Processing Of Speech Sounds Research Articles

Neural processing of speech sounds at premature and term birth: ERPs and MMR between 32 and 42 weeks of gestation

Prenatal listening experience reportedly modulates how humans process speech at birth, but little is known about how speech perception develops throughout the perinatal period. The present experiment assessed the neural event-related potentials (ERP) and mismatch responses (MMR) to native vowels in 99 neonates born between 32 and 42 weeks of gestation. The vowels elicited reliable ERPs in newborns whose gestational age at time of experiment was at least 36 weeks and 1 day (36 + 1). The ERPs reflected spectral distinctions between vowel onsets from age 36 weeks + 6 days and durational distinctions at vowel offsets from age 37 weeks + 6 days. Starting at age 40 + 4, there was evidence of neural discrimination of vowel length, indexed by a negative MMR response. The present findings extend our understanding of the earliest stages of speech perception development in that they pinpoint the ages at which the cortex reliably responds to the phonetic characteristics of individual speech sounds and discriminates a native phoneme contrast. The age at which the brain reliably differentiates vowel onsets coincides with what is considered term age in many countries (37 weeks + 0 days of gestational age). Future studies should investigate to what extent the perinatal maturation of the cortical responses to speech sounds is modulated by the ambient language.

Language networks of normal-hearing infants exhibit topological differences between resting and steady states: An fNIRS functional connectivity study.

Task-related studies have consistently reported that listening to speech sounds activate the temporal and prefrontal regions of the brain. However, it is not well understood how functional organization of auditory and language networks differ when processing speech sounds from its resting state form. The knowledge of language network organization in typically developing infants could serve as an important biomarker to understand network-level disruptions expected in infants with hearing impairment. We hypothesized that topological differences of language networks can be characterized using functional connectivity measures in two experimental conditions (1) complete silence (resting) and (2) in response to repetitive continuous speech sounds (steady). Thirty normal-hearing infants (14 males and 16 females, age: 7.8 ± 4.8 months) were recruited in this study. Brain activity was recorded from bilateral temporal and prefrontal regions associated with speech and language processing for two experimental conditions: resting and steady states. Topological differences of functional language networks were characterized using graph theoretical analysis. The normalized global efficiency and clustering coefficient were used as measures of functional integration and segregation, respectively. We found that overall, language networks of infants demonstrate the economic small-world organization in both resting and steady states. Moreover, language networks exhibited significantly higher functional integration and significantly lower functional segregation in resting state compared to steady state. A secondary analysis that investigated developmental effects of infants aged 6-months or below and above 6-months revealed that such topological differences in functional integration and segregation across resting and steady states can be reliably detected after the first 6-months of life. The higher functional integration observed in resting state suggests that language networks of infants facilitate more efficient parallel information processing across distributed language regions in the absence of speech stimuli. Moreover, higher functional segregation in steady state indicates that the speech information processing occurs within densely interconnected specialized regions in the language network.

Native language background affects the perception of duration and pitch

Estonian is a quantity language with both a primary duration cue and a secondary pitch cue, whereas Chinese is a tonal language with a dominant pitch use. Using a mismatch negativity experiment and a behavioral discrimination experiment, we investigated how native language background affects the perception of duration only, pitch only, and duration plus pitch information. Chinese participants perceived duration in Estonian as meaningless acoustic information due to a lack of phonological use of duration in their native language; however, they demonstrated a better pitch discrimination ability than Estonian participants. On the other hand, Estonian participants outperformed Chinese participants in perceiving the non-speech pure tones that resembled the Estonian quantity (i.e., containing both duration and pitch information). Our results indicate that native language background affects the perception of duration and pitch and that such an effect is not specific to processing speech sounds.

Individual differences in the consistency of neural and behavioural responses to speech sounds

There are documented individual differences among adults in the consistency of speech sound processing, both at neural and behavioural levels. Some adults show more consistent neural responses to speech sounds than others, as measured by an event-related potential called the frequency-following response (FFR); similarly, some adults show more consistent behavioural responses to native speech sounds than others, as measured by two-alternative forced choice (2AFC) and visual analog scaling (VAS) tasks. Adults also differ in how successfully they can perceive non-native speech sounds. Interestingly, it remains unclear whether these differences are related within individuals. In the current study, native English-speaking adults completed native phonetic perception tasks (2AFC and VAS), a non-native (German) phonetic perception task, and an FFR recording session. From these tasks, we derived measures of the consistency of participants’ neural and behavioural responses to native speech as well as their non-native perception ability. We then examined the relationships among individual differences in these measures. Analysis of the behavioural measures revealed that more consistent responses to native sounds predicted more successful perception of unfamiliar German sounds. Analysis of neural and behavioural data did not reveal clear relationships between FFR consistency and our phonetic perception measures. This multimodal work furthers our understanding of individual differences in speech processing among adults, and may eventually lead to individualized approaches for enhancing non-native language acquisition in adulthood.

Prismatic adaptation coupled with cognitive training as novel treatment for developmental dyslexia: a randomized controlled trial

Despite intense and costly treatments, developmental dyslexia (DD) often persists into adulthood. Several brain skills unrelated to speech sound processing (i.e., phonology), including the spatial distribution of visual attention, are abnormal in DD and may represent possible treatment targets. This study explores the efficacy in DD of rightward prismatic adaptation (rPA), a visuomotor adaptation technique that enables visuo-attentive recalibration through shifts in the visual field induced by prismatic goggles. A digital intervention of rPA plus cognitive training was delivered weekly over 10 weeks to adolescents with DD (aged 13–17) assigned either to treatment (N = 35) or waitlist (N = 35) group. Efficacy was evaluated by repeated measures MANOVA assessing changes in working memory index (WMI), processing speed index (PSI), text reading speed, and words/pseudowords reading accuracy. rPA treatment was significantly more effective than waitlist (p ≤ 0.001; ηp2 = 0.815). WMI, PSI, and reading speed increased in the intervention group only (p ≤ 0.001, ηp2 = 0.67; p ≤ 0.001, ηp2 = 0.58; p ≤ 0.001, ηp2 = 0.29, respectively). Although modest change was detected for words and pseudowords accuracy in the waitlist group only (words: p ≤ 0.001, d = 0.17, pseudowords: p = 0.028; d = 0.27), between-group differences were non-significant. rPA-coupled cognitive training enhances cognitive and reading abilities in adolescents with DD. This innovative approach could have implications for early remedial treatment.

Auditory-inspired adaptive frequency tracking

One of the best frequency trackers to date is the human (mammalian) auditory system, which has evolved through millions of years of resolving classification problems. It is a versatile, elegant and powerful sound processing unit. It excels in detecting, estimating, and classifying multiple targets simultaneously even in noisy environments. Hence, mimicking even some features of the auditory system could be beneficial in developing superior frequency tracking and classification algorithms. An auditory inspired adaptive synchrony capture filterbank (SCFB) signal processing architecture for tracking signal frequency components was proposed in a related paper [JASA-2013]. The SCFB architecture consists of a fixed array of traditional, passive linear, gammatone filters in cascade with a bank of three adaptively tunable bandpass filters that form a frequency-discriminator-loop (FDL). The SCFB exhibits many desirable properties for processing speech, music, and other complex sounds. In recent work [Dec2021], the algorithm was modified using adaptive tuning parameters, and a generalized way to determine/suppress voiced and unvoiced (silent) regions. This modified algorithm estimates frequencies with higher accuracy even in the presence of closely spaced input tones. Preliminary analysis with synthetic, human speech and humpback/whale-call signals demonstrates that the revised algorithm performs well. This talk will focus on follow-on work.

Resolving competing predictions in speech: How qualitatively different cues and cue reliability contribute to phoneme identification.

Listeners have many sources of information available in interpreting speech. Numerous theoretical frameworks and paradigms have established that various constraints impact the processing of speech sounds, but it remains unclear how listeners might simultaneously consider multiple cues, especially those that differ qualitatively (i.e., with respect to timing and/or modality) or quantitatively (i.e., with respect to cue reliability). Here, we establish that cross-modal identity priming can influence the interpretation of ambiguous phonemes (Exp. 1, N = 40) and show that two qualitatively distinct cues - namely, cross-modal identity priming and auditory co-articulatory context - have additive effects on phoneme identification (Exp. 2, N = 40). However, we find no effect of quantitative variation in a cue - specifically, changes in the reliability of the priming cue did not influence phoneme identification (Exp. 3a, N = 40; Exp. 3b, N = 40). Overall, we find that qualitatively distinct cues can additively influence phoneme identification. While many existing theoretical frameworks address constraint integration to some degree, our results provide a step towards understanding how information that differs in both timing and modality is integrated in online speech perception.

Can speech perception deficits cause phonological impairments? Evidence from short-term memory for ambiguous speech.

Poor performance on phonological tasks is characteristic of neurodevelopmental language disorders (dyslexia and/or developmental language disorder). Perceptual deficit accounts attribute phonological dysfunction to lower-level deficits in speech-sound processing. However, a causal pathway from speech perception to phonological performance has not been established. We assessed this relationship in typical adults by experimentally disrupting speech-sound discrimination in a phonological short-term memory (pSTM) task. We used an automated audio-morphing method (Rogers & Davis, 2017) to create ambiguous intermediate syllables between 16 letter name-letter name ("B"-"P") and letter name-word ("B"-"we") pairs. High- and low-ambiguity syllables were used in a pSTM task in which participants (N = 36) recalled six- and eight-letter name sequences. Low-ambiguity sequences were better recalled than high-ambiguity sequences, for letter name-letter name but not letter name-word morphed syllables. A further experiment replicated this ambiguity cost (N = 26), but failed to show retroactive or prospective effects for mixed high- and low-ambiguity sequences, in contrast to pSTM findings for speech-in-noise (SiN; Guang et al., 2020; Rabbitt, 1968). These experiments show that ambiguous speech sounds impair pSTM, via a different mechanism to SiN recall. We further show that the effect of ambiguous speech on recall is context-specific, limited, and does not transfer to recall of nonconfusable items. This indicates that speech perception deficits are not a plausible cause of pSTM difficulties in language disorders. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Probing the phonological-coding deficit in dyslexia with vowel length perception

There is an established linkage between dyslexia and anomalous processing of speech sounds. We probed this in a Danish language context with vowel length tasks based on “kugle” ball /ku:lə/ and sequential deletions of the vowel portion to yield “kulde” coldness /kulə/. Vowel length continua are methodologically advantageous as they do not involve the perception of a sudden phonetic change that may instead tap other auditory processing abilities. Identification and discrimination tasks were administered to tertiary (n = 28), reading impaired (n = 26), and lower secondary students (n = 20), and the latter were approximately aged matched to the reading-impaired group. Identification functions derived from regression modelling of the responses showed that the dyslexics had significantly flatter curves than the other groups. Moreover, the secondary and dyslexic groups differed at the long vowel extremity of the continuum. Discrimination results showed that mean peak sensitivity of the tertiary students was higher than that of the secondary and dyslexic students. These results indicate that the phonological-coding deficit observed in dyslexics may be indexed by vowel length identification. Furthermore, identification results suggest that the nature of the phonological-coding deficit concomitant with dyslexia may stem from a lack of precision in processing the minimally modified longer vowel stimuli.

Associations between rapid auditory processing of speech sounds and specific verbal communication skills in autism.

The ability to rapidly process speech sounds is integral not only for processing other's speech, but also for auditory processing of one's own speech, which allows for maintenance of speech accuracy. Deficits in rapid auditory processing have been demonstrated in autistic individuals, particularly those with language impairment. We examined rapid auditory processing for speech sounds in relation to performance on a battery of verbal communication measures to determine which aspects of verbal communication were associated with cortical auditory processing in a sample of individuals with autism. Participants were 57 children and adolescents (40 male and 17 female) ages 5-18 who were diagnosed with an Autism Spectrum Disorder (ASD). Rapid auditory processing of speech sounds was measured via a magnetoencephalographic (MEG) index of the quality of the auditory evoked response to the second of two differing speech sounds ("Ga" / "Da") presented in rapid succession. Verbal communication abilities were assessed on standardized clinical measures of overall expressive and receptive language, vocabulary, articulation, and phonological processing. Associations between cortical measures of left- and right-hemisphere rapid auditory processing and verbal communication measures were examined. Rapid auditory processing of speech sounds was significantly associated with speech articulation bilaterally (r = 0.463, p = 0.001 for left hemisphere and r = 0.328, p = 0.020 for right hemisphere). In addition, rapid auditory processing in the left hemisphere was significantly associated with overall expressive language abilities (r = 0.354, p = 0.013); expressive (r = 0.384, p = 0.005) vocabulary; and phonological memory (r = 0.325, p = 0.024). Phonological memory was found to mediate the relationship between rapid cortical processing and receptive language. These results demonstrate that impaired rapid auditory processing for speech sounds is associated with dysfunction in verbal communication in ASD. The data also indicate that intact rapid auditory processing may be necessary for even basic communication skills that support speech production, such as phonological memory and articulatory control.

Neurodevelopmental trajectories of letter and speech sound processing from preschool to the end of elementary school

Learning to read alphabetic languages starts with learning letter–speech-sound associations. How this process changes brain function during development is still largely unknown. We followed 102 children with varying reading skills in a mixed-longitudinal/cross-sectional design from the prereading stage to the end of elementary school over five time points (n = 46 with two and more time points, of which n = 16 fully-longitudinal) to investigate the neural trajectories of letter and speech sound processing using fMRI. Children were presented with letters and speech sounds visually, auditorily, and audiovisually in kindergarten (6.7yo), at the middle (7.3yo) and end of first grade (7.6yo), and in second (8.4yo) and fifth grades (11.5yo). Activation of the ventral occipitotemporal cortex for visual and audiovisual processing followed a complex trajectory, with two peaks in first and fifth grades. The superior temporal gyrus (STG) showed an inverted U-shaped trajectory for audiovisual letter processing, a development that in poor readers was attenuated in middle STG and absent in posterior STG. Finally, the trajectories for letter-speech-sound integration were modulated by reading skills and showed differing directionality in the congruency effect depending on the time point. This unprecedented study captures the development of letter processing across elementary school and its neural trajectories in children with varying reading skills.

Musical Aptitude and the Acquisition of Pitch and Duration in L2 Vowels

Music and language have been believed to be processed similarly in the brain. Studies have compared the processing of vowels and music, as they share spectral and temporal features such as pitch and timing. When comparing music and language skills the question arises whether music skills are inherent, measured with aptitude, or acquired, measured with experience. This article focused on musical aptitude studies. It was found that musical aptitude predicted non-native speech-sound processing better than musical experience. It is proposed that music and language skills are interconnected as greater overlaps in neural processing of music and speech were found for musicians than nonmusicians, and as individuals with no musical experience but with high L2 production skills were found to have high musical aptitude. Furthermore, spectral sensitivity was found to be the primary partial mediator of the link between musicality and non-native speech-sound processing, with temporal sensitivity having no role.

Speech in noise performance in individuals with misophonia and hyperacusis using behavioral and auditory brainstem response

Hyperacusis is reduced tolerance to everyday environmental sounds, whereas misophonia is a relatively unknown disorder in which characterized by intense and excessive emotional responses to specific “trigger” sounds. Few studies have investigated ability of speech in noise (SiN) and examined them using auditory brainstem responses (ABR)—which is used to measure early auditory processing and overall hearing function in human listeners, in particular for the sound sensitivity disorders of misophonia and hyperacusis. The primary aim of this study was to investigate the relationships between SiN performance and ABR components, specifically for group differences. A total of 60 participants were categorized into three groups: 13 with misophonia, 12 with hyperacusis, and 35 control group. Our findings indicated that (1) the misophonia group showed poor speech in noise performance than control and hyperacusis groups at 20 and 5 SNR; (2) both hyperacusis and misophonia groups showed enhanced wave I amplitude compared with the control group; and (3) the misophonia group showed prolonged wave V latency compared to the other two groups. Additionally, delayed ABR wave V latency in misophonia group was correlated with decreased speech-in-noise performance score, suggesting possible delays in processing of speech sounds in this group leading to poorer performance.

Auditory-inspired adaptive frequency tracking

One of the best frequency trackers to date is the human (mammalian) auditory system, which has evolved through millions of years of resolving classification problems. It is a versatile, elegant and powerful sound processing unit. It excels in detecting, estimating, and classifying multiple targets simultaneously even in noisy environments. Hence, mimicking even some features of the auditory system could be beneficial in developing superior frequency tracking and classification algorithms. An auditory inspired adaptive synchrony capture filterbank (SCFB) signal processing architecture for tracking signal frequency components was proposed in a related paper [JASA (2013)]. The SCFB architecture consists of a fixed array of traditional, passive linear, gammatone filters in cascade with a bank of three adaptively tunable bandpass filters that form a frequency-discriminator-loop (FDL). The SCFB exhibits many desirable properties for processing speech, music, and other complex sounds. In recent work (Dec 2021), the algorithm was modified using adaptive tuning parameters, and a generalized way to determine/suppress voiced and unvoiced (silent) regions. This modified algorithm estimates frequencies with higher accuracy even in the presence of closely spaced input tones. Preliminary analysis with synthetic, human speech and humpback/whale-call signals demonstrates that the revised algorithm performs well. This talk will focus on the latest updates.

Abstract WMP26: Acute Left Superior Temporal Damage Results In Phonological Errors In Word Comprehension

Right hemisphere deactivation by right carotid injection of sodium amytal (Wada testing) in chronic epilepsy results in few phonological errors in word comprehension. Based on this finding, recent models of language processing indicate that speech sound and spoken word processing is bilateral, so that unilateral stroke should not affect distinguishing phonologically similar words (e.g. boat/note) for word comprehension. We tested the alternative hypothesis that acute left hemisphere stroke in superior temporal gyrus (STG) results in phonological errors in a word-picture verification task, which would support the proposal that the right hemisphere does not support speech sound processing and word comprehension before recovery through reorganization of structure/function relationships. Methods: A series of 218 participants were tested with a word-picture verification test with targets and phonologically similar words and had brain MRI within 4 days of stroke onset. We evaluated associations between any damage to left STG and phonological errors using chi square, and the correlation between degree of damage to STG and number of errors using Spearman Rho. Results: We found a very strong association between infarct in left STG and the production of phonological errors in word comprehension (chi squared= 31.2; p <0.00001). Furthermore, the number of phonological errors was strongly correlated with the degree of infarct in left STG (rho=0.49; p<0.00001). Results indicate that left STG, when acutely damaged, often results in phonological errors in word comprehension. These results indicate that the paucity of such errors in comprehension in studies of chronic left hemisphere stroke or Wada testing is a result of reorganization following chronic left temporal dysfunction. Patients with acute left STG stroke should be tested with sensitive tests for word comprehension deficits, including phonologically similar words, before giving verbal direction on exam or obtaining informed consent for procedures.

Categorical perception of Mandarin lexical tones in language-delayed autistic children.

Some theories suggested that autistic people have better pitch perception skills than non-autistic people. However, in a context where pitch patterns are used to differentiate word meanings (i.e. lexical tones), autistic people may encounter difficulties, especially those with less language experience. We tested this by asking language-delayed autistic children to identify and discriminate two Mandarin lexical tones (/yi/ with Tone 1, meaning 'clothes'; /yi/ with Tone 2, meaning 'aunt'; /yi/: the standard romanization of Mandarin Chinese). On average, these autistic children were 7.35 years old, but their developmental age in language ability was 4.20, lagging behind 7-year-old non-autistic children in terms of language ability. Autistic children's performance in identifying and discriminating lexical tones was compared with two groups of non-autistic children: one group was matched with the autistic group on age, and the other was matched based on language ability. Autistic children performed differently from the non-autistic children matched on age, while autistic and non-autistic children matched on language ability exhibited seemingly similar performance. However, both the non-autistic groups have developed the perceptual ability to process lexical tones as different categories, but this ability was still developing in autistic children. Finally, we found autistic children who performed worse in identifying lexical tones had poorer language ability. The results suggest that language disability might have adverse influence on the development of skills of speech sound processing.

The development of cortical processing of speech differs between children with cochlear implants and normal hearing and changes with parental singing.

The aim of the present study was to investigate speech processing development in children with normal hearing (NH) and cochlear implants (CI) groups using a multifeature event-related potential (ERP) paradigm. Singing is associated to enhanced attention and speech perception. Therefore, its connection to ERPs was investigated in the CI group. The paradigm included five change types in a pseudoword: two easy- (duration, gap) and three difficult-to-detect (vowel, pitch, intensity) with CIs. The positive mismatch responses (pMMR), mismatch negativity (MMN), P3a and late differentiating negativity (LDN) responses of preschoolers (below 6 years 9 months) and schoolchildren (above 6 years 9 months) with NH or CIs at two time points (T1, T2) were investigated with Linear Mixed Modeling (LMM). For the CI group, the association of singing at home and ERP development was modeled with LMM. Overall, responses elicited by the easy- and difficult to detect changes differed between the CI and NH groups. Compared to the NH group, the CI group had smaller MMNs to vowel duration changes and gaps, larger P3a responses to gaps, and larger pMMRs and smaller LDNs to vowel identity changes. Preschoolers had smaller P3a responses and larger LDNs to gaps, and larger pMMRs to vowel identity changes than schoolchildren. In addition, the pMMRs to gaps increased from T1 to T2 in preschoolers. More parental singing in the CI group was associated with increasing pMMR and less parental singing with decreasing P3a amplitudes from T1 to T2. The multifeature paradigm is suitable for assessing cortical speech processing development in children. In children with CIs, cortical discrimination is often reflected in pMMR and P3a responses, and in MMN and LDN responses in children with NH. Moreover, the cortical speech discrimination of children with CIs develops late, and over time and age, their speech sound change processing changes as does the processing of children with NH. Importantly, multisensory activities such as parental singing can lead to improvement in the discrimination and attention shifting toward speech changes in children with CIs. These novel results should be taken into account in future research and rehabilitation.

One-to-One or One Too Many? Linking Sound-to-Letter Mappings to Speech Sound Perception and Production in Early Readers.

Effects related to literacy acquisition have been observed at different levels of speech processing. This study investigated the link between orthographic knowledge and children's perception and production of specific speech sounds. Sixty Spanish-speaking second graders, differing in their phonological decoding skills, completed a speech perception and a production task. In the perception task, a behavioral adaptation of the oddball paradigm was used. Children had to detect orthographically consistent /t/, which has a unique orthographic representation (〈t〉), and inconsistent /k/, which maps onto three different graphemes (〈c〉, 〈qu〉, and 〈k〉), both appearing infrequently within a repetitive auditory sequence. In the production task, children produced these same sounds in meaningless syllables. Perception results show that all children were faster at detecting consistent than inconsistent sounds regardless of their decoding skills. In the production task, however, the same facilitation for consistent sounds was linked to better decoding skills. These findings demonstrate differences in speech sound processing related to literacy acquisition. Literacy acquisition may therefore affect already-formed speech sound representations. Crucially, the strength of this link in production is modulated by individual decoding skills.

Attentional modulation of interhemispheric (a)symmetry in children with developmental language disorder

The nature of auditory processing problems in children with developmental language disorder (DLD) is still poorly understood. Much research has been devoted to determining the extent to which DLD is associated with general auditory versus language-specific dysfunction. However, less emphasis has been given to the role of different task conditions in these dysfunctions. We explored whether children with DLD demonstrate atypical interhemispheric asymmetry during the auditory processing of speech and non-speech sounds and whether this interhemispheric balance is modulated by attention. Magnetoencephalography was used to record auditory evoked fields in 18 children (9 to 10 years old), 9 with DLD and 9 with language typical development, during active or passive listening to speech and non-speech sounds. A linear mixed model analysis revealed a bilateral effect of attention in both groups. Participants with DLD demonstrated atypical interhemispheric asymmetry, specifically in the later (185–600 ms) time window but only during the passive listening condition. During the active task, the DLD group did not differ from the typically developed children in terms of hemispheric balance of activation. Our results support the idea of an altered interhemispheric balance in passive auditory response properties in DLD. We further suggest that an active task condition, or top–down attention, can help to regain leftward lateralization, particularly in a later stage of activation. Our study highlights the highly dynamic and interhemispheric nature of auditory processing, which may contribute to the variability in reports of auditory language processing deficits in DLD.

TDCS modulates speech perception and production in second language learners

Accurate identification and pronunciation of nonnative speech sounds can be particularly challenging for adult language learners. The current study tested the effects of a brief musical training combined with transcranial direct current stimulation (tDCS) on speech perception and production in a second language (L2). The sample comprised 36 native Hebrew speakers, aged 18–38, who studied English as L2 in a formal setting and had little musical training. Training encompassed musical perception tasks with feedback (i.e., timbre, duration, and tonal memory) and concurrent tDCS applied over the left posterior auditory-related cortex (including posterior superior temporal gyrus and planum temporale). Participants were randomly assigned to anodal or sham stimulation. Musical perception, L2 speech perception (measured by a categorical AXB discrimination task) and speech production (measured by a speech imitation task) were tested before and after training. There were no tDCS-dependent effects on musical perception post-training. However, only participants who received active stimulation showed increased accuracy of L2 phoneme discrimination and greater change in the acoustic properties of L2 speech sound production (i.e., second formant frequency in vowels and center of gravity in consonants). The results of this study suggest neuromodulation can facilitate the processing of nonnative speech sounds in adult learners.