a partner as well in the course of communication. In the process of learning, the communicative partner can play two separate roles: could be the main
33 source of information, and also the target of observation and learning. We would like to highlight that from very early on, children are able to exploit both roles of the partner in order to enrich their knowledge on different fields of the environment in an integrated format.
7. thesis. In their early months, children primarily learn FROM others and encode the content learnt as predictive, generic information.
The benefit of social learning can be characterized by the description that young children learn about objects in the world through the observation of others’ object directed behavior. This form of learning allows the transmission of information filtered by a knowledgeable partner. Thus, social learning exploits the mental achievements of an expert partner and supports fast acquisition of complex information.
The most prominent example of this learning situation is social referencing.
When an infant first encounters a novel, unfamiliar, and even surprising object, she first checks the behavior and emotional expression of a close partner and tries to read her/his emotional, referential expression whether the target object is approachable or avoidable (Walden és Ogan, 1988). We have reinterpreted this phenomenon arguing that the social partner’s emotional expression not simply modulates the infant’s behavior through emotion regulation, but also conveys information about the object itself.
Supposedly, valence information is transmitted that is already known by the expert individual and unknown for the infant novice partner: in this situation, the young observer learns generic, predictive information about the object quickly from the partner. So, it can be supposed that in social referencing situations infants assume that the other’s object-directed emotion manifestations conveys universally shared information about the referent that is available to all individuals.
In a study with 14-month-olds we have provided evidence that infants learn and remember the object valence information, and not the personal, consistent preferences of partners observed. Based on this, we argue that infants rely on their so-called ‘object-centred’ interpretations to form generalized expectations that all others will perform the same kind of object-directed actions that are appropriate given the objective valence quality of the referent that the infant’s newly formed object representation contains.
Indeed, object-directed emotion expressions provide two types of information:
they can be understood as communicative signals that convey culturally shared knowledge about referents that can be generalized to other individuals (as we have
34 shown above as object-centered interpretation), or they can convey the expressers’
person-specific, subjective disposition toward objects. In a further study, by presenting object-directed emotion expressions in communicative versus non-communicative contexts, we demonstrated that 18-month-olds could flexibly assign either a person-centered interpretation or an object-person-centered interpretation to referential emotion displays. The findings indicate that infants are prepared to learn shared knowledge from nonverbal communicative demonstrations addressed to them at an early age. While they are also capable to learn about the partner: they attribute person specific dispositions and preferences.
8. thesis. The possibility that children are able to apply both the object centered and the person centered interpretative schemas when interpreting an other person’s behavior, invites the reconsideration of findings on mindreading or naïve psychology competencies in young infants: we propose the primacy of the object-centered approach. In addition, we postulate that the emergence of flexible use of interpretative models might contribute to improvement in performance.
There is an ongoing debate on the availability, and nature of early mindreading (Naïve Psychology) competences (Perner and Ruffman, 2005; Rakoczy, 2012). The so called implicit minreading tests rely on robust behavioral measures, like looking time or anticipatory looks applied in simple object choice context. In most of the cases, these simple object choice scenarios can be interpreted by both the object-centered and by the person centered interpretative frames as well, and these possibilities cannot be disentangled in the classic, existing approaches. However, in the framework of mindreading, there is a potential hierarchy between the above two interpretative schemas: the person specific interpretation could be described as an object-centered content bound to a specific person, distinctively.
Revisiting the findings in the domain of implicit mindreading, recent findings claimed to suggest that infants understand others’ preferential choice and can use the perspectives and beliefs of others to interpret their actions. The standard interpretation in the field is that infants understand preferential choice as a dispositional state of the agent, so learn about that specific agent. In our view, it is possible, however, that these social situations trigger the acquisition of more general, object centered knowledge. We propose that early mindreading processes lack the binding of belief content to the belief holder. However, such limitation may in fact serve an important function, allowing
35 infants to acquire information through the perspectives of others in the form of universal access to general information.
In our view, binding mental states to specific persons, as a consequence of the simultaneous availability of object centered and person centered interpretative schemas, is a developmental milestone with respect to the emergence of mindreading capacities, as it makes possible the individuation of mental states approximately at the age of 18 months .
The integrated, flexible use of interpretative models might contribute to a more and more elaborate and complex performance, even inviting different domains. As an example, we show that a further developmental achievement in mindreading is brought about by the involvement of emerging episodic memory competences.
A study of ours aimed to investigate the contribution of episodic memory to mindreading, by proposing two different processes of belief attribution: prospective (online) belief tracking and retrospective inference based belief attribution. The experiment explored whether 18- and 36-month-old children could flexibly use episodic memory to attribute or update a protagonist’s belief. After a displacement event that an experimenter witnessed wearing sunglasses, 36-month-olds correctly attributed to her a false belief, when finding out after that event that the sunglasses were opaque. They successfully ascribed a false belief based on this new information and behaved accordingly. In contrast, 18-month-olds behaved as if the sunglasses were transparent. This suggests that 18-month-olds cannot use their memories to (re)compute a belief retrospectively, although they performed well when they could track false beliefs prospectively. This dissociation reflects that 18-month-olds rely primarily on prospective (online) belief tracking, while 36-month-olds can also flexibly compute beliefs retrospectively, based on episodic memories, well before they pass explicit tasks.
9. thesis. The primary function of mindreading (naïve psychology) is to enable the observer to monitor in the real time, say spontaneously and prospectively, the knowledge state of the partner. This capability facilitates both learning about the partner in the here and know and also evaluating the possibility whether learning from the partner would contribute to a valuable shared representational space for the long term or not.
36 Both children and adults are able to compute and monitor the perspective of their interactional partner, fostering the rich net of social interactions. However, due to the mindreading system’s limited capacity perspective taking was argued to occur spontaneously in a restricted manner, only for all or nothing type, so level-1 perspective taking, but not for the more sophisticated 2 perspectives. We proposed that level-2 perspectives (containing aspectual information) could also be computed spontaneously if participants have reason to assume that the partner is indeed aware of the objects’ aspectual properties. In a reaction-time study where adults, and school aged-children answered a simple number verification task, the partner’s inconsistent perspective was found to interfere with reaction times providing evidence for spontaneous level-2 perspective taking.
Overall, we propose that this capability is a necessary prerequisite for the establishment of common ground, both for the ongoing interaction, and also for the enrichment of a long term, knowledge base shared with a broader group.