1. Introduction

Raimo Tuomela has been doing research since the end of the 1970’s in the foundation of the social domain with the focus on collective intentionality, cooperation, and related subjects, for example, collective action, social groups, institutions, group responsibility, and commitments. At the same time, his extensive theoretical work is a contribution to an analytical philosophy of sociality and social ontology. Social ontology is the study of the fundamental elements of the social domain and thus an analysis of its ontological commitments, for example, the elementary components, of the best explaining scientific social theories and their postulated ontologies.

Tuomela’s research is relevant not only for philosophers and sociologists because there is interdisciplinary research between both, but also his studies overlap with problems in the philosophy of mind, psychology, and cognitive science as well. From the philosophical point of view, Tuomela has helped to extend the concept of practical philosophy and the theory of action which have so far analyzed the concept of intention, belief, intentionality with respect to individuals only.

In particular, it is accepted in the literature that Tuomela (1984, 1995, 2000, 2002, 2007, 2013), Tuomela & Miller (1988) together with Gilbert (1989, 2000, 2006, 2014), Bratman (1993, 1999, 2014), and Searle (1990, 1995, 1998, 2001, 2010) are the four “founders” of the contemporary focus on collective intentionality, beliefs, and intentions of the philosophy of sociality since the end of the 1980s. (Chant, Hindriks, Preyer 2014a) Some recent results in the area are found in, for example, Miller (2001, 2010), Ludwig (2014), Tollefsen (2015), and many others, the volumes in the Springer Series Studies in the Philosophy of Sociality, Chant, Hindriks, Preyer eds. (2014b), as well as papers in various journals such as Synthese, Economics and Philosophy, and ProtoSociology, on the critical examination of Tuomela’s philosophy of sociality, see, Preyer, Peter eds. (2017a). The accounts differ in particular on the ontology of the social domain, for example, is there a sense in which groups have their own intentions and beliefs? This problem takes effect in social ontology because its main question is whether there are such things like irreducible social system properties and processes? Also relevant is a debate between Gilbert and others concerning collective belief on her account of it (Gilbert 2002, Meijers 2002, 2003, Tollefsen 2002, 2003, McMahon 2003, Brad Way 2003, Preyer 2012a); the others, labelled by Gilbert the “rejectionists” argue that Gilbertian collective belief is not belief but rather acceptance, on Gilbert and Tuomela, see, Tollefsen 2015, on a review of Tollefsen 2015, see, Preyer 2017b).

Tuomela is orientated on a foundational analysis of the conceptual resources and the philosophical presuppositions of the social domain. He has extended his earlier studies to an integrated account of a collective acceptance theory of sociality and social ontology which brings together the analysis of collective intentionality, social group, cooperation, social institution, collective responsibility, and the socio-cultural evolution (Tuomela 2007). From the sociological point of view, Tuomela’s account is relevant because of its impacts on, for example, the theory of social systems and sociological theory in general. It can also be applied to the field of multi-agent artificial intelligence.

Tuomela’s core philosophy of sociality and social ontology is sketched below. The account is based on a comprehensive concept of rationality that entails the ends -, value -, and expressive rationality (Tuomela 2000, pp. 299-300). From the theoretical point of view, Tuomela argues that the rational choice theory based on action, preferences, and degree of belief, plus the cooperative game theory are powerful mathematical tools, but these proposals are conceptually and theoretically too weak and in need of modification since, for instance, rational choice theory does not seriously consider the concepts of goal, reason, and commitment. However, it also does not exclude that the mathematical results of cooperative game theory contribute with clear analytical tools for the analysis of collective goals.

Outlining Tuomela’s philosophy of sociality and social ontology, it is instructive to begin with his constructivist analysis of the philosophy of sociality (1.), starting with the key-concept of collective acceptance (2.), and his analysis of We- and I-attitudes (3.). Both are, in his account, applied to the characterization of the constitutive features of social groups (4.). The conceptual analysis of the shared point of view is also helpful for the analysis of the authority, cooperation, social norms, and institutions of the social domain. It considers Tuomela’s (2013) analysis of egalitarian and hierarchical group organization along the distinction between autonomous and non-autonomous groups that can be related to different aspects of collective acceptance (5., 6., 7.). Finally, some points are made about Tuomela’s philosophy of sociality as having as its core a sociology of membership because the basic concept of the analysis of all features of the social domain is a theory of membership, and at the end a conclusion is drawn about his view (8., 9.).

2. Philosophy of Social Practices

The idea of Tuomela’s philosophy of social practices is that these activities are constituted by collective intentionality, basically as shared we-mode activities. There are strongly analogous views, for example Gilbert’s (1989, 2006) on social conventions and rules and Searle’s (1995) as well.

Tuomela calls this view “wide program of social constructivism” because a central part of the systems of communication is constructed collectively by the concepts of collective acceptance. This is a matter of the ontology of the social domain as a subject of constructivist analysis. Tuomela also defends “the narrow program of constructivism”, that is, the collective intentionality as shared we-mode is for the conceptual construction and maintenance of social institutions. Strong collective intentionality is analysed through

1. group reason, that is, a unified reason as a condition of participation in constitutive group-activities,

2. the collectivity condition, that is, “necessarily being in the same boat”, and

3. collective commitment. Collective commitments are generally based on the members’ joint intentions that are comparable to those of the single-agent case in which intention entails commitment. Thus, joint intention entails collective commitment. Such commitments can, in some cases, be the members’ reasons to perform their parts or shares of the group’s action at stake. Compare with Gilbert’s on joint commitments: She has long emphasized the analogy between joint commitments involving two or more people and personal decisions and intentions which involve personal commitments.

The constructivist program is presupposed for the analysis of the conceptual activities and rule-following as a pattern-governed behaviour. Tuomela goes along in part with W. Sellars and L. Wittgenstein. The basic concept is oriented toward a collective pattern of behaviour, which is not satisfactorily analysed by Sellars. (Tuomela 2002, pp. 40-52) Contrary to a holistic and communitarian interpretation of rule-following, it is not assumed that thinking and speech are things that we cannot do on our own. This is emphasized by Tuomela, because fully developed conceptual activities, which require speech acts, have a social background.

The wide program of social constructivism must defend the collective acceptance theory of sociality. Therefore, the task is to show how the collective acceptance, analysed by the appropriate relevant we-mode, is the basic feature of the social domain ontologically, which cannot be reduced to psychological or biological properties. They are not informative about the structure of the social domain. Collective acceptance is based on the performative construction of social entities by the members of social groups and institutions and is self-referential (reflective). For example, the Euro is money only if the currency is accepted among the members of a particular social domain. (Tuomela 2002, on his “central account of social practices”, pp. 89-99.) These features of the social domain are also discussed, for example, by Barnes (1983), Bloor (1997), Kusch (1997) and Searle (1995, 2001, on Searle, Tuomela 2002, chapter 6.4, on Searle’s new account Searle (2010), see, Tuomela (2013), chapter 6.8.)

Tuomela contrasts the we-mode with the I-mode. The we-mode means thinking and acting as a member of a group. But collective intentionality is not enough for the analysis of institutionalized social practices and communication because the institutionalized components of the social domain are also structured by social norms, as expectations of expectations, sanctions, authority, and regulations of communication. The collective acceptance theory of sociality is seen as an aid in the analysis of institutions like language, money, marriage, ownership, and formal organisations.

The analytic framework of the core of Tuomela’s philosophy of sociality and social ontology is the structure that tentatively connects “we-mode ↔ collectivity ↔ collective sociality ↔ (broad) institutionality” by equivalences. (See Tuomela 2007, pp. 211, on collective acceptance and sociality, pp. 187-199.) We-mode items (for example, intentions, beliefs, groups, cooperation, when in the we-mode, and institutions) presuppose collective acceptance. The foundation of the structure is the membership in groups in the social domain because the collective acceptance of the members as members is a constitutive condition. Without that there is no collective sociality.

3. Collective Acceptance

The key-concept of Tuomela’s philosophy of sociality is we-mode collective acceptance. This concept has, for him, a paradigmatic relevance in the social philosophy and social ontology. Therefore, he has earlier called his account the collective acceptance theory of sociality (see (Tuomela 2013), chapter 5, for his most recent, slightly revised account emphasizing that collective acceptance presupposes the members’ acceptance as a group). Collective acceptance involves an agreement about we-attitudes, that is, a particular attitude of members of social groups to an attitude content p, entailing collective commitment of the group members as members to satisfy p together. Most collective commitments are group-social and intention-based. They need not be normative in a moral sense.

Collective acceptance is linguistically analysed because the concept implies that the accepting members hold a particular kind of collective attitude toward the respective sentence or proposition in question. Thereby collective acceptance involves instantiated expectations and expectations of expectations of members of the social domain and, in particular, its differentiation; a membership system. Actions of the members are therefore dependent on the correspondent attitude. The collective acceptance may come in different strengths. All we-mode collective activities involve a kind of joint action and cooperation directed towards satisfying a goal-content.

The central unit of collective intentionality is joint intention, that is, the members of a social group share as group the content of an intention x. A typical joint intention is often expressed in the form we as group will do x. On joint intentions, see (Tuomela 2002), chapter 2, (Tuomela 2007), chapter 4, (Tuomela 2000), chapter 3, and, in special, (Tuomela 2013), chapter 3. Joint intentions in the we-mode cannot be reduced to I-mode intentions. The content of such joint intentions is “the participants jointly intend to jointly see it as a group where x (x= state of affair or a joint action) comes about”. We-intentions are often “aim-intentions” (intentions that some state of affairs, possibly a collective one (for example, that there be peace in Syria), come about or be maintained. Such intentions are to be distinguished from ordinary by action intentions. One possible account of how joint intentions come about is given by the bulletin board view––see, for example, Chapter 5 of (Tuomela 2013) and the chapters that precede it.

The distinction between group-(we-) and I-goals is a strategy of the participants because the commonality of collective intents is a matter of the assumed agreement. It may be that the agreement about collective actions as means is to reach an underlying private aim of some of the participants. But collective goals presuppose collective cooperation and coordination of actions as a means.

Tuomela (2009) introduces the bulletin board view of collective acceptance in order to describe the general conceptual feature of forming joint intentions and beliefs: For instance, the organizing committee communicates, on a black board, “Members of group g will clean the park next Saturday” [(Tuomela 2007), chapters 4 and 5, (Tuomela 2002), Chapter 4]. “Those who will participate, please sign up here”. The “will” expresses an intention, and not a prediction. The account has the advantage:

1. It explains the collective and shared acceptance of attitudes (intentions, beliefs, and other voluntary attitudes) and exemplifies the acceptance with the presupposition of the shared task. If enough members come together cleaning the park then there exists a categorical, non-conditional intention, and a joint intention among the members, who sign the list;

2. The members have committed themselves to the goal in question;

3. The account does not assume a prior intention to form the joint intention cleaning the park;

4. The members can take different symmetric, asymmetric, and complementary social status functions and roles solving the task;

5. There is a strong epistemic intention because the members can verify the content of the intention; and

6. The account is also applicable to I-mode intention formation.

Tuomela revises his version of collective acceptance with respect to egalitarian and hierarchical groups. He gives an analysis of social structured groups with different roles and status positions in which not all members need to participate in the collective acceptance. (Tuomela 2013, Chapter 5, esp. pp. 127-130.)

He also discusses other accounts involving the formation of group attitudes in chapter 5 of the aforementioned book. The analysis of the social domain by the acceptance model is “philosophical (conceptual, metaphysical)”, and “design-theoretic”. Tuomela (Tuomela 2002, pp. 158) emphasizes that collective acceptance involves a “procedural” component. (Tuomela 2013, pp. 125.) The orientation of the possible participants of the “voluntary group attitude formation” is “Which attitude (e.g., want or belief, as the present case may require) should we accept for our group in this situation?” (Tuomela 2013, pp. 125.) Therefore the collective acceptance theory of sociality leads to the analysis of the collective we- and I-attitudes of members of social groups.

4. We- and I-Attitudes

Tuomela applies the conceptual tool of collective acceptance to various kinds of social practices. Social practices and communication can be of many different sorts: For example, working in a business company, participating in road traffic, meetings, and Christmas parties. All these communicative events are connected by a unified concept of a collective attitude. This attitude can be analysed by shared we-mode, collective intentionality.

Definition of we-attitude:

A believes that p and believes that the others in the group believe that p, and all this is mutually believed by the group. If all the members have a we-belief of this kind, then we are dealing with a (weak) kind of group belief that p.

The content of this mode is a shared social ground, which the participants have for the intended collective social action, that is, the we-mode members of the group act together as one agent. But this does not entail that there is a group, agent or person in a strong sense; only in a weak functional sense (Tuomela 2002), pp. 128-129, and (Tuomela 2013), chapter 2).

The feature distinguishing between the we- and the I-mode is collective, group-based, acceptance and commitment to the we-mode-content in contrast to a private commitment in the I-mode case.¹ With respect to this distinction, we always take into account asymmetrical conditions of the interaction: It may be that I am committed to a private attitude without being collectively committed and vice versa. It is to emphasize that Tuomela makes the distinction between

1. collective sociality respectively the collective social features of activities, and

2. mere collectivity because there are many social actions that are not performed collectively. For example: How one thinks about others.

3. There are also collective activities that are not social, for example, when people are walking in the street opening their umbrellas (M. Weber).

The predicate “collective” applies to a set of people, but the predicate “social” is instantiated by attitudes that take into account the attitudes of others.

The relevant conceptual component of collective attitudes is the collective shared goal. Tuomela makes the ideal typical distinctions between:

1. collective goals, which are based on a mutual we-goal (we-want), and

2. intended collective goals.

The first type is the weakest, and the second type is a strong concept because it entails an intended collective goal, although there is no plan-based joint intention.

3. Participants have a basic joint intention to jointly achieve a particular goal. The third type equates ‘every participant intends to contribute to the joint action and the participants form a we-intention to achieve the goal’. This intention is held by the participants as members of a group, and they are mutually committed, here a strong version of a collective commitment. 4. Goals which are ascribed to members of a social collective: For example, goals of organizations. The fourth type shows the members as the basis for the ascription of goals because the ascription is instantiated by the members of a collective. If someone is a member of a collective, then he satisfies particular conditions that make his membership possible. The structure of a collective includes:

(a) a decision system, which fixes the goals of members. This implies that not all members have the fixed collective goals, but they are obliged to go along with the goals.

(b) A collective can pursue a goal, although the members have a weak we-intention only. This is the statistical feature of the concept of collective goals. These matters also concern the authority system.

Tuomela analyses strong collective intentionality in terms of we-mode thinking and reasoning that is connected to group members’ we-mode activities, based on group reason, the satisfaction of the collectivity condition, and collective commitment. Collective intentionality is connected to cooperation because the latter is a significant example of collective intentional mental states and action. Both are based on a we, that is, a social group the members of which engages in collective reasoning and acting.

5. Social Groups

A social group is a part of a social action-communication system, that is, a social system with different environments. Groups are not a person, they have no body, but they are composed of members. Besides, they do not possess intrinsic, but only extrinsic intentionality. In Tuomela’s recent account, social groups are viewed as functional group agents to which mental states predicates, actions, and responsibility can be attributed to group members and others. This is relevant for his theory of corporation.² A group-based reasoning is: “When functioning as group members, we want X and take this to require that we jointly do Y and hence do it as a group”. Groups (Tuomela 2013, chapter X, pp. 99) consist of group members and their interrelations, and, in the case of organizations, also necessary material tools. Therefore, Tuomela’s version of group agent refers to the membership unit. Accordingly, the social domain is constituted by members and non-members of groups, their status functions, roles, and expectations. Group membership presupposes mutual knowledge of this membership (on the concept of social group, (Tuomela 2007), chapter 1, and (Tuomela 2013), chapter 2; on the concept of social group in sociological theory, (Preyer 2012b).

Tuomela’s substantial characterization of sociality is that the social domain is based on the point of view of the social group, that is, the shared we-perspective (attitudes) of the members of groups. The full we-mode presupposes a social group in the strong sense. Therefore, membership is presupposed as a constitutive condition. But this also presupposes collective commitments and their public access. The strong collective intentionality (we-mode) is that acting together intentionally as a group has to satisfy authoritative group reasons, collective commitments, and the collectivity condition. The core of Tuomela’s philosophy of sociality is the collective acceptance theory of sociality. Therefore, it is to show how this theory works together with the shared point of view and the group concepts.

The we-mode entails group concepts of the members of social groups. The lingual equivalent for that is collective-referential expressions: For example, we, our or one of us. This is not trivial because taking the group point of view is connected at the same time with the acceptance of group goals, values, beliefs, and norms to which the members are committed. These are commitments as a self-binding of members of groups. Therefore, the foundation of the perspective of a social group is the self-binding of their members. Tuomela calls this the we-mode attitude. Thereby is expressed the thinking and doing of a member of social groups. The I-mode is contrary to the attitude of a private person. But the I-mode can be connected with the we-mode if the speaker addresses determinations of others and he intends to cooperate: For example, “I want travel with you to London, how do you think about …?” The crucial point here is that we-attitudes are not to reduce to I-attitudes. Tuomela (2009) analyses shared we-modes by the concept of collective intentions and shared mutual beliefs. Thereby is to answer the ascription of attitudes to social groups (collectivities) as distinguished form individuals as single person.

The ideal-typical description of we-attitudes of person specified to social groups is:

1. The attitude exists only if the group has this attitude,

2. It is presupposed the attitudes are shared among the members,

3. It is assumed the mutual shared attitude is shared factually.

This does not mean that there are no exclusions of the members of groups. The shared we-attitudes are a social ground of the members of groups for the correspondent intentional collective action and the participants of cooperation are committed collectively to the correspondent goal, which they have not to their disposal voluntarily. This account names Tuomela collective goal theory of cooperation.³ Thereby it is connected to the analysis of we-attitudes and social groups with his analysis of cooperation.

Tuomela analyses the concept of the I-attitude and we-attitude. The we-attitude-belief means we, the members of the group g, believe that p (Tuomela 2002), on a summary of acting as a member of group, pp. 41). This belief is accepted by the members of the group. The belief has a benefit for the group, and the members are committed to this belief ideally. A we-mode-attitude or action presuppose the satisfaction of the collectivity condition:

[T]he members as members of the social group who peruse together a joint goal performing x respectively have a joint intention, which satisfies the truth condition of quasi-conceptual reasons to the joint agreed perused intention iff the satisfaction of one member is at the same time satisfied of every member of the group. (Tuomela 2002, pp. 29-36)

The general assumption that the members of the group have a shared knowledge about the collectivity condition is presupposed. The collectivity condition is predicated collectively, and in the chorus to every single member distributively. The we-mode collective goal and the instantiated token of the collective intention of the single members are connected with the collective acceptance necessarily (Tuomela 2007), pp. 47-51).

Collective acceptance means, in the normal case, that:

1. every participant agrees with the intention and holds true that he himself has the intention to do x,

2. there is a shared belief about that the single members of the group coincide about their agreement, and

3. the participants are committed to their agreement. Collective agreements can vary in its strength.

The collectivity condition is a version of the golden rule applied to social groups. The member of the group in the I-mode-attitude is committed to its private goals and beliefs. The collectivity condition takes as a basis the full-blown shared point of view as a membership condition (Tuomela 2007), chapter 2).

The I-mode of the shared we-attitude occurs in different version. A person who is in a group connection peruses a particular goal p in a situation knowing that others peruse also this goal. This is a simple example for a conform attitude, which is shared by the group. The we-mode is diluted in the I-mode. In this case, what represents the we-mode is an I-mode. The case is hence relevant because just any collectivity, which has some intersubjective identity properties, bare based on a mutually shared knowledge of beliefs, and can have a we-attitude. For example, people at the train station who are waiting for the train share beliefs in a we-mode: “why the expected train is delayed”. In the case of proper we-mode/-intention is the group reasons not necessary (internal) contrary to the correspondent I-mode because the reasons are contingent (external). In the I-mode may happen a circularity, for example, “I want if you want …”. In such cases is happened a coordination dilemma: I want x (one of my option to do something) if you want to do x, but you want x if I want x and the on (Tuomela 2007), chapter 3, has analysed solutions to this problem).

This is related to the question, what is an intentional joint action?

A joint action is an action thereby the participants are determined as a group. (Tuomela 2007), chapter 5, uses the expressions “joint action” and “acting together” as interchangeable. The intentional we-mode is fundamental for the joint action because the joint intentional and the relevant belief are the guarantee about the existence of a joint action (on the final version of joint actions, see (Tuomela 2007), pp. 108-112). Joint actions are valid thereby that they are performed by the members as members of groups, that is, in the we-mode.⁴ This does not contradict that the core of Tuomela’s philosophy of sociality classifies groups as social systems.⁵

6. Authority

Tuomela’s view is collective acceptance among the members of groups and the group-attitudes are the conceptual core of the philosophy of sociality and social ontology. The analysis leads to the definition of the function of authorization and authorized members of social groups.⁶ For Tuomela authority and authorization is basic for his account of group reasons, group action, group belief, and the accepted attitude in general. (Tuomela 2007, pp. 129-134) The normative power structure of social groups is based on authorization of some group members.

Tuomela’s positional account of group attitudes means that there is a differentiation within a social system, group, or organization between operative and nonoperative members with respect to building of attitudes, and the operative members are internally authorized for types of given task solution (Tuomela 2007), pp. 129-130, (Tuomela 2013), pp. 130-136). From the sociological point of view, it is of particular relevance the shared we-modes of the decision makers of social groups as their operative members because they are significant for the self-binding of a social group as a whole and are, at the same time, responsible for the group belief, that is, they take the view of the group, and the operative members accept p and are committed to the proposition. The operative members are determined by status functions, which formally define roles and tasks of a group in corporations or informally does so in groups without written membership conditions in corporate handbooks or charters. An operative person for a group g can be a group member or in some cases a non-member. In general, an operative in a we-mode group is authorized by the group members for either decision making in g or for acting, for example, carrying out the group’s decisions on behalf of the group. A person can be operative in one sense (for example, for decision) without being an operative in another sense (for example, action).

A non-operative person (member) may thus be one concerning either decision making or action – or one that is “in reserve” for group tasks. Two aspects should be emphasized:

1. in a we-mode group all members may be operative members (for decision and action) if none has been specially authorized for a task, and

2. the authorization means that the nonoperative members are obligated to obey only if the members decide to keep their authorization in force—the authorization is meant to be for the use and benefit of the group and their members and is justified only if this function is fulfilled.

The distinction between operative and non-operative members divides group intentions because not all members decide on the intentions of the group.

Tuomela goes along with Raz’s view: That authority is a matter of preemption because the authoritative directive excludes and replaces the addressee’s own judgment (Raz 1986, 1975, on normative power, pp. 98-104). With the “positional account” goes along that a group is structured by a division of their members into operative and non-operative members with different status functions, social roles, and tasks. Authority is to instantiate of the operative members. These members decide and/or act for the group, that is, in the name of the group or other instances. It may be their decisions are also psychologically motivated, but they are desire independent decisions and reasons of the members in principle. Analysans is the concept of group attitude as a membership attitude.

The group attitude has an authoritative function for the practical relevant reasoning in groups. (Tuomela 2000), on cooperation and practical reasoning, pp. 141-156) The authorized members share particular attitudes in the we-mode as members of the group, that is, the joint intention formed by a pre-emptive reason is an authoritative group reason. Thereby are constituted the ontological status of groups and of membership because the “positional (authority-based account) of group action”, which basically refers to a we-mode action of a group member, constitutes the social domain as such. Tuomela’s analysis of authority relates to his “positional view of groups and social institutions (including organizations)”, that is, positionally functioning as a group member. The communication in this status functions is based on I- or we-mode thinking and reasoning. Authority is based on position holders, which are authorized performing actions, that is, to act and speech for the social unit.

Tuomela has continued the analysis of authority with respect of hierarchical social units (formal organizations). Analysans is the function of internally and externally authorized leaders. The first is intrinsically cooperative because the action in question is an intrinsic component of the group or membership action. These intragroup/-member actions are done in the we-mode as a full-blown member of a group agent which are intrinsically cooperative. The authority is given by the members, that is, by their collective acceptance. The authoritative power is over the members as members. The second is a non-autonomous we-mode cooperation in formal organizations (Tuomela: theory of corporation). They exist as a communication system, which requires unifying actions. The speech and actions of members of this type of social systems are the “limbs of a collective body, to adopt an apt metaphor” (Tuomela 2013), pp. 22).

In the case of external authorized leaders, there is a dominance of one group over the other. That is when the interrelation of individuals and their observation is not to be applied to the structure. In both cases, the leaders can give new directive and goals (see (Tuomela 2013) on external leaders, chapters 2, 4. 6). Tuomela (2013) goes along with Raz: That the power of the leaders works or is justified as long as the members of the groups are following the advice of the leaders, that is, for as long as those have normative power.⁷ Every authority system, as a system of communication and decision, as well as the operative-non-operative differentiation for the task-orientated system, can be a multi-layered structured by hierarchies.

Tuomela gives a new account of autonomous and non-autonomous groups.⁸ Basically an autonomous group is one that is governed by the group itself, that is, collectively by their members or by persons authorized by them, for example, to make decisions and accept beliefs for the group or to act with the purpose of realizing the group goals. In contrast, a non-autonomous group is governed by an external authority (for example, another group or person that has the powers of a dictator or something of the kind) or by an internal one not authorized by the group. In a non-autonomous group its ethos (viz. its constitutive or most central goals, beliefs, norms, standards, practices, and so on) is determined by others than the group members (in contrast to how they generally are determined in an autonomous group). This is an important feature because the ethos is of course highly relevant to explain the activities of groups members.

The activities of the operatives and other members presuppose in general that they act as proper group members in accordance with the normative obligations and recommendations that the ethos of the group imposes. If in a non-autonomous group the dictator (irrespective of whether he is a group member) changes the ethos, the original explanation is not valid (because it refers to the old rather than the new ethos). Of course, in the case of an autonomous group the operative members for decision making can collectively change the ethos, but then the explanandum will typically be different from what it is in the non-autonomous case—it is different because acting as a group member then has a different content (Tuomela 2013), chapter 2). Therefore, the autonomous-non-autonomous distinction is an important feature in an account of the group’s activities.⁹ This is the link to the features of cooperation in social systems, groups, and organizations.

This turn is not a contingent step in Tuomela’s philosophy of sociality because hierarchical and authoritative we-mode groups of formal organizations, which function as group agent, are a significant feature of the social domain. A group agent is not intrinsically an intentional agent, but extrinsically with respect to joint attitudes and actions of members of a group. It is a weak collectivistic conception of groups as intentional agents.

7. Cooperation

There are philosophical, mathematical, game-theoretical, and experimental accounts, which study the collective action dilemma in the analysis and research of cooperation. It should be also noted biological (ethnological) researches of cooperation among animals. The economic and game-theoretical accounts assume that cooperation is based on individual decider. Tuomela corrects these accounts because group reasons are required for the explanation of the most cases, and the game theory has not consider the institutional frame of reference significantly, therefore the prisoner dilemma is not the paradigmatic example of the analysis of cooperation. It is to conclude that the individualist basic theory of cooperation is not to generalize (Tuomela, 2000), chapters 6-7, (Tuomela 2007), chapter 7).

A structural feature of the social domain is constructed by collective acceptance as a “weakly cooperative activity”. Tuomela (2007) emphasizes the function of “collective social construction” in his philosophy of sociality. Cooperation is an instructive example for the analysis of collective intentionality because acting together requires joint intention, shared belief, and shared collective goals. Tuomela (2007) has analysed basic features and sorts of cooperation. He makes the self-explaining distinctions between cooperation:

1. as g-cooperation, that is “based on shared collective goals” as “full-blown cooperation” (institutional and non-institutional) and

2. i-cooperation as cooperation as coaction, that is, “based on compatible private goals” (institutional and non-institutional).

The distinction is made from the goals (intentional content), which the participants are orientated to achieve (Tuomela 2000), p. 10).

Tuomela’s main theses for cooperation are:

1. Basic Thesis of Cooperation: two or more participants cooperate if and only if they have a joint goal and act together to achieve their goals (Tuomela 2000), chapter 12, pp. 12-14).

2. Commonality Thesis: Ceteris paribus, that is, there is a (definition of) situation within is to prefer and to initiate a successful cooperation for the commonality of preferences of the participants (Tuomela 2000), 12, pp. 12-14).

3. Closeness of Given and Final Preferences Theses: Ceteris paribus, that is, “the closer (and the higher) an actor’s given and final preferences, viz. preferences_giv and preferences_fin, are, the more likely he is to cooperate rationally in the long run (in a sense respecting his given preferences)”. This is defined in terms of measured utilities in terms of given preferences. (Tuomela 2000, pp. 15)

4. Reward Thesis: Ceteris paribus, that is, if the participants expect a more reward than in the case of non-cooperations then a rational agent cooperates (Tuomela 2000), 16, pp. 16-17).

5. Institutional Thesis, that is, cooperation and its organization is central for institutions (Tuomela 2000), pp. 17).

Tuomela’s concept of cooperation is: cooperative working together has as core full-blown cooperations (Tuomela 2000), chapters 1-4). These are cooperations based on shared collective goal (g-cooperation (g-/we-mode, we-ness perspective). The preference correlation has a leading role in particular in the case of full-blown cooperations. These correlations have a stabilizing function, but they contribute also something for the flexibility and motivation engaging oneself and continuing cooperations. For a theory of cooperation, there are not interesting collective actions as such, but cooperation, which are based on social reasons. It is assumed that the collective social action is strongly full-blown cooperative action. The achievement of objectives presupposes a we-mode, which is not to reduce to I-mode and its aggregation. Institutional cooperation is based on group cooperation. Tuomela’s turn is that normative reasons take effect in the paired institutional preferences, which contribute to cooperations. This entails also normative authorities, which are relevant in this cases.

The collective acceptance theory of sociality assumes that a relative weak sort of cooperation is also enough for a collective acceptance producing and maintaining collective social events and social institutions. Social norms and agreement are cooperative activities. Rituals and ceremonies are extreme cases of these activities.¹⁰

Cooperative preferences are factors out in basic components of social control, whose underline types of social situations of interactions within the group of participants have a mutually shared knowledge about collective activities.¹¹ The subject is the analysis of the “semi-motivational” component of social control. The components are the structural or situational determinations (“multi-agent action” in a technical sense). The account is exemplified to simple two-person-interactions, that is, the participants could be factoring out the component of the benefit of their mutual result expressed in:

1. the absolute control of their own actions,

2. respect to the actions of the other participants, and

3. their conditional and interactive control.

The relevant correlation between the preferences of the participants of their cooperation can be measured and is dependent in particular on the component of the social control. Group- and I-cooperations presuppose correspondent preferences among the participants. Tuomela’s interprets, in the cases of conflicts with respect to public good, collective goals with the collectivity condition, that is, the intended collective goals are determined by the mutual goal of the participants and are fixed therefore by collective commitments.¹²

Tuomela classifies cooperations under different—also overlapping—features.

1. It is to distinguish between group- and I-cooperation.

2. It is to take into play the similarities of preferences, that is, the degree of correspondence, which measures the result of the planed mutual actions in principle. These are the collective results, which result from the respective contributions of the participants.

3. Another factor is the followed strategy. A good example for that is the prisoner dilemma.

4. The interface between the components of social control, and the given situation of cooperation.

5. The institutional respectively non-institutional feature of cooperation, for example, in the social subsystems the directives in corporations, monetary operations, or organizational cooperation in the politic and scientific system and in different communities (Tuomela 2000), Chapters 3-4, 6, 8-10, Part V).

The Basic Thesis of Cooperation is that cooperative acting together builds the core of full-blown cooperative actions based on group-mode/we-mode. Cooperations are the link to the existence and maintenance of institution, and hence, sociality in general. But not all cooperations are institutional.

Tuomela turns to a new version of the theory of cooperation, which takes in the analysis the authority system (Tuomela 2013), chapter 6.3). He concedes that his account is open to different ontological interpretations. His analysis is to interpret by a difference scheme of the participations of cooperations because we-intention and the collectivity condition are to be instantiated by the membership in groups and status functions. It is to emphasize that he argues for the centrality of collective reasons for cooperation, which are illustrated with game theoretical means (Tuomela 2000), chapter 11).

8. Institutions

Tuomela’s claim is a general theory of social institutions abstracting from their varied features.¹³ This is to emphasize because the concept of institution is not clear enough in the literature. Institutions as social artefact are “collectively man-made”. The concept of institution is a reflective one, which is constituted by a system of norms based on collective acceptance (Tuomela 2000), chapter 6), “Appendix: Institution Concepts as Reflexive Concepts”) Institutions involve the we-mode and not an I-mode activity of their members only. They define “ground rules” respectively expectations of expectations for their members. (This is also the account of Eisenstadt 1995c, pp. 344-348). The activities of their members are determinated by social (status) functions. Tuomela outlines institutions collectively (collective pattern-governed behaviour), which are also an account for the analysis of social organisations. They are not necessarily intended. Institutions’ work also depend on routines and non-intentional behaviour (Tuomela 2013), pp. 215-216). Typically, institutions have an authority order and authorized members give devices to their members. The authorized members established rule-norms which are valid for the members in question, and involve rights and duties. These norms regulate the behaviour of the members of institutions, which are based on mutual expectations and have to known of the members. This is to emphasize because the mutual knowing of the rule norms is a particular feature of institutions. This is not valid in other social systems generally.

The collective acceptance theory of sociality has analysed institutions, thereby concluding that:

1. They are norm-governed social practices established by authorized members (government, governing board). Thereby they have strong sanctions regulating the behaviour of members. The main feature of social organizations as organization is that there are power relations based on authority.

2. They confer a new conceptual and social status on some entities and events: For example, members, activities, communication, or objects.

3. They confer a new deontic and status functions on their members. This goes along with the institutionalization of the authority system of communication.

4. They entail as organization social status and role position of their members and a task right system.

5. They do not need I-mode only, but involve ideally we-mode intentionality, that is, the members are committed to conceptual status, which creates the status of the membership and full-blown institutional acting requires we-mode action and communication. Institutional communication without collective commitments and acting in the I-mode are not excluded in structured collectives.

6. Linguistic abilities and skills are required to participate in institutional communication and acts of the members of institutions on a fundamental level.¹⁴

Tuomela’s view of institutions:

1. They are constituted by constructive/performative acceptance, which is a “reflective collective acceptance”,

2. The actions and communications are regulated by norms, making new sorts of behaviour conceptually possible when those do not exist in the pre-institutional situation, and

3. The members have more or less correct beliefs on, for example, the institutional regulations, business company, and school, as functional imperative.

4. One of the main features of the institutions is that they establish social order, which goes along with the satisfaction of the basic needs of people. But conflicts between individual and collective interests are not to exclude in provision institutionalized process, for example, coordination dilemmas on which side of the road people should drive, and situations of full conflict, so say zero-sum situations in game-theoretic terms.¹⁵

Yet the dynamic of social practices and social institutions require a particular analysis. Tuomela and Balzer implement the analysis with a general mathematical model. The model goes along with Giddens (1984) structuration theory. But contrary to Giddens the analysandum of the model are repeated social activities and the maintenance of the relevant social structures by joint we-attitudes. From the model theory results also a scientific exchange to the distributive artificial intelligence research, and the computer simulation of social systems (Giddens 2002), chapter 7).

The application of the collective acceptance model, thus collective intentionality in the form of shared we-mode/we-attitude, basically we-intention and/or we-belief, shows social institutions are collections of position-involving normative regulated structure, which is made causally effective by the group members’ minds, actions, and communications only. It is a constraint of membership that all members have in different extent correct beliefs about the social status function that they perform. It is not required that the members acts be a contribution to the achievement or the renewal of the institution. But it is not disputed that social institutions generate social order (regulations) of communication, solve coordination problems, and collective action problems entailing also conflicts between their members and collective rationality.

9. Membership and Social Ontology

Ontology is the study of what exists, that is, of our ontological commitments. Tuomela is an advocate of a theory of social systems because groups, communication of decisions, authorized advises, cooperation, institutions, and organizations are a part of a social action-communication system and its ontology. It is not ‘lug in’ that he systemizes social practices and collective pattern-governed behaviour by membership in social systems as the basic component of the social domain. Therefore, Tuomela’s philosophy of sociality and social ontology have as a core a sociology of membership (Preyer 2018a, b, c). The status function of membership and its indication are the shared point of view, which is not supposed to be determined by a single person, individualistically. The indication of this status function is observer dependent, in principle.

Searle has made a fruitful distinction between ontological objectivity and subjectivity, and epistemological objectivity and subjectivity (Searle 2010), chapter 1 VII). The social domain is not a part of the physical world. But this does not mean that objective judgements about this domain are not possible. These judgements are observer dependent and the observer is a member of the domain or a participant of the communication process. This goes along with that we-attitudes presuppose a social frame of reference in a strong sense. Therefore, these attitudes presuppose membership and the collective commitments, which are instantiated by members as members and, at the same time, by their public access, that is, the observation of the members in their status functions by their indication. This is the ontological distinction between the social domain and the rest of the nature.

Tuomela’s analysis of the ontology of the social domain is that sociality is “man-made” from the beginning by the performative construction, the self-referential (reflective) feature, and the we-mode. Tuomela’s, Searle’s (1995, 1998, 2001, 2010), and others’ view is that the social domain itself is not a part of the ontology of the physical world, that is, it is not a natural property; it is not made up of atomistic individuals, and is not a regional being, ontologically speaking, but is an artificial being instead. Sociologists would agree with that, in principle. Therefore, it is ontologically subjective. But the domain is epistemological objective because this does not mean it is a fiction because it is to reproduce by objective status functions and their indications, which establish a particular mode of observation of the status function indication. There is a social ontology only if a person functioning as a group member and decision of membership is established in the communication system of the members of social systems. For members as members, it is valid that group reasons override individual reasons (we-moders).

Social groups are not agents or persons in the literal meaning of the words, but the social domain is constituted by group memberships, ontologically speaking. Tuomela makes the conceptual distinction between single person and group member. Therefore, he ascribes also to members the attitudes of the group. But membership units like groups are not to classify as agents or singular entities. This does not exclude a metaphorical speech about collectives, which is not to reject. The ontological status of a social group is to characterize thereby that groups as social systems are supervenient on their members.¹⁶ Tuomela (2013, pp. 91) roughly speaking on “supervenience”. It means a necessary condition: A group property changes with the group members’ individual properties or their interrelations. This goes along with our everyday life understanding because we make in ordinary live the assumption that members of groups have a belief, intentions, desire, and do something. This means that members of social groups by their membership status have correspondent attitudes, which they share or not with other members. Tuomela analyses group attitudes in relation to the concept of collective acceptance. This is the connection to the analysis of the group authorities and the cause of the formation of attitudes and action of groups because, through the function of these authorities, the basis concept of having an attitude as a member of groups is introduced. These attitudes are based on group reasons (Tuomela 2007), chapter 6). Tuomela (2007, pp. 19-21) assumes that membership is defined by the acceptance of a group ethos.

The bulletin board view of collective acceptance and the positional account of group attitudes play together.¹⁷ Membership is determined by status functions (positions) and roles, and the members take a commitment to the ethos of the group. (Searle: local background) Thereby the both accounts are connected by the authority order and the distinction between operative (autonomous) and non-operative (non-autonomous) membership units in social systems and corporations (formal organizations). This is linked with full-blown cooperative actions based on “group-mode/we-mode”, and “acting together” as maintenance function of institution.

Tuomela’s turn to the collective acceptance theory of sociality goes along with the sociology of membership because the collective acceptance is to specify to the members of the social domain and its differentiation in membership-systems with open (permeable) and closed (formal regulated, written down) membership conditions. There are social groups within these systems, which are put together because of different motives and reasons, or are formed because of formal requirements of organizations. The ontology of a collective agent is always something which is not given in nature, also not in animal behaviour, but is instantiated by membership status functions, decisions, and authority instances by operative and autonomous groups. The decision of membership is the status function elementary operation thereby there is a social domain only, which is to reproduce in the time dimension continuously. This presupposes an observer as an instance that has, as authority to its disposal, normative power to establish for all members’ social expectations and sanctions. All members of social systems are equal as members, but they are distinct by status functions, which form cooperation, decisions, and collective goals. Therefore, we conclude that Tuomela’s social ontology is neither an individualistic nor collectivistic ontology.

10. Conclusions

Summing up, the core of Tuomela’s philosophy of sociality as a sociology of membership is that collective acceptance is analysed by the appropriate relevant we-mode of members as members of social systems. Analysans is the distinction between the “we mode” thinking, feeling, and acting, its conceptual and functional relation to the “I-mode” thinking and feeling as private person. Collective intentionality is connected with one of the main features of cooperation. Cooperations as acting together themselves are based on “group-mode/we-mode” as constitutive condition. They are a general feature of sociality and are linked to the maintenance of institutions. The basic feature of social institutions is that they are collections of position-involving normative regulated social systems. But the social domain is also determinated by asymmetrical conditions of interaction. The operative and non-operative members (autonomous and non-autonomous groups), with respect to the authority, make us split the social domain into social groups, cooperations, institutions, and formal organizations. But the group members shared we-mode concept has a partial priority in the domain of sociality, which is not reducible to the I-mode concept because we-mode collective intentionality is “ultimately needed for understanding social life”.¹⁸ This is a weak resonance of a moderate Durkheimian sociology. It is to conclude that the membership decision and its specification of roles, status functions, their tasks, obligations, and commitments are the basic decision and selection thereby is constituted by the social domain and its self-observation as such.

From the sociological observation of collective intentionality, it is fruitful to turn to the distinction between member versus non-member of a social domain as leading theoretical distinctions, and what is entailed thereby for the structure and continuation of communication. It is a condition of participation in the social domain that the members work together in different social positions. This contributes to its reproductive success. But the members of the social domain are also fighting against each other, compete on the market of the economic, political, and scientific system; are self-seeking and ambitious. In the meantime, we have evidence enough that the fight and regulation about the control of the flow of resources is an evolutionary universal (Eisenstadt 1995c, pp. 344-345).

The analysis of cooperation is relevant for a sociology of membership because it is a necessary condition of social systems that the members in different social positions are willing to cooperate achieving a joint collective goal. Social systems have enhancement power, and increased selective advantage by the decision of membership resolves problems of cooperative tasks. These are limited negations thereby the social domain is capable of learning in general. If we make the assumption that the analysis of collective acceptance among the members of the social domain and the group attitudes are the conceptual and core of the philosophy of sociality, then we should conclude that all social systems are self-constituted by the decision of membership and its implementation. This makes their efficiency clear, but also their fragility. The turn to a membership sociology will also result in a consequential problem in Tuomela’s more integrated account involving his Collective Acceptance View of social systems because the membership decision is a continuous self-irritation of social systems which programs exclusions from communication systems and disintegration of members. But it is to take into play that his integrative account is differentiated by his analysis of autonomous and non-autonomous groups and corporation. An account of social system is required which shows that nomic and anomic dimensions of the self-constitution of social systems play together in their constitution. When this membership turn works then we have another account of social norms, institutions, cooperation, and the authority order of social systems. More research on social ontology is desirable, so keeping in touch with the founders and the researches that connect to the foundation of the social domain in a continuous manner.

Notes

1. On I- and we-mode, see also, Tuomela 2002, chapter 2, 2007, chapter 2. – 4., 2000, chapter 2. V, 2013, chapter 2, 2012, on Individualism and Collectivism in Social Science, pp. 129-143, 2013, he agrees with Searle about that the we-mode is not to reduce to the I-mode. His claim is to give a stronger conceptual and ontological account, 83-85; on conceptual, explanatory, and ontological reduction, pp. 90-93.

2. The term “corporate agent” as a membership system goes back to the ancient Roman law. The terms universitas, corporatio, and collegium referred to “group persons” (intentional group agents). The member as member of this social system were authorized making agreement entailing promises, which goes along with a “corporate responsibility”. This social system was contrasted with societas (koinoniá/communitas) as a generic sameness of humans distinguished from animals. The town as civitas sive societas civilis (political society) is a system among others and at the same time independent from other social systems. (Tuomela 2013), pp. 233-240, on his “theory of corporation” in the context of Searle’s (2010) new account Tuomela (2013). Corporations have not a physical existence at all. Tuomela gives a detailed analysis of group agents as a functional concept.

3. The prisoner dilemma is a case of a cooperation in a weak sense. This is a I-cooperation with a shared private goal. On I-cooperations, see (Tuomela 2000), chapters 10-12.

4. Tuomela (2007), on Miller 117-119, Schmitt 119-20, 122, Gilbert 121-22, Bratman 100-101, 120-121, 2013, on List and Pettit 53-54, 140-144.

5. Tuomela, Social Ontology, 22.

6. Basic elites, leadership and thereby authoritative decisions are emerged ‘early’ in the evolutionary differentiation of social division of labour and the political function in primitive societies with respect to “ (1) the definition of the main collective goals that can be implemented and determination of their order (institutions in the society); (2) the allocation of prestige, influence, and an authorized use of power and facilities to various groups in the society; (3) distribution of various facilities, benefits, and rights to such groups and individuals”, Eisenstadt 1971, 11.

7. Raz 1975, on power-conferring norms, 85-106, Preyer 2013, on the authority system and the decision of membership, membership and social norms, and normative power, 519-527.

8. Tuomela, Social Ontology, chapter 2, see, especially the detailed account in Appendix 1. See also, with some variations and qualifications explained in the appendix.

9. Tuomela, 2013, see, the “spy example”, 51, for an illustration of explanation of group member action.

10. Tuomela (2000), on the game theory chapter 7. The power of this account is limited because it is not to analyse—without artificiality—concrete goals, for example, the edition of a book and the building of a bridge, and the commitment, which the participants accept. But limited is not useless, Tuomela (2000), chapter 7 I, II, he uses game theoretic notions in Artificial Intelligence describing problems of cooperation and he analyses a concept of “strongly rational cooperation”.

11. On factors of social control, for example, Thibuat and Kelly (1959), Wilson and Bixenstine 1962, pp. 92-102, (Tuomela 2000), Chapter 8.

12. Tuomela 2000, Capt. 4, see also 27. On a technical analysis with game theoretical structures, Capt. 9, on cooperation and conflict in the context of the dilemma of collective action, the conflict between individual and collective rationality, for example, the prisoner dilemma, Capt. 10.

13. Tuomela (2002), chapters 6, 7.9., 2007, chapter 8, 2000, chapter 6, 2013, chapter 8, on Searle’s concept of institution, (Tuomela 2002), chapter 6.4., see also, (Tuomela 2009), pp. 272-306, on Searle V. It is to mention in this context of theorizing that in the sociological theory since the 1950s years is analysis of institutions switched to the processes of institutionalization (institution building), see, on a summary with respect to the charismatic dimension, Eisenstadt 1995a, b, pp. 86-105, pp. 167-201.

14. Tuomela 2013, on social institutions as a normative action system and their basic functions, 229, on Searle’s analysis of status function declaration and institutions, 233-240.

15. Tuomela 2013, these dilemmas are solute often be sanction norms, 224.

16. (Tuomela 2013), roughly speaking on “supervenience”. It means as a necessary condition: a group property changes with the group members, individual properties or their interrelations, 91.

17. Tuomela 2007, on the entitative (singular entity view) and non-entitative ontological characterization of groups, 145-48, 2013, on methodological individualism, 9-13. The status (position) function and its relation to social roles is an old sociological categories going back to Linton 1936.

18. Tuomela 2007, VII. It is to mention that this is near by Habermas 1981 because he makes the basic distinction between acts formed a mutual agreement (Verständigung) based on an argumentative reasoning and acts of exertion influence in strategic action systems.