Thèse Comment les Croyances des Parents Concernant l'Esprit de leur Nourrisson Influencent leurs Premiers Échanges H/F

Doctorat.Gouv.Fr

The first aim of this project will be to design and validate a questionnaire to measure parents' beliefs about their Infants' Minds and Learning (the BIMLq). BIMLq will be based on past measures1-4, but expand on them in order to build a psychometrically sound questionnaire that can be used in the general population. It will combine both open-ended and closed questions in order to obtain a complete picture of parents' intuitive theories about the ontogeny of thoughts. With the data collected with BIMLq, we will then explore how parents' intuitive theories relate to the genealogy of innatism and empiricism in philosophy, and to the way these theories changed over time with the spread of scientific knowledge, progressively incorporating elements from both traditions to reach more pluralistic views in contemporary cognitive sciences and philosophy.

In the second axis of our project, we will conduct a large-scale study with an online distribution of the BIMLq in several countries to ask: (i) are there systematic biases towards innatism or empiricism in the general population, and do they vary with parenthood? (ii) how diverse are these beliefs? (iii) do they systematically vary, for instance as function of parents' levels of education, gender, country or residence, or whether there are some neurodevelopmental disorders in their family? (iv) where do these parental BIMLs come from? (v) how do they change as infants' age?

Finally, in a third axis of research, we will examine whether parents' BIMLs impact how much, how contingently, and how specifically they speak with their child while at home, and whether this in turn impacts infants' vocal and linguistic development. To do so, we will conduct a longitudinal study in which we will collect home recordings, as well as the BIMLq and measures of infants' vocal (babbling) and lexical (vocabulary size) development between 4- and 18-months. Axis 1. Conception and validation of the BIML. We will design a mixed-methods questionnaire (combining both open-ended and closed questions) which can be completed in an interview format or autonomously (e.g., online). The BIMLq is based on previous measures of mind-mindedness, parental reflective functioning, caregiving practices and experimental philosophy, as well as novel items that we designed ourselves. We identified a typology of BIMLs that are widely present in a varied corpus of texts (scientific, philosophical, parenting websites...) to design these novel items. We are currently testing some of them in a pilot study (alongside established measures of mind mindedness for a comparison), which is part of a wider study where we collect home recordings of parents and infants who are 4- and 7-months-old. The sample size for this preregistered study is 50, and its aim is to assess infants' vocal development and self-recognition, and their link to caregivers' input (social contingency, quantity of input, etc.). We will use this pilot data to further inform our design of the BIML, by drawing on an automatic transcription of parents' speech while talking about their child, and their response on a beta version of the BIML. BIML is currently designed with 4 sub-items, from which we can extract 4 scores: (1) the cognitive agency score measures parents' tendency to attribute various types of cognitive states (e.g., memories...) to their own infant; (2) the communicative agency score measures parental interpretations of their infant's vocalizations as communicative acts; (3) the emotional versus cognitive contents ratio reflects the relative focus parents make on their infant's emotional experiences versus cognitive activities; (4) the innatism-empiricism score measures parents tendency to conceive their infants' development in terms of maturation versus learning.

Once we have a satisfactory version of the BIMLq, we will validate it by running an online study on Prolific with a sample size that will be determined with a power analysis using the pilot data.

To relate parents' intuitive theories to philosophical theories of innateness and empiricism, we will draw on an in-depth review of the philosophical literature on concept innatism and empiricism, and the experimental philosophy literature, and draw parallels between these theories and our BIMLq data. We will write a theoretical paper to summarize these theoretical contributions. This work will also be informed by our second axis of research.

Axis 2. Online study to describe parental BIML.
We will conduct an online study to collect descriptive data using the BIMLq. We will collect data from parents when the mother is in her third trimester of pregnancy, and then when the infant is 4, 7, 12, 18, 24, and 36-months-old (these ages are chosen because they correspond to specific developmental milestones in typically developing infants). We will collect sociodemographic information (level of education, origin, gender) to examine their potential links with BIMLs. We will also examine whether developmental milestones (e.g., the first steps, words...) and/or chronological age impact parental BIMLs, and whether BIML systematically vary across countries . We will start with a comparison between France, the United Kingdom, and Spain , but hope to be able to expand to other countries and languages later on. The BIMLq takes less than 30 minutes to fill in, and parents will be remunerated 5€ for their participation.

Axis 3. Longitudinal study with home recordings. We will follow families longitudinally when their infants are 4, 7, 12, and 18-months-old. At the age of 4, 7 and 12-months of age, we will do a home visit. During each visit parents will complete the BIMLq, and record themselves during a typical weekend with their child. From these recordings, infant vocalizations and parental speech will be automatically extracted and we will compute: (i) the proportion of speech directed to the infant; (ii) social contingency, and (iii) the extent to which parents change their register when addressing their infant. When infants are 18-month-old, parents will fill in a standardized questionnaire to evaluate their language acquisition.