Poster Session 1 (with coffee break)
Thursday, June 20th, 2013 [16:00 - 17:30]
The acquisition of low applicatives and dative se in L1 Spanish
Escobar Álvarez, M. &. & Teomiro García, I. I.
Universidad Nacional de Educación a Distancia (UNED)
Anticausatives require the clitic ?se? in Spanish. In contrast, other kinds of verbs (consumption and certain intransitives like ?caer(se)? and ?morir(se)?) optionally allow for this clitic. Data from L1 acquisition show that children have no difficulty with obligatory ?se?. Besides, experimental data support the hypothesis that children find the clitic ?se? easier with anticausative than with transitive verbs. Furthermore, the acquisition of optional clitic ?se? is also problematic, since children tend not to use it at all, while there are no such omissions when ?se? is obligatory. It could be argued that the difficulties with clitic ?se? are due to its optional character. However, other configurations with obligatory ?se? are also acquired at later stages provided data from CHILDES. In adult syntax, clitic ?se? with anticausatives is argued to be an inflexion element inserted in the derivation due to formal reasons. In contrast, other syntactic configurations with ?se? have in common the fact that ?se? is within a low applicative phrase and assigned dative Case. If we consider the Czech clitic ?si? as the equivalent counterpart of Spanish ?se?, we observe that it is also marked as dative in this second language. Hence, we think that there is also a low applicative phrase in the later acquired configurations. Hence, we want to put forward a twofold analysis of clitic ?se?: as an expletive and as a dative clitic argument. We assume that the difficulty we found with the latter is due to the fact that it is within a low applicative phrase. We assume that this complex derivation needs some time to be acquired. Actually, if the optional character of ?se? was the reason for children?s difficulty, some acquisition data would be unexpected since ?se? in ?romperse la quisma? (break your head) is equally compulsory in adult grammar.
Discrimination learning in learning Korean as a second language
Chen, J. , Liu, Y. , Li, A. & Wang, T.
National Taiwan Normal University, Taiwan
Discrimination learning is one of several implicit learning mechanisms not only available for learning one?s first language, but, we hypothesize, also available for learning a second language. The hypothesis was tested by engaging 20 Chinese speakers in a perceptual discrimination task involving Korean minimal word pairs. Three categories of stop consonants were selected, each with a minimal contrast distinguishing a lenis and a fortis member. Six minimal word pairs so contrasted were chosen for each category. Half of the pairs in each category were randomly selected for each participant to serve as the trained pairs, and the remaining half as the untrained ones. Each pair was permuted to produce all four possible arrangements. All 72 pairs, spoken by a female native Korean speaker, were presented to the participants in a pre-test to determine if the members of each pair were the same or different words. Following the pre-test, the trained pairs were administered to the participants in the same way as in the pre-test. Feedback was given after each trial. A pair was repeatedly administered until the participants scored correctly on that pair three times in a row. Across participants and pairs, each pair took an average of 5.3 repetitions before it was learned to criterion. A post-test involving the same 72 pairs followed the training. For the trained pairs, the proportion of correct responses was .68 on the pre-test and .79 on the post-test. For the untrained pairs, these proportions were .66 and .74, respectively. These results show that training improved performance by about 16%. In addition, the outcome of discrimination learning achieved during training could transfer to benefit the performance on the untrained pairs to almost the same extent (12%). We conclude that discrimination learning of linguistic forms can be a powerful mechanism for learning a second language.
Phonological wildness in early language development: Exploring the role of onomatopoeia
Smith, C.
University of York, UK
Onomatopoeic forms are often disregarded from the phonological analysis of infant data (e.g., Fikkert & Levelt (2008)), seen as a temporary and irrelevant aspect of the developing lexicon which is superfluous to the adult speech model of Indo-European languages. However, synesthetic forms known as mimetics are abundant in languages such as Japanese, and have been found to facilitate novel-verb learning in Japanese amongst infants acquiring either Japanese or English as a first language (Kantartzis et al. (2011)). Onomatopoeic forms often constitute an important portion of infants` earliest word forms in a range of languages (Menn & Vihman (2011)) despite their limited role in most adult lexica. These are thought to provide a linguistic scaffold in language development through the perception of phonologically wild segments (Rhodes (1994)): wildness in the input, whereby the vocal tract`s full capacity is used in order to approximate sounds of non-human origin, is found to serve as an attention-marker, as well as aiding phonological recall in production. This study uses eye-tracking to explore the role that onomatopoeia play in language development. Infants are presented with recordings of onomatopoeic forms produced in familiar and unfamiliar languages. Forms are presented in a phonologically wild or tame manner, and infants` response times and eye movements are measured. It is hypothesised that wild onomatopoeic forms in both familiar and unfamiliar languages will elicit a quicker looking time and more accurate response than tame forms in familiar and unfamiliar languages. Results reflect the role that onomatopoeia play in early language development: cross-linguistic similarities in the prosody of onomatopoeic forms may prompt understanding and early word production, driving the infant towards further phonological development. Furthermore, the wild versus tame paradigm highlights the contrast between prosodic and phonological learning, reflecting which of the input`s linguistic queues are most relevant in early language development.
Effects of development on cross-language speech perception
Dar, M.
University of York
The present study tested infants from English-speaking homes to examine effects of development on cross-language speech perception. Werker (1981, 1984) showed that 6-8-month-old infants are able to discriminate non-native speech sounds that adults cannot discriminate, but by 10-12 months the infants were no longer able to make the discrimination. According to Kuhl (2004, 2008), this decline in discrimination is due to infants? increase in native-language exposure, which leads to ?neural commitment? to the native language at this age. Many studies have shown a decline in discriminatory abilities of infants for non-native contrasts between 6 - 12 months of age but no study to date has tested a contrast in affricates in a cross-language perception test. Also, very little attempt has been made to show whether the experimental order of presentation of stimuli affects infants? performance. An aspirated - unaspirated contrast of Urdu /t?/ - /t?h/ was selected based on a pilot study with 20 English-speaking adults who were tested on a number of Urdu contrasts not found in English to identify the most difficult. Twenty-four 7- and 11-month olds were tested in a habituation procedure. Half of the infants were habituated to the voiceless aspirated affricate and tested on the contrasting voiceless unaspirated affricate while the remaining infants experienced the reverse pattern. Discrimination was assessed by comparing mean looking time during the last two habituation trials to mean looking time during the first two trials of the test phase. In agreement with the literature, the results indicated that 6-8-month-olds could discriminate the affricate pair but 11-month-olds could not. Infants presented with the non-prototypical consonant (the aspirated affricate, which does not occur in English) in the habituation phase showed better discrimination in the test phase than the infants presented with the prototypical consonant in the habituation phase.
Infants' audiovisual speech integration does not hinge on phonetic knowledge
Baart, M. 1 , Vroomen, J. 2 , Shaw, K. 3 & Bortfeld, H. 3, 4
1 Basque Center on Cognition, Brain and Language, Donostia, Spain
2 Tilburg University, Dept. of Cognitive Neuropsychology, Tilburg, the Netherlands
3 University of Connecticut, Dept. of Psychology, Storrs, CT, the United States of America
4 Haskins Laboratories, New Haven, CT, the United States of America
Infants and adults are well able to match auditory and visual speech, but the cues on which they rely (viz. temporal, energetic and phonetic correspondence in the auditory and visual speech streams) may differ. Here we assessed the relative contribution of the different cues using sine-wave speech (SWS). Adults (N=52) and infants (N=30) matched 2 trisyllabic speech sounds (?kalisu? and ?mufapi?), either natural speech or SWS, with visual speech information. On each trial, adults saw two articulating faces and matched a sound to one of these faces, while infants were presented the same stimuli in a preferential looking paradigm. Adults were almost flawless with natural speech, but significantly less accurate with SWS. In contrast, infants looked longer at the articulating face that matched the sound irrespective of whether it was natural speech or SWS. These findings are in-line with a multi-stage view on audiovisual speech integration and suggest that during development, phonetic knowledge becomes perceptually more important.
DBIfNIRS: An online database of infant functional Near InfraRed Spectroscopy studies
Cristia, A. 1, 2 , Dupoux, E. 1 , Hakuno, Y. 3 , Lloyd-Fox, S. 4 & Minagawa-Kawai, Y. 3, 5
1 Laboratoire de Sciences Cognitives et Psycholinguistique, CNRS, IEC-ENS, EHESS
2 Neurobiology of Language, Max Planck Institute for Psycholinguistics
3 Graduate School of Human Relations, Keio University
4 Centre for Brain and Cognitive Development, Birkbeck
5 Institut d'Etude de la Cognition, ENS
Until recently, imaging the infant brain was very challenging. Functional Near InfraRed Spectroscopy (fNIRS) is a promising, relatively novel technique, whose use is rapidly expanding. As an emergent field, it is particularly important to share methodological knowledge to ensure replicable and robust results. We present DBIfNIRS (sites.google.com/site/dbifnirs), a community-augmented database which will facilitate precisely this exchange. We tabulated articles and theses reporting empirical fNIRS research carried out on infants below three years of age along several methodological variables. The resulting spreadsheet has been uploaded in a format allowing individuals to continue adding new results, and download the most recent version of the table. Thus, this database is ideal to carry out systematic reviews. We illustrate its academic utility by focusing on the factors affecting three key variables: infant attrition, the reliability of oxygenated and deoxygenated responses, and signal-to-noise ratios. We then discuss strengths and weaknesses of the DBIfNIRS, and conclude by suggesting a set of simple guidelines aimed to facilitate methodological convergence through the standardization of reports.
Tracking phonological representations in the first year of life
Teickner, C. 1 , Becker, A. 1 & Friedrich, C. K. 2
1 University of Hamburg
2 University of Tübingen
Adult spoken word processing is gradually sensitive to mismatch between the speech signal and phonological word form representations. In the current study we investigated the development of phonological representations in the first year of life. We recorded event-related potentials (ERPs) while infants listened to single syllables (primes) followed by complete German words (targets). Common disyllabic words, frequently used in caregiver-infant interactions, were chosen from an early words screening inventory (German version of the McArthur Communicative Development Inventories). Primes and targets were presented in three different conditions. In an identity condition, primes match the onsets of the target words (e.g., ma - Mama). In a variation condition, primes deviate from targets in their initial place of articulation (e.g., na - Mama). In a control condition, phonemes of primes and targets differ completely (e.g., vo - Mama). We tested 30 infants from German speaking parents at the age of three and six months. ERPs suggest that in contrast to adults, infants at the age of three months do not group the identity and the variation condition together. This indicates that phonological processing at three month does not tolerate mismatch in place of articulation. On the contrary, we found no differences in the ERPs for the identity and the variation condition of at the age of six months. Thus it appears that phonological word form representations established around six month after birth tolerate place variation.
Content words in Child Directed Speech of Hebrew-speaking parents
Adi-Bensaid, L. 1, 2 , Tubul-Lavy, G. 1, 3 & Ben-David, A. 4, 5
1 Department of Communication Sciences and Disorders, Ono Academic College, Israel
2 Speech and Hearing Center Sheba Medical Center, Tel Hashomer, Ramat Gan, Israel
3 Department of Special Education, Achva Academic College, Israel
4 Department of Communication Disorders, Hadassah Academic College, Israel
5 Department of Communication Disorders, Tel-Aviv University, Israel
Recent studies show the importance of the input in the process of language development. Motherese, baby-talk, Infant-directed Speech (IDS) and Child-Directed Speech (CDS) are terms that refer to special characteristics that exist in the speech that parents display to infants, that differ in key ways from Adult-Directed Speech (ADS). These unique characteristics were described for all language components. Such characteristics may include: shorter and less complex sentences, more restricted vocabulary, a slower rate of speech, a higher fundamental frequency and repetitive utterances (Ferguson, 2004; Owens, 2008). It is assumed that these modifications are designed to facilitate language acquisition and to withdraw the infant's attention to the adult speech. The purpose of the present study was to describe the parents' use of content words in Hebrew-CDS at different stages of language development. The words were classified into six lexical categories: nouns, verbs, adjectives, adverbs, prepositions and pronouns. Participants included 55 parents to children at different stages of language development: a) early pre-verbal, b) late pre-verbal, c) single words and d) early grammar. Children's age range was 5.5-33 months. Forty five minute parent-child-dyads were videotaped at the child's home during eating, bathing, playing, dressing or diaper changing activities. For the purpose of this study, a ten minute randomly selected sample, was transcribed and analyzed both quantitatively and qualitatively. The findings reveal that there are some significant differences in the use of content words in CDS throughout the different developmental stages. While there was a significant decrease in the number of tokens (amount of words per session) with child's age, there was a significant increase in the type of words (diversity of content words) with age. Thus, parents of infants at the early period of the pre-verbal stage used fewer nouns, verbs, pronouns, adjectives and prepositions compared to the following stages.
Effects of SES and maternal talk on early language: New evidence from a direct assessment of vocabulary comprehension
DeAnda, S. 1 , Deák, G. 1 , Poulin-Dubois, D. 2 , Zesiger, P. 3 & Friend, M. 1
1 SDSU/UCSD Joint Doctoral Program In Language and Communicative Disorders
2 Concordia University
3 University of Geneva
Mounting evidence indicates positive relations exist between language input (quantity and quality), socioeconomic status (SES), and vocabulary acquisition in young children. Since these studies focus on lexical production rather than comprehension, little is known about the effects of input and SES at the earliest stages of language acquisition. Further, studies that do evaluate lexical comprehension use parent report, such as the MacArthur-Bates Communicative Inventory: Words and Gestures (MCDI, Fenson et al., 1993). Although useful for assessing early language, use of the MCDI across SES has been questioned, with some studies finding a tendency of lower SES parents to over-report early comprehension (Feldman et al., 2000; Fenson et al., 1993; Reznick, 1990). Consequently, gaps exist in our understanding of the effects of SES and maternal input in early life. This study investigates the relationship between hours of maternal talk, SES, and vocabulary comprehension in 67 English-dominant children between the ages 15;15 and 18;5 (M = 16;21). Lexical comprehension was assessed using a touch-screen behavioral test: the Computerized Comprehension Task (CCT, Friend & Keplinger, 2003; Friend & Keplinger, 2008; Friend, Schmidt, & Simpson, 2012), and the MCDI. Results reveal main effects, but no interaction, of SES and maternal talk on the CCT: higher SES and hours of maternal talk predicted higher comprehension scores, consistent with research using language samples. Conversely, MCDI comprehension scores revealed an interaction of SES and maternal talk: lower SES parents? estimates did not reflect differences in vocabulary as a function of maternal talk, whereas among higher SES parents, more hours of maternal talk predicted higher vocabulary comprehension. Thus, CCT results suggest that both SES and maternal talk have effects on early comprehension before the onset of word combinations in production. Further, the interaction on the MCDI suggests SES influences parent reports of early comprehension, consistent with previous literature.
Segmentation of IDS and ADS in 12-month-olds: An ERP study
Von Holzen, K. , Wolff, D. & Mani, N.
Georg-August-Universität Göttingen
To build up a vocabulary, the infant must extract, or segment, individual words from the language being spoken around them. Research has shown that at 7.5-months-of-age, infants are able to segment words from sentences (Jusczyk & Aslin, 1995) presented in infant-directed speech (IDS). However, only 15% of the speech in the infants' environment is infant-directed (van der Weijer, 2002). At what age, then, do infants begin to segment words from "normal" or adult-directed speech (ADS)? The current study tested German 12-month-olds on their ability to segment both IDS and ADS. Infants were presented with 40 familiarization-test blocks, half in IDS and the other half in ADS. In the familiarization phase, infants heard eight sentences, each containing the same (familiarized) word. In the test phase, infants heard the familiarized word and an unfamiliar control word. ERPs time-locked to the onset of the words in the test phase were examined for differences between familiarized and control words in IDS and ADS blocks. If infants extract the phonological form of the familiarized word from the sentences, recognition of this word-form should be indicated by more negative ERPs for familiarized compared to control words. As predicted ERPs were more negative following familiarized compared to control words, 200-350ms after word onset in the test phase: infants successfully segmented the familiarized words from the sentences and recognized these words later in the test phase. The effect for IDS was greater than ADS; however, this difference was not significant. In summary, the current study demonstrates that German infants, as young as 12-months-of-age, are able to segment words from both ADS and IDS. This may provide further support to infants' ability to begin producing words by 12-months of age and their rapid vocabulary development later in the second year of life (Bloom, 1973).
Preverbal infants show social preferences for native-dialect speakers
Okumura, Y. & Itakura, S.
Kyoto University
Language is one of the most important tools of communication. Recent research shows that social preferences for native language or native-accented speakers emerge remarkably early in development, indicating that infants prefer the speakers in the society to which they belong. The dialect also may be a reliable cue to group membership because it provides information about an individual's social and ethnic group identity. The present study investigated whether infants show social preferences toward native-dialect speakers rather than unfamiliar-dialect ones. Infants at 10 months of age (n = 43) were shown videos in which a native- and an unfamiliar-dialect speaker each talked for 15 seconds. Then, the two speakers offered the infants a toy as real toys appeared in front of them. In addition, we investigated the relationship between infants' social preferences for native-dialect speakers and their own experience with the native dialect by asking the infants' parents how often the infants had been exposed to the dialect. The results demonstrated that the infants preferentially reached for toys that were offered by the speaker who spoke their native dialect (30 of 43 infants, binomial test, two-tailed, p = 0.0137). Moreover, there was a relationship between the proportion of choosing the toy from the native-dialect speaker and the amount of experience infants had with that dialect (r = 0.364, p = 0.016). Therefore, infants showed social preferences for a native-dialect speaker depending on their language experience through daily interactions with the native dialect. These results indicate that native dialects are sufficient to evoke social preferences even in preverbal infants, despite both dialects are within the same native language. Our findings suggest that dialects may be a reliable cue to group membership and infants' orientation toward members of their native community can guide their social and cultural learning.
InPhonDB: A developing meta-analysis of infant vowel perception
Tsuji, S. 1, 2 & Cristia, A. 3
1 Radboud Universiteit Nijmegen
2 International Max Planck Research School for Language Sciences
3 CNRS LSCP
The development of native language phoneme perception is a central focus in research on language development. Yet, no recent source provides a comprehensive overview over decades of data, let alone by organizing them in a qualitative and/or quantitative way. InPhonDB aims to provide such an overview, starting with infant vowel discrimination. Infant vowel perception is particularly interesting to focus on because vowels vary along dimensions that are easily parametrized. Moreover, it can shed light on two key theoretical questions: infants' innate perceptual abilities (e.g., Polka & Bohn, 2011) and their subsequent experience-modulated development (e.g., Kuhl et al., 2008). Therefore, we qualitatively reviewed xx studies on infant vowel discrimination including results from behavioral, electrophysiological, and neuroimaging methods applied to infants 0-12 months of age. Of these, 35 (124 unique experiments) provided sufficient information to be included in a subsequent quantitative meta-analysis. The quantitative data provide statistical support for early perceptual asymmetries in vowel discrimination (cf. Polka & Bohn, 2011) and a role of acoustic distance between vowels (cf. Sebastian-Galles & Bosch, 2009), but reveal little evidence for perceptual attunement based on language exposure (cf. Kuhl et al., 1992). We discuss these new findings in light of recent evidence suggesting that there might be differences between vowel and consonant perception (e.g., Pons, Sabourin, Cady, & Werker, 2006), a possibility not reflected in most current models of infant phoneme perception. Thus, the free, public repository InPhonDB already provides a unique overview of the main factors structuring infant vowel discrimination. Furthermore, it points to important limitations in current available data. Finally, this online resource can accommodate extensions of the vowel discrimination database as new results come to light, as well as the addition of other topics of infant perception. Thus, it allows the research community to better keep up with infants' phonological development.
Bilingual infants show advantage in perceiving non-native tonal contrasts
Liu, L. & Kager, R.
Utrecht Insititute of Linguistics - OTS, Utrecht University
Infants are born with an initial sensitivity to speech prosody, and undergo tonal perceptual reorganization (PR) between 4 and 9 months, after which non-tone-learning infants' sensitivity to tones sharply decreases. However, non-tone-language adults show a sharp psycho-acoustic sensitivity to linguistic pitch. A gap occurs regarding the transition from non-tone-language-learning infants' loss of tonal sensitivity to adults' recovery, calling for an inspection of the crucial periods in this perceptual change, and the factors that influence it. The research questions are: 1) When is the perceptual turning point for non-native pitch perception? 2) Do monolingual and bilingual infants follow the same tonal perceptual trajectory in this process? 3) Does the acoustic salience of the tonal contrast influence monolingual and bilingual infants' tone perception? 140 monolingual Dutch infants and 100 bilingual infants of five age groups (5 to 18 months) were tested on their discrimination of an acoustically salient tonal contrast (/ta/, high level vs. high falling) in Mandarin Chinese and a manipulated less-salient version of the same contrast differing solely on F0 direction. Results display a U-shaped pattern for bilingual infants by the end of the first year of life, and for monolingual infants at 17-18 months. In particular, bilingual infants follow the same developmental trajectory as their monolingual peers in tonal tuning (between 5 and 9 months), but recover their acoustic sensitivity earlier (at 11-12 months). To answer the research questions: The turning point for non-native pitch perception occurs early on in the first two years of life. Bilingual infants' enhanced sensitivity to non-native tonal contrasts at the end of the first year may be due to the more challenging language environment they encounter in general, and/or specifically regarding more complex intonation input. Acoustic salience influences both mono- and bilingual infants' perception, so that acoustically salient contrast may survive PR.
The words of our parents: How parental language use influences early language processing efficiency
Odean, R. , Abad, C. , Bradley, C. A. & Pruden, S. M.
Florida International University
Research on early word learning in recent decades has focused on how children learn nouns (e.g., Fernald et al., 2008; Pruden et al., 2006). However, a comprehensive theory of word learning will require that we understand the development of other word types, including words that describe our spatial world (e.g., ''big,'' ''under,'' ''beside''). We seek to understand whether the quantity of spatial language children hear in the home predicts individual differences in the processing and mapping of a spatial word to its referent. Bilingual Spanish-English children (36-38 months) and a parent participated in a total of 60 minutes of in-home observation, including 30 minutes of everyday activities, and 30 minutes of play with provided spatial toys - a puzzle, a shape sorter puzzle and a Mega Bloks set. The home visits were recorded and transcribed. Transcripts were coded for spatial and non-spatial words produced by the parent and child. Children also participated in an Intermodal Preferential Looking Paradigm in the lab, which measured their ability to map spatial terms onto images, and their efficiency in doing so. Children saw two images depicting a spatial relation simultaneously on a split-screen, while hearing a female speaker ask for the target spatial relation. A Tobii X60 eyetracker recorded children's gaze to each stimuli. Only those spatial terms parents indicated on a checklist that their child knew were included in the final analyses as we were interested in assessing children's efficiency in processing a familiar spatial term. Our predictions are that children whose parents (a) use more spatial language in the home setting, and (b) engage children in spatial activities for longer periods of time will: (1) have larger receptive and productive spatial vocabularies, and (2) be more efficient in processing and mapping spatial words to their correct referents. Preliminary data will be reported.
Non-verbal referential cues aid early word learning of consonant but not vowel minimal pairs
Escudero, P. 1 , Nguyen, B. 2 , Gillespie-Lynch, K. 3, 2 & Johnson, S. P. 2
1 MARCS Institute, University of Western Sydney
2 Department of Psychology, University of California, Los Angeles
3 College of Staten Island, City University of New York
Research has demonstrated that, before 17 months of age, children have substantial difficulty in learning to associate novel words to their picture referents when the novel words form phonologically minimal pairs, i.e., when they differ in a single consonant (e.g. 'bin'-'din'). It has also been shown that children under 17 months can associate objects' referents to some novel minimal pairs involving vowels, namely 'deet'-'dit', suggesting that the difficulty depends on the type of phoneme contrast presented to the learner. Yet other studies show that 14-month-olds can successfully learn minimal pairs if words are presented in clear sentential contexts and word reference via verbal cues is given during learning, demonstrating that the apparent difficulty is only found when the referential status of the novel word is unclear. We examined whether early word learning of minimal pairs can be aided by non-verbal cues, i.e., eye-gaze direction to the referent during word naming. During training, twenty-two 15- and 17-month-olds (N = 11 per group) saw twelve videos with two novel objects on a table and a model naming one of the objects while directing her gaze at it. In each video, one of four novel words ('bon', 'don', 'deet' and 'dit') was named. Testing trials presented two repetitions of one of the spoken words together with its referent next to that of another word. A repeated-measures ANOVA with Word as within-subject factor and Age-Group as between-subjects factor revealed a main effect of Word, with longer looks at the target referent for 'bon' and 'don' but not for 'deet' and 'dit'. No effect of Age-Group was found. We will discuss the implications of our findings for the effect of non-verbal cues on early word learning and for the debate on the differential role of consonants and vowels in lexical processing and its development.
Monodialectal and bidialectal infants use of the Mutual Exclusivity Bias in early word learning
Durrant, S. , Gunning, L. , Jenkins, H. & Floccia, C.
Plymouth University
At around 18 months infants are entering a period of rapid vocabulary growth employing a range of strategies to support this, one being the Mutual Exclusivity Bias (Liittschwager & Markman, 1994). This bias explains how infants match a novel word with an object in their environment for which they currently have no name. This strategy has been repeatedly identified in monolingual infant populations, however, not surprisingly, multilingual infants performance lies on a continuum with bilingual infants occasionally and trilingual infants never applying it (Byers-Heinlein and Werker, 2009). Indeed for these infants this rule is often inappropriate as each object they encounter has at least two nouns to which it can be mapped. An interesting case for further exploration are infants hearing multiple dialects of a single language. It has been demonstrated that these infants store dialect specific representations of familiar words. This could then impact on their use of Mutual Exclusivity as they have multiple representations - one for each dialect to which they are exposed. Following the procedure of Halberda (2003), here we presented 18 month bidialectal infants with pairs of images with one being named (the target). The target was either a name-known object paired with a name known or name-unknown distracter, or a name-unknown object paired with a name-known distracter. Looking times to the intended target following naming were recorded and analysed. Preliminary results show no evidence of Mutual Exclusivity in either group, as revealed by no naming effect for name-unknown target trials. A naming effect was found for name-known trials irrespective of distracter for monodialectal infants, however bidialectal infants only identified the target correctly when both images were name-known. This suggests that the presence of an unknown image is impacting on name-known object naming for bidialectal but not monodialectal infants.
Learning: The L1 acquisition process of IP-internal accusative clitic doubling
Ungureanu, M.
Universite de Moncton, Canada
This presentation centers on the default setting of the Accusative Clitic Doubling (henceforth [ACD]) parameter in first language acquisition, by considering data from Romanian, a [+ACD] language. [+ACD] is a syntactic phenomenon that allows/requires certain overt and covert direct object Determiner Phrases (DPs) to be co-indexed with a pronominal clitic within the same clause (IP). Conversely, in [-ACD] languages (e.g. French), only covert DPs in direct object position may be co-indexed with a pronominal clitic (IP-internally). I adopt the syntactic analysis proposed in Sportiche (1993), according to whom accusative pronominal clitics are functional categories that head their own functional projection. Furthermore, following the Full Competence Hypothesis (FCH), I assume that functional categories (including clitics) are present in the child?s grammar from the beginning. The results of this preliminary study suggest that the default value of the [ACD] parameter is the [-ACD] value, while the [+ACD] value is acquired subsequently, most probably on the basis of positive evidence. As predicted by the hypothesis tested, in the early stages of L1 acquisition, children entertain the [-ACD] parameter: even those learning a [+ACD] language. This pilot experiment was designed to investigate the acquisition process of the [+ACD] value; thus, children at different stages of acquisition and adult control participants were tested. Two tasks were used to test the ACD values: an elicited production task and an imitation task, both of which attempt to minimize complexity from modules other than syntax. The study concludes that, in the process of L1 acquisition of [ACD], the default value is [-] and provides further evidence for the presence of functional categories in the earlier stages of L1 acquisition. The results of this study also raise questions regarding the necessity, or lack thereof, for binary features in the representation of certain syntactic parameters, as opposed to privative features.
Relationships between phonological and lexical development: A longitudinal study with typically developing and late talking children
Rujas, I. 1 , Mariscal, S. 2 , Casla, M. 1 & Galera, N. 1
1 Universidad Autónoma de Madrid
2 Universidad Nacional de Educación a Distancia
Research on early language acquisition has found a bidirectional relationship between phonological and lexical development (Gathercole, 2006; Gathercole & Baddeley, 1989, 1990; Hoff, Core, & Bridges, 2008; Stoel-Gammon, 1991, 2011; Velleman & Vihman, 2006; Vihman, 1981, 1993). In order to examine this relationship we conducted a one-year longitudinal study with children aged 2;06-3;06 who were assessed at three different times. Twenty-three typically developing children (TD) and 15 late talking children (LT) participated in this study. Late talking children scored below percentile 15 in the Spanish version of the MacArthur-Bates Communicative Development Inventories (López-Ornat, Gallego, Gallo, Karousou, Mariscal, & Martínez, 2005). We designed and administered two experimental tasks to assess phonological and lexical development: a word and nonword repetition task, and a fast mapping task (split into a noun task and verb task that included singular and verbal morphemes). Results show that: a) late talking children perform worse in both tasks compared to typically developing children; b) children perform better in time 3 compared to time 1; and c) performance in the repetition task correlates with performance in the fast mapping task. Interactions between assessing times, vocabulary levels and frequency of words and morphemes were found. These results encourage the discussion about the predictive relationships between word and nonword repetition tasks - as measures of phonological development - and fast-mapping tasks - as measures of early lexical development -. Moreover, we will examine the different developmental patterns of TD and LT children, and we will go into some of the possible sources that may help us understand these differences.
The error-driven ranking model of the child's early acquisition of phonotactics
Magri, G.
CNRS, University of Paris 8
Nine-month-old infants are already sensitive to the distinction between licit and illicit forms drawn by the target adult phonotactics (Jusczyk et al. 1993). Yet, they are still blind to alternations, as morphology still lags behind (Hayes 2014). How can they manage to acquire phonotactics from licit forms only? According to the error-driven phonotactic learning model, the child is trained on a stream of licit forms, starts from the most restrictive initial phonotactics, and slightly relaxes it whenever it fails on the current piece of data. The model does not require a lexicon (which might still be unavailable at this age) and predicts a sequence of grammars that can be matched with child acquisition paths (Gnanadesikan 2004). Being trained on licit forms, it is easy to guarantee that the phonotactics learned by the model is consistent: it correctly rules forms which are indeed licit according to the target phonotactics. Yet, the learned phonotactics could be unrestrictive: it could rule in also forms which are instead illicit according to the target phonotactics (Prince and Tesar 2004). In this talk, I tackle the problem of restrictiveness of error-driven learning within the framework of Optimality Theory (OT; Prince and Smolensky 2001). Knowledge of phonotactics is modeled in OT through a ranking of markedness and faithfulness constraints. I show that the relative ranking of the faithfulness constraints turns out to be irrelevant for phonotactics in the vast majority of cases. I then develop an implementation of the error-driven phonotactic learning model within OT and prove that it is restrictive in this vast majority of cases. These results illustrate an approach to cognitive modeling based on the systematic investigation of the algorithmic implications (e.g., implications for restrictiveness) of peculiar typological properties (e.g., the fact that the relative ranking of the faithfulness constraints does not matter).
Word extraction from continuous speech in newborns based on statistical information
Fló, A. , Ferry, A. & Mehler, J.
Cognitive Neuroscience Sector, SISSA, International School for Advanced Studies, Trieste Italy
In this experiment we ask if newborns can extract words from a continuous speech using statistical information. Previous work (Teinonen et al, 2009) demonstrated that newborns could detect word boundaries using just statistical information. Yet it is an open question if they actually extract the words or if they are only detecting boundaries. From previous studies using near-infrared spectroscopy (NIRS) we know that newborns show increased brain activity for a novel word than from a previously learned word (Benavides-Varela et al, 2011). Here, we used a similar methodology, asking if infants show different patterns of brain activity for words compared to part-words from continuous speech with only statistical cues. Infants were exposed to 3 minutes and 40 seconds of continuous speech. The speech stream consisted of four three-syllabic words of the form CVCVCV (e.g., ?lamipeduvokanubefi?). After familiarization, infants were presented with four test phases, each consisting of a short re-familiarization block followed by a test block, with silences of 25-30 seconds interposed between each block. In two of test blocks the four words from the stream were randomly played (e.g., ?lamipe?duvoka?), spaced by 0.5 - 1.5 seconds and repeated twice each, whereas in the other two, four part-words were presented (e.g., ...mipedu?.kanube...). If the infants extract and remember the words, we expect increased activity for the part-word blocks compared to the whole-word blocks. Initial results (N=27) show a differential activation in the test with words respect to the ones with part-words. Whereas part-words elicited an increase in oxyHb spread all over the temporal right hemisphere and in some fronto-temporal channels of left hemisphere, no increase was observed for words. These findings suggest that newborns can extract and remember four different words from continuous speech using only statistical information.
Fast phonetic learning occurs already in two-to-three-month olds
Wanrooij, K. 1 , Boersma, P. 1 & van Zuijen, T. L. 2
1 Amsterdam Center for Language and Communication (ACLC), University of Amsterdam
2 Department of Child Development and Education, University of Amsterdam
An important mechanism for phonetic learning in the first year of life is 'distributional learning', i.e., learning by simply listening to the frequency distributions of the speech sounds in the environment. In the lab, where exposure to speech sound distributions typically lasts only a few minutes, distributional learning has been reported for infants of four months and older. The present study examined whether such fast distributional learning can also be demonstrated before this age. Two-to-three-month-old Dutch infants were presented with either a unimodal or a bimodal vowel distribution based on the English /?/~/æ/ contrast (as in 'bet' vs. 'bat'), for only twelve minutes. Subsequently, mismatch responses (MMRs) were measured in an oddball paradigm, where one half of the infants in each group heard a representative [?] as the standard and a representative [æ] as the deviant, and the other half heard the opposite pattern. The results disclosed a larger MMR for bimodally trained infants than for unimodally trained infants, thus extending an effect of distributional learning found in previous behavioral research to a younger age group and a new method (MMRs as measured from event-related potentials in the electroencephalogram). Moreover, the results revealed a robust interaction between the distribution (unimodal vs. bimodal) and the identity of the standard stimulus ([?] vs. [æ]), which provides direct evidence for an interplay between a previously reported asymmetry in vowel perception and distributional learning. These results were obtained when infants were awake, drowsy or in active sleep during the test; when infants were in quiet sleep there was no effect of distributional learning or perceptual asymmetry.
Modelling the headturn preference procedure: Insights from an end-to-end model of speech processing and behaviour generation
Bergmann, C. 1, 2 , ten Bosch, L. 1 & Boves, L. 1
1 Centre for Language Studies, Radboud University Nijmegen, The Netherlands
2 International Max Planck Research School for Language Sciences
Early language perception and processing abilities are usually investigated using a variety of procedures that rely on a neural or a behavioural response as indirect measure of infants' abilities. Computational models offer a different perspective on language acquisition, since modelling requires precise statements about all assumptions and hypotheses. By the use of computational modelling, we can closely examine (1) which underlying factors need minimally be present to give rise to neural responses or behavioural patterns usually observed during experiments and (2) which factors, be it linguistic or extralinguistic, these methods are susceptible to. We present a model of the headturn preference procedure (HPP) that simulates infant behaviour based on minimal assumptions, such as the lack of language-specific phoneme representations, no explicit segmentation of continuous speech and a module directly converting recognition into overt behaviour, the eponymous headturns. Headturns are only modulated by recognition and infant attention span. Individual attention span accounts for some of the frequently observed between participant variability in HPP studies. According to our model, too short or too long attention span can even mask underlying abilities. As a case study, we replicated HPP experiments on cross-speaker generalisation. Infant data show a mixed pattern. Our model shows that cross-speaker generalisation depends on acoustic properties of the test speaker. This result can help reunite seemingly conflicting infant data on cross-speaker generalisation. Thanks to our modelling result, we can make concrete predictions on the influence of stimulus material on infant performance during a typical HPP experiment on segmentation. Based on our model, we predict that infant performance depends on which speaker infants are being tested with in the lab. Future studies are necessary to investigate this issue by providing detailed acoustic comparisons of test speakers and linking those to infant performance.
Newborns' sensitivity to the biological motion of speech
Guellai, B. 1 , Streri, A. 2 & Kitamura, C. 3
1 SISSA-Language Cognition and Development Lab-Trieste
2 University Paris Descartes CNRS UMR 8158
3 MARCS Lab University of Western Sydney
Among the stimuli encountered soon after birth talking faces constitutes one of the most important source of social information. Here we questioned newborns sensitivity to the biological motion of speech. We presented to newborns, only a few hours after birth, two point-light displays representing talking faces moving at the same time side-by-side. Newborns heard only one sentence that belonged to one of the two displays. Analyses of looking behaviors indicated that newborns looked more at the matching stimulus than at the mismatch. Implications in terms of early multimodal speech perception are discussed.
Anaphoric dependencies in child L1 and L2 Basque
Iraola, M. 1, 2 & Ezeizabarrena, M. 2
1 Universität Konstanz
2 UPV/EHU
Studies on the acquisition of anaphoric dependencies in null-subject languages have shown that the distribution of null and overt pronouns is not optional. Moreover, they have revealed some pragmatic deviance in the use of overt pronouns, which are overaccepted in topic maintenance contexts in developing grammars. In this regard, extralinguistic factors such as the age of acquisition and/or input have been considered to affect children's performance. The present study deals with the interpretation of Basque null and overt pronouns in transitive sentences by different Basque-Spanish bilinguals (L1 and L2 6-7-year-olds and native adults) in order to test the effect of age and input in their performance. Basque is a language lacking true third-person pronouns where the demonstrative hura "that" is said to fulfill the function of the third-person pronoun in other languages. Performance data from a Picture Selection Task revealed that only adults show evidence for a division of labour: whereas null pronouns are coreferent with the preceding subject, overt pronouns are coreferent with the extrasentential referent. In contrast, both child groups, who do not differ significantly in their antecedent choices, show a preference for the preceding subject, regardless of pronoun type. Hence, children's overextension of overt pronouns referring to topic antecedents converges with the general outcome from crosslinguistic developmental studies. Similar to what has been observed in child Italian and Spanish, both child groups are still on the way to acquire the specific discursive properties of the low frequent overt pronoun hura. Input rather than age of acquisition seems to explain the target-deviant interpretations of both groups of bilingual children.
The effect of hearing loss on the perception and production of Infant-Directed Speech
Hay, J. 1 , Robertson, V. S. 2 & von Hapsburg, D. 2
1 Department of Psychology, University of Tennessee
2 Department of Audiology & Speech Pathology, University of Tennessee, Health Science Center
Infant-directed speech (IDS) is an exaggerated form of speech caregivers use with infants. It has higher pitch, longer duration, an enlarged vowel space, and is more rhythmic than adult-directed speech (ADS). IDS is thought to capture infant attention, and infants with normal hearing (NH) show a preference for listening to IDS over ADS. This preference may be adaptive as IDS also facilitates early language learning. Infants with hearing impairment (HI), who received a degraded speech signal, may not have sufficient access to pitch contours present in IDS to benefit from it during early language learning. Thus, the purpose of this study is to 1) determine whether infants with HI also prefer IDS over ADS, and 2) to compare the acoustic characteristics of speech directed to NH and HI infants in their home environment. Thirty-six infants, 9 HI infants, 9 NH hearing-matched controls and 9 NH chronological age-matched controls were tested on their listening preference for IDS over ADS using a central fixation preference procedure. Three to six hours of spontaneous infant-parent conversational interactions were audio recorded in the home with a digital recorder (LENA), for each infant. Five-minute segments were randomly selected and acoustically analyzed. Results confirmed that HI infants significantly preferred listening to IDS over ADS. The preference for IDS was also seen in younger hearing-matched NH controls, but not older NH controls. Preliminary production data from the mothers of 5 HI and 5 NH infants suggests that mothers of HI infants produce their utterances with a higher maximum pitch. Our results suggest that HI infants appear to have sufficient access to the speech signal to display a developmentally appropriate preference for IDS over ADS. Exaggerated pitch contours seen in parents' speech to their HI infants may support preferences for IDS at this critical time in language development.
Positive affect in infant-directed speech: Expressed through and recognized from the vowels?
Benders, T.
Center for Language Studies, Radboud University Nijmegen
It has been claimed that the intonation in infant-directed speech (IDS) expresses affect and that the enhanced auditory contrast between caregivers' corner vowels is to help their baby's language acquisition, but is unrelated to affect (Kuhl et al., 1997; Uther et al., 2007). However, our Study 1 shows that the vowels in Dutch IDS have reduced contrast, and that their acoustic properties are shaped by positive affect (on Norwegian, Englund & Behne, 2005/2006). Study 2 tests whether Dutch infants prefer speech with such happy-sounding vowels. Study 1 (completed). Eighteen Dutch mothers were recorded while speaking to their infant at 11 and 15 months and an adult. The results showed that: 1) The vowel space was smaller in IDS than in adult-directed speech (ADS); 2) The second formant (F2) of the vowels and center of gravity (CoG) of /s/ were higher in IDS; 3) F2 was higher to 11- than to 15-month olds. Because Dutch mothers reduce the auditory contrast between vowels in IDS, they do not seem to promote their infant's language development. A higher F2 and CoG are associated with the expression of positive emotions (Tartter & Braun, 1980; Kienast & Sendlmeier, 2000) and affect is expressed stronger to younger infants (Bornstein et al., 1992). Therefore, Dutch mothers' infant-directed vowels are primarily carriers of positive affect. Study 2 (in progress). Stimuli are normal sentences, sentences with enhanced intonation, and sentences with raised formants. In a preferential-listening procedure, Dutch 6-month-olds are expected to prefer the speech with enhanced intonation over the normal speech (Fernald & Kuhl, 1987; Singh et al., 2002). If infants also listen longer to speech with raised formants, they recognize positive affect from the vowels. Together, these studies test whether the affective character of IDS is, partly, contained in the realization of the vowels.
Distributional information in English and Brazilian-Portuguese Infant-Directed Speech: Language differences and their implications for acquisition
Echols, C. 1 , Souza, A. 2 , Barbosa, P. 3 & Cardoso-Martins, C. 3
1 University of Texas
2 Concordia University, Montréal
3 Universidade Federal de Minas Gerais, Brazil
Building on research documenting the potential value, for language acquisition, of information in the input, we analyze distributional information in Brazilian Portuguese (BP), and compare it to English. BP is of potential interest because adult speakers often omit pronouns (which is grammatical) and other grammatical elements from their speech, raising questions about the effects on distributional information of grammatical and ungrammatical omissions from speech directed to children. Maternal speech samples collected from 18 Brazilian and 18 United States mother-child dyads during semi-naturalistic interactions when children were 9, 13 and 18 months old were transcribed and coded for various grammatical properties. Initial analyses from this sample showed that 61% of the BP mothers' productions included omissions (with only 13% being grammatical) contrasted with 16% incomplete sentences for the English mothers, verifying important differences between languages in the completeness of utterances directed to infants. To evaluate distributional information, we built on frequent frame (Mintz, 2003) and Construction Grammar (Langacker, 1992) approaches. We identified, for all possible 3-word sequences, a set of frequent frames-sequences following either a word-X-word or word-word-X format that met a frequency criterion. We conducted collostructional analyses to measure the strength of association between the frequent frames and particular grammatical categories, focusing here on noun and verb categories. In both languages, the frames predicting nouns are virtually non-overlapping from those predicting verbs, indicating that both languages contain sufficient information to distinguish basic grammatical categories. The results of the collostructional analyses suggest that although the frequent frames provide statistical information that distinguishes noun from verbs in each language, relative strengths of association vary across language, grammatical category and frame type (word_X_word, word_word_X). Results will be discussed in relation to differences across languages in the distributional contexts of linguistic information and will be compared to vocabulary acquisition data from the children.
Smile as a 'hook' for early communication between mothers and their two-month-old infants during peek-a-boo play
Szufnarowska, J. & Rohlfing, K. J.
Bielefeld University
The mechanism of turn-taking is the basic rule in dialogue and seems to be present even in the earliest interactions between mothers and their infants. At the age of about 2 months, a developmental change takes place which increases the scope of mutual responsiveness: Infants become able to maintain visual attention for a longer time and start to smile socially. Infants? gazing at the interaction partners provides framework for taking turns, and the emergence of social smiling seems to be crucial for the mutual engagement in the face-to-face exchange. An in-game situation can be a good opportunity to practice and develop communication skills in infants as it provides them with an enjoyable context motivating to take a turn. In our study, we contribute to our understanding of the nature of early dialogues. 21 mother-infant dyads from Poland were visited at home, where the mother-infant peek-a-boo play was videotaped from three video camera perspectives. We examined whether the two-month-olds are willing to engage in a peek-a-boo play and under what condition turn-taking in form of social smiling can be observed. We observed that 50 % of the infants reacted with smiles during the play. All infants who were able to establish mutual gaze for over 65 % of interaction time responded with a smile. Every interaction consisted of several units of play. The infants were willing to respond contingently with smile in a given unit of play only when the mothers were able to recruit their attention within 1 second after uncovering. If the infants looked at the mother after longer break, they did not smile. Our results confirm the findings that the mutual attention is a necessary condition for infants? turn-taking but also suggest that infant?s smile follows only the contingent gazing at the mother.
Tone perception abilities of typically developing and infants at risk of dyslexia from monolingual and bilingual language backgrounds
Molnar, M. , Lallier, M. & Carreiras, M.
Basque Center on Cognition, Brain and Language (BCBL). Donostia. Spain
It has been established that stress cues (e.g., duration or pitch) are particularly important for accurately segmenting speech as early as 8 months of age in monolingual infants (e.g., Jusczyk and Aslin, 1995); also, stress cues seem to contribute to the early development of phonological organization (e.g., Jusczyk, Goodman & Baumann). As performance on stress discrimination and identification tasks has been linked with reading and phonological processing (Clin, Wade-Woolley and Heggie, 2009), the current project was designed to investigate whether infants who are at risk of reading disorders would exhibit disadvantage in a stress processing task as compared to their not at risk peers. Since it is unclear whether early exposure to two languages as opposed to one would alter stress processing skills in the first year of life, we also recruited infants raised in bilingual families. Using the habituation preference procedure, we have tested 9-month-old infants with four different backgrounds: Spanish monolingual and Basque-Spanish bilingual infants with no reported family history of dyslexia, and Spanish monolingual and Basque-Spanish bilingual infants with known family history of dyslexia. We have examined infants? ability of perceptually grouping tones that varied in terms of duration and pitch. Previous cross-language experiments with monolingual infants and adults also showed that listeners have different perceptual grouping preferences based on the specific phrasal stress patterns of their native language (Iversen et al., 2008; Yoshida et al., 2010). Our preliminary results suggest that linguistic background and at risk status affect stress perception in infancy.
The role of consonants and vowels in lexical recognition at 5 months
Bouchon, C. 1, 2, 3 & Nazzi, T. 1, 2, 3
1 University Paris Descartes ? Laboratoire de Psychologie de la Perception UMR 8158
2 CNRS ? UMR 8158
3 LabEx Empirical Foundations of Linguistics
At 6 months infants already know a few sound patterns (Mandel et al., 1995; Tincoff & Juszcyk, 1999; 2011;
Bergelson & Swingley, 2012), and use familiar words to segment unfamiliar words from fluent speech (Bortfeld et al., 2005; Mersad & Nazzi, 2012). This study investigates the relative contribution of consonants (Cs) and vowels (Vs) to lexical recognition before 6 months. Cs carry more information at the lexical level (Nespor et al., 2003), yielding to a consonantal bias in lexical tasks in adults and infants older than 14 months. While a consonantal bias was observed at 12 and 8 months (respectively: Hochmann et al., 2011; Nishibayashi & Nazzi, 2012), 6?montholds exhibited a reversed vocalic bias in a lexical task (Hochmann, 2010). It is still unknown how much speech exposure infants need before Cs, although acoustically less salient and phonetically less defined than Vs, become preferential cues in word processing. 5 month?olds already recognize their own names (Mandel et al., 1995) and ERPs show that this recognition occurs from the first phoneme (Parise et al., 2010). Using HPP, we tested the sensitivity of French?learning 5?month?olds to a minimal initial phonetic change in their name, by presenting them with repetitions of their correctly vs. mispronounced names. To determine the relative weight of Cs and Vs, this change was either consonantal (n=30, e.g. victor/zictor) or vocalic (n=26, e.g. eliot/øliot). The vocalic change yielded a strong preference for the name (p = .002), whereas the consonantal change did not (p = .26). Results in control groups ruled out acoustic effects. Thus the consonantal bias might not have emerged yet by 5 months of age, and infants may be relying mostly on acoustic saliency to detect phonetic changes in familiar words.
Distributional learning in infants at familial risk of dyslexia
Kerkhoff, A. , de Bree, E. & Wijnen, F.
Utrecht University
At 18 months infants are sensitive to distributional patterns in the input, such as the relationship between ?is? and ?ing? in ?is running? (Santelmann & Jusczyk, 1998) and nonadjacent dependencies in an artificial ?aXb? language (Gómez, 2002). English ?frequent word frames? (e.g. you X it) were used by 12-month-olds for categorisation of novel nouns and verbs (Mintz, 2006), and Dutch ?frequent morpheme frames? (e.g. we X-en ?we X-plur?) were used by 16-month-olds (Kerkhoff et al., 2010). However, 19-month-old infants at family-risk (FR) of dyslexia (i.e. infants with a dyslexic parent) were not sensitive to a morphosyntactic dependency in Dutch (Wilsenach & Wijnen, 2004), and 18-month-old FR infants could not track non-adjacent dependencies in an artificial language (Kerkhoff et al., 2013). The current study investigates whether FR infants can use Dutch frequent morpheme frames for categorisation. Two groups of typically developing (TD) children aged 16 months (n=29) and 19 months (n=27) were compared to two groups of FR infants aged 16 months (n=35) and 19 months (n=21). In a headturn study, infants were exposed to sentences such as we freppen niet hoor ?we do not frep? in the training phase, and heard both a consistent sentence (wat frept er nou ?what freps now?) and an inconsistent one (ik zie een frepje ?I see a frep-dim?) in the test phase. Results showed that infants listened longer to sentences that were consistent with training (e.g. a novel ?noun? in a noun frame). There was also an interaction between Consistency, Group and Attention, indicating that for the TD group, only the infants who were ?attentive? during the test phase (i.e. those not showing restless behaviour) showed a familiarity preference, whereas this pattern of results was reversed for the FR children. This suggests that attentional factors may explain individual differences in children?s language learning skills.
The acquisition of aspect and clitics in spanish
Torrens, V. & Escobar, L.
Facultad de Psicologia, Universidad Nacional de Educación a Distancia
Telicity is a property of accomplishment verbs with a direct object (Krifka, 1998). Tense has also a clear effect and the present perfect is usually part of telic predicates. In some languages we can also identify telicity markers as in English (Brinton, 1988) or Spanish (Nishida, 1994; Sanz 1999). In this paper we present empirical research on the acquisition of Spanish ?reflexive? clitic se. We applied a truth-judgment-value task in order to see whether Spanish children are able to acquire the Spanish telicity marker se at a similar early age. The results of our study indicate that this is not the case. Spanish children from 3;0 to 6;0 have more difficulty with the telicity marker se than with other ordinary telic predicates without the clitic. We now argue that the difficulty lies in the operator character of aspectual se. Clitics are challenging in language acquisition since they behave partly like independent words and partly like affixes. They seem to be more autonomous than affixes but they attach phonologically to a host, contrary to words. Likewise, clitic se contrasts with the English particle up which needn?t be attached phonologically to a host. Note that unlike up, clitic se in the second person singular te form is enclitic with the imperative. In our analysis clitic se is an operator, and it undergoes movement. We propose that se moves to check its operator feature against the aspectual head Asp of the Spanish clause and suggest that ?merge? is preferred over ?move? in children?s grammars as a more economical operation, which falls within the minimalist guidelines (Chomsky 1995, 1998). In this way we can finally explain the statistically significant difference (p < 0.001) in the acquisition of telic predicates with and without se in our experiment.
A multimodal advantage in statistical and artificial grammar learning
Glicksohn, A. & Cohen, A.
The Hebrew University
Our environment is richly structured with objects producing correlated information within and across sensory modalities. A prominent challenge faced by our perceptual system is to learn such regularities. Numerous studies reveal both infants? and adults? exceptional ability in learning various types of regularities. Moreover, during initial development multimodal information benefits certain forms of learning. We ask (1) whether learning occurs separately within modalities (e.g., vision, audition) or whether learning is a-modal, and (2) whether multimodal information benefits learning. We explored these questions using two prominent learning paradigms: statistical learning (SL), referring to learning of statistical regularities between elements such as transitional probabilities, and artificial grammar learning (AGL), referring to learning the underlying grammar of a set of exemplars. Adult participants were familiarized with one novel environment: (a) visual, (b) auditory, or (c) an audiovisual environment consisting of one type of regularities (transitional probabilities, or exemplars adhering to a grammar). Two main findings are noted: (a) learning occurred in all three environments and for both types of regularities, supporting the existence of an a-modal learning mechanism(s); (b) Learning for the audiovisual regularities was significantly higher than learning of the visual- or auditory- regularities, suggesting an advantage for multimodal learning. We suggest that learning of regularities is accomplished via a general, a-modal, learning mechanism which is sensitive to multimodal information. We further suggest that multimodal information plays a crucial role in learning throughout our lives.