Claudia Männel

Infants acquire the sound categories of their native language within the first year of life, despite the lack of invariant acoustic features that define these categories across speakers and contexts. Given infants’ ability to process voice information and their preference for the familiar maternal voice, we investigated whether voice familiarity supports the acquisition of phonemes. Specifically, we tested whether infants learn a difficult native vowel contrast better from a familiar speaker compared to an unfamiliar speaker. During a 5-minute phoneme training, 71 3-month-old German-learning infants were exposed to a difficult native vowel contrast (/ɪ/–/ɛ/), spoken either by a previously familiarized speaker or by an unfamiliar speaker. In the subsequent test phase, measuring the electrophysiological mismatch response, we assessed infants’ ability to differentiate the vowels when spoken by either the training speaker or a novel speaker. Using Bayesian analysis, we found evidence against improved vowel discrimination after training with a familiar speaker. Instead, it was training with an unfamiliar speaker that resulted in facilitated vowel discrimination for a novel speaker at test. Our findings suggest that infants may not necessarily benefit from a familiar speaker during phoneme learning, particularly when the speaker is not a caregiver and thus lacks an emotional connection. Learning from an unfamiliar speaker, however, may enhance generalization across speakers, potentially due to increased attention driven by novelty.

Starting in utero, infants perceive voice information and preferentially attend to the maternal voice. In adults, speaker familiarity facilitates word and phoneme perception. The current study investigated whether a similar voice-familiarity benefit supports phoneme perception in early infancy, and whether this benefit is driven by acoustic familiarity with speaker voice, prior exposure to phoneme exemplars, or their combination. We tested 72 German-learning 3-month-olds, who were familiarized at home with two speakers. In the lab, infants received 5 minutes of training on a difficult native vowel contrast (/ɛ/–/ɪ/) from either one of the familiarized speakers or an unfamiliar speaker. Discrimination was tested with the electrophysiological mismatch response using three talkers: (1) the phoneme training speaker (familiarized or unfamiliar, depending on group), (2) the familiarized non-training speaker, and (3) a novel speaker (i.e., an unfamiliarized non-training speaker). This design dissociated acoustic familiarity (at-home familiarization) from phoneme-exemplar exposure (training tokens). Preregistered contrasts provided evidence against discrimination differences across test speakers when training was delivered by the unfamiliar speaker; this training might promote speaker-independent generalization, as predicted by abstractionist accounts of speech perception. By contrast, exploratory analyses revealed a voice-familiarity benefit in the group trained by the familiar speaker; this training may drive infants to rely more strongly on speaker-specific cues. Thus, the voice-familiarity benefit only arose when acoustic familiarity and exemplar-based familiarity were combined, suggesting that neither form alone was sufficient. Together, these results are consistent with hybrid accounts of speech perception in which exemplar-like and abstract representations coexist.

Auditory discrimination can be electrophysiologically measured by the mismatch response (MMR). Whereas in adults, this response occurs with a negative polarity (mismatch negativity; MMN), in infants it may either occur with a positive (pMMR) or a negative polarity (nMMR). Infant age has been identified as a key factor influencing MMR polarity, as the pMMR seems to change towards an adult-like nMMR with increasing age. It has also been suggested that pMMR and nMMR reflect different underlying processes, and growing evidence suggests that age triggers these processes in concert with other individual and methodological factors. The current quantitative review assessed the factors influencing the discrimination of speech sound contrasts in 0- to 24-month-old typically developing monolingual infants, with a focus on vowels, consonants as well as lexical tones. We found that the MMR polarity is affected not only by age but also by interindividual, experimental, and data-analysis factors. Based on our findings, we offer methodological considerations for developmental EEG researchers for each study phase. In sum, our systematic approach disentangling the factors affecting the MMR polarity elucidates the processes underlying pMMR and nMMR and thus informs future study design and data interpretation in the widespread application of the MMR in developmental studies.

Objective: Sex hormones testosterone and estradiol have been related to children’s language development. Expanding the focus on dehydroepiandrosterone (DHEA), which has not yet been considered as a biological marker of language ability, may provide novel insights, as emerging evidence suggests that fetal DHEA plays a critical role in the organization of the neonatal brain, potentially shaping later language development. The present study investigated whether fetal DHEA, compared to fetal testosterone, is associated with infant language development. Design and methods: DHEA and testosterone concentrations were measured in newborn hair strands ( n = 63) collected two weeks after birth, capturing fetal long-term hormone secretions during the third trimester of pregnancy. At six months of age, children’s language abilities were assessed using the German version of the Bayley Scales of Infant Development. Results. Linear regression analysis revealed fetal DHEA levels to be significantly associated with language abilities at six months of age, with lower DHEA levels corresponding to higher language scores. Control analyses assessing general cognitive abilities showed no association of fetal DHEA levels with infant cognitive function. Testosterone levels were not associated with language. Conclusions: The current study identifies fetal DHEA levels extracted from newborn hair samples as a potential biological factor influencing infant language development.

Language outcomes of children with hearing loss remain heterogeneous despite recent advances in treatment and intervention. Consonants with high frequency, in particular, continue to pose challenges to affected children’s speech perception and production. In this review, we evaluate findings of how enriched child-directed speech and song might function as a form of early family-centered intervention to remedy the effects of hearing loss on consonant acquisition already during infancy. First, we review the developmental trajectory of consonant acquisition and how it is impeded by permanent pediatric hearing loss. Second, we assess how phonetic-prosodic and lexico-structural features of caregiver speech and song could facilitate acquisition of consonants in the high frequency range. Finally, recommendations for clinical routines and further research are expressed.

Evidence for sequential associative word learning in the auditory domain has been identified in infants, but adults have proven to have difficulties in this type of word learning. To investigate the role of auditory expertise and auditory stimuli salience in the association of auditory objects and their labels, we tested in the first experiment auditorily-trained musicians versus athletes as a high-level control group and in the second experiment stimulus ordering, contrasting object-label versus label-object presentation. Learning was evaluated from electrophysiological (EEG) recordings during training and subsequent testing phases, as well as accuracy-judgement responses during test. Results revealed for musicians a late positive component in the EEG during testing, neither N400 nor behavioral effects were found at test, while athletes did not show any effect of learning. Moreover, the object-label-ordering group only exhibited emerging association effects during training, while the label-object-ordering group showed an EEG effect during test as well as above chance accuracy-judgement scores. Thus, our results suggest an advantage of musical expertise and more salient label-object-ordering in auditory associative word learning in adults. Implications with respect to developmental differences in auditory word learning are discussed.

During language acquisition, infant speech perception becomes selective for the speech sounds of their native language. It is unclear how infants can infer native speech sounds given a critical neurobiological limitation:The infant brain is too slow to match the rate of speech sounds-or phonemes-in natural speech. Infant brains are characterized exclusively by slow electrophysiological activity. Here we show that early categorical acquisition hinges on the timing of phonological features. In particular, the order of acquisition seems to depend on the duration of features in speech. We employ deconvolution modeling of EEG recorded during speech processing from a large sample of children across the first five years of life. From about 14 months of age, the EEG indicates neural sensitivity to categories that are continuous across sequences of multiple subsequent phonemes, thus fitting infants' long temporal receptive windows. Features that alternate at faster rates are added step by step, with the fastest rates acquired last.Categorical sensitivity was not observed for a non-native baseline. Trajectories of acoustic processing did not differ across native- and non-native languages, suggesting that emerging categorical processing was not driven by an enhancement of auditory processing abilities.Category-continuous stretches may thus allow categorical knowledge to emerge in spite of slow electrophysiology. Early language acquisition may rely on a symmetric relationship between electrophysiological resolution and the timing of the to-be-acquired linguistic units.

Infants master temporal patterns of their native language at a developmental trajectory from slow to fast:Shortly after birth, they recognize the slow acoustic modulations specific to their native language before tuning into faster language-specific patterns between 6-12 months of age. We here propose that this trajectory is constrained by neuronal maturation-in particular, the gradual emergence of high-frequency neural oscillations in the infant electroencephalogram. Infants' initial focus on slow prosodic modulations is consistent with the prenatal availability of slow electrophysiological activity (i.e., theta-- and delta-band oscillations). Our proposal is consistent with the temporal patterns of infant-directed speech, which initially amplifies slow modulations, approaching the faster modulation range of adult-directed speech only as infants' language has advanced sufficiently. Moreover, our proposal agrees with evidence from premature infants showing maturational age is a stronger predictor of language development than ex-utero exposure to speech, indicating that premature infants cannot exploit their earlier availability of speech because of electrophysiological constraints.In sum, we provide a new perspective on language acquisition emphasizing neuronal development as critical driving force of infants' language development.