the_neural_basis_of_language_develo años pasado

8/7/2019 The_Neural_Basis_of_Language_Develo aos pasado

1/12

Neuron 52 , 941952, December 21, 2006 2006 Elsevier Inc. DOI 10.1016/j.neuron.2006.12.002

ReviewThe Neural Basis of LanguageDevelopment and Its Impairment

Angela D. Friederici 1 ,*1 Max Planck Institute for Human Cognitive

and Brain SciencesLeipzigGermany

The neural correlates of early language developmentand language impairment are described, with the adultlanguage-related brain systems as a target model.Electrophysiological and hemodynamic studies indi-cate that language functions to be installed in thechilds brain are similar to those of adults, with lateral-ization being present at birth, phonological processesduring the rst months, semantic processes at 12months, and syntactic processes around 30 months.These ndings support the view that the brain basisof language develops continuously over time. Discon-tinuities are observed in children with language im-pairment. Here, the observed functional abnormalitiesare accompanied by structural abnormalities in infe-rior frontal and temporal brain regions.

IntroductionLanguage is acquired without much effort and seems todevelop as the brain matures. The milestones of normallanguage development as evident from behavior can bedened as follows. In therst days after birth,infants areable to discriminate between different phonemes anddistinguish the sentence melody (prosody) of their mother tongue from that of other languages. By theage of 9 months, they have acquired the inventory of phonemes as well as the specic stress patterns of their mother tongue (for a review, see Jusczyk, 1997 ). Thechild now understands the rst words and starts toproduce rst words between 11 and 13 months, witha lexicon of about 50to 75items by the age of 16monthsand a clear vocabulary spurt between the age of 18 and24 months (see Bates and Goodman, 1999 ). Syntacticstructures are acquired continuously in the secondand third year of life, with rst productions of two-word utterances at the age of 18 to 24 months, and later more-word utterances. The basic knowledge of the syn-tactic word order constraints is present around the ageof 2.5 years ( Ho hle et al., 2001 ).

The complexity of the language system is often onlydetected when examined developmentally and realizedwhen language development derails. The way in whichlanguage development can be impaired is multifaceted.It can manifest itself in the inability to acquire phonolog-ical, semantic, and syntactic information (specic lan-guage impairment [SLI]) ( Levy and Schaeffer, 2003 ), or grammatical aspects selectively (grammatical-SLI)( Van der Lely, 2005 ), the inability to read (Dyslexia)( Shaywitz et al., 1990 ), or to understand others (autism)( Baron-Cohen et al., 1985, 1997 ). As the former twotypes of impairments are the most relevant when it

comes to explaining early language development andits impairment, the present review will focus on these.

The neural parameters of the neural basis of normallanguage development, however, are still not under-stoodentirely, and the description of its impairment trulyremains incomplete. This is due to the fact that there aredramatic constraints on the investigation of the neuralcorrelates of language development. Direct animalmodel, cellular, and molecular approaches are impossi-ble, and genetic approaches can only describe correla-tions between phenotype and genotype. At the macro-scopic level, functional magnetic resonance imaging(fMRI) and structural imaging (diffusion tenor imaging[DTI]) and magnetencephalography (MEG) are notalways applicable in young childrenand infants. Electro-encephalography (EEG) and near-infraredspectroscopy(NIRS) are the only methods that can readily be appliedto this age group without much constraint.

The review consists of two parts. The rst partdiscusses the neural correlates of normal languagedevelopment using adult data as the reference model,and the second part covers the available neurophysio-logical and neuroanatomical studies on specic lan-guage impairment.

Neural Correlates of Normal Language DevelopmentStudies on the neural basis of normal language develop-ment have focused on different aspects of the languagesystem, namely (1) phonological processes concerningsuprasegmental information (i.e., prosody, sentencemelody) and segmental information (phonemes, i.e.,speech sounds relevant for word meaning), (2) lexical-semantic processes (i.e., processes that deal withword forms and word meanings), and (3) syntactic pro-cesses concerning the grammatical relation of differentwords in a sentence.

Phonological Processes An infants rst exposure to language is based onacoustic-phonetic and phonological information (for a review, see Kuhl, 2004 ).As an initial step into language,theinfant must be able to differentiate speech from non-speech sounds. When comparing forward to backwardplayed speech in a NIRS experiment, sleeping newbornsshowed a larger increase in cerebral blood volume over the left temporal brain regions for forward speech ( Penaet al., 2003 ). In an fMRI experiment with 3-month-olds,Dehaene-Lambertz et al. (2002) also found that strongleft hemispheric activation comprised the superior tem-poral gyrus for speech sounds as measured by forwardand backward speech compared to silence. This activa-tion included Heschls gyrus and extended to the supe-rior temporal sulcus andthe temporalpole(see Figure1 ).Differences between forward and backward speechwere found in the left angular gyrus and precuneus. Anadditional right frontal brain activation observed onlyin awake infants was interpreted to reect attentionalfactors. Both studies suggest a left hemispheric domi-nance for speech in early infancy which is similar tothat of adults.*Correspondence: [email protected]
mailto:[email protected]:[email protected]


2/12

One language-relevant aspect of the speech input arefeatures of its intonational contour, which allow the seg-mentation of the input into structural units. Crucially,prosodic breaks (intonational phrase boundaries) signalsyntactic phrase boundaries, thus allowing an easy en-trance into the target language. Using NIRS, Homaeet al. (2006) investigated the brain basis for prosodicprocesses in 3-month-old infants. Examining the hemo-dynamic response to normal speech and speech withattened intonational contours, they found bilateralactivation in the temporo-parietal and the frontal cor-tices for both conditions. A direct comparison of normaland attened speech revealed the right temporo-parietal cortex, suggesting a right hemispheric domi-nance for the processing of sentential prosody (pitchinformation) similar to adults ( Meyer et al., 2004 ) to bepresent already at 3 months of age.

Using a temporally moresensitivemethod, namely themeasurement of event-related brain potentials (ERP), itwas shown that the specic ERP component, identiedas an adult brain response to intonational phraseboundaries ( Steinhauer et al., 1999; Pannekamp et al.,2005 ), was present in 8-month-olds ( Pannekamp et al.,2006 ). Its topographic distribution over the scalp wassimilar to adults, but its peak was somewhat delayed.This suggests that the brain systems involved in pro-cesses of identifying intonational phrase boundariesare similar in infants and in adults, but take more timein earlier development.

The other language-relevant aspect in the speechinput is acoustic information concerning the differentphonemes and stress patterns of words in a givenlanguage. The acoustic parameter of duration is most

relevant during spoken language perception, as it differ-entiates not only between different phonemes (i.e., shortvowel [ full ] versus a long vowel [ fool ]) but also betweenwords with different stress patterns (i.e., long rst sylla-ble [ the a bstract ] versus short rst syllable [ to abstra ct ]).There are a number of ERP studies that provideevidence for the early discrimination of differentlanguage-relevant sounds and sound patterns. Thesestudies used the auditory oddball paradigm in whichparticipants are presented with a sequence of stimuli,most of which are identical (standard) but interruptedby a deviant stimulus differing on one or more acousticparameters.

ERP studies in adults have reported a mismatchnegativity (MMN) as a neural correlate of auditorydiscrimination in such a paradigm ( Na a ta nen, 1990 ; for a review, see Na a ta nen et al., 2001 ). By means of MEG, the MMN in adults was localized in the auditorycortices bilaterally (e.g., Alho et al., 1998 ). fMRI studiesthat applied a mismatch paradigm also reported activa-tion in the auditory cortices bilaterally with stronger activation in the right superior temporal gyrus to reectthe discrimination of pitch and duration in nonspeechsounds and additional frontal activations to be attribut-able to attentional processes ( Opitz et al., 2002; Doeller et al., 2003; Schall et al., 2003; Molholm et al., 2005 ).When directly comparing speech and music withrespect to duration discrimination, signicantly moreactivation was found in the left superior temporal gyrus(BA 22/42) for speech and in the right Heschls gyrus(BA 41) for music ( Tervaniemi et al., 2006 ).

While the mismatch response in adults is alwaysexpressed in a negativity, the response in infants is

Figure 1. Brain Activation in Response toSpeech Sounds in 3-Month-Olds

(A) Averaged brain activation in response tospeech sounds (forward speech and back-ward speech against rest) in 3-month-old in-fants.L, left hemisphere; R, right hemisphere.(B)(Leftpanel) Averaged brain activation in 3-month-old awake infants for forward speechagainst backward speech. (Right panel) Aver-aged hemodynamic responses of forwardspeech and backward speech in awake andasleep infants.

Neuron942


3/12

reported as either a negativity ( Cheouret al., 1998; Kush-nerenko et al., 2001 ) or a positivity ( Dehaene-Lambertzand Dehaene, 1994; Friederici et al., 2002; Leppa nenet al., 1999; Weber et al., 2004 ) and sometimes with dif-ferent scalp distributions and latencies (for a review, seeCheour et al., 2000 ).

Presenting a sequence of ve syllables with the lastone either being the same or a different one, Dehaene-Lambertz and Dehaene (1994) reported a mismatchresponse in the form of a posterior positivity for thedeviant syllable (/ga/) compared to the standard syllable(/ba/) in 2- to 3-month-old infants. An MMN-like negativ-ity was found for vowel discrimination in therst monthsof life ( Cheour et al., 1997 ) and even in newborns( Cheour-Luhtanen et al., 1995 ). Mismatch responseseither expressed as a negativity or as a positivity havebeen observed in different languages, such as Finnish,German, and English, for vowel contrasts ( Cheour et al., 1997; Leppa nen et al., 1999; Pihko et al., 1999;Friederici et al., 2002, 2004 ) and for consonant contrasts( Dehaene-Lambertz and Baillet, 1998; Rivera-Gaxiolaet al., 2005 ) early during development.

Interestingly, evidencefor a language-specicphone-mic discrimination seems to establish between the ageof 6 and 12 months ( Cheour et al., 1998; Rivera-Gaxiolaet al., 2005 ). While younger infants aged 6 and 7 monthsshow discrimination for the phonemic contrast relevantand not-relevant for their target language, older infantsaged 11 and 12 months only display a discriminationresponse for the phonemic contrast in their targetlanguage.

The importance of word stress for word recognitionduring speech perception was shown in a recent ERPstudy with infants learning Dutch ( Kooijman et al.,2005 ). In this study, 10-month-olds recognized two-syllable words with stress on the rst syllable in contin-uous speech after they had heard the words in isolation.Recognition was reected in a greater negativitybetween 350 and 500 ms over the left hemisphere for familiar words than for unfamiliar words.

Lexical-Semantic ProcessesIn adults, a particular ERP component, the N400 (i.e., anegative going wave form peaking at around 400 ms),has been identied to correlate with lexical-semanticprocesses. The semantic N400 effect is reected ina larger amplitude forwords that aresemantically incon-gruous to a given context than words that arecongruousand it is taken to indicate semantic integration difcul-ties. Moreover, the N400 amplitude is found to be larger for pseudowords than for words. This nding is inter-preted to reect the difculty in identifying a lexicalrepresentation for the pseudoword in the mental lexicon(for reviews, see Kutas and Federmeier, 2000; Kutas andvan Petten, 1994 ).

Concerning the neural basis of semantic processes,MEG studies with adults localized the sentential N400in the auditory cortex bilaterally ( Halgren et al., 2002;Ma kela et al., 2001 ) or with an additional left inferior frontal source ( Maess et al., 2006 ). fMRI experimentswith adults applying comparable sentential paradigmsindicate an involvement of the middle and superior temporal gyri bilaterally with a larger activation in theleft hemisphere ( Kuperberg et al., 2000; Ni et al., 2000;

Friederici et al., 2003 ) and sometimes the basal gangliabilaterally ( Friederici et al., 2003 ).

The adequate description of the brain basis of theN400 observed for pseudowords requires that theparadigms used in the fMRI studies and the ERP ex-periments are comparable. Such fMRI studies founda variety of activations including anterior and inferior portions of the left temporal lobe ( Mummery et al.,1999 ), anterior and middle portions of the left superior temporal gyrus ( Kotz et al., 2002 ), and the inferior frontalgyrus/sulcus bilaterally ( Mummery et al., 1999; Rossellet al., 2001; Kotz et al., 2002 ). The frontal activationsare discussed as being related to task demands,whereas the temporal activations are taken to reectlexical processes.

Lexical and Semantic Processes at the Word Level An ERP study with 11-month-old infants suggests a dif-ferential brain reactionto known andunknownwords ex-pressed as a negativity around 200 ms after word onsetwithlonger amplitude to familiar versus unfamiliar words( Thierry et al., 2003 ). Testing 14- to 20-month-olds, Millset al. (2004) also found a negativity between 200 and400ms that was larger for known than for unknown words.The distribution of which, however, seems to changefrom bilateral at 13 months to left hemisphere dominantat 20 months of age ( Mills et al., 1997 ). These data dem-onstrate that by the end of their rst year infants are ableto differentiate between familiar and unfamiliar words,but it is not clear whether infants at this age processthe semantics of words similar to adults.

In another word-processing study, the effects of wordexperience (training) and vocabulary size (word produc-tion) were tested ( Mills et al., 1997 ). In this word-learningparadigm, 20-month-olds acquired novel words either paired with a novel object or without an object. After training, the infants ERPs showed a repetition effectindicated by a reduced N200-500 amplitude to familiar and novel unpaired words, whereas ERPs indicated anincreased bilaterally distributed N200-500 for novelpaired words. This nding is taken to indicate that theN200-500 is linked to word meaning. The interpretationof this early effect as a semanticone is challenged, giventhat semantic effects in adults are observed later, i.e., inthe N400.

Friedrich and Friederici (2004, 2005a, 2005b) used anERP paradigm appropriate to test for semantic knowl-edge and knowledge about lexical forms in youngchildren between 12 and 19 months. In this paradigm,a picture of an object is presented together with anauditory stimulus that is either a word matching theobjects name or not (semantic knowledge) or witha pseudoword that is phonotactically legal or illegal(phonological knowledge). In 12-month-olds, an earlyfronto-central negativity between 100-400 ms wasfound for words that were congruous with a picturecompared to words that were incongruous. This earlyeffect was interpreted as a familiarity effect reectingthe fulllment of a lexical expectation after seeing thepicture of an object. At the age of 14 months, an N400effect was observed for semantically incongruouswords in addition to the early negativity effect for thecongruous words. By 19 months, an N400 was foundfor semantically incongruous and for phonotactically

Review943


4/12

legal pseudowords, but not for phonotactically illegalpseudowords. This indicates that at this age both realwords and phonotactically legal pseudowords are con-sidered as word candidates, whereas phonotacticallyillegal pseudowords are not. The observed semanticN400 effect in 14- and 19-month-olds reached signi-cance later,lasted longer, andhad more frontal distribu-tion when compared to adults (compare Figure 3 ). Thelatency differences suggest slower lexical-semanticprocessing in children than in adults. The more frontaldistribution could mean either that childrens semanticprocesses are still more image-based, as adults showa frontal distribution when pictures instead of wordsare processed ( West and Holcomb, 2002 ), or that chil-dren recruit frontal brain regions associated with in-creased attentional processes (e.g., Opitz et al., 2002 )in addition to those subserving semantic processing.

There are a few fMRI studies with children investigat-ing semantic processes at the word level. Usinga semantic judgment task requiring the evaluation of the semantic relatedness of two auditory words, a studywith 9- to 15-year-old children revealed activation in thetemporalgyri bilaterally (BA22),in theleft middle tempo-ralgyrus (BA 21), and in the inferior frontal gyri bilaterally(BA 47/45) ( Chou et al., 2006 ). Correlations with agewereobserved in the left middle temporal (BA 21) and theright inferior frontal gyrus (BA 47). The increased frontalactivation was interpreted to reect a broader semanticsearch and the temporal activation to relate to moreefcient access to lexical-semantic representations.Using a semantic categorization task, a recent fMRIstudy with 5- to 10-year-old children found activationin similar frontal and temporal regions of the left hemi-sphere and in the left fusiform gyrus (BA 37, BA 20), sug-gesting language to be left-lateralized as early as 5 years( Balsamo et al., 2006 ).

Semantic Processes at the Sentential LevelThe processing of semantic information at the sententiallevel has only recently been investigated in childrenyounger than 4 years of age ( Silva-Pereyra et al.,2005a; Friedrich and Friederici, 2005b ). Previous studieswith 5- to 26-year-olds ( Holcomb et al., 1992 ), 8- to13-year-olds ( Atchley et al., 2006 ), and 6- to 13-year-olds ( Hahne et al., 2004 ) reported N400-like negativitiesfor semantically anomalous sentences in children of allage groups. In a study with 2.5-, 3-, and 4-year-olds,English learning children listened passively to sentencestimuli that were either semantically correct or anoma-lous (e.g., My uncle will watch/blow the movie ) ( Silva-Pereyra et al., 2005a, 2005b ). Children at 2.5 yearsshowed an early starting frontally distributed negativitypeaking around 500 ms for semantically anomalouscompared to normal sentences ( Silva-Pereyra et al.,2005a ). In contrast, 3- and 4-year-olds demonstrated anegative slow wave negativity peaking at around 400ms,600 ms,and 800 ms.According to theauthors, theseanteriorly distributed negativities may reect either onegeneral semantic integration mechanisms or differentsemantic mechanisms, which, however, are not speci-ed further, except the late negativity, taken to reectsentence closure.

An N400-like semantic effect at the sentence level hasbeen reportedevenin younger German learning children

aged 19 and 24 months ( Friedrich and Friederici, 2005c ).This study used sentences in which there was thepresence or absence of a semantic mismatch betweenthe verb and its object argument (e.g., The cat drinksthe ball/the milk ). For 19-month-olds, a rst negativitywas observed between 400 and 500 ms, followed bya sustained negativity between 600 and 1200 ms. In24-month-olds, the negativity was found to start at300 ms and to last until 1200 ms, whereas in adultlisteners, the N400 effect was present between 300and 800 ms. These data indicate that the processesunderlying the N400 effect in children are similar toadults, but that the integration of the object noun intothe sentence context requires more processing time inyoung children.

Semantic processes of word integration into senten-tial context have been tested in English and Germanchildren between the ages of 1 year and 26 years usingdifferent sentence constructions. The studies availablesuggest an early sensitivity to semantic anomalies asreected by an N400-like negativity (1) whose durationdecreases with age, as shown in children between19 months and 2 years ( Friedrich and Friederici, 2005c )and between 5 and 15 years ( Holcomb et al., 1992 ), indi-cating faster integration processes, (2) whose amplitudedecreases linearly between 5 and 15 years, indicatingless reliance on sentential context with older age( Holcomb et al., 1992 ), and (3) whose distribution iswider in younger children than in older children or adults( Friedrich and Friederici, 2005c; Atchley et al., 2006;Holcomb et al., 1992 ). Its general morphology, however,appears to remain similar across theage from childhoodto adulthood.

Given the ERP evidence, it is likely that thesame brainregions observed in adults are recruited during child-hood. The available fMRI studies conducted with chil-dren at the sentential level, however, have not lookedat semanticprocesses in isolation butrather at text com-prehension. fMRI data from 6- to 10-year-olds listeningto short texts (compared to resting baseline) revealedbilateral activation in the auditory cortices, specicallyin the superior temporal gyrus and Heschls gyrus, theplanum temporale, as well as in theinferior frontal gyrus,the anterior cingulate regions, and parietal regions( Ulualp et al., 1998 ). When listening to stories (comparedto reversed speech), 6-year-old children showed activa-tion in the left superior temporal gyrus and sulcusextending backto theangulargyrusand inthe left middletemporal gyrus ( Ahmad et al., 2003 ). When listening tostories in which content words were missing (comparedto resting baseline), 6- to 14-year-olds displayed activa-tion in anterior superior temporal and posterior temporalareas bilaterally and in the classical inferior-frontallanguage areas ( Wilke et al., 2005 ). In this study, theinferior frontal activation was interpreted to result fromsemanticmemoryretrieval necessaryto ll in themissingword. Future fMRI studies separating semantic and syn-tacticprocesses at thesentential level will have to revealwhich of the activations found for text comprehensionare semantic in nature.

Syntactic ProcessesERP studies of syntactic processing in adults haveshown that violations of syntactic rules are associated

Neuron944


5/12

with two ERP components: a late, centro-parietalpositivity (P600) preceded by a left anterior negativity(LAN), or an early LAN (ELAN). The P600 is not onlyseen for syntactically anomalous sentences requiringsyntactic repair, but alsofor temporally syntacticambig-uous sentences requiring syntactic reanalysis ( Osterh-out and Holcomb, 1992; Osterhout et al., 1994 ) andmoreover appears to modulate as a function of syntacticcomplexity ( Kaan et al., 2000 ). According to a recentneurocognitive model of auditory language comprehen-sion ( Friederici, 2002 ), the ELAN component is viewedto reect difculties in the process of initial localstructure building. The LAN is taken to reect difcultiesin morpho-syntactic processing during the assignmentof grammatical relations in a second stage. The P600is seen to reect difculties in the stage of integrationwhere syntactic and thematic structure have to bemapped onto each other.

The brain basis of these processes has been investi-gated in adults using space- and time-sensitive imagingmethods. Studies investigating syntactic processesusing fMRI usually report superior temporal activation(including the anterior and posterior portion) as well asactivation in the inferior frontal gyrus (including Brocasarea and/or the frontal operculum) (for reviews, seeBookheimer, 2002; Friederici, 2002; Grodzinsky andFriederici, 2006 ). The anterior portion of the superior temporal gyrus and the frontal operculum have beenidentied as those brain regions that are involved insyntactic phrase structure building processes reectedby the ELAN ( Friederici et al., 2000, 2006 ).

Currently, no fMRI study and only very few ERP stud-ies on syntactic processing in children are available.

Studies investigating morphosyntactic violations inEnglish (e.g., My uncle will watch/watching the movie ),which usually elicit an LAN-P600 pattern in adults,reported no LAN effects but only a P600-like positivityfor 3- and 4-year-olds ( Silva-Pereyra et al., 2005a ). For slightly younger 30-month-old children, the positivityobservedbetween 600 and 1000 ms didnot reach signif-icance ( Silva-Pereyra et al., 2005b ).

A recent ERP study in German investigating theprocessing of local phrase structure violations (e.g.,The lion in the roars instead of The lion in the zoo roars ),which elicits an ELAN-P600 in adults, found a biphasicERP pattern consisting of a delayed ELAN and a lateP600 in children at 2.5 years ( Oberecker et al., 2005 ).When applying the same phrase structure violationparadigm to 2-year-olds, the ERPs revealed a lateP600, but no ELAN component nor any other left-lateralized negativity preceding the P600 ( Oberecker and Friederici, 2006 ) (see Figure 2 ). The combinedresults indicate that the neural basis for the ELAN andthe P600 follow a different developmental pace. Whatcould be the explanation for the nding that the ELANestablishes later than the P600? Adult data indicatethat the ELAN reects highly automatic processes of initial structure building that are unaffected bystrategy-inducing factors such a instruction ( Hahneand Friederici, 2002 ) or a variation of the amount of correct versus incorrect sentences in an experimentalset ( Hahne and Friederici, 1999 ). In contrast, the P600modulates as a function of these factors and thus seemsto reect late controlled processes. On the basis of these adult ndings, the ERP data of the children wouldsuggest that thosehighly automatic processes reected

Figure 2. Averaged Event-Related Brain Po-tentials to Auditorily Presented Correct andSyntactically Incorrect Sentences in 24-Month-Olds, 32-Month-Olds, and Adults

Vertical line indicates the onset of the violat-ing word. (A) ERPs at left anterior electrodeF7 for the early syntactic effect (ELAN). (B)Distribution maps for the difference betweencorrect and incorrect sentences in the timewindows in which the ELAN effect is ex-pected. Dark blue indicates negativity, whichis clearly present at left anterior site for 32-month-olds and for adults. (C) ERPs at cen-tro-parietal electrode PZ for the late syntacticeffect (P600).

Review945


6/12

by the ELAN are not yet established in 2-year-olds,whereas processes of syntactic and thematic integra-tion arereected by theP600 are. By pure logic, thend-ing of a P600 in the absence of the ELAN would meanthat online syntactic structure building either was pres-ent by not reected in the ERP or that the build up of thesentences structure was not entirely based on syntacticfeatures (as the preposition in not only constrains thenext word syntactically [i.e., noun instead of verb] butalso semantically [i.e., location instead of action]). Thepresence of a delayed but adult-like ELAN componentat 2.5 years appears to indicate that the neural mecha-nisms of online syntactic structure building are presentat this age, although the processes are clearly slower in children than in adults.

In summary, the studies on normal language develop-ment indicate that the language-related ERP compo-nents reecting (1) lexical-semantic processes (N400by ERP) that involve the middle and superior temporalgyri and possibly theleft inferior frontal gyrus,(2) syntac-tic processes (ELAN/LAN-P600 by ERP) which in adultsrely on the left superior temporal gyrus mainly its ante-rior and posterior portion and the left inferior frontalgyrus, and (3) prosodic processes (CPS by ERP) whichrely on the superior temporal and frontal opercular cortex of the right hemisphere appear to change in their latency and duration, but not in their basic morphologyfrom childhood to adulthood.

Neural Correlates of Language ImpairmentBefore turning to the studies on individuals diagnosedwith SLI, I will briey discuss a unique family, spanningfour generations, in which about half of the membersof the rst three generations are affected by a severedisorder of language production.

The KE family was initially described as a severe formof verbal apraxia ( Hurst et al., 1990 ). Subsequently, thefamilys decit has been described as a linguistic decitconcerning inectional morphology ( Gopnik, 1990;Gopnik and Crago, 1991 ). Further studies indicatedthat the disorder is not specic to language, but affectsthe processing of phonology and syntax as well as non-linguistic oral praxia ( Vargha-Khadem et al., 1995;Watkins et al., 1999 ). Thus, it appears that the primarydecit is a nonlinguistic one that affects language ina secondary manner, and therefore, studies of this fam-ily may not directly reect true language abnormalities.

Members of theKE family have been investigatedwithrespect to abnormalities in brain function and brainstructure. The functional dataavailable haveused differ-ent methods (PET, fMRI) and different tasks (wordrepetition, verb generation). These report overactiva-tions and underactivations for particular brain regions.Liegeois et al. (2003) using verb generation tasks in thefMRI reported underactivation for Brocas area (BA 44/45), and Vargha-Khadem et al. (1998) found overacti-vation in Brocas area for a word-repetition task in a PETstudy. The latter study registered underactivation inthe left posterior middle temporal gyrus, whereas theformer found overactivation in the superior and middletemporal gyrus bilaterally. Abnormalities in the basalganglia (again over and underactivation) were reportedfor the caudate nucleus and the putamen. As the func-tional interpretation of overactivation and underactiva-

tion is not yet clear, our conclusion will only consider deviances from the normal pattern and not their direc-tion. Functional abnormalities are seen in speech/ language-relevant areas in the temporal and inferior frontal cortex but also for those structures in the basalganglia, namely caudate and putamen known to becrucial for controlling and selecting motor sequencesnecessary for articulation ( Volkmann et al., 1992 ).

Structural studies indicate reduced gray matter volume and density of Brocas area (BA 44/45) andincreased gray matter volume and density for the poste-rior temporal gyrus/sulcus bilaterally ( Vargha-Khademet al., 1998; Belton et al., 2002 ). All studies availablefound reduced volume or gray matter density in thehead of the caudate nucleus ( Vargha-Khadem et al.,1998; Belton et al., 2002; Watkins et al., 2002 ). This univ-ocal nding for the subcortical structures has led to thesuggestion that the genetic abnormality in the KE familymay selectively affect the development of the caudatenucleus or the basal ganglia in general ( Watkins et al.,1999 ).

The genetic analyses conducted initially proposedregion 7q31 at chromosome 7 ( Fisher et al., 1998 ) andlater more specically FOXP2 located on chromosome7q31 ( Lai et al., 2001 ) as being functionally related tothe observed pathology of speech and language in theKE family. In further studies, FOXP2 expression wascorrelated with the development of motor-related braincircuits in the mouse and the human ( Lai et al., 2003 ).This latter nding is in line with the phenotype of theKE family, which is characterized by severe facial dys-praxia, affecting an articulatory movements accompa-nied by expressive and receptive language impairmentsand which can clearly be differentiated from the SLIphenotype. A differentiation of these two phenotypesis supported by genetic evidence that FOXP2 is nota major susceptibility gene for SLI ( Newbury et al.,2002; OBrien et al., 2003 ) and therefore cannot beconsidered to be the language gene.

Specic Language ImpairmentSpeciclanguage impairment (SLI) is dened as a devel-opmental disorder that selectively affects the domain of language processing. Children with SLI perform belowtheir age on language tasks requiring the processingof phonological, semantic, or syntactic information( Levy and Schaeffer, 2003 ), despite normal intelligence,an adequate learning environment, and the absence of peripheral hearing problems or emotional problems( Bishop, 1992 ). Its genetic basis has been describedby the SLI Consortium (2002, 2004) as involving chromo-somes 16q and 19q.

Different views on the underlying decit of SLI havebeen put forward. One view posits that children withSLI, or at least certain subgroups, suffer from a decitspecic to the domain of language or a subdomain,i.e., grammar ( Clahsen, 1989; Rice and Oetting, 1993; Van der Lely, 1994 ). Another perspective holds thatSLI is not due to a processing decit specic to lan-guage but to a decit in temporal auditory processing( Merzenich et al., 1993; Tallal et al., 1993 ) or even to def-icits that aregeneral in nature, such as a reduced capac-ity in processing ( Bishop, 1994; Kail, 1994 ), or a decitinthe procedural memory system ( Ullman and Pierpont,

Neuron946


7/12

2005 ). All of these views are based on empirical datamostly, however, from behavioral studies. Here, I willnot advocate one or the other view, but rather reviewthe neurophysiological studies conducted with childrenwith SLI.

Phonological and Lexical-Semantic ProcessesNeurophysiologically, 2-month-old infants from familiesat risk for SLI demonstrate a massively delayed brainresponses for the discrimination of two syllables withvowels of different durations, i.e., long versus short( Friedrich et al., 2004 ). Note that a mismatch discrimina-tion response for long versus short vowels in CV sylla-bles even differentiated newborns with risk for dyslexiafrom those without risk ( Leppa nen et al., 1999; Pihkoet al., 1999 ).

By the age of 45 months, infants from families at riskdisplayed a reduced brain response for the discrimina-tion of the stress pattern of a two-syllable word ( Weber et al., 2004 ). A longitudinal study showed in a retrospec-tive analysis that the reduced brain response observedby the age of 45 months correlated with the childrensimpaired language abilities diagnosed at the age of 4 years ( Weber et al., 2005 ). These data suggestthat early insufciencies in the processing of durationinformation can lead to language impairments in pro-duction and comprehension later in life.

With respect to semantic processes, children whowere diagnosed as having an advanced risk for SLIbasedon behavioral tasks at theage of 2.5 years alreadyby 19 months of age demonstrate ERP patterns for lexical-semantic processes that differed from their age-matched controls. These children did not show an N400as found with normally developing children at this age( Friedrich and Friederici, 2006 ) and even at the age of 14 months ( Friedrich and Friederici, 2005b ) (see Fig-ure 3 ). The absence of the N400 effect in 19-month-oldSLI children may reect insufcient lexical-semanticrepresentations that prohibit the normal detection of the

semantic mismatch between the word and the picturecontext.

An ERP study with 10- to 12-year-old children diag-nosed with SLI on the processing of semantic informa-tion at the sentence level still reveals a pattern of brainresponses that are different from age-matched typicallydeveloping children. The latter show an N400-P600pattern in response to a semantic violation, whereasthe SLI children show no N400 component, but onlya P600 (see Figure 4 ). The absence of the N400 effectsuggests weaker lexical-semantic representation of the critical words ( Sabisch et al., 2006 ).

Thus, ERP studies with SLI children indicate thatthose who are diagnosed with impaired verbal languageskills at the age of 3 or later fail to show normal age-adequate ERP patterns already during their rst monthsof life. During the rst months of life, infants at risk for SLI are characterized by a massively delayed mismatchresponse for the discrimination of syllable duration.Such a delay may affect processes at the level of a words phonological stress pattern, at the level of word form and word semantics, and even semantic pro-cesses at the sentential level. The studies reviewed indi-cate that impairments can indeed be observed at all of these levels. Whether the N400 decits observed in thelatter two studies, however, are based on a primarydecit in temporal auditory processing cannot be deter-mined on the basis of the data presented.

The decit in phonological and lexical-semanticprocesses as revealed by the SLI childrens ERP patterncan be related to a recent fMRI study with a Finnishfamily diagnosed to suffer from SLI. Hugdahl et al.(2004) conducted an auditory fMRI experiment withve members of a family with SLI (1070 years) and sixage-matched controls. The controls brain activation tovowels, pseudowords, and words was bilateral in thesuperior and middle temporal gyri with a slight laterali-zation to the left, areas usually involved in auditory, pho-nological, and word processing. There was also a small

Figure 3. Averaged ERP for Spoken Words Congruous and Incongruous with a Picture Presented Vertical line indicates the word onset. (A) ERPs of typically developing children and adults showing a semantic N400 effect except for theyoungest group. (B) ERPs of 19-month-old infants who were diagnosed with risk for SLI at the age of 2.5 years.

Review947


8/12

activation in therightinferior frontal lobe, a regionknownto support memory and attention ( Demonet et al., 1994 ).In the SLI family, the activation was also bilateral in thetemporal lobes, but it was smaller and weaker than incontrols. The reduction of the activation in the STS/ MTG, in areas usually involved in speech perceptionand the mapping of acoustic-phonetic cues onto lexicalrepresentations ( Scott and Johnsrude, 2003 ), couldlikely be related to the impairments in decoding thephonological structure of pseudowords and words. Thisinterpretation would also be in line with the observedERP abnormalities in SLI children.

Syntactic ProcessesWith respect to syntactic processes, Van der Lely (2005)hypothesizes a specic subgroup of SLI characterizedby selective impairment of a syntactic processing com-ponent in the core language system. Others view thegrammatical impairment to result from a decit in proce-dural memory ( Ullman and Pierpont, 2005 ).

In one of the rst neurophysiological studies, Nevilleet al. (1993) compared children with language impair-ment with typically developing children. SLI childrenpresented a normal ERP pattern for the processing of content words, but not for function words that carrygrammatical information. Function words that normallyelicit a LAN evoked a more bilateral or even right-lateralized negativity in language-impaired children.

In an ERP study with 12- to 14-year-old children diag-nosed with a selective grammatical impairment (G-SLI)on the basis of behavior tests (e.g., Van der Lely et al.,1998; Van der Lely, 2005 ), it was found that these chil-dren do not demonstrate the syntax-related ELAN com-ponent for violations of nonlocal syntactic violationsobserved in normal age-matched controls. As thesechildren showed an N400 in relation to semantic aspectsof processing, the results were taken as evidence for a selective impairment of grammatical aspects of lan-guage processing ( Van der Lely and Fonteneau, 2006 ).

Neuroanatomy of the SLI BrainThe brain basis of SLI is not well investigated. There are,however, a number of MRI studies assessing structuralbrain abnormalities in children with SLI (for an overview,see Ullman and Pierpont, 2005 ). Regional abnormalitieshave been reported for language-related and other areas. Investigating 9-year-olds, Gauger et al. (1997)found a volume decrease in the left pars triangularis aspart of the Brocas area, and Jernigan et al. (1991)reported a volume decrease in the left posterior tempo-ral region. Moreover, it appears that children with SLIhave a less leftward structural asymmetry for frontaland temporal language-related regions ( Gauger et al.,1997; Plante et al., 1991 ).

In addition to abnormalities in the temporal, orbito-frontal, dorsolateral, and medial frontal cortex, abnor-malities in the inferior frontal regions in SLI childrenhave been reported in some studies ( Gauger et al.,1997; Jernigan et al., 1991 ). In adults, the inferior frontalgyrus (Brocas area) has been functionally dened tosupport phonological processes (dorsal portion) andsemantic processes (anterior ventral portion) as wellas syntacticprocesses (posterior ventral portion) ( Book-heimer, 2002; Friederici, 2002 ). In the absence of de-tailed fMRI studies in normal children, it is not entirelyclear to which functional aspects the observed struc-tural abnormalities should be related. There is, however,a recent proposal which takes frontal abnormality tosupport the notion of a decit in procedural memory( Ullman and Pierpont, 2005 ).

Leonard et al. (2002) compared the neuroanatomy of children with SLI and those suffering from dyslexia.Theyreported a reduced leftward asymmetry in the planumtemporale for the SLI group. Like most other anatomicalstudies, this one suggests a reduced volume in the leftplanum temporale. In adults, this area is known to be in-volved in thesegregationandidentication of sound pat-terns in speech and other acoustic patterns containingspectotemporal information ( Grifths and Warren, 2002 ).

Figure 4. Averaged ERPs to Auditorily Presented Correct and Semantically Incorrect Sentences Vertical line indicates the onset of the violating word. (A) ERPs of typically developing children and adults showing a semantic N400 effect. (B)ERPs of 10- to 12-year-old children diagnosed with SLI.

Neuron948


9/12

A structural abnormality in the planum temporale couldexplain the decit in temporal auditory processing ob-served behaviorally( Tallal et al.,1993 )and inERP studies( Friedrich et al., 2004 ), which in turn could have major impact on phonological and therefore on semantic andsyntactic processes.

Given the limited data available to explain SLI, we arenot in the position to decide which of the views putforward is the most valid. It may well be that there areat least two different subgroups, one that affects gram-matical processes specically, be it due to a decit inthe core language system ( Van der Lely, 2005 ) or toa decit in the procedural memory system ( Ullman andPierpont, 2005 ), and one that affects all aspects of language processing due to a decit in auditory tempo-ral processing ( Tallal et al., 1993 ). If this is the case, theformer group should show abnormalities in the leftinferior frontal cortex, whereas the latter group shoulddemonstrate abnormalities in the superior temporalgyrus, in particular the planum temporale.

ConclusionsThe available studies on the neural basis of normallanguage development suggest that the brain systemsunderlying language processing are in place already inearly development. The processes supported by thesedeveloping brain systems change quantitatively thoughnot qualitatively over time. In contrast, impaired lan-guage development is correlated with abnormalities inthe neurophysiological patterns of different aspectsof language processing and with abnormalities in thestructures of areas known to support language pro-cesses in the healthy adult brain.

References

Ahmad, Z., Balsamo, L.M., Sachs, B.C., Xu, B., and Gaillard, W.E.(2003). Auditory comprehension of language in young children -Neural networks identied with fMRI. Neurology 60 , 15981605.

Alho, K., Winkler, I., Escera, C., Huotilainen, M., Virtanen, J., Jaaske-lainen, I.P., Pekkonen, E., and Ilmoniemi, R.J. (1998). Processing of novel sounds and frequency changes in the human auditory cortex:Magnetoencephalographic recordings. Psychophysiology 35 ,211224.

Atchley,R.A.,Rice,M.L.,Betz,S.K.,Kwasny, K.M., Sereno, J.A., andJongman, A. (2006). A comparison of semantic and syntactic eventrelated potentials generated by children and adults. Brain Lang.99 , 236246.

Balsamo, L.M., Xu, B., and Gaillard, W.D. (2006). Language laterali-zation and the role of the fusiform gyrus in semantic processing inyoung children. Neuroimage 31 , 13061314.

Baron-Cohen, S., Baldwin, D., and Crowson, M. (1985). Do childrenwith autism use the Speakers Direction of Gaze (SDG) strategy tocrack the code of language? Child Dev. 68 , 4857.

Baron-Cohen, S., Leslie, A.M., and Frith, U. (1997). Does the autisticchild have a theory of mind? Cognition 21 , 3746.

Bates, E., and Goodman, J.C. (1999). On the emergence of grammar from the lexicon. In Emergence of Language, B. McWhinney, ed.(Hillsdale, NJ: Lawrence Earlbaum Associates), pp. 2980.

Belton, E., Salmond, C., Watkins, K., Vargha-Khadem, F., andGadian, D. (2002). Bilateral grey matter abnormalities in a familywith a mutation in FOPX2. Neuroimage 16 , 10144.

Bishop, D.V.M. (1992). The underlying nature of specic languageimpairment. J. Child Psychol. Psychiatry 33 , 366.

Bishop, D.V.M. (1994). Is specic language impairment a valid diag-nostic category? Genetic and psycholinguistic evidence. Philos.Trans. R. Soc. Lond. B. Biol. Sci. 346 , 105111.

Bookheimer, S. (2002). Functional MRI of language: New ap-proaches to understanding the cortical organization of semanticprocessing. Annu. Rev. Neurosci. 25 , 151188.

Cheour, M., Alho, K., Sainio, K., Reinikainen, K., Renlund, M., Aaltonen, O., Eerola, O., and Na a ta nen, R. (1997). The mismatchnegativity to changes in speech sounds at the age of three months.Dev. Neuropsychol. 13 , 167174.

Cheour, M., Ceponiene, R., Lehtokoski, A., Luuk, A., Allik, J., Alho,K., and Na a ta nen, R. (1998). Development of language-specicpho-neme representations in the infant brain. Nat. Neurosci. 1, 351353.

Cheour, M.,Leppanen, P.,and Kraus,N. (2000). Mismatch negativity(MMN) as a tool for investigating auditory discrimination and sen-sory memory in infants and children. Clin. Neurophysiol. 111 , 416.

Cheour-Luhtanen,M., Alho, K., Kujala, T., Sainio, K., Reinikainen, K.,Renlund, M., Aaltonen, O., Eerola, O., and Na a ta nen, R. (1995).Mismatch negativity indicates vowel discrimination in newborns.Hear. Res. 82 , 5358.

Chou, T.L., Booth, J.R., Burman, D.D., Bitan, T., Bigio, J.D., Lu, D.,and Cone, N.E. (2006). Developmental changes in the neural corre-lates of semantic processing. Neuroimage 29 , 11411149.

Clahsen, H. (1989). The grammatical characterization of develop-

mental dysphasia. Linguistics 27 , 897920.Dehaene-Lambertz, G., and Baillet, S. (1998). A phonological repre-sentation in the infant brain. Neuroreport 9, 18851888.

Dehaene-Lambertz, G., and Dehaene, S. (1994). Speed and cerebralcorrelates of syllable discrimination in infants. Nature 370 , 292295.

Dehaene-Lambertz, G., Dehaene, S., and Hertz-Pannier, L. (2002).Functional neuroimaging of speech perception in infants. Science 298 , 20132015.

Demonet, J.F., Price, C., Wise, R., and Frackowiak, R.S.J. (1994).Differential activation of right and left posterior sylvian regions bysemantic and phonological tasksa positron-emissiontomographystudy in normal human subjects. Neurosci. Lett. 182 , 2528.

Doeller, C.F., Opitz, B., Mecklinger, A., Krick, C., Reith, W., andSchro ger, E. (2003). Prefrontal cortex involvement in preattentive

auditory deviance detection:neuroimaging and electrophysiologicalevidence. Neuroimage 20 , 12701282.

Fisher, S.E., Vargha-Khadem, F., Watkins, K.E., Monaco, A.P., andPembrey, M.E. (1998). Localisation of a gene implicated in a severespeech and language disorder. Nat. Genet. 18 , 168170.

Friederici, A.D. (2002). Towards a neural basis of auditory sentenceprocessing. Trends Cogn. Sci. 6 , 7884.

Friederici, A.D., Wang, Y., Herrmann, C.S., Maess, B., and Oertel, U.(2000). Localization of early syntactic processes in frontal and tem-poral cortical areas: A magnetoencephalographic study. Hum. BrainMapp. 11 , 111.

Friederici, A.D., Friedrich, M., and Weber, C. (2002). Neural manifes-tation of cognitive and precognitive mismatch detection in early in-fancy. Neuroreport 13 , 12511254.

Friederici, A.D., Ru schemeyer, S.-A., Hahne, A., and Fiebach, C.J.(2003). The role of left inferior frontal and superior temporal cortexin sentence comprehension: Localizing syntactic and semantic pro-cesses. Cereb. Cortex 13 , 170177.

Friederici, A.D., Gunter, T.C., Hahne, A., and Mauth, K. (2004). Therelative timing of syntactic and semantic processes in sentencecomprehension. Neuroreport 15 , 165169.

Friederici, A.D., Bahlmann, J., Heim, S., Schubotz, R.I., and Anwan-der, A. (2006). The brain differentiates human and non-human gram-mars: Functional localization and structural connectivity. Proc. Natl. Acad. Sci. USA 103 , 24582463.

Friedrich, M., and Friederici, A.D. (2004). N400-like semantic incon-gruity effect in 19-month-olds: Processing known words in picturecontexts. J. Cogn. Neurosci. 16 , 14651477.

Friedrich, M., and Friederici, A.D. (2005a). Phonotactic knowledgeand lexical-semantic processing in one-year-olds: Brain responsesto words and nonsense words in picture contexts. J. Cogn.Neurosci. 17 , 17851802.

Review949


10/12

Friedrich, M., and Friederici, A.D. (2005b). Lexical priming andsemantic integration reected in the ERP of 14-month-olds.Neuroreport 16 , 653656.

Friedrich, M., and Friederici, A.D. (2005c). Semantic sentence pro-cessing reected in the event-related potentials of one- and two-year-old children. Neuroreport 16 , 18011804.

Friedrich, M., and Friederici, A.D. (2006). Early N400 developmentand later language acquisition. Psychophysiology 43 , 112.

Friedrich, M., Weber, C., and Friederici, A.D. (2004). Electrophysio-logical evidence for delayed mismatch response in infants at-riskfor specic language impairment. Psychophysiology 41 , 772782.

Gauger, L.M., Lombardino, L.J., and Leonard, C.M. (1997). Brainmorphology in children with specic language impairment.J. Speech Lang. Hear. Res. 40 , 12721284.

Gopnik, M. (1990). Genetic basis of grammar defect. Nature 347 , 26.

Gopnik, M., and Crago, M. (1991). Familial aggregation of a develop-mental language disorder. Cognition 39 , 150.

Grifths, T.D., and Warren, J.D. (2002). The planum temporale asa computational hub. Trends Neurosci. 25 , 348353.

Grodzinsky, Y., and Friederici, A.D. (2006). Neuroimaging of syntaxand syntactic processing. Curr. Opin. Neurobiol. 16 , 240246.

Hahne, A., and Friederici, A.D. (1999). Electrophysiological evidencefor two steps in syntactic analysis: Early automatic and late con-trolled processes. J. Cogn. Neurosci. 11 , 194205.

Hahne, A., and Friederici, A.D. (2002). Differential task effect on se-mantic and syntactic processes as revealed by ERPs. Brain Res.Cogn. Brain Res. 13 , 339356.

Hahne, A., Eckstein, K., and Friederici, A.D. (2004). Brain signaturesof syntactic and semantic processes during childrens language de-velopment. J. Cogn. Neurosci. 16 , 13021318.

Halgren, E., Dhond, R.P., Christensen, N., Van Petten, C., Marin-kovic, K., Lewine, J.D., and Dale, A.M. (2002). N400-like magnetoen-cephalography responses modulated by semantic context, wordfrequency, and lexical class in sentences. Neuroimage 17 ,11011116.

Ho hle, B., Weissenborn, J., Schmitz, M., and Ischebeck, A. (2001).Discovering wordorder regularities: The role of prosodic informationfor earlyparametersetting. In Approaches to Bootstrapping: Phono-logical, Lexical, Syntactic and Neurophysiological Aspects of EarlyLanguage Acquisition, J. Weissenborn and B. Ho hle, eds. (Amster-dam: John Benjamins), pp. 249265.

Holcomb, P.J., Coffey, S.A., andNeville,H.J. (1992). Visualand audi-tory sentence processing: A developmental analysis using event-related brain potentials. Dev. Neuropsychol. 8, 203241.

Homae, F., Watanabe, H., Nakano, T., Asakawa, K., and Taga, G.(2006). The right hemisphere of sleeping infant perceives sententialprosody. Neurosci. Res. 54 , 276280.

Hugdahl, K., Gundersen, H., Brekke, C., Thomsen, T., Rimol, L.M.,and Ersland, L. (2004). fMRI brain activation in a Finnish familywith specic language impairment compared with a normal controlgroup. J. Speech Lang. Hear. Res. 47 , 162172.

Hurst, J.A., Baraitser, M., Auger, E., Graham, F., and Norrell, S.(1990). An extended family with a dominantly inherited speech disor-der. Dev. Med. Child Neurol. 32 , 347355.

Jernigan, T.L., Hesselink, J.R., Sowell, E., and Tallal, P.A. (1991).Cerebral structure on magnetic-resonance-imaging in language-impaired and learning-impaired children. Arch. Neurol. 48 , 539545.

Jusczyk, P.W. (1997). The Discovery of Spoken Language (Cam-bridge: MIT Press).

Kaan, E., Harris, A., Gibson, E., and Holcomb, P. (2000). The P600 asan index of syntactic integration difculty. Lang. Cogn. Process. 15 ,159201.

Kail, R. (1994). A method for studying the generalized slowing hy-pothesis in children with specic language impairment. J. SpeechHear. Res. 37 , 418421.

Kooijman, V., Hagoort, P., and Cutler, A. (2005). Electrophysiologicalevidence for prelinguistic infants word recognition in continuousspeech. Brain Res. 24 , 109116.

Kuhl, P.K. (2004). Early language acquisition: Cracking the speechcode. Nat. Rev. Neurosci. 5, 831843.

Kushnerenko, E., Cheour, M., Ceponiene, R., Fellman, V., Renlund,M., Soininen, K., Alku, P., Koskinen, M., Sainio, K., and Naatanen,R. (2001). Central auditory processing of durational changes in com-plex speech patterns by newborns: An event-related brain potentialstudy. Dev. Neuropsychol. 19 , 8397.

Kuperberg, G.R., McGuire, P.K., Bullmore, E.T., Brammer, M.J.,Rabe-Hesketh, S., Wright, I.C., Lythogoe, D.J., Williams, S.C.R.,and David, A.S. (2000). Common and distinct neural substrates for pragmatic, semantic, and syntactic processing of spoken senten-ces: an fMRI study. J. Cogn. Neurosci. 12 , 321341.

Kutas, M., and Federmeier, K.D. (2000). Electrophysiology revealssemantic memory use in language comprehension. Trends Cogn.Sci. 4, 463470.

Kutas, M., and van Petten, C.K. (1994). Psycholinguistics electried:Event-related brain potential investigations. In Handbook of Psy-cholinguistic, M.A. Gernsbacher, ed. (New York: Academic Press),pp. 83143.

Kotz, S.A., Cappa, S.F., von Cramon, D.Y., and Friederici, A.D.(2002). Modulation of the lexical-semantic network by auditory se-mantic priming: an event-related functional MRI study. Neuroimage17 , 17611772.

Lai, C.S.L., Fisher, S.E., Hurst, J.A., Vargha-Khadem, F., and Mon-aco, A.P. (2001). A novel forkhead-domain gene is mutated in a se-vere speech and language disorder. Nature 413 , 519523.

Lai, C.S.L., Gerrelli, D., Monaco, A., Fisher, S., and Copp, A. (2003).FOXP2 expression during brain development coincides with adultsites of pathology in a severe speech and language disorder. Brain126 , 24552462.

Leonard, C.M., Lombardino, L.J., Walsh, K., Eckert, M.A., Mockler,J.L., Rowe, L.A., Williams, S., and DeBose, C.B. (2002). Anatomicalrisk factors that distinguish dyslexia from SLI predict reading skillin normal children. J. Commun. Disord. 35 , 501531.

Leppa nen, P.H.T., Pihko, E., Eklund, K.M., and Lyytinen, H. (1999).Cortical responses of infants with and without a genetic risk for dys-lexia: II. Group effects. Neuroreport 10 , 969973.

Levy, Y., and Schaeffer, J. (2003). Language Competence acrossPopulations: Toward a Denition of Specic Language Impairment(Mahwah, NJ: Lawrence Erlbaum Associates, Publishers).

Liegeois, F., Connelly, A., Baldeweg, T., Connelly, A., Gadian, D.G.,Mishkin, M., and Vargha-Khadem, F. (2003). Language fMRI abnor-malities associated with FOXP2 gene mutation. Nat. Neurosci. 6 ,12301237.

Maess, B., Herrmann, C.S., Hahne, A., Nakamura, A., and Friederici, A.D.(2006). Localizingthe distributedlanguage network responsiblefor the N400 measured by MEG during auditory sentence process-ing. Brain Res. 1096 , 163172.

Ma kela , A.M., Makinen, V., Nikkila, M., Ilmoniemi, R.J., and Tiitinen,H. (2001). Magnetoencephalographic (MEG) localization of the audi-tory N400m: effects of stimulus duration. Neuroreport 12 , 249253.

Merzenich, M.M., Schreiner, C., Jenkins, W., and Wang, X.Q. (1993).Neural mechanisms underlying temporal integration, segmentation,and input sequence representationsome implications for the ori-gin of learning-disabilities. Ann. N Y Acad. Sci. 682 , 122.

Meyer,M., Steinhauer, K.,Alter,K., Friederici, A.D., andvon Cramon,D.Y. (2004). Brain activity varies with modulation of dynamic pitchvariance in sentence melody. Brain Lang. 89 , 277289.

Mills, D.L., Coffey-Corina, S., and Neville, H.J. (1997). Languagecomprehension and cerebral specialization from 13 to 20 months.Dev. Neuropsychol. 13 , 397445.

Mills, D.L., Prat, C., Zangl, R., Stager, C.L., Neville, H.J., and Werker,J.F. (2004). Language experienceand the organization of brainactiv-ity to phonetically similar words: ERP evidence from 14-and 20-month-olds. J. Cogn. Neurosci. 16 , 14521464.

Molholm, M.,Martinez,A., Ritter,W., Javitt,D.C., andFoxe,J.J. (2005).The neural circuitry of pre-attentive auditory change-detection: anfMRI study of pitch and duration mismatch negativity generators.Cereb. Cortex 15 , 545551.

Neuron950


11/12

Mummery, C.J., Patterson, K., Wise, R.J.S., Vandenbergh, R., Price,C.J., and Hodges, J.R. (1999). Disrupted temporal lobe connectionsin semantic dementia. Brain 122 , 6173.

Na a ta nen, R.(1990). Theroleof attentionin auditory information pro-cessing as revealed by event-related potentials and other brainmeasures of cognitive function. Behav. Brain Sci. 13 , 201232.

Na a ta nen, R.,Tervaniemi,M., Sussman, E.,Paavilainen, P.,and Win-kler, I. (2001). Primitive intelligence in the auditory cortex. TrendsNeurosci. 24 , 283288.

Neville, H.J., Coffey, S.A., Holcomb, P.J., and Tallal, P. (1993). Theneurobiology of sensory and language processing in language-impaired children. J. Cogn. Neurosci. 5, 235253.

Newbury, D.F., Bonora, E., Lamb, J.A., Fisher, S., Lai, C.S.L., Baird,G., Jannoun, L., Slonims, V., Scott, C.M., Merricks, M.J., et al., andInternational Molecular Genetic Study of Autism Consortium.(2002). FOXP2 is not a major susceptibility gene for autism or spe-cic language impairment. Am. J. Hum. Genet. 70 , 13181327.

Ni, W., Constable, R.T., Mencl, W.E., Pugh, K.R., Fulbright, R.K.,Shaywitz, S.E., Shaywitz, B.A., Gore, J.C., and Shankweiler, D.(2000). An event-related neuroimaging study distinguishing formand content in sentence processing.J. Cogn.Neurosci. 12 , 120133.

OBrien, E.K., Zhang, X., Nishimura, C., Tomblin, J.B., and Murray,J.C. (2003). Association of Specic Language Impairment (SLI) tothe region of 7q31. Am. J. Hum. Genet. 72 , 15361543.

Oberecker, R., and Friederici, A.D. (2006). Syntactic ERP compo-nents in 24-month-olds sentence comprehension. Neuroreport 17 ,10171021.

Oberecker, R., Friedrich, M., and Friederici, A.D. (2005). Neural cor-relates of syntactic processing in two-year-olds. J. Cogn. Neurosci.17 , 16671678.

Opitz,B., Rinne,R., Mecklinger, A.,von Cramon, D.Y., andSchroger,E. (2002). Differential contribution of frontal and temporal cortices toauditory change detection: fMRI and ERP results. Neuroimage 15 ,167174.

Osterhout, L., and Holcomb, P.J. (1992). Event-related brain poten-tials elicited by syntactic anomaly. J. Mem. Lang. 31 , 785806.

Osterhout,L., Holcomb, P.J., andSwinney,D.A. (1994). Brain poten-tialselicited by garden-pathsentencesevidence of the applicationof verb information during parsing. J. Exp. Psychol. Learn. Mem.Cogn. 20 , 786803.

Pannekamp, A., Toepel, U., Alter, K., Hahne, A., and Friederici, A.D.(2005). Prosody-driven sentence processing: An event-related brainpotential study. J. Cogn. Neurosci. 17 , 407421.

Pannekamp, A.,Weber, C.,and Friederici, A.D. (2006). Prosodic pro-cessing at sentence level in infants. Neuroreport 17 , 675678.

Pena, M., Maki, A., Kovacic, D., Dehaene-Lambertz, G., Koizumi, H.,Bouquet, F., and Mehler, J. (2003). Sounds and silence: An opticaltopography study of language recognition at birth. Proc. Natl. Acad. Sci. USA 1000 , 1170211705.

Pihko, E., Leppa nen, P.H.T., Eklund, K.M., Cheour, M., Guttorm,T.K., and Lyytinen, H. (1999). Cortical responses of infants withand without a genetic risk for dyslexia: I. Age effects. Neuroreport10 , 901905.

Plante, E., Swisher, L., Vance, R., and Rapcsak, S. (1991). MRI nd-ings in boys with specic language impairment. Brain Lang. 41 ,5266.

Rice, M.L., and Oetting, J.B. (1993). Morphological decits of chil-dren with SLI Evaluation of number marking and agreement.J. Speech Hear. Res. 36 , 12491257.

Rivera-Gaxiola, M., Klarman, L., Garcia-Sierra, A., and Kuhl, P.K.(2005). Neural patterns to speech and vocabulary growth in Ameri-can infants. Neuroreport 16 , 495498.

Rossell, S.L.,Bullmore, E.T.,Williams, S.C.R., and David,A.S. (2001).Brain activation during automatic and controlled processing of semantic relations: a priming experiment using lexical-decision.Neuropsychologia 39 , 11671176.

Sabisch, B., Hahne, A., Glass, E., von Suchodoletz, W., and Frieder-ici, A.D. (2006). Lexical-semantic processes in children with SpecicLanguage Impairment. Neuroreport 17 , 15111514.

Schall, U., Johnston, P., Todd, J., Ward, P.B., and Michie, P.T.(2003). Functional Neuroanatomy of auditory mismatch processing:an event-related fMRI study of duration-deviant oddballs. Neuro-image 20 , 729736.

Scott, S.K., and Johnsrude, I.S. (2003). The neuroanatomical andfunctional organization of speech perception. Trends Neurosci. 26 ,100107.

Shaywitz, S.E., Shaywitz, B.A., Fletcher, J.M., and Escobar, M.D.(1990). Prevalance of reading-disability in boys and girlsresultsof the Connecticut Longitudinal Study. JAMA 264 , 9981002.

Silva-Pereyra, J., Klarman, L., Lin, L.J., and Kuhl, P.K. (2005a).Sentence processing in 30-month-old children: An event-relatedpotential study. Neuroreport 16 , 645648.

Silva-Pereyra, J., Rivera-Gaxiola, M., and Kuhl, P.K. (2005b). Anevent-related brain potential study of sentence comprehension inpreschoolers: semantic and morphosyntactic processing. BrainRes. 23 , 247258.

SLIConsortium.(2002). A genomewide scan identiestwo novel lociinvolved in specic language impairment. Am. J. Hum. Genet. 70 ,384398.

SLIConsortium.(2004). Highlysignicant linkage to theSLI1 locus in

an expanded sample of individualsaffectedby specic language im-pairment. Am. J. Hum. Genet. 74 , 12251238.

Steinhauer, K., Alter, K., and Friederici, A.D. (1999). Brain potentialsindicate immediate use of prosodic cues in natural speech process-ing. Nat. Neurosci. 2, 191196.

Tallal, P., Miller, S., and Fitch, R.H. (1993). Neurobiological basis of speech a case for the preeminence of temporal processing. Ann.N Y Acad. Sci. 682 , 2747.

Tervaniemi, M., Szameitat, A.J., Kruck, S., Schro ger, E., Alter, K., DeBaene, W., and Friederici, A.D. (2006). From air oscillations to musicand speech: Functional magnetic resonance imaging evidence for ne-tuned neural networks in audition. J. Neurosci. 26 , 86478652.

Thierry, G., Vihman, M., and Roberts, M. (2003). Familiar words cap-ture the attention of 11-month-olds in less than 250 ms. Neuroreport14 , 23072310.

Ullman, M.T., and Pierpont, E.I. (2005). Specic language impair-ment is not specic to language: The procedural decit hypothesis.Cortex 41 , 399433.

Ulualp, S.O., Biswal, B.B., Yetkin, F.Z., and Kidder, T.M. (1998).Functional magnetic resonance imaging of auditory cortex in chil-dren. Laryngoscope 108 , 17821786.

Van der Lely, H.K.J. (1994). Canonical linking rules: Forward vs Re-verse linking in normally developing and specically language im-paired children. Cognition 51 , 2972.

Vander Lely,H.K. (2005). Domain-specic cognitive systems: insightfrom Grammatical-SLI. Trends Cogn. Sci. 9, 5359.

Van der Lely, H.K., and Fonteneau, E. (2006). ERP signatures in lan-guage-impaired children reveal a domain-specic neural correlateof syntactic dependencies. Poster presented at The 19th AnnualCUNY Conference on Human Sentence Processing.

Van derLely,H.K.,Rosen, S.,and McClelland, A. (1998). Evidence for a grammar-specic decit in children. Curr. Biol. 8, 12531258.

Vargha-Khadem, F., Watkins, K., Alcock, K., Fletcher, P., and Pas-singham, R. (1995). Praxicand nonverbalcognitive decits in a largefamily with genetically transmitted speech and language disorder.Proc. Natl. Acad. Sci. USA 92 , 930933.

Vargha-Khadem, F., Watkins, K.E., Price, C.J., Ashburner, J., Al-cock, K.J., Connelly, A., Frackowiak, R.S., Friston, K.J., Pembrey,M.E., Mishkin, M., et al. (1998). Neural basis of an inherited speechand language disorder. Proc. Natl. Acad. Sci. USA 95 , 1269512700.

Volkmann, J., Hefter, H., Lange, H.W., and Freund, H.J. (1992). Im-pairment of temporal organization of speech in basal ganglia dis-eases. Brain Lang. 43 , 386399.

Watkins, K.E., Gadian, D.G., and Vargha-Khadem, F. (1999). Func-tional and structural brain abnormalities associated with a geneticdisorder of speech and language. Am. J. Hum. Genet. 65 ,12151221.

Review951


12/12

Watkins, K.E., Vargha-Khadem, F., Ashburner, J., Passinham, R.,Connelly, A., Friston, K., Frackowiak, R., Mishkin, M., and Gadian,D. (2002). MRI analysis of an inherited speech and language disor-der: Structural brain abnormalities. Brain 125 , 465478.

Weber, C., Hahne, A., Friedrich, M., and Friederici, A.D. (2004).Discrimination of word stress in early infant perception: Electro-physiological evidence. Brain Res. Cogn. Brain Res. 18 , 149161.

Weber, C., Hahne, A., Friedrich, M., and Friederici, A.D. (2005).Reduced stress pattern discrimination in 5-month-olds as a marker of risk for later language impairment: Neurophysiological evidence.Brain Res. Cogn. Brain Res. 25 , 180187.

West, W.C., and Holcomb, P.J. (2002). Event-related potentials dur-ing discourse-level semantic integration of complex pictures. Cogn.Brain Res. 13 , 363375.

Wilke, M., Lidzba, K., Staudt, M., Buchenau, K., Grodd, W., andKrageloh-Mann, I. (2005). Comprehensive language mapping in chil-dren, using functional magnetic resonance imaging: whats missingcounts. Neuroreport 16 , 915919.

Neuron952

the_neural_basis_of_language_develo años pasado

Documents