the adaptive brain - a neurophysiological perspective

13
The adaptive brain: A neurophysiological perspective Teija Kujala a,b, *, Risto Na ¨a ¨ta ¨nen a,c,d a Cognitive Brain Research Unit, Institute of Behavioural Sciences, University of Helsinki, Helsinki, Finland b Department of Psychology, 20014 University of Turku, Turku, Finland c Department of Psychology, University of Tartu, Tartu, Estonia d Center for Integrative Neuroscience (FIN), University of Aarhus, Aarhus, Denmark Contents 1. Introduction ...................................................................................................... 55 1.1. Involuntarily elicited neurophysiological responses ................................................................. 56 2. Plastic changes in the mechanism of cortical discrimination ............................................................... 57 2.1. Short-term adaptation ........................................................................................ 57 2.2. Longer term plastic changes.................................................................................... 58 3. Long-term auditory experience shapes the brain ......................................................................... 59 3.1. Learning effects on language acquisition .......................................................................... 59 3.2. Plasticity induced by musical training ............................................................................ 61 4. Brain plasticity in recovery and remediation ............................................................................ 61 4.1. Reorganization after brain lesion ................................................................................ 61 4.2. Recovery from sensory deprivation .............................................................................. 62 4.3. Plasticity of brains with developmental deficits .................................................................... 62 5. Concluding remarks ................................................................................................ 63 Acknowledgements ................................................................................................ 64 References ....................................................................................................... 65 1. Introduction Intensive research during the last decades has changed our view of the brain’s capacity to adapt to new or special circumstances. By now, it has been proven that in its most extreme forms, brain plasticity may even extend from one sensory Progress in Neurobiology 91 (2010) 55–67 ARTICLE INFO Article history: Received 16 July 2009 Received in revised form 22 December 2009 Accepted 21 January 2010 Keywords: Brain plasticity Learning Recovery Auditory evoked responses Auditory perceptual skills Mismatch negativity P1 N1 MMN P3a P2 ABSTRACT When an individual is learning a new skill, recovering from a brain damage, or participating in an intervention program, plastic changes take place in the brain. However, brain plasticity, intensively studied in animals, is not readily accessible in humans to whom invasive research methods cannot be applied without valid clinical or therapeutic reasons. Animal models, in turn, do not provide information about higher mental functions like language or music. Evoked neural responses have shed new light to the mechanisms underlying learning and recovery, however. Of particular interest are those higher order neural responses that can be recorded even with absence of attention, such as the mismatch negativity (MMN) and N1. They enable one to determine plastic neural changes even in patients who are unable to communicate and in infants learning a language. ß 2010 Elsevier Ltd. All rights reserved. Abbreviations: CI, cochlear implant; EEG, electroencephalography; EMF, event- related magnetic field; ERP, event-related brain potential; MEG, magnetoencepha- lography; MMN, mismatch negativity; MMNm, magnetic mismatch negativity. * Corresponding author at: Cognitive Brain Research Unit, Institute of Beha- vioural Sciences, University of Helsinki, P.O.Box 9, FIN-00014 Helsinki, Finland. E-mail address: teija.m.kujala@helsinki.fi (T. Kujala). Contents lists available at ScienceDirect Progress in Neurobiology journal homepage: www.elsevier.com/locate/pneurobio 0301-0082/$ – see front matter ß 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.pneurobio.2010.01.006

Upload: gustavocl2000

Post on 22-Nov-2014

49 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: The Adaptive Brain - A Neurophysiological Perspective

Progress in Neurobiology 91 (2010) 55–67

The adaptive brain: A neurophysiological perspective

Teija Kujala a,b,*, Risto Naatanen a,c,d

a Cognitive Brain Research Unit, Institute of Behavioural Sciences, University of Helsinki, Helsinki, Finlandb Department of Psychology, 20014 University of Turku, Turku, Finlandc Department of Psychology, University of Tartu, Tartu, Estoniad Center for Integrative Neuroscience (FIN), University of Aarhus, Aarhus, Denmark

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

1.1. Involuntarily elicited neurophysiological responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

2. Plastic changes in the mechanism of cortical discrimination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

2.1. Short-term adaptation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

2.2. Longer term plastic changes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

3. Long-term auditory experience shapes the brain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

3.1. Learning effects on language acquisition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

3.2. Plasticity induced by musical training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

4. Brain plasticity in recovery and remediation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

4.1. Reorganization after brain lesion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

4.2. Recovery from sensory deprivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

4.3. Plasticity of brains with developmental deficits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

5. Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

A R T I C L E I N F O

Article history:

Received 16 July 2009

Received in revised form 22 December 2009

Accepted 21 January 2010

Keywords:

Brain plasticity

Learning

Recovery

Auditory evoked responses

Auditory perceptual skills

Mismatch negativity

P1

N1

MMN

P3a

P2

A B S T R A C T

When an individual is learning a new skill, recovering from a brain damage, or participating in an

intervention program, plastic changes take place in the brain. However, brain plasticity, intensively

studied in animals, is not readily accessible in humans to whom invasive research methods cannot be

applied without valid clinical or therapeutic reasons. Animal models, in turn, do not provide information

about higher mental functions like language or music. Evoked neural responses have shed new light to

the mechanisms underlying learning and recovery, however. Of particular interest are those higher order

neural responses that can be recorded even with absence of attention, such as the mismatch negativity

(MMN) and N1. They enable one to determine plastic neural changes even in patients who are unable to

communicate and in infants learning a language.

� 2010 Elsevier Ltd. All rights reserved.

Contents lists available at ScienceDirect

Progress in Neurobiology

journa l homepage: www.e lsev ier .com/ locate /pneurobio

Abbreviations: CI, cochlear implant; EEG, electroencephalography; EMF, event-

related magnetic field; ERP, event-related brain potential; MEG, magnetoencepha-

lography; MMN, mismatch negativity; MMNm, magnetic mismatch negativity.

* Corresponding author at: Cognitive Brain Research Unit, Institute of Beha-

vioural Sciences, University of Helsinki, P.O.Box 9, FIN-00014 Helsinki, Finland.

E-mail address: [email protected] (T. Kujala).

0301-0082/$ – see front matter � 2010 Elsevier Ltd. All rights reserved.

doi:10.1016/j.pneurobio.2010.01.006

1. Introduction

Intensive research during the last decades has changed ourview of the brain’s capacity to adapt to new or specialcircumstances. By now, it has been proven that in its mostextreme forms, brain plasticity may even extend from one sensory

Page 2: The Adaptive Brain - A Neurophysiological Perspective

T. Kujala, R. Naatanen / Progress in Neurobiology 91 (2010) 55–6756

system to another (for reviews, see Rauschecker, 2002; Kujalaet al., 2000; Bavelier and Neville, 2002). In this case, the corticalareas deprived of input due to a complete damage to sensoryreceptors (e.g., peripheral blindness or deafness) leads to theresponsiveness of these areas to the input originally received bythe other sensory receptors. Reorganization taking place within thedifferent modalities has been even more intensively investigated,however. For example, the amputation of a finger does not leavethe corresponding brain representation area ‘‘silent’’. Instead, thisarea will process input from the neighbouring fingers, which isreflected as an increased responsiveness of the neurons in this areato the stimulation of the adjacent fingers (Merzenich et al., 1984).Analogous phenomena were also demonstrated in the visual (Kaaset al., 1990) and auditory modalities (Robertson and Irvine, 1989).Furthermore, intensive sensory training, too, causes plastic corticalchanges. For example, sound-frequency discrimination trainingincreases the cortical representation area of the correspondingfrequencies (Recanzone et al., 1993). In this case, a larger numberof neurons become responsive to the trained frequencies.

Our understanding of the cortical-map plasticity within thesensory systems is largely based on animal models (for a review,see Kaas, 2001) even though some of these phenomena, such astraining-induced enlargement of a representation area or thespreading of adjacent representations to an area deprived of input,were demonstrated in humans, too (Pascual-Leone and Torres,1993; Rossini et al., 1994; Jancke et al., 2001; Weiss et al., 2004).Research on animal neurophysiology has given the basis forunderstanding physiological changes underlying learning. At thecellular lever, a corner stone of learning is long-term potentiation(LTP). LTP is an enhancement of synaptic transmission, which maylast from hours to lifetime (Kaas, 2001). LTP is suggested to have arole in experience-dependent plasticity because of the strongcorrelation between critical periods for LTP induction andnaturally occurring plasticity. Furthermore, procedures disruptingLTP also disrupt experience-dependent plasticity. LTP appears tobehave according to the principles suggested by Hebb (1949):synapse is strengthened if the pre- and postsynaptic cells fire insynchrony. Learning affects the receptive-field size of neurons(Jenkins et al., 1990; Recanzone et al., 1993) as well as the width ofcortical columns (Recanzone et al., 1992) and representationalareas (Recanzone et al., 1992), that is, the number of neuronsdriven by the stimuli.

However, animal models are not fully informative on plasticityassociated with higher order cognitive functions, like neuralreorganization caused by the acquisition of language or musicalskills, which are inherently human phenomena. Although someauthors see strong parallels between human and animal languages(Rauschecker and Scott, 2009) and the underlying neuronalsubstrates do indeed share some of the relevant structures, it isclear that the degree of complexity reached by the human language(numerous phonemes, ten thousands of words, flexible syntacticrules) goes beyond the capacities of non-human primates, and it isnoteworthy that this difference may have a correlate in theunderlying neuroanatomical substrate (Rilling et al., 2008).

1.1. Involuntarily elicited neurophysiological responses

A feasible approach to the dynamics of plastic brain changes inhumans is to record stimulus- or event-locked synchronousactivity of neural populations giving rise to neural evokedresponses (or event-related brain potentials, ERPs; and event-related magnetic fields, EMFs; Hari et al., 2000), which permit oneto address stimulus processing with a millisecond’s temporalaccuracy. With these responses, stimulus reception and discrimi-nation as well as stimulus recognition can be studied (Naatanen,1990, 1992). Some of these responses can be elicited even

involuntarily, irrespective of the individual’s primary task ordirection of attention. Such responses are of particular significancewhen one is interested in brain functions of individuals whocannot adequately communicate, like severely aphasic patients,sleeping individuals, or infants.

Neural stimulus reception can be addressed by recordingstimulus-elicited P1, N1, and P2 responses of which the N1response has been most extensively studied. The N1 peaks at about100 ms after stimulus onsets, offsets, or changes in stimulusenergy. Therefore, the N1 can be used to monitor plastic changes inthe central afferent system induced by stimulation. It has beensuggested that N1 reflects stimulus representation area in thecortex (Naatanen, 1992). Stimulus repetition diminishes the N1amplitude, which appears to reflect the refractoriness of the neuralpopulations stimulated by the input (for a review, see Naatanenand Picton, 1987). Additionally, the N1 is modulated by latentinhibition mechanisms (Sable et al., 2004). The N1 has a negativepolarity and reaches its amplitude maximum over the fronto-central scalp areas for sounds (Naatanen and Picton, 1987). Itsgenerators for auditory stimuli are located in the supratemporalplane and lateral areas of the auditory cortices and in frontal areas(Giard et al., 1994; Naatanen and Picton, 1987).

The neural basis and plasticity of stimulus discrimination, inturn, can be best investigated with the mismatch negativity(MMN) response (Naatanen et al., 1978), which is elicited by anydiscriminable auditory change (Sams et al., 1985; Tiitinen et al.,1994; Kujala et al., 2001a; for reviews, see Naatanen et al., 2005,2007). When a deviant stimulus is presented after a string ofsimilar sounds, then an MMN is elicited at 150–250 ms afterchange onset (Naatanen et al., 2007). The main MMN generatorsare located in the supratemporal auditory cortices and frontalareas (for a review, see Alho, 1995), but also parietal-cortex sourceshave been found (Lavikainen et al., 1994). Acoustic changes elicitstronger MMNs in the right than left temporal lobe, whereasMMNs for speech sound changes are left-hemisphere preponder-ant (Kujala et al., 2007).

The MMN is based on a memory representation of the auditorypast up to about 4–15 s (Mantysalo and Naatanen, 1987; Cowanet al., 1993; Ulanovsky et al., 2004), reflecting an auditory cortexresponse to any violation of regularities in the auditory scene(Winkler et al., 2009). That is, the MMN is even elicited in theabsence of a physical sound change if the incoming stimulusviolates some aspect of the regularity present in the past auditoryinput. Thus, the MMN reflects higher order memory processes thanthe N1, which is elicited by a change in the energy or physicalproperties of a stimulus (Naatanen and Picton, 1987). Furthermore,the MMN and N1 generation seem to have at least partly distinctneurochemical mechanisms, since the blockage of NMDA receptorsabolishes the MMN but does not affect the generation of obligatorysound-elicited responses in the auditory cortex (Javitt et al., 1996).At the neuronal level, both N1 and MMN have been associated withstimulus-specific adaptation (SSA), which is a stronger reduction ofa neuron’s response to a repetitive stimulus than to a rare stimulus(Ulanovsky et al., 2004). It may underlie the maintenance andupdate of auditory representations, and its high sensitivity to smalldeviations and fast time course suggests that it encodes relation-ships between sounds and detects deviations (Winkler et al., 2009).Subcortical and cortical SSA recorded in animals occurs earlier thanthe N1 or MMN (Nelken and Ulanovsky, 2007), and, therefore,presumably does not reflect identical processes but rather theirprecursors.

A strong relationship has been found between the MMNparameters and the discrimination ability in behavioural tests(Fig. 1; Naatanen et al., 1993; Kujala et al., 2001a; Novitski et al.,2004; for reviews, see Kujala et al., 2007; Naatanen et al., 2007).MMN is large for large and easily-discriminable sound differences,

Page 3: The Adaptive Brain - A Neurophysiological Perspective

Fig. 1. The MMN reflects cortical discrimination accuracy and its amplitude parallels with perceptual test results. The MMN and behavioral responses were recorded for silent

interval differences between tone pips. In the repetitive pair, the interval between the tones was 120 ms, whereas in the occasional deviant pairs, it was 20, 60, or 100 ms (the

stimuli are illustrated under the difference waves of the ERPs). The 20-ms ISI deviant, which differed most from the standard pair, elicited a robust MMN (upper panel, left).

The MMN was also present, but smaller, for the 60-ms ISI deviant, whereas the 100-ms ISI deviant elicited no MMN. The results replicated well between two recording

sessions carried out on separate days (continuous and dashed lines). Consistently with these results, the hit rate (lower panel) was highest for the 20-ms ISI deviant, second

highest for the 60-ms ISI deviant, and close zero for the 100-ms ISI deviant. For this deviant, the false-alarm rate was highest, decreasing by the increasing stimulus difference.

Figure adapted from Kujala et al., 2001a.

T. Kujala, R. Naatanen / Progress in Neurobiology 91 (2010) 55–67 57

whereas its amplitude diminishes when the discriminationbecomes harder. This has been shown for physical stimulusdifferences, such as pitch (Baldeweg et al., 1999; Novitski et al.,2004), duration (Baldeweg et al., 1999; Amenedo and Escera,2000), and silent interval between tones (Kujala et al., 2001a).However, this relationship between the MMN and performance indiscrimination tests is not straightforward. For example, the MMNcan be elicited by violations of regularities based on complex rulesin the auditory input, whereas individuals displaying the MMNmight not consciously identify these violations (van Zuijen et al.,2006; Paavilainen et al., 2007).

The MMN is followed by a positive P3a peaking at about 250 msfrom change onset if the sound change causes an attention switch(Donchin and Coles, 1988; Escera et al., 2000). Very intrusive soundchanges are associated with a large-amplitude and short-latencyP3as, whereas minor, non-distractive sound changes might elicitno P3a. Distractable individuals, such as ADHD children havelarger-amplitude P3as than control subjects (Gumenyuk et al.,2005).

2. Plastic changes in the mechanism of cortical discrimination

Learning and memory, which are seamlessly connected toeach other, include different stages or types of processes.Learning and memory involve a range of processes from short-term adaptation to long-term memory representations some ofwhich are life-long. By recording responses to stimulusrepetition and change in the primary auditory cortex of theanesthesized cat, Ulanovsky et al. (2004) showed that the short-term memory system spans from hundreds of milliseconds to

tens of seconds, which is consistent with electrophysiologicalstudies on humans (Bottcher-Gandor and Ullsperger, 1992;Cowan et al., 1993).

2.1. Short-term adaptation

Our sensory systems not only constantly adapt to the changinginput from the environment, but also make predictions upon theforthcoming events (Naatanen et al., 2001). The central auditorysystem extracts regularities of the sound scene and stores thisinformation for short time periods (Paavilainen et al., 2001, 2007;van Zuijen et al., 2006; Bendixen et al., 2007, 2008; for reviews, seeNaatanen et al., 2001; Winkler et al., 2009). This information isthen used for automatically predicting future sound events. Whena sound violating this predictive model occurs, an MMN response iselicited. For example, if a sequence of sounds varies across a widerange of intensity and frequency dimensions but always followsthe rule ‘‘the higher the frequency the louder the intensity’’, then ahigh frequency soft sound elicits an MMN (Paavilainen et al., 2001).This type of adaptation is very rapid: already two sound exemplarsare sufficient to represent a regularity, as suggested by the MMNelicited by a stimulus violating the regularity represented by thesetwo sounds (Bendixen et al., 2007). Third violation also elicited aP3a and prolonged the reaction time in the primary task, indicatingan attention switch towards auditory stimuli. However, thesecomplex regularity analyses may occur without our awareness,being actually disconnected from consciousness (van Zuijen et al.,2006; Paavilainen et al., 2007), as suggested, for example, byMMNs for violations, which subjects are unable to learn to detect inan associative learning task (van Zuijen et al., 2006).

Page 4: The Adaptive Brain - A Neurophysiological Perspective

T. Kujala, R. Naatanen / Progress in Neurobiology 91 (2010) 55–6758

2.2. Longer term plastic changes

Sound and sound-change elicited responses can also be used asindicators of long-term plastic changes caused by learning.Naatanen et al. (1993) demonstrated that learning-inducedmemory traces can be probed with the MMN response. Theauthors hypothesized that since high discrimination accuracy isassociated with large MMN amplitudes (Lang et al., 1990), sound-discrimination learning should increase the MMN amplitude. Thiswas tested by using two complex sound patterns difficult todiscriminate from one another, implementing discriminationtraining sessions between the passive MMN recording sessions(Naatanen et al., 1993). In most of the subjects, no MMN wasoriginally elicited. However, in case the subject learned todiscriminate the stimuli in the intervening training session, thenan MMN emerged in the subsequent passive recording session (seealso Atienza et al., 2001; Gottselig et al., 2004). Similar trainingresults were also demonstrated by using sinusoidal tones(Menning et al., 2000) or speech sounds (Kraus et al., 1995;Menning et al., 2002; Tremblay et al., 1997, 1998). Plastic changesin neural processes caused by learning may even be detectableprior to the improvement of discrimination performance. Trem-blay et al. (1998) trained subjects to discriminate voice-onset timedifferences in speech sounds in four sessions and followed up thetraining effects with the MMN and identification tests. Changes inMMN were observed either concomitantly or, in some subjects,even prior to the improvement of behavioural discrimination.These results suggest that there are different stages in the learningprocess and, further, that MMN changes, reflecting the initialstages of this learning, may precede the actual improvement inperceiving sound differences.

Consistent with these results, neurophysiological responses inthe A1 of monkeys were shown to become stronger to target soundfrequencies during discrimination learning (Blake et al., 2002).Monkeys had to identify target frequencies among standardfrequencies, receiving juice reward for correct responses. Therewere frequencies which the monkeys could not originallydiscriminate, but after learning to discriminate them, the neuralresponses were elevated to all frequencies involved in training, buteven more to target frequencies.

In addition, Naatanen et al. (1993) found that learning does nottake place unless the subject actively tries to discriminate sounds,with the mere passive exposure to the stimuli not causing suchlearning effects. This is consistent with animal results showing

Fig. 2. Learning-induced plasticity and the consolidation of memory representations can b

small pitch change in one segment of a complex tone pattern (Pretraining responses on t

shading). The MMN amplitude became larger 48 and 72 h post-training in normally sl

deprived group (panel B), no such amplitude change could be observed, which indicates

Atienza et al., 2004.

expansions of the cortical reprensentation areas of the skin as aresult of attentive tactile discrimination, which did not occur as aconsequence of the mere passive exposure to the same stimuli(Jenkins et al., 1990). However, after learning, sound discriminationoccurs automatically, judging from the fact that after discriminationlearning, the MMN is elicited even when sounds are ignored(Naatanen et al., 1993). Strikingly, the MMN was elicited even insubjects in REM sleep on the third day after the discriminationtraining (Atienza and Cantero, 2001). Hence, this type of learningclearly differs from the short-term adaptation discussed in thebeginning of this review. While that adaptation occurs rapidly andpre-attentively so that some subjects are not even aware of howtheir auditory processing has changed (van Zuijen et al., 2006;Paavilainen et al., 2007), the type of learning introduced above isattention-governed and may take variable periods of time.

In the consolidation of the neural changes subserving learnedstimulus distinctions, sleep seems to play a central role. UsingNaatanen et al.’s (1993) complex sound patterns, Atienza et al.(2004) compared the consolidation of the discrimination trainingeffects between sleep-deprived and normally sleeping groups.Subjects were investigated before and after a 45–90 min trainingsession and again 48 and 72 h post-training. Both rapid and slowchanges were found in discrimination ability and neural responses.Firstly, the discrimination performance improved during thetraining session and became even more accurate 48 h after thetraining. Secondly, the MMN emerged right after the trainingsession and its amplitude continued to grow until 72 h post-training in the normally sleeping group, whereas the sleep-deprived group showed no such increase over time (Fig. 2). TheP3a, in turn, was not present after the training but was found at 48and 72 h post-training but in the normally sleeping group only(Fig. 2). A slow change, emerging 48 h post-training, was also foundin positive responses, presumably corresponding the P2, elicited bystandard and deviant sound patterns. Unlike the MMN and P3a, thechange of P2 was not affected by sleep deprivation. Thus, thesedifferent neural responses reflect different types of learningprocesses and stages. The MMN emerges rapidly after learning,reflecting learning process the consolidation of which depends onsleep. The P3a emerges later but its emergence is also sleep-dependent. Since it is usually elicited by easily-perceivable soundchanges (Escera et al., 2000), it could be assumed that a sufficientlystabilized MMN process should precede the emergence of P3a. Thelearning-related P2 differs from these two components in that itreflects sleep-independent consolidation.

e directly observed from neural responses. The MMN was originally not elicited by a

he left), but after discrimination training, it emerged (Post-training responses, black

eeping subjects (panel A), suggesting consolidation of memory traces. In a sleep-

that sleep deprivation interferes with memory consolidation. Figure adapted from

Page 5: The Adaptive Brain - A Neurophysiological Perspective

T. Kujala, R. Naatanen / Progress in Neurobiology 91 (2010) 55–67 59

3. Long-term auditory experience shapes the brain

Spoken language and music, as inherently related to the humanbrain, are based on the ability to effectively analyze complexauditory information and to form memory representations of therelationships between the elements of the auditory input. Thememory representations formed during long-term learning suchas the acquisition of language or professional-level competency inplaying an instrument have an immense modulating effect onneural networks involved and perception.

3.1. Learning effects on language acquisition

During the normal development, the human brain acquires alarge number of perceptual skills and rules concerning the nativelanguage. In a proficient language user, these skills and rulesoperate in an automatic fashion. For instance, phonetic, lexical,semantic, and syntactic information is processed even when theindividual attends other information than speech (Pulvermullerand Shtyrov, 2006). An important foundation in languageacquisition is to adopt the speech sound system of the nativelanguage. During the development, the human brain becomestuned to the speech sounds of the native language, with thesensitivity to other than native-language speech contrasts beingdiminished (Kuhl, 2004). This is evident when considering howwell two familiar speech sounds can be discriminated, whereastwo unfamiliar ones, even when the acoustic difference is thesame, may not be differentiated (Aaltonen et al., 1997; Sharma andDorman, 1999). The accuracy of making distinctions between thenative-language phonemes is particularly high at the borders ofphonetic categories (Liberman et al., 1957). Within these catego-ries, the speech sound that is considered as an ideal representativeof the given phonetic category is regarded as a phonetic prototype.They serve as perceptual magnets, that is, the nonprototypicmembers of the category are perceived as more similar to theprototype than to each other even when the physical differencesbetween the sounds are equal (Kuhl, 1991).

Naatanen et al. (1997) addressed the neural substrate ofprocessing familiar vs. unfamiliar speech sounds by comparingresponses of subject groups speaking two different but closelyrelated languages, Estonian and Finnish. These subjects werepresented with vowel contrasts /e/ vs. /o/, /o/, and /o/ of which allbelonged to the Estonian and all but /o/ to the Finnish language. InEstonians, all these contrasts elicited MMNs that were increasinglylarger (in this order) as the physical difference between the deviantvowel and the standard /e/ became larger. In Finns, in turn, MMNshowed a similar behavior, except for an amplitude decrement forthe Estonian /o/, which had no representation in their long-termmemory. Further, the neural sources for these responses wereevaluated in Finns with magnetoencephalography (MEG), whichhas a high spatial accuracy in locating activated neural populations(Salmelin and Baillet, 2009). It was found that MMNm (the MEGequivalent of MMN) responses for native-language sounds werepredominant in the left hemisphere, consistent with the postulat-ed primary role of the left hemisphere in speech perception (for areview, see Tervaniemi and Hugdahl, 2003). In a subsequent study(Shestakova et al., 2002), an MMNm subcomponent specific forvowel-category changes was identified, originating from the lefttemporal lobe, distinct from the MMNm activity reflectingaccompanying acoustic change. Furthermore, over and abovethe level of speech sounds, the existence of memory traces forspoken words in the human brain could also be demonstratedusing MEG, EEG and fMRI (Pulvermuller and Shtyrov, 2006;Pulvermuller et al., 2009). Although these word-related memoryactivations are, similar to speech sounds, mainly of left hemi-spheric origin (Shtyrov et al., 2005), they appear to be generated by

neuronal populations distributed over a range of cortical areas(Shtyrov et al., 2004; Pulvermuller et al., 2005). These memoryrepresentations could be probed from inattentive subjects byrecording the MMN process.

These native-language-specific phonetic representations areacquired during the infancy. The brain of a newborn does not havea preference for mother-tongue speech sounds to the others.However, after being exposed to the language used by the parentsand other caretakers, the infant’s brain gradually gets ‘‘specialized’’to the language at hand (Kuhl et al., 1992; Kuhl, 2004), the earlystages of which occur with the aid of prosodic cues (Mehler andChristophe, 1994). By recording head-turning responses (Kuhl,1985), it was shown that already 6-month old infants show the so-called magnet effect, which is stronger for their native languagethan for foreign vowels (Kuhl et al., 1992). American infantsperceived the American English /i/ prototype as identical to itsnon-prototypical variants more often (66.9%) than the Swedish /y/prototype as identical to its non-prototypical variants (50.6%).Swedish infants showed a pattern of perception consistent withthis, perceiving a higher similarity between the Swedish thanEnglish vowels. However, the problem of this approach is that anactive role is required from the infant, whereby the data arecontaminated by fluctuations of attention and interest. Here theinvoluntarily elicited neural responses are of particular help,providing direct information on neural development as well as onthe timing of the emerging perceptual skills (He and Trainor, 2009).

By comparing the MMN responses to native (Finnish /e/ vs. /o/)vs. non-native (Estonian /e/ vs. /o/) vowel changes in a cross-linguistic design, Cheour et al. (1998) demonstrated that the neuralrepresentations of the native phonemes develop at 6–12 monthsafter birth. At 6 months of age, no MMN facilitation was found inFinnish infants in discriminating the native-language vowel /o/from the Estonian /o/, whereas it was present at the age of 12months (even though the acoustic difference is smaller for thelatter contrast). In contrast, Estonian infants at 12 months showedquite similar MMNs to the two contrasts as both vowels belongedto their language. Thus, by the age of one year, the speech system ofthe infant had adapted to the mother tongue and had become lesssensitive to contrasts between speech sounds that do not belong tothis language. However, even the brain of a neonate can extractphonetic information from the wide physical variation included inspeech. MMN responses were elicited in neonates by phonemechanges irrespective of whether or not irrelevant speaker variationwas present in speech stimulation (Dehaene-Lambertz and Pena,2001). These results suggest that the different stages and theirprogress in speech-perception development can be identified withMMN.

Further, the neural representations of phonetic categories,established in the infancy for the native language, may bemodulated during the acquisition of a foreign language, as shownboth by studies with discrimination-training exercises (Menninget al., 2002; Tremblay et al., 1997, 1998, 2001; Tremblay and Kraus,2002; Reinke et al., 2003) and by studies on the effects of exposureto natural language environments (Winkler et al., 1999). Discrimi-nation exercises aiming at improving the identification ofpreviously unfamiliar speech sounds modulates their receptionin the central auditory system. For example, Tremblay et al. (2001)and Tremblay and Kraus (2002) recorded neural responses elicitedby /ba/ with various voice-onset times before and after identifica-tion training of the different exemplars. This training resulted bothin improved identification of the stimuli and enhancements of theN1-P2 peak-to-peak amplitudes (Tremblay et al., 2001) and P1, N1,and P2 amplitudes (Tremblay and Kraus, 2002). Consistently,Reinke et al. (2003) reported learning-induced identificationimprovement and enhanced P2 amplitude and shortened N1and P2 latencies for unfamiliar vowels, which were created by

Page 6: The Adaptive Brain - A Neurophysiological Perspective

T. Kujala, R. Naatanen / Progress in Neurobiology 91 (2010) 55–6760

summing two different vowels. These results show that learning toidentify speech sounds facilitates their cortical representationsand speeds up their neural processing.

Winkler et al. (1999) determined the effect of foreign-languageacquisition in an authentic language environment on phoneme-category representations in Hungarian adult immigrants living inFinland since they were 13–32 years old. They compared thediscrimination of the Finnish phonetic contrast /e/ vs. /ae/ betweenthese subjects, Hungarians with no Finnish command, and Finns. Itwas found that these vowels, not present in the dialect of theHungarian subjects, were poorly identified by Hungarians withoutFinnish command, whereas the identification rate was high in thetwo other groups. Consistent with this, in these two groups, thiscontrast elicited very similar MMNs, whereas MMN was absent inHungarians not knowing Finnish (Fig. 3). Thus, the phonemerepresentation system can be adapted to foreign speech soundsafter extensive exposure to the language even after the childhood.

In fact, language has such a powerful effect on brain processesthat it may even modify basic auditory non-linguistic functions.Tervaniemi et al. (2006) compared duration and frequencydiscrimination of harmonical tones between quantity language

Fig. 3. Language learning modifies phoneme representations of the auditory system. The

Hungarian, was tested by recording MMN and behavioural responses from Finns, Hungar

The contrast elicited an MMN in Finns and Hungarians with a good command of Finnish b

showed corroborating results, with the lowest hit rate and longest RT in the Hungaria

speakers (Finns) and non-quantity language speakers (Germans).In quantity languages like Finnish, word meaning may be changedby only changing the duration of a speech sound, for instance, /tuli/(fire), /tuuli/ (wind), /tulli/ (customs). Whereas no group differ-ences were found in MMNs or just-noticeable-difference discrimi-nation for sound-frequency differences, the MMNs anddiscrimination performance were enhanced for duration differ-ences in Finns compared to Germans. These results show thatexperience in discriminating a particular speech-sound feature (inthis case duration) generalizes to perceiving accurately such afeature in other types of sounds, too. This type of generalizationcould be possible since the duration cue has to be perceived in veryvariable circumstances, that is, in the contexts of speechproduction rates greatly varying from one speaker to another.

As evident above, intensive language training modulatesauditory perception and neural representation of phonetic catego-ries. Further, the native-language phoneme representations pre-dominantly involve the neural networks of the left temporal areas(Naatanen et al., 1997). However, when new phonetic categories areacquired, are they represented in the language-dominant or non-dominant hemisphere? This issue was investigated by A. Kujala et al.

discrimination of a Finnish phoneme contrast (/e/ vs. /ae/), which does not exist in

ians with no Finnish command, and in Hungarians with a good command of Finnish.

ut not in Hungarians not knowing Finnish. The behavioural discrimination condition

ns not knowing Finnish. Figure adapted from Winkler et al., 1999.

Page 7: The Adaptive Brain - A Neurophysiological Perspective

Fig. 4. Brain tunes into the musician’s principal instrument. Left and right temporal

lobe N1m responses were recorded for trumpet and violin tones from trumpeters

and violinists. In the upper panel, left- and right-hemisphere response strengths

from two representative subjects for trumpet (thin line) and violin (thick line)

sounds. In the lower panel, average data (left and right hemisphere responses

combined) from these two groups. The results show increased neural responses for

the principal instrument suggesting selective tuning into the sound of training.

Figure adapted from Pantev et al., 2001.

T. Kujala, R. Naatanen / Progress in Neurobiology 91 (2010) 55–67 61

(2003a) by determining the representations of Morse codes beforeand after a Morse-code training course. MMNm was recorded, bothto syllables and to the corresponding Morse-coded sound contrastsbefore and after an intensive 3-month training course. It was foundthat before the training period, the Morse-coded syllables were firstprocessed primarily in the non-language dominant hemisphere,whereas after the learning period, the cerebral dominance in Morse-code discrimination was shifted to the subjects’ language-dominanthemisphere. These results imply that after learning to associate non-speech sounds with speech, sound representations become estab-lished in the neural network specialized in speech processing.

This Morse-code discrimination training also modulated theattention-switching sensitivity in the trainees. Uther et al. (2006)recorded responses for frequency, duration, and timing changes ofa sinusoidal repetitive sound before, during (halfway through), andafter the Morse code training course. It was found that the latencyof the P3a elicited by these deviant sounds became shorter duringthe training, indicating that training speeded up the attention-switching mechanism. Furthermore, the frequency change eliciteda larger P3a after than before the training period. This could resultfrom the fact that the standard stimulus had the same frequency asthe Morse-coded stimuli used in the training, whereby the traineeshad become very sensitive to deviations from this highly trainedcarrier frequency.

3.2. Plasticity induced by musical training

Like language learning in an authentic environment, playing aninstrument at a professional level also involves extensive training.The N1, MMN, and P3a responses, recorded for a variety of auditorystimuli, indicated superior auditory cortical processes in musicianscompared with those in non-musicians. For instance, the magneticN1 (N1m), which reflects the cortical representation area of asound (Naatanen, 1992), is larger in amplitude for piano tones thanfor sinusoidal tones in musicians, whereas the N1ms for these twotone types did not differ in amplitude from one another in non-musicians (Pantev et al., 1998). Thus, intensive training with musicenhances the neural processing of sounds specific to music.Moreover, the type of musical training has specific effects: Pantevet al. (2001), comparing the N1ms elicited by trumpet and violinsounds in players of these two instruments with one another,found that the timbre of the instrument selectively modulatessound processing in the auditory cortex. In both groups, the N1mwas larger for the timbres of the instrument of training (Fig. 4).Thus, musical training enhances neural processing in the auditorycortex and, further, the neurons become selectively tuned to thetimbre of the trained instrument.

In addition, the type of musical training may also modulateattention networks of the brain. While focusing on one musician,the conductor has to simultaneously monitor the entire orchestra.One could, therefore, expect that this type of training facilitates theattention-shifting mechanisms of the brain. This is exactly whatwas found by recording neural responses of conductors and controlsubjects to sounds with different spatial origins (Nager et al.,2003). The conductors’ P3a responses, reflecting the efficiency ofattention-switching (Escera et al., 2000), were stronger than thoseof control subjects for occasional sound changes in unattendedspatial locations (Nager et al., 2003).

Moreover, the cortical accuracy in discriminating chords is alsobetter in musicians than non-musicians. When slightly mistunedchords are presented in chord sequences, they elicit larger-amplitude MMNs in musicians than non-musicians (Koelsch et al.,1999). In addition, stronger MMNms in musicians than non-musicians were found for melodic-contour and interval-structuredifferences (Fujioka et al., 2004). Furthermore, musical experiencealso affects the efficiency in learning melodic patterns (Tervaniemi

et al., 2001). Musicians who primarily play without a score werefound to be superior to those playing with a score in learning todifferentiate between small differences in melodic patterns. Thislearning effect was evident in their enhanced MMNs after thepattern-discrimination session.

4. Brain plasticity in recovery and remediation

Sensory-information processing may be impaired by severaldifferent types of deficits or injuries. Such impairments may becaused by local lesions to neural tissue subserving these functions,by lesions to the receptors of sensory input, or by genetic factorsleading to abnormalities in the sensory-neural apparatus underly-ing perception. Both the neural basis of these deficits as well asplastic changes resulting from recovery and intervention can beinvestigated with neurophysiological approaches.

4.1. Reorganization after brain lesion

A large body of evidence suggests that after brain injury, themost extensive plastic changes take place during the first 3 months(Heiss et al., 1999; Cao et al., 1999). While it would be important todetermine the loss and subsequent recovery of perceptual abilitiesand neural function soon after the damage, patients are oftenunable to cooperate or understand instructions particularly if theneural network involved in speech perception is affected.

So far, the majority of neurophysiological studies on compen-satory mechanisms in aphasia utilized attentive paradigms andinvestigated patients whose recovery had already reached aplateau (see, for example, Hagoort et al., 2003; Pulvermuller et al.,2005; Laganaro et al., 2008). In order to determine damaged

Page 8: The Adaptive Brain - A Neurophysiological Perspective

T. Kujala, R. Naatanen / Progress in Neurobiology 91 (2010) 55–6762

functions and the progresses of recovery from the early stagesinvolving diaschisis onwards, responses that can be acquiredwithout patient’s active role are helpful. Some scholars suggestthat it is important to determine deficits in basic auditoryprocessing in aphasia since they may underlie impairments inspeech perception (Pettigrew et al., 2004). For example, in a studyemploying a wide variety of psychophysical tests in determininghow they predict verbal comprehension skills in left-hemispherelesioned patients, frequency discrimination, frequency-sweepdiscrimination, and frequency uncertainty effect in tone-in-noisedetection were the best predictors (Divenyi and Robinson, 1989)Neurophysiological responses requiring no active role from thepatient are attractive for investigating such basic auditoryprocesses in aphasic patients. Studies with this approach havereported diminished sound-frequency and duration MMNs (Ilvo-nen et al., 2001, 2003) and different MMN patterns for speechsound vs. non-speech stimuli (Aaltonen et al., 1993; Wertz et al.,1998; Ilvonen et al., 2004) in aphasic patients. Furthermore, thepresence of MMN for syllable stimuli was significantly associatedwith auditory comprehension performance in aphasic patients(Auther et al., 2000).

Ilvonen et al. (2003) used MMN for determining perceptuallosses and recovery in a follow-up study of aphasic left-hemisphere stroke patients. The recordings were carried out at4 and 10 days, and then at 3 and 6 months after stroke onset.Language-comprehension tests (Boston Diagnostic Aphasia Exam-ination and Token test) were also administered in other sessionsexcept for the first one, during which the patients could notcooperate. In the first 2 recording sessions, MMN was quite small inamplitude for sound duration and frequency changes, but grewthereafter, reaching a normal-like amplitude at 3 monthspoststroke. Furthermore, there was a significant correlationbetween the MMN-amplitude increase for sound-durationchanges and the improvement in Boston Diagnostic AphasiaExamination test scores from 10 days to 3 months.

Sarkamo et al. (2008, in press) investigated the effect of musicand audio book listening on the recovery of left- or right-hemisphere stroke patients in follow-up MMNm recordings at 3and 6 months poststroke. Both the growth of the MMNm elicitedby sound duration and frequency changes (Sarkamo et al., in press)and improvement in neuropsychological tests (Sarkamo et al.,2008) suggested greater recovery in these groups than in thecontrol group with no intervention. Music therapy facilitated morethe recovery of verbal memory and focused attention than theaudio book intervention or lack of intervention (Sarkamo et al.,2008). Furthermore, there was a significant correlation betweenthe increment of the right-hemisphere frequency MMNm changesand improvement in delayed story recall and mental subtraction ofnumbers (Sarkamo et al., in press).

These results suggest that listening to complex auditorystimulation, such as stories or music, enhances the neural changesassociated with recovery from stroke. These changes might bebased on structural and molecular plasticity associated withenriched stimulation. Enriched environment was shown toimprove motor and cognitive recovery, decrease infarct volume,and increase neurogenesis in rodents with an experimentally-induced stroke (Nithianantharajah and Hannan, 2006).

4.2. Recovery from sensory deprivation

The neural basis of perception faces a different type of achallenge when sensory receptors are damaged, e.g., in deafness orblindness. In this case, the sensory-specific brain areas aredeprived of their normal input, which changes the functional roleof these areas at least when the deprivation starts early in infancy(for reviews, see Kujala et al., 2000; Rauschecker, 2002; Bavelier

and Neville, 2002). The auditory cortex, deprived of acoustic input,may recover even after a long period of deafness after cochlearimplantation (CI), as suggested by the progressive improvement ofperceptual auditory skills in cochlear-implant patients (e.g., Laskeet al., 2009). Cortical speech-sound discrimination of such patientswas followed-up by recording MMN to vowel changes at 1 and at2–3 years after the implantation (Lonka et al., 2004). MMN waspresent in the 1-year recording, and its amplitude was increasedboth for a small and large vowel contrast between the tworecording times, suggesting a gradual improvement in voweldiscrimination over time in these patients. Consistent with this,there was also an improvement in the speech-discrimination testscores of these individuals. This result suggests progressive plasticchanges over a long time period in the auditory cortex, reflectingre-learning to process auditory input. Furthermore, a goodcommunication competence with the CI is associated with a largeand a poor one with a small MMN amplitude (Kraus et al., 1993;Ponton et al., 2000; Groenen et al., 1996; Singh et al., 2004). Thus,with MMN, it is possible to determine both sound-discriminationaccuracy in CI users and to follow up the progress of neural plasticchanges after the implantation.

4.3. Plasticity of brains with developmental deficits

Perceptual and language deficits may also be caused byabnormal central nervous system development due to geneticfactors. For alleviating these types of developmental disorders,carefully designed intervention programs are necessary. In fact, alarge proportion of children suffer from some type of develop-mental language deficit, the estimates being, for example, forchildren with specific language impairment and children withdyslexia 3–10% of child population (Bishop and Snowling, 2004).The high prevalence and the devastating effects of these disorderson the individual’s success in society and academia, and therebyeven on self esteem, call for effective early identification andremediation methods for minimizing the problems with preven-tion. Fortunately, recent research has shed light on the types ofneural processing impairments of children with language deficits.Abnormal MMNs obtained from children with specific language/learning impairment suggest that their central auditory system isdeficient in distinguishing consonants (Kraus et al., 1996; Bradlowet al., 1999; Shafer et al., 2005) and pitch differences (Korpilahtiand Lang, 1994), which also applies to dyslexia (Baldeweg et al.,1999; Kujala et al., 2003a,b; Renvall and Hari, 2003; Lachmannet al., 2005). Furthermore, diminished MMNs at the age of 5months were found in children who at the ages of 1–2 years werediagnosed as language impaired (Weber et al., 2005). In addition,abnormal auditory processing was evident in 6-month old infantsat familial risk for dyslexia, judging from their attenuated MMNs tospeech-sound changes over the left hemisphere (Leppanen et al.,2002).

Fortunately, the aberrant neural substrate of language deficitscan be ameliorated with appropriate intervention. Languageimpaired children at the age of 5–6 years benefited from anintervention program including a variety of language and speechexercises as compared with a group having motor-exercisetraining (Pihko et al., 2007). In addition, accompanying MEGrecordings indicated that language and speech training alsochanged the neural processing of speech sounds in both temporallobes. The P1m responses to syllables became stronger in bothhemispheres after training, suggesting the strengthening ofspeech-sound representations. In addition, the MMNm becamestronger for syllable changes in the left hemisphere.

Moreover, intervention-induced amelioration of readingdifficulties, concurrently with neural plastic changes, were alsoshown in dyslexic first-grade children (Kujala et al., 2001b).

Page 9: The Adaptive Brain - A Neurophysiological Perspective

Fig. 5. Audiovisual training enhances auditory cortical discrimination accuracy and concurrently improves reading skills of first-grade dyslexic children. The training group

did computerized audiovisual exercises requiring matching of visually- and auditorily-presented elements for 7 weeks. Both the MMNs and reading-skill test scores, which

originally did not differ between this and a matched control dyslexic group, significantly changed after the training period. The MMN for tone-order reversals became larger in

the training than control group. Concurrently, the training group became more accurate in reading and tended to read faster than the control group. Figure adapted from

Kujala et al., 2001b.

T. Kujala, R. Naatanen / Progress in Neurobiology 91 (2010) 55–67 63

Dyslexic 7-year old children used audiovisual training programwith no linguistic items (Audilex program; Karma, 1999) for 7weeks, and reading skills as well as MMN were assessed beforeand after this training period. The audiovisual exercises includedmatching of sounds with different frequencies, durations, orintensities with visually-presented rectangles varying in theirvertical position, length, or thickness on the computer screen. Insome exercises, the child had to decide from two rectanglepatterns which of them corresponded to the sound pattern theyheard. In the rest of the exercises, one visual pattern waspresented and when the sound pattern was played, the task was topress the space bar when the last sound was expected to occur.Controls were matched dyslexic children receiving no training.Tone-order reversals were used in the MMN experiment, sincedyslexic individuals have pronounced problems in temporaldiscrimination, such as sound-order judgement (for a review, seeFarmer and Klein, 1995). After the training period, the traininggroup read words more correctly and somewhat faster than didthe control group (Fig. 5). Furthermore, the MMN amplitude fortone-order reversals was increased in the training but not in thecontrol group (Fig. 5). Moreover, there was a significantcorrelation between the MMN amplitude change and the changein the reading-skill scores between the first and second recordingsessions. These results show that reading deficits can beameliorated with audiovisual training and, further, that theconcurrent plastic changes can be determined by recording MMN.Furthermore, the presence of these positive training results withno linguistic stimuli suggests that impaired perceptual processesother than deficient phonology may also contribute to readingimpairments. This is consistent with the observations thatdyslexic children have also other than phonological deficits,which has led to the suggestion that the underlying cause ofdyslexia might be the impaired discrimination of rapidly changingor brief non-speech sounds (Farmer and Klein, 1995; Tallal, 2004).However, the stimuli of Kujala et al. (2001b) were neither rapidlypresented (200 ms onset-to-onset pace) nor included rapidtransitions. Thus, the results do not in essence support the theoryof Tallal (2004), whereas, however, they indicate that dyslexia alsoat least partly involves other problems than those directly relatedto speech processing. Currently, the view of several scholars is thatthere are subgroups of dyslexic individuals having differentprofiles of deficits (e.g., Ramus et al., 2003).

Evidently, it is possible to identify abnormalities in the neuralbasis of auditory and speech discrimination at very young ages,even in the infancy. Moreover, the neural responses indicatingthese abnormalities also serve as a means to monitor the

effectiveness of intervention. These results provide remarkableprospects for detecting deficits in speech development in at-riskinfants and children and in evaluating the effectiveness of veryearly intervention. In this way, for instance, deficits in learning tounderstand and produce speech, could be tackled with. During theso called critical or sensitive period, the neural substrate for theskill to be acquired is most malleable, and missing the correct timeto learn will result in the poor mastering of that skill (e.g., Ruben,1999; Kaas, 2001). Therefore, it would be important to apply theinvoluntarily-elicited neural responses for determining the typesof speech deficits and subsequently, most effective means toameliorate them at this early age.

5. Concluding remarks

The possibility to observe neural plasticity directly from brainactivity without the subject’s behavioural response has manyadvantages. This approach diminishes sources of artefacts such asthe effects of the subject’s motivation on the results. The functionalroles of the early involuntary neural responses in informationprocessing are also fairly well known (Naatanen and Picton, 1987;Escera et al., 2000; Naatanen et al., 2007; Kujala et al., 2007). Thus,the modulation of these responses during learning or recoveryindicates functions that have changed.

As presented in the afore-going, plastic changes are reflected bythe enhancements of the P1, N1, P2, and MMN responses, with theP1, N1 and P2 reflecting such changes in the afferent sensorysystem underlying perception and the MMN changes in themechanisms of sensory discrimination. The N1 enhancementmight be caused by further differentiation of sensory receptivefields, resulting in an increasingly fine-grained cortical mappingand representation of different stimulus features received by theafferent sensory system (Naatanen and Alho, 1997). These changes,in turn, account for increasingly fine-grained stimulus discrimina-tion, reflected by enhanced MMN responses. Stimulus changes aremore accurately detected when the standard stimulus represen-tation is sharper whereby the deviant stimulus is distinguishedmore accurately. The improved stimulus-change detection accu-racy may lead to higher sensitivity to stimulus changes, which inturn is reflected in increased P3a amplitudes suggesting enhancedattention switching (e.g., Uther et al., 2006). Since scalp-recordedneurophysiological responses reflect synchronous activity ofneural populations, the training-induced enhancement of theresponses is probably based on the involvement of new neuronsdue to learning. This is supported by the recordings of intracranialneural responses in animals which have shown that after sensory

Page 10: The Adaptive Brain - A Neurophysiological Perspective

T. Kujala, R. Naatanen / Progress in Neurobiology 91 (2010) 55–6764

discrimination training, a larger number of neurons in the sensoryrepresentation area become responsive to the stimuli used intraining (Jenkins et al., 1990; Recanzone et al., 1992).

Neural responses with which cortical sound representationscan be investigated even without subject’s attention haveprovided novel understanding of neuroplasticity in several ways,as shown by the current review. Since these responses enable oneto disentangle the automatic stages of perception and the furtherprocessing stages associated with attention and consciousnessfrom one another, it is possible to determine, for instance, thestability of the learned distinctions and mechanisms of rapidadaptation vs. long-term plasticity. Both short- and long-termplastic changes occur in the neural processes after sensory input.Short-term adaptation differs from long-term changes both by itstime-course and in that it requires no attention to stimuli,whereas for inducing long-term plastic changes, one has toattentively try to identify or differentiate stimuli. Follow-up ofneural responses after discrimination learning has revealed thatneural plastic processes go on for several days after thediscrimination training session. There are at least two stages inthis learning-associated neuroplasticity. First, concomitantlywith the behavioural discrimination improvement, an MMNemerges (Naatanen et al., 1993; Tervaniemi et al., 2001; Atienzaet al., 2002, 2004). This stage in learning might be associated withrapid modulation of the receptive-fields of cortical neurons(Atienza et al., 2002; Xerri et al., 1996; Gilbert, 1994). It has beenshown that soon after monkey has learned to discriminate targetsounds, neurons in A1 increase responsiveness and showdifferentiation to frequency ranges, which indicates shifts ofindividual neurons’ responsiveness (Blake et al., 2002). Second, afollow-up of MMN showed further amplitude increments still 48and 72 h post-training (Atienza et al., 2004). These results suggestconsolidation of the learned distinction between auditory stimuli.At the neural level, these changes might correspond modulationsof neural representations (Recanzone et al., 1993; Atienza et al.,2002).

These neural responses also enable one to assess how long-term experience or expertise modulates neural representations ofacoustic features. For example, as the comparisons of neuralresponses between musicians using different instrumentsshowed, the neural networks become selectively tuned to soundsincluded in training (Pantev et al., 2001). The same was evident inlanguage-learning studies: the brain adapts to the surroundinglanguage environment. Phonetic representations, establishedduring the infancy (Cheour et al., 1998), may, however, assimilatenew phonetic representations during foreign-language learningeven after the childhood (Winkler et al., 1999). Thus, theseresponses serve as indicators of the specificity of learning effects.They also enable the comparison of the learning process indifferent ages.

Recording responses generated even in passive subjects givesus access to brain processes of patients having severe commu-nication problems and those of infants, whose perceptualabilities are very difficult to reliably address with any othermethod. In brain-injured patients with communication pro-blems, for instance, the early identification of impaired neuralfunctions could help to determine which perceptual functionsshould be rehabilitated before the patient has recovered to anextent that neuropsychological tests can reliably be applied. Thisway the rehabilitation could be started when the neural tissuestill is highly plastic after the injury, which is within threemonths from, e.g., stroke onset (Cao et al., 1999; Heiss et al.,1999). However, in order to apply this approach to investigationsof individual patients or subjects, the reliability of recording andanalyzing neural responses in single subjects should be notablyimproved. For example, for acquiring the MMN, which has a

small amplitude especially for small sound differences, a largenumber of trial averages is needed for extracting it from thelarge-amplitude background EEG activity and noise (e.g., muscleartefacts or eye blinks). This may make the recording sessionsintolerably long for patients or children, especially if one wishesto get a comprehensive picture on the discrimination of multiplesound differences. However, recent paradigm improvements arepromising, providing the possibility to record MMNs for even fivedifferent sound attributes within the time in which MMN waspreviously acquired for one sound attribute only (Naatanen et al.,2004).

While our knowledge on the adult human brain is alreadyextensive due to the great number of feasible research approaches,developmental neuroscience still is an evolving research areacurrently making dramatic progress. Evoked neural responses havenotably contributed to this progress. Development involvesconstant, at times quite rapid, changes in brain processes. Thepossibility to observe them directly from brain responses hasduring the recent years shed light to range of issues, includingquestions like the time-course of native-language acquisition(Cheour et al., 1998). The brain processes of even younger infants,fetuses, can be investigated due to recent methodologicaldevelopments (Huotilainen et al., 2005). This opens unforeseenperspectives to the earliest stages of auditory perception. Thepossibility to identify, for instance, speech-sound discriminationdeficits in early infancy would enable one to select appropriatesound material for early intervention. As the young, rapidlydeveloping brain is highly plastic, this approach could haveremarkable effects in preventing later language problems.

The present review shows that by observing certain neuralresponses, we can answer some fundamental questions concerningthe human brain. The evolution of the processes leading toperception can be identified with an accuracy on the order ofmilliseconds. By identifying these processes and their relationshipswith experimental manipulations we can find answers toquestions such as ‘‘what’’ happens during perception and ‘‘howand when’’ it happens. Other methods (e.g., positron emissiontomography, functional magnetic resonance imaging), albeitaccurate in locating activated neural structures, are poor for thistype of questions due to their low temporal resolution. Theneurophysiological response has a rich temporal structuresometimes indicating profound activity changes and differencesin cognitively-related dynamics changing at a pace of 100 ms oreven faster. The haemodynamic brain response, in contrast,provides a temporal average over seconds, thus making itimpossible, for example, to discriminate the early effects ofMMN from the fundamentally different ones reflected by the lateP3. Furthermore, the data acquisition is expensive with thesemethods, and the machinery is immobile. The EEG, in turn, isinexpensive, can be carried, e.g., to bed-side, is subject-friendly,non-invasive and safe, its use involves no ethical problems, and itcan be easily used for studying extensive subject and patientpopulations. When there is a need to locate the neurophysiologicalactivity in more detail, the MEG is helpful (Salmelin and Baillet,2009). Its advantage compared to EEG is also the rapid preparationtime for experiments. When these methodological properties ofneurophysiology are combined with optimal paradigmaticapproaches, even the secrets of some of the most challengingareas of human cognition, that of an infant, or a sleeping, anunconscious, or otherwise uninteractive individual, can beunveiled.

Acknowledgements

This work was supported by the Academy of Finland (grantnumbers 128840 and 122745) and the University of Helsinki.

Page 11: The Adaptive Brain - A Neurophysiological Perspective

T. Kujala, R. Naatanen / Progress in Neurobiology 91 (2010) 55–67 65

References

Aaltonen, O., Eerola, O., Hellstrom, A., Uusipaikka, E., Lang, A.H., 1997. Perceptualmagnet effect in the light of behavioral and psychophysiological data. J. Acoust.Soc. Am. 101, 1090–1105.

Aaltonen, O., Tuomainen, J., Laine, M., Niemi, P., 1993. Cortical differences in tonalversus vowel processing as revealed by an ERP component called mismatchnegativity (MMN). Brain Lang. 44, 139–152.

Alho, K., 1995. Cerebral generators of mismatch negativity (MMN) and its magneticcounterpart (MMNm) elicited by sound changes. Ear Hear. 16, 38–51.

Amenedo, E., Escera, C., 2000. The accuracy of sound duration representation in thehuman brain determines the accuracy of behavioural perception. Eur. J. Neu-rosci. 12, 2570–2574.

Atienza, M., Cantero, J.L., 2001. Complex sound processing during human REM sleepby recovering information from long-term memory as revealed by the mis-match negativity. Brain Res. 901, 151–160.

Atienza, M., Cantero, J.L., Dominguez-Marin, E., 2002. The time course of neuralchanges underlying auditory perceptual learning. Learn. Mem. 9, 138–150.

Atienza, M., Cantero, J.L., Escera, C., 2001. Auditory information processing duringhuman sleep as revealed by event-related potentials. Clin. Neurophysiol. 112,2031–2045.

Atienza, M., Cantero, J.L., Stickgold, R., 2004. Post-training sleep enhances automa-ticity in perceptual discrimination. J. Cog. Neurosci. 16, 53–64.

Auther, L.L., Wertz, R.T., Miller, T.A., Kirshner, H.S., 2000. Relationships among themismatch negativity (MMN) response, auditory comprehension, and site oflesion in aphasic adults. Aphasiology 14, 461–470.

Baldeweg, T., Richardson, A., Watkins, S., Foale, C., Gruzelier, J., 1999. Impairedauditory frequency discrimination in dyslexia detected with mismatch evokedpotentials. Ann. Neurol. 45, 495–503.

Bavelier, D., Neville, H.J., 2002. Cross-modal plasticity: where and how? Nature Rev.Neurosci. 3, 443–452.

Bendixen, A., Roeber, U., Schroger, E., 2007. Regularity extraction and application indynamic auditory stimulus sequences. J. Cog. Neurosci. 19, 1664–1677.

Bendixen, A., Prinz, W., Horvath, J., Trujillo-Barreto, N.J., Schroger, E., 2008.Rapid extraction of auditory feature contingencies. NeuroImage 41, 1111–1119.

Bishop, D.V.M., Snowling, M.J., 2004. Developmental dyslexia and specific languageimpairment: same or different? Psychol. Bull. 130, 858–886.

Blake, D.T., Strata, F., Churchland, A.K., Merzenich, M.M., 2002. Neural correlates ofinstrumental learning in primary auditory cortex. Proc. Natl. Acad. Sci. U.S.A. 99,10114–10119.

Bradlow, A.R., Kraus, N., Nicol, T.G., McGee, T.J., Cunningham, J., Zecker, S.G., Carrell,T.D., 1999. Effects of lengthened formant transition duration on discriminationand neural representation of synthetic CV syllables by normal and learning-disabled children. J. Acoust. Soc. Am. 106, 2086–2096.

Bottcher-Gandor, C., Ullsperger, P., 1992. Mismatch negativity in event-relatedpotentials of auditory stimuli as a function of varying interstimulus interval.Psychophysiol. 29, 546–550.

Cao, Y., Vikingstad, E.M., George, K.P., Johnson, A.F., Welch, K.M.A., 1999. Corticallanguage activation in stroke patients recovering from aphasia with functionalMRI. Stroke 30, 2331–2340.

Cheour, M., Ceponiene, R., Lehtokoski, A., Luuk, A., Allik, J., Alho, K., Naatanen, R.,1998. Development of language-specific phoneme representations in the infantbrain. Nature Neurosci. 1, 351–353.

Cowan, N., Winkler, I., Teder, W., Naatanen, R., 1993. Memory prerequisites of themismatch negativity in the auditory event-related potential (ERP). J. Exp.Psychol. Human Percept. Perform. 19, 909–921.

Dehaene-Lambertz, G., Pena, M., 2001. Electrophysiological evidence for automaticphonetic processing in neonates. NeuroReport 12, 3155–3158.

Divenyi, P.L., Robinson, A.J., 1989. Nonlinguistic auditory capabilities in aphasia.Brain Lang. 37, 290–326.

Donchin, E., Coles, M.G.H., 1988. Is the P300 component a manifestation of contextupdating? Behav. Brain Sci. 11, 355–425.

Escera, C., Alho, K., Schroger, E., Winkler, I., 2000. Involuntary attention anddistractibility as evaluated with event-related brain potentials. Audiol.Neuro-otol. 5, 151–166.

Farmer, M.E., Klein, R.M., 1995. The evidence for a temporal processing deficit linkedto dyslexia: a review. Psychon. Bull. Rev. 2, 460–493.

Fujioka, T., Trainor, L.J., Ross, B., Kakigi, R., Pantev, C., 2004. Musical trainingenhances automatic encoding of melodic contour and interval structure. J.Cog. Neurosci. 16, 1010–1021.

Giard, M.H., Perrin, F., Echallier, J.F., Thevenet, M., Froment, J.C., Pernier, J., 1994.Dissociation of temporal and frontal components in the human auditory N1wave: a scalp current density and dipole model analysis. Electroen. Clin.Neurophysiol. 92, 238–252.

Gilbert, C.D., 1994. Early perceptual learning. Proc. Natl. Acad. Sci. 91, 1195–1197.Gottselig, J.M., Hofer-Tinguely, G., Borbely, A.A., Regel, S.J., Landolt, H.-P., Retey, J.V.,

Achermann, P., 2004. Sleep and rest facilitate auditory learning. Neuroscience127, 557–561.

Groenen, P., Snik, A., van den Broek, P., 1996. On the clinical relevance of mismatchnegativity: results from subjects with normal hearing and cochlear implantusers. Audiol. Neuro-otol. 1, 112–124.

Gumenyuk, V., Korzyukov, O., Escera, C., Hamalainen, M., Huotilainen, M., Hayrinen,T., Oksanen, H., Naatanen, R., von Wendt, L., Alho, K., 2005. Electrophysiologicalevidence of enhanced distractibility in ADHD children. Neurosci. Lett. 374, 212–217.

Hagoort, P., Wassenaar, M., Brown, C., 2003. Real-time semantic compensation inpatients with agrammatic comprehension: electrophysiological evidence formultiple-route plasticity. Proc. Natl. Acad. Sci. 100, 4340–4345.

Hari, R., Levanen, S., Raij, T., 2000. Timing of human cortical activation sequencesduring cognition: role of MEG. Trends Cog. Sci. 4, 455–462.

He, C., Trainor, L.J., 2009. Finding the pitch of the missing fundamental in infants. J.Neurosci. 29, 7718–8822.

Hebb, D.O., 1949. The Organization of Behavior: A Neuropsychological Theory.Wiley, New York.

Heiss, W.D., Kessler, J., Thiel, A., Ghaemi, M., Karbe, H., 1999. Differential capacity ofleft and right hemispheric areas for compensation of poststroke aphasia. Ann.Neurol. 45, 430–438.

Huotilainen, M., Kujala, A., Hotakainen, M., Parkkonen, L., Taulu, S., Simola, J.,Nenonen, J., Karjalainen, M., Naatanen, R., 2005. Short-term memory functionsof the human fetus recorded with magnetoencephalography. NeuroReport 16,81–84.

Ilvonen, T.-M., Kujala, T., Kiesilainen, A., Salonen, O., Kozou, H., Pekkonen, E., Roine,R.O., Kaste, M., Naatanen, R., 2003. Auditory discrimination after left hemi-sphere stroke: an MMN follow-up study. Stroke 34, 1746–1751.

Ilvonen, T.-M., Kujala, T., Kozou, H., Kiesilainen, A., Salonen, O., Alku, P., Naatanen, R.,2004. The processing of speech and non-speech sounds in aphasic patients asreflected by the mismatch negativity (MMN). Neurosci. Lett. 366, 235–240.

Ilvonen, T.-M., Kujala, T., Tervaniemi, M., Salonen, O., Naatanen, R., Pekkonen, E.,2001. The processing of sound duration after left hemisphere stroke: event-related potential and behavioral evidence. Psychophysiology 38, 622–628.

Javitt, D.C., Steinschneider, M., Schroeder, C.E., Arezzo, J.C., 1996. Role of cortical N-methyl-D-aspartate receptors in auditory sensory memory and mismatch neg-ativity generation: implications for schizophrenia. Proc. Natl. Acad. Sci. U.S.A.93, 11962–11967.

Jenkins, W.M., Merzenich, M.M., Ochs, M.T., Allard, T., Guic-Robles, E., 1990.Functional reorganization of primary somatosensory cortex in adult owl mon-keys after behaviourally controlled tactile stimulation. J. Neurophysiol. 63, 82–104.

Jancke, L., Gaab, N., Wuestenberg, T., Scheich, H., Heinze, H.-J., 2001. Short-termfunctional plasticity in the human auditory cortex: an fMRI study. Cognit. BrainRes. 12, 479–485.

Kaas, J.H., 2001. The Mutable Brain. Harwood Academic Publishers, Amsterdam, TheNetherlands.

Kaas, J.H., Krubitzer, L.A., Chino, Y.M., Langston, A.L., Polley, E.H., Blair, N., 1990.Reorganization of retinotopic cortical maps in adult mammals after lesions ofthe retina. Science 248, 229–231.

Karma, K., 1999. Auditory structuring in dyslexia. In: McKevitt, P., Mulvihill, C.,Nuallain, S.O., Riordan, C. (Eds.), Proceedings of the 8th International Workshopon the Cognitive Science of Natural Language Processing (CSNLP-8). August 9–11, 1999, Galway, Ireland. John Benjamins, Amsterdam/Philadelphia.

Koelsch, S., Schroger, E., Tervaniemi, M., 1999. Superior pre-attentive auditoryprocessing in musicians. NeuroReport 10, 1309–1313.

Korpilahti, P., Lang, H.A., 1994. Auditory ERP components and mismatch negativityin dysphasic children. Electroen. Clin. Neurophysiol. 91, 256–264.

Kraus, N., McGee, T.J., Carrell, T.D., King, C., Tremblay, K., 1995. Central auditorysystem plasticity associated with speech discrimination learning. J. Cog. Neu-rosci. 7, 27–34.

Kraus, N., McGee, T.J., Carrell, T.D., Zecker, S.G., Nicol, T.G., Koch, D.B., 1996. Auditoryneurophysiologic responses and discrimination deficits in children with learn-ing problems. Science 273, 971–973.

Kraus, N., Micco, A.G., Koch, D.B., McGee, T., Carrell, T., Sharma, A., Wiet, R.J.,Weingarten, C.Z., 1993. The mismatch negativity cortical evoked potentialelicited by speech in cochlear-implant users. Hear. Res. 65, 118–124.

Kuhl, P.K., 1985. Methods in the study of infant speech perception. In: Gottlieb, G.,Krasnegor, N. (Eds.), Measurement of Audition and Vision in the First Year ofPostnatal Life: A Methodological Overview. Ablex, Norwood, NJ, pp. 223–251.

Kuhl, P.K., 1991. Human adults and infants show a ‘‘perceptual magnet effect’’ forthe prototypes of speech categories, monkeys do not. Percept. Psychophys. 50,93–107.

Kuhl, P.K., 2004. Early language acquisition: cracking the speech code. Nature Rev.Neursci. 5, 831–843.

Kuhl, P.K., Williams, K.A., Lacerda, F., Stevens, K.N., Lindblom, B., 1992. Linguisticexperience alters phonetic perception in infants by 6 months of age. Science255, 606–608.

Kujala, A., Huotilainen, M., Uther, M., Shtyrov, Y., Monto, S., Ilmoniemi, R.J., Naa-tanen, R., 2003a. Plastic cortical changes underlying learning to communicatewith non-speech sounds. NeuroReport 14, 1683–1687.

Kujala, T., Alho, K., Naatanen, R., 2000. Cross-modal reorganization of humancortical functions. Trends Neurosci. 23, 115–120.

Kujala, T., Kallio, J., Tervaniemi, M., Naatanen, R., 2001a. The mismatch negativity asan index of temporal processing in audition. Clin. Neurophysiol. 112, 1712–1719.

Kujala, T., Karma, K., Ceponiene, R., Belitz, S., Turkkila, P., Tervaniemi, M., Naatanen,R., 2001b. Plastic neural changes and reading improvement caused by audio-visual training in reading-impaired children. Proc. Nat. Acad. Sci. U.S.A. 98,10509–10514.

Kujala, T., Belitz, S., Tervaniemi, M., Naatanen, R., 2003b. Auditory sensory memorydisorder in dyslexic adults as indexed by the mismatch negativity. Eur. J.Neurosci. 17, 1323–1327.

Kujala, T., Tervaniemi, M., Schroger, E., 2007. The mismatch negativity in cognitiveand clinical neuroscience: theoretical and methodological considerations. Biol.Psychol. 74, 1–19.

Page 12: The Adaptive Brain - A Neurophysiological Perspective

T. Kujala, R. Naatanen / Progress in Neurobiology 91 (2010) 55–6766

Lachmann, T., Berti, S., Kujala, T., Schroger, E., 2005. Diagnostic subgroups ofdevelopmental dyslexia have different deficits in neural processing of tonesand phonemes. Int. J. Psychophysiol. 56, 105–120.

Laganaro, M., Morand, S., Schwitter, V., Zimmermann, C., Schnider, A., 2008.Normalisation and increase of abnormal ERP patterns accompany recoveryfrom aphasia in the post-acute stage. Neuropsychologica 46, 2265–2273.

Lang, H.A., Nyrke, T., Ek, M., Aaltonen, O., Raimo, I., Naatanen, R., 1990. Pitchdiscrimination performance and auditory event-related potentials. In: Brunia,C.H.M., Gaillard, A.W.K., Kok, A., Mulder, G., Verbaten, M.N. (Eds.), Psycho-physiological Brain Research. Tilburg University Press, Tilburg, the Netherlands,pp. 294–298.

Laske, R.D., Veraguth, D., Dillier, N., Binkert, A., Holzmann, D., Huber, A.M., 2009.Subjective and objective results after bilateral cochlear implantation in adults.Otol. Neurootol. 30, 313–318.

Leppanen, P.H.T., Richardson, U., Pihko, E., Eklund, K.M., Guttorm, T.K., Aro, M.,Lyytinen, H., 2002. Brain responses to changes in speech sound durations differbetween infants with and without familial risk for dyslexia. Dev. Neuropsychol.22, 407–423.

Lavikainen, J., Huotilainen, M., Pekkonen, E., Ilmoniemi, R.J., Naatanen, R., 1994.Auditory stimuli activate parietal brain regions: a whole-head MEG study.NeuroReport 6, 182–184.

Liberman, A.M., Harris, K.S., Hoffman, H.S., Griffith, B.C., 1957. The discrimination ofspeech sounds within and across phoneme boundaries. J. Exp. Psychol. 54, 358–368.

Lonka, E., Kujala, T., Lehtokoski, A., Johansson, R., Rimmanen, S., Alho, K., Naatanen,R., 2004. The mismatch negativity (MMN) brain response as an index of speechperception recovery in cochlear-implant recipients. Audiol. Neurotol. 9, 160–162.

Mehler, J., Christophe, A., 1994. Language in the infant’s mind. Phil. Trans. R. Soc.Lond. B. 346, 13–20.

Menning, H., Imaizumi, S., Zwitserlood, P., Pantev, C., 2002. Plasticity of the humanauditory cortex induced by discrimination learning of non-native, mora-timedcontrasts of the Japanese language. Learn. Mem. 24, 253–267.

Menning, H., Roberts, L.E., Pantev, C., 2000. Plastic changes in the auditory cortexinduced by intensive frequency discrimination training. NeuroReport 11, 817–822.

Merzenich, M.M., Nelson, R.J., Stryker, M.P., Cynader, M.S., Schoppmann, A., Zook,J.M., 1984. Somatosensory cortical map changes following digit amputation inadult monkeys. J. Comparat. Neurol. 224, 591–605.

Mantysalo, S., Naatanen, R., 1987. The duration of a neuronal trace of an auditorystimulus as indicated by event-related potentials. Biol. Psychol. 24, 183–195.

Nager, W., Kohlmetz, C., Altenmuller, E., Rodriguez-Fornells, A., Munte, T.M.,2003. The fate of sounds in conductors’ brain: an ERP study. Cog. Brain Res. 17,83–93.

Nelken, I., Ulanovsky, N., 2007. Mismatch negativity and stimulus-specific adapta-tion in animal models. J. Psychophysiol. 21, 214–223.

Nithianantharajah, J., Hannan, A.J., 2006. Enriched environments, experience-de-pendent plasticity and disorders of the nervous system. Nature Rev. 7, 697–709.

Novitski, N., Tervaniemi, M., Huotilainen, M., Naatanen, R., 2004. Frequency dis-crimination at different frequency levels as indexed by electrophysiological andbehavioral measures. Cog. Brain Res. 20, 26–36.

Naatanen, R., 1990. The role of attention in auditory information processing asrevealed by event-related potentials and other brain measures of cognitivefunction. Behav. Brain Sci. 13, 201–288.

Naatanen, R., 1992. Attention and Brain Function. Erlbaum, Hillsdale.Naatanen, R., Alho, K., 1997. Mismatch negativity–The measure for central sound

representation accuracy. Audiol. Neuro-otol. 2, 341–353.Naatanen, R., Gaillard, A.W.K., Mantysalo, S., 1978. Early selective-attention effect

on evoked potential reinterpreted. Acta Psychol. 42, 313–329.Naatanen, R., Jacobsen, T., Winkler, I., 2005. Memory-based or afferent processes in

mismatch negativity (MMN): a review of the evidence. Psychophysiology 42,25–32.

Naatanen, R., Lehtokoski, A., Lennes, M., Cheour, M., Huotilainen, M., Iivonen, A.,Vainio, M., Alku, P., Ilmoniemi, R.J., Luuk, A., Allik, J., Sinkkonen, J., Alho, K., 1997.Language-specific phoneme representations revealed by electric and magneticbrain responses. Nature 385, 432–434.

Naatanen, R., Paavilainen, P., Rinne, T., Alho, K., 2007. The mismatch negativity(MMN) in basic research of central auditory processing: a review. Clin. Neu-rophysiol. 118, 2544–2590.

Naatanen, R., Pakarinen, S., Rinne, T., Takegata, R., 2004. The mismatch negativity(MMN)—towards the optimal paradigm. Clin. Neurophysiol. 115, 140–144.

Naatanen, R., Picton, T., 1987. The N1 wave of the human electric and magneticresponse to sound: a review and an analysis of the component structure.Psychophysiology 24, 375–425.

Naatanen, R., Schroger, E., Karakas, S., Tervaniemi, M., Paavilainen, P., 1993. Devel-opment of a memory trace for a complex sound in the human brain. NeuroRe-port 4, 503–506.

Naatanen, R., Tervaniemi, M., Sussman, E., Paavilainen, P., Winkler, I., 2001. ’Primi-tive intelligence’ in the auditory cortex. Trends Neurosci. 24, 283–288.

Paavilainen, P., Arajarvi, P., Takegata, R., 2007. Preattentive detection of nonsalientcontingencies between auditory features. NeuroReport 18, 159–163.

Paavilainen, P., Simola, J., Jaramillo, M., Naatanen, R., Winkler, I., 2001. Preattentiveextraction of abstract feature conjunctions from auditory stimulation asreflected by the mismatch negativity (MMN). Psychophysiology 38, 359–365.

Pantev, C., Oostenveld, R., Engelien, A., Ross, B., Roberts, L.E., Hoke, M., 1998.Increased auditory cortical representation in musicians. Nature 392, 811–814.

Pantev, C., Roberts, L.E., Schulz, M., Engelien, A., Ross, B., 2001. Timbre-specificenhancement of auditory cortical representations in musicians. NeuroReport12, 169–174.

Pascual-Leone, A., Torres, F., 1993. Plasticity of the sensorimotor cortex represen-tation of the reading finger in Braille readers. Brain 116, 39–52.

Pettigrew, C.M., Murdoch, B.E., Chenery, H.J., Kei, J., 2004. The relationship betweenthe mismatch negativity (MMN) and psycholinguistic models of spoken wordprocessing. Aphasiology 18, 3–28.

Pihko, E., Mickos, A., Kujala, T., Pihlgren, A., Westman, M., Alku, P., Byring, R.,Korkman, M., 2007. Group intervention changes brain activity in bilinguallanguage-impaired children. Cereb. Cortex 17, 848 858.

Ponton, C.W., Eggermont, J.J., Don, M., Waring, M.D., Kwong, B., Cunningham, J.,Trautwein, P., 2000. Maturation of the mismatch negativity: effects of profounddeafness and cochlear implant use. Audiol. Neuro-otol. 5, 167–185.

Pulvermuller, F., Hauk, O., Zohsel, K., Neininger, B., Mohr, B., 2005. Therapy-relatedreorganization of language in both hemispheres of patients with chronicaphasia. Neuroimage 28, 481–489.

Pulvermuller, F., Shtyrov, Y., 2006. Language outside the focus of attention: themismatch negativity as a tool for studying higher cognitive processes. Prog.Neurobiol. 79, 49–71.

Pulvermuller, F., Shtyrov, Y., Hauk, O., 2009. Understanding in an instant: neuro-physiological evidence for mechanistic language circuits in the brain. BrainLang. 110, 81–94.

Pulvermuller, F., Shtyrov, Y., Ilmoniemi, R., 2005. Brain signatures of meaning accessin action word recognition. J. Cog. Neurosci. 17, 884–892.

Ramus, F., Rosen, S., Dakin, S.C., Day, B.L., Castellote, J.M., White, S., Frith, U., 2003.Theories of developmental dyslexia: insights from a multiple case study ofdyslexic adults. Brain 126, 841–865.

Rauschecker, J.P., 2002. Cortical map plasticity in animals and humans. Prog. BrainRes. 138, 73–88.

Rauschecker, J.P., Scott, S.K., 2009. Maps and streams in auditory cortex: nonhumanprimates illuminate human speech processing. Nature Neurosci. 12, 718–724.

Recanzone, G.H., Merzenich, M.M., Jenkins, W.M., 1992. Frequency discriminationtraining engaging a restricted skin surface results in an emergence of a cutane-ous response zone in cortical area 3a. J. Neurophysiol. 5, 1057–1070.

Recanzone, G.H., Schreiner, C.E., Merzenich, M.M., 1993. Plasticity in the frequencyrepresentation of primary auditory cortex following discrimination training inadult owl monkeys. J. Neurosci. 13, 87–103.

Renvall, H., Hari, R., 2003. Diminished auditory mismatch fields in dyslexic adults.Ann. Neurol. 53, 551–557.

Reinke, K.S., He, Y., Wang, C., Alain, C., 2003. Perceptual learning modulates sensoryevoked response during vowel segregation. Cog. Brain Res. 17, 781–791.

Rilling, J.K., Glassler, M.F., Preuss, T.M., Ma, X., Zhao, T., Hu, X., Behrens, T.E.J., 2008.The evolution of the arcuate fasciculus revealed with comparative DTI. NatureNeurosci. 11, 426–428.

Robertson, D., Irvine, D.R.F., 1989. Plasticity of frequency organization in auditorycortex of guinea pigs with partial unilateral deafness. J. Comparat. Neurol. 282,456–471.

Rossini, P.M., Martino, G., Narici, L., Pasquarelli, A., Peresson, M., Pizzella, V., Tecchio,F., Torrioli, G., Romani, G.L., 1994. Short-term brain ‘‘plasticity’’ in humans:transient finger representation changes in sensory cortex somatotopy followingischemic anesthesia. Brain Res. 642, 169–177.

Ruben, R.J., 1999. A time frame of critical/sensitive periods of language develop-ment. Indian J. Otolaryngol. Head Neck Surg. 51, 85–89.

Sable, J.J., Low, K.A., Maclin, E.L., Fabiani, M., Gratton, G., 2004. Latent inhibitionmediates N1 attenuation to repeating sounds. Psychophysiology 41, 636–642.

Salmelin, R., Baillet, S., 2009. Electromagnetic brain imaging. Hum. Brain Mapp. 30,1753–1757.

Sams, M., Paavilainen, P., Alho, K., Naatanen, R., 1985. Auditory frequency discrimi-nation and event-related potentials. Electroenceph. Clin. Neurophysiol. 62,437–448.

Shafer, V.L., Morr, M.L., Datta, H., Kurtzberg, D., Schwartz, R.G., 2005. Neurophysio-logical indexes of speech processing deficits in children with specific languageimpairment. J. Cog. Neurosci. 17, 1168–1180.

Sharma, A., Dorman, M.F., 1999. Cortical auditory evoked potential correlates ofcategorical perception of voice-onset time. J. Acoust. Soc. Am. 106, 1078–1083.

Shestakova, A., Brattico, E., Huotilainen, M., Galunov, V., Soloviev, A., Sams, M.,Ilmoniemi, R.J., Naatanen, R., 2002. Abstract phoneme representations in the lefttemporal cortex: magnetic mismatch negativity study. NeuroReport 13, 1813–1816.

Shtyrov, Y., Pihko, E., Pulvermuller, F., 2005. Determinants of dominance: is lan-guage laterality explained by physical or linguistic features of speech? Neuro-image 27, 37–47.

Shtyrov, Y., Hauk, O., Pulvermuller, F., 2004. Distributed neuronal networks forencoding category-specific semantic information: the mismatch negativity toaction words. Eur. J. Neurosci. 19, 1083–1092.

Singh, S., Liasis, A., Rajput, K., Towell, A., Luxon, L., 2004. Event-related potentials inpediatric cochlear implant patients. Ear Hear. 25, 598–610.

Sarkamo, T., Pihko, E., Laitinen, S., Forsblom, A., Soinila, S., Mikkonen, M., Autti, T.,Silvennoinen, H.M., Erkkila, J., Laine, M., Peretz, I., Hietanen, M., Tervaniemi, M.,(2009). Music and speech listening enhance the recovery of early sensoryprocessing after stroke. J. Cog. Neurosci. in press.

Sarkamo, T., Tervaniemi, M., Laitinen, S., Forsblom, A., Soinila, S., Mikkonen, M.,Autti, T., Silvennoinen, H.M., Erkkila, J., Laine, M., Peretz, I., Hietanen, M., 2008.Music listening enhances cognitive recovery and mood after middle cerebralartery stroke. Brain 131, 866–876.

Page 13: The Adaptive Brain - A Neurophysiological Perspective

T. Kujala, R. Naatanen / Progress in Neurobiology 91 (2010) 55–67 67

Tallal, P., 2004. Improving language and literacy is a matter of time. Nature Rev.Neurosci. 5, 721–728.

Tervaniemi, M., Hugdahl, K., 2003. Lateralization of auditory-cortex functions. BrainRes. Rev. 43, 231–246.

Tervaniemi, M., Jacobsen, T., Rottger, S., Kujala, T., Widmann, A., Vainio, M.,Naatanen, R., Schroger, E., 2006. Selective tuning of cortical sound-featureprocessing by language experience. Eur. J. Neurosci. 23, 2538–2541.

Tervaniemi, M., Rytkonen, M., Schroger, E., Ilmoniemi, R.J., Naatanen, R., 2001.Superior formation of cortical memory traces for melodic patterns in musicians.Learn. Mem. 8, 295–300.

Tiitinen, H., May, P., Reinikainen, K., Naatanen, R., 1994. Attentive novelty detectionin humans is governed by pre-attentive sensory memory. Nature 372, 90–92.

Tremblay, K., Kraus, N., 2002. Auditory training induces asymmetrical changes incortical neural activity. J. Speech, Lang., and Hear. Res. 45, 564–572.

Tremblay, K., Kraus, N., Carrell, T.D., McGee, T., 1997. Central auditory systemplasticity: generalization to novel stimuli following listening training. J. Acoust.Soc. Am. 102, 3762–3773.

Tremblay, K., Kraus, N., McGee, T., 1998. The time course of auditory perceptuallearning: neurophysiological changes during speech-sound training. NeuroRe-port 9, 3557–3560.

Tremblay, K., Kraus, N., McGee, T., Ponton, C., Otis, B., 2001. Central auditoryplasticity: changes in the N1-P2 complex after speech-sound training. Ear Hear.22, 79–90.

Ulanovsky, N., Las, L., Farkas, D., Nelken, I., 2004. Multiple time scales of adaptationin auditory cortex neurons. J. Neurosci. 24, 10440–10453.

Uther, M., Kujala, A., Huotilainen, M., Shtyrov, Y., Naatanen, R., 2006. Training inMorse code enhances involuntary attentional switching to acoustic frequency:evidence from ERPs. Brain Res. 1073–1074 417–424.

van Zuijen, T.L., Simoens, V.L., Paavilainen, P., Naatanen, R., Tervaniemi, M., 2006.Implicit, intuitive, and explicit knowledge of abstract regularities in a soundsequence: an event-related brain potential study. J. Cog. Neurosci. 18, 1292–1303.

Weber, C., Hahne, A., Friedrich, M., Friederici, A.D., 2005. Reduced stress patterndiscrimination in 5-month-olds as a marker of risk for later language im-pairment: neurophysiologial evidence. Cog. Brain Res. 25, 180–187.

Weiss, T., Miltner, W.H.R., Liepert, J., Meissner, W., Taub, E., 2004. Rapid functionalplasticity in the primary somatomotor cortex and perceptual changes afternerve block. Eur. J. Neurosci. 20, 3413–4342.

Wertz, R., Auther, L., Burch-Sims, P., Abou-Khalil, R., Kirshner, H., Duncan, G.,1998. A comparison of the mismatch negativity (MMN) event-related poten-tial to tone and speech stimuli in normal and aphasic adults. Aphasiology 12,499–507.

Winkler, I., Denham, S.L., Nelken, I., 2009. Modeling the auditory scene: predictiveregularity representations and perceptual objects. Trends Cog. Sci. 13, 532–540.

Winkler, I., Kujala, T., Tiitinen, H., Sivonen, P., Alku, P., Lehtokoski, A., Czigler, I.,Csepe, V., Ilmoniemi, R.J., Naatanen, R., 1999. Brain responses reveal thelearning of foreign language phonemes. Psychophysiology 36, 638–642.

Xerri, C., Coq, J.O., Merzenich, M.M., Jenkins, W.M., 1996. Experience-inducedplasticity of cutaneous maps in the primary somatosensory cortex of adultmonkeys and rats. J. Physiol. 90, 277–287.