Abstract

Auditory evoked potentials (AEPs) were recorded from eight developmental dyslexic adults with persistent reading, spelling and phonological deficits, and 10 non-dyslexic controls to voiced (/ba/) and voiceless (/pa/) consonant–vowel syllables. Consistent with previous data, non-dyslexics coded these stimuli differentially according to the temporal cues that form the basis of the voiced/voiceless contrast: AEPs had time-locked components with latencies that were determined by the temporal structure of the stimuli. Dyslexics were characterized by one of two electrophysiological patterns: AEP pattern I dyslexics demonstrated a differential coding of stimuli on the basis of some temporal cues, but with an atypically large number of components and a considerable delay in AEP termination time; AEP pattern II dyslexics demonstrated no clear differential coding of stimuli on the basis of temporal cues. These data reveal the presence of anomalies in cortical auditory processing which could underlie persistent perceptual and linguistic impairments in some developmental dyslexics. Furthermore, scalp AEP distribution maps showing the difference observed between /ba/ and /pa/ activity over time suggest that the regions implicated in the processing of crucial time-related acoustic cues were not systematically lateralized to the left hemisphere like they were for non-dyslexics. These findings may be conducive to a better understanding and treatment of perceptual dysfunctions in developmental language disorders.

Introduction

Developmental dyslexia, a language-based reading disorder, is a heterogeneous neurological syndrome affecting 5–8% of otherwise intelligent school-aged children. Most often, reading difficulties will persist into adulthood, although the degree to which subjects continue to suffer from their reading and spelling disabilities is highly variable among individuals. Some cases of dyslexia may have distinct genetic causes (see Francks et al., 2002; Fischer and DeFries, 2002). Among several theories currently proposed to account for developmental reading disorders, the most widely accepted suggests that phonological processing impairments, generally regarded as the core deficit in dyslexia, are rooted in a more general auditory deficit present from early life (Tallal, 1980; McArthur and Bishop, 2001; Goswami, 2000, 2003).

Evidence has accumulated over the last decade showing that developmental dyslexics are impaired on a number of different psychoacoustic measures of auditory processing: categorical perception tests involving natural phonemes or speech-like stimuli (Manis et al., 1997; Breier et al., 2001; Serniclaes et al., 2001, 2004), auditory temporal order judgements involving pure tone or syllable pairs (Tallal, 1980; Reed, 1989; Heath et al., 1999; Heirvang et al., 2002; Brethelton and Holmes, 2003), frequency discrimination (Witton et al., 1998; Talcott et al., 1999; Hill et al., 1999; McAnally et al., 2000), binaural processing (McAnally and Stein, 1996), backward masking (Rosen and Manganari, 2001) and perception of amplitude modulation (Goswami et al., 2002). Electroencephalographic (Schulte-Korne et al., 1999) and magnoencephalographic (Nagarajan et al., 1999) data, too, suggest that dyslexics suffer an enduring dysfunction in low-level auditory processing. Likewise, anatomical and MRI evaluations of dyslexic brains suggest striking alterations in regions that straddle the temporal bank of the left sylvian fissure, such as the planum temporale (Galaburda et al., 1985; Larsen et al., 1990) and Heschl's gyrus (Leonard et al., 2001). These findings militate strongly in favour of a central auditory deficit in dyslexic children and adults (Amitay et al., 2002; Bailey and Snowling, 2002), but considerable controversy exists over the mechanisms linking auditory dysfunction to phonological and reading disorders. Whereas several researchers have emphasized the high incidence of auditory deficits and suggest a causal link (Tallal, 1980; Goswami, 2003; King et al., 2003), others argue that these deficits cannot be considered a major factor in dyslexia because not all dyslexics display them (Ramus et al., 2003; Rosen, 2003).

Here, we investigate the integrity of auditory processing in developmental dyslexic adults with persistent reading, spelling, and phonological deficits using auditory evoked potentials (AEPs). AEPs can provide an objective means of evaluating, with high temporal precision, how the auditory cortex codes acoustico-phonetic cues crucial to speech and language processing. One such cue is the voice onset time (VOT), which constitutes the basis of the voiced/voiceless speech contrast and refers to the temporal relationship between voicing and another phonetically relevant supralaringeal event. In a study using intracerebral AEPs in non-language-impaired subjects, Liégeois-Chauvel et al. (1999) found that VOT-differing voiced and voiceless stop consonant–vowel (CV) syllables were coded in a temporal fashion in the left auditory cortex, with AEP components time-locked to the successive acoustic events in these sounds. In the present study, AEPs were used to explore how these same speech sounds, which developmental dyslexics often have difficulty distinguishing during development, are coded in dyslexic adults and non-dyslexic controls.

Materials and Methods

Subjects

Eight male adult developmental dyslexics with persistent reading, spelling and phonological deficits (aged 22–49 years; mean 32 years) and ten non-dyslexic male controls (aged 21–38 years; mean 28.2 years) participated in the study. All subjects reported themselves as strongly right-handed on the Edinburgh Handedness Inventory (Oldfield, 1971). All were native French speakers, had normal hearing on an audiogram and had an IQ [converted from Raven's Standard Progressive Matrices (PM) (Raven et al., 1992)] of >90, with no known history of neurological, psychiatric, motivational or sensorial antecedents. All gave written informed consent to participate in the study.

Dyslexics were referred to our clinic by the National Association for Parents of Dyslexic Children in France or by a clinical neuropsychologist who had made a previous diagnosis of developmental dyslexia. Dyslexic respondents were administered a questionnaire on personal and family history of developmental language-specific and other impairments, scholastic background and medical history, and were retained in the study if they (i) suffered a long history of debilitating developmental learning impairments specific to reading and writing with a previous diagnosis of developmental dyslexia documented or reported; (ii) reported significant persistent reading, spelling, and/or language deficits; and (iii) achieved a reading age (RA) of more than two years short of the adult RA (i.e. <12.3 years, adult RA being >14.3 years) on a French-language-standardized reading test, with at least 2 SD lower than the adult norm on either a word or pseudoword spelling test despite having a normal non-verbal IQ as evaluated by Raven's PM. All but two dyslexics also reported having persistent difficulties with oral speech sounds (i.e. substitutions involving primarily voiced and voiceless sounds of a minimal pair) and all but two reported having more than one close family member with declared developmental language deficits.

Non-dyslexics were administered the same questionnaire as dyslexics and were included in the study if they (i) had no personal or family history of reading or language impairment; (ii) reported consistently achieving scores of ‘average or better’ on spelling tests and in French language lessons in grade school; (iii) demonstrated normal reading and language function as assessed by a neuropsychologist; and (iv) had normal non-verbal IQ as evaluated by Raven's PM.

RAs were calculated on the basis of reading speed and number of errors on the Alouette standardized reading test (Lefavrais, 1967). Other language tests administered to dyslexics and non-dyslexics were: (i) word spelling (12 items); (ii) pseudoword spelling (11 items); (iii) a phoneme deletion task (40 items); (iv) an ‘odd word out’ phoneme task, in which one of four orally presented words differs from the other three with respect to one speech sound (19 items); and (v) a syllable deletion task (20 items). Performance on all of these tests was highly correlated with RAs across all subjects (r = 0.68 or greater; P < 0.05, Pearson product-moment). Dyslexics were significantly impaired on reading and all other language tests compared with non-dyslexics, in spite of normal non-verbal intellectual function (see Table 1).

Table 1

Mean chronological age, reading age, spelling and phonological performance, IQ and memory span for dyslexics and non-dyslexic controls (standard deviations shown in brackets)


 
Dyslexics (n = 8)
 
Non-dyslexics (n = 10)
 
P-level (Mann–Whitney U)
 
Chronological age (years) 32.0 (10.2) 28.2 (6.7) ns 
Reading age (years) 9.6 (1.6) 14.0+ (0.4) 0.0003 
Word spelling (%) 48 (28.5) 87 (6.1) 0.002 
Pseudoword spelling (%) 57 (19.9) 96 (6.6) 0.0005 
Phoneme deletion (%) 79 (18.7) 98 (2.1) 0.047 
Odd-word-out phoneme task (%) 65 (17.3) 95 (5.3) 0.0006 
Syllable deletion (%) 87 (5.3) 99 (2.2) 0.0004 
IQ (Raven's PM) 118 (12.5) 124 (7.0) ns 
Verbal memory span 5.6 (1.4) 7.1 (1.2) 0.05 
Working memory span 3.3 (1.0) 4.6 (1.1) 0.03 
Visuospatial memory span
 
5.6 (0.8)
 
6.4 (0.7)
 
ns
 

 
Dyslexics (n = 8)
 
Non-dyslexics (n = 10)
 
P-level (Mann–Whitney U)
 
Chronological age (years) 32.0 (10.2) 28.2 (6.7) ns 
Reading age (years) 9.6 (1.6) 14.0+ (0.4) 0.0003 
Word spelling (%) 48 (28.5) 87 (6.1) 0.002 
Pseudoword spelling (%) 57 (19.9) 96 (6.6) 0.0005 
Phoneme deletion (%) 79 (18.7) 98 (2.1) 0.047 
Odd-word-out phoneme task (%) 65 (17.3) 95 (5.3) 0.0006 
Syllable deletion (%) 87 (5.3) 99 (2.2) 0.0004 
IQ (Raven's PM) 118 (12.5) 124 (7.0) ns 
Verbal memory span 5.6 (1.4) 7.1 (1.2) 0.05 
Working memory span 3.3 (1.0) 4.6 (1.1) 0.03 
Visuospatial memory span
 
5.6 (0.8)
 
6.4 (0.7)
 
ns
 

The dyslexics included in this study suffered significant selective impairments in reading, spelling, phoneme and syllable manipulation, and working memory, while non-linguistic analytical functions (IQ) were normal.

Electrophysiological Recordings

Data Acquisition

AEPs were recorded using acquisition software and a 64-channel SynAmps EEG/EP amplication system from Neuro Scan Labs (Neurosoft, Inc.). A 64-electrode cap [QuickCap (Neuromedical Supplies, Inc.)] with scalp locations based on the 10–10 International System was used. The site of the reference was the right mastoid.

During acquisition, the EEG signal was amplified with a digital band-pass filter (0.5–200 Hz) and digitized at a rate of 1 kHz per channel (and a resolution of 1 ms per sample). Recordings were carried out while subjects sat comfortably in a chair in an electrically isolated and sound attenuating room and read a self-chosen book. Subjects were instructed to refrain from making movements and clenching muscles.

Stimuli and Procedure

The stimuli used were the endpoints taken from the voiced (/ba/) and voiceless (/pa/) CV continuum described above. During EEG acquisition, each subject was administered five 8 min blocks of 450 trials of one of two stimuli, followed by the same number of blocks and presentations of the other stimulus. Testing order of /ba/ and /pa/ block groups was counterbalanced across subjects. The stimuli in each block were presented binaurally through headphones (Sennheiser HD 25) at a comfortable listening level adjusted to 70 dB SL (sensation level, relative to the threshold at 1000 Hz) for each subject and were administered at regular intervals of 1030 ms using Labview. Subjects were instructed not to attend to stimuli while reading and were allowed to rest between blocks.

Data Analysis

EEG recordings were sectioned into 820 ms epochs (82 ms pre-stimulus and 738 ms post-stimulus) and a baseline correction using the pre-stimulus portion of the signal was carried out. AEPs for each stimulus were averaged for each subject and grand-averaged across subjects. Epochs that were contaminated by muscle or eye-blink artefacts were automatically rejected from the averaging procedure. Averaged AEPs were then mean-referenced and digitally band-pass filtered (high-pass cut-off frequency = 1 Hz, slope = 24 dB/octave; low-pass cut-off frequency = 30 Hz, slope = 48dB/octave).

Analysis of AEP Timecourses

AEP timecourses for /ba/ and /pa/ were analyzed using ASA (Advanced Source Analysis, A.N.T. Software BV). For each subject, AEP activity from all 63 electrodes was first superposed and the global field power (GFP) calculated, in order to isolate and identify AEP components. The number and latencies of components were subsequently recorded for each stimulus and subject. Scalp potential topographic maps were generated using a standard three-dimensional head model and a three-dimensional spherical spline interpolation.

Categorical Perception (CP)

All subjects were administered identification and discrimination subtests of a CP task using CV stimuli taken from a synthetically created, natural speech voiced–voiceless (/ba/–/pa/) continuum. To obtain the stimuli in this continuum, the voiced CV, /ba/, was recorded by a female native French speaker then modified by extracting progressively longer segments of the initial low frequency activity characterizing voicing (i.e. <500 Hz) from the spectrum. In this way, 14 stimuli differing in VOT and ranging from /ba/ (total duration = 380 ms; VOT = −110 ms) to /pa/ (total duration = 270 ms; VOT = +40 ms) were obtained. These stimuli were tested and validated in a previous study (Laguitton et al., 1997).

The identification task consisted of a training phase with /ba/ and /pa/ endpoint stimuli and 140 test trials of continuum stimuli (14 items × 10 repetitions) presented randomly through headphones. Subjects were instructed to identify each stimulus by pressing a ‘BA’ or ‘PA’ button with their right index finger. In the training phase, subjects were required to score correctly on at least eight consecutive trials to demonstrate their understanding of the task. Feedback was given on training trials, but not on test trials.

The discrimination task consisted of a training phase with same or different endpoint pairs and 100 test trials of continuum item pairs (10 pairs × 10 repetitions). After adequate training (correct performance on eight consecutive endpoint pairs with feedback), subjects were asked to make a ‘same’ or ‘different’ judgement (AX paradigm) upon each presentation by pressing either a ‘SAME’ or a ‘DIFFERENT’ button with their right index finger.

Results

Electrophysiological Data

Table 2 summarizes latencies of /ba/ and /pa/ AEP components identified for non-dyslexics and dyslexics on the basis of the GFP.

Table 2

Individual and mean non-dyslexic and dyslexic AEP component latencies for the voiced CV stimulus /ba/ (left) and its voiceless counterpart /pa/ (right)

 /ba/ (voiced CV)
 
      /pa/ (voiceless CV)
 
     

 
P50
 
N1a
 
N1b
 
P2
 
N240 (*)
 
Off-response
 
 P50
 
N1a
 
N1b
 
P2
 
Off-response
 
 
Non-dyslexic              
    AF   98  257 361   78 121    
    DD   96 158 213 343    94 214   
    DM  75 108 175 224    69 102 162   
    JG   130 195 267     112    
    JM  83 129 172 248     90 158   
    JP  81  177 223 315   68 161 201   
    MT  72 131  230 356   57 101    
    OT  86  181 245    74 113    
    RC  84  180 243 291   70 128    
    YA  77  180 260 338   70  170   
    Mean  79.7 115.3 177.3 241.0 334.0   69.4 113.6 181   
    SD  5.15 16.59 10.33 17.89 26.53   6.48 21.64 25.00   
Dyslexic pattern I              
    NR  70 108 177 247 346 480  61 101 161 260 299 
    AB 55 116 144 171 192 235 380 59  111  260 320 
    AS  73 110 173 231 278 347  61 101 161 260 299 
    SC 56  108 167 223 369 446  67 104 200 250 350 
Dyslexic pattern II              
    PH   101 171 – 320   76  190   
    FL  76  184 – 350   70  190   
    CM  72  178 –    71 131    
    MD  83 127 180 –    67 104 200   
    Mean 55.5 81.7 116.3 175.1 223.3 316.3 413.3 59.0 67.6 108.7 183.7 257.5 317.0 
    SD
 
0.71
 
17.42
 
16.08
 
5.59
 
23.10
 
50.76
 
60.62
 

 
5.41
 
11.54
 
18.12
 
5.00
 
24.12
 
 /ba/ (voiced CV)
 
      /pa/ (voiceless CV)
 
     

 
P50
 
N1a
 
N1b
 
P2
 
N240 (*)
 
Off-response
 
 P50
 
N1a
 
N1b
 
P2
 
Off-response
 
 
Non-dyslexic              
    AF   98  257 361   78 121    
    DD   96 158 213 343    94 214   
    DM  75 108 175 224    69 102 162   
    JG   130 195 267     112    
    JM  83 129 172 248     90 158   
    JP  81  177 223 315   68 161 201   
    MT  72 131  230 356   57 101    
    OT  86  181 245    74 113    
    RC  84  180 243 291   70 128    
    YA  77  180 260 338   70  170   
    Mean  79.7 115.3 177.3 241.0 334.0   69.4 113.6 181   
    SD  5.15 16.59 10.33 17.89 26.53   6.48 21.64 25.00   
Dyslexic pattern I              
    NR  70 108 177 247 346 480  61 101 161 260 299 
    AB 55 116 144 171 192 235 380 59  111  260 320 
    AS  73 110 173 231 278 347  61 101 161 260 299 
    SC 56  108 167 223 369 446  67 104 200 250 350 
Dyslexic pattern II              
    PH   101 171 – 320   76  190   
    FL  76  184 – 350   70  190   
    CM  72  178 –    71 131    
    MD  83 127 180 –    67 104 200   
    Mean 55.5 81.7 116.3 175.1 223.3 316.3 413.3 59.0 67.6 108.7 183.7 257.5 317.0 
    SD
 
0.71
 
17.42
 
16.08
 
5.59
 
23.10
 
50.76
 
60.62
 

 
5.41
 
11.54
 
18.12
 
5.00
 
24.12
 

For non-dyslexics, a second negative component at ∼240 ms (asterisk) occurs uniquely for /ba/ and reflects the processing of the release burst that follows the onset of voicing in the case of French voiced stop CVs. Dyslexic /ba/ and /pa/ AEPs demonstrate either a greater number of components with a delay in AEP termination time (pattern I) or an absence of a clear voiced-CV-specific supplementary component (pattern II).

Non-dyslexics' scalp AEPs are similar to those obtained from the auditory cortex using intracerebral recordings (Liégeois-Chauvel et al., 1999) with respect to the number of components and latencies observed for voiced and voiceless stimuli. Following the N1/P2 complex, a negative component peaking at ∼240 ms can be observed for the voiced CV /ba/ but not the voiceless CV /pa/. This component is followed by an off-response at ∼334 ms.

Dyslexic AEPs were similar to non-dyslexic AEPs with respect to the N1/P2 complex, but markedly different in several other respects. Two distinct patterns of atypical responses were observed. The first (AEP pattern I: subjects NR, AB, AS, SC) was characterized by (i) a greater number of components for both /ba/ and /pa/; (ii) an AEP termination lag of ∼130 ms in all subjects for /pa/ and in two subjects (NR, SC) for /ba/; and (iii) an appearance of earlier-latency components (P50), which were never present in non-dyslexics, in two subjects (AB, SC). The second dyslexic pattern (AEP pattern II: subjects PH, FL, CM, MD) was characterized by a non-appearance of the voiced CV-specific N240 supplementary component for /ba/ after the N1/P2 complex and an absence of stimulus differentiation on the basis of temporal cues. In other words, these latter subjects exhibited the same number of components before the off-response, irrespective of whether the eliciting stimulus was /ba/ or /pa/.

Figure 1A (upper panel) shows grand-averaged non-dyslexic AEPs for the voiced CV /ba/ recorded over the scalp. In the lower panel, the same responses are shown superposed from all 63 electrodes, along with the /ba/ oscillogram. As in Table 2, several components are identifiable. The first corresponds to the N1a (80 ms) and N1b (120 ms) and the second to the P2 (180 ms). This N1a,b/P2 complex is observed with the onset of voicing (or, rather, pre-voicing in the case of French voiced consonants), while the N240 component that follows (asterisk) corresponds to the processing of the other acoustic cue necessary in determining stop consonant VOT (the release burst) and vowel onset. This component was observed in all non-dyslexic /ba/ but not /pa/ AEPs, and is thought to be indicative of the sequential processing of phonetically relevant cues in voiced stop consonant perception (Liégeois-Chauvel et al., 1999). Figure 1B (upper panel) shows grand-averaged AEPs for the voiceless CV /pa/ for the same subjects. In the lower panel, the same responses are shown as the superposition of all 63 electrodes along with the /pa/ oscillogram. In this case, only an N1a,b/P2 complex is present. Here, the N1a,b culminate slightly earlier, at 70 and 115 ms, respectively. For both /ba/ and /pa/, AEP durations are roughly time-locked to stimulus duration.

Figure 1.

Grand-averaged non-dyslexic /ba/ (A) and /pa/ (B) AEPs from 63 electrodes. In the top panel, AEPs are shown distributed over the scalp (negative polarity is up). In the bottom panel, the same AEPs are shown superimposed from all electrodes to faciliate component identification. /Ba/ and /pa/ AEPs are shown alongside corresponding /ba/ and /pa/ oscillograms to illustrate how non-dyslexic AEPs are determined by the temporal structure of these two VOT-differing speech stimuli. A supplementary voiced-CV-specific component peaking at ∼240 ms (asterisk) is observed uniquely for /ba/ and corresponds to the processing of the release burst which occurs long after the onset of voicing in the case of French voiced stop CVs.

Figure 1.

Grand-averaged non-dyslexic /ba/ (A) and /pa/ (B) AEPs from 63 electrodes. In the top panel, AEPs are shown distributed over the scalp (negative polarity is up). In the bottom panel, the same AEPs are shown superimposed from all electrodes to faciliate component identification. /Ba/ and /pa/ AEPs are shown alongside corresponding /ba/ and /pa/ oscillograms to illustrate how non-dyslexic AEPs are determined by the temporal structure of these two VOT-differing speech stimuli. A supplementary voiced-CV-specific component peaking at ∼240 ms (asterisk) is observed uniquely for /ba/ and corresponds to the processing of the release burst which occurs long after the onset of voicing in the case of French voiced stop CVs.

Figures 2 and 3 show the two different patterns of dyslexic /ba/ and /pa/ responses (grand-averages over the scalp and superposed from 63 electrodes.) The first (Fig. 2: AEP pattern I) is characterized by several additional peaks following the voiced-CV-specific component at 230 ms (asterisk) for /ba/. /Ba/ AEPs from these subjects did not clearly terminate before 400 ms and identification of an off-response was difficult. /Pa/ AEPs also terminated later (after 255 ms) for this group compared with non-dyslexic AEPs (which terminated after 180 ms). In spite of this, /ba/ and /pa/ AEPs were distinguishable on the basis of response duration and number of components. A second pattern (Fig. 3: AEP pattern II) did not demonstrate a clear negative component at or near 240 ms for /ba/. Subjects from this group differed most markedly from non-dyslexics and AEP pattern I dyslexics in that, although a more pronounced off-response (320 ms) was observed for /ba/ than for /pa/, /ba/ and /pa/ AEPs were not distinguishable on the basis of the voiced-CV-specific N240 component or response duration.

Figure 2.

Grand-averaged AEP pattern I dyslexic responses for /ba/ (A) and /pa/ (B). In the top panel, AEPs are shown distributed over the scalp (negative polarity is up). In the bottom panel, the same AEPs are shown superimposed from all 63 electrodes to faciliate component identification.

Figure 2.

Grand-averaged AEP pattern I dyslexic responses for /ba/ (A) and /pa/ (B). In the top panel, AEPs are shown distributed over the scalp (negative polarity is up). In the bottom panel, the same AEPs are shown superimposed from all 63 electrodes to faciliate component identification.

Figure 3.

Grand-averaged AEP pattern I dyslexic responses for /ba/ (A) and /pa/ (B). In the top panel, AEPs are shown distributed over the scalp (negative polarity is up). In the bottom panel, the same AEPs are shown superimposed from all 63 electrodes to faciliate component identification.

Figure 3.

Grand-averaged AEP pattern I dyslexic responses for /ba/ (A) and /pa/ (B). In the top panel, AEPs are shown distributed over the scalp (negative polarity is up). In the bottom panel, the same AEPs are shown superimposed from all 63 electrodes to faciliate component identification.

Lateralization of the /ba/ Supplementary Component

Figure 4 shows scalp distribution maps of differences in /ba/-minus-/pa/ potential over time for non-dyslexics and for AEP pattern I and II dyslexics. For non-dyslexics, a highly significant difference (t = 3.7; P < 0.001) in potential was observed at latencies around the voiced-CV-specific N240 component over left-sided temporal electrodes only, suggesting a lateralization of sequential processing underlying voiced stop consonant perception. A significant difference (t = 3.4, P < 0.001) was also observed over frontal electrodes, primarily on the left, at latencies around the P2 component (shown on the figure at 190 ms), reflecting a larger P2 amplitude in /ba/ compared with /pa/ responses over these regions. Other significant /ba/minus-/pa/ differences were observed at the latency of the /ba/ off-response over left temporal electrodes (t = 3.2; P < 0.001) and between 399 and 418 ms over bilateral fronto-central electrodes (not shown). The latter are likely related to differences in the time required for /ba/ and /pa/ responses to return to baseline (/ba/ AEPs return to baseline later than /pa/ AEPs).

Figure 4.

t-test distribution maps showing the difference observed between /ba/ and /pa/ grand-averaged activity over time for non-dyslexics, AEP pattern I dyslexics and AEP pattern II dyslexics. Bright blue denotes significant negative differences; bright red denotes significant positve differences. For non-dyslexics, significant negative /ba/-minus-/pa/ activity is observed over temporal electrodes on the left (top row shows left side; bottom row shows right side) at latencies around the voiced-CV-specific supplementary component (red square shows negative activity; significant differences are indicated by an asterisk). For AEP pattern I dyslexics, significant /ba/-minus-/pa/ activity is observed over temporal electrodes on the right side only at the latency of the voiced-CV-specific supplementary component (230 ms for this group). For AEP pattern II dyslexics, no significant /ba/minus-/pa/ activity is observed at any latency over left or right temporal electrodes.

Figure 4.

t-test distribution maps showing the difference observed between /ba/ and /pa/ grand-averaged activity over time for non-dyslexics, AEP pattern I dyslexics and AEP pattern II dyslexics. Bright blue denotes significant negative differences; bright red denotes significant positve differences. For non-dyslexics, significant negative /ba/-minus-/pa/ activity is observed over temporal electrodes on the left (top row shows left side; bottom row shows right side) at latencies around the voiced-CV-specific supplementary component (red square shows negative activity; significant differences are indicated by an asterisk). For AEP pattern I dyslexics, significant /ba/-minus-/pa/ activity is observed over temporal electrodes on the right side only at the latency of the voiced-CV-specific supplementary component (230 ms for this group). For AEP pattern II dyslexics, no significant /ba/minus-/pa/ activity is observed at any latency over left or right temporal electrodes.

For AEP pattern I dyslexics, a highly localized significant difference in /ba/-minus-/pa/ potential was observed at the latency of the voiced-CV-specific negative component (230 ms for this group) over right (but not left) temporal electrodes (t = 3.8; P < 0.001). This observation suggests differences in these dyslexics in the functional lateralization and organization of regions underlying the sequential processing of phonetic cues in voiced stop consonants. For AEP pattern II dyslexics, no significant difference in /ba/-minus-/pa/ potential was observed at the expected latency of the N240 component (or at any other latency) over left or right temporal electrodes. Neither dyslexic group exhibited the /ba/-minus-/pa/ difference at the latency of the P2 component that was observed for non-dyslexics.

Categorical Perception Data

Figure 5A shows grand-averaged /ba/–/pa/ identification functions for non-dyslexics and the two dyslexic subgroups characterized on the basis of AEP data. No significant group differences were observed in boundary location (50% /ba/–/pa/ cross-over point) or categorization accuracy (indexed by slope).

Figure 5.

Grand-averaged categorical perception data for non-dyslexics and dyslexics of the two AEP pattern types. (A) /Ba/–/pa/ identification functions. Category boundary position (50% cross-over point) is indicated by the arrows. (B) /Ba/–/pa/ discrimination functions. Inter-categorical discrimination peaks for all three groups and within-category discrimination for AEP pattern II dyslexics are indicated. Compared with non-dyslexics, pattern I and II dyslexics exhibit flatter and wider discrimination functions, as well as significantly smaller inter-minus-intra-categorical discrimination differences. Pattern II dyslexics in particular discriminate within-category pairs (/ba/ pairs 2;5, 3;6, 4;7, 5;8, /pa/ pair 11;14) better than non-dyslexics.

Figure 5.

Grand-averaged categorical perception data for non-dyslexics and dyslexics of the two AEP pattern types. (A) /Ba/–/pa/ identification functions. Category boundary position (50% cross-over point) is indicated by the arrows. (B) /Ba/–/pa/ discrimination functions. Inter-categorical discrimination peaks for all three groups and within-category discrimination for AEP pattern II dyslexics are indicated. Compared with non-dyslexics, pattern I and II dyslexics exhibit flatter and wider discrimination functions, as well as significantly smaller inter-minus-intra-categorical discrimination differences. Pattern II dyslexics in particular discriminate within-category pairs (/ba/ pairs 2;5, 3;6, 4;7, 5;8, /pa/ pair 11;14) better than non-dyslexics.

Figure 5B shows discrimination functions for the same groups. Dyslexics made markedly poorer discrimination judgements: inter-categorical discrimination peaks were flatter and wider compared with non-dyslexics. The difference between inter- and intra-categorical performance was significantly smaller for both AEP pattern types of dyslexics (pattern I: χ2 = 13.7, P < 0.0002; pattern II: χ2 = 21.9, P < 0.00001) compared with non-dyslexics, suggesting that the discrimination of voiced and voiceless speech sounds is less categorical for these adult dyslexics with persistent reading and language deficits.

Discussion

AEPs recorded to voiced and voiceless stop CV syllables from dyslexic adults with persistent reading, spelling and phonological deficits were markedly and systematically different from non-dyslexic AEPs. The latter were highly comparable to those observed in previous scalp (Roman et al., 2004) and intracerebral (Liégeois-Chauvel et al., 1999; Steinschneider et al., 1999) AEP and magnetoencephalographic (Simos et al., 1998a; Ackermann et al., 1999) studies in non-language-impaired humans. In non-language-impaired subjects, voiced and voiceless stop CV syllables are processed in a temporal fashion according to the sequential phonetic markers constituting the voiced/voiceless contrast and this differential processing is quantifiable on the basis of electroencephalographic scalp recordings. In French, voiced stop consonants (/b/, /d/, /g/) have long negative VOT values with the onset of voicing preceding release by ∼100–120 ms, while voiceless stops have short positive values. For our non-dyslexic controls, the long VOT value voiced CV /ba/ elicited an N1/P2 (100–180 ms) complex occurring with the onset of voicing, followed by a negative component occurring at ∼240 ms. The delay between the first and second complex corresponded roughly to VOT duration (120 ms). The short VOT value voiceless CV /pa/, on the other hand, elicited only a single clear complex (N1/P2). AEPs elicited to these two stimuli were thus clearly distinguishable on the basis of a supplementary negative component that was time-locked to consonantal release and vowel onset in /ba/. This voiced-CV-specific N240 component may be thought of as a correlate of the sequential processing implicated in voiced/voiceless stop CV perception and discrimination. Pisoni (1977) was the first to propose that the differential perception of voiced and voiceless stops consonants is based on whether consonantal release and voicing onset are perceived as occurring sequentially. Animal studies have revealed a characteristic pattern of activity wherein syllables with a short VOTs (i.e. where the salient acoustic cues of consonantal release and voicing onset are less temporally distinct) evoke a single response burst that is time-locked to consonantal release, while syllables with a longer VOT (i.e. where acoustic cues are more temporally distinct) evoke response bursts that are time-locked to both consonantal release and voicing onset (Steinschneider et al., 1994, 1995, 2003; McGee et al., 1996; Eggermont, 1999). These findings, which are similar to those obtained directly from the non-language-impaired human auditory cortex (Liégeois-Chauvel et al., 1999; Steinschneider, 1999, 2004), suggest that the human and non-human animal auditory cortex is tuned to process the sequentially occurring acoustically salient cues that constitute speech sounds. In humans, time-locked sequential neuronal activity may underlie VOT processing and contribute to voiced/voiceless speech sound discrimination in the left hemisphere. Significantly, /ba/-minus-/pa/ scalp distribution maps for the non-dyslexics in our study showed the voiced-CV-specific N240 component to be the most consistently left-lateralized of components, consistent with previous intracerebral data demonstrating its origin in the left auditory cortex (Liégeois-Chauvel et al., 1999) and MEG data suggesting a left-hemispheric origin (Simos et al., 1998b; Papanicolaou et al., 2003).

Dyslexic AEPs were characterized by one of two distinct atypical patterns of response. AEP pattern I dyslexics demonstrated a differential coding of voiced and voiceless stimuli, but with a larger number of components and an overall delay in AEP termination time. Off-responses occurred much later than for non-dyslexics, consistent with the idea that a lag or sluggishness in auditory processing may be associated with reading impairment (Hari and Kiesila, 1996; Hari and Renvall, 2001). The presence of additional components in this group may be due to an increase in synchronized onset responses coding frequency modulations along the speech signal. In a recent study, Steinschneider et al. (2004) showed that onset responses evoked by both the first and second tones of a two-tone complex are reliably detected from tonotopically organized neuronal populations in the primary auditory cortex of the monkey at tone onset time separations of as short as 20 ms. In the same study, they demonstrated that the representation of VOT in the human auditory cortex varied in a manner that reflected the spectral composition of voiced/voiceless speech sounds and the tonotopic organization of the auditory cortex in spectral processing. Numerous distinct temporo-spectral components thus contribute to the scalp-recorded N1 (Liégeois-Chauvel 1994), but they are generally difficult to distinguish on the basis of surface recordings. It is possible that the multiple components observed in AEP pattern I dyslexics reflect an over-synchronization of activity from the sources of these different components: because they would not overlap in time, each component could be identifiable from electrophysiological recordings. This may be due to anatomically related differences affecting the orientation of sources or may reflect the processing of extraneous cues within the speech signal (Serniclaes et al., 2001).

Previous studies have shown a rightward functional asymmetry in spectral processing in the human auditory cortex (Liégeois-Chauvel et al., 2001; Zatorre and Belin, 2001). Interestingly, t-test distribution maps of /ba/-minus-/pa/ activity for AEP pattern I dyslexics suggest that the regions generating the N240 component were lateralized to the right hemisphere, unlike for non-dyslexics. On a more general level, this observation suggests differences for these dyslexics in the functional organization of regions implicated in VOT perception and voiced/voiceless consonant discrimination. Numerous PET (Rumsey et al., 1992; Brunswick et al., 1999), fMRI (Temple et al., 2001) and MEG (Simos et al., 2000a,b) studies have found decreased activation, or attenuated asymmetry, of left-hemispheric temporal or temporoparietal regions in dyslexics during phonology or reading tasks. Perhaps the atypical pattern of asymmetry observed for AEP pattern I subjects has to do with the fact they constituted part of a population of adult-aged dyslexics with persistent deficits: in these subjects, a ‘normalization’ of the left-hemispheric neural circuits underlying speech, phonological and reading functions through rehabilitation is not likely to have occurred. In a recent MEG study, Simos et al. (2002) reported that, before successful remedial training, dyslexic children differed from their non-dyslexic counterparts in that they exhibited strong activity in the right and little or no activity in the left posterior portion of the superior temporal gyrus (STGp) during pseudoword reading. After training, however, this pattern normalized and activity in the left STGp increased markedly. It is also possible that these dyslexics have developed compensatory mechanisms or strategies to aid them in speech and language processing and that this process is associated with a functional reorganization of neural systems subserving speech perception.

AEP pattern II dyslexics did not demonstrate the differential coding based on sequential phonetic cues that is typically observed for voiced and voiceless /ba/–/pa/ stimuli. AEPs from these dyslexics appeared almost identical due to the absence of the supplementary voiced CV-specific N240 component for /ba/. This observation may reflect a difficulty in some developmental dyslexics to process rapidly occurring acoustic events. In a magnetoencephalographic study, Nagarajan et al. (1999) found that reading-impaired adults exhibited diminished responses to the second of two rapidly successive acoustic stimuli when inter-stimulus intervals were 100–200 ms, but not when they were 500 ms. Given that voiced and voiceless consonants of a minimal pair differ predominantly according VOT values which in any case fall within a relatively small (i.e. <200 ms) time window, it is conceivable that such a deficit could make a differential coding of these sounds difficult. The data for this group are evocative of the ‘rapid auditory processing’ deficits cited in some behavioral studies (Tallal, 1980; Reed, 1989). These authors have suggested that such deficits could constitute a causal mechanism in developmental reading disorders because they could compromise the formation of clear phonemic representations necessary for sound-to-letter learning.

Both groups of dyslexics also differed from non-dyslexics with respect to the P2 component: while /ba/-minus-/pa/ activity for P2 was observed over (primarily left) frontal regions for non-dyslexics, no such activity was observed for dyslexics. The P2 is a positive-going component thought to reflect both exogenous and endogenous processing (Dunn et al., 1998; see also Crowley and Colrain, 2004), and may be an indicator of processes underlying feature detection (Luck and Hillyard, 1994) and early sensory stage item encoding (Dunn et al., 1998). Its amplitude is typically maximal over fronto-central regions of the scalp and its generators have been localized to the auditory cortex (secondary and associative areas) (Hari et al., 1987; Verkindt et al., 1994; Godey et al., 2001). That dyslexics of both groups in the present study exhibited anomalous P2 activity in the passive listening to voiced and voiceless speeech may suggest auditory processing deficits in feature detection and item encoding in these subjects. Similar findings have been reported in other studies (Breznitz and Meyler, 2003).

Consistent with our AEP data, psychophysical data from the categorical perception task revealed impaired discrimination functions in our dyslexics — especially those with pattern II AEPs. In all dyslexics, poor categorical discrimination was the result of flatter and wider discrimination peaks observed at category boundaries and/or better within-category discrimination performance. Previous investigations on the categorical perception of stop consonants have similarly demonstrated that both dyslexic children (Godfrey et al., 1981; Werker and Tees, 1987; Serniclaes et al., 2001) and adults (Steffens et al., 1992) often exhibit flatter or wider inter-categorical discrimination peaks and, in some cases, better within-category discrimination. These findings corroborate our AEP data suggesting that developmental dyslexics with persistent reading and language deficits may also exhibit auditory processing anomalies in VOT discrimination. It is important to note, however, that, while the dyslexics in the present study exhibited impaired discrimination functions, they were not significantly impaired at identifying the same stimuli. The reasons for this are unclear, although similar findings have been reported in previous studies (Maassen et al., 2001).

The electrophysiological and psychophysical anomalies reported in this study likely reflect differences in the functional organization of speech and language regions in dyslexics. These may ultimately arise from, or be associated with, differences in structural organization. Post-mortem (Galaburda et al., 1985) and brain imaging data (Hynd et al., 1990; Humphreys et al., 1990; Kushch et al., 1993) have suggested a reduction in, or an absence of, the usual left-greater-than-right asymmetry of the planum temporale (PT) and/or neighbouring areas. More recently, Leonard et al. (2001) reported a duplicated Heschl's gyrus (HG) in seven out of nine adult dyslexics with a specific phonological deficit. A duplicated left-hemispheric HG is uncommon in the normal population (Penhune et al., 1996) and has been associated, along with other perisylvian anomalies, with poor phonological performance (Leonard et al., 1993). Given that all the dyslexic subjects in the present study exhibited a specific phonological impairment (in addition to their reading and spelling deficit), it may be conceivable that their deficits are related to underlying anatomical anomalies. It should kept in mind, however, that other studies have failed to demonstrate anomalies in PT asymmetry and/or HG duplication in dyslexics [although this may due to differences in subject selection criteria and to heterogeneity in this population (see Eckert, 2004)]. Nonetheless, the use of brain imaging techniques in a future study may provide a neuroanatomical correlate to the auditory/speech processing anomalies observed here.

The mechanisms underlying auditory processing deficits and their role in dyslexia are subject to much debate (Habib, 2000), but most current views tend to implicate a single one. Our observations suggest, instead, that more than one dysfunctional mechanism may be implicated in different dyslexics: one related to the processing of extraneous acoustic cues in the speech signal and/or a ‘sluggishness’ in auditory processing (AEP pattern I), and another to an inability to code crucial, sequentially occurring cues differentiating voiced/voiceless speech sounds (AEP pattern II). Both of these were associated with significant reading, spelling, and phonological impairments persisting into adulthood. Other auditory processing dysfunctions likely exist, such as would underlie impairments in rhythm timing — a slow-moving feature of the speech signal (Goswami et al., 2002). Put together, these observations may explain some inconsistencies found in the literature (Ramus et al., 2003; Rosen, 2003), since the tasks used in different studies to demonstrate auditory deficits may only reveal one subtype of dysfunction. Our data also suggest that auditory processing dysfunctions can be identified objectively on the basis of AEP recordings. It is hoped that future investigations in this direction will shed light on the role of auditory processing deficits in developmental language disorders and provide a plausible explanation as to why only a proportion of dyslexics show evidence of speech perception deficits and/or central auditory dysfunction (Tallal, 1980; King et al., 2003; Rosen, 2003), when virtually all suffer phonological impairments.

This work was supported by a grant ‘PROGRES’ from INSERM and the French ministry of Health program PHRC.

References

Ackermann H, Lutzenberger W, Hertrich I (
1999
) Hemispheric lateralization of the neural encoding of temporal speech features: a whole-head magnetencephalography study.
Brain Res Cogn Brain Res
 
7
:
511
–518.
Amitay S, Ahissar M, Nelken I (
2002
) Auditory processing deficits in reading disabled adults.
J Assoc Res Otolaryngol
 
3
:
302
–320.
Bailey PJ, Snowling MJ (
2002
) Auditory processing and the development of language and literacy.
Br Med Bull
 
63
:
135
–146.
Breier JI, Gray L, Fletcher JM, Diehl RL, Klaas P, Foorman BR, Molis MR (
2001
) Perception of voice and tone onset time continua in children with dyslexia with and without attention deficit/hyperactivity disorder.
J Exp Child Psychol
 
80
:
245
–270.
Bretherton L, Holmes VM (
2003
) The relationship between auditory temporal processing, phonemic awareness, and reading disability.
J Exp Child Psychol
 
84
:
218
–243.
Breznitz, Z, Meyler, A (
2003
) Speed of lower-level auditory and visual processing as a basic factor in dyslexia: electrophysiological evidence.
Brain Lang
 
85
:
166
–184.
Brunswick N, McCrory E, Price CJ, Frith CD, Frith U (
1999
) Explicit and implicit processing of words and pseudowords by adult developmental dyslexics: a search for Wernicke's Wortschatz?
Brain
 
122
:
1901
–1917.
Crowley KE, Colrain IM (
2004
) A review of the evidence for P2 being an independent component process: age, sleep and modality.
Clin Neurophysiol
 
115
:
732
–744.
Dunn BR, Dunn DA, Languis M, Andrews D (
1998
) The relation of ERP components to complex memory processing.
Brain Cogn
 
36
:
355
–376.
Eckert M (
2004
) Neuroanatomical markers for dyslexia: a review of dyslexia structural imaging studies.
Neuroscientist
 
10
:
362
–371.
Eggermont JJ (
1999
) Neural correlates of gap detection in three auditory cortical fields in the cat.
J Neurophysiol
 
81
:
2570
–2581.
Fisher SE, DeFries JC (
2002
) Developmental dyslexia: genetic dissection of a complex cognitive trait.
Nat Rev Neurosci
 
3
:
767
–780.
Francks C, MacPhie IL, Monaco AP (
2002
) The genetic basis of dyslexia.
Lancet Neurol
 
1
:
483
–490.
Galaburda AM, Sherman GF, Rosen GD, Aboitiz F, Geschwind N (
1985
) Developmental dyslexia: four consecutive patients with cortical anomalies.
Ann Neurol
 
18
:
222
–233.
Godey B, Schwartz D, de Graaf JB, Chauvel P, Liégeois-Chauvel C (
2001
) Neuromagnetic source localization of auditory evoked fields and intracerebral evoked potentials: a comparison of data in the same patients.
Clin Neurophysiol
 
112
:
1850
–1859.
Godfrey JJ, Syrdal-Lasky AK, Millay KK, Knox CM (
1981
) Performance of dyslexic children on speech perception tests.
J Exp Child Psychol
 
32
:
401
–424.
Goswami U (
2000
) Phonological representations, reading development and dyslexia: towards a cross-linguistic theoretical framework.
Dyslexia
 
6
:
133
–151.
Goswami U (
2003
) Why theories about developmental dyslexia require developmental designs.
Trends Cogn Sci
 
7
:
534
–540.
Goswami U, Thomson J, Richardson U, Stainthorp R, Hughes D, Rosen S, Scott SK (
2002
) Amplitude envelope onsets and developmental dyslexia: a new hypothesis.
Proc Natl Acad Sci USA
 
99
:
10911
–10916.
Habib M (
2000
) The neurological basis of developmental dyslexia: an overview and working hypothesis.
Brain
 
123
:
2373
–2399.
Hari R, Kiesila P (
1996
) Deficit of temporal auditory processing in dyslexic adults.
Neurosci Lett
 
205
:
138
–140.
Hari R, Renvall H (
2001
) Impaired processing of rapid stimulus sequences in dyslexia.
Trends Cogn Sci
 
5
:
525
–532.
Hari R, Pelizzone M, Makela JP, Hallstrom J, Leinonen L, Lounasmaa OV (
1987
) Neuromagnetic responses of the human auditory cortex to on- and offsets of noise bursts.
Audiology
 
26
:
31
–43.
Heath SM, Hogben JH, Clark CD (
1999
) Auditory temporal processing in disabled readers with and without oral language delay.
J Child Psychol Psychiatry
 
40
:
637
–647.
Heiervang E, Stevenson J, Hugdahl K (
2002
) Auditory processing in children with dyslexia.
J Child Psychol Psychiatry
 
43
:
931
–938.
Hill NI, Bailey PJ, Griffiths YM, Snowling MJ (
1999
) Frequency acuity and binaural masking release in dyslexic listeners.
J Acoust Soc Am
 
106
:
L53
–L58.
Humphreys P, Kaufmann WE, Galaburda AM (
1990
) Developmental dyslexia in women: neuropathological findings in three patients.
Ann Neurol
 
28
:
727
–738.
Hynd GW, Semrud-Clikeman M, Lorys AR, Novey ES, Eliopoulos D (
1990
) Brain morphology in developmental dyslexia and attention deficit disorder/hyperactivity.
Arch Neurol
 
47
:
919
–926.
King WM, Lombardino LJ, Crandell CC, Leonard CM (
2003
) Comorbid auditory processing disorder in developmental dyslexia.
Ear Hear
 
24
:
448
–456.
Kushch A, Gross-Glenn K, Jallad B, Lubs H, Rabin M, Feldman E, Duara R (
1993
) Temporal lobe surface area measurements on MRI in normal and dyslexic readers.
Neuropsychologia
 
31
:
811
–821.
Laguitton V, DeGraaf JB, Cazals Y, Marquis P, Chauvel P, Liégeois-Chauvel C (
1997
) Traitement d'un indice acoustique de la parole naturelle: le délai de voisement. In: Forum Annuel de Neuropsychologie (Platel H, Lambert J, & Eustache F, Eds.), pp. 99–113. Marseille: Solal.
Larsen JP, Hoien T, Lundberg I, Odegaard H (
1990
) MRI evaluation of the size and symmetry of the planum temporale in adolescents with developmental dyslexia.
Brain Lang
 
39
:
289
–301.
Lefavrais P (
1967
) Test de l'Alouette. Paris: Edition du Centre de Psychologie Appliquée.
Leonard CM, Voeller KK, Lombardino LJ, Morris MK, Hynd GW, Alexander AW, Andersen HG, Garofalakis M, Honeyman JC, Mao J, et al. (
1993
) Anomalous cerebral structure in dyslexia revealed with magnetic resonance imaging.
Arch Neurol
 
50
:
461
–469.
Leonard CM, Eckert MA, Lombardino LJ, Oakland T, Kranzler J, Mohr CM, King WM, Freeman A (
2001
) Anatomical risk factors for phonological dyslexia.
Cereb Cortex
 
11
:
148
–157.
Liégeois-Chauvel C, Musolino A, Badier JM, Marquis P, Chauvel P (
1994
) Evoked potentials recorded from the auditory cortex in man: evaluation and topography of the middle latency components.
Electroencephalogr Clin Neurophysiol
 
92
:
204
–214.
Liégeois-Chauvel C, DeGraaf JB, Laguitton V, Chauvel P (
1999
) Specialization of left auditory cortex for speech perception in man depends on temporal coding.
Cereb Cortex
 
9
:
484
–496.
Liégeois-Chauvel C, Giraud K, Badier JM, Marquis P, Chauvel P (
2001
) Intracerebral evoked potentials in pitch perception reveal a functional asymmetry of human auditory cortex.
Ann NY Acad Sci
 
930
:
117
–132.
Luck SJ and Hillyard SA (
1994
) Electrophysiological correlates of feature analysis during visual search.
Psychophysiology
 
31
:
291
–308.
Maassen B, Groenen P, Crul T, Assman-Hulsmans C, Gabreëls F (
2001
) Identification and discrimination of voicing and place-of-articulation in developmental dyslexia.
Clin Linguist Phon
 
15
:
319
–339.
McAnally KI, Stein JF (
1996
) Auditory temporal coding in dyslexia.
Proc R Soc Lond B Biol Sci
 
263
:
961
–965.
McAnally KI, Castles A, Stuart GW (
2000
) Visual and auditory processing impairments in subtypes of developmental dyslexia: a discussion.
J Dev Phys Disabil
 
12
:
145
–156.
McArthur GM, Bishop DVM (
2001
) Auditory perceptual processing in people with reading and oral language impairments: current issues and recommendations.
Dyslexia
 
7
:
150
–170.
McGee T, Kraus N, King C, Nicol T (
1996
) Acoustic elements of speechlike stimuli are reflected in surface recorded responses over the guinea pig temporal lobe.
J Acoust Soc Am
 
99
:
3606
–3614.
Manis FR, McBride-Chang C, Seidenberg MS, Keating P, Doi LM, Munson B, Petersen A (
1997
) Are speech perception deficits associated with developmental dyslexia?
J Exp Child Psychol
 
66
:
211
–235.
Nagarajan S, Mahncke H, Salz T, Tallal P, Roberts T, Merzenich MM (
1999
) Cortical auditory signal processing in poor readers.
Proc Natl Acad Sci USA
 
96
:
6483
–6488.
Oldfield RC (
1971
) The assessment and analysis of handedness: the Edinburgh inventory.
Neuropsychologia
 
9
:
97
–113.
Papanicolaou AC, Castillo E, Breier JL, Davis RN, Simos PG, Diehl RL (
2003
) Differential brain activation patterns during perception of voice and tone onset time series: a MEG study.
Neuroimage
 
18
:
448
–459.
Penhune VB, Zatorre RJ, MacDonald JD, Evans AC (
1996
) Interhemispheric anatomical differences in human primary auditory cortex: probabilistic mapping and volume measurement from magnetic resonance scans.
Cereb Cortex
 
6
:
661
–672.
Pisoni DB (
1977
) Identification and discrimination of the relative onset time of two component tones: implications for voicing perception in stops.
J Acoust Soc Am
 
61
:
1352
–1361.
Ramus F, Rosen S, Dakin SC, Day BL, Castellote JM, White S, Frith U (
2003
) Theories of developmental dyslexia: insights from a multiple case study of dyslexic adults.
Brain
 
126
:
841
–865.
Raven JC, Court JH, Raven J (
1992
) Standard progressive matrices. Oxford: Oxford Psychologists Press.
Reed MA (
1989
) Speech perception and the discrimination of brief auditory cues in reading disabled children.
J Exp Child Psychol
 
48
:
270
–292.
Roman S, Canevet G, Lorenzi C, Triglia JM, Liégeois-Chauvel C (
2004
) Voice onset time encoding in patients with left and right cochlear implants.
Neuroreport
 
15
:
601
–605.
Rosen S (
2003
) Auditory processing in dyslexia and specific language impairment: is there a deficit? What is its nature? Does it explain anything?
J Phonetics
 
31
:
509
–527.
Rosen S, Manganari E (
2001
) Is there a relationship between speech and nonspeech auditory processing in children with dyslexia?
J Speech Lang Hear Res
 
44
:
720
–736.
Rumsey JM, Andreason P, Zametkin AJ, Aquino T, King AC, Hamburger SD, Pikus A, Rapoport JL, Cohen RM (
1992
) Failure to activate the left temporoparietal cortex in dyslexia: an oxygen 15 positron emission tomographic study.
Arch Neurol
 
49
:
527
–534 [published erratum appears in Arch Neurol 1994;51:243].
Schulte-Korne G, Deimel W, Bartling J, Remschmidt H (
1999
) Pre-attentive processing of auditory patterns in dyslexic human subjects.
Neurosci Lett
 
276
:
41
–44.
Serniclaes W, Sprenger-Charolles L, Carre R, Demonet JF (
2001
) Perceptual discrimination of speech sounds in developmental dyslexia.
J Speech Lang Hear Res
 
44
:
384
–399.
Serniclaes W, Van Heghe S, Mousty P, Carre R, Sprenger-Charolles L (
2004
) Allophonic mode of speech perception in dyslexia.
J Exp Child Psychol
 
87
:
336
–361.
Simos PG, Breier JI, Zouridakis G, Papanicolaou AC (
1998
a) Identification of language-related brain activity using magnetoencephalography.
J Clin Exp Neuropsychol
 
20
:
705
–722.
Simos PG, Breier JI, Zouridakis G, Papanicolaou AC (
1998
b) Assessment of cerebral dominance for language using magnetoencephalography.
J Clin Neurophysiol
 
15
:
364
–372.
Simos PG, Breier JI, Fletcher JM, Bergman E, Papanicolaou AC (
2000
a) Cerebral mechanisms involved in word reading in dyslexic children: a magnetic source imaging approach.
Cereb Cortex
 
10
:
809
–816.
Simos PG, Breier JI, Fletcher JM, Foorman BR, Bergman E, Fishbeck K, Papanicolaou AC (
2000
b) Brain activation profiles in dyslexic children during non-word reading: a magnetic source imaging study.
Neurosci Lett
 
290
:
61
–65.
Simos PG, Fletcher JM, Bergman E, Breier JI, Foorman BR, Castillo EM, Davis RN, Fitzgerald M, Papanicolaou AC (
2002
) Dyslexia-specific brain activation profile becomes normal following successful remedial training.
Neurology
 
58
:
1203
–1213.
Steffens ML, Eilers R, Gross-Glenn K, Jallad B (
1992
) Speech perception in adult subjects with familial dyslexia.
J Speech Hear Res
 
35
:
192
–200.
Steinschneider M, Schroeder C, Arezzo JC, Vaughan HG (
1994
) Speech-evoked activity in primary auditory cortex: effects of voice onset time.
Electroenceph Clin Neurophysiol
 
92
:
30
–43.
Steinschneider M, Schroeder C, Arezzo JC, Vaughan HG (
1995
) Physiologic correlates of the voice onset time (VOT) boundary in primary auditory cortex (A1) of the awake monkey: temporal response patterns.
Brain Lang
 
48
:
326
–340.
Steinschneider M, Volkov IO, Noh MD, Garell PC, Howard MA (
1999
) Temporal encoding of the voice onset time phonetic parameter by field potentials recorded directly from human auditory cortex.
J Neurophysiol
 
82
:
2346
–2357.
Steinschneider M, Fishman YI, Arezzo JC (
2003
) Representation of the voice onset time (VOT) speech parameter in population responses within primary auditory cortex of the awake monkey.
J Acoust Soc Am
 
114
:
307
–321.
Steinschneider M, Volkov IO, Fishman YI, Oya H, Arezzo JC, Howard MA 3rd (
2004
) Intracortical responses in human and monkey preimary auditory cortex support a temporal processing mechanism for encoding of the voice onset time phonetic parameter. Cereb Cortex Advance Access, July 6.
Talcott J, Witton C, McLean M, Hansen P, Rees A, Green G, Stein J (
1999
) Auditory FM sensitivity predicts children's phonological and reading skills.
Neuroreport
 
10
:
2045
–2050.
Tallal P (
1980
) Auditory temporal perception, phonics, and reading disabilities in children.
Brain Lang
 
9
:
182
–198.
Temple E, Poldrack RA, Salidis J, Deutsch GK, Tallal P, Merzenich MM, Gabrieli JD (
2001
) Disrupted neural responses to phonological and orthographic processing in dyslexic children: an fMRI study.
Neuroreport
 
12
:
299
–307.
Verkindt C, Bertrand O, Thevenet M, Pernier J (
1994
) Two auditory components in the 130–230 ms range disclosed by their stimulus frequency dependence.
Neuroreport
 
5
:
1189
–1192.
Werker J, Tees R (
1987
) Speech perception in severely disabled and average reading children.
Can J Psychol
 
41
:
48
–61.
Witton C, Talcott JB, Hansen PC, Richardson AJ, Griffiths TD, Rees A, Stein JF, Green GG (
1998
) Sensitivity to dynamic auditory and visual stimuli predicts nonword reading ability in both dyslexic and normal readers.
Curr Biol
 
8
:
791
–797.
Zatorre RJ, Belin P (
2001
) Spectral and temporal processing in human auditory cortex.
Cereb Cortex
 
11
:
946
–953.

Author notes

1INSERM EMI-U 9926, Faculté de Médecine, Marseilles, France, 2INSERM U455, Hôpital Purpan, Toulouse, France and 3Hôpital La Timone, Department of Pediatric Neurology, Marseilles, France