Early Word Segmentation in Naturalistic Environments: Limited Effects of Speech Register

We examined 7.5‐month‐old infants' ability to segment words from infant‐ and adult‐directed speech (IDS and ADS). In particular, we extended the standard design of most segmentation studies by including a phase where infants were repeatedly exposed to target word recordings at their own home (extended exposure) in addition to a laboratory‐based familiarization. This enabled us to examine infants' segmentation of words from speech input in their naturalistic environment, extending current findings to learning outside the laboratory. Results of a modified preferential‐listening task show that infants listened longer to isolated tokens of familiarized words from home relative to novel control words regardless of register. However, infants showed no recognition of words exposed to during purely laboratory‐based familiarization. This indicates that infants succeed in retaining words in long‐term memory following extended exposure and recognizing them later on with considerable flexibility. In addition, infants segmented words from both IDS and ADS, suggesting limited effects of speech register on learning from extended exposure in naturalistic environments. Moreover, there was a significant correlation between segmentation success and infants' attention to ADS, but not to IDS, during the extended exposure phase. This finding speaks to current language acquisition models assuming that infants' individual attention to language stimuli drives successful learning.     

Early Word Segmentation by Infants and Toddlers With Williams Syndrome

This study tested the ability of English infants and toddlers with Williams syndrome to segment, that is, to extract from fluent speech, bisyllabic nouns that had either a strong–‐weak stress pattern (predominant in English), or a weak–‐strong stress pattern. The testing procedure was the same for both types of words: Children were familiarized with instances of isolated nouns, and then tested on their recognition of these nouns embedded in passages. In English, typically developing infants start segmenting strong–‐weak nouns by 7.5 months of age, and weak–‐strong nouns by 10.5 months. Our clinical population was able to segment strong–‐weak nouns, but failed, despite chronological ages above 15 months, to segment weak–‐strong words. These results suggest that the development of word segmentation is seriously delayed in Williams syndrome. This deficit in early phonological processing may contribute to a fuller understanding of the late lexical onset in this population, a phenomenon that had hitherto only been explained in terms of cognitive and semantic deficits.     

Can a Click be a Word?: Infants' Learning of Non‐Native Words

Forms that are nonlinguistic markers in one language (i.e., “tsk‐tsk” in English) may be part of the phoneme inventory—and hence part of words—in another language. In the current paper, we demonstrate that infants' ability to learn words containing unfamiliar language sounds is influenced by the age and vocabulary size of the infant learner, as well as by cues to the speaker's referential intent. When referential cues were available, infants at 14 months learned words with non‐native speech sounds, but at 20 months only those infants with smaller vocabularies succeeded. When no referential cues were present, infants at both 14 and 20 months failed to learn the same words. The implications of the relation between linguistic sophistication and non‐native word learning are discussed.     

Functional Parallelism of Detailed and Rough Speech Processing at the End of Infancy

Although realization of the same speech sound is far from being consistent across different contexts, speech recognition has to rely on phonetic detail in order to detect words. So far, it appeared that young infants cannot avoid noticing subtle speech sound variation whenever it occurs. Only later on, they are able to tolerate speech sound variation in some word recognition tasks. Here, we test whether this ability is associated with the time infants start storing their first word forms. We recorded event‐related potentials (ERPs) in a priming paradigm. German words (targets) followed syllables (primes) with a different amount of phoneme overlap. We tested infants at three, six, and nine months after birth. ERPs reflected sensitivity to prime‐target variation in a single phoneme in three‐month‐olds, tolerance to this in six‐month‐olds, and both processing aspects in nine‐month‐olds. Our findings reveal individual developmental priorities for different aspects of speech processing, with very detailed speech processing dominating at around 3 months, rough processing dominating at around half a year after birth, and an architecture of parallel rough and detailed processing at around 9 months. Functional parallelism at the end of infancy might explain the heterogeneous pattern of results regarding the degree of acoustic detail that toddlers appear to consider at different ages and across different paradigms.     

Successful Word Recognition by 10‐Month‐Olds Given Continuous Speech Both at Initial Exposure and Test

Most words that infants hear occur within fluent speech. To compile a vocabulary, infants therefore need to segment words from speech contexts. This study is the first to investigate whether infants (here: 10‐month‐olds) can recognize words when both initial exposure and test presentation are in continuous speech. Electrophysiological evidence attests that this indeed occurs: An increased extended negativity (word recognition effect) appears for familiarized target words relative to control words. This response proved constant at the individual level: Only infants who showed this negativity at test had shown such a response, within six repetitions after first occurrence, during familiarization.     

English‐Learning Infants' Representations of Word Forms With Iambic Stress

Retaining detailed representations of unstressed syllables is a logical prerequisite for infants' use of probabilistic phonotactics to segment iambic words from fluent speech. The head‐turn preference study was used to investigate the nature of English‐learners' representations of iambic word onsets. Fifty‐four 10.5‐month‐olds were familiarized to passages containing the nonsense iambic word forms ginome and tupong. Following familiarization, infants were either tested on familiar (ginome and tupong) or near‐familiar (pinome and bupong) versus unfamiliar (kidar and mafoos) words. Infants in the familiar test group (familiar vs. unfamiliar) oriented significantly longer to familiar than unfamiliar test items, whereas infants in the near‐familiar test group (near‐familiar vs. unfamiliar) oriented equally long to near‐familiar and unfamiliar test items. Our results provide evidence that infants retain fairly detailed representations of unstressed syllables and therefore support the hypothesis that infants use phonotactic cues to find words in fluent speech.     

Language Experience Shapes the Development of the Mutual Exclusivity Bias

Halberda (2003) demonstrated that 17‐month‐old infants, but not 14‐ or 16‐month‐olds, use a strategy known as mutual exclusivity (ME) to identify the meanings of new words. When 17‐month‐olds were presented with a novel word in an intermodal preferential looking task, they preferentially fixated a novel object over an object for which they already had a name. We explored whether the development of this word‐learning strategy is driven by children’s experience of hearing only one name for each referent in their environment by comparing the behavior of infants from monolingual and bilingual homes. Monolingual infants aged 17–22 months showed clear evidence of using an ME strategy, in that they preferentially fixated the novel object when they were asked to “look at the dax.” Bilingual infants of the same age and vocabulary size failed to show a similar pattern of behavior. We suggest that children who are raised with more than one language fail to develop an ME strategy in parallel with monolingual infants because development of the bias is a consequence of the monolingual child’s everyday experiences with words.     

Boosting the input: 9-month-olds’ sensitivity to low-frequency phonotactic patterns in novel wordforms

To learn their first words, infants must attend to a variety of cues that signal word boundaries. One such cue infants might use is the language-specific phonotactics to track legal combinations and positions of segments within a word. Studies have demonstrated that, when tested across statistically high and low phonotactics, infants repeatedly reject the low-frequency wordforms. We explore whether the capacity to access low-frequency phonotactic combinations is available at 9 months when pre-exposed to wordforms containing statistically low combinations of segments. Using a modified head-turn procedure, one group of infants was presented with nonwords with low-frequency complex onsets (dr-), and another group was presented with zero-frequency onset nonwords (dl-). Following pre-exposure and familiarization, infants were then tested on their ability to segment nonwords that contained either the low- or the zero-frequency onsets. Only infants in the low-frequency condition were successful at the task, suggesting some experience with these onsets supports segmentation.     

Robust effects of stress on early lexical representation

This study aims to elucidate the factors that affect the robustness of word form representations by exploring the relative influence of lexical stress and segmental identity (consonant vs. vowel) on infant word recognition. Our main question was which changes to the words may go unnoticed and which may lead the words to be unrecognizable. One‐hundred 11‐month‐old Hebrew‐learning infants were tested in two experiments using the Central Fixation Procedure. In Experiment 1, 20 infants were presented with iambic Familiar and Unfamiliar words. The infants listened longer to Familiar than to Unfamiliar words, indicating their recognition of frequently heard word forms. In Experiment 2, four groups of 20 infants each were tested in each of four conditions involving altered iambic Familiar words contrasted with iambic Unfamiliar nonwords. In each condition, one segment in the Familiar word was changed—either a consonant or a vowel, in either the first (unstressed) or the second (stressed) syllable. In each condition, recognition of the Familiar words despite the change indicates a less accurate or less well‐specified representation. Infants recognized Familiar words despite changes to the weak (first) syllable, regardless of whether the change involved a consonant or a vowel (conditions 2a, 2c). However, a change of either consonant or vowel in the stressed (second) syllable blocked word recognition (conditions 2b, 2d). These findings support the proposal that stress pattern plays a key role in early word representation, regardless of segmental identity.     

Object Familiarity Enhances Infants’ Use of Phonetic Detail in Novel Words

Infants greatly refine their ability to discriminate language sounds by 12 months, yet 14‐month‐olds appear to confuse similar‐sounding novel words. Two explanations could account for this phenomenon: infants initially have incomplete phoneme representations, suggesting developmental discontinuity; or word‐learning demands interfere with use of established phonetic detail. These hypotheses were tested at 14 months by pairing a novel word with an object preexposed to half the infants and novel to the other half. If demands are key, only preexposed infants should efficiently use phonetic detail; there is no need to concurrently learn object details with the word. If representations lack detail, object familiarity should not matter. Only infants preexposed to the object noticed a change in its label, thus challenging the discontinuity position and demonstrating the impact of object familiarity on early word learning.     

From Syllables to Syntax: Multilevel Statistical Learning by 12‐Month‐Old Infants

To successfully acquire language, infants must be able to track multiple levels of regularities in the input. In many cases, regularities only emerge after some learning has already occurred. For example, the grammatical relationships between words are only evident once the words have been segmented from continuous speech. To ask whether infants can engage in this type of learning process, 12‐month‐old infants in 2 experiments were familiarized with multiword utterances synthesized as continuous speech. The words in the utterances were ordered based on a simple finite‐state grammar. Following exposure, infants were tested on novel grammatical and ungrammatical sentences. The results indicate that the infants were able to perform 2 statistical learning tasks in sequence: first segmenting the words from continuous speech, and subsequently discovering the permissible orderings of the words. Given a single set of input, infants were able to acquire multiple levels of structure, suggesting that multiple levels of representation (initially syllable‐level combinations, subsequently word‐level combinations) can emerge during the course of learning.     

A meta-analysis of infants’ word-form recognition

Recognizing word forms is an important step on infants’ way toward mastering their native language. The present study takes a meta-analytic approach to assess overarching questions on the literature of early word-form recognition. Specifically, we investigated the extent to which there is cross-linguistic evidence for an early recognition lexicon, and how it may be influenced by infant age, language background, and familiarity of the selected stimuli (approximated by parent-reported word knowledge). Our meta-analysis—with open data access on metalab.stanford.edu—was based on 32 experiments in 16 different published or unpublished studies on infants 5–15 months of age. We found an overall significant effect of word-form familiarity on infants’ responses. This effect increased with age and was higher for infants learning Romance languages than other languages. We further found that younger, but not older, infants showed higher effect sizes for more familiar word lists. These insights should help researchers plan future studies on word-form recognition.     

Influences of Infant‐Directed Speech on Early Word Recognition

When addressing infants, many adults adopt a particular type of speech, known as infant‐directed speech (IDS). IDS is characterized by exaggerated intonation, as well as reduced speech rate, shorter utterance duration, and grammatical simplification. It is commonly asserted that IDS serves in part to facilitate language learning. Although intuitively appealing, direct empirical tests of this claim are surprisingly scarce. Additionally, studies that have examined associations between IDS and language learning have measured learning within a single laboratory session rather than the type of long‐term storage of information necessary for word learning. In this study, 7‐ and 8‐month‐old infants' long‐term memory for words was assessed when words were spoken in IDS and adult‐directed speech (ADS). Word recognition over the long term was successful for words introduced in IDS, but not for those introduced in ADS, regardless of the register in which recognition stimuli were produced. Findings are discussed in the context of the influence of particular input styles on emergent word knowledge in prelexical infants.     

Early Segmentation Abilities in Preterm Infants

Infants born preterm have higher risks of developing linguistic deficits. Considering that the ability to segment words from fluent speech is crucial for lexical acquisition, Experiment 1 tested the ability of healthy extremely‐to‐late preterm infants to segment monosyllabic words at 6 months of postnatal age. Results establish basic segmentation skills in these infants. While we failed to find an effect of the degree of prematurity, this issue will need further exploration. Future studies will also have to specify the scope of these early segmentation skills, both in terms of the types of words segmented, the cues used to do so, and in terms of possible differences in performance between subgroups of preterm infants (e.g., based on gestational age or medical risks). Lastly, given that the preterm infants tested had a mean maturational age of 4 months, Experiment 2 explored monosyllabic segmentation in full‐term 4‐month olds. Infants succeeded at the task, providing the earliest developmental evidence of word segmentation in full‐term infants. These findings better specify the early trajectory of segmentation abilities in both full‐term and healthy, low‐risk preterm infants and support the proposal that prematurity might have a differential effect on the early acquisition of various linguistic levels.     

The First Slow Step: Differential Effects of Object and Word-Form Familiarization on Retention of Fast-Mapped Words

Recent research demonstrated that although twenty-four month-old infants do well on the initial pairing of a novel word and novel object in fast-mapping tasks, they are unable to retain the mapping after a five-minute delay. The current study examines the role of familiarity with the objects and words on infants' ability to bridge between the initial fast mapping of a name and object, and later retention in the service of slow mapping. Twenty-four-month-old infants were familiarized with either novel objects or novel names prior to the referent selection portion of a fast-mapping task. When familiarized with the novel objects, infants retained the novel mapping after a delay, but not when familiarized with the novel words. This suggests familiarity with the object versus the word form leads to differential encoding of the name-object link. We discuss the implications of this finding for subsequent slow mapping.     

Statistical word segmentation: Anchoring learning across contexts

The present experiments were designed to assess infants' abilities to use syllable co-occurrence regularities to segment fluent speech across contexts. Specifically, we investigated whether 9-month-old infants could use statistical regularities in one speech context to support speech segmentation in a second context. Contexts were defined by different word sets representing contextual differences that might occur across conversations or utterances. This mimics the integration of information across multiple interactions within a single language, which is critical for language acquisition. In particular, we performed two experiments to assess whether a statistically segmented word could be used to anchor segmentation in a second, more challenging context, namely speech with variable word lengths. The results of Experiment 1 were consistent with past work suggesting that statistical learning may be hindered by speech with word-length variability, which is inherent to infants' natural speech environments. In Experiment 2, we found that infants could use a previously statistically segmented word to support word segmentation in a novel, challenging context. We also present findings suggesting that this ability was associated with infants' early word knowledge but not their performance on a cognitive development assessment.     

Infant Recognition of Hebrew Vocalic Word Patterns

The ability of infants to recognize phonotactic patterns in their native language is widely acknowledged. However, the specific ability of infants to recognize patterns created by nonadjacent vowels in words has seldom been investigated. In Semitic languages such as Hebrew, groups of multisyllabic words are identical in their nonadjacent vowel sequences and stress position but differ in the consonants interposed between the vowels. The goals of this study were to assess whether infants learning Hebrew show a preference for (1) a nonadjacent vocalic pattern or template, common in Hebrew nouns (CéCeC), over a nonattested nonadjacent vocalic pattern (CóCoC), and (2) a nonadjacent vocalic pattern common in Hebrew words (CaCóC) over an existing but less common pattern (CaCéC). Twenty Hebrew‐learning infants aged 8 to 11 months were presented with lists of nonsense words featuring the first two patterns (Experiment 1), and 20 were presented with nonsense words featuring the second two patterns (Experiment 2). The results showed longer listening to CéCeC than to CóCoC lists and to CaCóC than to CaCéC lists, suggesting that infants recognized the common nonadjacent vocalic patterns in both cases. The study thus demonstrates that Hebrew‐learning infants are able to disregard the intervening consonants within words and generalize their vocalic pattern to previously unheard nonwords, whether this pattern includes identical or different vowels and regardless of the rhythmic pattern of the word (trochaic or iambic). Analysis of the occurrence of the relevant vowel patterns in input speech in three Hebrew corpora (two addressed to children and one to adults) suggests that exposure to these patterns in words underlies the infants' preferences.     

Hierarchical Processing in Seven‐Month‐Old Infants

Hierarchical structures are crucial to many aspects of cognitive processing and especially for language. However, there still is little experimental support for the ability of infants to learn such structures. Here, we show that, with structures simple enough to be processed by various animals, seven‐month‐old infants seem to learn hierarchical relations. Infants were presented with an artificial language composed of “sentences” made of three‐syllable “words.” The syllables within words conformed to repetition patterns based on syllable tokens involving either adjacent repetitions (e.g., dubaba) or nonadjacent repetitions (e.g., du ba du ) . Importantly, the sequence of word structures in each sentence conformed to repetition patterns based on word types (e.g., aba‐ abb ‐ abb ). Infants learned this repetition pattern of repetition patterns and thus likely a hierarchical pattern based on repetitions, but only when the repeated word structure was based on adjacent repetitions. While our results leave open the question of which exact sentence‐level pattern infants learned, they suggest that infants embedded the word‐level patterns into a higher‐level pattern and thus seemed to acquire a hierarchically embedded pattern.     

Beneficial effects of the mother's voice on infants’ novel word learning

The maternal voice appears to have a special role in infants’ language processing. The current eye‐tracking study investigated whether 24‐month‐olds (n = 149) learn novel words easier while listening to their mother's voice compared to hearing unfamiliar speakers. Our results show that maternal speech facilitates the formation of new word–object mappings across two different learning settings: a live setting in which infants are taught by their own mother or the experimenter, and a prerecorded setting in which infants hear the voice of either their own or another mother through loudspeakers. Furthermore, this study explored whether infants’ pointing gestures and novel word productions over the course of the word learning task serve as meaningful indexes of word learning behavior. Infants who repeated more target words also showed a larger learning effect in their looking behavior. Thus, maternal speech and infants’ willingness to repeat novel words are positively linked with novel word learning.     

Seeing a talking face matters: Infants' segmentation of continuous auditory-visual speech

Visual speech cues from a speaker's talking face aid speech segmentation in adults, but despite the importance of speech segmentation in language acquisition, little is known about the possible influence of visual speech on infants' speech segmentation. Here, to investigate whether there is facilitation of speech segmentation by visual information, two groups of English-learning 7-month-old infants were presented with continuous speech passages, one group with auditory-only (AO) speech and the other with auditory-visual (AV) speech. Additionally, the possible relation between infants' relative attention to the speaker's mouth versus eye regions and their segmentation performance was examined. Both the AO and the AV groups of infants successfully segmented words from the continuous speech stream, but segmentation performance persisted for longer for infants in the AV group. Interestingly, while AV group infants showed no significant relation between the relative amount of time spent fixating the speaker's mouth versus eyes and word segmentation, their attention to the mouth was greater than that of AO group infants, especially early in test trials. The results are discussed in relation to the possible pathways through which visual speech cues aid speech perception.