Statistical learning of tone sequences by human infants and adults

Previous research suggests that language learners can detect and use the statistical properties of syllable sequences to discover words in continuous speech (e.g. Aslin, R.N., Saffran, J.R., Newport, E.L., 1998. Computation of conditional probability statistics by 8-month-old infants. Psychological Science 9, 321-324; Saffran, J.R., Aslin, R.N., Newport, E.L., 1996. Statistical learning by 8-month-old infants. Science 274, 1926-1928; Saffran, J., R., Newport, E.L., Aslin, R.N., (1996). Word segmentation: the role of distributional cues. Journal of Memory and Language 35, 606-621; Saffran, J.R., Newport, E.L., Aslin, R.N., Tunick, R.A., Barrueco, S., 1997. Incidental language learning: Listening (and learning) out of the corner of your ear. Psychological Science 8, 101-195). In the present research, we asked whether this statistical learning ability is uniquely tied to linguistic materials. Subjects were exposed to continuous non-linguistic auditory sequences whose elements were organized into 'tone words'. As in our previous studies, statistical information was the only word boundary cue available to learners. Both adults and 8-month-old infants succeeded at segmenting the tone stream, with performance indistinguishable from that obtained with syllable streams. These results suggest that a learning mechanism previously shown to be involved in word segmentation can also be used to segment sequences of non-linguistic stimuli.     

A Bayesian framework for word segmentation: Exploring the effects of context

Since the experiments of Saffran et al. [Saffran, J., Aslin, R., & Newport, E. (1996). Statistical learning in 8-month-old infants. Science, 274, 1926-1928], there has been a great deal of interest in the question of how statistical regularities in the speech stream might be used by infants to begin to identify individual words. In this work, we use computational modeling to explore the effects of different assumptions the learner might make regarding the nature of words - in particular, how these assumptions affect the kinds of words that are segmented from a corpus of transcribed child-directed speech. We develop several models within a Bayesian ideal observer framework, and use them to examine the consequences of assuming either that words are independent units, or units that help to predict other units. We show through empirical and theoretical results that the assumption of independence causes the learner to undersegment the corpus, with many two- and three-word sequences (e.g. what’s that, do you, in the house) misidentified as individual words. In contrast, when the learner assumes that words are predictive, the resulting segmentation is far more accurate. These results indicate that taking context into account is important for a statistical word segmentation strategy to be successful, and raise the possibility that even young infants may be able to exploit more subtle statistical patterns than have usually been considered.     

(Non)words, (non)words, (non)words: evidence for a protolexicon during the first year of life

Previous research with artificial language learning paradigms has shown that infants are sensitive to statistical cues to word boundaries (Saffran, Aslin & Newport, 1996) and that they can use these cues to extract word‐like units (Saffran, 2001). However, it is unknown whether infants use statistical information to construct a receptive lexicon when acquiring their native language. In order to investigate this issue, we rely on the fact that besides real words a statistical algorithm extracts sound sequences that are highly frequent in infant‐directed speech but constitute nonwords. In three experiments, we use a preferential listening paradigm to test French‐learning 11‐month‐old infants' recognition of highly frequent disyllabic sequences from their native language. In Experiments 1 and 2, we use nonword stimuli and find that infants listen longer to high‐frequency than to low‐frequency sequences. In Experiment 3, we compare high‐frequency nonwords to real words in the same frequency range, and find that infants show no preference. Thus, at 11 months, French‐learning infants recognize highly frequent sound sequences from their native language and fail to differentiate between words and nonwords among these sequences. These results are evidence that they have used statistical information to extract word candidates from their input and stored them in a ‘protolexicon’, containing both words and nonwords.     

Lexical and sublexical units in speech perception

Saffran, Newport, and Aslin (1996a) found that human infants are sensitive to statistical regularities corresponding to lexical units when hearing an artificial spoken language. Two sorts of segmentation strategies have been proposed to account for this early word-segmentation ability: bracketing strategies, in which infants are assumed to insert boundaries into continuous speech, and clustering strategies, in which infants are assumed to group certain speech sequences together into units ( Swingley, 2005 ). In the present study, we test the predictions of two computational models instantiating each of these strategies i.e., Serial Recurrent Networks: Elman, 1990 ; and Parser: Perruchet & Vinter, 1998 in an experiment where we compare the lexical and sublexical recognition performance of adults after hearing 2 or 10 min of an artificial spoken language. The results are consistent with Parser's predictions and the clustering approach, showing that performance on words is better than performance on part-words only after 10 min. This result suggests that word segmentation abilities are not merely due to stronger associations between sublexical units but to the emergence of stronger lexical representations during the development of speech perception processes.     

Computation of Conditional Probability Statistics by 8-Month-Old Infants

A recent report demonstrated that 8-month-olds can segment a continuous stream of speech syllables, containing no acoustic or prosodic cues to word boundaries, into wordlike units after only 2 min of listening experience (Saffran, Aslin, & Newport, 1996). Thus, a powerful learning mechanism capable of extracting statistical information from fluent speech is available early in development. The present study extends these results by documenting the particular type of statistical computation–transitional (conditional) probability–used by infants to solve this word-segmentation task. An artificial language corpus, consisting of a continuous stream of trisyllabic nonsense words, was presented to 8-month-olds for 3 min. A postfamiliarization test compared the infants' responses to words versus part-words (trisyllabic sequences spanning word boundaries). The corpus was constructed so that test words and part-words were matched in frequency, but differed in their transitional probabilities. Infants showed reliable discrimination of words from part-words, thereby demonstrating rapid segmentation of continuous speech into words on the basis of transitional probabilities of syllable pairs.     

Segmentation of the speech stream in a non-human primate: statistical learning in cotton-top tamarins

Previous work has shown that human adults, children, and infants can rapidly compute sequential statistics from a stream of speech and then use these statistics to determine which syllable sequences form potential words. In the present paper we ask whether this ability reflects a mechanism unique to humans, or might be used by other species as well, to acquire serially organized patterns. In a series of four experimental conditions, we exposed a New World monkey, the cotton-top tamarin (Saguinus oedipus), to the same speech streams used by Saffran, Aslin, and Newport (Science 274 (1996) 1926) with human infants, and then tested their learning using similar methods to those used with infants. Like humans, tamarins showed clear evidence of discriminating between sequences of syllables that differed only in the frequency or probability with which they occurred in the input streams. These results suggest that both humans and non-human primates possess mechanisms capable of computing these particular aspects of serial order. Future work must now show where humans' (adults and infants) and non-human primates' abilities in these tasks diverge.     

Statistical computations over a speech stream in a rodent

Statistical learning is one of the key mechanisms available to human infants and adults when they face the problems of segmenting a speech stream (Saffran, Aslin, & Newport, 1996) and extracting long-distance regularities (G6mez, 2002; Pe?a, Bonatti, Nespor, & Mehler, 2002). In the present study, we explore statistical learning abilities in rats in the context of speech segmentation experiments. In a series of five experiments, we address whether rats can compute the necessary statistics to be able to segment synthesized speech streams and detect regularities associated with grammatical structures. Our results demonstrate that rats can segment the streams using the frequency of co-occurrence (not transitional probabilities, as human infants do) among items, showing that some basic statistical learning mechanism generalizes over nonprimate species. Nevertheless, rats did not differentiate among test items when the stream was organized over more complex regularities that involved nonadjacent elements and abstract grammar-like rules.     

The statistical signature of morphosyntax: A study of Hungarian and Italian infant-directed speech

Does statistical learning (Saffran, Aslin, &; Newport, 1996) offer a universal segmentation strategy for young language learners? Previous studies on large corpora of English and structurally similar languages have shown that statistical segmentation can be an effective strategy. However, many of the world’s languages have richer morphological systems, with sometimes several affixes attached to a stem (e.g. Hungarian: iskoláinkban: iskolá-i-nk-ban school.pl.poss1pl.inessive ‘in our schools’). In these languages, word boundaries and morpheme boundaries do not coincide. Does the internal structure of words affect segmentation? What word forms does segmentation yield in morphologically rich languages: complex word forms or separate stems and affixes? The present paper answers these questions by exploring different segmentation algorithms in infant-directed speech corpora from two typologically and structurally different languages, Hungarian and Italian. The results suggest that the morphological and syntactic type of a language has an impact on statistical segmentation, with different strategies working best in different languages. Specifically, the direction of segmentation seems to be sensitive to the affixation order of a language. Thus, backward probabilities are more effective in Hungarian, a heavily suffixing language, whereas forward probabilities are more informative in Italian, which has fewer suffixes and a large number of phrase-initial function words. The consequences of these findings for potential segmentation and word learning strategies are discussed.     

The Temporal Dynamics of Regularity Extraction in Non‐Human Primates

Extracting the regularities of our environment is one of our core cognitive abilities. To study the fine‐grained dynamics of the extraction of embedded regularities, a method combining the advantages of the artificial language paradigm (Saffran, Aslin, & Newport, 1996 ) and the serial response time task (Nissen & Bullemer, 1987 ) was used with a group of Guinea baboons (Papio papio) in a new automatic experimental device (Fagot & Bonté, 2010 ). After a series of random trials, monkeys were exposed to language‐like patterns. We found that the extraction of embedded patterns positioned at the end of larger patterns was faster than the extraction of initial embedded patterns. This result suggests that there is a learning advantage for the final element of a sequence that benefits from the contextual information provided by previous elements.     

Infant-directed speech supports phonetic category learning in English and Japanese

Across the first year of life, infants show decreased sensitivity to phonetic differences not used in the native language [Werker, J. F., & Tees, R. C. (1984). Cross-language speech perception: evidence for perceptual reorganization during the first year of life. Infant Behaviour and Development, 7, 49-63]. In an artificial language learning manipulation, Maye, Werker, and Gerken [Maye, J., Werker, J. F., & Gerken, L. (2002). Infant sensitivity to distributional information can affect phonetic discrimination. Cognition, 82(3), B101-B111] found that infants change their speech sound categories as a function of the distributional properties of the input. For such a distributional learning mechanism to be functional, however, it is essential that the input speech contain distributional cues to support such perceptual learning. To test this, we recorded Japanese and English mothers teaching words to their infants. Acoustic analyses revealed language-specific differences in the distributions of the cues used by mothers (or cues present in the input) to distinguish the vowels. The robust availability of these cues in maternal speech adds support to the hypothesis that distributional learning is an important mechanism whereby infants establish native language phonetic categories.     

Infants are sensitive to within-category variation in speech perception

Previous research on speech perception in both adults and infants has supported the view that consonants are perceived categorically; that is, listeners are relatively insensitive to variation below the level of the phoneme. More recent work, on the other hand, has shown adults to be systematically sensitive to within category variation [McMurray, B., Tanenhaus, M., & Aslin, R. (2002). Gradient effects of within-category phonetic variation on lexical access, Cognition, 86 (2), B33-B42.]. Additionally, recent evidence suggests that infants are capable of using within-category variation to segment speech and to learn phonetic categories. Here we report two studies of 8-month-old infants, using the head-turn preference procedure, that examine more directly infants' sensitivity to within-category variation. Infants were exposed to 80 repetitions of words beginning with either /b/ or /p/. After exposure, listening times to tokens of the same category with small variations in VOT were significantly different than to both the originally exposed tokens and to the cross-category-boundary competitors. Thus infants, like adults, show systematic sensitivity to fine-grained, within-category detail in speech perception.     

Learning words’ sounds before learning how words sound: 9-Month-olds use distinct objects as cues to categorize speech information

One of the central themes in the study of language acquisition is the gap between the linguistic knowledge that learners demonstrate, and the apparent inadequacy of linguistic input to support induction of this knowledge. One of the first linguistic abilities in the course of development to exemplify this problem is in speech perception: specifically, learning the sound system of one’s native language. Native-language sound systems are defined by meaningful contrasts among words in a language, yet infants learn these sound patterns before any significant numbers of words are acquired. Previous approaches to this learning problem have suggested that infants can learn phonetic categories from statistical analysis of auditory input, without regard to word referents. Experimental evidence presented here suggests instead that young infants can use visual cues present in word-labeling situations to categorize phonetic information. In Experiment 1, 9-month-old English-learning infants failed to discriminate two non-native phonetic categories, establishing baseline performance in a perceptual discrimination task. In Experiment 2, these infants succeeded at discrimination after watching contrasting visual cues (i.e., videos of two novel objects) paired consistently with the two non-native phonetic categories. In Experiment 3, these infants failed at discrimination after watching the same visual cues, but paired inconsistently with the two phonetic categories. At an age before which memory of word labels is demonstrated in the laboratory, 9-month-old infants use contrastive pairings between objects and sounds to influence their phonetic sensitivity. Phonetic learning may have a more functional basis than previous statistical learning mechanisms assume: infants may use cross-modal associations inherent in social contexts to learn native-language phonetic categories.     

Infant memory for musical experiences

Recent findings suggest that infants can remember words from stories over 2 week delays (Jusczyk, P. W., & Hohne, E. A. (1997). Infants' memory for spoken words. Science, 277, 1984-1986). Because music, like language, presents infants with a massively complex auditory learning task, it is possible that infant memory for musical stimuli is equally powerful. Seven-month-old infants heard two Mozart sonata movements daily for 2 weeks. Following a 2 week retention interval, the infants were tested on passages of the familiarized music, and passages taken from similar but novel music. Results from two experiments suggest that the infants retained the familiarized music in long-term memory, and that their listening preferences were affected by the extent to which familiar passages were removed from the musical contexts within which they were originally learned.     

Songs to syntax: the linguistics of birdsong

Unlike our primate cousins, many species of bird share with humans a capacity for vocal learning, a crucial factor in speech acquisition. There are striking behavioural, neural and genetic similarities between auditory-vocal learning in birds and human infants. Recently, the linguistic parallels between birdsong and spoken language have begun to be investigated. Although both birdsong and human language are hierarchically organized according to particular syntactic constraints, birdsong structure is best characterized as 'phonological syntax', resembling aspects of human sound structure. Crucially, birdsong lacks semantics and words. Formal language and linguistic analysis remains essential for the proper characterization of birdsong as a model system for human speech and language, and for the study of the brain and cognition evolution.     

A computational model of word segmentation from continuous speech using transitional probabilities of atomic acoustic events

Word segmentation from continuous speech is a difficult task that is faced by human infants when they start to learn their native language. Several studies indicate that infants might use several different cues to solve this problem, including intonation, linguistic stress, and transitional probabilities between subsequent speech sounds. In this work, a computational model for word segmentation and learning of primitive lexical items from continuous speech is presented. The model does not utilize any a priori linguistic or phonemic knowledge such as phones, phonemes or articulatory gestures, but computes transitional probabilities between atomic acoustic events in order to detect recurring patterns in speech. Experiments with the model show that word segmentation is possible without any knowledge of linguistically relevant structures, and that the learned ungrounded word models show a relatively high selectivity towards specific words or frequently co-occurring combinations of short words.     

Newborns are sensitive to multiple cues for word segmentation in continuous speech

Before infants can learn words, they must identify those words in continuous speech. Yet, the speech signal lacks obvious boundary markers, which poses a potential problem for language acquisition (Swingley, Philos Trans R Soc Lond. Series B, Biol Sci 364 (1536), 3617–3632, 2009). By the middle of the first year, infants seem to have solved this problem (Bergelson & Swingley, Proc Natl Acad Sci 109 (9), 3253–3258, 2012; Jusczyk & Aslin, Cogn Psychol 29 , 1–23, 1995), but it is unknown if segmentation abilities are present from birth, or if they only emerge after sufficient language exposure and/or brain maturation. Here, in two independent experiments, we looked at two cues known to be crucial for the segmentation of human speech: the computation of statistical co‐occurrences between syllables and the use of the language's prosody. After a brief familiarization of about 3 min with continuous speech, using functional near‐infrared spectroscopy, neonates showed differential brain responses on a recognition test to words that violated either the statistical (Experiment 1) or prosodic (Experiment 2) boundaries of the familiarization, compared to words that conformed to those boundaries. Importantly, word recognition in Experiment 2 occurred even in the absence of prosodic information at test, meaning that newborns encoded the phonological content independently of its prosody. These data indicate that humans are born with operational language processing and memory capacities and can use at least two types of cues to segment otherwise continuous speech, a key first step in language acquisition.     

Effects of visual information on adults' and infants' auditory statistical learning

Infant and adult learners are able to identify word boundaries in fluent speech using statistical information. Similarly, learners are able to use statistical information to identify word-object associations. Successful language learning requires both feats. In this series of experiments, we presented adults and infants with audio-visual input from which it was possible to identify both word boundaries and word-object relations. Adult learners were able to identify both kinds of statistical relations from the same input. Moreover, their learning was actually facilitated by the presence of two simultaneously present relations. Eight-month-old infants, however, do not appear to benefit from the presence of regular relations between words and objects. Adults, like 8-month-olds, did not benefit from regular audio-visual correspondences when they were tested with tones, rather than linguistic input. These differences in learning outcomes across age and input suggest that both developmental and stimulus-based constraints affect statistical learning.     

The minimal unit of phonological encoding: prosodic or lexical word

Wheeldon and Lahiri (Journal of Memory and Language 37 (1997) 356) used a prepared speech production task (Sternberg, S., Monsell, S., Knoll, R. L., & Wright, C. E. (1978). The latency and duration of rapid movement sequences: comparisons of speech and typewriting. In G. E. Stelmach (Ed.), Information processing in motor control and learning (pp. 117-152). New York: Academic Press; Sternberg, S., Wright, C. E., Knoll, R. L., & Monsell, S. (1980). Motor programs in rapid speech: additional evidence. In R. A. Cole (Ed.), The perception and production of fluent speech (pp. 507-534). Hillsdale, NJ: Erlbaum) to demonstrate that the latency to articulate a sentence is a function of the number of phonological words it comprises. Latencies for the sentence [Ik zoek het] [water] 'I seek the water' were shorter than latencies for sentences like [Ik zoek] [vers] [water] 'I seek fresh water'. We extend this research by examining the prepared production of utterances containing phonological words that are less than a lexical word in length. Dutch compounds (e.g. ooglid 'eyelid') form a single morphosyntactic word and a phonological word, which in turn includes two phonological words. We compare their prepared production latencies to those syntactic phrases consisting of an adjective and a noun (e.g. oud lid 'old member') which comprise two morphosyntactic and two phonological words, and to morphologically simple words (e.g. orgel 'organ') which comprise one morphosyntactic and one phonological word. Our findings demonstrate that the effect is limited to phrasal level phonological words, suggesting that production models need to make a distinction between lexical and phrasal phonology.     

Perceiving speech rhythm in music: Listeners classify instrumental songs according to language of origin

Recent evidence suggests that the musical rhythm of a particular culture may parallel the speech rhythm of that culture’s language (Patel, A. D., & Daniele, J. R. (2003). An empirical comparison of rhythm in language and music. Cognition, 87, B35-B45). The present experiments aimed to determine whether listeners actually perceive such rhythmic differences in a purely musical context (i.e., in instrumental music without words). In Experiment 1a, listeners successfully classified instrumental renditions of French and English songs having highly contrastive rhythmic differences. Experiment 1b replicated this result with the same songs containing rhythmic information only. In Experiments 2a and 2b, listeners successfully classified original and rhythm-only stimuli when language-specific rhythmic differences were less contrastive but more representative of differences found in actual music and speech. These findings indicate that listeners can use rhythmic similarities and differences to classify songs originally composed in two languages having contrasting rhythmic prosody.