The Conditions of Collectivity: Joint Commitment and the Shared Norms of Membership

Part of the Studies in the Philosophy of Sociality book series (SIPS, volume 2)


Collective intentionality is one of the most fundamental notions in social ontology. However, it is often thought to refer to a capacity which does not presuppose the existence of any other social facts. This chapter critically examines this view from the perspective of one specific theory of collective intentionality, the theory of Margaret Gilbert. On the basis of Gilbert’s arguments, the chapter claims that collective intentionality is a highly contingent achievement of complex social practices and, thus, not a basic social phenomenon. The argument proceeds in three steps. First, Gilbert’s thesis that certain kinds of collective intentionality presuppose joint normative commitments is introduced. Second, it is argued that, on this view, individual commitments can only constitute the relevant kinds of collective intentional states if there are socially shared “principles of membership” that connect the force of individual commitments to a shared content. Third, it is shown that strong collective intentionality depends on the practical acceptance of shared norms and on the establishment of authority relations through mutual recognition.


collective intentionality joint commitment membership Margaret Gilbert Robert Brandom 

1 Introduction

A large part of contemporary debates about social ontology is primarily concerned with two questions about collective intentional states: (i) are there any collective intentional states, i.e., do we have to refer to collective intentionality in the best possible account of social reality; and, (ii) if we have to, what exactly are collective intentional states and how are they connected to individual intentionality? Although these questions are, of course, reasonable and important, they tend to encourage a view of collective intentionality in abstraction from other social phenomena. The most extreme result of looking at the question of collective intentionality from this perspective is the attempt to understand it as some brute “given”, as an ability which humans possess naturally (Searle 1995, pp. 23ff.), or as an ability which is fundamental in at least the sense that it does not require any further social preconditions to be met in order for (fully developed) individuals to be able to attempt to exercise it.

As an alternative to this view, it can be argued that ascribing collective intentional states to groups of persons is part of a complex general social practice of ascribing to the verbal and non-verbal behaviour of groups a specific status, namely the status of this behaviour counting as intentionally governed by collective beliefs, desires or other intentional states. If we adopt this perspective, other questions become more important: what conditions must a society fulfil in order for its members to be able to describe themselves and others in intentional terms? And are there any specific conditions to be met for them to be able to describe themselves as being engaged (or trying to engage) in collective activities and as having collective beliefs in a strong sense?

In other words, by adopting this perspective, it might turn out that there are very specific and demanding social conditions which have to obtain in a community of speakers for them to be able to attribute certain collective intentional states to each other. To appreciate this point, at least a certain form of collective intentionality—not only in individual instances but in regard to its very possibility—should be understood as an achievement, as something which not only needs to be examined with respect to how it can be explained, but also with respect to how it is possible at all.

To provide some arguments in favour of this perspective, it is helpful to focus on the specific normative role that ascriptions of collective intentional states play in social life. For this purpose, the following discussion focuses on Margaret Gilbert’s account of collective intentional phenomena (Gilbert 1989). According to Gilbert’s approach, collective intentional states necessarily involve joint normative commitments, either of a plural subject or of the members of some group.

This chapter will argue that Gilbert’s concept of joint commitment, which she takes to be central for at least certain kinds of collective intentionality (Sect. 2), cannot be understood without noting that the very possibility of the relevant kind of joint commitment depends on the social institution of inferential rules—that is, rules that govern what follows from such joint commitments for individuals—which give joint commitments practical significance. This claim is justified by discussing two objections against Gilbert’s account of joint commitment (Sect. 3). The chapter then shows that if one accepts her description of joint commitments, the relevant collective intentional states cannot be understood in abstraction from the social practices in which these rules are instituted (Sects. 4 and 5). In particular, the chapter argues that the kinds of social practices that must be presupposed to make sense of strong forms of collective intentionality should be understood as instances of those inferential practices which Robert Brandom analyses in his “normative pragmatics” (Sect. 6). Finally, it is suggested that this amounts to understanding strong collective intentionality as dependent on an underlying structure of mutual recognition (Sect. 7).

2 The Centrality of Normative Commitments for Strong Collective Intentional States

Descriptions of social phenomena in terms of collective intentional states try to capture those features of social reality which cannot be explained by mere descriptions of how persons influence each other’s individual intentional states, but only in virtue of them sharing certain intentions, attitudes or beliefs in a strong sense.

Of course, you and I can share an intention to push a car up a hill individually, in the sense that we both have the intention individually. We might even know of each other’s intentions, might individually intend to make the intention of the other person effective, and might recognise that the effectiveness of the other person’s intention is a condition for the realisation of our own intention. However, according to some accounts in contemporary social ontology (Gilbert 1989, pp. 161f.), there is still a difference between such cases and the case where, in an emphatic sense, we do something as our project. This difference is spelled out as follows. For a goal to become “our” goal in the strong sense of being the object of a collective intention, each of us must not only be individually bound to our (suitably related) individual intentions, we must be bound to a goal as a group. We do not need to look as far as to patriotism or class solidarity to find examples of this phenomenon: if a friend and I jointly intend to push a car up the hill, a possible failure of either of us (or even a mutually known simultaneous failure of both of us) to act on this intention would not just be a failure to realise our individual intentions, but would amount to us—collectively—failing a joint project in the sense of an enterprise with a joint goal (Gilbert 1989, pp. 163, 421ff.). Furthermore, insofar as collective projects in this strong sense also establish mutual obligations between group members or shared normative commitments to act in a certain way, these obligations and commitments cannot be understood as just individual obligations, but are rather obligations that we have together (in other words, obligations of the group as subject), and joint commitments. In participating in a collective project, one could say, we jointly accept a goal together and thus are collectively responsible for its attainment.

This strong notion of collective intentionality, which has mainly been developed by Margaret Gilbert and Raimo Tuomela, presents a challenge to individualist theories of social ontology. Since individualist theories often assume that collective intentionality supervenes on individual intentional states and phenomena on the one hand, and on certain non-normative relations between these states or between the individual agents on the other hand, they are at a loss to explain how the collectivised normativity described above—beyond the normativity involved in individual intentional attitudes—comes into play (cf. for example Bratman 1999).

There is more than one way to describe the normative features of collective intentional states: Raimo Tuomela, for example, identifies the “reproachability” induced by the existence of joint commitments as one of the core conceptual dimensions of strong collective notions (Tuomela 2007, p. 37). A very similar notion of joint commitment is central to Margaret Gilbert’s theory on which this paper will focus.1 The most concise summary of Gilbert’s view is that “genuine [acting] together involves rights and duties that are something other than moral rights and duties” (1989, p. 162), and that the same feature holds for collective beliefs. Putting the question aside whether all kinds of collective intentionality depend on such joint commitments, she at least claims that certain kinds of collective intentional states—which I will call “strongly collective intentional states”—involve normative commitments. According to Gilbert, such commitments are clearly to be distinguished from both general moral duties and, more importantly, from the “personal” normative commitments we can undertake via individual intentionality.

As evidence for this claim, Gilbert cites the fact that we intuitively feel entitled to “punitive criticism” of fellow members of those groups with which we share collective beliefs and intentions. Whenever those members act inconsistently with collectively shared attitudes, we feel entitled to rebuke them (1987, 1990). We take such reactions to be appropriate, according to Gilbert, because the joint acceptance of a proposition or goal necessarily involves a joint commitment as a body or as one (Gilbert 1996, p. 8). A shared commitment is, furthermore, not a genuinely joint commitment if the parties to the commitment are only committed individually—which would imply that they could leave or abrogate the commitment as individuals—but only if they are committed in such a way that they can only rescind the commitment together (Gilbert 1999, pp. 152f.). Because this relation of the commitment to the group as a whole is typical for strongly collective intentional states, Gilbert assumes that such joint commitments are necessary to make groups of individuals into strongly collective subjects (“plural subjects”, cf. Gilbert 1989, p. 163), or into group agents in a sense that supports diverse explanatory projects.

3 Two Objections against Joint Commitment Accounts

If we are interested in the conditions that make it possible to ascribe intentional states to groups in a way that is not just metaphorical, Gilbert’s claim that strongly collective intentional states must be understood as necessarily involving joint commitments, or commitments of plural subjects, is a useful starting point. But as two kinds of objections raised to this claim show, it is defensible only if we understand such commitments as being made possible by the existence of a complex social structure. Firstly, the idea of a joint commitment seems to leave us with a paradox of constitution in regard to its subject, since joint commitments are taken to be at the same time constitutive of plural subjects and a state of already given plural subjects. Secondly, it is unclear how the collective obligations deriving from joint commitments can become effective obligations of individual group members.

In regard to the first objection, the obvious dilemma we face when talking about joint commitments is that the subject of a joint commitment does not exist independently of the commitment itself. Joint commitments are not commitments of individuals, but are rather commitments of “plural subjects”. However, plural subjects do not exist independently of their relevant joint commitments but are rather constituted by such commitments (Gilbert 1996, p. 348). Although it can subsequently undertake additional joint commitments, the emergence of a plural subject in Gilbert’s sense requires an initial joint commitment. But who undertakes this initial joint commitment? A plural subject cannot emerge through individuals undertaking a joint commitment, since, by definition, individuals are the wrong kind of subject to undertake such commitments. Thus, plural subjects and joint commitments must emerge simultaneously, as part of the same process. But how do we have to understand this process?

According to Gilbert, a joint commitment emerges whenever two or more persons express their readiness to be jointly committed to each other and their doing so becomes common knowledge among them (Gilbert 2002a, p. 65, 2006, pp. 138f.). Yet this explanation is somewhat puzzling: if an individual expresses a readiness to be jointly committed, she might thereby commit herself individually to undertake (a part of) a joint commitment. If there is common knowledge about such commitments within a group, then she can also acquire information about the respective individual commitments of others. But this does not explain what she has committed herself to be a part of. In other words, if one expresses one’s readiness to be jointly committed, one is individually committed to undertake a joint commitment. But one is not yet jointly committed. One is only jointly committed if one is not only ready but actually undertakes a joint commitment and does not fail to do what one has expressed oneself to be ready to do. The readiness to join a joint commitment, in other words, cannot itself already be the joint commitment, for the joint commitment is the commitment of a plural subject whereas the readiness expresses the commitment of an individual subject. But if it is necessary for a plural subject already to exist in order for there to be a joint commitment (Gilbert 2002b), then the account of individual readiness does not provide a satisfying explanation of the emergence of plural subjects.

The second objection to Gilbert’s account starts from the observation that, in her view, joint commitments become effective in governing the behaviour of social agents by way of individual commitments that “flow” from joint commitments. Such derived individual commitments are not “personal” (i.e. “normal”, non-derived) commitments of an individual person because they have a special feature that they do not share with non-derived personal commitments; they cannot be abrogated unilaterally (Gilbert 1999, pp. 145ff.). Gilbert explains this feature by pointing out that joint commitments are commitments of a group and that, consequently, an individual is not in a position to abrogate them because they are not her own commitments. Accordingly, by virtue of being a member of a plural subject, an individual is also not in a position to abrogate those of her individual commitments that are derived from joint commitments of this group.

Gilbert thus introduces individual commitments that are derived from joint commitment as a special class:

Though it is clearly not appropriate to speak of the parts of a joint commitment, each of the parties to a joint commitment is committed through it. It is therefore tempting to refer to the parties’ ‘individual commitments’. (1996, p. 11)

And similarly:

Though no one of them independently constitutes the subject of their joint commitment, each of the committed persons is committed through it. Each is bound at least in the way a personal intention binds its subject: each has sufficient reason to act in a certain way. Bearing this in mind, one might speak of the parties’ derived or associated ‘individual’ commitments. (2006, p. 136)

Kenneth Shockley (2004) argues convincingly that this explanation puts Gilbert’s account in a dilemma. If joint commitments are commitments of a plural subject which can be distinguished from the corresponding set of individuals insofar as this subject can have obligations which the individuals do not have as a mere set, it is necessary to explain how individual commitments “flow” from this joint commitment. Such an explanation is required because individual agents seem to have no obvious reason to adhere to the commitments of another agent (the group). In other words, how can derived commitments be at once the individual’s own commitments (insofar as she has reason to let her actions be guided by them) and not her own commitments (insofar as she is not in a position to abrogate them)?

Furthermore, not only does the joint commitment of a group seem to fail to provide individual members with reasons to conform to them, it also fails to put individuals in a position to rebuke their interaction partners for non-compliance, because, as individuals, they are the wrong sort of subject to take the joint commitment as a reason for action. In absence of a convincing explanation of how joint commitments can become relevant for the individuals which are not their subjects, the introduction of joint commitments cannot do the work required in Gilbert’s account.

Intuitively, Shockley argues, we would think that if persons are party to the commitments of groups, they are so in virtue of some aspect of their membership in those groups. What we need to find then are facts establishing the right kind of membership that can provide the necessary link between joint and individual commitments, and that can explain how individuals can come to be bound by the commitments of the group.

In other words, we must find facts which make it the case that individuals have individual commitments derived from the commitments of the group. Shockley (2004, p. 552) calls this condition a “principle of membership” (M):

(M): If Group A is (jointly) committed to X and B is a member of Group A, then B is (individually) committed to Y.

(M) is a different way of describing what has to be the case in order for Gilbert’s account of joint commitment to be plausible. If there is anything that can make this principle true, it has to be the relation of membership which binds individuals to groups. But what kind of relation does the truth of (M) depend on? (M) must either be true in virtue of a relation between the individual and other individuals or in virtue of a relation between the individual and the group as plural subject.

Shockley subsequently argues that Gilbert’s plural subjecthood account rests on an equivocation between these two possible truth-conditions. While (M) might seem plausible because we hold each other responsible as individuals for failing to fulfil shared commitments, the other possible relation, described by the part-whole sense of (M), is—according to Shockley—not only mysterious but unnecessary to understand our intuitions about the truth of (M). However, Gilbert seems to rely on this second interpretation of the truth conditions of (M) in order to justify her introduction of joint commitments as commitments of “plural subjects”.

If we come to accept the plausibility of the first interpretation of (M), Shockley argues, we can relinquish the talk of plural subjects and joint commitments and acknowledge that the normative character of collective intentionality depends on non-collective, inter-personal commitments alone. This solution, however, would also mean giving up Gilbert’s plural subject account of collective intentionality.

4 Two Kinds of Commitment

It remains to be determined whether the two objections just described can be answered. The following arguments will try to show that there is at least one way to think about the social preconditions for the possibility of plural subjecthood that makes sense of the thought that the truth of the principle of membership might be given by a relation between the individual and the group in a non-mysterious way.

Even though Shockley convincingly argues that it is unnecessary to introduce the concept of a plural subject in order to understand the force of the interrelated commitments of group members, this argument does not show that we can dispense with the concept of joint commitment in analysing collective intentionality. Even though we can and should understand the force of the basic commitments in virtue of which members of groups hold each other accountable—and on which collective intentionality consequently supervenes—as established by commitments between (individual) persons, an analysis of the content of these commitments may still license or even require the concept of a plural subject (for the distinction between normative force and normative content, see Brandom 2009, p. 71). In other words, individual-individual relations may be sufficient to establish that group members are committed to each other, but these relations might be insufficient to fully understand to what they are thereby committed.

In order to appreciate this point, it is useful to distinguish between two possible ways of analysing collective intentional states in terms of individual commitments. The first option is to analyse a collective intention to do A as a joint commitment to do A, which is in turn constituted by individual commitments to do A between group members, possibly in combination with other individual commitments. On this account, joint commitments are at least partly constituted by individual commitments with the same content (or at least with a content that can be derived from the content of the joint commitment by understanding what the group has committed itself to). On this account, whenever a group is collectively committed to doing A or believing B, and one understands what A and B mean, one does not need any further knowledge or interpretation to know what the individual members have committed themselves to via their membership in the group.

The second option is to analyse a joint commitment to do A in terms of individual commitments with a different content. In this case, the question of what individuals are individually committed to as members of a joint commitment to do A cannot, in normal cases, be answered without engaging in some interpretation of what it means for them to collectively do A or believe B. On this account, a joint commitment to do A may essentially involve distinct individual commitments to do C. In most cases, a joint commitment of a group (“to take a walk”) in some specific social context will involve individual commitments to accept criticism for behaviour that violates some shared standard of social behaviour which may or may not be derived from the object of the joint commitment. A joint commitment to take a walk together might, in some social context, involve distributively shared commitments to wear hats, because there is a socially shared rule “taking a walk with others is incompatible with not wearing a hat”. This can be true even in cases in which this standard is not entailed by the concept of taking a walk.2 While we may then perfectly understand what the group has jointly committed itself to, we might—according to this understanding—not know the full extent of what the individual members are individually committed to (if, for example, we are not familiar with the rule that the membership in a group which is jointly taking a walk is incompatible with not wearing a hat or that persons become open to criticism if they do not wear hats while taking a walk with others).3

Thus, the content of a joint commitment does not straightforwardly entail the content of individual commitments, even if the binding force of the joint commitment can be fully explained by a particular structure of individual commitments.

Rather, the content of the individual commitments is only indirectly connected to the content of the joint commitment through a shared understanding of the meaning of the joint commitment. So, if we want to know what a group collectively intending to take a walk entails for its members, we have to take account of it being an open question (within some boundaries) what its members are individually committed to—and this holds true even if we understand perfectly what the group is collectively committed to.

On this second account of the connection between individual and joint commitment, the concept of a plural subject is irreplaceable because the individual commitments involved in a joint commitment can only be understood in their relation to a joint commitment with some content of its own. The reference to this joint commitment cannot be eliminated and replaced with a reference to individual commitments. In this case, it must be the commitment of another subject, namely, the group.4

As Gilbert’s walking-together example suggests (Gilbert 1990), the content of this connection between individual and joint commitment is highly contingent: in different social settings or in different cultures, we might expect different standards of what is appropriate when walking together, and consequently of what minimal set of individual commitments is constitutive of two people counting as collectively taking a walk together. In general, we can only say that whenever there are individual commitments between persons to criticise and accept criticism according to some standard that is inferentially connected to a collective action type, they fulfil one of the conditions for attributing to them a joint commitment to the content of the relevant collective intentional state, and thereby for an ascription of the collective intentional state to a plural subject constituted by these persons.

5 The Social Constitution of Plural Subjects

Having answered one of the two objections to Gilbert’s account of collective intentionality, the argument for the claim that strongly collective phenomena are not basic (if they have the normative structure that Gilbert describes), but depend on complex social preconditions, can be spelled out. Plural subject phenomena involve two types of commitments whose contents are not analytically connected (in the sense that there is no open question whether the connection is true), namely: (1) the individual-individual commitments of group members; and (2) the joint commitment of the group as a whole. The existence of the appropriate set of elements of the first type is constitutive for the force (and, thus, the very existence) of the second, but the content of the individual commitments between the members involved in the first element is to be fully understood only in relation to the (separate and non-trivially related) content of the joint commitment of the group as a whole. If the content of individual and collective commitments is independent in this way, and if some individual commitments with some content constitute the force of any collective commitment, then the relation between specific individual and joint commitments cannot be assumed to be given a priori, but must rather be understood as socially created.

Taking note of the significance of the socially shared standards that regulate what joint commitments entail for the commitments of individuals helps us to better understand the role of the “principle of membership”. An analysis of the membership of individuals in a group requires both, an understanding of how individual commitments acquire their content from a shared interpretation of what membership in a group with a specific joint intention requires, and an understanding of how these individual commitments, if they are undertaken, establish the force of the joint commitment and constitute the plural subject. While the interpretation of the “principle of membership” given above is not supposed to spell out the full significance of the relation of membership in all its aspects, it can still serve the more limited function of establishing a relation between joint and individual commitments. To do so, it must express the (contingent, non-analytical, socially established) connection between individual and joint commitment in both these dimensions: it is only possible for a group of individuals to form a plural subject with this commitment, if there is some socially accepted principle of membership which specifies the connection between some specific individual commitments and some specific joint commitment.

To put it more technically, principles of membership can be described as parts of constitutive rules (cf. Searle 1995, p. 28) that define what kinds of social groups with which properties count as plural subjects exhibiting which collective intentional state or attitude. The general form of such a constitutive rule is:
  • A group of persons with properties Y counts as a plural subject with the strongly collective intentional state or attitude S in context C.

Because principles of membership as discussed here do not necessarily exhaust the properties and context that are constitutive for plural subjects, the fact that such principles are accepted should only be taken as a necessary condition for their constitution. Thus, we get at least a necessary part of the relevant constitutive rule:
  • A group of persons counts as a plural subject with the strongly collective intentional state or attitude S only if it displays a structure of mutual interlocking commitments to accept criticism and criticise each other according to some set of normative standards N.

This general rule is only a “template” for the relevant parts of more specific constitutive rules: each socially instituted collective action type or attitude type depends on the social acceptance of some concrete constitutive rule, that is, on a socially accepted inferential connection between some N and some S in regard to plural subjects.

As these parts of constitutive rules specify that for any plural subject, each of its members must accept their respective part of a structure of individual commitments obliging them to accept criticism and to criticise according to some set of standards of what counts as correct for members of this plural subject, these rule parts thereby also acquire a regulative-normative dimension. In other words, the rule parts connect the status of membership to the individual acceptance of certain obligations and entitlements.

If Gilbert’s account is reformulated in this manner, it can answer Shockley’s objection: even if joint commitments supervene on a structure of individual commitments in regard to their force, we still need plural subjects, because we can understand the content of the individual commitments only in terms of plural subjecthood. That is, the normative status instituted by the principle of membership can only be understood as being the status of a group that is committed as a plural intentional subject.5

This answer also lays the foundation for the beginnings of a solution to the first problem, the problem of the constitution of plural subjects. If the acceptance of a principle of membership is a precondition of plural subjecthood, and insofar as the principle of membership specifies that a specific structure of interrelated individual commitments constitutes a plural subject, individuals just displaying their readiness to enter into a joint commitment is insufficient to constitute this commitment. Rather, each individual must independently undertake the required individual commitments in order to create a specific plural subject, but they can do so only given a background of an already socially accepted principle of membership. Consequently, the fact that all initial joint commitments are created by the individual commitments of their prospective members does not mean that the content of these joint commitments is derivable from the individual intentions to undertake such individual commitments. Rather, it means that undertaking individual commitments is a precondition of some joint commitment acquiring normative force, the content of which can only be inferred if one knows the relevant socially shared background rules. That means that the joint commitment can acquire its content only on the condition that there is a socially accepted constitutive rule which specifies that a group with some structure of individual commitments counts as a plural subject with commitments of its own.

Thus, a specific structure of individual commitments is necessary for the constitution of a plural subject and for its having a certain joint commitment. This structure can also plausibly be said to be constitutive for something further; namely, for the type of relation between the individual members and the plural subject that is sufficient to establish an obligation of the individual members towards the plural subject.

In other words, we can neither deny the significance of the individual-individual nor the significance of the individual-collective relationship, although the normative individual-collective relationship (the commitment of the members to the plural subject) derives its normative force from the standards that are instituted by the individual-individual relations.

6 An Intersubjectivist Model of Joint Commitments

To summarise the argument so far, there is at least one possible answer to both objections that have been made against the joint commitment theory of Gilbert. It involves the claim that the relation between individual commitments and strong plural subjecthood can only be usefully understood if we assume that there is a socially accepted background of principles of membership that specify which structures of individual commitments can legitimately be counted as establishing strong plural subjects of some kind.

What then are some of the social preconditions which must obtain in a community in order for its members to be able to enter into collective intentional states? First of all, there must be some established standards that define the content of the individual commitments that form the structure constituting a plural subject. In the normal case, these will be commitments to criticise and accept criticism for the violation of the socially accepted inferential properties of a given collective intentional state. These rules define, for example, the socially shared understanding of which individual behaviour is appropriate for you if you are member of a group that has the intention to take a walk together, or if you are a member of a group that collectively believes that a poem is beautiful. Without such standards, the relevant principle of group membership will be empty.

However, we should not understand these standards as themselves instituted by some sort of collective or individual belief. This would either, in the collective case, lead into an infinite regress or, in the case of individual beliefs, undermine the justification for the claim that joint commitment has an independent normative force. We should rather understand the background standards as a part of the implicit “institutional fabric” of a community.6

Understanding background standards in this way means, however, that strongly collective intentionality can not be understood as a simple, basic building block of institutional reality. It rather depends on the existence of other social facts, particularly on the implicitly shared acceptance of relevant normative standards. One necessary precondition of strongly collective intentionality is therefore a social practice in which certain normative standards are implicitly instituted.

But how should we understand this precondition? Fortunately, the extended discussion of the Wittgensteinian problem of rule-following has already produced useful ways of examining this question, of which Robert Brandom’s (1994) normative pragmatism is perhaps the best known. In the case of individual intentionality Brandom argues that we can understand the ability of rational agents to follow explicit rules, that is, to regulate their behaviour according to normative constraints, only if we presuppose forms of shared, implicit, practically instituted propriety (1994, pp. 18ff.). By taking this implicitly governed, practically instituted ability to correctly undertake, evaluate and ascribe normative commitments as fundamental, we can, according to Brandom, get a grip on the issue of intentionality: we have to understand the ascription of intentional content as the attribution of a specific normative status, of a commitment to a position in the “space of reasons”, the significance of which is defined by collectively instituted inferential rules that lay out the further entitlements and commitments connected to this position.

Without going into the details of his model, I propose to extend Brandom’s programme of understanding intentionality as the ascription of a normative status from individual to (strongly) collective intentionality. Just like we can treat an individual as an intentional agent by attributing to her intentional states that can be spelled out as the entitlements and obligations that follow from the status of having undertaken a certain kind of commitment, so we can also treat collectives (in the strong sense) as intentional agents by taking them to be able to undertake and successfully act on joint commitments.7

There is, however, an important difference between the individual and the collective case. In the individual case, we attribute, for example, a belief to a specific person by ascribing a complex normative status to the same person. Roughly, we might say that we take her behaviour to be governed by an obligation that does forbid her from assenting to any proposition that is incompatible with the attributed belief-content, or to anything that entails such a proposition, an obligation to provide reasons for her belief when challenged, an entitlement to commit herself to the consequences of her belief (and so on). According to Brandom, even the very idea of a subject can be understood as the reference point for the ascription of such normative commitments which need to be consistently integrated with one another (Brandom 2009, pp. 48f.).

The collective case, in contrast, both establishes a new kind of subject position (a plural subject) and builds upon an already established individual subjectivity. When there is a strongly collective intentional state, there is, similarly to the case of individual subjects, a commitment attributed to the (plural) subject of the intention or belief, as being a subject with a “mind of its own”. This plural subject is taken to be governed by norms prescribing, for example, that the inferential consequences of the relevant beliefs and intentions be rationally consistent. There is, however, another set of norms. In the case of strongly collective intentionality, there are also norms that specify commitments of the individual members of the plural subject, prescribing what they ought to do or to believe in their institutional role as group members.

To take up Gilbert’s example once again, if two persons jointly intend to take a walk together, there is one set of norms that apply to them jointly: They cannot without any further reason stay at home and play chess without inviting doubts about their (perhaps professed) joint intention. But there is also a second set of norms that apply to them individually. As individuals, they must not, for example, run away from the other person without a reason, and they have to conform to a huge number of culturally and socially implicit requirements to count as members of a group that intends to jointly take a walk. These two sets of norms have not only different content but also different subjects, whereby the subject of the first set is constituted by the individuals being in social relations which are defined by the second set.

It is essential to distinguish between these two sets of norms in order to avoid the pitfalls of either reductive individualism or strong collectivism. If we assume that the inferential properties of the social status attributed in the course of the ascription of collective intentional states and the resulting commitments are strictly congruent with their equivalents in the individual case, then we have either to attribute all relevant commitments to the individual members of the group, rendering the talk of a plural subject unnecessary but also missing the normative consequences of strongly collective intentionality, or we have to attribute all relevant commitments only to the plural subject and not to the individual members, making the reality and effectiveness of joint commitments mysterious.

The question which has been asked at the beginning of this chapter can now receive a tentative answer: If we do not assume that strongly collective intentional states are to be explained as simple building blocks of institutional reality, but rather as an achievement of practices of the ascription of normative statuses that turn out to be rather complex, we can say more about their social preconditions. Strongly collective intentional states presuppose—as a necessary condition, though certainly not as a full explanation—the existence of certain social practices of attributing both individual and joint commitments. More specifically, they presuppose social practices which incorporate shared rules and understandings about the connection between collective and individual commitments, and shared rules and understandings concerning the conditions of membership in plural subjects with certain types of strongly collective intentional states. This formulation does not in itself tell us everything we need to know about collective intentional states, as it is compatible with a wide range of theories. However, it does establish constraints that any plausible theory of strongly collective intentional states must satisfy.

7 Recognition as the Condition of Possibility of Collective Commitment and Collective Self-Governance

If collective intentionality presupposes socially instituted implicit principles of membership, then collective intentionality is not a simply given, free-standing feature of social reality. The ability of individual persons to enter into joint commitments, their ability to be subjects of strongly collective intentional states, depends on the existence of shared inferential rules—or principles of membership—which establish a shared understanding of the connection between collective and individual commitments. The existence of such an understanding is—as I have attempted to show—constitutive for the possibility of strongly collective intentional states.

To present a slightly more concrete picture, it might be appropriate to comment further on the social preconditions of strong plural subjecthood—even if these remarks are somewhat speculative.

If the practical proprieties which establish shared inferential rules are necessary preconditions for collective intentional states, then the practical interactions (assessments, sanctions and so on) which establish their normativity must be underwritten by some kind of authority for these rules to be in effect in any meaningful sense. Thus, being strongly jointly committed to a goal requires that the participants accept the authority of their co-members regarding their conformity to those rules which are the inferential consequences of the collective acceptance of that goal.

The ability to enter into a strongly joint commitment thus presupposes an attitude of the potential participants towards their prospective co-participants that attributes to them the entitlement to judge the behaviour of their fellow (potential or actual) group members and, if necessary, to respond to it with criticism. But this authority ascription only covers the application (again, the force) of certain shared rules and not their meaning (content). One does not have to accept the reactions of others as the last word on the meaning of the shared rules. Rather, the meaning of these rules and standards depends on the acceptance of the relevant principles of membership, which in turn get their meaning from more widely shared understandings of a linguistic or cultural community. These principles therefore involve acceptance of the authority of a large number of other persons (the relevant community) which normally includes the more limited number of those persons with whom one enters into a joint commitment. Thus, we can understand normative commitments in both the narrow sense in which they are necessary for collective intentionality and in the wider sense in which they are fundamental to the shared understanding of membership rules as presupposing structures of mutual authority ascription.

These structures of mutual authority ascription are recursive and defeasible. They are recursive insofar as being a co-member of a collective intentional state not only requires me to authorise you to evaluate my actions according to some rule which we both accept, and which defines what follows from our being jointly committed; I must also accept your evaluation of my interpretation of that rule according to some further linguistic or interpretative rules that we both share with a larger community. The authority is defeasible insofar as the acceptance of a person’s authority—her recognition by others as a fellow member of a “we”—accords her not an indefeasible and absolute but only a standard authority, which is usually kept in check by rules governing exceptions. Being party to a joint commitment as well as being a member of a social practice entails having to accept evaluations by co-members, but this obligation can always be overridden by an a priori unspecified number of exceptions. Strongly collective intentionality thus supervenes on structures of pragmatic, defeasible authority ascription which can be called—in line with the neo-Hegelian theory of language and mind proposed by Brandom, Pippin and others—“mutual recognition”.

As the individual preconditions of joint commitments are always only intelligible in their embedding into this whole structure of recognition, the theory of collective intentionality cannot be reduced to a project concerned only with an analysis of the phenomenal properties of collective intentional states in terms of relations or properties of individuals; rather, it must attempt to achieve an understanding of these properties in the context of systems of social practices. If we understand the possibility of collective intentionality as an achievement of recognitive communities, collective intentionality theory must not be conducted as a mere extension of the philosophy of mind but at least to the same degree as social analysis.


  1. 1.

    I accept Gilbert’s claim to the effect that strong collective intentional states have the relevant normative features. This claim is contested—for an overview about the normativity question, see Tollefsen (2004); for more specific arguments for a normative view, see Meijers (2003).

  2. 2.

    It is implausible to claim that the content of the derived individual commitment can always be straightforwardly derived from the content of the joint commitment without further information about the social context (Gilbert 2006, p. 136). Firstly, there are types of collective commitments that do not have the promotion of a goal as their content, and secondly, the individual commitments flowing from the joint commitment to a goal need not be commitments to promote that goal, although functional constraints usually guarantee that they do not diverge too far.

  3. 3.

    In the case of an individual intention, fully understanding a person’s intention entails understanding what this intention commits her to. In the collective case, however, even a full understanding of the collective intention leaves open the further question of what individual commitments one has to accept to count as a member of the relevant group.

  4. 4.

    As Philip Pettit (2003) notes, we can attribute to groups minds of their own if they collectivise reason in an appropriate way. Thus, the individual commitments necessary for plural subjecthood could be understood as commitments to collectivise reason without it being necessary for the participants to be personally committed to the result of this process.

  5. 5.

    Of course, nothing keeps us from only talking about the individual membership commitments (as part of an explanatory story, for example). But we will miss the point of these commitments if we do not see that they socially institute the group as a plural subject.

  6. 6.

    This argument connects to a point frequently made by Tuomela: collective action types have to be available for members of a community in order for concrete collective actions to be possible. This issue is also discussed in Stekeler-Weithofer (2002).

  7. 7.

    This approach is very similar to Tollefsen’s (2002) analysis, which describes the ascription of collective intentional states from the perspective of Dennett’s and Davidson’s “interpretationism”.


  1. Brandom, R. 1994. Making it explicit. Cambridge: Harvard University Press.Google Scholar
  2. Brandom, R. 2009. Reason in philosophy. Animating ideas. Cambridge: Belknap/Harvard University Press.Google Scholar
  3. Bratman, M. 1999. Shared intention. In Faces of intention, 109–129. Cambridge: Cambridge University Press.CrossRefGoogle Scholar
  4. Gilbert, M. 1987. Modelling collective belief. Synthese 73(1): 185–204.CrossRefGoogle Scholar
  5. Gilbert, M. 1989. On social facts. London: Routledge.Google Scholar
  6. Gilbert, M. 1990. Walking together: A paradigmatic social phenomenon. Midwest Studies in Philosophy 25: 1–14.CrossRefGoogle Scholar
  7. Gilbert, M. 1996. Living together. Rationality, sociality and obligation. New York: Rowman and Littlefield.Google Scholar
  8. Gilbert, M. 1999. Obligation and joint commitment. Utilitas 11(2): 143–163.CrossRefGoogle Scholar
  9. Gilbert, M. 2002a. Acting together. In Social facts and collective intentionality, ed. G. Meggle, 53–72. Frankfurt a. M: Dr. Hänsel-Hohenhausen AG.Google Scholar
  10. Gilbert, M. 2002b. Considerations on joint commitment: Responses to various comments. In Social facts and collective intentionality, ed. G. Meggle, 73–10. Frankfurt a. M: Dr. Hänsel-Hohenhausen AG.Google Scholar
  11. Gilbert, M. 2006. A theory of political obligation: Membership, commitment, and the bonds of society. Oxford: Oxford University Press.CrossRefGoogle Scholar
  12. Meijers, A. 2003. Can collective intentionality be individualized? The American Journal of Economics and Sociology 62(1): 167–183.CrossRefGoogle Scholar
  13. Pettit, P. 2003. Groups with minds of their own. In Socializing metaphysics, ed. F. Schmitt, 172–175. New York: Rowman and Littlefield.Google Scholar
  14. Searle, J. 1995. The construction of social reality. London: The Penguin Press.Google Scholar
  15. Shockley, K. 2004. The conundrum of joint commitment. Social Theory and Practice 30(4): 535–557.CrossRefGoogle Scholar
  16. Stekeler-Weithofer, P. 2002. Zur Logik des ‘Wir’. Formen und Darstellungen gemeinsamer Praxis. In Kultur – Handlung – Wissenschaft, ed. M. Gutmann, D. Hartmann, and W. Zitterbarth. Weilerswist: Velbrück Wissenschaft.Google Scholar
  17. Tollefsen, D. 2002. Organizations as true believers. Journal of Social Philosophy 33(3): 395–410.CrossRefGoogle Scholar
  18. Tollefsen, D. 2004. Collective intentionality. Internet Encyclopedia of Philosophy. Accessed 19 Sept 2011.
  19. Tuomela, R. 2007. The philosophy of sociality. Oxford: Oxford University Press.CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2014

Authors and Affiliations

  1. 1.Department of PhilosophyGoethe UniversitätFrankfurtGermany

Personalised recommendations