dyslexia and word reading problems
TRANSCRIPT
ToappearinK.Cain,D.Compton,&R.Parrila(Eds.),Theoriesofreadingdevelopment.JohnBenjamins
Dyslexiaandwordreadingproblems
RaunoParrila&AthanassiosProtopapas
1.Introduction
Developmentaldyslexiaisthemostcommonlearningdisabilityinchildren,
withprevalenceestimatesvaryingbetween3%and20%ofallschoolage
children(seee.g.,Shaywitz,1996;Snowling,2013).Itisacknowledgedtoaffect
childrenacrosslanguages,writingsystems,andeducationalapproachesto
readinginstruction.Developmentaldyslexiaisalsothemostwidelystudied
behaviourallydefineddevelopmentaldisorder,witharapidlyexpanding
evidencebaseonassociatedgenetics,neuralfunctioning,cognitiveskills,and
environmentalinfluences.Inthischapter,weprovideanoverviewofwidely
availablecognitivetheoriesofdevelopmentaldyslexia.Ourreviewisbyno
meansexhaustiveintermsoftheoriesincludedortheevidenceforandagainst
eachofthem—notevenabook-lengthtreatment(see,e.g.,Elliot&Grigorenko,
2014)couldachievethat.Wehope,however,thatwecoverthemaintheories
andreferencesdrivingthecognitiveresearchondyslexiaatthemoment.
Further,welimitourdiscussionofdevelopmentaldyslexiatoalphabetic
orthographies,andmainlytoEuropeanalphabeticorthographiesthathavebeen
studiedmostextensively.Wemakenoclaimsabouttheuniversalityofideas
presented(seeMcBride-Chang,thisvolume;McCardle,Miller,Lee,&Tzeng,
2011;Nag,thisvolume;Share,2008,forcross-linguisticissuesindyslexiaand
readingresearch)butacknowledgethatasthetheoreticalmodelsof
developmentaldyslexiadevelopandmovefromsingle-deficitmodelsto
multiple-deficitandhybridmodels,theirpotentialforaccommodatingspecific
featuresofdifferentwritingsystemslikelyimproves.
2
Below,wewillfirstdefinedevelopmentaldyslexiaandthenreviewwhat
wecallthesingle-causetheoriesofdyslexiathathavetraditionallydominated
thefield.Wethenadvancetomorerecentdouble-andmultiple-deficitdeficit
models,andconcludewithadiscussionconsideringtheplaceofdevelopment
andoftheindividualwithdyslexiaindevelopmentaldyslexiaresearch.
2.DevelopmentalDyslexiaDefined
Wedefinedevelopmentaldyslexiaasapersistentandunexpecteddifficulty
indevelopingage-andexperience-appropriatewordreadingskills.Forus,word
readingskillsincludebothaccuracyandefficiency,definedascorrectlyread
wordsperunitoftime.Difficultyincludesbothperformancethatfallsatthelow
endofthedistributionforanygivenageandeducationallevelaswellas
performancethatmaynotbeclinicallylowbutcanbemaintainedonlywith
extraordinaryeffort.Wetakeitforgrantedthatwordreadingabilitiesare
continuouslydistributedinapopulationandourdefinitionpotentiallyidentifies
asdyslexicsthoseindividualswhoconsistentlyfallatthelowerendofthe
distributiononsomewordreadingtest(s).Thecut-offbetweendyslexicsand
non-dyslexicsisarbitraryandsignifiesnoqualitativedifferencesbetweenthose
ondifferentsidesoftheborder.
Byfocusingonwordreadingproblems,wedistinguishbetweendyslexia
andreadingdisabilityandarguethatdyslexiaisasubsetofthelatter.According
toICD-10,forexample,specificreadingdisabilityrequiresimpairmentinreading
comprehension,wordrecognition,oralreading,orintasksthatrequirereading.
Whiledyslexiafrequentlyleadstooralreadingandreadingcomprehension
problems,wesuggestthatdyslexiaispresentwhentheprimaryreadingproblem
3
isatthelevelofwordsandtheadditionalproblemsareeithercomorbidor
secondarytothewordreadingdifficulty.
Bypersistentwemeanthatthedifficultyhastobepresentoversome
periodoftimeandnoteasilyremediedbyanalternativeinstructionalmethod.
Forexample,ifsomechildrenfailtolearntoreadwordsingrade1withone
instructionalmethodbutthenmakeclearprogresswithadifferentinstructional
method,theywouldnotqualifyasdyslexics(seee.g.,Vellutinoetal.,1996).
Instead,wewouldcallthem“teachingdisabled”(followingTunmer&Greaney,
2010).Note,however,thatpersistencedoesnotnecessarilyrequireearlyonset
(seee.g.,Catts,Compton,Tomblin&Bridges,2012;Torppa,Eklund,vanBergen
&Lyytinen,2015).
Finally,ourdefinitionincludestheelementofunexpectedness.Despite
potentialproblemsinoperationalization,thisisnecessaryfordistinguishing
dyslexiafromwordreadingdifficultiesingeneral.Unexpectednessrequiresthat
wecanestablishreasonableexpectationsnotsimplybasedonage.Onesuch
basiscouldbeorallanguagecomprehension(cf.Tunmer&Greaney,2010)but
thisisnottypicallyincludedindefinitionsofdyslexiaandmaybeproblematicin
thatlanguageskillsandwordreadingareintertwined(e.g.,Nation&Snowling,
1998;Ricketts,Nation&Bishop,2007).Instead,manywidelyadopted
definitionsofdyslexia,suchasthoseinDSM-5andinICD-10,thatincludethe
ideaofunexpectednessstatethatpoorgeneralcognitiveability,sensory
perceptionproblems,orinadequateeducationalopportunitiesmustberuledout
aspossiblecausesofpoorreadingbeforeadiagnosisofspecificlearningor
readingdisordercanbeascertained(seealsoInternationalDyslexiaAssociation,
2002).
4
Contrarytomanydefinitionsofdyslexia,wehavenoinclusioncriteria.The
mostcommoninclusioncriterionisanassociatedphonologicalprocessing
deficit.Weacknowledgethatmostindividualswithdyslexiawillexhibita
phonologicalprocessingdeficit.However,aphonologicaldeficitdoesnotseemto
beanecessaryconditionfordyslexiaandmanyindividualswithconsiderable
wordreadingproblemsdonotexhibitdepressedphonologicalawarenessscores
(e.g.,Georgiou,Parrila,Cui,&Papadopoulos,2013;Penningtonetal.,2012;
Torppaetal.,2013;vanBergen,Bishop,vanZuijen&deJong,2015).Thus,it
seemsproblematictoconfinetheuseoftheterm–andtheaccesstoadditional
resourcesandaccommodationsthatmaycomewithit–toonlythosewith
phonologicalprocessingdeficits,nomatterhowlargeaproportiontheyformof
thetotaldevelopmentaldyslexiapopulation.
3.Single-CauseTheoriesofDevelopmentalDyslexia
Atraditionalandmostcommontheoreticalapproachtodevelopmental
dyslexiahasbeentopositaspecificdeficitinsomecognitiveorperceptual
processtoaccountforwordreadingdifficulties.Thedeficitistypicallyobserved
onasignaturenonreadingtask(oranarrowsetoftasks)thatismeanttoexpose
somecrucialunderlyingweakness.Inmostcases,thepositeddeficitismeantto
accountforimpairmentsinlearningtoreadratherthanfordysfunctioninthe
cognitivemechanismofmaturereading;therefore,suchproposalsarebest
viewedasdevelopmental,ratherthanneuropsychological,accountsofreading
difficulties.IntheterminologyofCastlesandColtheart(2004),thesetheories
concerndistalcausesratherthanproximalcauses,necessitatingadditional
theoreticalsteps(andempiricaldemonstrations)tolinkthemwithobserved
readingperformance.Thefollowingdiscussionconsiderssomeimportant
5
aspectsofasubsetofinfluentialapproaches.Furtherinformationcanbefoundin
recentreviewsbyRamusandAhissar(2012)andElliotandGrigorenko(2014).
3.1.PhonologicalDeficits
Thecurrentlydominatingtheoryofdevelopmentaldyslexiapositsa
“phonologicaldeficit”atthecoreoftheproblemforalloralargemajorityof
childrenwithdifficultieslearningtoreadwords(Bishop&Snowling,2004;
Ramusetal.,2003;Vellutino,Fletcher,Snowling,&Scanlon,2004).The
phonologicalfamilyofapproachestoreadingdifficultiesisempiricallybasedon
asetoftasksknownasphonologicalawarenesstasksinwhichchildrenare
askedtosegment,blend,delete,orotherwisemanipulatephonemesinoraltasks
notdirectlyassociatedwithreading.Aspecificcausallinkbetweendeficitsin
phonologicalawarenessandwordreadingdifficultieshasbeendifficultto
demonstrateconclusivelyduetointeractionsbetweenreadingandphonological
awarenessskills,suchastheimplicationoforthographicprocessingin
phonologicalawarenesstasks(Castles&Coltheart,2004).However,thisshould
notdetractfromthefactthatpoorperformanceinphonologicalawarenesstasks
isconcurrentlyandlongitudinallyassociatedwithdyslexiaacrosslanguages
(Ziegler&Goswami,2005).
Inmostphonologicaldeficittheories,phonologicalawarenessisassumed
tobecausallyrelatedtowordreadingbecausebeingabletodeliberately
individuateandidentifyphonemesinaspokenwordisaprerequisiteto
consciouslylinkinggraphemestothosephonemes.However,additional
theoreticalstepsarerequiredtoexplainwhyreadingdifficultiespersistpastthe
initialstagesorwhyphonologicalawarenesspredictsadvancedorthographic
6
knowledgeasmeasuredbywordrecognitionandspellingtests.Moreover,
difficultyinphonemeindividuationandidentification,impedinggrapho-
phonemicdecoding,wouldbeexpectedtocausemajordifficultiesinreading
unfamiliarletterstrings,suchasnovelwordsandpseudowords.Indeed,poor
pseudoworddecodinghaslongbeenconsideredahallmarkofdyslexia.Itis,
however,increasinglyacknowledgedthatwordrecognitionisthemostseverely
affecteddomain,withlargereffectsizesbetweentypicalandpoorreadersthan
pseudoworddecoding(VandenBroeck&Geudens,2012).
Alternativeapproachestoacausallinkbetweenphonologicalawareness
tasksandreadingdevelopmentfocusonphonologicalrepresentations,assuming
thattherepresentationsofphonemesareimpaired,somehowimproperlyor
insufficientlyspecified(Perfetti,1992;Ramus,2003;Snowling,2000).Poor
phonemicrepresentationsaccountforpoorphonologicalawarenessandgrapho-
phonemicdecodingbecausephonemicrepresentationsareneededbothfor
consciousmanipulationandforefficientmappingtographemes.
However,de-emphasizingtheroleofawarenessoverlooksthefactthat
tasksinwhichpoorreadersexhibitpoorperformanceareconsistentlythosein
whichphonologicalrepresentationsmustbeexplicitlymanipulated.Thereis
littleevidencethatspeechperceptionorproductionareaffected,aswouldbe
expectedifphonemicrepresentationswereimpaired(Ramus&Ahissar,2012).
Perceptionstudieshavereportedinconsistentfindings,includingsomewhat
poorer(e.g.,Rosen&Manganari,2001),nodifferent(e.g.,Hazan,Messaoud-
Galusi,Rosen,Nouwens,&Shakespeare,2009),orenhanced(e.g.,Serniclaes,Van
Heghe,Mousty,Carré,&Sprenger-Charolles,2004)discriminationofspeech
sounds.Theonlineuptakeofacousticinformationinmatchinglexicalcandidates
7
alsoappearsnormal(Magnusonetal.,2011).Giventhesechallengestothe
representationaccount,adeficitinphonologicalaccess,ratherthan
representation,hasbeenproposed(Boetsetal.,2013;Ramus&Szenkovits,
2008,2009).Accordingtothisaccount,phonemesareproperlyspecifiedbut
theyarenotefficientlyaccessibleforoperationssuchasthoserequiredfor
explicitphonologicalawarenesstasksandformappingbetweenvisualand
phonologicalcodes.Itremainstobeseenhowthisideamightaccountforthe
observeddeficitsoncemoldedintospecifictheoreticalhypotheseswith
associatedempiricalimplications.
Thedominanceofthe“phonologicaldeficit”intheorizingaboutreading
difficultiesremainsundisputedandmostindividualswithdevelopmental
dyslexiaperformpoorlyinthesignaturetasks;atthesametime,manychildren
withphonologicaldeficitsdevelopintoadequatereaders.Othersingle-cause
approachestendtoacknowledgethesefactsandfallintotwocategories:Inthe
first,complementaryoralternativedomainsofimpairmentarepositedto
explainwordreadingproblemsofdifferenttypes.Inthesecondcategory,the
phonologicaldeficititselfisattributedtoamoregeneralorlower-level
dysfunction.
3.2.RapidNamingDeficit
Asecondmajorbranchoftheorizingisbasedon“rapidautomatized
naming”(RAN)tasks(Denckla&Rudel,1976;Norton&Wolf,2012;Wolf&
Bowers,1999;Wolf,Bowers&Biddle,2000).Inthesetasks,participantsare
shownanarrayofsymbols(letters,digits,colorpatches,orobjects)andaskedto
namethemaloudsequentiallyasfastaspossible.RANtaskshavebeendescribed
as“anearly,simplerapproximationofthereadingprocess,”including“rapid,
8
serialprocessingandintegrationofattentional,perceptual,conceptual,lexical,
andmotoricsubprocesses”(Wolfetal.,2000,p.393).Thetimetogothroughthe
entirearraydifferentiatespoorreadersfromgoodreadersandismoderatelyto
stronglycorrelatedwithwordreadingfluency,bothconcurrentlyand
longitudinally,acrosslanguagesandoveralargeagerange(seereviewinKirby,
Georgiou,Martinussen,&Parrila,2010).LowperformanceinRANtasksis
termeda“namingdeficit”andisconsideredcausaltoreadingperformanceasan
additionalcriticalfactornotsubsumedunderphonologicalprocessing.
RANtasksarecommonlythoughttoexpose“rateproblems”becauseof
theirmultiple“processingspeedrequirements”(Wolf&Bowers,1999),thereby
constitutingapreferredpredictorforreadingfluency.Statedthisway,itsounds
likeamethodvarianceissue,withatimedpredictoraccountingfortime-limited
measures,butRANalsopredictsreadingaccuracy(Kirbyetal.,2010;Parrila,
Kirby&McQuarrie,2004).Despitepronouncementsregardingtaskcomplexity,
inpracticethetheoreticalemphasishasbeenplacedonageneral“processing
speed”constructthataffectscognitivecomponentsrequiredforsymbol
processing.However,namingsinglesymbolsdisplayedindividuallyisnota
strongpredictorofreadingperformanceorreadingdifficulty(Jones,Branigan,&
Kelly,2009;Zoccolottietal.,2013;Zoccolotti,DeLuca&Spinelli,2015).The
arraypresentationformat,withmultiplestimulidisplayedsimultaneously,is
criticaltothepredictivepowerofthetask(Georgiouetal.,2013).Therefore
processesinvolvedinnamingindividualsymbolscannotaccountfortheRAN-
readingrelationship,regardlessoftheirspeedrequirements.Instead,itmustbe
theefficiencyinsequentiallynaminganarrayofsymbolsthatbringsout
9
varianceuniquelyrelatedtoreading(Gordon&Hoedemaker,inpress;
Protopapas,Altani,&Georgiou,2013).
Atthemoment,the“namingdeficit”theoryconcernsprimarilyan
assessmentissueratherthanacoherentexplanatoryapproachtoreading
development.PerformanceonRANtaskshelpsidentifychildrenwithreading
difficultythatcannotbeattributedtophonologicaldeficits.Moreover,children
withlowRANperformanceinadditiontopoorphonologicalawarenesstendto
bethepoorestreaders(Kirby,Parrila&Pfeiffer,2003;Torppaetal.,2013)who
likelybenefitleastfromtraditionalintervention(Kirbyetal.,2010).
Thewell-documentedrelationshipsbetweenRANandreadingremain
largelyunexplained(seeGeorgiou&Parrila,2012,forareview).Although
currentresearcheffortsfocusonthecrucialaspectofthetaskformat(deJong,
2011;vandenBoer,Georgiou,&deJong,2016),itisnottheoreticallynecessary
thatthedifficultiesarisedirectlyfromthecognitivecomponentsresponsiblefor
processingsequences(Zoccolottietal.,2015).Perhapsarelativelyminor
difficultyinnamingindividualsymbolsbecomesexacerbatedduetothe
relentlessrequirementsforrapidintegrationamongcognitiveprocesseswhen
goingthroughasequenceofsymbols.Accesstophonologicalrepresentationsor
visualsymbolidentificationcannotprovidetheexplanationbecausearticulation
ofthesymbolnamesisnecessaryforthecrucialindividualdifferencestoemerge
(Georgiouetal.,2013).Yet,articulationratesarenottheanswereitherbecause
silentintervalsbetweensymbols(“pausetimes”)withintheRANtaskarealso
correlatedwithreading(Georgiou,Aro,Liao,&Parrila,2015).WhileRAN
performancehasbeenactivelystudiedforyearsacrossmultiplelaboratoriesand
languages,atpresentthe“namingdeficit”approachremainsaplaceholderfor
10
thefutureidentificationofputativecognitiveandneuralprocessesunderlying
efficientwordandtextreading.
3.3.Auditoryprocessing
Thelackofdirectevidenceforimpairedphonologicalrepresentations
notwithstanding,anumberofresearchprogramshavesoughttoaccountforthe
deficientphonologicalrepresentationsbyaddressingeithergeneralauditory
processingorspeech-specificprocessingunderlyingphoneticperception.Eachof
theseapproacheshasbeenbasedonasignaturemeasure,oranarrowsetof
measures,inwhichsignificantdifferencesareoftenfoundbetweengroupswith
developmentaldyslexiaandgroupsoftypicallydevelopingreaders.Forexample,
“rapidauditoryprocessing”(Gaab,Gabrieli,Deutsch,Tallal&Temple,2007;
Tallal,1980)isassessedwitha“repetitiontest,”inwhichtwobriefstimuliare
presentedinrapidsuccessionandtheparticipantmustreporttheminthe
correctsequence;“temporalsampling”(Goswami,2011,2015)isassessedwitha
“risetimeperceptiontest,”inwhichstimulidifferinginonsetabruptnessmustbe
distinguished;an“allophonicmodeofspeechperception”(Noordenbos,Segers,
Serniclaes,&Verhoeven,2013;Serniclaesetal.,2004)isassessedwith
categoricalperceptiontasksincludingidentificationanddiscriminationof
syntheticspeechsyllables;andsoon.
Atthemoment,robustnessandinterpretationoftheinitialresultsremains
controversial(seeProtopapas,2014,andRamus&Ahissar,2012,fordiscussion
andreferences).First,thepurporteddeficitsdonotreliablyemergeinevery
study;failurestoreplicateandpartialreplicationsabound.Second,despitethe
significantgroupdifferences,whenindividualperformanceisexamineditis
invariablyfoundthatamajorityofparticipantsinthereadingimpairedgroup
11
performwithintherangeofperformanceofthetypicallydevelopinggroup.This
isunlikephonologicalawareness,lowperformanceinwhichistypically
observedforthemajorityofreadingimpairedindividuals.Moreover,besides
groupdifferencesinthetargettasks,differencesaresystematicallyobservedin
tasksthat,accordingtothetheorybeingtested,shouldnotbeaffected.Because
thetasksposecomplexcognitiverequirementsforsuccessfulperformance,it
remainsplausiblethatperformancedifferencesmaybeattributableto
perceptualorcognitiveaspectsofcarryingoutthetaskotherthanthe
hypothesizedauditoryprocessingrequirements.
Moreover,foracausalinterpretationofdifferencesinauditoryprocessing,
thehypothesizeddeficitsmustdemonstrablyprecedeandpredictmediating
deficits,whichinturnmustprecedeandpredictreadingdifficulties.Precedence
canonlybeestablishedinlongitudinalstudiesbeginningatpre-readingages
(Boetsetal.,2011),agoalnoteasytoachieveinpractice.Theexisting
comparisonstoage-matchedcontrolgroupsconfoundperformancewithall
kindsofexperienceandexpertiseassociatedwithreading.Farfromsolvingthis
problem,thealternativereading-levelmatchdesignsconfoundgroupwithage
andcanonlyrevealdevelopmentalanddistributionalaspectsofthemeasures
ratherthantheoreticallyimportantdifferencesamongindividualchildren(Van
denBroeck&Geudens,2012).Asitisbecomingclearthatphonological
representationsmaynotbeimpairedinthesenseoriginallythought,itremains
tobedeterminedwhetherandhowlowperformanceinvariouspsychophysical
tasksmaybeinvolvedintheformationofandaccesstophonological
representationsorotherwiseinlearningtoread.
12
3.4.Visualattention
Anentirelydifferentsetofalternativeapproachestoexplaining
developmentaldyslexiahavefocusedonvisual-spatialattention(Vidyasagar&
Pammer,2010).Oneproposalisbasedona“letterspan”task,inwhichasetof
fivelettersisflashedbrieflyonthescreenandtheparticipantisaskedtoreport
eitherallofthelettersorasingleletterinapositioncuedaftertheir
disappearance.Itishypothesizedthatthistaskassessesthenumberofvisual
elementsthatcanbeprocessedsimultaneously,asrequiredforefficientreading.
A“visualattentionspan”deficitispositedascomplementarytothephonological
deficitapproachoritcanexistindependentlyandaccountforreadingproblems
intheabsenceofphonologicaldeficits(Bosse,Tainturier,&Valdois,2007;
Zoubrinetzky,Bielle,&Valdois,2014).
Theuseoflettersasstimuliinthecriticaltaskadmitsalternative
interpretationsbesidesvisualattention.Forexample,uptakeofvisualletter
informationmaybelimitedduetoinsufficientreadingexperience,orinefficient
graphophonemicconnectionsmayslowdownactivationofphonologicalcodes
forthelettersthroughfeedbackloops.Diminishedeffectshavebeenreported
withstimuliotherthanlettersordigits(Ziegler,Pech-Georgel,Dufau,&Grainger,
2010),thoughsubsequentvisualcategorizationdata(Lobier,Zoubrinetzky,&
Valdois,2012)andneuroimagingdata(Lobier,Peyrin,Pichat,LeBas,&Valdois,
2014)wereunaffectedbystimulustype.Still,thecausaldirectionalityofthe
visualattentionspanremainstobeindependentlyverifiedbecausereading
practicemayconceivablyaffectvisualattentionalefficiencyandmulti-element
processingofnon-alphanumericstimuliaswell(Dehaeneetal.,2010).
13
Amorerecentproposalisbasedonaspatialcueingtask,inwhichthe
locationofasymbolonthescreenisbrieflycuedpriortoitsappearance.Italian
childrenwithdyslexiawerefoundtobenefitlessfromcorrectcueingthan
typicallydevelopingreaders(Facoettietal.,2006,2010).Inaprospective
longitudinalstudy,theperformanceofpreschoolersonthistaskwasfoundto
predicttheirreadingperformanceingrades1and2.Thishasbeeninterpreted
asevidenceforafundamentaldeficitinorientingvisualattention,termed
“sluggishvisualattention,”theorizedtounderlieletterandwordrecognition
(Franceschini,Gori,Ruffino,Pedrolli,&Facoetti,2012).However,thisproposalis
notspecificallyrelatedtoletterandwordidentificationperformance.Rather,the
linktowordrecognitionisthroughphonologicaldecoding,viaamultimodal
attentionmechanismthatmediatesefficientorthographic–phonologicalbinding
(Gori&Facoetti,2015).
3.5.Discussionofsingle-deficittheories
Mostofthesingle-deficittheoriesofdyslexiaofferlittlemorethan
observationofanassociationbetweenpoorreadingperformanceandlow
performanceinasignaturenonreadingtask(orinanarrowsetoftasks).The
mostsuccessfulamongthem,thephonologicaldeficitandthenamingspeed
deficittheories,includeevidenceforlongitudinalassociationsandfindingsthat
aremorerobustacrossstudies.However,muchmorewillbeneededbeforea
“theoryofdyslexia”canbeproclaimed.Wehighlightheretwoobstaclesevident
incurrenttheorizingaboutdyslexia,namely,(a)understandingtask
performanceand(b)constructingcausaltheoriesofreadingdevelopmentthat
involvethenecessarytheoreticalconstructstoconnectreadingwiththe
signaturetasks.
14
Signaturetaskperformanceisoftentakentoindexaspecifictarget
construct,ignoringothercognitiveandperceptualrequirementsforsuccessful
performance.Forexample,performanceonphonologicalawarenesstasksmay
betakentoindexthequalityofphonologicalrepresentationseventhough
successfultaskperformancealsorequiresaccurateperception,retentionin
short-termandworkingmemory,manipulation,andformationandexecutionof
anarticulatoryresponse.Weaknessesinanyofthesestepsorintheirintegration
willimpairtaskperformancewithoutnecessarilyinvolvingpoorqualityofthe
phonologicalrepresentations.Similarly,performanceinrapidnamingtasks,
risetimeperceptiontasks,letterreporttasks,etc.,willnecessarilyinvolvea
multitudeofperceptualandcognitiveprocessesandrepresentations,anyof
whichmightbeimplicatedinpoorperformance.Importantly,the“weaklink”
neednotbeasinglesteporprocess:perhapstwoormoreelementsmightneed
tobecompromisedbeforeperformancedecrementscanbeobserved.This
possibilitycannotbeaddressedintheabsenceofin-depthtaskanalyses.
Thevalidityofataskindexingaconstructcannotbedeterminedapriori.
Convergentanddivergentvaliditymustbedemonstratedbyreferenceto
additionaltasksthatdoordonotsharethepurportedcriticalrepresentationor
process.Foratheoreticalproposaltostandonsolidground,thecrucial
theoreticalconstructsmustbeproperlyoperationalizedbytheirempirical
indices.Thismustincludeawiderangeoftaskshypothesizedtoinvolvethe
constructinquestion.Crucially,itmustalsoexcludetasksofsimilarformand
comparabledifficultythatdonotinvolvetheconstruct.Intheabsenceofwell-
definedconstructs,theoreticalconnectionsarepositedinavacuum.
15
Second,thetheoreticalconstructsmustbeunambiguouslyimplicatedina
causaltheorythatclearlyshowshowtheyunderliepoorreadingperformance.
Admittedly,itisnotclearhowthiscanbeachievedintheabsenceofawell-
definedtheoryofreadingdevelopmentandreadingperformance.Still,it
behoovestheproponentsofspecifictheoriesofdyslexiatoexplainhowlearning
toreaddependsonthehypothesizedtheoreticalconstructsandtodemonstrate
thedependenceinproperlycontrolledlongitudinalandexperimentalstudies
thatsimultaneouslyassessconvergentanddivergentconstructvalidityandalso
includetasksassessingalternativehypothesesforthesameindividual
differences.Interventionstudiesinparticularcanbeeffectiveiftrainingina
distaldomaincanbeshowntoaffectreadingskill.However,itmustbeclearly
demonstratedthattheeffectarisesspecificallyduetothetheoretically
hypothesizedaspectoftraining.Thisisonlyachievableinthecontextofnot
simplyactivecontrolgroupsbutofawell-matchedcontroltrainingregimethat
differsonlyinthetheoreticallycriticalfeaturefromtheproposedintervention.
4.AcknowledgingHeterogeneity:SubtypesofDyslexia
Oneprincipledapproachtotheheterogeneityofwordreadingdifficulties
hasbeenthroughsubtyping.Thatis,ifasingle-deficittheorycannotaccountfor
allcasesofreadingfailure,perhapsacombinationoftheorieswill.Theremaybe
two(ormore)kindsofproblemspotentiallyimpedingwordreading
development,thereforetwo(ormore)kindsofdevelopmentaldyslexia.Asa
consequence,childrencanstillbe“dyslexic”iftheirwordreadingisbelowpar,
butassignedto“typeA,”“typeB,”ora“combinedA+B”dyslexicgroupdepending
onsomerelevantprofiling.Thecategorizationmaybebasedonmeasuresof
readingperformanceoronothercognitive,linguistic,orperceptualindices,
16
includingthesignaturetasksdiscussedabove.Wereviewherethemost
prominentexamplesofsuchsubtyping.
4.1.Phonologicalvs.SurfaceDyslexia
Perhapsthebest-establishedapproachtosubtypingisbasedona
theoreticaldistinctionoriginallydrawninneuropsychologicalEnglish-speaking
patients(Coltheart,2012).Specifically,itwasobservedthat,followingbrain
damage,certainpatientshavereadingdifficultiesthatmanifestdifferentially
withdifferenttypesofletterstrings:Somepatientshavemoredifficultywith
unfamiliarormade-upwords(pseudowords)whereasotherpatientshavemore
difficultywithfamiliarbutinconsistentwords,thatis,wordsthatare
pronounceddifferentlyfromotherwordswithsimilarspellingpatterns
(Woollams,2014).
Thedistinctionbetweenconsistentandinconsistentwordsisespecially
relevantfortheEnglishorthography,inwhichthereisagreatrangeof
graphophonemicconsistency,includingwordswithpronunciationhardly
licensedbytheirspelling,suchasyachtandrough.Anabsolutedivisionbetween
“regular”and“irregular”wordshasbeenimposedonacontinuumof
consistency,basedonthetheoreticalassumptionofcontent-independent“rules”
forgraphophonemicconversion.Anywordnotfullypronounceablebytherules
istermed“irregular”regardlessofitsrelationstootherwords.Forexample,the
wordpintisdeemedirregularforfailingtoadheretothesamepatternasmint
andhint,eventhoughtheusualmappingsholdforthreeoutofitsfourletters
andpronunciationoftheletteriinthiscontextisconsistentwiththewordpine
(Plaut,McClelland,Seidenberg,&Patterson,1996).Thedevelopmental
17
plausibilityandcross-linguisticrelevanceofthisrule-baseddistinctionremain
controversial.
Thedifferentialpatternsofimpairmentseeninpatientswithacquired
dyslexia,inconjunctionwiththetheoreticalhypothesisofabsolute
graphophonemicrules,haveledtoatwo-prongedapproachtowordreading,
includinga“nonlexical”routeassignedtoapplyingrulesanda“lexical”routeto
recognizefamiliarwords(Coltheart,Rastle,Perry,Langdon,&Ziegler,2001).
Accordingly,twoalternativeroutestofailurecanbeposited:Damagetothe
nonlexicalrouteimpedesgraphophonemicconversionwhereasdamagetothe
lexicalrouteimpedesrecognitionoffamiliarwords.Theeffectsofdamagewould
bemostobviousonpseudowordsandirregularwords,respectively,because
thesecanonlybereadbythecorrespondingroute.Incontrast,regularwords
canbereadcorrectlybyeitherroute,andthereforetheyarenotdiagnostic.
Childrenwithdifficultiesinpseudowordreadingaretermed“phonological”
dyslexicswhereasthosewithdifficultiesinirregularwordreadingaretermed
“surface”dyslexics,consistentwiththeclassificationofneuropsychological
patients(Castles&Coltheart,1993).
Althoughtheoreticallyattractiveinitssimplicity,thisproposalhasmet
withempiricaldifficultyinthatthevastmajorityofchildrenwithreading
difficultyexhibitlowperformancewithallkindsofwordsandpseudowords
(Manis,Seidenberg,Doi,McBride-Chang,&Petersen,1996),raisingconcerns
abouttheparsimonyofpositingsimultaneousimpairmentinbothroutes.The
surfacesubtypehasalsobeenelusiveininvestigationsusingreading-levelmatch
designs(Manisetal.,1999;Stanovich,Siegel,&Gottardo,1997).Asforthe
phonologicalsubtype,ithaslongbeenknownthatthepseudowordreading
18
deficitiscontingentonitemdifficulty,emergingmainlywithcomplex
pseudowordsthatdonotresemblewords(Rack,Snowling,&Olson,1992).
Recentadvancesinunderstandingthepsychometricissuesincomparing
performancedecrementsacrossdomains,suchasbetweenwordsand
pseudowords(VandenBroeck&Geudens,2012),havefurtherunderminedthe
potentialfordefiningsubtypesonthebasisofrelativeperformanceinsuchtasks.
4.2.Double-DeficitTheories
Proponentsofvisualandnamingdeficittheoriesreviewedabovehave
takenaslightlydifferentapproachtosubtyping.Lowperformanceinthe
signaturetask(letterspanorrapidnaming)isattributedtoanunderlyingdeficit
thatimpedesreadingdevelopmentindependentlyofphonologicaldeficits.
Therefore,thevisualornamingdeficitsareseenasalternativeoradditionalto
thephonologicaldeficit,supportingclassificationintofourquadrants:Children
withoutdifficulty,childrenwithaphonologicaldeficit,childrenwitha
visual/namingdeficit,andchildrenwithadoubledeficit(Bosseetal.,2007;Wolf
&Bowers,1999).Thispermitsexplanationofpoorreadingintheabsenceof
phonologicalproblems,whichremainsathornyissueforproponentsofthe
phonologicalcoreapproach.Thedoubledeficitapproachesalsoprovidean
additionaldimensionofseverity,possiblyassociatedwithpoorerresponseto
intervention,insofaraschildrenwithadoubledeficitwouldsufferfrommore
pervasiveandseveredifficulties,whichmayalsobemoredifficulttoameliorate
(Wolfetal.,2000).
Howdeficitsareassociatedwithreadingperformancevariesacrossthe
doubledeficitapproaches.IntheoriginaldoubledeficittheoryofWolfand
Bowers(1999),namingspeeddeficitsimpedethedevelopmentofefficientword
19
recognition,whereasphonologicaldeficitsaffectthedevelopmentofaccurate
decoding;combineddeficitsimpedereadingdevelopmentonbothfronts,
resultinginmoreseveredifficulties.Bosseetal.(2007),inturn,suggestedthat
visualattentiondeficitsimpairvisualwordprocessingmoregenerally,
manifestingthemselvesindecodingaswell,evenintheabsenceoffrank
phonologicaldeficits.Thus,thevisualattentionspanisproposedtoaccountfor
readingproblemspotentiallyincludingthosetypicallyattributedtoimpaired
phonologicalprocessing.Nevertheless,visualattentionspanproblemsare
expectedtoaffectsightwordreadingmoreseverelythanphonologicaldecoding.
Finally,proponentsofvisualattentiondeficitsbasedonspatialcueingtaskshave
claimedthatdeficitsinspatialattentionspecificallyaffectprocessesrelatedto
phonologicaldecoding(Ruffino,Gori,Boccardi,Molteni,&Facoetti,2014).Inthis
approach,subtypesdonotdistinguishamongpatternsofreadingperformance
but,rather,amongpatternsofcognitiveskillsthatunderliesimilardifficultiesin
wordreading.
4.3.DiscussionofSubtypingTheories
Themovefromasinglephonological-coreapproachtowardsencompassing
alternativecognitivesubstratesofwordreadingdifficultiesseemswelcomein
thecontextoftheestablishedcognitiveheterogeneityinthelowendoftheword
readingperformancespectrum.Freedomfromthetyrannyofasinglecausemay
paveapathtowardamorepervasiveacceptanceofamultitudeofpotential
routestowordreadingdifficulties.However,theoriesdefiningsubtypesonthe
basisoflowperformanceinspecificsignaturetasksaremorestronglyrelatedto
single-causetheoriesthantomultipledeficitalternativesdiscussedbelowinthat
theyseektoidentifycircumscribed,distinctcausesforspecificpatternsofword
20
readingdifficulties.Thus,theyalsoinheritproblemsassociatedwithsingle-cause
approaches,includingthosestemmingfromtheheterogeneityofreading
problems,whichmaynotfitentirelywithinanyparticularclassification.For
example,thealternativesreviewedabovewouldalreadyleadonetoexpecta
futuretriple-orquadruple-deficittheoryofdyslexia.
Moreover,subtypingproposalsraiseadditionalconcernsregardingthe
reliabilityandstabilityoftheclassification,inadditiontothevalidityofthe
constructsunderlyingtheclassification.Reliabilityisdifficulttoassess
conclusivelyduetothearbitrarinessofcutoffpointsplacedoncontinuous
performancedistributions.Astaskperformanceisinherentlynoisy,the
reliabilityofclassificationdependsonthereliabilityofthesignaturetasks.Here,
issuesoftheoreticalimportance(e.g.,intraindividualvariability)maybe
dismissedastrivialmeasurementnoise.Stabilityofsubtypinghasbeenfoundto
bemoderateforsurfacevs.phonologicaldyslexia(Manisetal.,1999)andfor
namingspeedvs.phonologicaldeficit(Steacy,Kirby,Parrila,&Compton,2014).
Althoughstudiesaddressingaspecificdistinctionmaynotgeneralizetoother
subtypingapproaches,itbehoovestheproponentsofsubtypestodemonstrate
thereliabilityandstabilityofclassificationoverlargeandrepresentative
populationsacrosslanguages.
5.Multiple-DeficitModelsofDyslexia
Theevidencereviewedabovesuggeststhatsingle-,double-ortriple-cause
theoriesofdevelopmentaldyslexiaareunlikelytoprovidesatisfactory
explanationsofdyslexiaasabehaviorallydefineddevelopmentaldisorder.There
isnowwidespreadconsensusthattheterm“dyslexia”referstothelowendofa
wordreadingdistributionratherthantoadiscretecondition.Ifthereisno
21
discreteconditionthataccountsforthegreatmajorityofchildrenwithword
readingdifficulties,thesearchforspecificcausesisgreatlychallenged.Rather,
thefocusisturningtomultifactorialdevelopmentalpathwaysthatcangiverise
tobrainsthatdifferintheirpropensityforlearningtoread(orlearningmath,or
acquiringanyotherculturalartifactourevolutionhasnotspecificallyequipped
usfor).Individualdifferencesinlanguageandreadingdevelopmentare
increasinglyattributedtoamultitudeofinteractinggenetic,neural,cognitive,
behavioral,andenvironmentalfactors,potentiallyleadingtohighorlowreading
performanceviamultipledevelopmentalpathways.
Insometheoreticalmodels,cognitivemultiplicityisassumedtoreducetoa
singlegeneticorneuralcausefordevelopmentaldyslexia,butthesecallsfor
simplicityfacesizeableempiricalchallenges.Thegeneticstudiesofdyslexia
startedwiththeexpectationofdominantinheritancecontrolledbyasinglegene
(Hallgren,1950).Specificsusceptibilitygenesweresoughtwithsingle-gene
strategies(suchasgeneticlinkage,targetedassociation,andchromosome
translocationordeletion).Instead,multiplelociwithmultiplesusceptibility
geneshavebeenidentified(Kere,2014).Molecularandbehaviorgeneticstudies
ofdyslexianowagreethatthegeneticarchitectureassociatedwithdyslexiais
complex,polygenic(twoormoregenescontributetothephenotype),and
heterogenic(thesamebehavioraloutcomecanbeassociatedwithmultiple
differentcauses;Carrion-Castillo,Franke,&Fisher,2013;Elliot&Grigorenko,
2014).Thereareexamplesoffamilieswheretheinheritancepatternis
consistentwithararemutationofasinglegene(e.g.,deKoveletal.,2004;
Nopola-Hemmietal.,2000).However,allgenesidentifiedinrarefamilialforms
ofdyslexiajointlyexplainatinyfractionofthevarianceinreadingabilitywhen
22
testedwithlargersamples.Instead,mostcasesofdyslexiaareprobablyaffected
byaverylargenumberofgenes,eachwithonlyweakeffects(Carrion-Castilloet
al.,2013),furthercomplicatedbyamultitudeofgene×geneandgene×
environmentinteractions(Bishop,2015;Jablonka&Lamb,2014).
Neurallevelexaminationshavenotfaredmuchbetterinsimplifying
dyslexiatheories.Recentmeta-analysesoffunctionalneuroimagingstudies
(Maisog,Einbinder,Flowers,Turkeltaub,&Eden,2008;Richlan,Kronbichler,&
Wimmer,2009,2011)haveidentifiedmorethanahundredfociofdifferences
betweendyslexicandnormallydevelopingreaders.Theresultsfromthemeta-
analysesweremostlyconsistentwiththetypicalneurophysiologicalaccountof
developmentaldyslexiaforadults(e.g.,Pughetal.,2000)buthighlightedthe
needforrefinementinthedevelopmentalaccount.Inparticular,Richlanetal.
(2011)foundnobrainareastypicallyassociatedwithphonologicalcodingtobe
reliablyunderactivatedacrossstudies.Theirresultsdidnotsupportthe
assumptionthattheprimaryandearlyemergingdysfunctionresidesintheleft
temporo-parietalcortexhousingthedorsalreadingsubsystem.Instead,they
suggestedthatanearlyandlimitedleftoccipito-temporaldysfunctionbecomes
extendedovertimeandisaccompaniedbyalefttemporo-parietaldysfunctionby
adulthood.
Asindividualfunctionalimagingstudiescontinuetoproducewidely
varyingresults,itmaybenecessarytoexaminehowmuchofthisvariabilityis
relatedtoheterogeneity(beyondage)inthedyslexiasamplesandhowmuchis
relatedtovariabilityinimagingandanalysismethodsandthesignaturetasks
usedindifferentlaboratories.Notably,substantialvariabilityisnotlimitedto
functionalimagingbutisalsopresentinstructuralimaging,asdemonstratedin
23
meta-analysesofvoxel-basedmorphometrystudiesofgraymatter
(Linkersdörfer,Lonnemann,Lindberg,Hasselhorn,&Fiebach,2012;Richlan,
Kronbichler,&Wimmer,2013;seereviewinJednorogetal.,2015)andwhite
matter(Vandermosten,Boets,Wouters&Ghesquière,2012)comparingdisabled
andtypicallydevelopingreaders.Asstructureandfunctionofthebrainareboth
alteredbyexperience(Gabrieli,2009;Krafnick,Flowers,Luetje,Napoliello,&
Eden,2014;Simosetal.,2002),therelevantheterogeneityinthedevelopmental
dyslexiasamplesisnotlimitedtothereadingandcognitivemeasuresusually
usedtoidentifythedyslexics.
5.1.ProbabilisticMultipleDeficitModels
Theassumptionthatbehaviorallydefineddevelopmentaldisorderscan
haveasinglecauseatanylevelofanalysishasbeenfurtherchallengedby
multifactorialetiologicalmodels(e.g.,Gottlieb&Halpern,2002;Lyytinenetal.,
1998;Pennington,2006;vanBergen,vanderLeij,&deJong,2014).Inan
influentialpaper,Pennington(2006)reviewedgenetic,neural,cognitive,and
comorbiditystudiesofdevelopmentaldyslexiaandconcludedthatconverging
evidenceprecipitatesamajorreconceptualizationoftheexistingtheoretical
models.Hearguedthatprobabilisticmultipledeficitmodels(PMDM)areneeded
toproviderealisticaccountsofdevelopmentaldisorders,theircomorbidity,and
thenondeterministicrelationshipsbetweendisordersandtheirpresumed
causes.HesuggestedfurtherthatsuchPMDMsmustincludeprotectiveandrisk
factors,multiplelevelsofanalysis,bidirectionalconnectionsbetweenconstructs
withineachlevel(horizontalorintralevelinteractions),andbidirectional
connectionsbetweenlevels(verticalorinterlevelinteractions)toaccountfor
interactionsbetweenprotectiveandriskfactorsfunctioningatdifferentlevelsof
24
analysis(seealso,Ford&Lerner,1992;Gottlieb,1983,1997;Gottlieb,Wahlsten,
&Lickliter,2006;Lyytinenetal.,1998).
Figure1showsasimplifiedversionofPennington’s(2006)PMDMwith
interlevelconnectionsomitted.Theleftsideofthemodelshowsthelevelsof
analysis–etiological,neural,cognitive,andbehavioral–andtherightside
displaysthemechanismsthatunderliehorizontalinteractionsineachlevel.
AccordingtoPennington,theetiologicallevelofanybehaviorallydefined
developmentaldisorder–includingdyslexia–ismultifactorialandinvolves
interactionofmultipleriskandprotectivefactorsthatcanbeeithergeneticor
environmental;thesejointlyandprobabilisticallyinfluencethedevelopmentof
neuralsystemsand,further,thecognitiveprocessestheysupport.Atthe
behaviorallevel,thedisorderisjointlyandprobabilisticallyproducedby
multiplecognitiveriskandprotectivefactors,eachinfluencedbymultiple
etiologicalfactors.Someoftheetiologicalriskandprotectivefactorsinfluence
severaldisorders(causingcomorbidity)whereasothersarespecifictoone
disorder.Nosingleetiological,neural,orcognitivefactorissufficient;a
combinationofseveralmaybenecessarytoproducethebehavioralsymptoms
thatdefinethedisorder.Finally,theliabilitydistributionforanygivendisorder
iscontinuous.Anindividual’spositiononthedistributionisaffectedbyriskand
protectivefactorsatanylevel.
Recently,vanBergenetal.(2014)extendedPennington’sPMDMtoallow
forintergenerationaltransferofriskandprotectivefactors.Toexplicitlyaccount
forparentaleffects,theyproposedtheintergenerationalmultipledeficitmodel
(iMDM;seeFigure2),whichincludesnotonlygenetictransmissionfromparents
tochildrenbutalsopassiveandevocativegene-environmentcorrelationsand
25
culturaltransmissionfromparents.InFigure2,environmentasshapedand
selectedbytheparentsisseparatefromextra-parentalenvironmentandfrom
geneticeffectsintheetiologicallevelofPennington’sformulation.IntheiMDM,
parentalskills(asexpressedintheirphenotype,PT)aretransmittedboth
geneticallyandviathehomeenvironment.Thisextendedhomeenvironmentcan
exertadirecteffect(culturaltransmissioninFigure2)onchildreninthat
parents’cognitivephenotypesimpacttheenvironmentstheycanofferfortheir
children,includingtheprotectiveandriskfactorsintheseenvironments.Ifhome
environmentcorrelateswithparents’andchild’sgenotype,weobservepassive
gene-environmentcorrelation(rGEinFigure2);theseinfluencescould,for
example,includehighlyliterateparentshavingahigherincomeandthusaccess
tobetterschools.EvocativerGEariseswhensomeofthechild’shighlyheritable
characteristics(e.g.,goodphonologicalawareness)elicitaresponsefromthe
environment(e.g.,morerhymingandalliterationgamesandplayingwithletters)
thatfurtherstrengthensthechild’sreadingdevelopment.Sharedenvironmental
confoundcontributestoparent-childresemblancebyaffectingreadingabilityin
bothgenerations.Forexample,accesstoeducation,particularlyforfemales,isa
significantsharedenvironmentalconfoundinmanypartsoftheworld.Another
suchconfoundcouldbeasharedhomelanguagethatisdifferentfromthe
languageofeducation.
Separatingenvironmentfromgenesasetiologicalfactorsallowedvan
Bergenetal.toalsodifferentiatebetweenenvironmentasshapedbyparentsand
extra-parentalenvironmentthatparentshavelessinfluenceover;thiscould
includereadinginstructionmethod,accesstoprintanddigitalmedia,peer
influences,legislationofspecialprovisionsandresourcesintheschoolsfor
26
studentswithdyslexia,andthevalueofliteracyinthesocietyatlarge.Van
Bergenetal.alsodistinguishbetweenactiverGE(e.g.,achildwholearnstoread
easilyismotivatedtoreadmoreandseeksoutopportunitiestodosointhe
environment)andevocativerGE,wherechildren’sgeneticallyinfluencedability
elicitsdifferentialreactionsfromtheenvironment,suchasgoodreadersbeing
givenmoredemandingmaterialstoread.Intheeducationalliterature,asimilar
differenceismadebetweentheactiveandevocativeimpactachildcanhaveon
teachers’behavior(e.g.,Nurmi,2012).
5.2.DiscussionofMultipleDeficitModels
Multipledeficitmodelsprovideaninterestingmeta-theoreticalframework
toadvancedevelopmentaldyslexiatheorizingandresearch.Thesemodelsare
examplesofdynamicordevelopmentalsystemsmodelsthathavealonghistoryin
developmentalembryologyandbiology(seee.g.,Gottlieb,2002)andhave
permeateddevelopmentalsciencesforsometime(seee.g.,Ford&Lerner,1992,
andThelen&Smith,1994,forintroductionstoearlierapproaches,andMolenaar,
Lerner&Newell,2014,fornewerformulations).Whileseveralprominent
authorshaverecentlyacknowledgedthelimitationsoftraditionalmodels(e.g.,
Catts,thisvolume;Snowling&Melby-Lervåg,2016),systemsapproachesin
generalhavehadlittletractionindyslexiaresearch(however,seeMorrison&
O’Connor,thisvolume,foranexampleofasystemsapproachtoreading
development),perhapsbecausesystemsapproachesposeformidableempirical
challengesandtheoreticalquestionstheanswerstowhichpoorlymatchour
dominantresearchtraditionsandpresuppositions..
Whiletherearemultiplefamiliesofdevelopmentalordynamicsystems
theories,thePMDMmodelsasoutlinedbyPennington(2006)andvanBergenet
27
al.(2014)bearaclosetheoreticalresemblancetodevelopmental
psychobiologicalsystemstheories(e.g.,Gottliebetal.,2006).Inthesetheories,
developmentisconceptualizedasasequentialemergenceofnewstructuraland
functionalpropertiesandcompetenciesatalllevelsofanalysisasaconsequence
ofhorizontalandverticalinteractionsamongtheparts(Gottliebetal.,2006).
Thisimpliesthatanycausalexplanationofadevelopmentaloutcome,suchas
dyslexia,mustdescribethedevelopmentalsystemthat,overtime,ledtothe
observedoutcome.Whilewecanstudycomponentsofthesystem,suchasrapid
namingtasksorindividualgenes,inrelativeisolation,individualcomponents
neitherexplainnorcause(normalorabnormal)developmentinanymeaningful
sensewithoutanaccountoftherestofthesystem;suchanaccountmustinclude
theorganismandthephysical,biological,andsocialfactors(“developmental
niche”)thatinteractwithandshapeitovertime.Asthecomponents,or
interactants,inthesedevelopmentalnetworksarethemselveslargelyproducts
ofearlierdevelopment,developmentalexplanationsrequirethatwestudytheir
interactionsoveraperiodoftime.Asaresult,weneedtorethinkbothwhat
constitutesanexplanationandwhatkindsofobservationsarerequiredto
understandthedevelopmentalpathwaystotheobservedoutcomes.
Asdevelopmentalsystemsmodels,PMDMsinherittheideathatan
explanationrequiresunderstandingthedevelopmentalsystemwithallofitsrisk
andprotectivefactors.Thus,theunitofexplanationisnotanindividualbuta
relationalcausalnetwork.Relationalcausalnetworksincludetheideathatno
singleelementorlevelinthedevelopingsystemhascausalprimacy,andthe
functionalsignificanceofanyelementondevelopmentcanonlybeunderstoodin
thecontextofthedevelopmentalsystemofwhichtheyarepart.Ateachlevelof
28
thedevelopmentalsystem,theeffectofanyelementisdependentontherestof
thesystem,makingallfactorspotentiallyinterdependentandmutually
constraining(Gottlieb,1991).
ThePMDMsreviewedaboveimplythiskindofmultidimensionalitywhere
assigningcausalprioritytoanylevelisproblematic.However,theseideasseem
difficulttoreconcilewithcurrenttheorizingthatassignscausalpriorityatthe
geneticlevelandassumesthatgeneticandenvironmentalinfluencesare
additive.Incontrast,thedevelopmentalsystemsviewleadstoconceptualizing
geneticandenvironmentaleffectsasinterdependent.Atleastinprinciple,
genetic(andotherinherited)effectsleadingtoindividualdifferencescannotbe
understoodapartfromdevelopmentoccurringinaspecificenvironmental
context.Complexhumanbehaviors,suchasreading,“areinfluencedbyhundreds
orthousandsofproteinsencodedinhundredsorthousandsofgenesofsmall
effectthatinteractwithoneanother,theenvironment,andtheepigenomein
complexways”(Charney&English,2012,p.30).Whatthenconstitutesan
explanationofdevelopmentanddevelopmentaloutcomeismuchmorecomplex
thanmostdevelopmentalmodelsindevelopmentaldyslexiaresearchcurrently
acknowledge.
Methodologically,weneedtosupplementcurrentnomotheticvariable-
centeredstudieswithidiographicstudiesandperson-centeredanalyses(see
examplesinMolenaaretal.,2014).Nomotheticvariable-centeredstudiesthat
havedrivenmostofthetheorydevelopmentdescribedaboveareinformativeof
generaltendenciesandthecomponentsthatexplanatorymodelsneedtoinclude,
buttheycannotpredicthowtheprocessofdevelopmentunfoldsovertime.We
suggestthatweneedtodevelopandtest“dynamicmechanisticexplanations”
29
(Bechtel&Abrahamsen,2005,2010)ofhowdifferentinteractantsworktogether
inproducingobservableoutcomes.Whilewecurrentlyhavefewsuchmodelsto
buildon(seeGiraud&Ramus,2013,foranexception),thesekindsofmodelsare
notuncommoninotherfieldsofinquiry(see,e.g.,Becheretal.,2014,fora
complexmodelofhoneybeecolonydynamics).Dynamicmechanistic
explanationsrequirelongitudinalandexperimentalstudies,goingbeyond
individualdifferencestoobservationsofdevelopmentalmechanisms,aswellas
computationalmodeling(e.g.,agent-basedmodels,seeRailsback&Grimm,
2011)ofthesemechanismsgroundedinempiricalobservationsandaimingto
understandtheirfunctioningwhereobservationsandexperimentationarenot
possible.Forexample,ifweconceptualizegenetic,neural,andcognitive
interactantssuggestedbyGiraudandRamus(2013)asagentsandmodeltheir
functioningandinteractionswithagent-basedmodels(seee.g.,Railsback&
Grimm,2011;Wilensky&Rand,2015;examplesat
https://ccl.northwestern.edu/netlogo/)thatfirstsimulatewhatwealready
know,wecanthenstarttopositmechanismsandconditionsunderwhichthey
operate.Aswelearnmore,themodelswillgetmorecomplexbyhavingto
includesimulationsofnewempiricalfindingsandtheywillproducenew
hypothesestoexamineempirically(seeBechtel&Abrahamsen,2010,foran
exampleofincreasingmodelcomplexity).Thescopeofthestudiesdoesnothave
tobeanymoreexpansivethanthestudieswealreadyconduct,butwhenweuse
modelsasfirstapproximationsofthedevelopmentalinteractionsbetweenthe
componentsweobserve,theexplicitnessofourtheorieswillincreasebecause
wewillhavetofocusonthemechanismsandnotonlyonassociationsamong
measures.
30
Finally,thenotionofequifinalityhasbecomeanaxiomofdevelopmental
systemstheory(seee.g.,Ford&Lerner,1992;Gottliebetal.,2006).Inthisview,
organismswithdifferentearly—or“initial”—conditionscanreachthesame
endpointandorganismswiththesameinitialconditionscantakedifferent
routesorpathwaystoreachacommonendpoint.Equifinalityisanimportant
principleinpsychologicaldevelopment,buttheconceptisseldomdiscussedin
developmentaldyslexiaresearch.However,ifPMDMmodelsareinterpretedas
developmentalsystemstheories,itfollowsthatdevelopmentisinfluencedby
manyriskandprotectivefactorsthatinteracttoproducethereadingbehavior
weusetodiagnosedyslexia.Suchcomplexprobabilisticnetworksareboundto
producesimilarobservablestateswithdifferentinteractants(e.g.,different
explanationsofwordreadingfailure;Snowling&Melby-Lervåg,2016).Examples
areboundtoproliferatewithincreasedemphasisonperson-centeredand
idiographicmethods.
6.Conclusion
Theoriesofdevelopmentaldyslexiacannotsimplybetheoriesofindividual
differencesinwordreadingdevelopment;instead,theyneedtoprogress
towardsdynamicmechanisticexplanationsofvariousdevelopmentalpathways
toinaccurateorinefficientwordreading.Theformeraregeneralandmeantto
accountprimarilyforassociationsamonggeneralconstructs,effectively
describingtheaveragesituationthatmaynotapplytoanyindividual(seee.g.,
Velicer,Babbin,&Palumbo,2014).However,ifthegoalistodevelopatheoryof
dyslexiaasaspecificconditionratherthanadiagnosticlabel,thenweneeda
muchmorespecifictheory,ofamoreappliednature.Thistheoryshouldaccount
foreachandeverychildthatdeservedlyreceivesthediagnosticlabel“dyslexia”
31
followingextensivetesting,examination,andpossiblyfailedintervention.There
isnoroomforlettingsomechildrenslip.Howeverlargeandheterogeneous,
groupsofchildrenwithdyslexiamustbefullyaccountedforbyanytheory
purportingtobeatheoryofdyslexia.Otherwiseitisnotreallyatheoryof
dyslexiabutmaybeatheoryofsomeofthedifficultiesofsomeofthechildren
whofailtolearntoreadwords.Therequirementforfulldiagnosticcoverage
seemsextremelyunlikelytobesatisfiedbyanyapproachfocusingonsingle
causesorsingle-factorcharacterizations,anditseemsverylikelytoinclude
multiplepathwaystothesamebehavioralcondition.
Theabovediscussionhasindicatedanumberofpotentialinteractantsfrom
thegeneticthroughtheenvironmentallevelthatneedtobeconsideredina
developmentalsystemstheoryofdevelopmentaldyslexia.Theseinteractants
wereidentifiedinvariable-centeredstudiesfocusingonindividualdifferences
because(a)weknowtomeasurethem,and(b)theyco-varysufficientlywiththe
dependentvariableintheexaminedsamples.Wesuspectthattheinteractants
wecurrentlyknowofas“actualdifferencemakers”(Waters,2007)arebuta
smallsubsetofthoseneededforadynamicmechanistictheoryofdevelopmental
dyslexia.Wemayhavealreadyincludedsome“potentialdifferencemakers”
(Waters,2007;seealsoGriffiths&Tabery,2013,andTabery,2014)inour
empiricalstudiesbutfailedtorecognizetheirsignificanceforadevelopmental
theoryofdyslexiabecausetheyeitherdidnotvarysufficientlytoproducethe
statisticalassociation,ortheheterogeneityoftheirexpressioninthesamples
drownedthesignal.However,thereareundoubtedlymorepotentialdifference
makersatalllevelsofanalysisthatareyettobeidentified.Hereiswheresingle-
casestudiesandexploratorycomputermodelingcancaptureextremecasesto
32
enhancetheorydevelopment.Thetrulydevelopmentalscienceofdevelopmental
dyslexiarequiresthatwe“deconstruct”thephenomenonateachlevelintoits
constituents,butalsothatwethenattempttoreconstructthedevelopingsystem
totesthypothesesaboutinteractionsbetweenlevelsandmechanismsofeffect.
Thetheorieswewantaretheonesthatnotonlyexplainwhyachildwith
dyslexiareadsdifferentlyfromanotherchildwithorwithoutdyslexia,butalso
whereinthatdevelopmentalsystemwecanintervenesuccessfully.
33
References
Becher,M.A.,Grimm,V.,Thorbek,P.,Horn,J.,Kennedy,P.J.&Osborne,J.L.
(2014).BEEHAVE:Asystemsmodelofhoneybeecolonydynamicsand
foragingtoexploremultifactorialcausesofcolonyfailure.Journalof
AppliedEcology,51,470-482.
Bechtel,W.&Abrahamsen,A.(2005).Explanation:Amechanisticalternative.
StudiesinHistoryandPhilosophyofBiologicalandBiomedicalSciences,36,
421-441.
Bechtel,W.&Abrahamsen,A.(2010).Dynamicmechanisticexplanation:
computationalmodelingofcircadianrhythmsasanexemplarfor
cognitivescience.StudiesinHistoryandPhilosophyofScience,41,321-333.
Bishop,D.V.M.(2015).Theinterfacebetweengeneticsandpsychology:Lessons
fromdevelopmentaldyslexia.ProceedingsoftheRoyalSocietyB,282:
20143139.
Bishop,D.V.M.,&Snowling,M.J.(2004).Developmentaldyslexiaandspecific
languageimpairment:Sameordifferent?PsychologicalBulletin,130,858–
886.
Boets,B.,deBeeck,H.P.O.,Vandermosten,M.,Scott,S.K.,Gillebert,C.R.,Mantini,
D.,...&Ghesquière,P.(2013).Intactbutlessaccessiblephonetic
representationsinadultswithdyslexia.Science,342,1251–1254.
Boets,B.,Vandermosten,M.,Poelmans,H.,Luts,H.,Wouters,J.,&Ghesquière,P.
(2011).Preschoolimpairmentsinauditoryprocessingandspeech
perceptionuniquelypredictfuturereadingproblems.Researchin
DevelopmentalDisabilities,32,560–570.
34
Bosse,M.L.,Tainturier,M.J.,&Valdois,S.(2007).Developmentaldyslexia:The
visualattentionspandeficithypothesis.Cognition,104,198–230.
Carrion-Castillo,A.,Franke,B.&Fisher,S.E.(2013).Moleculargeneticsof
dyslexia:Anoverview.Dyslexia,19,214-240.
Castles,A.,&Coltheart,M.(1993).Varietiesofdevelopmentaldyslexia.Cognition,
47,149–180.
Castles,A.,&Coltheart,M.(2004).Isthereacausallinkfromphonological
awarenesstosuccessinlearningtoread?Cognition,91,77–111.
Catts,H.W.,Compton,D.,Tomblin,J.B.,&Bridges,M.S.(2012).Prevalenceand
natureoflate-emergingpoorreaders.JournalofEducationalPsychology,
105,166–181.
Charney,E.&English,W.(2012).Candidategenesandpoliticalbehavior.
AmericanPoliticalScienceReview,106(1),1-34.
Coltheart,M.(2012).Dual-routetheoriesofreadingaloud.InJ.S.Adelman(Ed.),
Visualwordrecognition,volume1:Modelsandmethods,orthographyand
phonology(pp.3–27).Hove,UK:PsychologyPress.
Coltheart,M.,Rastle,K.,Perry,C.,Langdon,R.,&Ziegler,J.(2001).DRC:Adual
routecascadedmodelofvisualwordrecognitionandreadingaloud.
PsychologicalReview,108,204–256.
deKovel,C.G.,Hol,F.A.,Heister,J.G.,Willemen,J.J.,Sandkuijl,L.A.,Franke,
B.,…Padberg,G.W.(2004).Genomewidescanidentifiessusceptibility
locusfordyslexiaonXq27inanextendedDutchfamily.JournalofMedical
Genetics,41(9),652–657.
35
Dehaene,S.,Pegado,F.,Braga,L.,Ventura,P.,NunesFilho,G.,Jobert,A.,…Cohen,
L.(2010).Howlearningtoreadchangesthecorticalnetworksforvision
andlanguage.Science330,1359-1364.
deJong,P.F.(2011).Whatdiscreteandserialrapidautomatizednamingcan
revealaboutreading.ScientificStudiesofReading,15,314–337.
Denckla,M.B.,&Rudel,R.G.(1976).Rapid‘automatized’naming(R.A.N.):
Dyslexiadifferentiatedfromotherlearningdisabilities.Neuropsychologia,
14,471-479.
Elliott,J.G.,&Grigorenko,E.L.(2014).Thedyslexiadebate.CambridgeUniversity
Press.
Facoetti,A.,Trussardi,A.N.,Ruffino,M.,Lorusso,M.L.,Cattaneo,C.,Galli,R.,...&
Zorzi,M.(2010).Multisensoryspatialattentiondeficitsarepredictiveof
phonologicaldecodingskillsindevelopmentaldyslexia.Journalof
CognitiveNeuroscience,22,1011–1025.
Facoetti,A.,Zorzi,M.,Cestnick,L.,Lorusso,M.L.,Molteni,M.,Paganoni,P.,...&
Mascetti,G.G.(2006).Therelationshipbetweenvisuo-spatialattention
andnonwordreadingindevelopmentaldyslexia.Cognitive
Neuropsychology,23,841–855.
Ford,D.H.&Lerner,R.M.(1992).Developmentalsystemstheory:Anintegrative
approach.NewsburyPark,CA:Sage.
Franceschini,S.,Gori,S.,Ruffino,M.,Pedrolli,K.,&Facoetti,A.(2012).Acausal
linkbetweenvisualspatialattentionandreadingacquisition.Current
Biology,22,814–819.
Gaab,N.,Gabrieli,J.D.E.,Deutsch,G.K.,Tallal,P.,&Temple,E.(2007).Neural
correlatesofrapidauditoryprocessingaredisruptedinchildrenwith
36
developmentaldyslexiaandamelioratedwithtraining:AnfMRIstudy.
RestorativeNeurologyandNeuroscience,25,295–310.
Gabrieli,J.D.E.(2009).Dyslexia:Anewsynergybetweeneducationand
cognitiveneuroscience.Science,325,280-283.
Georgiou,G.K.,Aro,M.,Liao,C.H.,&Parrila,R.(2015).ThecontributionofRAN
pausetimeandarticulationtimetoreadingacrosslanguages:Evidence
fromamorerepresentativesampleofchildren.ScientificStudiesof
Reading,19,135–144.
Georgiou,G.&Parrila,R.(2012).Rapidnamingandreading.InL.Swanson,K.
Harris&S.Graham(eds.),HandbookofLearningDisabilities(2nded.,pp.169-
185).NewYork:Guilford.
Georgiou,G.K.,Parrila,R.,Cui,Y.,&Papadopoulos,T.C.(2013).Whyisrapid
automatizednamingrelatedtoreading?JournalofExperimentalChild
Psychology,115,218–225.
Giraud,A.-L.&Ramus,F.(2013).Neurogeneticsandauditoryprocessingin
developmentaldyslexia.CurrentOpinioninNeurobiology,23,37-42.
Gordon,P.C.,&Hoedemaker,R.S.(inpress).Effectiveschedulingoflookingand
talkingduringrapidautomatizednaming.JournalofExperimental
Psychology:HumanPerceptionandPerformance.Advanceonline
publication.http://dx.doi.org/10.1037/xhp0000171
Gori,S.,&Facoetti,A.(2015).Howthevisualaspectscanbecrucialinreading
acquisition?Theintriguingcaseofcrowdinganddevelopmentaldyslexia.
JournalofVision,15(1):8,1-20.
Goswami,U.(2011).Atemporalsamplingframeworkfordevelopmental
dyslexia.TrendsinCognitiveSciences,15,3–10.
37
Goswami,U.(2015).Sensorytheoriesofdevelopmentaldyslexia:three
challengesforresearch.NatureReviewsNeuroscience,16,43–54.
Gottlieb,G.(1983).Thepsychobiologicalapproachtodevelopmentalissues.InM.
M.Haith&J.J.Campos(Eds.),Handbookofchildpsychology(4thed.,vol.2,
pp.1-26).NewYork:JohnWiley&Sons.
Gottlieb,G.(1991).Experientialcanalizationofbehavioraldevelopment:Theory.
DevelopmentalPsychology,27,4-13.
Gottlieb,G.(1997).Synthesizingnature-nurture:Prenatalrootsofinstinctive
behavior.Mahwah,NJ:LawrenceErlbaumAssociates.
Gottlieb,G.(2002).Individualdevelopmentandevolution.Mahwah,NJ:Lawrence
ErlbaumAssociates.
Gottlieb,G.&Halpern,C.T.(2002).Arelationalviewofcausalityinnormaland
abnormaldevelopment.DevelopmentandPsychopathology,14,421-435.
Gottlieb,G.,Wahlsten,D.,&Lickliter,R.(2006).Thesignificanceofbiologyfor
humandevelopment:Adevelopmentalpsychobiologicalsystemsview.In
R.M.Lerner&W.Damon(Eds.),Handbookofchildpsychology(6ed.,Vol.
1,pp.210-257).Hoboken,NJ:JohnWiley&Sons.
Griffiths,P.E.&Tabery,J.(2013).Developmentalsystemstheory:Whatdoesit
explain,andhowdoesitexplainit?InR.M.Lerner&J.B.Benson(eds.),
Advancesinchilddevelopmentandbehavior(vol.44,pp.65-94).Waltham,
MA:AcademicPress.
Hallgren,B.(1950).Specificdyslexia(congenitalword-blindness);aclinicaland
geneticstudy.ActaPsychiatricaetNeurologicaScandinavica
Supplementum,65,1-287.
38
Hazan,V.,Messaoud-Galusi,S.,Rosen,S.,Nouwens,S.,&Shakespeare,B.(2009).
Speechperceptionabilitiesofadultswithdyslexia:isthereanyevidence
foratruedeficit?JournalofSpeech,Language,andHearingResearch,52,
1510–1529.
InternationalDyslexiaAssociation(2002).Definitionofdyslexia.RetrievedMay
15,2015,fromhttp://eida.org/definition-of-dyslexia/
Jablonka,E.&Lamb,M.J.(2014).Evolutioninfourdimensions(rev.ed.).
Cambridge,MA:TheMITPress.
Jednorog,K.,Marchewka,A.,Altarelli,I.,MonzalvoLopez,A.K.,vanErmingen-
Marbach,M.,Grande,M.,Grabowska,A.,Heim,S.&Ramus,F.(2015).How
reliablearegraymatterdisruptionsinspecificreadingdisabilityacross
multiplecountriesandLanguages?InsightsfromaLarge-scalevoxel-
basedmorphometrystudy.HumanBrainMapping,36,1741-1754.
Jones,M.W.,Branigan,H.P.,&Kelly,M.L.(2009).Dyslexicandnondyslexic
readingfluency:Rapidautomatizednamingandtheimportanceof
continuouslists.PsychonomicBulletin&Review,16,567–572.
Kere,J.(2014).Themoleculargeneticsandneurobiologyofdevelopmental
dyslexiaasmodelofacomplexphenotype.BiochemicalandBiophysical
ResearchCommunications,452,236–243.
Kirby,J.R.,Georgiou,G.K.,Martinussen,R.,&Parrila,R.(2010).Namingspeed
andreading:Frompredictiontoinstruction.ReadingResearchQuarterly,
45,341–362.
Kirby,J.R.,Parrila,R.K.&Pfeiffer,S.L.(2003).Namingspeedandphonological
awarenessaspredictorsofreadingdevelopment.JournalofEducational
Psychology,95,453-464.
39
Krafnick,A.J.,Flowers,D.L.,Luetje,M.M.,Napoliello,E.M.&Eden,G.F.(2014)
Aninvestigationintotheoriginofanatomicaldifferencesindyslexia.The
JournalofNeuroscience,34,901-908.
Linkersdörfer,J.,Lonnemann,J.,Lindberg,S.,Hasselhorn,M.,&Fiebach,C.J.
(2012).Greymatteralterationsco-localizewithfunctionalabnormalities
indevelopmentaldyslexia:AnALEmeta-analysis.PLoSONE,7(8),e43122.
doi:10.1371/journal.pone.0043122
Lobier,M.A.,Peyrin,C.,Pichat,C.,LeBas,J.F.,&Valdois,S.(2014).Visual
processingofmultipleelementsinthedyslexicbrain:Evidencefora
superiorparietaldysfunction.FrontiersinHumanNeuroscience,8,479.
Lobier,M.,Zoubrinetzky,R.,&Valdois,S.(2012).Thevisualattentionspandeficit
indyslexiaisvisualandnotverbal.Cortex,48,768–773.
Lyytinen,H.,Ahonen,T.,Aro,M.,Aro,T.,Närhi,V.,&Räsänen,P.(1998).Learning
disabilities:Aviewofdevelopmentalneuropsychology.InR.Licht,A.
Bouma,W.Slot,&W.Koops(Eds.),Childneuropsychology:Reading
disabilityandmore.Delft,NL:Eburon.
Magnuson,J.S.,Kukona,A.,Braze,D.,Johns,C.L.,VanDyke,J.A.,Tabor,W.,Mencl,
W.E.,Pugh,K.R.,&Shankweiler,D.(2011).Phonologicalinstabilityin
youngadultpoorreaders:Timecoursemeasuresandcomputational
modeling.InP.McCardle,B.Miller,J.R.Lee,&O.J.L.Tseng(Eds.).Dyslexia
acrosslanguages(pp.184–201).Baltimore,MD:PaulH.Brookes.
Manis,F.R.,Seidenberg,M.S.,Doi,L.M.,McBride-Chang,C.,&Petersen,A.
(1996).Onthebasesoftwosubtypesofdevelopmentdyslexia.Cognition,
58,157–195.
40
Manis,F.R.,Seidenberg,M.S.,Stallings,L.,Joanisse,M.,Bailey,C.,Freedman,L.,...
&Keating,P.(1999).Developmentofdyslexicsubgroups.Annalsof
Dyslexia,49,105–134.
Maisog,J.M.,Einbinder,E.R.,Flowers,D.L.,Turkeltaub,P.E.&Eden,G.F.(2008).
Ameta-analysisoffunctionalneuroimagingstudiesofdyslexia.Annalsof
theNewYorkAcademyofSciences,1145,237–259.
McCardle,P.,Miller,B.,Lee,J.R.,&Tzeng,O.J.L.(2011).Dyslexiaacross
languages:Orthographyandthebrain-gene-behaviorlink.Baltimore,MD:
PaulH.Brookes.
Molenaar,P.C.M.,Lerner,R.M.,&Newell,K.M.(eds.)(2014).Handbookof
developmentalsystemstheoryandmethodology.NewYork:Guilford.
Nation,K.,&Snowling,M.(1998).Semanticprocessingandthedevelopmentof
word-recognitionskills:Evidencefromchildrenwithreading
comprehensiondifficulties.JournalofMemoryandLanguage,39,85–101.
Noordenbos,M.W.,Segers,E.,Serniclaes,W.,&Verhoeven,L.(2013).Neural
evidenceoftheallophonicmodeofspeechperceptioninadultswith
dyslexia.ClinicalNeurophysiology,124,1151–1162.
Nopola-Hemmi,J.,Taipale,M.,Haltia,T.,Lehesjoki,A-E.,Voutilainen,A.&Kere,J.
(2000).Twotranslocationsofchromosome15qassociatedwithdyslexia.
JournalofMedicalGenetics,37,771-775.
Norton,E.S.,&Wolf,M.(2012).Rapidautomatizednaming(RAN)andreading
fluency:Implicationsforunderstandingandtreatmentofreading
disabilities.AnnualReviewofPsychology,63,427–452.
Nurmi,J.-E.(2012)Students’characteristicsandteacher–childrelationshipsin
instruction:Ameta-analysis.EducationalResearchReview,7,177–197.
41
Parrila,R.K.,Kirby,J.R.,McQuarrie,L.(2004).Articulationrate,namingspeed,
verbalshort-termmemory,andphonologicalawareness:Longitudinal
predictorsofearlyreadingdevelopment?ScientificStudiesofReading,8,
3-26.
Perfetti,C.A.(1992).Therepresentationprobleminreadingacquisition.InP.B.
Gough,L.C.Ehri,&R.Treiman(Eds.),Readingacquisition(pp.145–174).
Hillsdale,NJ:LawrenceErlbaum.
Pennington,B.F.(2006).Fromsingletomultipledeficitmodelsofdevelopmental
disorders.Cognition,101,385-413.
Pennington,B.F.,Santerre-Lemon,L.,Rosenberg,J.,McDonald,B.,Boada,R.,
Friend,A.…Olson,R.K.(2012).Individualpredictionofdyslexiabysingle
versusmultipledeficitmodels.JournalofAbnormalPsychology,121,212-
224.
Plaut,D.C.,McClelland,J.L.,Seidenberg,M.S.,&Patterson,K.(1996).
Understandingnormalandimpairedwordreading:Computational
principlesinquasi-regulardomains.PsychologicalReview,103,56–115.
Protopapas,A.(2014)Fromtemporalprocessingtodevelopmentallanguage
disorders:mindthegap.PhilosophicalTransactionsoftheRoyalSocietyB:
BiologicalSciences,369(1634),1-11.
Protopapas,A.,Altani,A.,&Georgiou,G.K.(2013).Developmentofserial
processinginreadingandrapidnaming.JournalofExperimentalChild
Psychology,116,914–929.
Pugh,K.R.,Mencl,W.E.,Jenner,A.R.,Katz,L.,Frost,S.J.,Lee,J.R.,etal.(2000).
Functionalneuroimagingstudiesofreadingandreadingdisability
(developmentaldyslexia).MentalRetardation&Developmental
42
DisabilitiesResearchReviews,6,207–213.
Rack,J.P.,Snowling,M.J.,&Olson,R.K.(1992).Thenonwordreadingdeficitin
developmentaldyslexia:Areview.ReadingResearchQuarterly,27,29–53.
Railsback,S.F.&Grimm,V.(2011).Agent-basedandindividual-basedmodeling:A
practicalintroduction.Princeton,NJ:PrincetonUniversityPress.
Ramus,F.(2003).Developmentaldyslexia:specificphonologicaldeficitor
generalsensorimotordysfunction?CurrentOpinioninNeurobiology,13,
212–218.
Ramus,F.,&Ahissar,M.(2012).Developmentaldyslexia:Thedifficultiesof
interpretingpoorperformance,andtheimportanceofnormal
performance.CognitiveNeuropsychology,29,104–122.
Ramus,F.,Rosen,S.,Dakin,S.C.,Day,B.L.,Castellote,J.M.,White,S.,&Frith,U.
(2003).Theoriesofdevelopmentaldyslexia:insightsfromamultiplecase
studyofdyslexicadults.Brain,126,841–865.
Ramus,F.,&Szenkovits,G.(2008).Whatphonologicaldeficit?TheQuarterly
JournalofExperimentalPsychology,61,129–141.
Ramus,F.,&Szenkovits,G.(2009).Understandingthenatureofthephonological
deficit.InK.Pugh&P.McCardle(Eds.),Howchildrenlearntoread(pp.
153–169).NewYork,NY:PsychologyPress.
Richlan,F.,Kronbichler,M.,&Wimmer,H.(2009).Functionalabnormalitiesin
thedyslexicbrain:Aquantitativemeta-analysisofneuroimagingstudies.
HumanBrainMapping,30,3299–3308.
Richlan,F.,Kronbichler,M.,&Wimmer,H.(2011).Meta-analyzingbrain
dysfunctionsindyslexicchildrenandadults.Neuroimage,56,1735–1742.
Richlan,F.,Kronbichler,M.,&Wimmer,H.(2013).Structuralabnormalitiesinthe
43
dyslexicbrain:Ameta-analysisofvoxel-basedmorphometrystudies.
HumanBrainMapping,34,3055–3065.
Ricketts,J.,Nation,K.,&Bishop,D.V.M.(2007).Vocabularyisimportantforsome,
butnotallreadingskills.ScientificStudiesofReading,11,235–257.
Rosen,S.,&Manganari,E.(2001).Istherearelationshipbetweenspeechand
nonspeechauditoryprocessinginchildrenwithdyslexia?Journalof
Speech,Language,andHearingResearch,44,720–736.
Ruffino,M.,Gori,S.,Boccardi,D.,Molteni,M.,&Facoetti,A.(2014).Spatialand
temporalattentionindevelopmentaldyslexia.FrontiersinHuman
Neuroscience,8,331.
Serniclaes,W.,VanHeghe,S.,Mousty,P.,Carré,R.,&Sprenger-Charolles,L.
(2004).Allophonicmodeofspeechperceptionindyslexia.Journalof
ExperimentalChildPsychology,87,336–361.
Share,D.L.(2008).Ontheanglocentricitiesofcurrentreadingresearchand
practice:Theperilsofoverrelianceonan“outlier”orthography.
PsychologicalBulletin,134,584-615.
Shaywitz,S.E.(1996).Dyslexia.ScientificAmerican,275(5),98-104.
Simos,P.G.,Fletcher,J.M.,Bergman,E.,Breier,J.I.,Foorman,B.R.,Castillo,E.M.,
Davis,R.N.,Fitzgerald,M.,&Papanicolaou,A.C.(2002).Dyslexia-specific
brainactivationprofilebecomesnormalfollowingsuccessfulremedial
training.Neurology,58,1203-1213.
Snowling,M.J.(2000).Dyslexia(2nded.).Oxford,UK:Blackwell.
Snowling,M.J.(2013).Earlyidentificationandinterventionsfordyslexia:A
contemporaryview.JournalofResearchinSpecialEducationalNeeds,13
(1),7-14.
44
Snowling,M.J.&Melby-Lervåg,M.(2016,January4).Orallanguagedeficitsin
familialdyslexia:Ameta-analysisandreview.PsychologicalBulletin.
Advanceonlinepublication.http://dx.doi.org/10.1037/bul0000037.
Stanovich,K.E.,Siegel,L.S.,&Gottardo,A.(1997).Convergingevidencefor
phonologicalandsurfacesubtypesofreadingdisability.Journalof
EducationalPsychology,89,114–127.
Steacy,L.M.,Kirby,J.R.,Parrila,R.,&Compton,D.L.(2014).Classificationof
doubledeficitgroupsacrosstime:Ananalysisofgroupstabilityfrom
kindergartentosecondgrade.ScientificStudiesofReading,18,255–273.
Tabery,J.(2014).Beyondversus:Thestruggletounderstandtheinteraction
betweennatureandnurture.Cambridge,MA:TheMITPress.
Tallal,P.(1980).Auditorytemporalperception,phonics,andreadingdisabilities
inchildren.BrainandLanguage,9,182–198.
Thelen,E.&Smith,L.B.(1994).Adynamicsystemsapproachtothedevelopment
ofcognitionandaction.Cambridge,MA:TheMITPress.
Torppa,M.,Eklund,K.,vanBergen,E.,&Lyytinen,H.(2015).Late-emergingand
resolvingdyslexia:Afollow-upstudyfromage3to14.Journalof
AbnormalChildPsychology,43,1389-1401.
Torppa,M.,Parrila,R.,Niemi,P.,Poikkeus,A.-M.,Lerkkanen,M.-K.,&Nurmi,J.-E.
(2013).ThedoubledeficithypothesisinthetransparentFinnish
orthography:alongitudinalstudyfromkindergartentograde2.Reading
andWriting:AnInterdisciplinaryJournal,26,1353-1380.
Tunmer,W.&Greaney,K.(2010).Definingdyslexia.JournalofLearning
Disabilities,43,229-243.
vanBergen,E.,Bishop,D.V.M.,vanZuijen,T.L.,&deJong,P.F.(2015).Howdoes
45
parentalreadinginfluencechildren’sreading.Astudyofcognitive
mediation.ScientificStudiesofReading,19,325-339.
vanBergen,E.,vanderLeij,A.&deJong,P.F.(2014).Theintergenerational
multipledeficitmodelandthecaseofdyslexia.FrontiersinHuman
Neuroscience,8,346.
vandenBoer,M.,Georgiou,G.K.,&deJong,P.F.(2016).Namingofshortwords
is(almost)thesameasnamingofalphanumericsymbols:Evidencefrom
twoorthographies.JournalofExperimentalChildPsychology,144,152–
165.
VandenBroeck,W.,&Geudens,A.(2012).Oldandnewwaystostudy
characteristicsofreadingdisability:Thecaseofthenonword-reading
deficit.CognitivePsychology,65,414–456.
Vandermosten,M.,Boets,B.,Wouters,J.&Ghesquière,P.(2012).Aqualitative
andquantitativereviewofdiffucsiontensorimagingstudiesinreading
anddyslexia.NeuroscienceandBiobehavioralReviews,36,1532–1552
Velicer,W.F.,Babbin,S.F.,&Palumbo,R.(2014).Idiographicapplications:Issues
ofergodicityandgeneralizability.InP.C.M.Molenaar,R.M.,Lerner&K.
M.Newell(eds.),Handbookofdevelopmentalsystemstheoryand
methodology(pp.425-441).NewYork:Guilford.
Vellutino,F.R.,Scanlon,D.M.,Sipay,E.R.,Small,S.G.,Pratt,A.,Chen,R.S.,&
Denckla,M.S.(1996).Cognitiveprofilesofdifficult-to-remediateand
readilyremediatedpoorreaders:Earlyinterventionasavehiclefor
distinguishingbetweencognitiveandexperientialdeficitsasbasiccauses
ofspecificreadingdisability.JournalofEducationalPsychology,88,601–
638.
46
Vellutino,F.R.,Fletcher,J.M.,Snowling,M.J.,&Scanlon,D.M.(2004).Specific
readingdisability(dyslexia):Whathavewelearnedinthepastfour
decades?JournalofChildPsychologyandPsychiatry,45,2–40.
Vidyasagar,T.R.,&Pammer,K.(2010).Dyslexia:adeficitinvisuo-spatial
attention,notinphonologicalprocessing.TrendsinCognitiveSciences,14,
57–63.
Waters,C.K.(2007).Causesthatmakeadifference.JournalofPhilosophy,104,
551-579.
Wilensky,U.&Rand,W.(2015).Anintroductiontoagent-basedmodeling:
Modelingnatural,socialandengineeredcomplexsystemswithNetLogo.
Cambridge,MA:TheMITPress.
Wolf,M.,&Bowers,P.G.(1999).Thedouble-deficithypothesisforthe
developmentaldyslexias.JournalofEducationalPsychology,91,415–438.
Wolf,M.,Bowers,P.G.,&Biddle,K.(2000).Naming-speedprocesses,timing,and
reading:Aconceptualreview.JournalofLearningDisabilities,33,387–
407.
Woollams,A.M.(2014).Connectionistneuropsychology:Uncoveringultimate
causesofacquireddyslexia.PhilosophicalTransactionsoftheRoyalSociety
B:BiologicalSciences,369,20120398.
Ziegler,J.C.,&Goswami,U.(2005).Readingacquisition,developmentaldyslexia,
andskilledreadingacrosslanguages:Apsycholinguisticgrainsizetheory.
PsychologicalBulletin,131,3–29.
Ziegler,J.C.,Pech-Georgel,C.,Dufau,S.,&Grainger,J.(2010).Rapidprocessingof
letters,digitsandsymbols:whatpurelyvisual-attentionaldeficitin
developmentaldyslexia?DevelopmentalScience,13,F8–F14.
47
Zoccolotti,P.,DeLuca,M.,Lami,L.,Pizzoli,C.,Pontillo,M.,&Spinelli,D.(2013).
Multiplestimuluspresentationyieldslargerdeficitsinchildrenwith
developmentaldyslexia:AstudywithreadingandRAN-typetasks.Child
Neuropsychology,19,639–647.
Zoccolotti,P.,DeLuca,M.,&Spinelli,D.(2015).Discreteversusmultipleword
displays:Are-analysisofstudiescomparingdyslexicandtypically
developingchildren.FrontiersinPsychology,6,01530.
Zoubrinetzky,R.,Bielle,F.,&Valdois,S.(2014).Newinsightsondevelopmental
dyslexiasubtypes:Heterogeneityofmixedreadingprofiles.PloSone,9,
e99337.
48
Figure1.Pennington’s(2006)ProbabilisticMultipleDeficitModel.Notethat
causalconnections(andfeedbackloops)betweenlevelsnotshowninthefigure
butacknowledgedinthetext.
49
Figure2.vanBergen,vanderLeijanddeJong’s(2014)intergenerational
multipledeficitmodel(seevanBergenetal.,2014,fordetaileddescriptionsof
components).