of perceptual organization
7.4 Current data: Early, parallel and automatic access to linguistic information
Atthissta��e(�inter2008–2009),ex�eri�entalevi��enceusin��MMNan��non-atten����esi��nshasbeencollecte��inall�ajor��o�ainsoflin��uisticinfor�ation:
�honolo��ical,lexical,se�antican��s��ntactic(Fi��ures2–5).
7.4.1 Phonolo��ical�rocesses
Sin��le�hone�esan��theirsi��leco�binations,s��llables,�erethefirstlin��uis-tic�aterialstobestu��ie��usin��MMN�ara��i���.Theseex�eri�entssho�e��
thatnativelan��ua��esoun��s,e.��.vo�els,elicitlar��erMMNres�onsesthantheir
analo��uesthat��onothavecorres�on��in��re�resentationsinone’s�honolo��ical
s��ste�(Dehaene-La�bertz1997;Näätänenetal.1997)Thisha��ene��irres�ec-tiveofthe�a��nitu��eofacousticcontrastsbet�eenthestan��ar��san����eviants.
Further�ore,�hileMMNisusuall��abilateralorevenri��ht-��o�inantres�onse
(Paavilainenetal.1991),such�honeticMMNssho�e��left-he�is�heric��o�i-nance,�otentiall��lin�in��theirori��intothelan��ua��e-s�ecificstructureshouse��
inthelefthe�is�here(Näätänenetal.1997;Sht��rovetal.1998,2000).Further
ex�eri�ents even sho�e�� ho� chan��es in the �attern of the MMN res�onse
�a��reflect��evelo��entofane��honolo��icalre�resentationinthe�rocessof
learnin��alan��ua��eb��chil��renora��ults(Cheour etal. 1998; Win�leretal.
1999).Laterex�eri�ents��e�onstrate��MMNsensitivit��to�honotactic�rob-abilities(Dehaene-La�bertzetal.2000;Bonteetal.2005,2007),stress�atterns
(Honbol����oetal.2004;Weberetal.2004),au��io-visualinte��rationof�honetic
infor�ation(Colinetal.2002,2004)an��variousother�honolo��icalvariables.
The ti�in�� of these �honetic an�� �honolo��ical �is�atch ne��ativities ran��e��
fro�closeto100�s(Rinneetal.1999)tonearl��~200�s(Sht��rovetal.2000).
Acco��o��ationofsuch��atainthetheoreticalfra�e�or�oftheMMNre- quire��acertainrevisionofit.It�assu����este��that,ina����itiontochan��e-��etec-tionan��short-ter��e�or���rocesses,MMNissensitivetolon��-ter��e�or��
traces �re-for�e�� in the subject’s neural s��ste� in the �rocess of their �revi-ousex�erience�iths�o�enlan��ua��e(Näätänenetal.1997,2001;Sht��rovetal.
2000).I��ortantl��,thisi��lie��thatsuchlon��-ter��e�or��tracesforlan��ua��e
ele�entscanbeco�eactivate��inthebrainb��a��eviantsti�ulusinano����-ball
sequencean��thiss�ecificactivationcanberecor��e��neuro�h��siolo��icall��even
�ithoutattentiontothesti�ulusoran��sti�ulus-oriente��tas�.Thisle��tofur-
therMMNex�eri�ents�ithlan��ua��esti�ulithattar��ete��‘hi��her-or��er’lin-��uistic�rocesses.
Figure 2. MMNreflectionsofearl��auto�aticaccesstolin��uisticinfor�ation:
�honolo��icalenhance�entfors��llables(a��a�te��fro�Sht��rovetal.2000),lexicall��en-hance���a��neticMMNfor�eanin��ful�or��s(a��a�te��fro�Sht��rovetal.2005);fMRI
counter�artoflexicalenhance�entfor�or��s(a��a�te��fro�Sht��rovetal.2008).
7.4.2 Lexical�rocesses
Si�ilar to the �honolo��ical enhance�ent of the MMN in res�onse to native
lan��ua��e’s�hone�esan��s��llables,�efoun��that�is�atchne��ativit��evo�e��
b��in��ivi��ual�or��s�as��reaterthanthattoco��arable�eanin��less�or��-li�e
(i.e.,follo�in���honolo��icalrulesofthelan��ua��e)sti�uli.Inaseriesofstu��ies,
�e�resente��subjects�ithsetsofacousticall���atche���or��an���seu��o�or��
sti�ulian��foun��anincrease��MMNres�onse�heneverthe��eviantsti�ulus
�asa�eanin��ful�or��(Pulver�üller etal.2001,2004;Sht��rovan��Pulver�üller
2002;Sht��rovetal.2005).Thisenhance�ent,t���icall���ea�in��at100–200�s,is
bestex�laine��b��theactivationofcortical�e�or��tracesfor�or��srealise��as
��istribute��stron��l��connecte���o�ulationsofneurones(Pulver�ülleran��Sht��-rov2006).Thelexicalenhance�entoftheMMN,re�eate��l��confir�e��b��our
��rou�,�asalso��e�onstrate��b��other��rou�susin��varioussti�ulusset-u�san��
lan��ua��es(Kor�ilahtietal.2001;Kujalaetal.2002;Sitti�ra�a�ornetal.2003;
En��rassetal.2004;Petti��re�etal.2004)incu��in��arecentvali��ationofitssu�e-rior-(STC)an���e��io-te��oral(MTG)sourcesb��fMRI(Sht��rovetal.2008).
Ourstu��iesin��icate��thatthelexicalstatusofthe��eviantsti�ulus�asrelevant
forelicitin��theMMN,butthelexicalstatusofthestan��ar��sti�ulus��i��not
si��nificantl��affecttheMMNa��litu��e(Sht��rovan��Pulver�üller2002).Other
re�ortssu����este��thattheevent-relate��brainres�onseevo�e��b��thestan��ar��
sti�ulus�a��alsobeaffecte��b��itslexicalstatus(Dieschetal.1998;Jacobsenet
al.2004).More��etaile��investi��ationsalsosho�e��thatMMN�a��besensitive
to�orethana��enerallexicalit��,an��canserveasanin��exof�or��-cate��or��-s�ecific�rocessin��su��ortin��,forexa��le,thenotionofearl���rocessin��an��
re�resentational��ifferencesbet�eenverbsan��nouns(Hastin��etal.2008).
Further scrutin�� of �or��-elicite�� MMN activations usin�� state-of-the-art
technolo����,suchas�hole-hea��hi��h-��ensit��MEGco�bine���ithsourcerecon-structional��orith�s,allo�e��for��etailin���or��-relate���rocessesinthebrainat
un�rece��ente��s�atio-te��oralscale.Usin��MMNres�onsestosin��le�or��s,�e
coul����e�onstrateho�theactivationa��earsfirstinthesu�eriorte��orallobe
shortl��(~130�s)aftertheinfor�ationallo�sforsti�ulusi��entification,an��,
ina��efine��or��er(�itha��ela��of~20�s),s�rea��stoinferior-frontalcortices,
�otentiall��reflectin��theflo�ofinfor�ationinthelan��ua��e-s�ecific�eris��lvian
areasinthe�rocessoflexicalaccess(Pulver�ülleretal.2003).Further�ore,�e
foun��acorrelationbet�eenthelatenc��ofMMNres�onsesinin��ivi��ualsub-jectan��thesti�ulus-s�ecific�or��reco��nition�oints��eter�ine��behaviourall��,
�hichsu����eststhat�or��-elicite��MMNreflectsthe�rocessesofin��ivi��ual�or��
reco��nitionb��thehu�anbrainan��thet�o�rocesses�a��thereforebelin�e��
(Pulver�ülleretal.2006).
increasin��l��accu�ulatin��evi��ence,beca�econsi��ere��asa�or��’sneural“si��- nature”,its�e�or��traceactivate��inthebrain.This,inturn,offere��aun�aral-lele��o��ortunit��toinvesti��ateneural�rocessin��oflan��ua��einnon-attention
��e�an��in�� an�� tas�-in��e�en��ent fashion. In��ee��, investi��ations of lan��ua��e
functionusin��MMNthatfollo�e���ereableto�rovi��ea��reatlevelof��etail
onthes�atio-te��oral�atternsofactivation�otentiall��relate��to�rocessin��of
se�anticinfor�ationattache��toin��ivi��ual�or��s.
7.4.3 Se�antic�rocesses
Totar��etse�antic�or���ro�ertiesusin��MMN,thever��firststu��iesinthis��o-�ainutilise���re��ictionsofthese�anticso�atoto����o��el(SSM,Pulver�üller
1999,2005).SSMisa�artofa��eneralfra�e�or��aintainin��that�or��-s�ecific
�e�or��tracesexistinthebrainas��istribute��neuronalcircuitsfor�e��inthe
�rocessof�utualconnectionstren��thenin��bet�een��ifferent(even��istant)ar-eas,asactions,objectsorconce�tsareex�erience��inconjunction�iththe�or��s
use��to��escribethe�(Hebb1949;Pulver�üller1999).Thereferential�eanin��
isaninte��ral�artofa�or��’sse�antics(Fre��e1980).Action�or��s’re��ularusa��e
forreferrin��toar�/han��actions(�or��ssuchaspick, write)orle��/footactions
(kick, walk)isthereforeanessentialcharacteristicoftheir�eanin��(eventhou��h
theirse�antics�a��notbeexhauste��b��it).Iflexicalre�resentationsbeco�e
�anifestcorticall��asneuronalasse�bliesan��theactionsreferre��tob��these
�or��sareso�atoto�icall��re�resente��in�otorareasofthebrain(Penfiel��an��
Ras�ussen1950),these�anticlin�sbet�eenneuronalsetsinthesecorticalre-��ionsshoul��realisethese�anticrelationshi�bet�eenthe�or��for�san��their
actions(Pulver�üller2001).Cruciall��,thislea��stothes�ecific�re��ictionthat
action�or��s�ith��ifferentreference��o�ainsinthebo����alsoactivatethecor-res�on��in��areasof�otorcortex.Thisclai�,�hichreceive��stron��su��ortfro�
conventional neuroi�a��in�� stu��ies (Hau� et al. 2004; Hau� an�� Pulver�üller
2004),�as�uttotestinMMNex�eri�entsusin����ifferent�etho��s(EEG,MEG)
an����ifferentsti�ulioft�olan��ua��es(En��lish,Finnish).
Ina����itiontotheusuall��observe��su�eriorte��oralMMNsources,theac-tivationelicite��b��the�or��sreferrin��tofacean��/orar��ove�entsinvolve��
inferiorfronto-centralareasli�el��inclu��in��thecorticalre�resentationoftheu�- �erbo����(Sht��rovetal.2004;Pulver�ülleretal.2005).Further�ore,thele��-relate���or��selicite��astron��ersu�eriorcentralsourceco��atible�iththele��
sensori�otorre�resentation.Thisle��-�or��s�ecificsu�eriorfronto-centralac-tivation�asseenlater(~170�s)thanthe�orelateralactivationforface-an��