2016 DIVIDNORM Workshop

Metacognitive diversity across cultures: Advances and perspectives

Dividnorm 2011-2016

Paris, ENS, May 31-June 1, 2016

Organization: Joëlle Proust (Institut Jean Nicod)

Contact/information: Cet adresse mail est protégé contre les spambots. Vous avez d'activer le javascript pour la visualiser.

Location: Ecole normale supérieure, 45 rue d'Ulm, 75005 Paris

May 31, 2016: Amphithéâtre Rataud

June 1, 2016: Salle des Actes


Metacognition - the capacity to evaluate one's uncertainty in a cognitive task - has been classically studied in industrialized Western societies. Little is known, however, about epistemic sensitivity in non-Western, non-industrial societies. This workshop is meant both to present interdisciplinary research conducted in this domain by the DIVIDNORM project, and to explore with anthorpologist, philosophers, psychologists and neuroscientists the changes in method and experimental paradigm that future research about cognitive and metacognitive diversity calls for.

Tuesday, May 31

8:45 - Welcome coffee

9:00-9:30 - Presentation of the workshop (Joëlle Proust, IJN)

9:30-10:30 - Asher Koriat (Institute of Information Processing and Decision Making, University of Haifa) : "Confidence, consistency, and social consensus".

10:30-10:45 - Coffee break

10:45-11:30 - Stéphane Bernard & Fabrice Clément (UNINE, Neuchatel) : "The impact of consensus on belief formation: Some developmental and cross-cultural elements for a global reflexion".

11:30-11:55 - Comments by Daniel Haun (University of Leipzig)

11:55-12:15 - General discussion

12:15-14:15 - Lunch break

14:15-14:55 - Shoji Itakura (University of Kyoto) : "Nonverbal theory of mind: Evidence in Japanese children".

14:55-15:20 - Comments by Athanasios Chasiotis (Tilburg University)

15:20-15:40 - General discussion

15:40-16:15 - Coffee break 

16:15-17:15 - Sunae Kim, Markus Paulus & Beate Sodian (Ludwig Maximilian University of Munich) : "Metacognition and theory of mind among German and Japanese children".

17:15-17:25 - Comments by Asli Ozyurek (Max Planck Institute for Psycholinguistics, Nijmegen)

17:25-17:45 - General discussion

19:15 - Conference dinner (Bouillon Racine, 3 rue Racine, 75006 Paris)

Wednesday, June 1st

9:00-10:00 - Chris Frith (University College London): TBA

10:00-10:15 - Coffee break

10:15-11:15 - Laurence Conty, Terry Eskenazi & Amelie Jacquot: "Social influence on metacognition in France and in Japan".

11:15-11:40 - Comments by Dan Bang (Metacognition lab, Wellcome Trust Centre or Neuroimaging, University College London)

11:40-12:00 - General discussion

12:00-14:00 - Lunch break

14:00-14:30 - Olivier LeGuen (CIESAS, Mexico) : "Metacognition in Yucatec Mayan".

14:30-15:00 - Martin Fortier (IJN, ENS, Paris) : "Supernatural thinking as probabilistic reasoning: Complexity, fluency and informativeness".

15:00-15:25 - Comments by Pierre Deléage (Labratoire d'Anthropologie Sociale, EHESS, Paris)

15:25-15:45 - General discussion

15:45-16:15 - Coffee break

16:15-17:00 - Nicholas Shea (Kings College London) : "Metacognition about concepts".

17:00-17:45 - Joëlle Proust (IJN, ENS, Paris) : "How cross-culturally diverse is metacognition? A dual-system viewpoint".




Metacognitive diversity: An interdisciplinary approach

24, 25, 26 September 2014

Ecole normale Supérieure, 45 rue d'Ulm, 75005 Paris. Salle Dussane.

Organization : Joëlle Proust (ENS) & Cet adresse mail est protégé contre les spambots. Vous avez d'activer le javascript pour la visualiser. (IJN)


While cognitive diversity has become an important subject of research, metacognitive diversity is still  poorly  understood.  Metacognition  refers to the set of conscious or unconscious  processes  with  which agents contextually control their first-order cognitive activity (such as perceiving, remembering, learning, or problem solving) by assessing its feasibility or likelihood of success. Metacognitive skills are involved in many daily activities, such  as  conversing, reading, planning,  and forming collaborations. Some of these skills  are evenly distributed across cultures, e.g.,  distinguishing familiar from new environments, informative from repetitive messages,  difficult from easy cognitive tasks. Others, however, seem to be socially constructed, or differentially shaped by social norms, linguistic and conversational usage, educational methods, religious practices, and self-related attitudes. Sensitivity to  truth, epistemic authority,  social consensus,  evidentiality,  uncertainty, and thought coherence seems to vary significantly  across languages and cultures. Social norms, models of mind,  and stereotypical attitudes about the self might also deeply influence how epistemic self-evaluation is conducted, and how it affects behaviour.

The aim of this conference is to study  these and similar questions with  interdisciplinary methods: cross-cultural evidence from linguistics,  developmental, experimental,  and social psychology,  and  anthropology, will be discussed with the goal of  examining metacognitive diversity and its cultural, social, institutional, or psychological sources.


Rakefet Ackerman (Israel Institute of Technology), Stephanie Carlson (University of Minnesota), Athanasios Chasiotis (Tilburg University), Eve Danziger (University of Virginia), Paul Harris (Harvard University), Shinobu Kitayama (Univ. of Michigan), Ulrich Kühnen (Jacobs University), Olivier Le Guen (CIESAS, México), Cristine Legare (University of Texas), Tanya Luhrmann (Stanford University), Michael Morris (Columbia Univ.), Daphna Oyserman (USC), Anna Papafragou (University of Delaware), Rolf Reber (Univ. of Oslo), Norbert Schwarz (USC), Bradd Shore (Emory Univ.), Richard Sorrentino (University of Western Ontario), Steven Stich (Rutgers University).

Program (version PDF) - Abstracts (Version PDF) - Videos (vimeo)


Workshop Dividnorm, October 2014, Paris

Tuesday September 30th

ENS, 45 rue d'Ulm, Salon du restaurant 



Welcome and introduction. Laurence Conty (ENS, Paris).



Joëlle Proust, Institut Jean-Nicod, Fondation Pierre-Gilles de Gennes pour la Recherche, Ecole Normale Supérieure, Paris

Conformity and consensus: similarity and differences

Consensus is an epistemic norm: it sometimes is, or seems adequate to allow one's own beliefs and epistemic evaluations to be influenced by those of the majority, whether or not they look accurate or plausible to oneself. Deference is a particular case of consensus, where individuals choose to adopt the beliefs and epistemic decisions of experts, rather than to engage themselves in the evaluation of the contents involved. Conformity is a social norm, requiring to suppress one's own behavioral tendency in order to match the behavior of the majority. Experimental evidence from development, comparative and adult studies for sensitivity to consensus and to conformity will be discussed, with their respective trade-offs, motivations, associated physiological measures and distal function.


Stéphane Bernard, Cognitive Science Centre, University of Neuchâtel, Switzerland

Relations between epistemic norms: Consensus, accuracy and fluency

This presentation will focus on the relations between the epistemic norms of consensus, accuracy and fluency. We will present experimental results about 1) visual fluency and visual discrimination control 2) a conflict between consensus (expressed by four bystanders) and accuracy. 1) In the first experiment, the results clearly showed that 3.5-year-old children accurately controlled their own capacity to discriminate visual stimuli and that this control was not linked to metarepresentational abilities (measured with three false belief tasks and with the Dimensional Change Card Sort (DCCS) task). 2) In a second experiment, we tested whether 3- and 4-year-old children prefer to follow the testimony of an inaccurate informant approved by four bystanders (consensus) rather than the testimony of an accurate informant disapproved by the four bystanders. Results showed that 3- and 4-year-olds trusted more an inaccurate informant approved by the four bystanders than an accurate informant disapproved by the four bystanders.


Sid Kouider & Louise Goupil, Laboratoire de sciences cognitives et psycholinguistique (LSCP) - CNRS/EHESS/ENS Paris

Precursors of metacognitive abilities in preverbal infants

Past research documented a rather late development of metacognitive capacities. Yet, these studies were based on children’s report about their own mental states. Here, we investigated the possibility that even preverbal infants can demonstrate rudimentary forms of metacognition in a simple and non-verbal setting. We tested infants capacity to monitor decision confidence, a hallmark of metacognitive sensitivity. In two different studies, we tested whether, after performing a binary choice, infants show differential overt behavior (e.g. persistence in their choice) following correct vs. incorrect decisions, even when nothing in the external world provides information about the correctness of their responses. A first study showed that 18 month-old infants appropriately use confidence in their own decision in a manual search paradigm. In a second study, we show that 12 month-old infants can evaluate the correctness of their anticipatory eye movements. We propose that although explicit metacognition develops much later, it stems from implicit metacognitive abilities already present in infants.


Sunae Kim, Ludwig-Maximilians-Universität, Munich

Children’s assessment of their own knowledge states

Metacognitive abilities may be a late development. In particular, in a recent study by Rohwer, Kloo, and Perner (2012) not until 5 or 6 years of age did children accurately assess their own knowledge states. In this talk, I’m going to present a new study where 4-year-old German children were tested on an adaptation of Rohwer et al. task. We found that when 4-year-old German children were asked to inform another ignorant person, they seemed to accurately report their own lack of knowledge.   I will also present some cross cultural data concerning how Japanese children may or may not perform differently in their own assessment of their own knowledge.


Nike Tsalas, Ludwig-Maximilians-Universität, Munich

The neural correlates of metacognitive monitoring in children and adults

In the present talk, we present two studies that investigated the development and neural correlates of metacognitive monitoring in 8-year old children, and healthy students. We will present data from two EEG studies that investigated the neurocognitive underpinnings of Judgments of Learning in contrast to normal memory judgments. Participants had to learn picture-pairs and had to either estimate their likelihood of whether they could remember the target later on, or whether the colour yellow had been present in one of the two pictures.


Wednesday October 1st

ENS, 45 rue d'Ulm, Salon du restaurant 


Vasily Klucharev, National Research University – Higher School of Economics, Moscow, University of Basel

Neurobiological mechanisms of normative social influence.

Humans often change their beliefs or behavior due to normative behavior of others. We explored, with the use of various neuroimaging methods (fMRI, TMS, ERPs), whether conformity to social norms is an automatic process that is based on a general error detection and error correction mechanism (Klucharev et al., 2009; Klucharev et al., 2011; Shestakova et al., 2013). We hypothesized that conflicts with normative group opinion modulates activity of the brain regions often associated with automatic adjustment of behavior. Using fMRI we showed that conflicts with group opinion modulated neuronal activity in the posterior medial frontal cortex and the ventral striatum. Our findings were supported by a number of other fMRI studies that have demonstrated an activation of the same regions, when participants were exposed to the opinion of a reference group conflicting with their own opinions. We also demonstrated that the transient downregulation of the posterior medial frontal cortex reduced conformity to social descriptive norms. Overall, recent neuroimaging results support the hypothesis that some forms of normative influence are mediated by activity of the automatic mechanism correcting deviant behavior. Furthermore, our results suggest that social conformity is underlined by the neural (dopamine-related) activity which signals probably the most fundamental social mistake that of being “too different” from others.


Jérôme Sackur and Gabriel Reyes, Laboratoire de Sciences Cognitive et Psycholinguistique (ENS / CNRS / RHRSS)

Stress impairs introspection

Metacognition is known to differ across individuals, and this has been linked to anatomical differences in the prefrontal cortex. Here, we looked for psychological predictors of interindividual variability in introspection. Based on previous research showing that stress impacts working memory, and leads to less flexible learning, we predicted that higher stress responses would be associated with poorer metacognitive sensitivity. In two experiments, we thus engaged participants in elementary cognitive tasks, and collected metacognitive judgments on a trial-by-trial basis. Peak cortisol levels in response to the Trier Social Stress Test were measured beforehand, so as to assess participants' stress sensitivity. We show that stress sensitivity had no impact on first order cognitive performance, but that it consistently reduced metacognitive sensitivity. These results are consistent with the view that stress depletes higher order executive resources that are implicated in the control and monitoring of cognitive functions.


Terry Eskenazi, LNC INSERM U960, Ecole Normale Supérieure, Paris, France

Social influences on metacognitive evaluations I

Social influences on decisions and behaviour have long been studied, however little is known whether metacognitive processes are also susceptible to social influence. In a set of experiments we addressed this question using non-verbal social cues that convey implicit or explicit feedback. Participants made a perceptual decision before receiving the social cue, and then provided a metacognitive evaluation by means a post-decision confidence rating. Behavioural experiments revealed an impact of social feedback on decision confidence, even when the feedback was implicit and entirely unreliable, at the expense of metacognitive accuracy. We also found that this social influence on decision confidence can scale with the number of social sources providing feedback. To explore the neural underpinnings of the link between social feedback and metacognitive evaluations we conducted an fMRI study and specifically asked if valuations of social feedback, and the associated neural response, change as a function of one’s initial level of confidence. Preliminary analyses indicate a parametric relationship between decision confidence and response to feedback, and implicate posterior medial frontal cortex (pMFC), a region that has been linked to reward processing and conformity behaviour. Altogether our results indicate a tight link between metacognition and processing of social information, which could, in turn, imply a critical role for metacognition in social learning.


Benoit Montalan, Laboratoire de Psychologie et Neurosciences EA 4699, Normandie Université, France

Social influences on metacognitive evaluations II: An ERP study.

In an EEG study we investigated the neurophysiological processes underlying the social susceptibility of metacognitive evaluations. We recorded ERPs while the participants performed a 2AFC task, which was followed by the presentation of a face randomly gazing directly at the participant or towards either of the response alternatives. Participants were misled in believing that the observed gaze direction reflected a previous participant’s response. The results confirmed our previous findings: participants aligned their confidence judgements as a function of the information provided by the social cue. Moreover, we found an amplified early posterior negativity (EPN) in response to averted gaze (i.e. positive or negative feedback). EPN is a component related to the motivational relevance of a stimulus, which may reflect the “epistemic meaning” attributed to the social cue while making metacognitive evaluations. However, further investigations are needed to determine precisely the functional role of this component in metacognitive monitoring.


Amelie Jacquot, Laboratory of Psychopathologie and Neuropsychologie, Université Paris 8, France

Social influences on metacognitive evaluations III: Cross-cultural differences

In this study, we examined potential cross-cultural differences in the extent to which social information influences metacognitive evaluations. Given that the Eastern cultures are considered as more collectivistic than the Western cultures we conducted a study in Japan and in France to determine (1) whether Japanese participants’ metacognitive evaluations are more sensitive to social information than those of the French participants, (2) whether social sensitivity is stronger when the information is shared by several individuals as compared to only one, and finally (3) whether group identity modulates social sensitivity in metacognitive evaluations.

Thursday October 2nd

ENS, 45 rue d'Ulm, Salle Info 2 


Stephen M. Fleming, Center for Neural Science, New York University, Department of Experimental Psychology, University of Oxford

Toward a neuroscience of metacognition

Metacognition concerns the evaluation of one’s own cognitive processes and underscores the ability to “know that we know”. For example, a student who has high confidence that they have learnt enough for an exam may put away the books and stop studying. This “cognitive” aspect of self-awareness is often thought to be a hallmark of the human mind, and, as this example illustrates, is crucial for high-level control of behaviour. Despite the centrality of metacognition in human mental life, the neural and computational mechanisms that underpin metacognition, as opposed to primary cognitive functions such as perception and memory, remain poorly understood. A key obstacle for a true science of metacognition to overcome is the development of quantitative measures of metacognitive ability. Recent computational approaches, grounded in signal detection theory, are able to circumvent these potential confounds and provide a robust, objective measure of metacognitive “efficiency” in the lab by analyzing the correspondence between subjects’ confidence ratings and objective behavioural performance. Studies applying such measures in tandem with neuropsychology, functional brain imaging and brain stimulation techniques have begun to reveal how metacognition may operate in the human brain. For instance, by applying precisely targeted brain stimulation before and after a perceptual judgment we are able to selectively interfere with subjects’ confidence in their decisions, thereby altering metacognitive accuracy in a predictable fashion. Drawing together different sources of evidence, I will outline a model in which parallel neural architectures support metacognition across different domains, such as perception and memory. I will close by discussing the implications of this work for attempts to alter or improve metacognition.


Karim NDIAYE, Institut du Cerveau et de la Moelle épinière, Hôpital de la Pitié-Salpêtrière, Paris

Varieties of metacognition: Behavioral studies with healthy humans and OCD patients.

Metacognition is the psychological processes that enable to monitor and control one’s own cognitive functioning, e.g. to assess our confidence in a perceptual decision and how one may react to it by seeking additional information. I will describe a series of experimental studies in humans addressing the nature of metacognitive processes. Firstly, by comparing performance across multiple tasks, we showed that metacognitive monitoring (indexed by the ROC curve derived from post-decisional confidence ratings) was domain-dependent. And, in a second study, we observed dissociation between metacognitive monitoring and metacognitive control (implemented as checking rate). In obsessive-compulsive disorder (a psychiatric disorder notably marked by pathological doubt and compulsive rituals, such as checking) various cognitive models based on clinical and self-report observations have suggested that metacognitive dysfunctions might be central to OCD. However, our ongoing studies in OCD patients provide empirical evidence challenging the hypothesis of a mere introspective monitoring deficit in these patients.


Shoji Itakura1 & Yuko Okumura2, 1Kyoto University, Japan 2NTT Communication Science Laboratories, Japan

Social learning in infants: From a human or a robot?

Social learning enables infants to acquire information, especially through communication. However, it is unknown whether humans are the prime source of information for infant learning. Here we report that humans have a powerful influence on infants’ object learning compared with nonhuman agents (robots). Twelve-month-old infants were shown videos in which a human or a robot gazed at an object. The results demonstrated that the infants followed the gaze direction of both agents, but only human gaze facilitated their object learning: Infants showed enhanced processing of, and preferences for, the target object gazed at by the human but not by the robot. Importantly, an extended fixation on a target object without the orientation of human gaze did not produce these effects. Together, these findings show the importance of humanness in the gazer, suggesting that infants may be predisposed to treat humans as privileged sources of information for learning.

Infants can acquire much information by following the gaze direction of others. This type of social learning is underpinned by the ability to understand the relationship between gaze direction and a referent object (i.e., the referential nature of gaze). However, it is unknown whether human gaze is a privileged cue for information that infants use. Comparing human gaze with nonhuman (robot) gaze, we investigated whether infants’ understanding of the referential nature of looking is restricted to human gaze. In the current study, we developed a novel task that measured by eye-tracking infants’ anticipation of an object from observing an agent’s gaze shift. Results revealed that although 10- and 12- month-olds followed the gaze direction of both a human and a robot, only 12-month-olds predicted the appearance of objects from referential gaze information when the agent was the human. Such a prediction for objects reflects an understanding of referential gaze. Our study demonstrates that by 12 months of age, infants hold referential expectations specifically from the gaze shift of humans. These specific expectations from human gaze may enable infants to acquire various information that others convey in social learning and social interaction.



Poster #1-Mid term and long term consequences of cocaine addiction on reversal learning

Pauline Smith, Luc Mallet, Florence Vorspan, Karim N’Diaye

Institut du Cerveau et de la Moelle épinière, Hôpital de la Pitié-Salpêtrière, Paris

Several studies have demonstrated that cocaine users have less cognitive flexibility than controls in probabilistic reversal learning tasks, in which subjects have to learn to choose between a generally rewarded stimulus and generally non-rewarded one, while adjusting to sudden contingency reversals. Furthermore, it has been shown in monkeys that cocaine administration disrupts reversal learning. We thus wondered if the impaired flexibility exhibited by cocaine users could be a reversible consequence of their cocaine consumption. To answer this question, we compared the performances of cocaine users, ex cocaine users (abstinent for 2 months or more), and controls in a probabilistic reversal task.

Poster #2-Relationship between a false belief understanding and an assessment of one's own knowledge in both German and Japanese children. 

Sunae Kim1, Markus Paulus1, Beate Sodian1 , Shoji Itakura2, Mika Ueno2 and Joelle Proust3

1Ludwig-Maximilians-Universität, Munich; 2Kyoto University; 3ENS, Paris 

False belief tasks tap on a development of meta-cognitive abilities.  However, meta-cognitive abilities come in various forms and facets and, therefore it is important to understand whether and how different tasks measuring meta-cognitive abilities relate to one another.   In the present study,  we asked whether children's performance in false belief tasks predict their performance in another task in which children are asked to assess their own knowledge states.  We further sought to understand cultural differences in meta-cognitive abilities by comparing German and Japanese children. 

Poster #3-Metacognitively controlled spacing strategies in 7-year olds, 10- year olds and adults 

Nike Tsalas, Markus Paulus, Beate Sodian

Ludwig-Maximilians-Universität, Munich

The current study investigated metacognitively guided spacing choices in children and adults and how these are influenced by repeated study trials. We tested a group of 7-year-olds, 10-year-olds and adults in a task in which they had to monitor their learning through Judgments of Learning (JoL) and decide whether to space their study, mass it or terminate it. Contrary to previous findings, our study provides first evidence that already by 7 years of age children can make metacognitively controlled scheduling decisions. Furthermore, our study provides evidence that participants improved their relative monitoring accuracy in the second study trial and that they used the feedback gained after the first study trial to adjust their monitoring and control processes.   

Poster #4-The self-awareness effect of eye contact: Behavioral evidence and neural correlates

Matias Baltazar1, Julie Grèzes2 , Jean-Luc Picq1, Laurence Conty1

1 LPN, Paris 8; 2 ENS, Paris

We have recently shown that the perception of a face with a direct gaze (that establishes eye contact), as compared to either a face with averted gaze or a mere fixation cross, led adult participants to rate more accurately the intensity of their physiological reactions induced by emotional pictures (Baltazar et al., in press). We thus demonstrated that bodily self-awareness becomes more acute when one is subjected to another’s gaze. Here, using fMRI, we show that self-awareness correlates with a typical self network (i.e. anterior cingulate cortex and bilateral insula), but only if we contrast non eye contact conditions minus eye contact. The reverse contrast (eye contact minus other) reveals a correlation between ratings of intensity and superior colliculus activity, a subcortical area known to underpins integrated awareness of self and environment. In sum eye contact triggers a specific form of self-processing.

Poster #5-I know you can see me. Attributing mental states influences self-awareness

 Nesrine Hazem1, Nathalie George2 , Matias Baltazar1, Laurence Conty1

1LPN, Paris 8; 2 ICM, Paris

Eye contact is a typical human behaviour known to impact concurrent cognitive processes. In particular, it has been recently demonstrated that eye contact elicits bodily self-awareness in human adults. Here, we investigated if ‘the belief to be the target of another person’s perception’ is sufficient to trigger such an effect. We manipulated participants to believe that they were in video on-line connection with another individual. We demonstrated that the perception of this individual, when wearing non-obstructed sunglasses as compared to obstructed sunglasses, led participants to rate more accurately the intensity of their physiological reactions induced by emotional pictures. Our data suggest that the belief to be the target of another person’s perception elicits a self-awareness process, probably by enhancing self-focused attention. Taken together with other studies, they further suggest that the belief to be seen by another social agent is a basic form of mentalizing embedded in direct gaze perception.

Poster #6-Baboons know when they do not know, and what they do not know

Raphaëlle Malassis1,2, Gilles Gheusi1, Joël Fagot2

1Laboratoire d'Ethologie Expérimentale et Comparée, université Paris Nord ; 2Laboratoire de Psychologie Cognitive, université de Provence.

This study assessed metacognitive skills in the guinea baboon (Papio papio). Earlier studies have demonstrated that nonhuman primates can use a key to escape testing and receive a minimal amount of rewards when they are uncertain of the correct response to be given in the task, but the significance of this behavior remains debated. The current study tested three baboons on a task requiring to memorize and recall the position of two target stimuli on a screen (ie, two blue squares in a matrice of 12 squares). The baboons could either report the location of the target stimuli directly, or use a "restudy" key providing an additional opportunity to visually examine the display. Baboons selected the “restudy” key more often during the difficult trials than during the easy trials, and generalized their response to new test conditions in which the perceptual saliency of the targets or their duration was manipulated. The second experiment showed two re-study keys instead of one, and used target stimuli of unequal perceptual saliencies. Two out of the three baboons expressed a preference for the "restudy" key displaying the least salient target stimulus. This finding demonstrates that these baboons knew that they did not know, and also knew which information was lacking in their memory for correct responses. Taken together, these results show that nonhuman primates can exploit their own epistemic states to trigger information-seeking behaviors.