauthorshavesuggestedaqualitativelydifferentstrategyforlexi-cosemanticprocessinginautism(KamioandToichi2000;ToichiandKamio2001,2002,2003;Gaffreyetal.
2007):onethat,somewhatimmature,reliesexcessivelyonvisu-alisationandperceptualprocessingattheexpenseofdeepsemanticanalysisofthevisually-orverbally-presentedlinguisticmaterial.
Ja¨rvinen-Pasleyetal.
(2008)commentedthat,inautism,''semantic-levelprocessingisnottheprimary,or'default'speechprocessingmode''(pp.
117).
Indeed,processingthesemanticratherthansurfacevisualfeaturesofwordsdoesnotleadtostrongerrecallinpeoplewithautism,unlikeinthetypicalpopulation(the'levelsofprocessing'effect:ToichiandKamio2002;Harrisetal.
2006;Lombardoetal.
2007).
Furthermore,theydonotbenetfromsemanticcuesinrecall(Mottronetal.
2001)orsemanticprimesindecisiontasks(Kamioetal.
2007),andfailtousesemanticchunkingstrategiesduringprocessing(HermelinandO'Connor1970).
ThismightexplainwhysemanticprocessingabnormalitiesandsubtleimpairmentsareconsideredahallmarkofASCbymanyauthors(Harrisetal.
2006;Walenskietal.
2006;Gaffreyetal.
2007;Braeutigametal.
2008).
Theseabnormalitiesmaycontributetodifcultieswithreadingcomprehensionthatarerevealedbylowerscoresinstandardisedbatteries(Venteretal.
1992;Mylesetal.
2002;Nationetal.
2006;Newmanetal.
2007).
Whenreadingtextandphrases,severalstudiesreportedthatparticipantswithASCfailtoutilisesemanticcontexttoinfersometimesambiguousmeaning(FrithandSnowling1983;Happe1997;WahlbergandMagliano2004),makeerrorsthataresemanticallyinappropriate(thoughsyntacticallycorrect)whenllingblankspaces(FrithandSnowling1983;Snow-lingandFrith1986),andhavedifcultyansweringquestionsbasedonpassages(O'ConnorandHermelin1994).
Thesereportssuggestthatindividualswithautismmighthavedifcultyreadingformeaningand/orinactivatingsemanticprocesses,particularly,asinthewordsofJa¨rvinen-Pasleyetal.
(2008),whennotexplicitlyaskedtodoso(though,consistentwithgreaterstrengthsinperceptualprocessing,theycanbenetfromimplicitsemanticcuespresentedpic-torially[KamioandToichi2000;Sahyounetal.
2010]).
Incontrast,thepreviousliteraturesuggeststhatthetypicalpopulationshowsemanticactivationrelatedtosensorimotorassociationsofwordsevenwithoutexplicitprocessinginstructionsorfocusedattention—suggestingautomaticactivationofneuralcircuitsrepresentingwordmeaning(Pulvermu¨lleretal.
2005;Pulvermu¨llerandShtyrov2006;Gonzalezetal.
2006;Hauketal.
2008;Kieferetal.
2008;Shtyrovetal.
2004,2010;Barros-Loscertalesetal.
2011).
Intypicalindividuals,thisactivityreectsdifferentialbraintopographiesfortherepresentationofwordswithdifferentmeanings.
Actionwords,forexample,havebeenstronglyassociatedwiththecorticalmotorsystem,eveninaspecif-icallysomatotopicmannerthatreectstheirassociationwiththeeffectorsofthebody(Hauketal.
2004,2008;Pulver-mu¨lleretal.
2005;Tettamantietal.
2005;Aziz-ZadehandDamasio2008;Kemmereretal.
2008).
Incontrast,wordsforobjectswithstronglyvisualassociationsevokeactivityinthetemporo-occipitalobjectprocessingstream(Warburtonetal.
1996;Pulvermu¨lleretal.
1999;MartinandChao2001;Martin2007).
Sincepeopletendtolearnthewordforanactionorobjectinthecontextofexperiencing/interactingwithit,suchorganisationisproposedtoarisethroughHeb-bianprinciplesduetothesimultaneousactivationofsenso-rimotorperceptualregionsandcoreperisylvianlanguagecortex(Pulvermu¨ller2001).
Consequently,wordphonology,articulatoryfeaturesandmeaningarerepresentedatabrainlevelindistributedneuronalassembliesreachingintoactionandperceptionpartsofthebrain(''action-perceptioncir-cuits'':seePulvermu¨llerandFadiga2010).
WhathappensinthebrainwhenpeoplereadwrittenwordsOnetheorysuggeststhattherearetwoneuralroutesthroughwhichwrittensymbolsonthepagearetransformedintomeaningfulunits(Coltheartetal.
2001).
Inonestrategy,wholevisualword-formsaremappeddirectlyontotheircorre-spondinglexicalentries,thustransparentlymatchingsymboltomeaning.
Thislexicosemanticroute,otherwiseknownasthe'direct'pathwayfromwordtomeaning(McCarthyandWar-rington1986;Coltheartetal.
2001),isassociatedwithaventralpathway,whichinvolvesactivationofleft-hemisphericoccipito-temporalareassuchasthefusiformgyrus(Fiebachetal.
2002;Jobardetal.
2003;Levyetal.
2009),typicallyimplicatedinvisualword-processing(Cohenetal.
2002;Kronbichleretal.
2004).
Incontrast,adorsalpathwaypro-cesseswrittensymbolsinapiecemealmanner,convertinggraphemestotheirauditoryphonemecounterparts,whichcanthenbespokenaloudorfurtherprocessedformeaningviatheirpronunciations.
Thisgrapheme-phonemeconversion(ornon-lexical)routeisassociatedwithleftparietalcortex,includingsuperiorparietallobule,inferiorparietalandsupramarginalgyrus,andalsoparsopercularis(Fiezetal.
1999;Jobardetal.
2003;Mechellietal.
2003;Levyetal.
2009),knowntobeinvolvedingeneralphonologicalprocessing(Paulesuetal.
1993;Fiez1997;Poldracketal.
1999;McDermottetal.
2003).
Theexistenceofdorsalandventralroutesforlanguagepro-cessinghasbeenequallysupportedintheauditorydomain,where,likevisualletters,soundsaremappedtoarticulationviathedorsalconnectionsofthearcuateandsuperiorlongitudinalfasciculus;higher-levelmeaningcomprehensionisservedbytheextremecapsuleinaventralstreamlinkingtemporaltoinferiorfrontalstructures(Sauretal.
2008).
Whilstskilledreadersmayutiliseandshiftbetweeneitherpathway,modulatedbyfeaturesofthewrittenwordssuchasfrequency,transparencyandorthographicregularity138JAutismDevDisord(2014)44:137–153123(ZevinandBalota2000),thereisevidencethathighlyfre-quent,familiarwordsarepreferentiallyprocesseddirectlyviathelexicosemanticrouteinaholisticfashion(ColtheartandRastle1994).
However,theaforementionedproblemswithwordcomprehensionwouldsuggestthatthesamemaynotbetrueinautism.
Interestingly,thisdifcultyoftenpresentsinconjunctionwithhyperlexia(Healyetal.
1982;WhitehouseandHarris1984;Goldberg1987;SmithandBryson1988;PattiandLupinetti1993;O'ConnorandHermelin1994;Grigorenkoetal.
2002;Newmanetal.
2007),whichearlyaccountsdenedasa''compulsiontodecodewrittenmaterialwithoutcomprehensionofitsmeaning''(WhitehouseandHarris1984)butwhichisalsooftendenedasbeingabletoreadbeforetheageofstartingschool.
Compulsivehyperlexiahasalsobeenobservedinstrokepatientsasa''releasephe-nomenon''followingbraindamage(Berthieretal.
2006).
Inautism,thisdecodingskillpossessesasavant-likequality,generallyfaroutstrippingreadingcomprehension:alongwiththeabilitytoreadnovelpseudowords(FrithandSnowling1983;Nationetal.
2006;Newmanetal.
2007),thissuggeststheintegrityofthegrapheme-phonemecon-versionrouteinautism,andthatthisrouteisperhapsenhancedandrelieduponratherthanwhole-wordmatching(Arametal.
1984;GoldbergandRothermel1984;AramandHealy1988).
However,neuroscienticevidenceforanover-emphasisonasemanticreadinginASC,evenforfamiliarwords,isstillnotavailable.
Inordertoinvestigatetheneuralroutesforvisualword-processingintheautisticbrain,weusedcombinedelec-troencephalographyandmagnetoencephalography(EEG/MEGorEMEG)tocomparethetime-courseandlocali-sationofbrainactivityinsubjectswithanASCwithtypicalcontrols.
Apassivereadingtaskwasemployedtoinvesti-gatepathwaysactivatedbyreadingshort,simplewords.
Apassiveperceptualparadigmhaspreviouslybeenusedinthetypicalpopulationtoinvestigatetheprocessingofdif-ferentsemanticcategories,whichevokedifferentialpat-ternsofneuralactivity(Hauketal.
2004;Gonzalezetal.
2006;Barros-Loscertalesetal.
2011),evenatearlylaten-ciesandwithoutconsciousattention(Shtyrovetal.
2004;Pulvermu¨lleretal.
2001,2005;MoscosodelPradoMartinetal.
2006;Hauketal.
2008).
Giventheaforementionedliteratureonreadingcomprehensionandsemanticpro-cessinginASC,itisunclearwhetherthesameistrueinASC.
Asthepresentexperimentinvolvedthesamepassivereadingparadigmwithwordsofdifferentsemanticmean-ing,wethereforealsodecidedtolookatdifferencesbetweenwordcategorieswithinourstimulusset,inordertoinvestigatewhethersemanticcategory-specicdiffer-encesalsoariseautomaticallyinASCastheydointhetypicalpopulation.
MethodsParticipants14Participantswithhigh-functioningASC(13withAsper-gers'Syndrome,1withPDD-NOS)and17typically-devel-opedcontrolparticipantstookpartinthestudy,allmonolingualnativespeakersofEnglish.
Thegroupswerematchedforfull-scaleIQasmeasuredbytheCattellCultureFairtest(CattellandCattell1960)(115.
8forcontrolsand118.
5forASCrespectively:t[29]=.
389,p\.
700),andwith11malesinthecontrolgroupand7intheASCgroup,containedaroughlyequaldivisionofsexratio,withnosignicantdifferenceinthis(t[29]=.
808,p\.
430).
Bothgroupswereright-handed,thoughscoresontheEdinburghHandednessInventory(Oldeld1971)indicatedthattheASCgroupwereslightlylessstronglylateralised(t[29]=2.
249,p=.
032).
EligibilityforthestudyrequiredthatallASCparticipantshadreceivedaformalclinicaldiagnosisusingDSM-IVcri-teria.
OntheAutismSpectrumQuotient(AQ:Baron-Cohenetal.
2001),theyscoredsignicantlyhigher(37.
3±SD9.
9)thandidthecontrolgroup(13.
8±5.
7;t[29]=8.
126,p\.
001),indicatingasignicantlygreaternumberofautistictraits.
Itwasnotpossibletofullymatchthemeanageofthegroups,withtheASCgroupbeingslightlyolderthancontrols(31.
4±8.
2yearsvs.
25.
0±5.
1years;t[29]=2.
638,p\.
014).
MaterialsThestudyemployedanextensivecorpusof360wordsmatchedforlength,letterbigramandtrigramfrequencyandnumberoforthographicneighbours,alongwith120length-matchedhash-markstringsthatactedasalow-levelvisualcontrolcondition.
ThesepsycholinguisticpropertieswereretrievedfromtheCELEXdatabase(Baayenetal.
1993).
PriortotheEMEGexperiment,asemanticratingstudywasperformedbyagroupof10nativeEnglishspeakers(seePulvermu¨lleretal.
1999,forproceduraldetails)inordertoobtainparticipantratingsforeachwordonanumberofsemanticvariables,includingsensorimotorfeatures(im-ageability,concretenessandaction-relatedness)andaffec-tive-emotionalfeatures(arousalandvalence).
Inaccordancewiththesesemanticratings,the360experimentalwordsconsistedof120action-related(e.
g.
''knead'',''jog''),120object-related(e.
g.
''hawk'',''cheese''),and120abstract(e.
g.
''faze'',''uke'')wordswhichwereusedhereasllers.
Naturally,duetotheirsemanticassociations,thesewordcategoriesdifferedinaction-relatedness,imageability,andothersemanticvariables:pleaseseeOnlineResource1fordetailsoftheirpsycholinguisticandsemanticproperties.
JAutismDevDisord(2014)44:137–153139123ProcedureHavinggiveninformedconsent,participantscompletedtheCattellCultureFairtest(CattellandCattell1960),theEdin-burghHandednessInventory(Oldeld1971)andtheAQ(Baron-Cohenetal.
2001)priortoEMEGpreparation.
Oncepreparedfortherecording,participantsweremadecomfort-ableandrequestedtostayasstillaspossible,avoidingallunnecessarymovements,andtofocusonacentralxationpointwhilstattendingtothestimuliappearingonthescreen.
Theexperimentaltask,splitintothreeblocksofapproxi-mately7mineach,involvedpassivereadingoftheexperi-mentalstimuliwhichwerepresentedtachistoscopicallyfor150ms,inlightgreyfontonablackbackground,withaninter-stimulusintervalof2,500ms.
Soastoavoidordereffects,twopseudo-randomisedstimuluslistswerecounter-balancedbetweensubjectsinbothgroups.
Betweeneach7minblockoftheexperimentaltask,participantsweregivenacoupleofmomentstorestifrequired.
Followingtheexperimentalproceduretocheckatten-dancetothetask,participantsweregivenanunseenwordrecognitiontestcontainingacombinationof50experi-mentaland25noveldistractorwordschosenfromabankoflength-matchedwordswhichdidnotmakethenalstimulusselection.
Nodifferencesinperformanceemergedbetweenthetwogroups(t[29]=1.
721,p\.
110)andbothperformedabovechance(averagehitrateforcontrols:82±8.
6%;averagehitrateforASC:74±14.
8%).
EMEGRecordingandDataPre-processingElectroencephalogram(EEG)andmagnetoencephalogram(MEG)weresimultaneouslyrecordedinamagnetically-andacoustically-shieldedMEGbooth(IMEDCOCorp,Switzerland).
EEGwasrecordedfromelectrodecaps(EasyCap,FalkMinowServices,Herrsching-Breitbrunn,Germany)with70Ag/AgClelectrodesarrangedaccordingtotheextended10/10%system.
ForMEG,thestudyemployedawhole-head306-channelMEGsetupof204planargradiometersand102magnetometers(ElektaNeu-romag,Helsinki,Finland),whichcontinuouslyrecordedmagneticeldsandeldgradientsduringthetask.
Headpositionwastrackedthroughoutthesessionusing5mag-neticcoils,attachedtotheEEGcap,whosepositionwithrespecttothreestandardisedpoints(nasion,leftandrightpre-auricularpoints)wasdigitisedusingthePolhemusIsotrakdigitaltrackersystem(Polhemus,Colchester,VT,USA).
Furtheranatomicalco-registrationwithMRIscanswasmadepossiblethroughadditionaldigitisationofEEGelectrodesandrandomisedpointsdistributedoverthescalp.
Inordertorejecttrialsdisturbedbyblinkingoreyesaccades,eyemovementsweremonitoredbyfourEOGelectrodesplacedlaterallytoeacheye(horizontalEOG)andverticallyaboveandbelowthelefteye(verticalEOG).
RecordingswerepreprocessedofineusingMaxFiltersoftware(ElektaNeuromag,Helsinki),whichemploystheSignal-SpaceSeparationmethod(TauluandKajola2005;Tauluetal.
2004)tominimiseexternalnoiseandsensorartefacts,alongwithspatio-temporallteringandhead-movementcompensationtocorrectforbetween-blockmovements;anybadEEG/MEGchannelswereidentiedandre-interpolated.
TheMNE2.
7softwarepackage(A.
MartinosCenterforBiomedicalImaging,Charlestown,MA,USA)wasusedthroughouttherestoftheanalysis.
Datawereband-passlteredbetween0.
1and30Hz.
Foraver-aging,epochsof500msweretakenfrom50mspriortostimulusonset:forbaselinecorrection,meanamplitudeoverthis50-msintervalwaslatersubtractedfromthesignalatalltime-points.
Epochswithanamplitudeexceeding150lVinEEGandEOGchannelsand2,000fT/cmand3,500fTingradiometerandmagnetometerchannelsrespectivelywerediscarded,andremainingepochswereaveragedwithinindividualsforeachstimulustype.
Foranunbiasedestimateoftheoverallneuraldynamicsinresponsetoverbalstimuli,aglobalsignal-to-noiseratio(SNR)wascalculatedforallparticipantspooledbydividingamplitudeateachtime-pointbythestandarddeviationinthebaselineperiod(therst50ms)andthencomputingtheroot-meansquareofSNRacrossallsensors.
PeaksandtroughsonthisglobalSNRcurve,averagedacrossallparticipants(Fig.
1),wereiden-tied,andthesetimeperiodsweresubjectedtofurthersourcereconstructionandstatisticalanalysis.
MRIAcquisitionandEMEGSourceReconstructionInordertoexploretheneuronalgeneratorsunderlyingelectrophysiologicalandneuromagneticactivity,L2mini-mumnormsourceestimations(Ha¨ma¨la¨inenandIlmoniemi1994)forcombinedEEG/MEGdatawerecomputedusingMNEandFreesurfer4.
3software(MartinosCentreforBiomedicalImaging)inconjunctionwithindividualsubjectstructuralMRIimagesusedtomodelcorticalgreymattersurface.
High-resolutionstructuralT1scansforeachsubjectwereacquiredwitha3TSiemensTimTrioMRIscanner(parametersoftheMPRAGEsequencewereasfollows:eld-of-view256mm9240mm9160mm,matrixdimensions25692409160,1mmisotropicresolution,TR=2,250ms,T1=900ms,TE=2.
99ms,ipangle9°).
TheywerepreprocessedandcoordinatesalignedtoEMEGdatausingdigitisedpositionsoftheanatomicallandmarks,electrodesandtheheadsurface.
A3-shellboundary-elementmodelforeachsubject,usinginnerandouterskullandskinsurfaces,wascreatedusingawatershedalgorithm.
Sourceestimatesforeachstimulustypewerecomputedforeachsubjectandthenmorphedtotheaverage140JAutismDevDisord(2014)44:137–153123brain(averagedfromallsubjectspooled),andgrandaver-agesforcontrolandASCgroupswerethencomputedtobedisplayedontheinatedaveragecorticalsurface.
Sourceactivationsforwordscomparedwithcontrolconditioninthegrandaveragescalculatedforbothgroupswerestatisticallyexploredinaregions-of-interest(ROI)approach.
Secondly,category-specicdifferencesbetweenthebroken-downcategoriesofaction,objectandabstractwordswereinvestigated.
ROIswereanatomicallydenedbasedontheDesikan-KillianyAtlassubdivisionsofthebrain(Desikanetal.
2006)asimplementedintheFreesurferpackage.
Wethenanalysedsourcedynamicsinthoselobesofthebrainwherereading-relatedactivitycanbeexpected,namelyoccipital,parietal,temporalandfrontallobes,whichincludedthefollowingstructures:(1)frontalcortices(cov-eringsuperiorfrontal,middlefrontaldorsal,middlefrontalventral,caudalfrontal,BA47,BA45,BA44,precentral,paracentral),(2)temporalcortices(superiortemporal,mid-dletemporal,inferiortemporal,fusiform),(3)parietalcor-tices(postcentral,supramarginal,superiorparietal,inferiorparietal),and(4)occipitalcortices(BA17,BA18dorsal,BA18ventral,BA19dorsal,BA19ventral).
PleaseseeOnlineResource2foradepictionofregions.
Bothleft-hemisphericcorticesandtheirright-hemispherehomologueswereana-lysed.
Wheredifferencesappeared,individualregionswerefurtherexplored.
Inthismoredetailedanalysis,threelargeregions(middlefrontalcortex,precentralstripandoccipitalcortex)weresubdividedintodorsal–ventralportionsinaccordancewiththesameanatomicalguide,inordertoassessmorene-grainedgroupdifferences.
Amplitudesofthesourcecurrentswithintheselobes/ROIswerecalculatedinthetime-windowsofinterestdenedthroughinspectionoftheSNRcurve,asdescribedabove.
Withallstatisticalanalysis,Huynh–Feldtcorrectionwasappliedtocorrectforsphericityviolationswhereverappro-priate.
Correctedpvaluesarereportedthroughout.
ResultsVisualinspectionoftheglobalSNRcurverevealedseveralpeaksandwindowsforfocus(seeFig.
1).
ThesignalFig.
1GLOBALsignal-to-noiseratio(SNR,forallsubjectspooled)curveforallwordsduringthe500msepoch,andactivationforallwordsdepictedwithinthevetime-windowsoffocus.
Forthesourceestimations,activityintheleft(top)andright(middle)hemisphereshasbeenpooledforbothsubjectgroupsineachtime-windowJAutismDevDisord(2014)44:137–153141123demonstratedasharpincreasefrom*70msonwardswithapeakaround150msfollowedbyadownstrokeandaplateau.
Wethereforeanalysed(1)theupstrokeperiodbetween70and130ms,(2)thepeakintervalat140–160ms,and(3)thedeclineofthispeakandthestartofthefollowingplateauat170–250ms.
Wealsostudiedlaterperiodsoftheepoch,capturingthewavebetween300and375ms,andthenalstretch,between375and450ms,giventhepreviouslitera-tureonlexicalandsemanticeffectsinM350(Embicketal.
2001;PylkkanenandMarantz2003)andN400(KutasandFedermeier2011;Lauetal.
2008)timeranges.
AscanbeseeninthesourceestimationsinFig.
1,writtenwordstimulievokedwidespreadactivityacrossvisualareasandperisylvianlanguageregions,includingthelengthofthetemporalcortexandtheinferiorfrontalgyrus,alongsideadditionalmotor,parietalandfrontalactivity.
Withtheexceptionofthersttime-window,activityintheseregionsappearsslightlystrongerandmorewide-spreadinthelefthemisphere.
Initially,withinthe70–130time-windowwherebrainresponsesrstdifferentiatebetweengroups(seebelow),themajorityofactivityoccurredinprimaryvisualcortex,thoughactivationalsopresentsininferiorfrontalcortex.
Activitywasseentospreadinananteriorfashionalongthetemporalcortex,increasinginthetemporalpoleanddecreasinginposteriortemporalregionsbythelatetime-windowsin300–450msrange.
GeneralReadingwithDifferentPathwaysAninitialROIanalysisfocusedondifferencesinbrainactivationbetweengroupsinthesetime-windows(Fig.
2).
Fig.
2SOURCEestimationsreectcontrastsbetweenthetwogroups:areasofgreateractivityforcontrolthanASCparticipantsinblue,areasofgreateractivityforASCthancontrolparticipantsinred.
Sourceestimatesareaveragedacrosseachtime-windowoffocus.
Time-windowsinwhichgroupdifferencesweresignicantspeci-callyduringwordbutnothash-markreadingaremarkedbyanasterisk(*)142JAutismDevDisord(2014)44:137–153123Ineachtime-windowindependently,anANOVAwascon-ductedtoexamineactivityineachlobe(frontal,temporal,parietalandoccipital),eachofwhichwassplitintoanumberofindividualROIs(seeMethodsfordetails).
Wheregroupdifferenceswereindicatedwithinlobes(i.
e.
atthelevelofROIs),thesewithin-loberegionswereexploredinANOVAsincludingthelevelofGroup(2)andROI.
TheseANOVAswererunindependentlyineachtime-windowforboththeword-readingandthehash-markcondition,buttheresultsdiscussedbelowarefortheword-readingconditionunlessexplicitlystated.
Areaswheremaineffectsofgrouparoseineachtime-windowarereportedinTable1,boldedforresultswhichwerespecicfortheword-readingcondition.
Allresultsaresummarisedbelow.
Overall,themoststrikingobservationoftheanalysiswasacontrastbetweensubjectgroupsduringwordreadinginwhichcontrolparticipantsshowedaventralspreadofactivationwhilstthosewithASCexhibitedactivationofthedorsalparietalroute.
AscanbeseeninFig.
2(PartA)andTable1,thedorsaltrendfortheASCgroupwaspre-dominantlynon-specicforwordsuntilthepeakoftheSNRcurve(140–160ms:seeFig.
2,PartB),whereaninteractionofROI,hemisphereandgroupwasdrivenbygreateractivityinparietalcortexfortheASCgroup.
TheASCgroupalsoshowedgreaterword-specicactivitythancontrolsinleftparietalcortexinthe170–250ms(seeFig.
2,PartC)and300–375ms(seeFig.
2,PartD)time-windows.
Thisgreaterword-specicactivityinthelatterwindowalsoincludedotherpartsofthedorsalpathwayforphonologicalprocessing,namelyparsopercularis(BA44)anddorsalprecentralgyrus.
Inthe170–250mstime-window,however,signicantinteractionsemergedfromanANOVAincludingthelobesofthereadingroutes(tem-poralandparietalcortices)alongwiththefactorsROI(4:superiortemporal,middletemporal,inferiortemporal,fusiformgyrus;postcentralgyrus,supramarginalgyrus,superiorparietal,inferiorparietal),andgroup(2):thesereectedthatwhilstASCparticipantsshowedgreaterword-readingactivitythancontrolsinleftparietalcortex,thelattergroupshowedgreateractivityinlefttemporalcortexincontrasttoASCparticipants.
Thisdorsal/ventraldistinctionbetweengroupsbegantotailoffinthenaltime-window,375–450ms(Fig.
2,PartE),thoughaninteractioninleftfrontallobereectedatendencyfortheASCgrouptostillshowgreateractivationinmoredorsalregionssuchasBA44andforcontrolparticipantstoshowgreateractivationinmoreventralregions.
ThoughtherewasapatternofgreaterdorsalactivityintheASCgroupandgreaterventralactivityincontrols,asecondaryanalysisfocusedatawithin-grouplevelandcomparedactivationbetweeninparietalandtemporalcorticesineachtime-window.
Thecontrolgroupshowedgreateractivationofthetemporalthanparietalcortexforword-readinginthe170–250mstime-window(f(1,16)=38.
124,p\.
001).
TheASCgroup,incontrast,showednosignicantdifferencebetweentemporalandparietalcorti-ceswhilstreading.
SemanticCategory-SpecicitySemanticdifferencesbetweenwell-matchedwordcatego-rieshavebeenreportedintypically-developingsubjectsacrosstherangeoftimethatwestudied,beginningasearlyas100ms(Pulvermu¨lleretal.
2001).
Inaccordance,weinvestigatedourdatasetforcategory-specicgroupdif-ferencesbetweenaction-,object-andabstractwordsinthetime-windowspreviouslydened.
Asinthepreviousanalysis,wordcategoryeffectsforeachtime-windowwereexploredineachlobeforeachgroupindividually,usinganANOVAemployingthelevelsROI(individualROIsofeachlobe:seeMethodsfordetails),WordCategory(3levels:action,objectandabstractwords),andhemisphere(2).
Ofthetime-windowsdenedthroughinvestigationoftheSNRcurveforallwords,category-specicdifferenceswereseenonlyinthe140–160mspeakandthe170–250mstime-windows.
Aspreviousworkhasillus-tratedthatshorttime-windowsmaybebesttocapturefocalandtemporally-briefsemanticdifferences(Pulvermu¨lleretal.
2009),weattemptedtoadditionallyscrutinisecate-gory-specicdifferencesinshortwindows20msbeforeandafterthemainpeak(140–160ms).
Wordcategoryeffectsforeachtime-windowsofinterestarelistedinTable2,alongsideposthoct-testswhichdeterminedthenatureofsemanticeffectsindifferentbrainregions.
Typically-developedcontrolparticipantsshowedaclearpatternacrossalltime-windowswhereactionwordsdomi-natedinfrontalbrainregions.
Thiswasmostrobustintherightprecentralcortex,whereactionwordsevokedgreateractivitythanobjectorabstractwordsfrom140to180ms.
Incomparison,objectwordsactivatedposteriorbrainregionsmorestronglythanotherwordcategory.
Mostnotably,t-testsshowedthattheyevokedsignicantlygreateractivitythanbothactionandabstractwordsintheleftfusiformgyruswithinthe140–160mstime-window,thoughtheywerethedominantsemanticcategoryinallregionslisted.
Incomparisonwiththecontrolgroup,veryfewwordcat-egoryeffectswerefoundforASCparticipants.
Thesewerelimitedtoeffectsinfrontalregionswhichrevealedaverydifferentpatternofactivitytothatseeninthecontrolgroup:greateractivityforobjectwordsthanforotherwordcategories.
ThesestatisticalresultsarereectedinFig.
3,whichdisplaysactivationmapsforeachgroupduringthetime-windowsofinterest.
Asthemajorityofliteraturefocusesonthedistinctionbetweenaction-andobject-relatedwordsandthisisourkeyinteresthere,onlythiscomparisonisdisplayed(ratherthancomparisonswithabstractwords).
JAutismDevDisord(2014)44:137–153143123Table1Groupdifferencesforword-readingineachtime-windowMaineffectsofgroupInteractionsControl[ASCASC[Control70–130msNomaineffectsofgroupBilateralsuperiorparietalcortex(f[1,29]=5.
511,p\.
03)Bilateralinferiorparietalcortex([1,29]=8.
601,p\.
01)BilateraldorsalBA19(f[1,29]=2.
266,p\.
03)Bilateralprecentral*andparacentralgyrus(f[1,29]=4.
471,p<.
05)*t-testsofsegmentedprecentralregions:Dorsal(t[29]=2.
350,p<.
03)Middle(t[29]=2.
279,p<.
03)Nointeractions140–160msNomaineffectsofgroupNomaineffectsofgroupBilateralparietalcortex:ROI(4)3hemisphere(2)3Group:f[3,87]=2.
843,p<.
05170–250msLefttemporalcortex(f[1,29]=4.
590,p<.
05)L.
inferiortemporalgyrus(t[29]=2.
793,p<.
01)L.
fusiformgyrus(t[29]=1.
944,p<.
065)Leftparietalcortex(f[1,29]=4.
550,p<.
04)L.
supramarginalgyrus(t[29]=2.
173,p<.
04)L.
superiorparietalcortex(t[29]=2.
244,p<.
04)L.
inferiorparietalcortex(t[29]=1.
902,p<.
07)Leftparietalandtemporalcortices:Lobe(2)3Group:f[1,29]=12.
905,p<.
005Lobe(2)3ROI(4)3Group:f[3,87]=4.
883,p<.
01300–375msNomaineffectsofgroupLeftparietalcortex(f[1,29]=5.
081,p<.
04)L.
parsopercularis(BA44)(BA44:t[29]=2.
211,p<.
04)Precentral*andparacentralgyrus(f[29]=4.
249,p<.
05)*t-testsofsegmentedprecentralregions:Dorsal(t[29]=2.
771,p<.
02)Middle(t[29]=2.
612,p<.
02)Leftfrontallobe:ROI(9)3Group:f[8,232]=2.
813,p<.
005375–450msNomaineffectsofgroupNomaineffectsofgroupLeftfrontallobe:ROI(9)3Group:f[8,232]=3.
261,p<.
04Signicantinteractionsandmaineffectsofgroupfoundineachtime-windowwhilstreading.
ThesecondcolumnindicatesareaswherecontrolparticipantsshowedgreateractivationthanindividualswithASC,whilstthethirdcolumnreectstheoppositepattern.
Boldtextindicatesinteractionsandgroupdifferenceswhichwerespecicforwordsinthatparticularregionandwhichdidnotappearforthehash-markcondition.
Marginallynon-signicanteffectsareshowninitalicisedfont,boldedwheretheyreferspecicallytotheword-readingcondition144JAutismDevDisord(2014)44:137–153123Table2Maineffectsandposthoct-testsforwordcategoriesineachgroup120–140ms140–160ms160–180ms170–250msControlgroupBilateralsuperiorfrontalcortex:f[2,58]=7.
906,p\.
005L.
hemisphere:Action[object(t[16]=2.
154,p\.
05)Action[abstract(t[16]=2.
207,p\.
05)R.
hemisphere:Action[abstract(t[16]=2.
522,p\.
03)Bilateralfusiformgyrus:f[2,58]=5.
494,p\.
01L.
hemisphere:Object[abstract(t[16]=2.
460,p\.
03)Object[action(t[16]=2.
060,p\.
06)R.
hemisphere:Object[abstract(t[16]=2.
415,p\.
03)BilateralventralBA19:f[2,32]=4.
702,p\.
03L.
hemisphere:Object[abstract(t[16]=2.
435,p\.
03)R.
hemisphere:Object[abstract(t[16]=2.
747,p\.
02)Bilateralsuperiorfrontalcortex:f[2,32]=4.
221,p\.
03L.
hemisphere:Action[abstract(t[16]=3.
195,p\.
01)BilateralBA44:f[2,32]=4.
603,p\.
02R.
hemisphere:Action[abstract(t[16]=2.
674,p\.
02)Rightprecentralcortex:f[2,32]=4.
429,p<.
03Action[object(t[16]=2.
784,p\.
02)Action[abstract(t[16]=2.
295,p\.
04)Leftfusiformgyrus:f[2,32]=5.
705,p\.
01)Object[action(t[16]=3.
084,p\.
01)Object[abstract(t[16]=2.
718,p\.
01)LeftBA17:f[2,32]=3.
844,p\.
035Object[abstract(t[16]=2.
806,p\.
02)BilateralBA18:f[2,32]=3.
677,p\.
04Bilateral:Object[abstract(t[16]=2.
370,p\.
04)BilateralventralBA19:f[2,32]=4.
033,p\.
03Bilateral:Object[abstract(t[16]=2.
589,p\.
02)Rightprecentralgyrus:f[2,32]=3.
999,e=.
993,p\.
03Action[object(t[16]=2.
271,p\.
04)Action[abstract(t[16]=2.
293,p\.
04)Rightprecentralgyrus:f[2,32]=4.
718,e=.
796,p\.
02Action[abstract(t[16]=2.
777,p\.
02)JAutismDevDisord(2014)44:137–153145123Ascanbeseen,manymoreinstancesofcategory-speci-cityareevidentinthecontrolgroup:thesereectagreaterstrengthforobjectwordsinposteriorbrainregionsandgreateractivityforactionwordsinfrontalregions.
IntheASCgroup,thestrengthforobjectwordsinthe120–140mstime-windowwasquiteweak(ascanbeseeninTable2).
Astrongerdominanceforobjectwordsinprecentralgyruscanhoweverbeseeninthe170–250mstime-window.
DiscussionWhilstpooledsubjectdatainoursourceanalysisrevealedactivitytypicalduringvisualwordprocessing,furtherinves-tigationofthecombinedEEG/EMEGdatasetrevealedcleargroupdifferencesinseveraltime-pointsoftheepoch.
ByfarthemoststrikingobservationwasapatternwherebyactivationforcontrolsubjectsseemedtospreadinaventralfashionincontrasttothedorsalactivationpatternshownbyASCpar-ticipants.
Thelattergroupshowedsignicantlygreateracti-vationthancontrolparticipantsinparietalregionsacrosseachofthetime-windowsstudied.
Thiseffectdidnotinitiallydiscriminatebetweenwordsandthecontrolcondition(70–130ms),butfrom140msonwardswasword-specic.
Thissuggestsgreaterrecruitmentandrelianceonparietalregionsthanthatseeninthecontrolgroupandimplies,asinpreviousliterature(KamioandToichi2000;ToichiandKa-mio2001,2002,2003;Harrisetal.
2006;Kanaetal.
2006;Gaffreyetal.
2007),qualitativelydifferentprocessingandrecruitmentofneuralpathwaysinindividualswithautism.
Exploringdifferencesbetweensemanticcategoriesrevealedthatcontrolsubjectsshowedatypicalpatternofgreateractivityforobjectwordsinposteriortemporalregionsandgreateractivityforactionwordsinfrontalandmotorsystems,ashasalsobeenreportedinpreviousresearch.
TheASCgroupactuallyshowedareversalofthispattern,withgreateractivityforobjectthanactionwordsinbilateralsuperiorfrontalandprecentralgyrus,indicatingatypicalrepresentationofcon-ceptsinthebrain.
Thesewere,however,theonlycategorydifferencesseeninthispopulation.
PreviousliteraturehassuggestedthatpeoplewithASCmaynotautomaticallypro-cesswordsatasemanticlevelunlessexplicitlyaskedtodoso,andourdata,ndingveryfewcategoryeffectsfortheASCgroupincomparisontothetheory-congruentpatternseenincontrols,seemconsistentwiththisproposition.
Thesendingsarediscussedinmoredetailbelow.
ReadingPathwaysintheBrainareDifferentiallyRecruitedinASCSuccessfulreadinginvolvesexibleshiftingbetweentwopath-ways:aventral,lexicosemanticpathway(leftoccipito-temporalTable2continued120–140ms140–160ms160–180ms170–250msASCgroupBilateralsuperiorfrontalgyrus:f[2,26]=5.
116,e=1.
000,p\.
013L.
hemisphere:Object[action(t[13]=2.
074,p\.
06)R.
hemisphere:Object[action(t[13]=2.
216,p\.
05)Object[abstract(t[13]=2.
105,p\.
06)Nowordcategoryeffects.
Nowordcategoryeffects.
Leftprecentralgyrus:f[2,26]=4.
000,p\.
04Object[action(t[13]=3.
172,p\.
01)Statisticalresultsfromtheanalysisofdifferentsemanticwordcategory.
Maineffectsofwordcategoryarereportedinboldfont,whilstposthoct-tests,carriedouttoinvestigatethenatureofthesemanticdifferences,arereportedinstandardtext.
Marginallynon-signicantresultsaredisplayedinitalics146JAutismDevDisord(2014)44:137–153123cortex)engagedindirectmappingofwhole-wordformstotheirmeanings,andadorsal,grapheme-phonemeconversionroute(leftparietalcortex,parsopercularis)whichdecodeswrittenwordsinarule-driven,piecemealmanner(Coltheartetal.
2001;Jobardetal.
2003;Levyetal.
2009).
Intypicalreaders,previousresearchsuggeststhatthelexicosemanticpathwayispreferentiallyemployedintheprocessingofhighlyfrequentwords(ColtheartandRastle1994)which,beingveryfamiliar,canbematcheddirectlyandefcientlyontotheirlexicalentriesandtheirsemanticsretrievedwithoutthenecessityofpriorgrapheme-to-phonemedecodingoftheirphonologicalforms.
Thecurrentdatasupportthisinterpretation,ascontrolsubjectsshowedsig-nicantlygreateractivityinlefttemporalthanparietalregionsinthe170–250mstime-window.
Atthistime,word-specicactivityintheleftventralroutecomprisingofthetemporalcortexwassignicantlygreaterthandorsalrouteactivityforthecontrolgroup,andsignicantlyhigherthanintheASCgroupasreectedbyagroupdifference.
Thelattergroup,incontrast,showednopreferentialrecruitmentoftheFig.
3SOURCEestimationsforactionandobjectwordsforthecontrolandASCgroups(leftandrightrespectively)duringeachofthetime-windowsanalysedforcategory-specicity.
Activityinredreectsareasofgreateractivityforactionthanobjectwords,whereasactivityinbluereectsgreateractivityforobjectthanactionwords.
Asterisks(*)andcirclesreectareaswherewithin-groupposthoct-testsrevealedsignicantdifferencesbetweenactionandobjectwords:redcirclesindicatesignicantlygreateractivityforactionwordswhereasbluecirclesreectgreateractivityforobjectwordsJAutismDevDisord(2014)44:137–153147123lexicalrouteorindeedofeitherpathway,withnosignicantdifferencesbetweenthem.
Theyshowedinsteadanaddi-tionalrecruitmentoftheparietalcortex,signicantlydif-ferenttothecontrolgroup,fortheseregularly-spelt,familiarwords.
InitialparietalactivationbytheASCgroup(70–130ms)wasnotspecictowords—butinthesametime-window,activationindorsalprecentralareaswasword-specicforthisgroupandalsoindicatesutilisationofthisdorsalroutetoinferiorfrontalareas.
Indeed,asthetimefollowinginitialwordpresentationincreased,thistrendfordorsalactivityinASCbecamegreaterandword-specic,withgreaterword-elicitedacti-vationinpostcentralgyrus,supramarginalgyrus,superiorandinferiorparietalregionsintheASCgroupthanincon-trols.
Thistrendcontinuedtolatertime-windows,withgreaterword-specicASCactivationinleftparietalcortex,leftparsopercularisanddorsalprecentralgyrusat300–375ms,andgreaterword-specicactivationofparietalcortexat375–450ms.
Inthesameperiod,theASCsubjectsalsoshowedgreater,non-specicactivityinparsopercularis(BA44),aregionalsonotablyassociatedwiththenonlexicalroute(FiezandPetersen1998;Fiezetal.
1999;Fiebachetal.
2002;Jobardetal.
2003),givenitsroleinphonologicalprocessing(Paulesuetal.
1993;Fiez1997;Poldracketal.
1999;McDermottetal.
2003).
Activityinparietalregionssuggeststhatratherthanpreferentiallyrecruitingthelexicalroutetomapshort,familiarwordsaswholeunitsdirectlytotheirmeanings,ASCparticipantsperformtheindirectoperationofgrapheme-phonemeconversionwhilstreading.
Thisatypicalrecruitmentofthenonlexicalgrapheme-phonemeconversionroutewhilstreadingistheoreticallyconsistentwiththeprecocitythatsomeautisticchildrenshowtowardssoundingwordsaloud(Newmanetal.
2007),andtheaforementionedrelationshipbetweenASCandhyperlexia.
Thissuggeststhatthe'mechanical'skillsofgrapheme-phonemedecodingmayexceedthedirectmap-pingofletterstringstomeaning.
Giventhatphonologicalprocessingstrategiesplayacriticalroleinlearningtoread(Racketal.
1994),theover-relianceonthisindirectpho-nologicalroutewhichweobservehereinautismiscon-sistentwiththehyperlexiasometimesobservedinthispopulationandthefactthatreadingproblemsintheliter-atureappeartobemorerelatedtocomprehensionthantomechanicaldecodingandthelearningprocess(Venteretal.
1992;Mylesetal.
2002;Nationetal.
2006;Newmanetal.
2007).
SemanticStrooptasksrevealthatsemanticprocessesareoccurringatsomelevelinautism(Bryson1983;Eskesetal.
1990;Ozonoff1997;OzonoffandJensen1999;Russelletal.
1999),anditisclearthatautisticindividualscanreadformeaning—butourresultssuggestthat,inapassivetask,theydonotautomaticallydosoinpreferenceoverthenon-semanticroute.
Indeed,semantic-levelprocessingmaynotbethe'defaultmode'ofprocessinginASC(Ja¨rvinen-Pasleyetal.
2008)asbehaviouralandbrain-imagingstudiessuggestthattheseindividualsnaturallyfavourperceptualprocessingstrate-gies(KamioandToichi2000;ToichiandKamio2001,2003;Gaffreyetal.
2007),whichwouldappearconsistentwiththeirrecruitmentofthephonologicalpathwayinthepresentstudy.
Theconvergenceofelectrophysiologicaldatalikethiswithovertbehaviouralprocessingtasksisofcriticalimportanceforfutureresearchinordertocorrob-orateandelucidateourinterpretationofthesendings.
AutomaticSemanticsVersusaLackofCategory-SpecicityPreviousresearchhasshownearlysemanticdifferencesbetweenwordcategoriesinthetypicalpopulationthatareindependentoffocusedattention(Shtyrovetal.
2004;Pulvermu¨lleretal.
2005).
Likewise,despitenotbeingexplicitlyaskedto'readformeaning'(onlyto'attendandreadeachwordasitappears'),ourcontrolgroupshowedapatternofcategory-specicitywherebyobjectwordsevokedgreateractivityinposteriortemporo-occipitalregions(particularlyintheearlytime-windows)andactionwordsevokedgreateractivityinfrontalandmotorregionsthroughout(particularlyintherighthemisphere).
Thisisconsistentwithpreviousliterature,whichhasreportedrobustassociationsofvisualobjectwordswithposteriortemporo-occipitalregions(Warburtonetal.
1996;Pulver-mu¨lleretal.
1999;MartinandChao2001;Martin2007)andactionwordswithfrontalmotorregions(Pulvermu¨lleretal.
2001,2005,2009;Hauketal.
2004;Shtyrovetal.
2004;Tettamantietal.
2005;Aziz-ZadehandDamasio2008;Hauketal.
2008;Kemmereretal.
2008;Boulengeretal.
2009,2012).
SuchassociationsaresuggestedtoarisethroughtheformationofneuralassembliesthroughHeb-bianlearning(Pulvermu¨ller2001),wherebyobjectandaction-relatedwords,whicharegenerallylearntinthepresenceoftheirreal-worldreferent,cometoevokeactivityinthesameregionsinvolvedinexperienceswiththatconceptintheworld(e.
g.
executingtheactionorseeing/interactingwiththeobject).
Theactivationevokedbyactionwordsinprecentralmotorareaswasparticularlyrobustinourcontrolgroup,persistingbetween140and250ms.
Interestingly,thiseffect,thoughpresentinthelefthemisphere,onlyreachedsignicanceintherightprecen-tralgyrus,wheregreatestactivitywasseenforaction,followedbyobject,thenabstractwords.
GreateractivityforactionwordsinfrontalcortexforcontrolswasalsoseeninbilateralsuperiorfrontalcortexandBA44between120and140ms,thoughthesefrontaleffectswerenotaslong-lastingasthatseeninprecentralgyrus.
AscanbeseeninTable2andFig.
3,earlysemanticwordcategoryeffectswereextremelylimitedintheASC148JAutismDevDisord(2014)44:137–153123group,whichmightsupportaninterpretationconsistentwiththatgivenaboveregardingautomaticaccesstomeaning.
Wheneffectsdidappeartheywererestrictedtothefrontalcortex,unlikeinthecontrolgroup.
AtypicalrepresentationofsemanticcategoriesinASChasbeensuggestedbypreviousresearchers(Dunnetal.
1996;RapinandDunn1997),andautisticchildrenareknowntohavedifcultyextrapolatingsharedfeaturesamongcategorymemberstogenerateaprototype(KlingerandDawson2001),aprocesscriticalfortypicalcategoryformation.
Assuch,atypicalrepresentationofsemanticcategoriesisexpectedwithinthisgroupandconrmedinthepresentdata.
Withinthe120–140mstime-window,theASCgroupshowedawordcategoryeffectinsuperiorfrontalcortexthatwasdivergentinnaturetothatshownbythecontrolgroupinthesameregion:greateractivityforobjectthanactionwords.
Thesametrend,greateractivityforobjectthanactionwordsintheASCgroup,emergedagaininthe170–250mstime-windowinleftprecentralgyrus.
Thispatternofactivationiswidelydivergentfromtheactivationshowninthecontrolgroup,which,aspreviouslystated,istheoreticallyconsistentwithmodelspostulatinginvolve-mentandimportanceofmotorareasinactioncompre-hensionaswellasintheencodingofaction-relatedlanguage(Pulvermu¨ller2001;Barsalou2008;Pulvermu¨llerandFadiga2010).
InASC,thelackofcategory-specicityforactionwordsinfrontocentralcortex,andindeedtheapparentstrengthforobjectwordsinthesameregion,deviatesfromthenormandrequiresanexplanation.
WhilstgeneralabnormalitiesofsemanticstorageandprocessingmightindeedbeexpectedinASC,itispossiblethatpeoplewithASCshowparticulardeviancefromthenormintheprocessingandrepresentationofactioncon-ceptswithinfrontocentralmotorsystems.
Thecurrentstudylacksabehaviouraltestofthishypothesis,butitissug-gestedonthebasisofstructuralabnormalitiestocorticalmotorsystems(Mostofskyetal.
2006)andearlyandper-vasivemotordysfunctioninASC(Teitelbaumetal.
1998;Jansiewiczetal.
2006;Provostetal.
2007;Deweyetal.
2007;Espositoetal.
2009;Greenetal.
2009).
Diseaseordamagetomotorsystemsisassumedtodisrupttheverycircuitsimportantforactionwordprocessing(Pulvermu¨llerandFadiga2010),andhasbeenlinkedempiricallytocat-egory-specicdecitsforactionwords(NeiningerandPulvermu¨ller2001,2003;Baketal.
2001,2006;Boulengeretal.
2008;Grossmanetal.
2008;BakandChandran2011;Kemmereretal.
2012).
Whilsttheabovecouldexplaintheabsenceofthetypicalaction-wordactivationinthefrontalneocortex,itleavesopenthequestionastowhyweobservedgreateractivityforobjectwordsinthesefrontalregionsinASC.
Itisnotunusualforobjectwordstoactivatefrontalmotorsystemsinthetypicalpopulationduetotheiractionaffordances(Carotaetal.
2012).
Itmayassuchbethatsomeelementsofactionsemantics(suchasthelinkbetweenanobjectwordanditsactionaffordances)mayberelativelypreservedinASC,whilstthesemanticlinkbetweenanactionwordandthemotorsystemunder-lyingthatactionmightbeespeciallydegradedanddys-functional.
Whilstintheorythiskindofactionsemanticinformationwouldalsobejeopardisedbymotordysfunc-tion,anotherinterpretationisthatthesocialpragmaticnatureofwordstimuliwasprotectiveforobjectwordsandparticularlydetrimentalforactionwords,whichnaturallyimplyanactorandoftenrefertosocialactivities.
Socialdysfunctionisattheheartoftheautism(APA2000),andsowordsdenotingobjects,whichhavenorequirementforanysocialcontext,maystillbeencodedandprocessedinconjunctionwiththeiractionreferent.
Asthepresentdatacannotfullyconrmthishypothesis,futurestudieswillbenecessarytoinvestigateitfurther.
ConclusionsWerecordedEMEGactivityfromhigh-functioningadultswithASCandIQ-matchedcontrolswhilstreadingpas-sively.
Ourdatarevealedthat:1.
Whilsttypicalcontrolspreferentiallyrecruitthelexicaltemporalpathwayforreadingfamiliar,simplewords(asopposedtothedorsalgrapheme-phonemeconver-sionroute),participantswithASCshowreducedactivityinthispathway;2.
ParticipantswithASC,unlikecontrols,additionallyactivatethedorsalparietalprocessingroute,withnopreferentialdifferencebetweenpathways;3.
Semanticdifferencesbetweenwordstimuliaremorelimitedduringearlyprocessinginautism,andcontrastthoseseenintypicalcontrols.
ThesendingsareconsistentwithpreviousobservationswhichsuggestedthatASCparticipantsdonotutiliseoraccesssemanticinformationunlessexplicitlyinstructedtodoso.
Additionalrecruitmentoftheparietalgrapheme-phonemeconversionroutewhilstreadingisalsoconsistentwithreportsofsavantdecoding-skillsinautism.
AcknowledgmentsWethankClareCook,LucyMacGregorandOlafHaukattheMRCCognitionandBrainSciencesUnitfortheiradviceandassistanceatdifferentstagesofstimuluspreparation,MEG/EEGrecordingandanalysis;wewouldalsoliketothankCarrieAllisonandBonnieAuyeungattheAutismResearchCentrefortheirhelpwithparticipantrecruitment.
ThisworkwassupportedbytheMedicalResearchCouncil(MC_US_A060_0034,U1055.
04.
003.
00001.
01toF.
P,andMC_US_A060_0043,U.
1055.
04.
014.
00001.
01,MC_A060_5PQ90toY.
S.
).
OpenAccessThisarticleisdistributedunderthetermsoftheCreativeCommonsAttributionLicensewhichpermitsanyuse,JAutismDevDisord(2014)44:137–153149123distribution,andreproductioninanymedium,providedtheoriginalauthor(s)andthesourcearecredited.
ReferencesAmericanPsychiatricAssociation.
(2000).
Diagnosticandstatisticalmanualofmentaldisorders(DSM-IV-TR).
Washington:Amer-icanPsychiatricAssociation.
Aram,D.
M.
,&Healy,J.
M.
(1988).
Hyperlexia:Areviewofextraordinarywordrecognition.
InL.
Obler&D.
Fein(Eds.
),Theexceptionalbrain:Neuropsychologyoftalentandspecialabilities(pp.
70–102).
NewYork:Guilford.
Aram,D.
M.
,Rose,D.
F.
,&Horowitz,S.
J.
(1984).
Hyperlexia:Developmentalreadingwithoutmeaning.
InR.
M.
Joshi&A.
A.
Whitaker(Eds.
),Dyslexia:Aglobalissue(pp.
517–531).
Netherlands:MarinusNijhoff.
Aziz-Zadeh,L.
,&Damasio,A.
(2008).
Embodiedsemanticsforactions:Findingsfromfunctionalbrainimaging.
JournalofPhysiology—Paris,102,35–39.
Baayen,H.
,Piepenbrock,R.
,&vanRijn,H.
(1993).
TheCELEXlexicaldatabase(CD-Rom).
Pennsylvania:LinguisticDataConsortium.
Bak,T.
H.
,&Chandran,S.
(2011).
Whatwirestogetherdiestogether:Verbs,actionsandneurodegenerationinmotorneuronedisease.
Cortex,48(7),936–944.
Bak,T.
H.
,O'Donovan,D.
G.
,Xuereb,J.
H.
,Boniface,S.
,&Hodges,J.
R.
(2001).
SelectiveimpairmentofverbprocessingassociatedwithpathologicalchangesinBrodmannareas44and45intheMotorNeuroneDisease-Dementia-Aphasiasyndrome.
Brain,124,103–120.
Bak,T.
H.
,Yancopoulou,D.
,Nestor,P.
J.
,Xuereb,J.
H.
,Spillantini,M.
G.
,Pulvermu¨ller,F.
,etal.
(2006).
Clinical,imagingandpathologicalcorrelatesofahereditarydecitinverbandactionprocessing.
Brain,129,321–332.
Baron-Cohen,S.
,Wheelwright,S.
,Skinner,R.
,Martin,J.
,&Clubley,E.
(2001).
TheAutism-SpectrumQuotient(AQ):EvidencefromAspergerSyndrome/High-functioningautism,malesandfemales,scientistsandmathematicians.
JournalofAutismandDevelopmentalDisorders,31,5–17.
Barros-Loscertales,A.
,Gonzalez,J.
,Pulvermu¨ller,F.
,Ventura-Campos,N.
,Bustamante,J.
C.
,Costumero,V.
,etal.
(2011).
Readingsaltactivatesgustatorybrainregions:fMRIevidenceforsemanticgroundinginanovelsensorymodality.
CerebralCortex.
doi:10.
1093/cercor/bhr324.
Barsalou,L.
W.
(2008).
Groundedcognition.
AnnualReviewofPsychology,59,617–645.
Berthier,M.
L.
,Pulvermu¨ller,F.
,Green,C.
,&Higueras,C.
(2006).
ArereleasephenomenaexplainedbydisinhibitedmirrorneuroncircuitsArnoldPick'sremarksonechographiaandtheirrelevanceformoderncognitiveneuroscience.
Aphasiology,20(5),462–480.
Boulenger,V.
,Hauk,O.
,&Pulvermu¨ller,F.
(2009).
Graspingideaswiththemotorsystem:Semanticsomatotopyinidiomcompre-hension.
CerebralCortex,19,1905–1914.
Boulenger,V.
,Mechtouff,L.
,Thobois,S.
,Broussolle,E.
,Jeannerod,M.
,&Nazir,T.
A.
(2008).
WordprocessinginParkinson'sdiseaseisimpairedforactionverbsbutnotforconcretenouns.
Neuropsychologia,46,743–756.
Boulenger,V.
,Shtyrov,Y.
,&Pulvermu¨ller,F.
(2012).
WhendoyougrasptheideaMEGevidenceforinstantaneousidiomunder-standing.
Neuroimage,59,3502–3513.
Braeutigam,S.
,Swithenby,S.
J.
,&Bailey,A.
J.
(2008).
Contextualintegrationtheunusualway:Amagnetoencephalographicstudyofresponsestosemanticviolationinindividualswithautismspectrumdisorders.
EuropeanJournalofNeuroscience,27,1026–1036.
Bryson,S.
E.
(1983).
Interferenceeffectsinautisticchildren:Evidenceforthecomprehensionofsinglestimuli.
JournalofAbnormalPsychology,92,250–254.
Carota,F.
,Moseley,R.
,&Pulvermu¨ller,F.
(2012).
Body-part-specicrepresentationsofsemanticnouncategories.
JournalofCognitiveNeuroscience,24(6),1492–1509.
Cattell,R.
B.
,&Cattell,A.
K.
S.
(1960).
Testof''g'':Culturefair,scale2,formA.
Champaign:InstituteforPersonalityandAbilityTesting.
Cˇeponiene,R.
,Lepisto¨,T.
,Shestakova,A.
,Vanhala,R.
,Alju,P.
,Na¨a¨ta¨nen,R.
,etal.
(2003).
Speech-soundselectiveauditoryimpairmentinchildrenwithautism:Theycanperceivebutnotattend.
PNASUSA,100,5567–5572.
Cohen,L.
,Lehericy,S.
,Chochon,F.
,Lemer,C.
,Rivaud,S.
,&Dehaene,S.
(2002).
Language-specictuningofvisualcortexFunctionalpropertiesofthevisualwordformarea.
Brain,125,1054–1069.
Coltheart,M.
,&Rastle,K.
(1994).
Serialprocessinginreadingaloud:Evidencefordualroutemodelsofreading.
JournalofExper-imentalPsychology,6,1197–1211.
Coltheart,M.
,Rastle,K.
,Perry,C.
,Langdon,R.
,&Ziegler,J.
(2001).
DRC:Adual-routecascadedmodelofvisualwordrecognitionandreadingaloud.
PsychologicalReview,108(1),204–256.
Dawson,G.
,Meltzoff,A.
N.
,Osterling,J.
,Rinaldi,J.
,&Brown,E.
(1998).
Childrenwithautismfailtoorienttonaturallyoccurringsocialstimuli.
JournalofAutismandDevelopmentalDisorders,28,479–485.
Desikan,R.
S.
,Segonne,F.
,Fischl,B.
,Quinn,B.
T.
,Dickerson,B.
C.
,Blacker,D.
,etal.
(2006).
AnautomatedlabellingsystemforsubdividingthehumancerebralcortexonMRIscansintogyralbasedregionsofinterest.
Neuroimage,31,968–980.
Dewey,D.
,Cantell,M.
,&Crawford,S.
G.
(2007).
Motorandgesturalperformanceinchildrenwithautismspectrumdisorders,devel-opmentalcoordinationdisorder,and/orattentiondecithyper-activitydisorder.
JournaloftheInternationalNeuropsychologicalSociety,13,246–256.
Dunn,M.
,Gomes,H.
,&Sebastian,M.
J.
(1996).
Prototypicalityofresponsesofautistic,languagedisordered,andnormalchildreninaworduencytask.
ChildNeuropsychology,2(2),99–108.
Embick,D.
,Hackl,M.
,Schaeffer,J.
,Kelepir,M.
,&Marantz,A.
(2001).
Amagnetoencephalographiccomponentwhoselatencyreectslexicalfrequency.
CognitiveBrainResearch,10(3),345–348.
Eskes,G.
A.
,Bryson,S.
E.
,&McCormick,T.
A.
(1990).
Compre-hensionofconcreteandabstractwordsinautisticchildren.
JournalofAutismandDevelopmentalDisorders,20(1),61–73.
Esposito,G.
,Venuti,P.
,Maestro,S.
,&Muratori,F.
(2009).
Anexplorationofsymmetryinearlyautismspectrumdisorders:Analysisoflying.
BrainandDevelopment,31(2),131–138.
Fiebach,C.
J.
,Friederici,A.
D.
,Mu¨ller,K.
,&YvesvonCramon,D.
(2002).
fMRIevidencefordualroutestothementallexiconinvisualwordrecognition.
JournalofCognitiveNeuroscience,14(1),11–23.
Fiez,J.
A.
(1997).
Phonology,semantics,andtheroleoftheleftinferiorprefrontalcortex.
HumanBrainMapping,5,79–83.
Fiez,J.
A.
,Balota,D.
A.
,Raichle,M.
E.
,&Petersen,S.
E.
(1999).
Effectsoflexicality,frequency,andspelling-to-soundconsis-tencyonthefunctionalanatomyofreading.
Neuron,24,205–218.
Frith,U.
,&Snowling,M.
(1983).
Readingformeaningandreadingforsoundinautisticanddyslexicchildren.
JournalofDevelop-mentalPsychology,1,329–342.
Gaffrey,M.
S.
,Kleinhans,N.
M.
,Haist,F.
,Akshoomoff,N.
,Campbell,A.
,Courchesne,E.
,etal.
(2007).
Atypical150JAutismDevDisord(2014)44:137–153123participationofvisualcortexduringwordprocessinginautism:AnfMRIstudyofsemanticdecision.
Neuropsychologia,45(8),1672–1684.
Goldberg,T.
E.
(1987).
Onhermeticreadingabilities.
JournalofAutismandDevelopmentalDisorders,17,29–44.
Goldberg,T.
,&Rothermel,R.
(1984).
Hyperlexicchildrenreading.
Brain,107,759–785.
Gonzalez,J.
,Barros-Loscertales,A.
,Pulvermu¨ller,F.
,Meseguer,V.
,Sanjuan,A.
,Belloch,V.
,etal.
(2006).
Readingcinnamonactivatesolfactorybrainregions.
Neuroimage,15,906–912.
Green,D.
,Charman,T.
,Pickles,A.
,Chandler,S.
,Loucas,T.
,Simonoff,E.
,etal.
(2009).
Impairmentinmovementskillsofchildrenwithautisticspectrumdisorders.
DevelopmentalMed-icineandChildNeurology,51(4),311–316.
Grigorenko,E.
L.
,Klin,A.
,Pauls,D.
L.
,Senft,R.
,Hooper,C.
,&Volkmar,F.
(2002).
Adescriptivestudyofhyperlexiainaclinicallyreferredsampleofchildrenwithdevelopmentaldelays.
JournalofAutismandDevelopmentalDisorders,32,3–12.
Groen,W.
B.
,Zwiers,M.
P.
,vanderGaag,R.
-J.
,&Buitelaar,J.
K.
(2008).
Thephenotypeandneuralcorrelatesoflanguageinautism:Anintegrativereview.
NeuroscienceandBiobehavioralReviews,32(8),1416–1425.
Grossman,M.
,Anderson,C.
,Khan,A.
,Avants,B.
,Elman,L.
,&McCluskey,L.
(2008).
Impairedactionknowledgeinamyotro-phiclateralsclerosis.
Neurology,71(8),1396–1401.
Ha¨ma¨la¨inen,M.
S.
,&Ilmoniemi,R.
J.
(1994).
Interpretingmagneticeldofthebrain:Minimumnormestimates.
Medical&BiologicalEngineering&Computing,32(1),35–42.
Happe,F.
(1997).
Centralcoherenceandtheoryofmindinautism:Readinghomographsincontext.
BritishJournalofDevelop-mentalPsychology,15,1–12.
Harris,G.
J.
,Chabris,C.
F.
,Clark,J.
,Urban,T.
,Aharon,I.
,Steele,S.
,etal.
(2006).
Brainactivationduringsemanticprocessinginautismspectrumdisordersviafunctionalmagneticresonanceimaging.
BrainandCognition,61,54–68.
Hauk,O.
,Johnsrude,I.
,&Pulvermu¨ller,F.
(2004).
Somatotopicrepresentationofactionwordsinhumanmotorandpremotorcortex.
Neuron,41,301–307.
Hauk,O.
,Shtyrov,Y.
,&Pulvermu¨ller,F.
(2008).
Thetimecourseofactioncomprehensioninthebrainasrevealedbycorticalneurophysiology.
JournalofPhysiology—Paris,102(3),50–58.
Healy,J.
M.
,Aram,D.
M.
,Horwitz,S.
J.
,&Kessler,J.
W.
(1982).
Astudyofhyperlexia.
BrainandLanguage,17,1–23.
Hermelin,B.
,&O'Connor,N.
(1970).
Psychologicalexperimentswithautisticchildren.
Oxford:PergamonPress.
Jansiewicz,E.
M.
,Goldberg,M.
C.
,Neschaffer,C.
J.
,Denckla,M.
B.
,Landa,R.
,&Mostofsky,S.
H.
(2006).
MotorsignsdistinguishchildrenwithhighfunctioningautismandAsperger'sSyndromefromcontrols.
JournalofAutismandDevelopmentalDisorders,36(5),613–621.
Ja¨rvinen-Pasley,A.
,Wallace,G.
L.
,Ramus,F.
,Happe,F.
,&Heaton,P.
(2008).
Enhancedperceptualprocessingofspeechinautism.
DevelopmentalScience,11(1),109–121.
Jobard,G.
,Crivello,F.
,&Tzourio-Mazoyer,N.
(2003).
Evaluationofthedualroutetheoryofreading:Ametanalysisof35neuroim-agingstudies.
Neuroimage,20(2),693–712.
Kamio,Y.
,Robins,D.
,Kelley,E.
,Swainson,B.
,&Fein,D.
(2007).
Atypicallexical/semanticprocessinginhigh-functioningautismspectrumdisorderswithoutearlylanguagedelay.
JournalofAutismandDevelopmentalDisorders,37(6),1116–1122.
Kamio,Y.
,&Toichi,M.
(2000).
Dualaccesstosemanticsinautism:IspictorialaccesssuperiortoverbalaccessJournalofChildPsychologyandPsychiatry,41(7),859–867.
Kana,R.
K.
,Keller,T.
A.
,Cherkassky,V.
L.
,Minshew,N.
J.
,&Just,M.
A.
(2006).
Sentencecomprehensioninautism:Thinkinginpictureswithdecreasedfunctionalconnectivity.
Brain,129,2482–2493.
Kanner,L.
(1943).
Autisticdisturbancesofaffectivecontact.
NervousChild,2,217–250.
Kemmerer,D.
,Gonzalez-Castillo,J.
,Talavage,T.
,Patterson,S.
,&Wiley,C.
(2008).
Neuroanatomicaldistributionofvesemanticcomponentsofverbs:EvidencefromfMRI.
BrainandLan-guage,107,16–43.
Kemmerer,D.
,Rudrauf,D.
,Manzel,K.
,&Tranel,D.
(2012).
Behaviouralpatternsandlesionsitesassociatedwithimpairedprocessingoflexicalandconceptualknowledgeofaction.
Cortex,48,826–848.
Kiefer,M.
,Sim,E.
J.
,Herrnberger,B.
,Grothe,J.
,&Hoenig,K.
(2008).
Thesoundofconcepts:Fourmarkersforalinkbetweenauditoryandconceptualbrainsystems.
JournalofNeuroscience,19,12224–12230.
Klin,A.
(1991).
Youngautisticchildren'slisteningpreferencesinregardtospeech:Apossiblecharacterizationofthesymptomofsocialwithdrawal.
JournalofAutismandDevelopmentalDisor-ders,21,29–42.
Klinger,L.
G.
,&Dawson,G.
(2001).
Prototypeformationinautism.
DevelopmentandPsychopathology,13,111–124.
Kronbichler,M.
,Hutzler,F.
,Wimmer,H.
,Mair,A.
,Staffen,W.
,&Ladurner,G.
(2004).
Thevisualwordformareaandthefrequencywithwhichwordsareencountered:EvidencefromaparametricfMRIstudy.
Neuroimage,21,946–953.
Kuhl,P.
K.
,Coffey-Corina,S.
,Padden,D.
,&Dawson,G.
(2005).
Linksbetweensocialandlinguisticprocessingofspeechinpreschoolchildrenwithautism:behaviouralandelectrophysio-logicalmeasures.
DevelopmentalScience,8,1–12.
Kutas,M.
,&Federmeier,K.
D.
(2011).
Thirtyyearsandcounting:FindingmeaningintheN400componentoftheevent-relatedbrainpotential(ERP).
AnnualReviewofPsychology,62,621–647.
Lai,G.
,Pantazatos,S.
P.
,Schneider,H.
,&Hirsch,J.
(2012).
Neuralsystemsforspeechandsonginautism.
Brain,135,961–975.
Lau,E.
F.
,Phillips,C.
,&Poeppel,D.
(2008).
Acorticalnetworkforsemantics:(de)ConstructingtheN400.
NatureReviewsNeuro-science,9,920–933.
Levy,J.
,Pernet,C.
,Treserras,S.
,Boulanouar,K.
,Aubry,F.
,Demonet,J.
-F.
,etal.
(2009).
Testingforthedual-routecascadereadingmodelinthebrain:AnfMRIeffectiveconnectivityaccountofanefcientreadingstyle.
PLoSONE,4(8),1–13.
Lombardo,M.
V.
,Barnes,J.
L.
,Wheelwright,S.
J.
,&Baron-Cohen,S.
(2007).
Self-referentialcognitionandempathyinautism.
PLoSONE,2(9),e883.
Martin,A.
(2007).
Therepresentationofobjectconceptsinthebrain.
AnnualReviewofPsychology,58,25–45.
Martin,A.
,&Chao,L.
L.
(2001).
Semanticmemoryandthebrain:Structureandprocesses.
CurrentOpinioninNeurobiology,11,194–201.
McCarthy,R.
,&Warrington,E.
K.
(1986).
Phonologicalreading:Phenomenaandparadoxes.
Cortex,22,359–380.
McDermott,K.
B.
,Petersen,S.
E.
,Watson,J.
M.
,&Ojemann,J.
G.
(2003).
Aprocedureforidentifyingregionspreferentiallyactivatedbyattentiontosemanticandphonologicalrelationsusingfunctionalmagneticresonanceimaging.
Neuropsycholo-gia,41(3),293–303.
Mechelli,A.
,Corno-Tempini,M.
L.
,&Price,C.
J.
(2003).
Neuroimagingstudiesofwordandpseudowordreading:Con-sistencies,inconsistenciesandlimitations.
JournalofCognitiveNeuroscience,15,260–271.
MoscosodelPradoMartin,F.
,Hauk,O.
,&Pulvermu¨ller,F.
(2006).
Category-specicityintheprocessingofcolour-relatedandform-relatedwords:AnERPstudy.
Neuroimage,29,29–37.
JAutismDevDisord(2014)44:137–153151123Mostofsky,M.
P.
,Burgess,M.
P.
,&GidleyLarson,J.
C.
(2006).
Increasedmotorcortexwhitemattervolumepredictsmotorimpairmentinautism.
Brain,8,2117–2122.
Mottron,L.
,Morasse,K.
,&Belleville,S.
(2001).
Astudyofmemoryfunctioninginindividualswithautism.
JournalofChildPsychologyandPsychiatry,42(4),253–260.
Muller,R.
A.
(2007).
Thestudyofautismasadistributeddisorder.
MentalRetardationandDevelopmentalDisabilitiesResearchReview,13,85–95.
Myles,B.
S.
,Hilgenfeld,T.
D.
,Barnhill,G.
P.
,Griswold,D.
E.
,Hagiwara,T.
,&Simpson,R.
L.
(2002).
AnalysisofreadingskillsinindividualswithAspergerSyndrome.
FocusonAutismandOtherDevelopmentalDisabilities,17(1),44–47.
Nation,K.
,Clarke,P.
,Wright,B.
,&Williams,C.
(2006).
Patternsofreadingabilityinchildrenwithautismspectrumdisorder.
JournalofAutismandDevelopmentalDisorders,36(7),911–919.
Neininger,B.
,&Pulvermu¨ller,F.
(2001).
Therighthemisphere'sroleinactionwordprocessing:Adoublecasestudy.
Neurocase,7,303–317.
Neininger,B.
,&Pulvermu¨ller,F.
(2003).
Word-categoryspecicdecitsafterlesionsintherighthemisphere.
Neuropsychologia,41,53–70.
Newman,T.
M.
,Macomber,D.
,Naples,A.
J.
,Babitz,T.
,Volkmar,F.
,&Grigorenko,E.
L.
(2007).
Hyperlexiainchildrenwithautismspectrumdisorders.
JournalofAutismandDevelopmen-talDisorders,37,760–774.
O'Connor,N.
,&Hermelin,B.
(1994).
Twoautisticsavantreaders.
JournalofAutismandDevelopmentalDisorders,24,501–515.
Oldeld,R.
C.
(1971).
Theassessmentandanalysisofhandedness:TheEdinburghinventory.
Neuropsychologia,9,97–113.
Ozonoff,S.
(1997).
Componentsofexecutivefunctioninautismandotherdisorders.
InJ.
Russell(Ed.
),Autismasanexecutivedisorder(pp.
179–211).
NewYork:OxfordUniversityPress.
Ozonoff,S.
,&Jensen,J.
(1999).
Briefreport:Specicexecutivefunctionprolesinthreeneurodevelopmentaldisorders.
JournalofAutismandDevelopmentalDisorders,29(2),171–177.
Patti,P.
J.
,&Lupinetti,L.
(1993).
Briefreport:Implicationsofhyperlexiainanautisticsavant.
JournalofAutismandDevel-opmentalDisorders,23(2),397–405.
Paulesu,E.
,Frith,C.
D.
,&Frackowiak,R.
S.
J.
(1993).
Theneuralcorrelatesoftheverbalcomponentofworkingmemory.
Nature,362,342–345.
Poldrack,R.
A.
,Wagner,A.
D.
,Prull,M.
W.
,Desmond,J.
E.
,Glover,G.
H.
,&Gabrieli,J.
D.
(1999).
Functionalspecialisationforsemanticandphonologicalprocessingintheleftinferiorprefrontalcortex.
Neuroimage,10(1),15–35.
Provost,B.
,Lopez,B.
R.
,&Heimerl,S.
(2007).
Acomparisonofmotordelaysinyoungchildren:Autismspectrumdisorder,developmentaldelay,anddevelopmentalconcerns.
JournalofAutismandDevelopmentalDisorders,37,321–328.
Pulvermu¨ller,F.
(2001).
Brainreectionsofwordsandtheirmeaning.
TrendsinCognitiveSciences,5(12),517–524.
Pulvermu¨ller,F.
,Assadollahi,R.
,&Elbert,T.
(2001).
Neuromagneticevidenceforearlysemanticaccessinwordrecognition.
Euro-peanJournalofNeuroscience,13,201–205.
Pulvermu¨ller,F.
,&Fadiga,L.
(2010)Activeperception:Sensorimo-torcircuitsasacorticalbasisforlanguage.
NatureReviewsNeuroscience,11,351–360.
Pulvermu¨ller,F.
,Lutzenberger,W.
,&Preissl,H.
(1999).
Nounsandverbsintheintactbrain:Evidencefromevent-relatedpotentialsandhigh-frequencycorticalresponses.
CerebralCortex,9,498–508.
Pulvermu¨ller,F.
,&Shtyrov,Y.
(2006).
Languageoutsidethefocusofattention:Themismatchnegativityasatoolforstudyinghighercognitiveprocesses.
ProgressinNeurobiology,79(1),49–71.
Pulvermu¨ller,F.
,Shtyrov,Y.
,&Hauk,O.
(2009).
Understandinginaninstant:Neurophysiologicalevidenceformechanisticlan-guagecircuitsinthebrain.
BrainandLanguage,110(2),81–94.
Pulvermu¨ller,F.
,Shtyrov,Y.
,&Ilmoniemi,R.
J.
(2005).
Brainsignaturesofmeaningaccessinactionwordrecognition.
JournalofCognitiveNeuroscience,17(6),884–892.
Pylkkanen,L.
,&Marantz,A.
(2003).
TrackingthetimecourseofwordrecognitionwithMEG.
TrendsinCognitiveSciences,7(5),187–189.
Rack,J.
,Hulme,C.
,Snowling,M.
,&Wightman,J.
(1994).
Theroleofphonologyinyoungchildrenlearningtoreadwords:Thedirect-mappinghypothesis.
JournalofExperimentalChildPsychology,57(1),42–71.
Rapin,I.
(1997).
Autism.
NewEnglandJournalofMedicine,337,97–104.
Rapin,I.
,&Dunn,M.
(1997).
Languagedisordersinchildrenwithautism.
SeminarsinPediatricNeurology,4(2),86–92.
Russell,J.
,Jarrold,C.
,&Hood,B.
(1999).
Twointactexecutivecapacitiesinchildrenwithautism:Implicationsforthecoreexecutivedysfunctionsinthedisorder.
JournalofAutismandDevelopmentalDisorders,29(2),103–112.
Sahyoun,C.
P.
,Belliveau,J.
W.
,&Mody,M.
(2010).
Whitematterintegrityandpictorialreasoninginhigh-functioningchildrenwithautism.
BrainandCognition,73(3),180–188.
Saur,D.
,Kreher,B.
W.
,Schnell,S.
,Ku¨mmerer,D.
,Kellmeyer,P.
,Vry,M.
-S.
,etal.
(2008).
Ventralanddorsalpathwaysforlanguage.
PNAS,105(46),18035–18040.
Schultz,R.
T.
,Gauthier,I.
,Klin,A.
,Fulbright,R.
K.
,Anderson,A.
W.
,Volkmar,F.
,etal.
(2000).
AbnormalventraltemporalcorticalactivityduringfacediscriminationamongindividualswithautismandAspergerSyndrome.
ArchivesofGeneralPsychiatry,57,331–340.
Shtyrov,Y.
,Hauk,O.
,&Pulvermu¨ller,F.
(2004).
Distributedneuronalnetworksforencodingcategory-specicsemanticinformation:themismatchnegativitytoactionwords.
EuropeanJournalofNeuroscience,19(4),1083–1092.
Shtyrov,Y.
,Kujula,T.
,&Pulvermu¨ller,F.
(2010).
Interactionsbetweenlanguageandattentionsystems:EarlyautomaticlexicalprocessingJournalofCognitiveNeuroscience,22(7),1465–1478.
Smith,I.
M.
,&Bryson,S.
E.
(1988).
Monozygotictwinsconcordantforautismandhyperlexia.
DevelopmentalMedicineandChildNeurology,30,527–535.
Snowling,M.
,&Frith,U.
(1986).
Comprehensionin'hyperlexic'readers.
JournalofExperimentalChildPsychology,42,392–415.
Swettenham,J.
,Baron-Cohen,S.
,Charman,T.
,Cox,A.
,Baird,G.
,Drew,A.
,etal.
(1998).
Thefrequencyanddistributionofspontaneousattentionshiftsbetweensocialandnon-socialstimuliinautistic,typicallydeveloping,andnon-autisticdevel-opmentallydelayedinfants.
JournalofChildPsychologyandPsychiatry,39(5),747–753.
Teitelbaum,P.
,Teitelbaum,O.
,Nye,J.
,Fryman,J.
,&Maurer,R.
G.
(1998).
Movementanalysisininfancymaybeusefulforearlydiagnosisofautism.
PNASUSA,95,13982–13987.
Tettamanti,M.
,Buccino,G.
,Saccuman,M.
C.
,Gallese,V.
,Danna,M.
,Scifo,P.
,etal.
(2005).
Listeningtoaction-relatedsentencesactivatesfronto-parietalmotorcircuits.
JournalofCognitiveNeuroscience,17,273–281.
Toichi,M.
,&Kamio,Y.
(2001).
Verbalassociationforsimplecommonwordsinhigh-functioningautism.
JournalofAutismandDevelopmentalDisorders,31(5),483–490.
Toichi,M.
,&Kamio,Y.
(2002).
Long-termmemoryandlevelsofprocessinginautism.
Neuropsychologia,40(7),964–969.
Toichi,M.
,&Kamio,Y.
(2003).
Long-termmemoryinhigh-functioningautism:Controversyonepisodicmemoryinautism152JAutismDevDisord(2014)44:137–153123reconsidered.
JournalofAutismandDevelopmentalDisorders,33(2),151–161.
Venter,A.
,Lord,C.
,&Schopler,E.
(1992).
Afollow-upstudyofhigh-functioningautisticchildren.
JournalofChildPsychologyandPsychiatryandAlliedDisciplines,33,489–507.
Wahlberg,T.
,&Magliano,J.
P.
(2004).
Theabilityofhighfunctioningindividualswithautismtocomprehendwrittendiscourse.
DiscourseProcesses,38(1),119–144.
Walenski,M.
,Tager-Flusberg,H.
,&Ullman,M.
T.
(2006).
Languageinautism.
InS.
Moldin&J.
Rubenstein(Eds.
),Understandingautism:Frombasicneurosciencetotreatment(pp.
175–203).
NewYork:CRCPress/TaylorandFrancisGroupLLC.
Warburton,E.
,Wise,R.
J.
S.
,Price,C.
J.
,Weiller,C.
,Hadar,U.
,Ramsay,S.
,etal.
(1996).
Nounandverbretrievalbynormalsubjects:StudieswithPET.
Brain,119,159–179.
Whitehouse,D.
,&Harris,J.
C.
(1984).
Hyperlexiaininfantileautism.
JournalofAutismandDevelopmentalDisorders,14,281–289.
Zevin,J.
D.
,&Balota,D.
A.
(2000).
Primingandattentionalcontroloflexicalandsublexicalpathwaysduringnaming.
JournalofExperimentalPsychology.
Learning,Memory,andCognition,26,121–135.
JAutismDevDisord(2014)44:137–153153123