SAP presents results at the International Conference on Infant Studies, June 30 - 3 July 2018 Philadelphia, USA

 

Symposia and Posters, see abstracts below: 

 

 

Abstracts:

Humans are born in social groups organized hierarchically with two well-known roles: leaders and followers. Leaders are defined as individuals who have a significant influence on others' actions (followers). We have addressed the early representation of leader-follower relationships in infants. First, 12-, 15-, and 18-month-olds were familiarized with animations depicting two agents (the followers) selectively imitating the intransitive actions of another agent (the leader). During test, the infants observed either a follower failing to follow the leader's path (incongruent outcome), or the leader failing to follow a follower's path (neutral outcome). The participants looked significantly longer at the incongruent than at the neutral outcome from 15 months of age. These results demonstrate infants' capacity to represent a leader-follower relation as asymmetric and stable across time. Next, we explored infants' intuitions about who would be selected as a leader between two agents with different power (amount of control over limited resource).Eighteen -month-old infants were familiarized with one agent (the observer) watching two other agents successfully picking up a ball one after the other. However, when both agents tried to pick up the ball at the same time, the same agent always prevailed (the winner). Critically, we manipulated how the winner prevailed, through communication or using force. During test, the observer followed one of the agents - either the "winner" (congruent output) or "loser" (incongruent output). Participants looked significantly longer at the incongruent outcome than at the congruent one only when the force was not used by the winner. These results evidence infants' capacity to link social power with leadership considering the way the agent succeeded in conflict situations. Altogether, our results demonstrate infants' sensitivity to agents' different social status and how this leads to asymmetrical relationships.

 

From early on infants navigate the social world selectively, preferring some individuals to others. Some of the cues that predict infants' social preferences have been defined, such as familiarity with the language/accent an individual speaks (Kinzler et al., 2007; Begus et al., 2016) or reliability of the social partner when providing information (Tummeltshammer et al., 2014). Nevertheless, the underlying mechanisms guiding early life social selectivity are still a subject of debate. Here we highlight the importance of successfully interpreting others' intentional actions to interact and learn from them, which have been proposed to be based on the principle of rationality (Gergely & Csibra, 2003). From early infancy, humans implicitly assume agents to perform the most efficient means action to obtain a desired reward by minimizing the costs they expect to incur (Jara-Ettinger et al., 2016; Liu & Spelke, 2017). <br> In a set of two studies, we used a violation of expectation paradigm to investigate whether the assumption of efficiency guides 15-months-old infants' expectations about others' social preferences. In the first study, we familiarized 24 participants with two animated geometric figures that performed the same goal-directed action, while a third agent observed them ("observer"). After familiarization, the efficiency of the agents' actions was manipulated. In the final test, the "observer" approached the efficient agent ("expected condition") and the inefficient one ("unexpected condition") and we measured infants' total looking time to the screen in each condition. Infants looked significantly longer to the "unexpected condition" over the "expected condition" (p < 0.001). Study 2, as a control of Study 1, confirmed that infants' preferences were guided by the rationality of theagents' movements and not by the novelty of the rational agent's actions during the test 75 phase. <br> These results suggest that the principle of rationality has an important role in early life social selectivity, as a basis to interpret others' behaviours. This proposal sheds light on the mechanisms that might underlie previous findings in infants' social selectivity, such as the preference for accurate or reliable partners as individuals who behave rationally (Tummeltshammer et al., 2014; Zmyj et al., 2010). We are currently investigating infants' sensitivity to cultural goals, as a way of explaining actions that might seem apparently inefficient. Humans are constantly exposed to cultural behaviours or rituals that might involve inefficient actions to obtain a simple reachobject goal, but that are rational beyond a certain cultural goal.

 

As opposed to adults, when presented with a face talking in their native language, infants in their first year of development tend to shift their visual attention from the eyes to the mouth of the speaker. This preference is stronger in infants growing in a bilingual environment, suggesting they need to rely more on the audiovisual redundant cues provided by the mouth area of talking faces to deal with their dual language learning (Pons, Bosch & Lewkowicz, 2015). In a previous study computing mean preference scores on eye-tracking measures, we showed that the preference of 12-month old monolinguals and 15-months-old bilinguals for the mouth area of a speaker prevent them from anticipating subsequent non-speech gestures displayed in her eyes region (Eyebrow-raise movement as opposed to Lip-Protrusion movement, Figure 1). In this work, our aim was to measure the stability (within participants) of their patterns of visual exploration of the speaker's face (Eyes, Mouth, Rest of the Face and Elsewhere, see Figure 2A), independently of their mean scores. To do so, we computed Markov matrices of transition probabilities of each participant's gaze locations across the 19 trials of the experiment (Figure 2). Results show (see Figure 2B) that controlling for age, condition (Eyebrow-raise and Lipprotrusion) and mean preference scores, bilinguals tend to switch more to the irrelevant 146 regions of the speaker's face during the Speech Event (Rest of the Face & Elsewhere) while monolingual remained more on the relevant regions (Eyes and Mouth). These results suggest that bilingual infants, as compared to their monolingual peers, exhibit different attentional strategies when perceiving faces talking in their native language. More precisely, our observations can be interpreted as resulting from different equilibria between top-down attentional constraints and bottom-up saliency capturing attention. These results may provide new answers regarding the cognitive mechanisms at play in early language acquisition in bilinguals. Pons, F., Bosch, L., & Lewkowicz, D. J. (2015). Bilingualism modulates infants' selective attention to the mouth of a talking face. Psychological Science, 26(4), 490-498. 

 

The pervasive nature of culture makes it a unique lens for examining relationships between individuals and their environment. Multiple studies in the past decade have documented cross-cultural differences in temperament, but nearly all have focused on only two to four cultures (c.f., Putnam & Gartstein, 2017), limiting inference regarding the role of societal processes suggested as contributors to cross-cultural differences. The current study addresses this shortcoming through analyses of differences in infant temperament gathered as 60 independent samples in 31 countries. Methods: Researchers from Australia, Belgium, Canada, Chile, China, Czech Republic, Finland, Germany, Hungary,  Israel, Italy, Japan, Korea, Latvia, Malaysia, Malta, Mexico, Netherlands, New Zealand, Nigeria, Poland, Portugal, Romania, Russia, Spain, Switzerland, Taiwan, Turkey, United Kingdom, Uruguay and United States provided parent-report data on a total of 9244 infants (4390 female, 4799 male; 55 unknown; mean age = 8.42 months). Temperament was measured with the Infant Behavior Questionnaire - Revised (Gartstein & Rothbart, 2003), yielding scores for Surgency (SUR), Negative Affectivity (NEG) and Regulatory Capacity (REG). Results: Sex*Nation ANOVAs, with age included as covariate, indicated substantial effects of nation for the three dimensions. F-values, R2 and marginal means for the three dimensions across the 31 countries are shown in Table 1. Regarding SUR, post-hoc (Tukey) tests indicated infants from Mexico, Romania, Turkey, Malta and Malaysia scored significantly higher than > 15 other cultures, whereas U.K., Switzerland, Japan and Latvia scored lower than most others. Results concerning NEG suggested higher scores in Nigeria, Turkey, Portugal, Romania, Malaysia, Switzerland, Chile and Israel than more than half of other nations; with the Netherlands, Finland, Canada and the U.S demonstrating lower levels than most other countries. REG scores were higher than most nations in Malta, Mexico, Belgium, Romania, Turkey, Uruguay, Switzerland, Nigeria, Chile, Italy and Israel; and lower than most others in Japan and Taiwan. Nation-level marginal means for the three dimensions were then correlated with archival scores on Hofstede's six dimensions of cultural orientation and Gross National Income (GNI) per capita (see Table 2). Consistent with meta-analytic results reported by Putnam and Gartstein (2017), high levels of NEG were linked with Collectivism, high Power Distance and low GNI; high SUR with Short-Term Orientation; and REG with Indulgence. In addition, Surgency was associated with Collectivism, high Power Distance and low GNI. This investigation provides the most extensive perspective to date on worldwide patterns of individual differences in infants, and suggests multiple directions for future research. Primary among these is the pervasive issue of the degree to which parent reports represent objective differences or subjective interpretations of child characteristics. Current views (e.g., Rothbart & Bates, 2006) suggest temperament scores reflect both, both possibilities as useful in revealing implications of culture for human development. Another critical direction concerns exploration of proximal mediators of the distal force of societal differences. It is our hope that the Global Temperament Project database becomes a valuable tool for developmental scientists to better understand these connections.

 

When learning their native language, one of the first challenges infants are faced with is the establishment of the phonetic categories. It has been proposed that infants acquire these phonetic categories by computing the distribution of speech sounds in the acoustic  space (Maye, Werker and Gerken, 2002). Around the same age as the perceptual reorganization takes place, infants acquire their first words (Bergelson and Swingley, 2009; Tincoff and Jusczyk, 1998). However, the possible top-down influence of word-level information on the phonetic reorganization process has been poorly investigated. One potentially interesting group of learners to investigate this question is that of bilinguals learning typologically close languages, such as Spanish and Catalan. These languages share a high number of cognates among their translation equivalents. Moreover, cognates differ mainly in their vowels, inducing a high vocalic variability in the speech stream (for instance "pOrt@" vs. "pwerta" or "xukulat@" vs. "t∫okolate") and incrementing the number of minimal pairs in the stimuli. Feldman et al (2012) proposed that the nonminimal pairs present in the speech stream may help separate an overlapping vocalic category by providing a clearer word context. They explored the influence of word-level information on the discrimination of a native overlapping vocalic contrast with both adults and 8 month-olds. Using a corpus of pseudo-words containing the vocalic contrast, they familiarized their participants according to two word-context conditions: the Minimal Pair (MP) condition where the vocalic contrast appeared in all the pseudo-words and the NonMinimal Pair (NMP) condition where the contrast appeared in distinct word-context. Analyzing the two test blocks, they found that adults who have been familiarized with the NMP condition tended to assign the test syllables to different categories more often than the participants familiarized with the MP condition. Similar results were found in 8 montholds with a Head-turn Preference Procedure. We want to address the possible impact of bilingualism/cognateness in the establishing of phonetic categories by comparing Spanish-Catalan bilinguals and monolinguals. We adapted Feldman and al. (2012) procedure to test both adults and infants on their discrimination of a difficult to perceive non-native vocalic contrast (British English /ɒ-ʌ/ contrast). We have tested so far 46 adult participants, comparing exposure to minimal pairs (MP group) and to non-minimal pairs (NMP group) in a discrimination task. We calculated their sensitivity to the /ɒ-ʌ/ contrast (d'). As Feldman et al (2012), we analyzed the exposure and block interaction and added the language variable (mon vs. Bil). The analysis yielded a triple interaction of language group, exposure and block (p=0.003) and a main effect of block (p<0.01) (see figures). The analysis of the first block showed that exposure to NMP increased sensitivity in monolinguals (p= 0.02) but not in bilinguals (ns). Infant data collection is currently underway. Adult monolingual results replicate the pattern reported by Feldman et al (2012). Bilinguals seem to behave differently from the monolinguals but more data is needed before firm conclusions can be drawn.

 

To probably overcome the challenge of learning two languages at the same time, infants raised in a bilingual environment pay more attention to the mouth of talking faces than same-age monolinguals. Here, we examined the consequences of such preference on monolingual and bilingual infants' ability to perceive additional information coming from the eyes or the mouth region of talking faces. In a previous study using the same paradigm, we showed that at 15-months of age, both monolingual and bilingual 15- month-olds could detect the apparition of a visual cue appearing in the eyes region but only 15-month-old monolinguals and 18-month-old bilinguals could learn to anticipate its appearance during the sentence phase. One possible explanation for this result is that at 15 months of age, bilinguals, as compared to their monolingual peers, need to rely more on the cues provided by the mouth region of the speaker to cope for their challenging language environment. Using the same paradigm (Figure 1), we tested whether at a younger age (12-month-olds), both monolingual and bilingual infants, who are less expert to process their native language (and may thus rely more on the mouth region of the talker), fail to anticipate the visual cue in the eyes (Eyebrow-raise movement) as opposed to the mouth region (Lip-protrusion movement). Growth curve analysis was used to analyze the evolution (over the course of the 19 trials) of the Proportion of Total Looking Time to the eyes minus the mouth region of the speaker, during the last 50% of the Speech Event (Figure 2). Surprisingly, contrary to our previous results, bilinguals in the Eyebrow-raise condition (N=20), as opposed to the ones in the Lip-Protrusion condition (N=20), significantly anticipated the apparition of the Eyebrow-raise movement by increasing their looking time to the eyes region of the speaker (p < .05). However, no similar change was observed in same-age monolinguals (N=20 per condition, t < 1). We are now gathering new data with infants aged between 12 and 18 months of age to explore the respective role of vocabulary growth and maturation of attention to explain these different developmental trajectories.

 

Recent evidence suggests that infants exhibit increased attention to stimuli after listening to a native speaker's speech (Begus, Gliga, & Southgate, 2016; Marno et al., 2016). These findings have been interpreted as early attentional biases for native speakers (Kinzler, Dupoux, & Spelke, 2007) reflecting infants' preference for potential sources of information. However, the question of whether early social biases lead to better learning of new information remains unclear. This study aimed to ask whether infants' selective listening to novel tunes introduced by native speakers (Soley & Sebastián-Gallés, 2015) also extends to better learning of those tunes. To answer this question, three experiments were conducted (n = 16 per experiment) with 7-10 months old infants. Using a habituation paradigm, Experiment 1 first established that infants cannot readily discriminate two novel tunes, Tune A and Tune B (See Figure 1 for the notation). After being habituated to Tune A, infants' listening times to Tune A and Tune B during test trials did not differ (t(15)  .756, p > .4). Accordingly, in the next experiments infants were initially familiarized with Tune A, either by a native speaker (Experiment 2) or a foreign speaker (Experiment 3). During each one of the three familiarization trials, a female actor appeared on the screen, she spoke briefly in infant-directed way and then played Tune A by pressing a button on a music player standing next to her. She silently smiled while Tune A was played for ~21 s. After this brief familiarization phase, infants' discrimination of Tune A and Tune B was measured using a habituation paradigm as in Experiment 1. The female actor was bilingual and she spoke in infants' native language in Experiment 1, while she spoke in a foreign language in Experiment 2. Listening times to Tunes A and B differed significantly between Experiments 1 and 2 (F(1, 30 = 4.66, p < .05). In Experiment 2, infants listened longer to Tune B compared to Tune A during test (t(15) = -2.15, p < .05). In contrast, listening times did not differ between Experiments 1 and 3. In Experiment 3, infants listened equally to Tune A and B during test (t(15) = -.487, p >.6). Across Experiments 2 and 3, infants listened longer to Tune B compared to Tune A (F(1, 30 = 4.64, p < .05), and there was also a marginal interaction between the listening times and Experiment (F(1, 30 = 4.66, p = .09). Infants attended equally to familiarization videos in native and foreign conditions (p > .6). Thus, while some learning happened in both conditions, infants could discriminate the two tunes in the native language condition, but not in the foreign language condition. These findings suggest that early attentional biases might have important ramifications for young infants and that infants might more readily acquire information when it is introduced by native speakers.

 

Across cultures, adults engage in infant-directed speech (IDS) when they address infants. Compared to adult-directed speech (ADS), IDS is characterized as having higher pitch and more melodic features (e.g., Fernald, 1989). Several studies show that infants prefer IDS over ADS (e.g., Fernald, 1985; Werker & McLeod, 1989). Further, infants also attend more to individuals, who previously produced IDS (Schachner & Hannon, 2011). Here, we explore whether infants expect IDS to be directed at infants and ADS to adults. Spanish or Catalan hearing monolingual infants (12-15 months, N = 16) were first familiarized with three characters, two of which represented adults and one represented an infant (e.g., Johnson, Dweck, & Chen 2007; Spokes & Spelke, 2017). In each of three familiarization videos, these characters appeared one by one, each produced a vocalization (laughter) and then they all moved together in synchrony. After the familiarization phase, infants received 8 test trials (2 blocks). In these trials, infants saw either two adults or one adult and one infant facing each other and the same adult produced either IDS or ADS across trials (Speech segments were either in Spanish or Catalan depending on the native language of the infant). Thus, half of the trials were "inconsistent", such that infants saw the adult addressing the infant with ADS, or the other adult with IDS. The other half of the trials were "consistent", such that infants saw the adult addressing the infant with IDS, or the other adult with ADS (See Figure 1). Trial durations were fixed for familiarization phase (~40 s) and were infant-controlled for the test phase (up to 40 s). Overall, infants' looking times to consistent and inconsistent videos did not differ (F(1, 30 = 1.24, p > .2). However, we found a significant interaction between infants' looking times to consistent and inconsistent videos and the block (First or Second) (F(1, 30 = 7.6, p < .05). Separate analyses of the first and the second blocks revealed that infants did not look longer to inconsistent compared to consistent events in the first block (t(15) = -1.15, p > .2), but they did so in the second block (t(15) = 2.74, p < .05) (See Figure 2). These findings suggest that infants might have certain expectations regarding whom IDS and ADS should be directed to. The block effects might be due to familiarization and test trials being rather different (i.e., infants saw the characters speak for the first time during the test trials).  Infants' looking patterns might also be driven by the content of the speech in IDS and ADS. Thus, testing another group of infants with a language other than their native language would be informative. Testing younger infants might also help understanding the role of exposure in infants' expectations regarding IDS and ADS communications.

 

Previous research indicates that infants can discriminate between languages belonging to different rhythmic classes already at birth. However, within rhythmic class discrimination starts to take place around the fourth-fifth month of life (Nazzi, Jusczyk, & Johnson, 2000). What type of information infants use to perform such discrimination is yet to be determined. Although infants have not yet established vowel categories at such a young age, they may already have some rough knowledge about the distributional properties of their vowel system. Therefore, they might use such distributions to discriminate languages. We tested this hypothesis by investigating the discrimination capacities of 4.5 month-old infants when listening to sentences from two dialects of Catalan (Southern and Central) -in experiment 1- and when listening to sentences of Southern Catalan and Spanish -in experiment 2. Spanish and (Central-Standard) Catalan have quite different vowel distribution due to the existence of vowel reduction in Catalan (resulting in very few mid vowels), but not in Spanish (see figure 1). Interestingly, Southern Catalan does not bear vowel reduction, yielding a distribution of vowels very similar to the Spanish one. We used the same procedure as in Bosch and Sebastian-Galles (2001). In the first study, 21 4.5-month-old infants having Central Catalan as their dominant language were tested. In the second study, 13 4.5-month-old infants having Spanish as their dominant language were tested. The preliminary results of the two experiments are shown in figure 2. Infants were able to discriminate the two dialects of Catalan (t(20)= 3.5244 ; p<0.0021, two-tailed) but they could not discriminate Southern Catalan from Spanish (t<1). We are currently finishing the collection of the data. Current results support the hypothesis that infants are sensitive to differences in the distribution of vowels across languages and that they can use these differences to distinguish languages or dialects.