Abstracts of Confirmed Speakers

Minds as Social Institutions
Istituto di Scienze e Tecnologie della Cognizione, CNR, Italy

I will first discuss how social interactions organize, coordinate, and specialize into artifacts and tools; how these tools are not only for “coordination” but for achieving something, for some outcome (goal/function), for a collective activity and objective. In particular I will suggest that these artifacts specify (predict and prescribe) the mental contents of the participants, both in terms of beliefs and acceptances and in terms of motives and plans. In contrast with prevailing behavioristic views of scripts and roles, I will argue that when we play a role we wear a “mind”. Wearing such a public mind turns out to be necessary for collective action and for an important form of automatic “mind reading” (mind ascription).
Second, I will try to argue that what really matters is the ascribed/prescribed mind, not the real, private one. We have to “play” (like in symbolic play) as if we had those mental contents. This social convention and mutual assumption makes social interaction possible, and allows “the game we’re playing” (Garfinkel). Moreover such ascribed beliefs and goals are not necessarily explicitly there; they might be just implicit in the sense of “inactive” (we act just by routine and automatically) or implicit as “potential”. The coordination and social action works thanks to these “as if” (ascribed and pretended) minds, thanks to those conventional constructs. Our social minds for social interactions are social institutions.

Intentionality as Documentality
University of Turin, Italy

Let us imagine a marriage that takes place in the total absence of documents or with documents written in invisible ink. Let us also imagine that all of the videocameras, the photographic cameras, the cell phones and the official documents had, for some reason, not registered anything. And, to complete the scene, let us imagine that, immediately after the ceremony, the newlyweds, the registrar, the witnesses and all the guests had drunk, mixed in their champagne, a chemical called “amnesine” that makes them all forget every memory of the event. Should we really say that the couple is married? There are strong reasons for denying that they are, given that no-one – not even the people most directly concerned – know anything about it. Indeed, marriages, like promises, bets, parties, revolutions and economic crises, only exist if we are aware of them, and to be aware of them, we have, in the first instance, to remember them. This is why the contemporary world has seen a gigantic explosion of the means for writing and recording. They are not so much means for communicating, as means for ensuring the fundamental social good, namely recording.
The effort to explain such a situation by means of collective intentionality by itself looks helpless. Whereas, by positing documents and records as the key notion to understanding social reality, we reshape the theoretical framework of social ontology and provide the means for the recognition of the sphere of social objects: money, artworks, marriages, divorces, economic crises, research projects, lectures and university degrees... These objects fill up our world more than do stones, tress and coconuts, and they are more important for us, given that a good part of our happiness or unhappiness depends on them. Yet we do not always take account of them, and even more rarely do we ask what they are made of, taking them seriously only when we lose our wallet or train ticket, our passport or credit card and we set to searching, paying, phoning, writing e-mails and queuing in all sorts of offices. It is only then that we understand (too late, alas) that social objects are made of inscriptions, whether on paper or on some magnetic support, or even (in the case of the promises we make every day) in people’s heads.
The explanatory core of social ontology, hence, is not collective intentionality but documentality (http://en.wikipedia.org/wiki/Documentality). And the law that brings social objects into being is not X counts ad Y in C (also in Searle’s new version of 2010), but Object = Inscribed Act. What this means is that a social object is the result of a social act (one that involves at least two persons or a person and a deputed machine), which is characterised by being registered on a piece of paper, in a computer file or on some other digital support, or even simply in the heads of persons. It is only on this basis that it is possible to develop an ontology capable of classifying documents and their selective storage, beginning with the grand divide between strong documents (inscriptions of acts), which make up social objects in the full sense, and weak documents (recordings of facts), which are secondary derivatives and of lesser importance.

Testing Collective Intentionality Theories: Shared Goals, Preferences, and Counterfactuals
Francesco GUALA
University of Milan, Italy

Theories of collective intentionality have roots in three different disciplines: group identity theory in psychology, team preference theories in economics, and collective intentionality theories in philosophy. Although philosophers have often endorsed a naturalistic approach to collective intentions, they have mostly focused on purely conceptual issues such as the reducibility of collective to individual intentionality or the existence of logically distinct forms of collective intention. Psychologists and game theorists, in contrast, have provided most of the empirical data that is currently available on this topic.
This division of work is far from optimal, for various reasons. One of them – the main motivation for this paper – is that empirical research tends to promote conceptual clarification, inducing the operationalization of concepts that otherwise may remain detached from empirical reality. In this paper I give an example asking what it means to be cooperative and to share an intention, and how shared intentionality can be investigated in practice using empirical data.
A collective or “shared” intention, in Bratman’s sense, involves a common (shared) goal and a plan to achieve it that requires interdependence of action among the individual agents (see also Tomasello et al. 2005). It is easy to demonstrate that sharing the representation of a goal is not sufficient for CI. Using a game-theoretic model (Bacharach 2006), I argue that the key aspect of CI is not the “sharedness” of goal-representation but the robustness of cooperation to a set of possible perturbations in the payoffs of the game. The main difference between an individual and a collective intention lies neither in the state of affairs that is jointly represented as the goal of the action, nor in the path leading to the achievement of that state of affairs, but in what would happen if certain changes in payoffs were to occur. A collective intention is robust to some manipulations of individual payoffs that encourage deviation from the plan, in a way that an individual intention is not.
This has two important consequences: first, Michael Tomasello’s insight that cooperation and CI require a special motivational basis (the assignment of collective values to actions and/or states of affairs) is vindicated. Second, experiments that test CI theories must focus on counterfactual scenarios where payoff functions are manipulated, or role-reversal designs where different payoffs are associated with different roles. I conclude giving some examples of research that goes in this direction.

De Re We? Extended Mind, Normative Inferentialism and Shared Action
Hellenic Open University, Greece

The paper advances a dual thesis about shared action as bearing unique characteristics that enable it to realize deontic inferentialism (originally developed as a philosophy of language) and cognitive interactive systematicity suitable for plural (extended) minds. In viewing social action in this way, the indispensable element of sociality, implicit in language communication or distributed cognition, becomes explicit.
(a)    The concept of sharedness, involved in an essentially plural action within a social context, is here approached via philosophical semantics and a certain view of normative inferentialism developed earlier on by Brandom. While the latter version contains as its central core a substantive thesis about the social dimension of a speaker’s deontic commitment with regard to the meaningful use of language in communication, it does not – and cannot have – at its disposal the necessary robust conditions for a sharedness that can only be provided in the case of plural acting. Plural action can achieve a genuine normativity stemming from a set of corresponding deontic commitments embedded in its necessary conditions, in the sense of actualizing the model of normative inferential semantics. Speaker-ascriber duality requiring normative connection to close the gap between the two does not arise in the case of shared action where both perspectives are intermeshed. 
(b)    This essential feature of normative inferentialism thus extracted from the case of language use is then injected into the theory of the Extended Mind and Distributed Cognition. Shared social action modeled on normative inferentialism enhances the thesis of the extended mind in accounting for the emergence of interactive cognitive systems true of shared action. It thus realizes a central thesis of the theory of the extended Mind. If what underpins Distributed Cognition is the extension of individual cognition into an interactive system of cognizer plus epistemic tools, as it is claimed by this theory, then the required interactive systematicity is fully realized in the case of plural social action.
Combined together, the two theses, yield, as far as social action is concerned, the obverse of what inferential semantics require: in the social context we must specify the de dicto content of the belief(s) involved and not merely how a mental content is hooked onto an object of the world non-descriptively. It requires specifying how the concepts entering into the expressions used for the specification of the de re content involve (or not) shared concepts. This entails its congruence with the thesis of distributed cognition.

Transforming Agency: From Individual to Shared Intentions
Elisabeth PACHERIE
Institut Jean-Nicod, Italy

One central aim of philosophical analyses of shared intentions is to capture what makes joint actions intentionally joint. The accounts that have been offered tend to require great cognitive sophistication on the part of co-agents or to be highly normative. They also typically require that some form of communication, verbal or otherwise, be possible between agents, allowing them to influence each other's intentions or to form explicit or tacit agreements to act jointly. By making joint action so cognitively or normatively costly, these accounts make it hard to explain the development of shared agency in ontogeny or indeed its widespread occurrence in everyday life. The team-agency theory developed by economists may offer an alternative route to shared intention. I concentrate on Michael Bacharach's version of team-agency theory, according to which cooperation among agents is made possible by an agency transformation involving a switch from an approach to a situation where agents reason as individuals to an approach where they reason as a team or group. According to Bacharach this agency transformation is the result of context-dependent psychological processes of self-framing. I will address three main issues. Can Bacharach's view of agency transformation as a matter of framing yield an account of shared intention that is less cognitively demanding than other existing accounts? Can it suggest a different conception of the role of norms in shared agency? Can we use it to shed light on the development of shared agency in ontogeny?

The Ontogeny of Social Ontology: The Development of Collective Actions with Status Assignment
University of Göttingen, Germany

The development of collective intentionality will be discussed from an ontogenetic and comparative point of view, with a special emphasis on the ontogeny of institutional activities. 
First, recent research in developmental and comparative psychology will be reviewed that suggests that even simple we-intentionality seems to be unique to humans (contrary to claims by Searle (1995), for example) and develops in the course of the second year.  This is documented most clearly by experimental studies with human children and non-human primates in the areas of collaboration, cooperative communication and so-called “role reversal imitation” (see Rakoczy & Tomasello, 2007; Tomasello et al., 2005). 
Second, special emphasis will be put on the development of (proto-) institutional practices, that is, collective forms of actions involving the assignment of status functions and constitutive rules (sensu Searle, 1995, 2009).  Existing research with non-human animals strongly suggests that such forms of collective intentionality present a uniquely human achievement.  Regarding human ontogeny, recent research is reviewed and our own new research is reported enquiring into the developmental course of proto-institutional activities.  What emerges from this is that joint games are a plausible candidate for the ontogenetic cradle of institutional reality.  From their second year on, children begin to play joint games with others, in particular games of pretence.  In joint games of pretence, objects acquire fictional status and thereby ‘count as’ something else (Walton, 1990).  A series of our own studies (Rakoczy et al., 2004, 2006; Wyman et al., 2009a) documents that children from age 2 understand that such fictional status assignment implicitly sets up a normative-inferential framework of appropriate acts in the course of the game (in analogous ways as performative speech acts do): Children respond in inferentially appropriate, systematic and creative ways to play partners’ pretence overtures.  A new set of studies (Rakoczy, 2008; Rakoczy et al., 2008, 2009, in press; Wyman et al., 2009b) documents that young children are not only able to act in accordance with the implicit rules of the game, but understand them as normatively binding in the context of the game: when a third party announced an intention to enter the game and then acted inappropriately, confusing the fictional status of some objects, children responded with protest, critique and teaching. 
Third, some potential arguments are discussed why joint games of pretence might be a, perhaps the, ontogenetic cradle of institutional phenomena. One argument is the following: In pretence, children necessarily have to take a dual perspective towards objects –keeping in mind and separate their real and fictional identities- otherwise it would be delusion rather than pretence.  This is in contrast to language: early on children can see through language without grasping its status structure much as one sees through one’s eyes without seeing them (see Wittgenstein, 1961, § 5.633).  Furthermore, pretence might be an ideal entry gate into institutional reality in contrast to many institutional practices, because it has concrete, transient and action-based contexts, and is less holistically interwoven with the rest of institutional phenomena (a possibility discussed, but too quickly rejected regarding games by Searle, 1991).

Plural Self-Awareness

Hans-Bernhard SCHMID
Universität Basel, Switzerland

The existing accounts of collective intentional attitudes can be classified according to where exactly the “plural” (or "collective") element is placed in the analysis: content, mode, or subject of the intentional states in question. In this paper, I argue that plurality can neither be limited to the content, or the mode of the intentional states in question, and that the one existing account of plural subjectivity (Margaret Gilbert’s), while being right on target concerning the question where to place plurality, fundamentally misconceives of the kind of self-reference that constitutes plural subjectivity. I argue that plural subjects are constituted by plural self-awareness in the exact same way individual subjects are constituted by individual self-awareness.

Group Reasons
University of Helsinki, Finland

Some groups can act as units.  In my earlier work (e.g. Tuomela, 2007) I have spoken of “we-mode groups” that are autonomous, self-governing groups capable of acting specifically because of their members’ we-thinking and we-reasoning (reasoning and intending together).  These groups can be regarded as functional  group agents. They can be taken to act intentionally for a reason.  We must distinguish between group agent’s reasons and its members’ reasons qua group members. The latter are conceptually based on the group’ ethos, viz. on its basic goals, beliefs, standards, etc. I will here call these member-level reasons group reasons (and  we-mode reasons, by definition as it were).
Acting as a group member is ethos-based and differs from acting as a private person (i.e. in the “I-mode”). The fact that the group has accepted a particular ethos for itself will always be part of a members’ group reason—that the others will have to share, too. There may be other “ingredients” in a group reason, e.g. simply the fact that the members have decided to do something together, say, to paint a house.
Such a shared group reason is needed because the members cannot properly act jointly without there being such a unifying factor that guides their action that should respect the ethos. This group derives from the group’s ethos or its other acceptances, decisions, goals, directives, or something analogous. To be sure, they have been formed by group members perhaps much earlier or by different members altogether, etc.—direct circularity need not be involved here. To guarantee unified action, then, a shared group reason must be involved at least in broad sense, even when the members are performing different tasks and functioning in different roles. Functioning as a group member in a we-mode group (a democratic, self-governing group) entails giving up part of one’s “natural” authority to act (i.e. power and right to act) in general. In contrast to the I-mode case, the members must give that part (their “ethos-related” part, so to speak) to the group, for the members joint use, for otherwise they cannot act as a group in a group-guided way and as a part of a strong “we”. The members giving up that part of their authority means that the group gets full authority over group action and is responsible for it.  In the I-mode case a group member can act for the group for an I-mode (viz. privately accepted) group reason, but he does not give up any part of his natural authority to act to the group.