Master 2017 2018
Stages de la spécialité SAR
Perception de l’intentionalité dans la communication pré-verbale (psychoacoustique/traitement du signal/psychologie du développement)

Site : CREAM
Lieu : The internship is a collaboration between the Nanterre BabyLab ( and the CREAM project in IRCAM (“Cracking the Emotional Code of Music”, The student will be co-supervised by both teams, and the internship will be co-located in both laboratories: preparation of the psychoacoustics procedure and of the stimuli (first half of the project) will be done in the “Perception and Sound Design” team, IRCAM (1 Place Stravinsky, 75004 Paris), while data collection (second half) will be done in University of Nanterre (200 avenue de la République, 92000 Nanterre)
Encadrant : Maya Gratier (Babylab, Département de Psychologie, Université Paris Nanterre), Louise Goupil & Jean-Julien Aucouturier (Perception and Sound Design Team, IRCAM UMR9912)
Dates :du 01/03/2017 au 31/07/2017
Rémunération :standard ( 500€/m)
Mots-clés : Parcours ATIAM : Acoustique, Parcours ATIAM : Informatique musicale, Parcours ATIAM : Traitement du signal

Cliquer ici pour vous authentifier


Human infants produce adult-directed social vocalizations from the first days after birth, and by 6 weeks of age their social vocalizations have become melodic and resonant (Oller, 2000), with clearly audible prosodic patterns, intonations or "melodies". However, little is known about how adult caregivers process these signals. One topic of particular interest is how adults can disambiguate a baby’s intentions from the sound of his/her voice. For example, a baby pointing to an object ("dah" !) may be interpreted as e.g. assertive/imperative ("look, a toy !") or uncertain/interrogative ("what is this ?"). It is an open question how adults use the intonation of the baby’s voice to interpret such intentions. This question has important consequences to understand the normal development of preverbal (and then verbal) communication, and may also help understand disorders of attachment between parents and infants (Shai, D. & Belsky, 2017)

This question is actively studied by the Nanterre Babylab, which has recently shown e.g. that, even without knowing a baby’s age, adult listeners find vocalisations from 5-month-olds more intentional than vocalisations from 1-month-olds. In addition, adults can distinguish baby vocalisations that are produced when the baby is alone, or when the baby is interacting with an adult (Gratier et al., 2016). However, it has remained difficult to characterize what exact prosodic patterns drive such judgements in parents/adults.

The CREAM project in IRCAM has recently developped a new experimental paradigm, combining state-of-the-art signal processing (voice transformations with phase vocoder) and psychoacoustical reverse-correlation, that seems well suited to study this question. In this paradigm, adult participants are asked to listen to pairs of stimuli (e.g. baby vocalisations) manipulated algorithmically to present random prosodic variations, and to judge in each pair which is e.g. more assertive or interrogative. Using data-driven analysis techniques correlating the participants’ responses to the acoustical characteristics of the stimuli, one can then reveal what mental representation participants have of e.g. a given intention (Ponsot et al, 2017). The technique was recently applied to adult speech, but has never been applied to baby vocalisations.

Student role in the project : The project will consist of a series of psychoacoustical experiments to uncover adults’ mental prosodic representations of intentionality in baby preverbal vocalisations. The student will participate in stimulus selection (using corpuses of baby recordings collected at the Babylab Nanterre), stimulus manipulation (using skills in audio signal processing and a Python toolbox recently developped in the CREAM lab in IRCAM), experiment building (using skills in computer programming in Python), data collection (1-hour experiment on 20+ participants in Babylab Nanterre), data analysis (using skills in statistics, and tools like Python/pandas) and manuscript preparation.

Student’s profile : We are looking for a Master’s student with a strong interest in experimental psychology and psychoacoustics, as well as sound processing skills. Experience with the Python programming language, and in particular with the jupyter notebook environment, the pandas and psychopy libraries, will be appreciated. Any experience with experimental data collection and listening tests, while optional, will be also appreciated.

To apply, send a CV and cover letter by email to Maya Gratier, Louise Goupil & Jean-Julien AUCOUTURIER,


Oller, D.K. (2000). The emergence of the speech capacity. Mahwah, NJ : Lawrence Erlbaum Associates.

Shai, D., & Belsky, J. (2017). Parental embodied mentalizing : how the nonverbal dance between parents and infants predicts children’s socio-emotional functioning. Attachment & human development, 19(2), 191-219.

Gratier, M, Lacheret, A., Bobin-Bègue, A, Devouche, A & Apter, G. (2016). Judgments of Emotion and Communicative Intent in Infant Vocalization at 1 and 5 Months. International Conference on Infant Studies (ICIS 2016), May 2016, New-Orleans, United States.

Ponsot, E., Burred, JJ., Belin, P. & Aucouturier, JJ. (en relecture, 2017) Cracking the social code of speech prosody using reverse correlation