Institutional Agency:

• Istituto di Scienze e Tecnologie della Cognizione - CNR, Roma (IT) [website of the Institute]

Project Leader:

• Prof. Cristiano Castelfranchi [homepage]

Team members:

• Dr. F. Giardini
• Prof. R. Manzotti
• Dr. M. Miceli
• Dr. F. Paglieri
• Dr. A. Rissotto
• Dr. L. Tummolini
• Dr. G. Consoli

National Funding Agency:

• Consiglio Nazionale delle Ricerche [website of the CNR]

• from social mind-reading to self-consciousness;
• from phenomenal consciousness to mental simulation and empathy.

From social mind-reading to self-consciousness as meta-cognition

Regarding this first pattern of interaction, our IP will investigate what evidence can be raised in favour of the co-development of consciousness and social cognition. More precisely, we intend to gather evidence, both theoretical and empirical, to support the thesis of the social genesis of self-consciousness as meta-cognition, according to which cognitive skills that were first applied and honed with the aim of "reading the mind" of other agents, at a later stage come to be self-applied to one's own mind, giving raise to self-conscious experience (a trajectory first postulated by [24]). We will address this point also profiting of the debate with IP-02, which claims that mind-reading for self and others arrive together, thus disagreeing with our position. In this perspective, it is first necessary to focus better the functions of mind-reading, taking as a starting point the debate opposing Simulation Theory and Theory Theory. Mindreading is usually framed within the subject's Theory of Mind (ToM), albeit there is not yet full agreement on the exact nature of such set of skills, and two competing theories have been the object of much debate in the last twenty years. Theory Theory (TT), according to which the subject attributes internal states to other agents by applying to them a set of general rules and/or theoretical predictions ([2]; [4]; [5]; [9]; [13]; [20]) is opposed to Simulation Theory (ST), which claims the subject simulates the mental states of other agents, by putting himself in their shoes and running an off-line simulation (i.e. without actually acting) of his behaviour in the presence of the same inputs ([8]; [14]; on the controversial import of mirror neurons on this debate, see [11]; [15]). Our IP aims to show that (i) this distinction between different formats of ToM is relevant to the understanding of the social dimension of consciousness, while at the same time (ii) the distinction between phenomenal consciousness and self-consciousness is crucial to shed new light on the cognitive and evolutionary connections between TT and ST skills. More precisely, our IP will show that: TT and ST are complementary capacities, rather than competing ones, serving different socio-cognitive functions and undergoing different developmental processes; both TT and ST have crucial (and different) roles in the social circuit of consciousness; they are both connected with emotions, feelings, and the parallel development of phenomenal consciousness.

Moreover, we claim that there are at least three main functions of socially applied theory of mind, of which only the first two are usually stressed in the literature: explanation, prevision and influence of others' behaviour. The last point is by no means secondary: an intelligent agent reads the mind (and the behaviour) of other agents in finalistic terms, not only to causally explain their actions and anticipate their future conduct, but also (or even mainly) to achieve a key for influencing their behaviour [6], and both functions are crucial to achieve better social coordination. Other available forms of explanation and prevision, such as frequencies correlation, input-output models, normative theories, etc., either do not give any insights on how the subject could influence the other agent's behaviour, or they do provide only a limited access to this crucial information. In other words, theory of mind provides us with an influence power that other prevision tools cannot offer.

This view has crucial consequences on our understanding of consciousness: as soon as we apply our ToM skills to our own mental processes, we become not only better at explaining and anticipating our behaviour - we also acquire a new capacity for self-influencing our own conduct. This corresponds to a well-known and vital function of consciousness, that of allowing for greater degrees of flexibility in deliberation, planning, and control of one's own actions. This self-influencing ability leads consciousness, considered as self-applied theory of mind, to constitute a remedy against short-sightedness, rashness, and impulsiveness. On the other side, the fact that this representation ability is a distinctive human feature makes possible not only to influence the others, but also to be influenced by them. Such openness to influence is not less important, since it is essential in granting that our behaviours, emotions and intentions will be read and presumably understood by other people. In this sense, consciousness as tightly related to ToM works as an interface between individual and social aspects of agents' actions, bridging internal and external representations and boosting people's sociality. Our IP aims to achieve a precise and comprehensive map of this complex network of interactions between self-consciousness and social cognition, exploring in details the pivotal role played by mindreading capacities in these processes.

From phenomenal consciousness to mental simulation and empathy

Deeply related to the above issue is the analysis of different patterns of, respectively, phenomenal consciousness and self-consciousness. In our view, the latter undergoes a social genesis, along the lines described above. In contrast, phenomenal consciousness is by definition born from the inner, private experience of the subject, and yet it allows almost immediately for "social uses of subjectivity", as posited by the Simulation Theory. We will examine how to bridge the gap between phenomenal consciousness as an individualistic account, and social devices of consciousness, like coordination or empathy. In doing so, we will argue that the interaction of phenomenal consciousness with mental simulation (jointly with IP-02) provides new insights also for the debate on the format of mind-reading, i.e. Theory Theory vs. Simulation Theory. In fact, the capacity of "putting oneself in the shoes of someone else" for the purposes of mental simulation is useful not only for achieving a reconstruction of the motivating reasons of the other's behaviour, but also (and even mainly) for experiencing oneself in that particular situation. The qualia of such experience need not to be the same, but it may well be that I can use my phenomenal experience of the same situation as a blueprint for guessing relevant aspects of your subjective view on the matter (e.g., reconstructing your feelings in that situation). For the same reason, phenomenal consciousness is also related to the development of empathy, which is essential for the recognition and understanding of emotions, beliefs and desires of other people, as well as for the emotional linkage with past and future emotional experiences of the Self it provides. More specifically, one of the direction of research that we intend to explore concerns the role of empathy in the construction of self-image (together with IP-04). We will argue that being capable of feeling empathy with one's own self in the future provides a crucial reinforcement to several emotions closely connected with the subject's self-evaluation (such as shame, guilt, fear, etc.), and this self-applied empathy affects self-image, with several repercussions on the agent's behaviour. In collaboration with IP-04, we will analyse the empathic component in some of these emotions, building on previous works done by the ISTC-CNR team on the cognitive anatomy of emotions ([17]; [18]; [19]).

We will also investigate whether self-image, considered as the way an individual perceives herself, relates with phenomenal consciousness, and whether self-applied theory of mind can affect individual well-being and foster social relationships. In this respect, we will highlight what are the cognitive features of self-image that makes it such a powerful tool for social manipulation, i.e. for influencing the agent's actions by modifying her/his self-image. Moreover, we will investigate the validity of considering consciousness as a coherent self-image across different times (often referred to as the narrative self, as opposed to the minimal self of phenomenal consciousness; see [7]; [23]; [10]).

On the other side, we will claim that the self, as well as self-consciousness, emerges as an internalization of primarily social cognitive competences, i.e. cognitive skills first developed to cope with social interaction, and in particular group action, and only later to be applied to one's own mental processes. Under this hypothesis it can be advanced that the capacity of individual agents to think themselves as individual selves is a by-product of their capacity to act as members of groups [3]. To act as a group member and to pursue group goals, an agent needs the capacity to frame one's own decisions in a group perspective which means to reason from the perspective of the group (do team reasoning see [1]; [25]). In fact, Bacharach [1] has recently proposed to understand self-identity, and more generally the notion of "personhood", as emerging from the application of team reasoning to the same agent at different point in time. In other words, individual identity (the narrative self) is seen as the entification of the individual agent becoming a unifying self, the self would be the intertemporal coordination of an agent (for a different but related approach to self-identity, see [22]). This view has relevant consequences on the notions of agency and personhood, and on the distinction between them; it is also coherent with empirical findings in developmental psychology ([12]; [26]). In this perspective, to develop a coherent I is to recognize that the I-in-the-past and the I-in-the-future share the same interest and should act as one single subject with the I-in-the-present (see also [21]). Finally, jointly with IP-03, we will investigate the connection of such processes of intertemporal coordination with coherent bodily experiences in the interaction with the natural environment, which in a sense constitutes the physical correlate of selfhood.

Our IP will employ the following research methodologies:

• critical review of the relevant literature on the social genesis of self-consciousness;
  
• conceptual analysis of the key notions of "Selfhood", "Theory of Mind", "mental simulation" and "coordination", starting from cognitive approaches and building on a vast familiarity with these topics;
• critical comparison and evaluation of different approaches (e.g. Simulation Theory vs. Theory Theory), with special reference to their explanatory role in the genesis of interactive consciousness;
• cognitive modelling of specific processes and phenomena closely related to different modes of conscious experience, such as empathy, intentional action, team reasoning, and anticipation, with special emphasis on highlighting their relationships with interactive consciousness;
• cross-fertilization with other approaches to the genesis of consciousness, both within the consortium (e.g., Hurley's shared circuits model, Clark's active externalism, Tan's theory of engaged vs. detached emotions, Rambaldi's historical analysis of the concept of consciousness) and beyond it (e.g., systematic comparison with models and results taken from comparative anthropology, neurosciences, experimental psychology, theory of perception, studies on collective intentionality).

YEAR 1: Kick-off Meeting (month 0) and website construction (month 6)

YEAR 2: International Conference on Roles of Simulation (month 22) and editing of a special issue of an international journal

YEAR 3: Interdisciplines on-line conference, to whom all the partners will collaborate (month 31-34); CONTACT Final Report (month 36) reporting the activities and the achieved scientific results. In 2009 we are planning to organize the PPNB graduate/post-doc conference in Rome (further details will be arranged during the project time course).

Deliverables and/or milestones

The ISTC Unit, as Principal Investigator, will organize the CONTACT Project Kick-Off Meeting, aimed to discuss scientific and organizational aspects of the project (D1). This meeting will gather all PIs and their teams but there will be also two or three invited speakers to face research problems and issues addressed by the project. We will also arrange for the creation of a CONTACT portal (D2), online from month 6, which will serve as an internal coordination tool for those involved in the project, and as a dissemination device toward the research community. In month 22 we will organize and host an International Conference on Roles of Simulation, which will be preceded by an Project Meeting to discuss achievements and open questions of the project. Further details about this conference will be arranged at the Kick-Off and during the first year of activities. Another major event organized by this Unit will be the PPNB graduate/post-doc conference 2009 in Rome (see the section on documentation for further details).