The Bilingual Adaptation

30 Pages·2017·0.27 MB·English

PsychologicalBulletin ©2017AmericanPsychologicalAssociation
2017,Vol.143,No.3,233–262 0033-2909/17/$12.00 http://dx.doi.org/10.1037/bul0000099
The Bilingual Adaptation: How Minds Accommodate Experience
Ellen Bialystok
YorkUniversity
Accordingtosomeestimates,morethanhalfoftheworld’spopulationismultilingualtosomeextent.
Because of the centrality of language use to human experience and the deep connections between
linguistic and nonlinguistic processing, it would not be surprising to find that there are interactions
between bilingualism and cognitive and brain processes. The present review uses the framework of
experience-dependent plasticity to evaluate the evidence for systematic modifications of brain and
cognitivesystemsthatcanbeattributedtobilingualism.Thereviewdescribesstudiesinvestigatingthe
y.
s.adl relation between bilingualism and cognition in infants and children, younger and older adults, and
herbro patients,usingbothbehavioralandneuroimagingmethods.Excludedarestudieswhoseoutcomesfocus
publisnated prergimaradriinlygoenxpleirnigeunicset-icdeapbeinlidteienstbchecaanugseesotofcthoeginritmioonr.eAplethriopuhgehraml coosnttorifbtuhteionresteoarthche dceisncturaslseqdueinstitohne
dmi review reports some relation between bilingualism and cognitive or brain outcomes, several areas of
salliedisse rdeissecarercpha,ncnieostaabrleydbisechuasvsieodraalndstcuodniessidewreitdhinyoteurnmgsaodfumltse,thloadrgoelolygicfaalilantodcsohnocweptthueasleisseufefesc.tTs.heTfhiensael
eofittobe stoecsteiotnouptroapnosaegsenadnaacfocrouthnetbnaesxetdsotenp“seixnecthuitsivfeiealdtt.ention”toexplaintherangeofresearchfindingsand
onot
n
oris Keywords:bilingualism,executivefunction,attention,cognitivedevelopment,cognitiveaging
ationand
cier
os
su
s
Aal A transformational change in cognitive neuroscience in recent Theintriguingpossibilityisthatexperiencehasthepotentialto
aldu
ologicindivi dexecpaedrieesnceh-arselacteodmeneufrroopmlastwiciidtyesparnedaditsevroidleenicneunfodrersltiafneldoinngg sSiommilearelyxpmeroidenifcyesbraarienwstreullctkunreowanndtocoigmnpitaivctebsyrasitnemansdinchougmniatinvse.
chhe brain and cognitive systems (Pascual-Leone, Amedi, Fregni, & development, especially for children. Primary among these is
yt
Psof Merabet, 2005). Although it was known for a long time that socioeconomic status, where impoverished environments affect a
anse enriching experience had positive effects on rat behavior and varietyofcognitivesystemsaswellasbrainvolumeandstructure
cu
erial learning (Hebb, 1949), the extension and application of this ca- (Noble, Houston, Kan, & Sowell, 2012). For adults, formal edu-
mn
Aso pacity to humans, particularly in adulthood, was not recognized cationhasbeenshowntoaffectbothbrainstructureandcognitive
heper until recently. Research with animals has documented the details level,especiallyintermsofslowingcognitivedeclinewithaging
ythe of these adaptive brain changes and the experiences that lead to (Kramer,Bherer,Colcombe,Dong,&Greenough,2004).Beyond
bt
yrightedolelyfor techxoearmtmicpa(rlleed,veitehnwastitirynat(KsRrooelsaberneezdtwainle.i,gs2t,i0mK1ur2el)ac.thTi,nhgBeeseennnvseitrutotd,ni&emseDhnaitasvmehaosvhneodw,g1rne9,a6ft2oe)rr scinyacnsltueamdlseiocmemunsovidicrioftyrnamibnreiannitgnal(aLenfadfpepccoets,g,nTeirntaigoianngoerin,mHehneurtmhinoalnpzsa,.r&tTichPuealsanerteaavcc,ttii2vv0iitt1iiee1ss;
ps
coed and perform better in learning tasks (Petrosini et al., 2009) than Peretz&Zatorre,2005;Rauscher,Shaw,&Ky,1993),experience
isnd rats raised in standard lab cages. Similarly, rats reared in social inspatialnavigation(Maguireetal.,2000),extendedengagement
entnte groups show more hippocampal neurogenesis and better learning inactionvideogameplaying(Bavelier&Davidson,2013;Green
mi
docucleis tihnatenrarcattisonresarbeedtwineeinsoltahteionge(nLoumeet aaln.d, 2t0h0e3)e.nMviororenmdreanmtalteiacdalltyo, &200B4a)v,ealniedr,ev2e0n03b;riKefartlrea,inWinagttienr,ju&ggSlihnegdd(Denr,ag2a0n1s0k;iRetieasle.,n2h0u0b4e)r.,
hisarti changes in DNA methylation that allow these changes to be
Ts
Thi transmitted to the offspring of the rats who had been raised in The Uniqueness of Bilingualism
stimulatingenvironments,supportingthecrucialroleofepigenetic
factorsinbrainandcognitiveoutcomes(Mychasiuketal.,2012). If experience can shape brain structure and cognitive ability,
Itisabundantlyclearthattheenvironmentplaysacrucialrolein thenbilingualismisaprimecandidateforsucheffects.Language
thebrainandmentaldevelopmentofanimals. use is the most intense, sustained, and integrative experience in
whichhumansengage.Theintensityreflectstherolethatlanguage
hasinallouractivities,notonlyforverbalcommunicationbutalso
forconceptualizingandinterpretingongoingexperience.Semantic
ThisworkwaspartiallysupportedbygrantR01HD052523fromtheNIH
networks are invoked each time an event is understood or a
NationalInstituteofChildHealthandHumanDevelopment,grantA2559
memory is formed. Language use is sustained because of all
fromtheNaturalSciencesandEngineeringResearchCouncilofCanada,
human activities, none consumes the proportion of waking (and
andgrantR21AG048431fromtheNIHNationalInstituteonAging.
Correspondence concerning this article should be addressed to Ellen perhapsnonwaking)timethatlanguagedoes.Otheractivitieswith
Bialystok,DepartmentofPsychology,YorkUniversity,4700KeeleStreet, known neuroplastic benefits, such as musical performance, can
Toronto,ON,CanadaM3J1P3.E-mail:[email protected] onlybeundertakensomefinitenumberofhoursperday;language
233
234 BIALYSTOK
use in all its forms has no limit so from the perspective of fullyunderstood.Kroll,Dussias,Bice,andPerrotti(2015)provide
dose-relatedeffects,itisunrivalled.Finally,languageuseisinte- a detailed review of these studies and consider this evidence for
grative;usinglanguageengagesmostofthebrain,includingfron- jointactivationtobeamajordiscoveryintheeffortstounderstand
tal,temporal,andparietallobes,aswellassomeposteriorregions bilingualism.Jointactivationmeansthereisconstantcompetition
(Friederici, 2011). This extensive involvement of brain centers for selection, so bilinguals must control attention to language
increases the possibility that experience-related effects found for representationsandlanguageprocessinginawaynotrequiredfor
language use have the potential to generalize beyond language monolinguals. Without such control, there would be the constant
because the experience itself involves more than just language- riskofintrusionfromthenontargetlanguage,somethingthatrarely
specificprocesses. occurs.Toachievefluentlinguisticperformance,therefore,bilin-
Supportforthepossibilitythatbilingualismcanleadtochanges guals experience greater demands on a control system than do
inbrainandcognitivesystemscomesfromevidenceforstructural monolinguals, even when language production appears to be
brain changes associated with learning a foreign language. In an equivalent.
extensive review of the literature, Li, Legault, and Litcofsky Thefirstevidenceforjointactivationcamefromsuchlinguistic
y. (2014)describethereliabledifferencesinbrainstructureforboth tasksaslexicaldecisionorpicturenaminginwhichresponsetime
shers.broadl gpreeryiodmsaottfesrecdoenndsi-tlyanagnudagwehleitaernminagt.teTrhienitrergervitiyewafcteornseidveernsbsuricehf fnoorntbairlginegtulaalnsgwuaagsein(fel.uge.n,cGedrabinygethr,e1in9c9l3u)s.ioHnoowfecvoegr,namteosretodtrhae-
publinated factors as the timing of second-language acquisition, the interac- matic evidence was reported from studies that incorporated mea-
dmi tion between the languages, and the typological relation between sures of brain response. For example, in a study using electroen-
salliedisse tdheenclaengfruoamgesa. F9o-mroenxtahmplolen,giStuchdliengaellsetutdayl.o(2f0f1o2r)eirgenpolratnedgueavgie- cbeilpinhgaluoaglsrapwhhyo(EwEeGre)bstyudTehnietsrryatanandWEnug(l2is0h0-7sp),eCakhiinngeseu–nEivnegrlsiisthy
ite
ofob learningthatfoundchangesinwhitematterintegrityinlefthemi- andimmersedinanEnglishcontext,wereaskedtodecidewhether
oneott sphere language areas and their right hemisphere analogs as well or not pairs of English words presented either in print or orally
orsn as in the frontal lobe over the course of the study. Other studies weresemanticallyrelated.Theexperimentalmanipulationwasthat
i
ationand hetaavlerefoguionnd,iancprreiamseadrygcreeyntmerafttoerrldaenngsuiatygeinptrhoecelsesfitnign,fewriaosrppoasrii-- fuonrrehlaaltfedofcothnediwtioonrds)ptahiersC(ehqinueaslelytrdainstsrliabtuiotendoifntthheewreolardtesdcaonnd-
cier
sous tively correlated to the participant’s degree of bilingualism tainedarepetitionofoneoftheChinesecharacters.Suchrepetition
s
Aal (Mechelli et al., 2004) and increased with the time spent in a wouldonlybeapparentiftheEnglishwordsweretranslatedinto
ologicalindividu 250-m12o)n.thTofogreetihgenrltahnegsueasgteudcileasssdienmaolnosntgraittuedtihnaatlsthtuedeyx(pSetreiiennecteaol.f, Cthheindeesmeaannddsothfetnhewtaristkte.nRedaocwtino,natcimtiveistiaensdthaactcuwraecreywweelrlebseimyoilnadr
he learningasecondlanguageleavesstructuraltracesinthebrainin for all conditions, but for both oral and written presentation,
ch
yt
Psof thoseregionsresponsibleforlanguageacquisitionanduse. English word pairs whose translations contained an overlapping
ne Itisnotsurprisingthatbilingualismchangesthelanguagerep- Chinese character produced a significant reduction in the N400
as
cu
erial resentationsandbrainstructurethatunderlielanguageacquisition componentoftheevent-relatedpotential(ERP),indicatingrepeti-
Amson or processing. However, the present argument goes beyond these tion priming (Kutas & Federmeier, 2011). Chinese monolinguals
eer linguisticconsequencesandsuggeststhattheexperienceofbilin- performingthetaskinChineseshowedthesamepatternofpriming
hp
ythe guallanguageuseleadstomodificationsnotonlyinbrainstructure but English monolinguals performing the task in English under-
bt
yrightedolelyfor armengadniofcunens.cTtaihnoudnsa,pstrshooecceiasarstgeeudsmwieninvthtollivasnetdghuaaitngbeinlpoinrnogvcueearslbsiasinmlgc,iobsguantniatilevsxoepienpretihrefnoocsree- sEetdnagngedliasobhflyCbihlsiinhngoeusweae,lsdevwneeonretehufofneuccgothnosCcnhioitnuheselsyeNiwn4fa0lsu0e.irnrTceehleduvsba,nytthttheoeitCrhhiksinntoeawssekl--.
ps
coed thathasthepotentialtomodifybrainandcognitivesystemsmore Morford and colleagues (Morford et al., 2011) adapted this para-
isnd generally,muchasenrichedcagesdoforratsandsocioeconomic digmtoparticipantswhowerebilingualinEnglishandAmerican
entnte statusdoesforyoungchildren. SignLanguage(ASL)byrepeatingahandshapeinhalftheword
mi
hisdocuarticleis Language Processing in Bilinguals pshaaiFrpsue.r,tThehevreenerevtshiduoelutnsgchaegitfaowinrajisoniidrnirtcealaetcevtdiavnpatrttiioomntihnoegfsfterhmoemannotthincetajruregdpegetmaltaeenndgthutaaasngkde.
Ts
hi Centraltotheargumentforwhybilingualismhasthepotentialto comes from an eye-tracking study by Marian and Spivey (2003).
T
achieve these outcomes is an explanation of how bilingual lan- Russian-EnglishbilingualscompletedasimpletaskinEnglishfor
guage use is different from monolingual language use. The key which there was no reason to invoke Russian. Participants saw a
pointcomesfromoverwhelmingevidencethatbothlanguagesina displaywiththreepicturedobjectsandwereaskedtomakeaneye
bilingual’srepertoirearealwaysactivetosomeextent,evenifone movement to the one that matched a spoken word. There were a
ofthemisnotrequiredforthecurrentcontext.Insomesense,this variety of conditions but the crucial feature is that each stimulus
point was always known, at least intuitively: Weinreich (1953) displayincludedthreetypesofobjects:atargetobject(“marker”),
talked about the “interference” between a bilingual’s two lan- anobjectwhosenamewasphonologicallysimilartothetargetin
guages.Empiricalevidencecamemuchlater,withresearchshow- English (“marble”), and an object whose Russian translation was
ing cross-language facilitation and interference in simple lexical phonologically similar to the target even though there was no
tasks (e.g., Costa, Caramazza, & Sebastian-Galles, 2000; Costa, sharedmeaning(stamp,whichis“marka”inRussian).Again,this
Santesteban,&Ivanova,2006;Dijkstra,Grainger,&vanHeuven, cross-languageinformationwaswhollyirrelevanttoperformance,
1999; Hermans, Bongaerts, de Bot, & Schreuder, 1998), but the butRussian-Englishbilingualsmadeinitialeyemovementstothe
significance of the insight or of the empirical findings were not cross-languagedistractor.
BILINGUALADAPTATION 235
Mechanism for Bilingual Effects model provides an excellent framework for understanding these
questions and makes detailed predictions about the cognitive and
Jointactivationrequiresthatthereisamechanismforlanguage
brainchangesthatshouldfollowfromeachofthecontexts.Atthis
selection to assure that use of the target language proceeds flu-
point,however,researchsupportingtheclaimsfromthismodelis
ently. The assumptions are that this mechanism is part of a
extremelypreliminary,apointtheauthorsfullyacknowledge,soa
domain-general process and that the constant engagement of this
thoroughevaluationmustawaitfurtherstudy.TheAdaptiveCon-
process for language selection fortifies it for other purposes, in-
trol Hypothesis is discussed again briefly in Determining Bilin-
cludingnonverbalones(Bialystok,Craik,Green,&Gollan,2009).
gualism.
Support for the notion that a domain-general system is recruited
Nosingleexplanationofthecontrolmechanismsresponsiblefor
forlanguagecontrolcomesfromneuroimagingevidenceshowing
bilingual language processing and hence cognitive effects of bi-
overlap in brain networks involved in language selection and
lingualism has emerged as decisive. The position that will be
nonverbal task switching (Abutalebi & Green, 2007; De Baene,
advancedinthepresentreviewisthatultimatelythebestaccount
Duyck,Brass,&Carreiras,2015;Luk,GreenAbutalebi,&Grady,
willbeonethatiscenteredintheattentionsystem(seediscussion
y. 2012). The nature of that domain-general system, however, re- in Finding the Pattern: Executive Attention). What is clear at
shers.broadl maTinhseafimrstatsteprecoufladteiobnatse.on how this selection system might op- tahcitsivatitmioenoisftthheattwthoelabnigluinaggueaslanmdinsdoriesqucihraersascetelericzteiodnbtoyajvooiindt
dpubliminated eprraotpeowseedrethbeasinedhiobnitothryeocroentitcroallw(IoCr)kmboydGelrefeonrl(a1n9g9u8a)gienswelheiccthiohne. imntuesrtfeardeanpctetforothmisamneunntwalanctoendfilgaunrgautiaogne.thTahtuisn,ctlhuedebsilijnoginutalylmacintid-
salliedisse Idnowhinscmueosdetlh,aat lseudpetorvtihseoriynhaitbteitnitoinonofsythsteemnowntaasrggeutildaendgubaygetospo- vreactreuditeladngtouamgeasnafgoer actotmenmtiounnictoatitohne ttowoprloacnegeuda.gIefstihsepnaerttwoofrka
ofitobe thatlanguageprocessingcouldproceedfromthecontextuallyand nonverbalattentionorselectionsystem,thentheconfigurationisin
oneott linguisticallyappropriaterepresentations.Theimplicationwasthat placeforbilingualismtoimpactthenatureorqualityofnonverbal
orsn theseinhibitoryprocessesweremodifiedbytheiruseinlanguage cognitiveprocessingandpresumablythebrainregionsonwhichit
ationandi saellaercgtieonlitaenradtuarfefeicntevdesitnighaibtiintogrythceoinmtrpollicinatoiothnesrodfoGmraeienns.’sT(h1e9re98is) itshebamsaeidn. Dcheatlelremngineinfogrtfhuetudreetarielsseoafrcthh.at selection system will be
ociser modelforbilinguallanguageprocessingthatfocusesontheroleof
su
Asal inhibition.Adiscussionofthosestudiesisbeyondthescopeofthe
aldu present review (because of its focus on cognitive outcomes of Evidence for Bilingual Effects on Cognition
hologiceindivi bliitleirnagtuuraelis(mKr)o,lblu&tthGeorellaanre,2s0ev1e4r;aKlcroolml,pGreuhlelinfseirv,eMrecvCileawins,oRfotshsait, amTohuenrteanhdasdibveeernsitaynoefnroesrmeaorcuhs iinnvcerestaisgeatiinngrethceenqtueysetaiorsnionftthhee
ch
Psyoft &CruzMartin,2015). possibleimpactofbilingualismformindandbrain.Thesestudies
ne An account based on inhibition of the nontarget language be- have examined individuals at all stages of the life span and have
as
ericalu came the dominant explanation for bilingual effects on cognition usedavarietyofbehavioralandimagingmethods.Becauseofthe
Amson (e.g., Bialystok et al., 2009). The account was appealing both rangeoftheresearchandtounderstandthelifespantrajectoryof
eer becauseitwasaplausibleexplanationoftheefficiencywithwhich these effects, the evidence will be presented separately by age
hp
ythe bilingualsavoidedlinguisticinterferenceandbecauseitwascom- group. For each of children (Research With Infants and Chil-
yrightedbolelyfort pwdaiavtsiebrblseeitiynwgimtdhoedaveehllioogpfhelMdyaiiynroaflukuneednatntihadelcsmoalomldeeaegltiuomefsee,(xnMeacmiuyetailkvyee,tefhutenacul.tn,ioi2tny00tah0na)dt. dtfiirersentntsf)o,(lEalodvwuidletedsncb(EeyfFsfetrucodtmsiesoPftahBtaietilniinntscg)luubadeleihsanmveiuoirnroailAmrdeasguuilnltthgsomaordee)a,dsueasrnecdsri.bpead-
ps
coed Theirmodelproposedthreesubcomponentsofexecutivefunction,
entisntend Fonuencotfiownh).icThhweaasssiunmhipbtiitoionnth(saeteinIhnihbiibtiiotnionisaantdleathstepEaxrtecouftitvhee Research With Infants and Children
mi
docucleis mfraemcheanthisims refsoerarbcihli.ngDuiasplaerfiftieecstswoitnhcthoegniintihoinbithioans vcoienwtinwueildl btoe quTenhceesnfootrionnontvheartbablilcinoggunaitliivsmeabcioliutlydorhiagvineatgeednienrarleiszeeadrcchownsieth-
hisarti noted throughout the present review and a reevaluation of that children. Following widespread belief that bilingualism was det-
This account is provided in The Mechanism of Neuroplasticity in rimentalforintelligence(reviewinHakuta,1986),astudybyPeal
T
Bilingualism. andLambert(1962)reportedbetterperformancebybilingualchil-
Recognizing the limits of inhibition as an explanation, Green drenthanmonolingualsonbothverbalandnonverbaltasks.This
andAbutalebi(2013)recentlyexpandedtheICmodeltoprovidea studyprovidedthefirstcredibleevidencethatratherthanbeinga
moredetaileddescriptionoftheprocessesresponsibleforbilingual negativeforce,bilingualismmightinsteadhavesignificantpositive
languageselectionandtheimplicationsforcognition.IntheAdap- outcomes.AlthoughtherewereproblemswiththePealandLam-
tive Control Hypothesis, they identified eight control processes bert study (the language groups may not have been equivalent in
(goal maintenance, conflict monitoring, interference suppression, socioeconomic status or intelligence and the measures were
salientcuedetection,selectiveresponseinhibition,taskdisengage- broadlybasedintelligencetests),theresultscreatedinterestinthe
ment, task engagement, and opportunistic planning) that are dif- possibilitythatbilingualismcouldaffectnonverbalcognitionand
ferentially recruited as a function of the type of interactional that the effect could be positive. Few, if any, recent studies have
context for language use. The model described three such con- usedthemethodologiesemployedbyPealandLambert,focusing
texts—single language, dual language, and dense code-switch- instead on more precise aspects of cognitive performance using
ing—each of which makes different demands on selection. The moredetailedexperimentalmethods.Nonetheless,theirstudywas
236 BIALYSTOK
instrumental in introducing this issue as an important area for infantsfocusedonthemouthmorethanmonolingualinfantsbutby
research and theory. The initial research following their study 12monthsthisdifferencewasnolongerfound.
largelyfocusedonthedevelopmentofmetalinguisticawarenessin The second type of evidence extends the research beyond lan-
monolingual and bilingual children, with most studies reporting guageandexaminesresponsesofinfantstononverbalstimuli.Two
moreprecociousdevelopmentinbilinguals(forreview,Bialystok, studies by Kovacs and Mehler showed that bilingually raised
2001). However, the focus of the present review will be on the infantsat7-months(Kovács&Mehler,2009a)and12-monthsold
cognitiveconsequencesofbilingualism,andparticularlynonverbal (Kovács & Mehler, 2009b) were able to override a habitual re-
outcomes. sponse to look to one side to receive a visual reward (dancing
Multilingualenvironmentsininfancy. Arecentdirectionin clown) and replace it with an eye movement to the opposite side
investigations of the consequences of bilingualism is studies of whenthepositionoftherewardchanged.Infantsraisedinmono-
infantsbeingraisedinenvironmentswhereoneormorethanone lingual environments continued to look to the original location
languages are heard from birth. This environmental difference is evenwhentherewardnolongerappearedthere.Similarly,astudy
associated with differences in language acquisition; for example, of 6-month-old infants engaged in a visual habituation task dem-
publishers.natedbroadly. udbgniieslrtitinil(ngeacu.bgtaio.ol,uneWtsn7vtehirmrakoteonarnmr&tehersneTtloseelevmdsa,,anai1tnl9lftoa8iirn3nf,aalt1nlh9tniss8a4tauu)rn.reiaAvlseelrarnsensacgilteuinsvaetegnsesttsoiutbditvuyhitetbyinpyffhoaFornentlresojtaniinnc- oigannnufdsaatlnrlMaytsteerdrhaailisebsereeddtsttecuironduisenbtsitimelir(nup2gla0uur0sta9slae(,neSnc2iovn0dig0riho9nnbgem)t,aeatnnhld.te,ssr2ete0hcr1aoen5gs)un.blityLtsioiktinhneedmitirhceaemmteKoonrboyoevltáitbcneysr-
sallieddissemi Rtoaemncíerepzh,aRloagmraíprehzy, (CMlaErkGe),tToaudlouc,uamnedntKhuohwl (2th0e17in)fuansetdbimlianggnuea-l ftelemTxsihbeailnitdthyirrodevfseaoatultsercndetiifoofnferefeovnricdeiensnfacbneettsmwweoevintehsinbffuialrintnhtgseurfarilonmetoxptcheoergisenenitceienv.veirsoyns--
ite brain remains sensitive to phonetic distinctions from multiple
ofob languages. However, the present discussion is focused on the mentsintheirabilitytogeneralizememoryforeventsinthefirst
oneott possibility that these environmental differences in infancy have yearoflife.Infantmemoriesarelargelyspecifictothestimuliand
orisn measurableconsequencesfornonverbaloutcomes.Threetypesof contextinwhichtheywereformed(Hayne,2006)butthisspeci-
ationand recentevidenceconvergeontheconclusionthattheydo. fmiceimtyolriymaitnsdcocongcenpittivfeorsmyastteiomnsarnedquleiraersnignegn.eTrahleizdinegvealcorpomssenctuoesf
cier Thefirsttypeofevidenceexaminesvisualattentiontofacesas
ssous a source of linguistic information in the first year of life. In andcontexts,anachievementthatbeginstoemergeataround12
Aal months. In a series of studies, Brito and colleagues examined
ologicalindividu abmdilidinnietgiouwnahlteoennmvaairisonpntaemiankeiennrtgshaaatrtseensatwilosiontcthaoebpdlehtotoonesutpisceeadkviisinstigunacalticoduniefsfs,eirtneofnadtnetltsaenirn-- mbielimngouryalgeennveirraolnizmaetinotns.iInnfiannfatsntisnbtheiengexrpaeirsiemdeinntamloconnodliintigounalobo-r
he served a puppet perform three simple actions and then spent 30
ycth guage, even in the absence of auditory information. As with the
Psof ability to hear phonetic contrasts, monolingual infants can detect min playing with toys. Infants in the control condition did not
ne observe the initial actions and began the experiment by playing
caus suchchangesuntilabout7monthsold,butevidencefrominfants
eAmeriersonal wabaltechtoinngostiilceentwvhiedneoassspheoawkesrthswatitbcihliensgluaanlgiunafganestsocnonthtienubeastiosboef wbpiuiltpihnpgetuthawelaitsnofpyarsne.tsseTnhinteerdeeaacnwhderctehoenedqqiuuteiaoslnti.onnIunmwtbaheserwstehosetfthpmehroaosnero,nlioantgdcuihafiflledrarenenndt
ythhep thesevisualcuesupto1yearold.Thisfindinghasbeendemon- wouldgeneralizetheactionsthathadbeenperformedbythefirst
bt stratedforFrench–Englishbilingualbabieswatchingavideoofa
yrightedolelyfor aspnedaSkepranswishit–cChabteatlwaneebnilEinnggulaislhbaanbdiesFrwenactchh(inWgetihkeumsametealv.i,d2e0o0o7)f pclihunipglpdueraetlntcmohialtddhereetnhn,ieswacnodnpnutehpcaptteiot.nnoTsnihegenoifrfiecstauhnlettlsychsmhiloodrwreeenodftienthnattthheabniclmionnogtnurooa-ll
iscopndeds aAlsbpaeraekdear-Cswasitteclhlobt,etWweeeiknuEmn,g&lishWaenrdkeFrr,en2c0h12()S,etbwasotialann-Gguaallgeess, cinodnidciattieonmpeemrfoorrymegdentehreasleizaactitoionnsacornostshecuneeswbpyupbpileint.gTuhaleirnefsaunlttss
entnte theyhaveneverheard.Thus,theirresponseisbasedonacoherent thanmonolingualinfantsat6months(Brito&Barr,2014)andat
documcleisi cpoanrtciceupltaroflanagulaanggeuwagiethswyshtiecmh thaneydanroetfaomnilainar.aIsnsobcoiathtiosntudtoiesa, 1re8plmicoatnetdhsfo(Brr1it8o-m&onBtharro,ld20i1n2fa)n.tMs worheoosveerh,otmheeselarnegsuualtgsesweinre-
hisarti monolingual babies failed to detect the language switch. Thus, cluded two similar languages, two typologically different lan-
This infants who are building up representational systems for two guages,orthreelanguages;allthreegroupsoutperformedcompa-
T
languages are also learning that they are distinct and can be rable monolinguals on this task (Brito, Sebastian-Galles, & Barr,
discriminated. Corroborating these results Pons, Bosch, and 2015). This language group difference in memory generalization
Lewkowicz (2015) used eye-tracking to show that from 8- to was replicated in a study of 24-month old infants, with better
12-monthsold,bilinguallyraisedinfantspaymoreattentiontothe performancebythebilingualgroup,buttherewerenodifferences
mouth of a talking face than the eyes whereas monolingually betweeninfantsinthetwolanguagegroupsoncuedrecall,work-
raised babies generally pay more attention to the eyes. Mouth ing memory, emotional responsiveness, or productive vocabulary
movementprovidesmorerelevantlinguisticinformationthandoes (Brito,Grenell,&Barr,2014).Thus,acrossalltheseexperiments,
informationfromtheeyes,andinthefirstyear,infantsraisedwith bilingual infants showed greater memory generalization than
morethanonelanguageattendtothemoreinformativesource.A monolingualinfants,withnootherdifferencesbetweengroups.
recent study by Ayneto and Sebastian-Galles (2017) showed that Thesestudiesexaminingdetectionoflanguageswitches,visual
this bias for focusing on the mouth by bilingual children extends responsetospatialpositionforreward,andmemorygeneralization
beyond linguistic information and also characterized attention to convergeontheinterpretationthatinthefirst2yearsoflifeinfants
adult faces displaying emotional states; 8-month old bilingual being raised in multilingual environments display different pat-
BILINGUALADAPTATION 237
ternsofvisualattentionfromthoseraisedinmonolingualhomes. Executivefunctioninginmonolingualandbilingualchildren.
Thisresearchprovidesanimportantconstraintontheexplanations Most of these tasks assessing executive function in children are
for bilingual effects on cognition because these infants have es- interpreted as tests of inhibition so are particularly relevant for
sentially no productive proficiency in language. Nonetheless, bi- evaluatingtheoriginalinterpretationthatinhibitionisatthecoreof
lingualinfantsrealizethataspeakerhaschangedlanguageseven themechanismthatdistinguishesbilingualfrommonolingualcog-
iftheyhaveneverheardeitherofthelanguages,possiblybecause nition. One of the most common tasks of this type is the flanker
oftheirfocusofattentiononthemouthratherthantheeyeswhile taskintroducedbyEriksenandEriksen(1974)andadaptedbyFan
watchingspeech.Bilingualinfantsshowmoreflexiblecontrolover et al. (2002) as the attention network task (ANT) to distinguish
attentionandcanlearnanewresponsetoreplaceonethathasbeen amongthreeattentionnetworks,namely,executivecontrol,alert-
previously rewarded, showing as well better encoding of stimuli. ing, and orienting. The ANT task was subsequently adapted by
Finally, bilingual infants show better ability to extract relevant Rueda et al. (2004) to create a children’s version in which the
featuresfromvisualstimulisotheycanbegeneralizedtoformnew stimuliareahorizontallineoffishandthechild’staskisto“feed”
categories beginning at 6-months old. Together these findings the center fish by pressing a response key indicating whether the
publishers.natedbroadly. stiahsuttegengneoetatisrotlcnietl,hertaahdrteeevwhaeyhrlploayopttmehixseepsnieotsnsoruteafrgiealaetfrdoldeixbnaiygbbtlithelhienaestgstueuepnapeltirroeioonnprvosaisyrbeosidtnleimtmdyei.ofnfAfteblrpteihrlnoioncmueggsouhtaeilinsst fl(taoicstohitnohgniesrucifseeannctihttneatrgrtifaltiehlssfeh)t.ooforruorripgpfhloats.niAkteisn(giwnicftiohsnhtghcreuaensntatbnetdriaparoldsi)nttdaisinrkge,ctithnioenthmreealnsaaitpmivuee-
sallieddissemi tionhiibnihtiibnigtthatetennotniotanrgteotlmanigslueaagdeinigsnsottimauvliiabbleaseexdploannateixopnefroiernthcee pErneIgsnleinsahtne–deKatorhlreyeacsnhtuiclddhyrieldwnr’iestnhAitNnhiTtshtetoasU4kn.,5iY-teyadenaSgr-t,aoYtledassnawgn,hdaoncwdomeLrpeuasbrteil(di2n0tgh1ue1ai)rl
ite evidencefoundinthefirst2yearsoflife.
ofob Cognitiveprocessinginchildhood. Mostoftheresearchon performancetomonolingualEnglishspeakersintheUnitedStates
oneott the cognitive consequences of bilingualism began by studying and monolingual Korean speakers in both the United States and
orisn young children. This is a large body of literature and has been Korea. The bilingual children outperformed the two monolingual
ationand reviewed elsewhere (Barac, Bialystok, Castro, & Sanchez, 2014) gorldouKpsoroenanb–oEthngspliesehdbainlidngauccaulrcahcyil.dIrnenaifnolltohwe-UupnistteuddyS,ta5t.e5s-yoebar-
cier andsubjectedtoameta-analysisshowingeffectsizesrangingfrom
ssous small to large across studies (Adesope, Lavin, Thompson, & tainedaccuracyscoresthatwerecomparablewithadults(Yang&
Aal Yang, 2016). Similarly, Yoshida, Tran, Benitez, and Kuwabara
ologicalindividu Udpirnsecgsueensrlsteeeidddehwre,irt2eh.01tIan0s)ktshsoethmoantaljyionrvhitoiyglvhoelfidgshvttuasdriiooefus,sthykaoitunnldigsterocahftuilcrdoernewfnliiclwlteboreer (m2a0n1c1e)bsytubdiileindg3u-aylseawr-aosldalssoanredlaftoeudntdobtheattterbewttoerrdflleaanrkneinrgp.eKrfaopra-
he andColombo(2013)testedchildrenacrossawideagerangewho
ycth requiredsomeformofswitchingbetweentasksorresponses.The
Psof generalfindingisthatbilingualchildrenperformthesetasksbetter averaged 10-years old and showed that bilinguals performed the
ne task significantly faster than monolinguals but that accuracy was
caus thanmonolingualchildren,althoughthereareexceptionsthatwill
eAmeriersonal bneotidoinsctuhsastedth.eBseocuaurcseeothfebielainrlgyuarlesdeiafrfcehrenwcaessminotciovganteidtivbeyptrhoe- eestqaasunyidvafaorlderntathdeiunlottlhdveeerrtswicohonilgdorrfoeunap.fsFl,ainpnkoaleslsryi,btalsysokmbieencaswutuhsdeiciehthsethhateavssektiumwsuealdsittaohroee
ythhep cessing was generalization from practice with inhibition of the arrows or chevrons and reported similar effects. Poarch and Van
bt nontarget language, the assumption was that these effects would
yrightedolelyfor nthoetirbeexfpoeurniednicnecwhiiltdhrelannygouuagnegeurstehawnaasbloimutit4e-dy.eHaroswoeldvebr,ecmauosree HtahneedlltaB(s2ik0al1fya2ss)ttoetekrstt(he2ad0n71m-5y)oenateros-lotienlddgsu9aa-nlysde(afironuomnlddosstthaacntodnbdiloiitbnitoganuinsa)lesdapnaedrfPsoiomramirlceahdr
iscopndeds r2e-cyeenatrsre(sBeiaarlcyhsthoaks,Breaproacrt,eBdltahyees,e&paPtoteurlnins-iDnucbhoilids,re2n01a0say;oPuonuglinas- resAultt.ask similar to the flanker task but used less often in this
entnte Dubois, Blaye, Coutya, & Bialystok, 2011). Together with the research is the Simon task. Like the flanker, incongruent trials
documcleisi iannfaonnttasrtguedtielasn,gthueasgeeraessualltiskcealystsdoouurcbetoofnptrhoecersoslienogfdiinffheirbeinticoensoinf ianretroadssuocceiaatecduweitthhaat iknetyerpfereresss,wwiitthhtehaechcorerrsepcotnrseespkoeynspeo.sSittiiomnueldi
hisarti bilinguals. Although it is theoretically possible that infants and on one side of the display monitor. The stimuli are presented on
This toddlers actually inhibit the emerging nontarget representational onesideofthedisplaysuchthathalfthetimethestimulusdisplay
T
system while watching or listening to speech in one of their positioncorrespondstothepositionoftheresponsekey(congruent
environmentallanguages,theexplanationseemsunlikely. trials) and half the time it does not (incongruent trials). Effortful
Avarietyoftasksandmethodshasbeenusedinthisliterature attention and control are required to overcome the tendency to
but this review will focus on those that are most relevant for the respond on the same side as the stimulus presentation. Bilingual
hypothesis that bilingualism has consequences for cognitive de- childrenperformthistaskbetterthanmonolingualsat5-yearsold
velopment,inparticular,theimplicationsforexecutivefunctionor (Martin-Rhee&Bialystok,2008),5.5-yearsold(Morales,Calvo,
attention control. Thus, studies that are primarily concerned with & Bialystok, 2013), 6-years old (Bialystok, Martin, & Viswana-
the role of bilingualism in aspects of language acquisition, meta- than,2005a),7-yearsold(Poarch&VanHell,2012),and5–9year
linguistic awareness, or cognitive ability not broadly considered oldswhovariedintheirdegreeofbilingualism(second-language
part of executive function will not be reviewed. The following proficiency)withmorebilingualismassociatedwithbetterperfor-
threesectionswilldiscussevidencefromstudiesbasedonstandard mance(Tse&Altarriba,2014).
modelsofexecutivefunction,conceptsofflexibility,andasmall Some studies using these tasks have not found these effects.
setofstudiesthathaveusedneuroimagingmethods. Carlson and Meltzoff (2008) tested 6-year-olds who were mono-
238 BIALYSTOK
lingual,bilingual,orattendingalanguageimmersionprogramand predictedtheincreaseinlevelofexecutivecontrolwithinthesame
foundnosignificantdifferenceinperformanceontheflankertask, children(Riggsetal.,2014).Third,inasampleoflowsocioeco-
although they did show better bilingual performance on other nomic status bilingual children who were 8- to 10-years old,
conflict tasks in their battery. Two studies by another group also degree of bilingualism predicted performance on executive func-
failed to find performance differences attributable to language tion tasks with no contributing prediction from any other back-
backgroundonthechildren’sANT(Antonetal.,2014)oraStroop ground variable (Thomas-Sunesson, Hakuta, & Bialystok, in
task(Dunabeitiaetal.,2014).Thesestudiesbothincludedchildren press). Finally, a study of children in immersion education pro-
over a 6-year age range from approximately 6- to 12-years old gramsalsoindicatedthatperformanceonexecutivefunctiontasks
(possiblythesamechildreninbothstudiesreducingtheindepen- from about 7- to 10-years old was predicted by their degree of
dence of the contributions), and the unusually wide age range bilingualism(Bialystok&Barac,2012).
mightbepartofthereasonthattheresultsaredifferentfromthose Tosummarize,themajorityofresearchwithchildrenbasedon
foundinotherstudies.Similarly,Gathercoleetal.(2014)reported tasks that are used in the investigation of executive functioning
datafromalargesamplethatcoveredtheentirelifespan(3-year- show that bilinguals generally perform better than comparable
y. olds to older adults) representing four language “configurations” monolinguals. There are a few studies that show no differences
shers.broadl bspaeseadkinogncloamngmuuangietiedso.mTihneayncfeouannddbhetotmerepelarfnogrumaagneceinbyEnmgolinsho-- blaertgweeaegnegrraonugpeswbiuthtothurteceonovfinthcoinsegsctoundtireosloexvaemrtihneeraonleuonfuasguealilny
publinated linguals and English-dominant participants, although only for performance. Studies that are based on response inhibition (e.g.,
dmi someofthecomparisons.However,thestructureofthesamplein go/no-go task) rather than conflict resolution (e.g., flanker task)
salliedisse tneorminsfoorfmthaetiolanngwuaasgpercoovnidfiegduraabtoiountspwaratiscliapragnetlsyinuntceormntsroolflesdoacniadl gperonceerasslleyssahreowconnosiddeirfefedretoncbeebmetewaseuernesgroofuipnshiebviteinont.hough both
ite
ofob class,education,orotherrelevantmeasures,exceptthatmonolin- Flexibility,switching,andmonitoringofattentioninchildren.
oneott gualswerefromEnglandandallothergroupswerefromWales. Some studies have considered executive control in terms of con-
orsn The flanker and Simon tasks are generally considered to be structssuchasflexibility,switching,andmonitoring.Allofthese
i
ationand aassspeescstmoefntisnohfibiintihoinbitsihoonw,budtifoftehreernttarsekssutlhtsa.tImneathsuereflaandkieffreraenndt cpoanrtciteipotnsedenitnatiol itnhhoisbeitsiounb,puropcdeastsinesg,baencdaussheiftthinegtabsuktstuhseeydairnetnhoist
cier
sous Simontasks,inhibitionisintherequirementtoavoidattendingto typeofresearchinvolveallofthem.Inthesetasks,childrenneed
s
Aal a distracting cue, but for another group of executive function, toselectivelyattendtospecificinformation,ofteninthecontextof
ologicalindividu iannhdibfiotirotnheisseintatshkes,rethqeuirreesmuletnstatroesduipffperreesnst.aTphreespeotteansktsreisnpcolundsee m(inishliebaidteindg),oarndirrsewleitvcahntbeintwfoeremnartieosnpotnhsaetsm. ay need to be ignored
he such paradigms as gift delay (refrain from opening an attractive OnetaskthathasbeenusedinthisresearchistheDimensional
ch
yt
Psof gift box when left alone), go/no-go paradigms (resist pressing a Change Card Sort Task developed by Zelazo, Frye, and Rapus
ne key when a specific stimulus appears), and to some extent, day- (1996).Childrenareaskedtosortasetoftwo-dimensionalcards
as
cu
erial nighttasksinwhichthechildmustnameapicture(e.g.,sun)with (e.g., red trucks) by one dimension (e.g., color) then resort the
Amson anoppositename(“night”)andinhibittheoverlearnedassociation same card by the other dimension (e.g., shape). The ability to
eer torespondwith“day.”Thesetasksgenerallydonotshowperfor- perform the postswitch classification emerges gradually from
hp
ythe mance differences between monolingual and bilingual children about 3- to 5-years old. Although the task requires some type of
bt
yrightedolelyfor (ceBhtiaa,lr.&a,c2,C0Mo1n2ot;reeEnnstoop,,o&2si0to1B,1iBa;lCaykasetrolrs-koW,n2a&0rd1,M6&;eBlMtoznouifeffal,lce2cr0i,,02G80;1ioE3mn;gMbeinlarid,teiBnA-eRlblhroeecue- eidnxihmeicbeunittsiiviooenn.,fuiCnnhhciitlbidoirtnea,nttthenneeteirodenquttooiretshmeeleeoncttthieivsrelfnyeoattautctrleeen,adhrloytlodatattrhsieobrutrtieanlbgelverauntloet
ps
coed & Bialystok, 2008) although in all those studies, bilinguals did in mind and update it for the postswitch phase, and override the
isnd outperform monolinguals in tasks that assessed the type of inhi- responseestablishedinthepreswitchphasetoindividualstimulus
entnte bitioninvolvedintheflankerorSimontasks.Thisdissociationin cards. Several studies have demonstrated better performance in
mi
docucleis tehreedretsopobnesetepsatstteorfnsinbheitbwiteioenntcwaoststydpoeusbotfotanskthsewuidtielliytycoofnstihde- t(hBiisaltyasstkokb,y19b9il9i;ngBuiaallycshtoilkdr&enMtharatnin,by20t0h4e;irCmarolsnoonli&nguMaelltpzeoefrfs,
hisarti concept of inhibition as a mechanism for the bilingual effects on 2008; Kalashnikova & Mattock, 2014; Okanda, Moriguchi, &
Ts
hi cognition. It also challenges the notion that inhibition is a single Itakura,2010).
T
factorinconceptionsoftheexecutivefunction. A second integrative task is one in which children need to
These studies are almost exclusively based on between-groups reversetheirinitialinterpretationofanimageandseethedrawing
comparisonsthatrequireassumptionsabouttheequivalenceofthe assomethingelse.Thetaskusestheclassic“ambiguousfigures,”
groups.Fourstudieshavetakenadifferentapproachandexamined thatis,drawingsthathavetwointerpretationsdependingonyour
the relation between bilingualism and executive function perfor- point of view (e.g., duck–rabbit). The empirical question is to
mance within the same children as a function of their level of examinewhenchildrenareabletoabandontheiroriginalinterpre-
bilingualism.Thefirstisalongitudinalstudythattestedchildrenat tation of one of these images and assign a new meaning to it by
24 and 31 months and found that children who became more presenting increasingly explicit cues pointing to the other inter-
bilingual over this period showed the largest advantage over pretation. In two studies, bilingual children were able to do this
monolingual children performing a battery of executive function with fewer cues than were monolingual children (Bialystok &
tasks at 31 months (Crivello et al., 2016). The second is also a Shapero,2005;Wimmer&Marx,2014).
longitudinal study that studied children over a 1-year period be- Otherstudieshaveusedvariousmeasurestocomparemono-
tween 9- and 10-years old and showed that level of bilingualism lingual and bilingual children on tasks that require some un-
BILINGUALADAPTATION 239
specified notion of flexibility. Lee and Kim (2011) reported pathways associated with semantic processing and that myelina-
better scores from bilingual children than monolinguals on a tionofthesepathwayswasbetterforchildrenwhobecamebilin-
standardtestofcreativity,Adi-Japha,Berberich-Artzi,andLib- gual at an earlier age than for those who became bilingual later
nawi, (2010) demonstrated more flexibility in a task requiring (Mohadesetal.,2012,2015,seealsoBrainStructureinAdult-
children to draw an “impossible” object, and Greenberg, Bel- hood).
lana, and Bialystok (2013) found better ability for bilingual DellaRosaetal.(2013)conductedalongitudinalstudywith
thanmonolingualchildrentotaketheperspectiveofanobserver multilingualchildrenwhowereabout10yearsoldandtestedat
andcalculatetheappearanceofacomplexdisplay.Studiesthat two points in time over approximately 1 year to investigate
showbetterperformancebybilingualchildrenthancomparable
structuralgreymatterdensity.Theyfoundgreatervolumeinthe
monolinguals on false belief or theory of mind tasks may also
left inferior parietal gyrus (LIPG) in children who were more
fitintothiscategory(Bialystok&Senman,2004;Goetz,2003;
bilingual than their less bilingual peers. Children also per-
Kovacs, 2009; Nguyen & Astington, 2014). What is different
formedaflankertask,andascorefortheirabilitytoresolvethe
aboutthesetasksfromthosedescribedintheprevioussectionis
conflictonincongruentitemswasenteredintotheanalysiswith
y. that none of them is associated with a specific subprocess of
shers.broadl emxoerceutdiivfefufsuendcetipoenn,dseuncche oans ainsheibtiotifopnr,obcuestsreasththearteaarcehpahratsoaf tLhIePiGro.vTehraelrlemwualtsilaingnueaglataibvielitcyorarnedlatthioengrbeeytwmeaetnterthveolcuomnfeliocft
publinated thegeneralabilitytocontrolattentionandintegrateinformation sticloinrgeuaanldcthhieldcrehnildp’esrfmourmltiilninggbueatltecromonpettheencfela,nwkietrhtmasokr.eMmourle-
dmi from different sources. Moreover, evidence for bilingualism
salliedisse bbeiliinnggutahleismrelaenvdanctoufnatcrtyorofcoormigeisnfarroemmasntuipduielasteidn,awshsihchowbnoitnh stwtreikeinngth,ethtweointecsrteiansgetiimnegsrweyasmpaotsteitrivveolylurmeleatiendttohethLeIcPoGnflbiec-t
oneofitottobe tthioensitnugdyinbMy oYnaonlginegtuaall.a(2n0d1B1)ildinegscuraibleCdhiinldErexne.cuBtiiavleysFtouknce-t sjucsotreedaRn2d(cid:2)to0.t4h4e).cChihlidl’dsrednewgrheoewmeureltimlionrgeuamlucltoimlinpgeuteanlcoerw(ahdo-
orsn al. (2010a) included monolingual English-speaking children had more multilingual ability showed larger increases in grey
ationandi fFrroamnceC, aannaddab,ilimngounaollicnhgiuldarlenFrienncCha-nspadeaakwinhgo cshpiolkderenEngfrloismh maStotemrevsotluudmieesihnavtheealLsoIPsGho.wnarelationbetweenneuroimaging
sociuser plus a variety of other languages. Again, the two monolingual and behavioral results. In the study by Barac et al. (2016), better
s
Aal groups performed equivalently and the bilingual group outper- differentiation of the N2/P3 waveform for bilingual children per-
ologicalindividu ffournmcteiodnbtoatshksm. oBnioalliynsgtuoakl agnrdouVpsiswonanaathvaanrie(t2y00o9f)eixneccluutdievde fitoyrmofinggoaagnod/nnoo--ggoottarisaklsw,absotahssioncdiiactaetdivweiothfmbeotrteermdiastcurriempinearfboirl--
he bilingual children in Canada, bilingual children in India, and mance.KuipersandThierry(2013)recordedERPsandpupilsize
ch
yt
Psof monolingual English-speaking children in Canada. This time, in monolingual and bilingual toddlers who were presented with
anse thetwobilingualgroupsperformedsimilarlyandoutperformed word-picturepairs.Forunrelatedpairs,bilingualsshowedgreater
cu
erial the monolingual group. There is no doubt that factors such as pupil dilation than monolinguals that was correlated with a de-
Amson culture, national origin, and language affect performance on creaseinN400amplitude;thatis,thelargerthepupildilation,the
heper these tasks for children, but bilingualism appears to be inde- lessnegativetheN400.Theauthorsinterpretedthisasindicating
ythe pendent of those influences. better semantic integration for bilinguals. Monolinguals, in con-
bt
pyrightedsolelyfor dstautEdaviefirsdoetmhnactmeeoxfnirsootlmisnhgounwaeludarifnofdiemrbeanilgciniengsguinailbncohtchilhdsirtlredunrcetanur.realliNamneidtuerfdou,inmbcutaitgoitnnhagel tcbriiaalsitnte,gdsuhawoliwttohedddmltehorerseorpenppelogiscaiattiteveidptyatthtoeenrnretsihnueltthsNaft4ro0pm0u.paiTlghdreiolauptpiaotontefrwmnaosonfaosltsihone--
coed measures.Insomecases,thesedifferencesarenotinherentlybetter gual adults performing the same task (Kuipers & Thierry, 2011).
entisntend oorthwerosr,stehefodribffielirnegnucaelscahrieldarsesnoccioamtepdarweidthtommeaosnuorlaibnlgyuablest,tebrutbien- The interpretation was that paying attention to the unexpected
mi stimulusinterferedwithsemanticintegrationforthemonolinguals
docucleis hshaovwiorgarleoatuetrcosmtruecst.uIrnalgdeennesriatly,ionutbciolimngeusaflrsoomrfnuenucrtoioimnaalgpinagttethrnast butfacilitateditforbilinguals.Theseresultspointtodifferencesin
hisarti in bilingual children that resemble those obtained from older attention to environmental stimuli between these two language
Ts groups.
hi children or adults are evidence for better brain development in
T Thesummaryoftheresearchwithinfantsandchildrenisthat
bilinguals.
the development of abilities broadly associated with executive
TwostudiesusingEEGdemonstratethattheERPwaveformsof
functioning is more precocious in bilingual children than in
bilingual children reflect greater sensitivity to stimulus change
their monolingual counterparts. The majority of studies report
thandothoseofmonolingualchildren(Baracetal.,2016;Kuipers
better performance by bilingual children on a variety of tasks,
&Thierry,2012).Anotherstudythatusedfunctionalnear-infrared
spectroscopyshowedthatmonolingualandbilingualchildrenhad althoughnosingletaskservesasthegoldstandardandalltasks
differentpatternsofbrainactivationwhileperforminganonverbal include research showing exceptions to this majority outcome.
control task, with bilingual children relying more on left hemi- Uniformly, these exceptions are cases in which children in the
sphereregionsthanmonolingualchildren(Arredondo,Hu,Satter- two language groups did not differ from each other; no studies
field, & Kovelman, in press). Finally, two articles reporting data have reported cases in which bilingualism creates a disadvan-
fromalongitudinalstudytracingwhitematterdifferencesbetween tage for children. Although the tasks themselves are very dif-
monolingualandbilingualchildrenshowthatbetweentheagesof ferent, they all require controlled, selective, or effortful atten-
8-and13-yearsold,bilingualshadbetterstructuralconnectivityon tion to perform.
240 BIALYSTOK
Effects of Bilingualism in Adulthood Blumenfeld and Marian (2011, 2014) tested young adults and
demonstrated better Stroop performance by bilinguals, although
The first article reporting cognitive consequences of bilingual-
the size of the effect was small in the first study. Coderre, Van
isminadulthoodincludedthreeexperimentsthatcomparedmono-
Heuven,andConklin(2013)administeredaStrooptasktoyoung
lingualandbilingualadultswhoweremiddle-aged((cid:3)40years)or
adults who were English monolinguals and two groups of
older ((cid:3)70 years) performing a Simon task (Bialystok, Craik,
Chinese–English bilinguals who differed in which was the domi-
Klein,&Viswanathan,2004).Thetaskwastopressakeyonone
nant language. The results confirmed an advantage for the bilin-
side of the monitor if a red stimulus appeared and a key on the
guals but importantly showed as well significant effects of such
othersideifabluestimulusappearedregardlessofthepositionof
factors as language proficiency and writing system. Finally, a
the stimulus, creating congruent and incongruent trials. The first
study by Incera and McLennan (2016) used a novel method in
experiment was a small-scale pilot study including about 10 par-
which the responses were recorded by tracking the mouse move-
ticipants in each of the four groups (Age Group (cid:4) Language
ments toward the correct response. They tested three groups—
Group)andshowedfasterrespondingbybilinguals,especiallyfor
monolingual English speakers, English–Spanish bilinguals, and
s.adly. imncidodnlger-uaegnedt tardiaullst.s,T3h0eosldeceornaddueltxsp)earnimdethnet twasaks ilnacrgluedresdcavlaerio(6u4s English–Other bilinguals. The overall times to reach the correct
herbro conditionstocontrolforresponsespeedandworkingmemory.The responseweresimilarforallthreegroups,buttherewasasignif-
blised resultswerecomparablewiththoseinExperiment1.Experiment3 icantdifferenceinhowtheresponsewasachieved.Monolinguals
salliedpudisseminat wT10eanscoamnsosmencoaulliltnivsgceuaablleloascntkudsd1yo0fobfthihleiontwgaustkahl.eImsneidtphdaeltetee-arangrlseydhbealldoduculktpss,cobovimleirnpgtlieumtaeelds. hshtaaaddrtfefaadssttmeerrotvirnainivtgiealttithoiemnemtsiomtuoesstehteshoataonrngbeeorttahofntbecirelitntrhgieauylalobneggsreoatnu,.pbsTuhitnebrteihliwantgatsuhanelyos
ofitobe performedsignificantlyfasterthanmonolingualsbutbythetenth dthifafterselonwceerbeintwiteiaetniotnheantwdofabsitleirngturaavleglrotiumpes.iTshteheaustihgonrastuarreguoef
et block the monolinguals had caught up and the difference disap-
onot expert behavior. In a related approach, Singh and Mishra (2012,
n peared. These results are consistent with the notion of practice
ationorandis emffoetciotsnoarleafrfgiceieanmcoyudnitffoefrernesceeasrbcehtwanedendrgerwouwpsi.dTeshpisreaardticaltetesnettioinn 2g0u1al3s)ocnomaSpatrroeodphtiagshk-inanwdhliocwh-tphreofriecsipeonncsyeHwiansdii-nEdnicgaltieshdbbyiliann-
ociser tobilingualism,particularlyintermsofadultcognition. eyemovementtothecorrecttarget.Bilingualsshowedmorerapid
Assalu Behavioral studies of executive function in adults. In the saccadestothetargetinallconditions.Note,however,thatastudy
aldu studybyBialystoketal.(2004),the“younger”grouphadamean by Bialystok, Craik, and Ryan (2006) using an antisaccade task
hologiceindivi awgiethoyfoaubnoguetr4a2dyueltasrsw,hboutamreosatropusnydch2o0loygeyarressoealdrc,hsoisscuobnsedquucetendt swhhoewnetdheberettsepropnesrefowrmasainncdeicbaytebdobthyayokuenygperreasnsdbuotldneordbiiflfienrgeunacles
Psycofth research turned its attention to this group. The majority of this beyetewmeeonvelmanegnuta.gMeogrreoruepsseawrchhenisthreeqrueisrpeodntsoecwlaarsifyintdhiecsaetecdobnyflitchte-
ne researchusedtasksthatwereadaptedfromtheexecutivefunction
eAmericaersonalus lfiltaePnrkoatesusrribeta.lyskthoermitsosAtNcoTmvmaroinaltyionu.seIdn taaslkarginetshciasleressteuadrych, Cisostthae, ignegnSeoprmaatlteiezrsentdusd.cioenscheapvteoffoecxuesceudtinvoetfounncstpioecni,fincamtaseklys,bcuotnoflnicat mmoonre-
ythhep Hernandez,andSebastian-Galles(2008)administeredtheANTto itoring. Costa, Hernandez, Costa-Faidella, and Sebastian-Galles
yrightedbolelyfort 1Cfa0as0ttaelrmanoovnspeoerlaianlklg,euratso,lowSkipthamnaoinsrheavsaepdreavagakenetaragsgeeaonofdf221th0ye0eaabrslie.lirBntiginluignaglcuuSaeplssa,nwiasenhrde- (oth2fe0m0li9ok)neoalildhimnogoiundaislotfearnaeddbbldoilicifnkfegcrueoannlttyavioenurisnniggonaasdnuoilfntsca.oTfnlgharenukmeenratnttairpsikuall,atotaiogdnriofwfueparss-
iscopndeds dliinsgpulaaylse.dSliemssilairntreersfuelrtesncsheofwroinmgibnectotenrgrpueernfotrsmtiamnucelibthyabnilminognuoa-l efonrcmeetdhatthcehmanogneoslitnhgeunaelsedinfocronmdoitniiotnosrinwgh.eTrehethbeilimngounaitlosrionugtpdeer--
entnte youngadultsusingaversionofaflankertaskhavebeenreported mandswerehigh,forexample,equallikelihoodofcongruentand
documcleisi icnuldtuifrfeesre(nCtacloaburnitar,ieHse,rwniatnhddeizff,eMreanrttilna,n&guCagoestgar,o2u0p1s1,;aMndadrzifefceorevnat, iwnhcoenngmruoesnttoftrtihaelst,ribaulstinthtehreebwloacsknhoaddtihffeerseanmceevbaeletwnceee.nCgornofulipcst
hisarti Asanowicz,Kriva,&Wodniecka,2013;Pelham,&Abrams,2014; monitoring was also the explanation given for better bilingual
This Tao, Marzecova, Taft, Asanowicz, & Wodniecka, 2011; Verreyt, performanceintheAXContinuousPerformanceTask(AX-CPT;
T
Woumans, Vandelanotte, Szmalec, & Duyck, 2016; Woumans, Braver et al., 2001). Participants are asked to identify the occur-
Ceuleers, Van der Linden, Szmalec, & Duyck, 2015; Yang & rence of the pattern A followed by X in a continuous stream of
Yang,2016). letters.ThecriticaltrialsarethoseinwhichanAisfollowedbyY
OtherstudieshaveusedtheStrooptasktoexamineperformance orXisprecededbyB,andsuccessfulperformancerequiresboth
differencesbetweenmonolingualandbilingualyoungadults.Bia- proactive and reactive control. Morales and colleagues (Morales,
lystok, Craik, and Luk (2008) administered both a Stroop and Gomez-Ariza, & Bajo, 2013; Morales, Yudes, Gomez-Ariza, &
Simontasktoyoungerandolderadultswhoweremonolingualor Bajo,2015)reportedbetterperformancebybilingualsthanmono-
bilingualandfoundbilingualsinbothagegroupsoutperformedthe linguals on this task and argued that the bilinguals had greater
monolinguals,althoughtherewasnolanguagegroupdifferencefor capacity to monitor these components of executive function. Fi-
youngadultsontheSimontask.TheresultsfromtheStrooptask nally, as part of a larger study of conflict monitoring in sentence
were replicated in a subsequent study testing older and younger processing,Teubner-Rhodesetal.(2016)demonstratedbetterper-
adultswithbilingualsinbothagegroupsoutperformingthemono- formancebybilingualsthanmonolingualsonanonverbalN-back
linguals(Bialystok,Poarch,Luo,&Craik,2014).Twostudiesby taskthattheydescribeasrequiringconflictmonitoring.
BILINGUALADAPTATION 241
The final group of studies falls broadly into a category of 2015)showedthesameresults(althoughthesestudiesincludethe
attentiontasks.OneexampleistheantisaccadetaskbyBialystok sameparticipantssodonotprovideanindependentcontribution).
et al. (2006) described above in which younger and older bilin- Otherstudieshavealsoshownnobehavioraldifferencesinexec-
guals were better able to direct attention to a target when there utive function in monolingual and bilingual young adults (Gath-
weredistractingcues(althoughthesametaskbasedoneyemove- ercole et al., 2014; Kalia, Wilbourn, & Ghio, 2014; Kousaie &
mentsshowednogroupdifferences).Colzatoetal.(2008)admin- Phillips, 2012b; Kousaie et al., 2014; Prior & Gollan, 2013;
isteredstopsignalinhibitionofreturn,andattentionalblinktasks Scaltritti, Peressotti, & Miozzo, 2017; von Bastian, Souza, &
toidentifytheelementsofinhibitorycontrolthatwereaffectedby Gade,2016)andolderadults(Antón,FernándezGarcía,Carreiras,
bilingualism.Althoughthereweresomelimitationsofthedesign & Dunabeitia, 2016; de Bruin, Bak, & Della Sala, 2015; Kirk,
(small sample and monolinguals and bilinguals lived in different Fiala,Scott-Brown,&Kempe,2014).
countries),theresultsshowedsignificantdifferencesbetweenthe The majority of studies showing no group differences were
groupsinapatternthattheauthorsinterpretasshowingbettergoal conductedwithyoungadultsperformingsimpletasks.Whywould
maintenanceandreactiveinhibitioninbilinguals.Continuingwith there be behavioral differences on these tasks for children and
y. the notion of inhibitory control, Treccani, Argyri, Sorace, and olderadultsbutnotyoungadults?Onepossibilitycomesfromthe
shers.broadl DbielilnlaguSalaslath(a2n00m9o)norelipnogruteadls,larergfleerctinneggamtiovreepinrhimibiintigone.ffLeocotsseilny athveesreaggerosuppese.dToafkirnegsptohneseflaannkdersttaansdkarads daneveiaxtaiomnploeb,ttahineemdefaonr
publinated fitting the category of attention tasks, Prior and MacWhinney reaction time (RT), including congruent and incongruent condi-
dmi (2010) reported smaller local costs for bilinguals in a task- tions,isabout500ms,aresponsetimethatisdifficulttoimprove
salliedisse stiwonitcbheitnwgepeanraddimigemn:siBonilsintgousaelsleccotuthldemcoorrreecetffriecsipeonntlsye.shHifotwaettveenr-, tbhortohugchhieldnrgeangeamndenotlidneraastdiumltuslaitsintgypeixcpaellryienlocneg.eTrhethmaneatnhoRsTe ffoorr
ite
ofob asubsequentstudyusingtaskswitchingconfirmedthataspectsof young adults allowing more room for individual differences re-
oneott performanceonthistaskwerebetterforbilingualsthanmonolin- lated to a group experience to influence the outcomes. In this
orsn gualsbuttherewasnooverallreductioninswitchcosts(Hernán- sense,theresultsfromyoungadultsforsimpleexecutivefunction
i
ationand adnedz,LMaianretin(2,0B1a1rc)efloóu,n&dCsmosatlale,r2m01ix3i)n.gSocvoesrtsi,tRhaotdwrigeureezn-eFgoartniveelllsy, tvaisdkusalmvaayriabbeilaittycefoilringgroleuvpeldsifafnerdenccoenstationeinmseurfgfiec.ieFnotrinretesreianrdcih-
cier
sous correlatedwiththeamountofdailylanguageswitchingreportedby withyoungadultsthatattemptstounderstandthecognitivestruc-
s
Aal participants, perhaps revealing an important underlying factor in turesinvolvedinexecutivefunctionandtheeffectoftaskmanip-
ologicalindividu tmheaskeinrge,sualntsd. Tweosrtkininggomldeemr oadryu)l,tsGwoirtahl,thCreaemptaasnkesll(i,Siamnodn,Sptriariol uenlacteiohnassolintttlheoesfefeocut.tcTohmatems,aiytibseboenneefriecaiaslonthwathiyndthiveitdausakldeexspigernis-
he (2015)reportedbetterperformancebybalancedbilingualsbutnot aresosimple.However,ifthequestionistounderstandindividual
ch
yt
Psof unbalanced bilinguals on the Simon task. However, in a larger differences more fully, then the task needs to provide the oppor-
ne study of 108 participants that included younger and older mono- tunityforawiderrangeofperformance.
as
cu
erial lingualsandbilinguals,bilingualsoutperformedmonolingualsona Otherfactorsthatmayalsopreventtheappearanceofsignificant
Amson Simon task in both age groups; subdividing the bilinguals into differences between monolingual and bilingual young adults per-
eer balancedandunbalanceddidnotchangethisoutcome(Salvatierra forming executive control tasks have been discussed elsewhere
hp
ythe &Rosselli,2011).Finally,asinthestudywithchildrendescribed and include the selection and designation of the participants as
bt
yrightedolelyfor efaeamwrlbieiergruc(ouGuessreftehingabunerermg(oCenthoualinln.g,g-u2Fa0al1ts-3Yt)o,imbfi,inliSdnogtrhugeael,ay&ltoeurBnniagatliyvasdetuoilmkts,a2gre0eq1iu7ni)r.aend my(Bsoeansk,o,tl2hin0eg1fu5ea,al2tuo0rr1e6bs;iolBifnigtahulyeasltt,aotskhk,es2,s0at1ant6di;sttKihceraolilnlmt&eertpBhroieadtlasytisuotsonekdo,f2in0th1teh3e;reZasnhuaoltlus-
ps
coed These studies show that both younger and older bilinguals & Krott, 2016). A full discussion of these factors is beyond the
isnd generally outperform monolinguals on a range of tasks that fall scope of the present review. On average studies investigating
entnte broadlywithinthecategoryofexecutivefunction.However,many youngadultstestedwithsimpleexecutivefunctiontasksoftenfail
mi
docucleis sBtuiadliyesstodkoetnoatl.s(h2o0w05tah)esaedmefifneicsttes,reedspaecSiiamllyonfotarskyotuongchailddurletns., tiontedripsrcertimthiensaeteresbuelttwseaenndwthheasteimgrpoluicpast,iobnustththeseeinsusullereissulhtsowhavtoe
hisarti youngadults,middle-agedadults,andolderadultsandfoundthat forthemorerobusteffectsofbilingualismfoundatotherpointsin
Ts
hi bilingualsoutperformedtheirmonolingualage-matesinallgroups the life span and using different approaches, including neuroim-
T
except young adults, where performance was equivalent. Thus, aging, even with young adults. If it turns out that there is abso-
early evidence showed that young adults performing these tasks lutelynotraceofaneffectofbilingualismonyoungadultsassome
frequentlyshowednoeffectoflanguageexperience.Morerecent have argued (e.g., Hilchey et al., 2015) one would still need to
studieshavealsofailedtofinddifferencesbetweenmonolinguals explainhowanexperiencethatimpactscognitiveandbrainstruc-
andbilingualsperformingexecutivefunctiontasks.Thefirstofthe tureinchildhoodandolderagerecedesinyoungadulthoodwhere
recentstudiesofthistypewasconductedbyPaapandGreenberg itappearstohavenoconsequence.Moreimportant,thestudiesthat
(2013). They reported three studies in which young adults who do show language group differences on behavioral measures of
self-identifiedasmonolingualorbilingualperformedfourstandard executivefunctionwouldstillneedtobeexplained.Thisisalarge
executivefunctiontasks,namely,antisaccade,Simon,flanker,and anddiversebodyofresearch;failurestoreplicateplacelimitson
switching. No language group contrasts were significant. A later thegeneralizabilityoftheseeffects,buttheydonoterasethem.
study following the same procedure and using the same tasks Evidence from neuroimaging in adults. Unlike evidence
(Paap & Sawi, 2014) and a third publication that presented the from behavioral studies, neuroimaging studies comparing mono-
combineddatafromthefirsttwostudies(Paap,Johnson,&Sawi, lingual and bilingual adults routinely report group differences in
242 BIALYSTOK
brain structure and function. Much of the neuroimaging research bilinguals. There was again a difference between monolinguals
withbilingualsinvestigateslanguageprocessingandtypicallyex- andbilinguals,butunlikethepreviousstudies,thegreatestdiffer-
aminesbilingualsperformingaverbaltaskintheirtwolanguages enceinthiscasewasfoundforlatebilinguals.Inaprepoststudy
or switching between them. These studies point to the way two ofadultsundergoingintenseforeignlanguagelearningtobecome
languages are represented and accessed in bilingual minds and interpreters, Mårtensson et al. (2012) reported increases in hip-
inform psycholinguistic theories more broadly (for review see pocampalvolumeandcorticalthicknessoftheLIPG(aswellasin
Abutalebi & Green, 2007). The present review, however, is fo- theleftmiddlefrontalgyrusandsuperiortemporalgyrus)withno
cused on the cognitive consequences of bilingualism, so these changesinareaforthecontrolgroup.NotethattheLIPGwasthe
studieswillnotbeincluded. region identified by Della Rosa et al. (2013) in their study with
In some of this research, neuroimaging differences between childrenandthatinthatstudy,greymattervolumeinthisregion
monolingualsandbilingualsarereportedintheabsenceofbehav- was associated with more multilingual competence and better
ioral differences in performance. Some researchers have argued performanceonanexecutivefunctiontask.
thatbrainresultscannotbeinterpretedinthatcase(García-Pentón The LIPG has been described as being involved in linguistic
y. etal.,2016;Paapetal.,2015).Researchincognitiveneuroscience, access and fluency (Poline et al., 1996), mediation of attention
shers.broadl hCoawbeezvaer,(20ta1k1e)srtehpeortoipnpgosthitee rveiseuwlt.s Foofranexfaumnpclteio,naDlenmnaigsnaentidc ((EBlumchers,bMauemyeer,tMala.,rra2m00a5,)&, aJnandckaes,E20lm11e)r,,aHndanwggoir,kianngdmJeamncokrye
publinated resonanceimaging(fMRI)studyoflearninginyoungerandolder (2014)pointout,wasidentifiedlongagobyPötzlin1925asbeing
dmi adults point out that equivalent behavioral performance was an the “language talent area” (cited by Elmer et al., 2014). These
salliedisse ahsasveitngbeccoanutrsoellietdaflolorwbeehdavthioerm.Ltuockin(v2e0s1ti4g)amteakneesuraalsidmifilfaerrepnocienst sintugdaiefsusnudpapmoertntiatslrsotrluecitnuraaclqrueigriionngoafstehceonbdraliannfgourabgielianngduaclhsatnhga-t
ite
ofob aboutinterpretingdifferencesinERPwaveforms:Electrophysiol- is involved in essential cognitive processes. While it would be
oneott ogycanprovideevidenceof“covert”mentalactivitythatprovides expected that acquisition of a new complex skill will have brain
orsn a reliable measure of discrimination, processing, and interpreta- correlates,theimportantpointinthepresentcontextisthatthese
i
ationand tiinotner.pBreytedcoinnttreorlmlinsgoffporrocpeesrsfoinrgmeafnfcicei,enbcrayin(e.dgi.f,fAerbeuntcaelsebciaentabl.e, btirvaeinpecrhfaonrmgeasnacree. also associated with aspects of superior cogni-
cier
sous 2012; Gold et al., 2013a), something not possible if both brain Otherstudieshavefoundstructuraldifferencesbetweenmono-
s
Aal functionandbehavioralperformancearedifferent.Notably,recent lingualsandbilingualsingreymatterdensityinmorefrontalareas,
ologicalindividu rbeesheaavrciohrailndcaotgan,iitnivveesntieguartoinscgiefnocreetxaakmesplpelatcheeinreltahteioanbsbeentwceeeonf aclotmhoeusgchouvnelrikaemthoerecodnicveenrstreatsieotnooffrreegsiuolntss.foFrroLnItPaGl ,retghieosnesoaurte-
he resting state connectivity and potential cognitive function (e.g., plausible candidates for neuroplastic changes from bilingualism
ch
yt
Psof Spreng et al., 2013; Vincent et al., 2008). However, in a recent becauseoftheroleofthefrontallobeinthecontrolledprocessing
ne commentary, García-Pentón et al. (2016) argued that evidence in which bilinguals typically excel. In one of the first articles to
as
cu
erial frombraindifferencesbetweenmonolingualsandbilingualswere show structural differences in frontal regions, Abutalebi et al.
Amson too“hazy”foranyconclusionsandthatbehavioraldifferencesare (2012)usedvoxel-basedmorphometrytodemonstrategreaterden-
eer a prerequisite to the interpretation of neuroimaging data. That sityinthedorsalanteriorcingulatecortex(ACC)forbilingualthan
hp
ythe assumptionisoutofstepwithcurrentcognitiveneuroscience. monolingual young adults. In their model describing the role of
bt
yrightedolelyfor taul.rBa(l2ra0bi0rna4i)ns.trcTuhhcaetnuygreuessienidnavbdoiulxilnetghl-uobaoaldsise.md TmwhoaerspfchiorosnmtdseuttucrtydeydtotbomyreeMapsoeucrrtheestlgrliruecey-t aYrettageminotanisosa,nkipn,aeNrttwaictoaurllkeansrklyion,tathhneedbPfrirolainnttg(a2ul0al1st4er)ifaffetouccmtussoaonnndccohAgannCigtCieosn(i,sneSeftroocancltsoaol,
ps
coed matterdensityinthreegroupsofyoungadults:monolinguals,early Becker,Prat,&Stocco,2016;Stocco&Prat,2014).Importantly,
isnd bilingualswhohadlearnedtheirsecondlanguagebeforetheageof they also implicate the basal ganglia in their model, a region
entnte 5 years, and late bilinguals who had learned a second language explored in greater depth by Krizman and colleagues (Krizman,
mi
docucleis bdeetnwsieteynwthasesaiggensifoicfa1n0tlyangdre1a5teyrefaorrs.bTilhinegyufaolusntdhatnhamtognreoylinmgautatelsr SKkraoues,,M20a1ri2a)nt,h&atKisraalusso,c2e0n1t4ra;lKtoriaztmteannti,oMnaarnidanc,oSnhtroool.kS,iSmkiolaer,l&y,
hisarti in the LIPG, with larger effects for early bilinguals and for con- Hosodaetal.(2013)conducteda16-weekforeignlanguagetrain-
Ts
hi trastsinthelefthemisphere.Inafollow-upstudy,anewgroupof ingstudywithuniversitystudentsandfoundthatthetrainedgroup
T
bilingualsprovideddetailedassessmentsoftheirsecond-language showedincreasedgreymattervolumeintheinferiorfrontalgyrus
proficiency. The results showed that grey matter density in this andthatthisincreaseinvolumewascorrelatedwithanincreasein
sameregionintheLIPGwaspositivelycorrelatedwithproficiency vocabularyfollowingtheinstruction.
in the second language and negatively correlated with age of Becausegreymatterdensitydeclineswithage,anotherapproach
acquisition of the second language. Thus, more proficiency and istocomparecorticalvolumeinrelevantregionsformonolingual
more experience in using a second language was associated with and bilingual older adults. Abutalebi et al. (2014) reported that
increased density of the LIPG. Extending this pattern into older bilingualsshowedbetterpreservedgreymatterdensitythancom-
age, Abutalebi et al. (2015) found that older bilinguals also had parable monolinguals in the anterior temporal pole, a region im-
increased grey matter volume in the LIPG than did comparable portantforconceptualization.AstudybyOlsenetal.(2015)found
monolinguals, although in this study the cortical density was not nosignificantdifferencebetweenmonolingualandbilingualolder
relatedtoproficiencyorageofacquisition.Consistentwiththese adultsintheanteriortemporalpolebutthegreymattervolumeof
results, Klein et al. (2014) compared grey matter density in the this region was significantly negatively correlated with age for
LIPG for young adults who were monolingual or simultaneous monolinguals but not for bilinguals; in other words, there was a