European Commission logo
italiano italiano
CORDIS - Risultati della ricerca dell’UE
CORDIS

How do infants mentalize? Bringing a neuroimaging approach to the puzzle of early mindreading.

Periodic Reporting for period 4 - DEVOMIND (How do infants mentalize? Bringing a neuroimaging approach to the puzzle of early mindreading.)

Periodo di rendicontazione: 2022-08-01 al 2023-07-31

The project aimed to test a novel theory about the uniqueness of infant social intelligence. The project was grounded in the assumption that early human cognition evolved to meet the unique challenges of the infancy period, which include the need to acquire a vast amount of information through observation. The project tested a novel theory of early perspective taking which claims that a) human infants have a bias to track what other people are attending to and b) they can do this without interference from their own perspective because self-representation emerges slowly in the first years of life.

There is a widely held belief that infants are egocentric, and the theory tested in the project proposes the opposite, specifically that in the first year of life, infants are altercentric. Infants rely on others to gather information about what they should attend to and learn about. Characterizing early social cognition is important for understanding the origins of our uniquely human social cognitive abilities but also for understanding developmental disorders which implicate developing social cognition (e.g. autism).

The overall objectives of the project were to test multiple predictions generated by the theory (Southgate, 2020, Psychological Review) using different methods to garner sufficient evidence during the 5-year-period to evaluate the value of the theory to the field. At the conclusion of the project, we believe the data that we have accumulated provides solid evidence that the theory well-describes early social cognition and its development.
Several months at the beginning of the project were spent building our lab (the Centre for Early Childhood Cognition) at the University of Copenhagen https://psychology.ku.dk/ecc-en/). We built the lab from scratch, including purchasing important equipment (EEG, fNIRS, eye-tracking) and set up and installation. We also spent significant time recruiting participants from the Danish national registry, and now have a database of approximately 10,000 infants. We have tested over 2000 infants as part of this project, both in our lab and in a public science museum in Copenhagen. We have also spent time ensuring that our studies comply with Open Science principles and all of our studies have been preregistered at the Open Science Framework (OSF). In addition, the PI has worked a lot on developing the theory that is being tested, and expanding the predictions that can be made.

The results from the project support the theory that was proposed. We found considerable evidence that, in the first year of life, infants have an altercentric bias meaning that they better remember the targets of others' attention than events that they have witnessed alone. By 12 months, this bias is receding, and we think this is a transition period during which some infants still show this bias (in this task) and others do not. In another study, we showed that whether or not 18-month-old infants evidence self-representation is associated with self vs. other memory bias. Those infants who show evidence for self-representation (by passing the mirror self-recognition task) show a bias towards better memory for objects that were assigned to them whereas infants who do not show evidence for self-representation (who do not pass the mirror self-recognition task) show a bias towards better memory for objects that were assigned to others. In adults, a 'self-reference' effect (as we found in the 18-month-olds who pass the mirror self-recognition task) is well-documented but this is the first time that the opposite effect (an other-reference effect) has been documented and shows a profound difference in the cognition of infants vs. adults. We have disseminated these data widely at conferences and meetings, as well as in publications. Many data sets are still waiting to be written up because the project has been so labour intensive that the majority of the post-docs time has been spent on data collection and analysis. In addition, we have built our own pupillometer eye tracker that is more suited to collecting pupillometry data from infants than commercially available solutions and we believe that this has much potential for exploitation in the future.
It has long been held that young children are egocentric, a position usually attributed to Piaget. The theory that we sought to test with the ERC project turned this long-held assumption on its head, proposing instead that human infants are altercentric and preferentially attend to the behaviour of other people. We further hypothesized that this is possible because young infants do not yet have a sense of self, which would otherwise compete for their attention. Without a developed self-representation, infants' attention is externally-oriented which we think enables them to learn better all the things that a young baby has to learn. At the end of the project, we have found considerable evidence in support of this idea. Three major findings are: 1) at 8 months, infants remember better the location of an object who's hiding was witnessed by another agent than the location of an object that the infant has witnessed alone (we call this an altercentric bias in memory); 2) this memory bias is receding by 12 months, but has not completely disappeared; but at 18 months, those infants with a less developed sense of self show a memory bias for objects associated with others whereas those infants who have a more developed sense of self show a memory bias for objects associated with themselves; and 3) the emergence of self-representation at 18 months brings with it an experience of perspective conflict: now infants have 2 perspectives (self and other) and these compete and cause conflict when they diverge. At the end of the project, we have accumulated over 10 large data sets, each of which contribute to testing the theory proposed in the ERC project.
Infant participant in project