(Forthcoming in
JOINT INTENTION, WE-MODE AND I-MODE
by
Raimo Tuomela
I INTRODUCTION
The central topic of this paper is to study joint intention to perform a joint action or to bring about a certain state. Here are some examples of such joint action: You and I share the plan to carry a heavy table jointly upstairs and realize this plan, we sing a duet together, we clean up our backyard together, and I cash a check by acting jointly with you, a bank teller, and finally we together elect a new president for our country. In these cases the participants can be said to have a joint intention jointly or as a group to carry the table upstairs: the content of the intention involves our performing something together and the pronoun “we” refers to us, viz. you and me and the possible other participants considered together. When we jointly intend to carry the table, each of us can be said to we-intend to do it. The purpose of the present paper is to discuss joint intention and we-intention primarily in a strong sense, the “we-mode” sense. In this sense the participants intend as group members because of a group reason (as will be clarified later). This contrasts with “I-mode” joint or shared intention. When the participants intend in the I-mode they intend solely as private persons – in contrast to the we-mode case where they must functioning as group members and where intending for a group reason must be at play.[1]
In my earlier work I have often taken (we-mode) joint intentions to be expressible by means of locutions like “We will do X”, where the word “will” is used conatively (rather than predictively, in the future tense) and X is a joint action type (see Section V below for the content of a joint intention).[2] However, joint intentions can also have other contents, and we may also speak of the jointly intending agents’ jointly seeing to it or (what is different) bringing it about that a state or event obtains. They may thus jointly see to it that they jointly build a house, that one of them builds it, or that some outsiders are hired to do the job, etc.[3] In this paper I will, however, focus on the we-mode case of the performance of an action where the agents jointly intend to perform the action together.
The content of a (we-mode) joint intention has, as it were, two parts. In the case of single-agent intention I take the intention to have the form “A intends, by his actions, to perform X”. Analogously with this, we have in the case of joint agency (here the dyadic case) “A and B jointly intend, by their actions, to perform joint action X ” or “A and B jointly intend jointly to see to it that X” (or, from their perspective, “We will perform X together”). There is joint action and there is content of the joint action (e.g. jointly bringing about something is the joint action and what is being brought about is the content of the action). Joint action involves that each participant of a joint intention is in principle “actionally” involved: he has a share or part in the participants’ performance of the joint action in question. Jointly intentional joint action is joint action performed on the basis of a joint intention, where each participant (or at least most of them) performs his part intentionally on the basis of his we-intention to perform that joint action (or a closely related action). Accordingly, his intention to perform his part of the joint action can be derived from, and be justified by, the participants’ joint intention.[4]
All intentions are necessarily related to one’s own actions. In the single-agent case, an agent may intend to see to it that his car is fixed. This intention has as its satisfaction condition that the agent by his own actions sees to it that his car becomes fixed – e.g. he can get a mechanic to fix the car or fix it himself. If the intention concerns the direct performance of an action (e.g. when an agent intends to open the window) the agent must himself bring about the satisfaction solely by his own action. This kind of intention I will call (direct) “action intention”. A minimal rationality condition for an action intention, at least a prior intention, is that the agent must at least believe that it is not impossible for him to perform the action. Assuming that at least a prior intention involves commitment to the content of the intention, we can ground the previous claim by saying that if he would not so believe it would be pointless for him to commit himself to his task.
An action intention contrasts with an “aim-intention”. In the latter case it is not conceptually required that the agent believes that he with nonzero probability can alone bring about or see to it that the action or its “result” event comes about, rather the agent is assumed by his action to contribute to the aimed result.[5] The kind of aim-intention that will concern us in the present paper is we-intention. A we-intention is a participant’s “slice” of their joint intention, so to speak. Or the other way around, it can technically be said that a joint intention consists of the participants’ we-intentions about the existence of which the participants have mutual belief. As is the case of ordinary action intention, also in the case of joint intention the participants have to bring it about, by their actions, that the intended state or event comes about. We-intentions are different from ordinary action intentions not only in being aim-intentions but also in that they conceptually depend on the joint intention in question. The central condition of satisfaction of the we-intention is that the we-intending agent intends to participate and accordingly intentionally participates in the joint action. That is, he intends by his own action, his part or share, to contribute to the joint action. Thus, the agent’s having the we-intention to perform a joint action together with the others (or that the participants perform that joint action) entails his participation intention, which is an action intention.
There is the rationality constraint on we-intention that an agent cannot rationally we-intend unless he believes not only that he can perform his part of their joint action, X, with some probability, but also that he together with his fellow participants can perform X jointly at least with some nonzero probability. The jointly intending agents must believe that the “joint action opportunities” for an intentional joint performance of X are (or will be) there at least with some probability. Yet another property of a we-intention is that, in each rational participant’s view, it must be mutually believed by the participants that the presuppositions for the jointly intentional performance of X hold or will hold with some probability.
The formation of a joint intention (and hence a we-intention, a personal “slice” of the joint intention) requires that the participants jointly and typically intentionally make up their minds to bring about something, thus exercising joint control over the possible courses of action and settling for a particular content. The formation of a joint intention (or plan) is based on their various personal and, especially, shared desires and mutual and other beliefs. In this sense a joint intention can be said to “summarize” or reflect the motivation underlying joint action. Of course, this final motivation underlying the joint intention need not be anything like an aggregation of private motivations but may instead be a compromise based on discussion, negotiation, or bargaining. In contrast, joint desires and wants do not similarly involve making up one’s mind and do not necessarily lead to intentional joint action, to attempts to function rationally in a coordinated way, and to fulfill the already formed plan. Furthermore, rational desires and wants do not require success beliefs as their normal (or “normal-rational”) accompaniments.
Concentrating on the case of the direct performance of a joint action, in joint-intention formation each agent accepts to participate in the participants’ joint performance as a group (or as one agent) of an action X. We can think of the situation in analogy with the case of a single agent: An agent intends X and performs X on the basis of the intention. Now assume that the agent is collective, and then the same can be said of it. A collective agent’s (group’s) intending and acting amounts to the group members’ jointly intending X and jointly acting on the basis of their joint intention. They must act on their joint intention for their satisfying action to be jointly intentional and not only intentional in the weak sense that all the participants act intentionally. If that satisfying action were not jointly intentional the analogy with the group agent acting intentionally would not hold true, and they would not act as a unit.
In the present case the agents jointly intend as a group and thus in the we-mode to realize the content of the joint intention that they – by their forming of the joint intention – have jointly accepted as the group’s intended goal (broadly understood). The joint intention entails a collective (viz. joint) commitment in analogy with a single-agent’s intention entailing a commitment to the content of the intention. A group is analogously committed to satisfying its intention, and the group members are collectively committed to satisfying their joint intention and are also socially committed to each other to performing their parts of the satisfying joint action. The collective commitment serves to keep the group members together as a unit. They function as group members towards achieving the content of the joint intention. Collective commitment need not be stronger than what joint intention conceptually entails: When our agents jointly intend X (e.g. a joint action) they must collectively bind themselves to X and what its satisfaction requires. This I call the instrumental group sense of collective commitment. This minimal group sense is intention-relative and, strictly speaking, non-normative or, if your prefer, only “technically” normative. In this sense my non-normative account sides with Michael Bratman and Seumas Miller against Margaret Gilbert.[6]
In the joint intention to perform a joint action X, it is precisely the content of the intention that is shared, viz., performing X jointly is shared. Because of the very idea of participation, each participant must intend to perform his part of the joint action, and, furthermore, that is typically required for the joint intention to become satisfied. The part performance must be intentional, of course, and thus based on the agent’s intention (and indeed, ultimately, on the joint intention in question, as will be seen later).
II I-MODE VERSUS WE-MODE REASONS FOR INTENDING
Consider the case where you and I jointly intend to go to a museum. We have jointly accepted that going to the museum is our, viz., our dyad’s, goal, and we are collectively committed to satisfying this joint goal with a social commitment (bond) to each other to this. Accordingly, we-mode joint intention concerning a joint action X entails that the participants must have collectively accepted “We together will perform X” (or one of its variants) for their group and hence for themselves, and consequently they must have collectively committed themselves to performing X jointly. Here “We together will perform X” applies to each participant, and in the case of a single participant it expresses his we-intention. An agent’s we-intention then is his part of the agents’ joint intention, and conversely a joint intention can, upon analysis, be said to consist of the participants’ mutually known we-intentions.[7] To recall, a we-intention is not by itself an action intention but an aim-intention involving that the agent intentionally aims at X (viz., at that the group performs X)and is “aim-committed” to X, while his “action commitment” is to performing his part of X. The agent’s intention to perform his part of the joint action accordingly is a proper action intention (that presupposes, of course, that the others perform their parts). We have seen that when the agents jointly intend to perform a joint action together, this involves performing one’s part of the joint action, and the concept of joint action in the general case cannot be composed out of individual action concepts independent of the joint action concept. Indeed, in many cases the joint action cannot antecedently be decomposed into parts in a more informative way than saying that you and I will both have to participate without a fully clear idea of what these parts will amount to (cf. the pattern of reasoning (W2) assumed in Section III to apply to we-intenders).
Focusing on full-blown joint intentions, a central subclass of them is formed by joint intentions based on the participants’ (explicit or implicit) agreement to act jointly. The making of an agreement in the full sense (viz., accepting a jointly obligating plan) is a joint action which is necessarily intentional. What does a plan-based joint intention presuppose? It must obviously be required that the participants understand – at least in some rudimentary sense – that a joint action in some sense, however weak, is involved. The joint action must be taken to include a “slot” for each participant’s intention. In general, all the relevant generic action concepts need to be possessed to a relevant extent by the participants – a kind of “hermeneutic circle” is at play. Thus the notion of joint action opportunity needs to be available to the participants together with the background knowledge, most of it culture-dependent, that it in turn presupposes. Next, certain situation-specific information must be presupposed. If the performance of a joint action X in a situation S is at stake, the concept of X must be possessed at least in a functional sense (not necessarily in a cognitive and reflective sense) by the potential participants, and they must also understand what S involves concerning the performance of X. It is also required that, at least ideally, each participant believes that the participants mutually believe that the joint action opportunities for X hold in S.[8] Some direct or indirect communication (or signaling) between the participants is needed for the reason that the participants are autonomous agents who, nevertheless, must make up their minds depending on what the others are thinking and doing. More concretely, communication is required for them rationally to arrive at unconditional intentions (we-intentions as well as the entailed intentions to perform one’s part of the joint action) in the presupposition-dependent sense to be discussed. The indirect communication may be previously “codified” and may relate to certain kinds of situations (e.g. “in situation S we always have a joint plan of a certain kind and act together appropriately”).
Jointly intending as a group thus amounts to jointly intending to realize a joint plan (e.g. simply to achieve a goal or to perform an action together) when functioning as a member of the jointly intending group of persons. Intuitively, the participants must be suitably bound together for proper collective action purporting to realize the shared plan to come about (and accordingly for their acting as a group).[9] In the general case we must allow that special authorized members act as operative members for the group and form intentions for it, but below I will mostly ignore this feature. When some persons jointly intend and act as a group I speak of “we-mode” intending and acting. It contrasts with I-mode intending and acting which is based on intending and acting as a private person. Let me now try to state a more precisely what an I-mode reason and a we-mode reason amount to. I will give the following characterization for a simple case relating to the context of a specific group, assuming that a reason is a fact or a fact-like entity expressible by a that-clause:
(IMR) Reason R is a group member’s motivating I-mode reason for performing an action X if and only if R is the agent’s main private (“non-group”) reason for his performing X. Typically, R is a state that the agent wants or intends to be the case or a state that, according to his belief, obtains; and X is an action that is a means to R or an action that R requires for its obtaining such that the agent is privately committed to performing X on the basis of R.
(WMR) Reason R is a group member’s motivating we-mode reason for performing an action X if and only if R is the agent’s main motivating group reason for his performing X. Typically, R is a state that the group in question wants or intends to be the case or a state that, according to the group’s belief, obtains; and X is an action that is the individual’s part of a collective action that is a means to R or a collective action that R requires for its obtaining, where the group members are collectively committed to performing the collective action for reason R and mutually believing so.
In (WMR) X can be e.g. a collective (or group) action with multiple tokens (e.g. going to church on Sundays) or a joint action like cleaning up a park as a many-person action. Functioning as group members, the participants have collectively accepted a the performance of a joint action as an aim for the group and are collectively committed to performing it for group reason R (a state expressible by a that-clause), and they are also socially committed to the group members to performing their parts of the collective action for reason R. Having a private commitment means in (IMR) that the person privately (rather than as a group member) has psychologically bound himself to a “content”, e.g. to performing an action for a reason. In general, private commitment is dependent on an intention, here the intention to reach a goal. Notice that functioning in the we-mode is necessarily connected to a group reason while functioning in the I-mode is at most only contingently connected to the group reason when one is involved. ((WMR) and (IMR) can be generalized to cover any voluntary attitude and action.)
One can now say that an attitude is in the I-mode or in the we-mode if, respectively, it is held for an I-mode reason or for a we-mode reason, and the analogous applies to actions.
III WE-INTENTIONS AND JOINT INTENTIONS ANALYTICALLY ELUCIDATED
In accordance with the grounds presented in the previous section, I will next discuss social we-intentions and joint intentions in a more analytical and precise fashion. Although the conceptual distinction between a we-mode and I-mode intention is clear in view of what was said above, both kinds of social intentions can in a sense be viewed in terms of the notion of a shared we-attitude.[10] Concisely put, a we-attitude is an attitude a person has relative to some collective if and only if he has that attitude in part because, according to his belief, the others in the collective have it, there being mutual belief among the participants about their having it. We can apply this to intentions and say somewhat loosely that a person has the we-intention to do X (or that action or outcome X will come about) if and only if he has the intention to do X (or that action or outcome X will come about) and has it in part because, according to his belief, the others in the collective have that intention, there being mutual belief among the participants about this. When there is shared we-intention in this sense we can speak of at least a weak joint intention, one in the I-mode. In the we-mode case we can say that the participant intends to participate in X at least in part because the group intends to perform X – recall that the group’s intending can be based on the operative agents’ intending rather than all the members so intending. I intend to participate in our joint performance of X in part because we intend to perform X and only derivatively because the others intend to participate. The group reason here is a conceptual condition for a participant’s we-intention while the other participants’ intention to participate (his we-intention) is a presupposed rationality condition (“it would not be rational for me to participate unless the others participate”). We may also say that in the we-mode case the participants are functioning as group members but not in the I-mode case, where they are functioning merely as private persons.
A we-mode joint intention is expressible by “We as a group will perform X jointly”. Here “we” is assumed to refer to a group or to some persons collectively considered and indeed the participants are assumed to take themselves to form a “we”, that is, “we-together” or “we-as-a-group” which is not a mere aggregative “we” that signifies “the two of us” (the basic I-mode sense of “we”). The “we-as-a-group” notion indicates the we-mode sense of “we”. We can say that the intentional subject of a we-intention is “we” while the ontological subject of a we-intention is a single agent.[11]
Deferring the group ascription case to the next section, I will here be concerned with the case in which the joint intention to perform X together is attributed to m agents, say A1,...,Ai,...,Am and in which case the participants assume that those participants really exist and understand what the joint action X at least in general terms involves. In this case, every participant Ai has the we-intention to perform X together with the others, and indeed the conatively used expression “We as a group will perform X jointly” can be taken to express the we-intention and also the group reason that each Ai here has for participating in the performance of the joint action X. The participants form a group here in the rather trivial sense that their having a joint intention means that they take themselves to be capable of acting as a group, and this will make them a social group on almost any view.[12]
I will next present the account of (we-mode) we-intention I have advocated for a long time.[13] This analysis will now be summarily formulated for the case of X being a joint action and on the presupposition that a we-intention expression “We will do X together” (or one of its variants or related phrases such as “member Ai we-intends that his group performs X together”) has been accepted by the participants A1,...,Ai,...,Am – entailing, under normal conditions, that a joint intention exists:
(WI) A member Ai of a collective g we-intends to do X if and only if
(i) Ai intends to do his part of X (as his part of X);
(ii) Ai has a belief to the effect that the joint action opportunities for an intentional performance of X will obtain (or at least probably will obtain), especially that a right number of the full-fledged and adequately informed members of g, as required for the performance of X, will (or at least probably will) perform their parts of X, which under normal conditions will result in an intentional joint performance of X by the participants;
(iii) Ai believes that there is (or will be) a mutual belief among the participating members of g (or at least among those participants who perform their parts of X intentionally as their parts of X there is or will be a mutual belief) to the effect that the joint action opportunities for an intentional performance of X will obtain (or at least probably will obtain);
(iv) (i) in part because of (ii) and (iii).
I have assumed that the participants actually exist, but I allow that a participant might under some outlandish conditions be mistaken in his beliefs (ii) and (iii). (On pain of not we-intending at all, he cannot be mistaken about the general situation at stake, viz. that there are some agents about to perform a joint action X.) Thus, in such abnormal circumstances a single agent can in principle have a we-intention. In such a case a we-intention is not a “slice” of a joint intention. Below I will, however, assume that the acceptance of “We will do X together” must be veridical and entail the existence of a joint intention. Thus it will be assumed below that all the agents in question really have the we-intention.[14]
I would like to emphasize three things here. Firstly, the general structure of the above account shows that we are dealing with a social we-intention. The agent intends to perform X together with the others, and his having this intention is based in part, according to his belief, on the others’ also intending to participate (or at least sufficiently many of them so intending that an intentional performance of X will result), there being mutual belief about this. That also the others collectively taken have the intention is a conceptual presupposition of a joint intention, but that they actually have formed the intention to participate in X is a contingent fact that can be an agent’s proximate factual reason for taking part. His so undertaking to participate is a matter of rationality rather than merely a conceptual feature. However, it must be emphasized that the principal proximate reasons for a participant’s intention to perform his part of X in the we-mode case – on conceptually grounds – is the group’s intention, although this group reason might not in fact exist before the participants’ forming their intentions to perform their parts. The group’s proattitude towards X precedes an agent’s intention to perform his part and is accordingly conceptually involved in this intention. In contrast, the fact that the others also intend to participate and have the relevant we-intention is a derivative rationality reason. Secondly, they are collectively committed to at least trying to see to it that their intention to perform X together will be satisfied, this collective commitment being a conceptual feature involved in their intention. Thirdly, the part division of X can be taken to indicate that a we-mode we-intention is at stake on the ground that the part action is a participant’s part of the agents’ intentionally performed action X based on their joint intention and thus proattitude to perform X for the group.[15]
I will comment on clause (i) later, but let me here say that this is not an “I-intention” in the sense that e.g. Searle uses the notion, because this is a we-mode personal intention essentially of the kind “I, as a group member, intend to participate”. The presupposed beliefs (ii) and (iii), expressing the minimal rationality of the we-intender concerning what a we-intention conceptually involves, and condition (iv) will not be commented on in detail here.[16] Let me only say that the joint action opportunity conditions include, besides the relevant mental and physical abilities of the participants, also that the others (or at least sufficiently many of the “right” kinds of them, as required for an intentional performance of X) will indeed participate.
It is presupposed by my analysis that a minimally rational we-intender should in the standard case of direct joint action be disposed to reason in accordance with the following two schemas (W1) and (W2) of practical inference (or in terms of their variants):
(W1) (i) We will do X.
Therefore:
(ii) I will do my part of X.
(W2) (i) We will do X.
(ii) X cannot be performed by us unless we perform action Z (for instance, teach agent A, who is one of us, to do something required of him for X).
Therefore,
(iii) We will do Z.:
(iv) Unless I perform Y we cannot perform Z.
Therefore (because of (iii) and (iv)),
(v) I will do Y (as my contribution to Z).
The first of these schemas in an obvious way connects we-intending to the we-intender’s own action - to his performance of his part or share of joint action X. The second of the schemas applies to all “normally rational” we-intenders, too, but of course only when the contingent clauses (ii) and (iv) apply, and it is to be exhibited by the we-intenders’ dispositions to reason in appropriate circumstances. This schema expresses part of what is involved in saying that a we-intention involves a joint commitment to contribute to the realization of the content of the we-intention. This joint commitment also involves social commitment, viz. that the participants are committed to one another to participating in the joint performance of X. Accordingly, (W2) clearly makes we-intentions cooperative to a considerable extent by requiring relevant interaction – or at least disposition to interact – of the participants. (W2) also indicates that the notion of one’s part of joint action X cannot be fully fixed in advance and constructed of antecedently available individual actions.
Supposing that joint intentions can be expressed by “We as a group will do X” or its variants, in order to cover “standing” intentions in addition to directly action-generating intentions we must also take into account dispositions to we-intend.[17] The following can accordingly be regarded as a true claim:
(JI) Agents A1,...,Ai,...,Am have the joint intention to perform a joint action X if and only if
(a) these agents have the we-intention (or are disposed to form the we-intention) to perform X; and
(b) there is a mutual belief among them to the effect that (a).
In the case of joint intention the conatively used “We will do X” is true of each participant Ai.[18]
The above account of we-mode joint intention and we-intention is conceptually non-reductive, although it is ontically individualistic or, rather, interrelational.[19] In my account (WI) the participants are assumed to believe that they will perform joint action X intentionally – as arguably an intention cannot be satisfied by means of non-intentional action. Assuming the standard case that the beliefs are true and assuming also that intentional action X requires a relevant intention (here typically X), we can in the present case take this (joint) intention to be one with the content to perform X , and we arrive at the view that (i) of (WI) entails for the first person case:
(*) I intend to perform my part of our joint action X as my part of X that we perform jointly intentionally as a group on the basis of our joint intention to perform X.
(*) is perhaps more circular than one may want to tolerate.[20] The notion of joint intention can here be understood in a somewhat vague, preanalytic sense, viz. in terms of what the agents actually “have in mind” when they jointly intend in everyday life, and this is what analytical theories of collective intentionality analyze and elucidate in terms of (partly) technical notions such as we-intentions. There is circularity but not vicious circularity here, because the elements in my analysis are not independently existing “building blocks” of joint intentions but only analytically isolated parts that presuppose the whole of which they are parts.
To approach circularity from a slightly different perspective, I claim that the italicized part that creates the circularity can be omitted without change of truth value (as indeed clause (i) of (WI) does). Let us thus go back to the following claim (in view of the second clause of (WI)):
(**) I intend to perform my part of our joint action X (that we will perform as a group) as my part of our performing X intentionally.
The central thing here is that the participants are assumed to be able to perform X intentionally as a group with an adequate functional understanding of joint action X. Performing X as a group is understood to involve acting fully as a group member, thus for a group reason (“forgroupness”) and being collectively committed to the performance of X and also to the joint intention to perform X.
Here intentional performance of X as a joint action should be understood at least in the practice sense, but not necessarily conceptually. The participants are taken to be able to perform X intentionally (in the most rudimentary sense as a kind of “collective pattern-governed action”), even if they need not have the full concept of intentionally performed joint action and may lack the ability to use it in their practical reasoning. Neither need the participants have the concept of joint intention in a reflective sense.
We can analyze intentional performance of joint action X further, and it could be done on the basis of the participants’ intentions to perform their parts of X as intended (or intentionally). We do get regress, if we try to analyze intentional joint action directly in terms of joint intention. But we can stop – in a functional sense – with just the requirement of intentional performance of joint action X. The participants must simply be able to act – to perform X – or at least try to act or to initiate action before forming the intention to act and similarly to act together or try to act together before being able to form the joint intention to perform X. So practice is central – the intention requires the practice. It can accordingly be suggested that we-intention-in-action is prior to future-directed intention.[21]
IV INTENTIONS ASCRIBED TO GROUPS
Another point that still needs to be made here concerns broader cases in which a group can be said to intend to perform an action. My account makes a distinction between group members who are somehow authorized (typically by the group members) to form intentions for the group and possibly to act on these intentions. These authorized members are called operative members.[22] One can accordingly speak of the ”authority system” for a group. It consists of a set of operative members that have the authority to bring about a group will or, more generally, the group’s acceptance of proposition e.g. of the kind “This group (“we”) will do X”, or “We ought to do X” or “We believe that p”. The notion of an authority system as such does not take a stand on from where such social group authority comes, but basically it can be due to internal authorization by the group members, external authorization from outside, or coercion by a powerful set of operatives (e.g. a junta or dictator).[23]
Consider now this account:[24]
(GINT) Group g intends to perform action X (or intends that a state X obtain) if and only if there are specific operative members of g such that
(1) these agents acting as group members intentionally collectively accept the conative expression “We will do X” (or one of its cognates) for g and accordingly jointly intend p and are collectively committed to achieving it;
(2) there is mutual belief among the operative members to the effect that (1);
(3) because of (1), the (full-fledged and adequately informed) non-operative members qua members of g tend to tacitly accept with collective commitment – or at least ought so to accept – that their group g has accepted “We will do X” for the group (as specified in clause (1));
(4) there is mutual belief in g to the effect that (3).
Clauses (1) and (3) express the most central ideas here. (2) and (4) express rationality conditions that might not be properly fulfilled while there still is a group intention on the basis of the fulfillment of (1) and (3).
We can see from this account that the nonoperative members can in a central way take part in the group’s intention simply by functioning as group members and tacitly accepting the operatives’ joint intention – or at least being normatively obligated to such acceptance. This tacit acceptance can amount to the nonoperatives’ endorsing the operatives’ joint intention qua members of g but they can well be “in reserve” e.g. concerning the execution of the intention. For instance, the sales personnel of a department store (an organization) can take part in the organization’s intentions and actions just by doing their work and without perhaps knowing very much about what the operatives for decision making are doing. Notice that the analysans of (GINT) can be satisfied even when some non-operatives fail to tacitly accept “We will do X” for g and the participation intentions that this intention expression entails for them (together with other relevant information). In such a case it does not seem appropriate to say that they really participate in the group’s intention.
Christopher
Kutz has recently argued that in a “minimal” joint action the participants need
not share a collective end but need only have a “participatory intention”, viz.,
an intention to do one’s part in joint action (accompanied by the belief that
the part action indeed will contribute to the collective end or goal).[25]
As just seen, I accept this for the case of actions performed by structured
groups (which actions Kutz includes among joint actions), for in the case of
group action in my account there need not be even such a participatory
intention in the case of all participants. However I claim, seemingly contrary
to Kutz, that in the case of many-person joint action a relevant joint
intention is required for jointly intentional performance (except perhaps in
the case of some kinds of defective joint actions). I generally require
collective commitment for joint intentionality, but Kutz’s account does not
require it or even shared private commitment. This feature makes his minimal
joint actions to some extent functionally deficient.[26]
V I-MODE AND WE-MODE JOINT INTENTIONS CONTRASTED
In this section I-mode and we-mode joint intentions will be revisited and their relationship considered. I will present an argument for the irreducibility of the we-mode to the I-mode in the case of joint intention. As above, I will mean by a full-blown, we-mode joint intention an intention expressible by “We together will do X”, where X may be a joint action or just any state that can be (jointly) intended to be brought about. When a joint intention is about performing a joint action, X will be a joint action the performance of which is expressible by means of “We are doing X” (or “We did X”, etc.). My point here is that the we-perspective must apply as indicated. Indeed, I assume that a joint intention and a joint action expressible in this way with a “thick we” expressing togetherness is necessarily in the we-mode, viz. the participants are jointly intending and joint acting as a group.[27] We-mode joint action is to be distinguished from interpersonal, dependent action directed towards the same goal. For instance, Seumas Miller’s theory is concerned with such interpersonal action directed towards an individualistically conceived “collective end”, and similarly Michael Bratman’s account belongs to the I-mode realm.[28] No matter how workable in a functional sense these accounts are or may be, they deal with interpersonal, “I-thou” action that falls short of being based on a group reason. The most central distinction in the case of discussing in some sense joint or shared or same attitudes and actions is whether the participants function as a group (as one agent, as it were) or whether there is only interpersonal action. The former does not entail the latter even if also (actual or potential) interaction is included, because the we-mode involves functioning as a group member and not as a private person while the I-mode is concerned only with functioning as a private person. Nor does the latter entail the former. The kind of “oneness” involved in acting as a group member in the case of the former is not as such involved in the kind of interaction pattern that the latter contains.
My
starting point here is to consider preanalytic, common-sense examples like the
multi-agent action of lifting a table or the multi-agent action of going to
To briefly comment on other approaches, Seumas Miller would say about the table lifting case that the participants shared the collective end of having the table lifted and they acted dependently to achieve that. His theory gives an I-mode answer. There is no joint intention, there is no we-action, no “we-together”, and there is no joint commitment (unless perhaps such a requirement is added as a separate requirement). Michael Bratman’s theory says for the case with you and me as participants that I intend that we lift the table (not necessarily as a joint action) and similarly for you, our intentions being suitably dependent; and then we carry out our intentions (I-mode intentions, in my classification).[29] Margaret Gilbert would say that the participants are jointly committed to lifting the table and therefore form a plural subject, and somehow act in light of their joint commitment to lift the table.[30] Her “plural subject” account can be viewed as a we-mode account, and so can John Searle’s view in principle (although the latter it is rather individualistic at bottom).[31]
As to my own account, it is clear that forgroupness (thus group reason) and collective (or joint) commitment need not be explicit in the content of the intention, nor do they appear in the surface of common sense parlance and thinking. They can be regarded as underlying conceptual presuppositions that philosophical analysis has revealed.
Given the above, the following simple argument for the ontological irreducibility of we-mode states can be proposed:
(a) The concepts of joint intention as a group and joint action as a group (in technical terms, “in the we-mode”) are not reducible to concepts expressing private intentions and actions (viz. intentions and actions based on a private or purely personal reason) and what can in general be conceptually constructed out of I-mode resources.
(b) If there actually are joint intentions and joint actions that can be correctly described as joint intentions and actions as a group, then they differ ontologically from private intentions and actions with the same content (or with a “counterpart” content).
I will here assume that the antecedent of (b) is true and that we-mode concepts are not idle:
(c) Some joint intentions and joint actions as a group in fact exist (and indeed are needed in many cases, e.g. for suitable tasks or many-person activities such as tennis, dialogue).
Therefore,
(d) Some ontologically irreducible joint intentions and actions as a group that exemplify the corresponding conceptually irreducible concepts in fact exist (and are, in addition, needed in many cases). (From (a), (b), and (c).)
What is at stake in premise (a) is basically the concepts of joint intention and joint action as a group. (My explication of the rather colloquial phrase ‘as a group’ is precisely the we-mode.) If concepts are irreducible then it is reasonable to think in the present case that the states and events to which they correctly apply are ontologically irreducible as well. The conceptual side so to speak induces naturalistically different ways of thinking and acting in the cases under comparison, and this means ontological difference (that very likely exists in the brain as well). Premise (b) refers to really existing activities. As such, (a) does not entail (b). There are (in principle) ontologically idle concepts.
To defend (a), one central argument for the claim that joint intention and action in the we-mode are irreducible to I-mode interpersonal intention and action is simply that the former are based on a group reason while the latter are based on private reasons. A group reason (e.g. “Our group intends to lift this heavy table”) for a participant’s intention and action to participate is different from private, I-mode reason (e.g. “I intend that you and I lift this table merely on my private grounds”, and similarly for you). Supposing that the relevant epistemic requirements for the participants are in place, in the we-mode case the participant’s reason (at least when reconstructed by a theoretician) involves a group (thus forgroupness and collective commitment) in a sense in which the I-mode reason does not. The I-mode reason here is based on the idea that I participate because of my private reason, given that you also participate. There need be no collective commitment here, only your and my private commitments, and you and I do not act for the use or benefit of a group, e.g. the dyad formed by you and me, but for our own personal use or benefit. E.g. if you also benefit from the I-mode joint action, that is only a contingent consequence from my point of view, not something that I have aimed at.[32] To be sure, in real life one often acts both for group reasons and private reasons when performing an action. In such cases the central problem is whether the group reason is effective enough to make the agent satisfy the conditions for acting as a group member.[33]
Assumption (b) can be argued on the basis of plain common sense. When two agents jointly lift the table that must be based on their joint intention, and each agent performs his part of the joint lifting, where one’s part of table lifting is ontologically and functionally dependent on the notion of joint table lifting and its ontic content might be fully determinable only after the performance of the joint action. However, a table can be lifted also in the I-mode in terms of interpersonal action directed towards the shared end of the table being lifted (cf. aforementioned Miller’s and Bratman’s accounts). My central point here, though, is that there are full-blown we-mode cases. For instance we make an agreement or shared plan to lift this table and perform our parts because of our joint plan for the simple reason that we have the joint intention to lift the table or for the reason that we are cleaning our house and take that to require that we suitably lift the table. This clearly can be regarded as a full-blown we-mode case. (In cases like pushing a broken bus up the hill, there might be I-mode actors involved, but if it is true that we together are seeing to it that the bus moves in the right way there must be some we-mode joint acting involved.)
To deepen the discussion, we consider Bratman’s analysis of shared intentions in terms of the locutions "We intend to J," where J is a joint action. I claim that this account is an I-mode account of joint intention. Indeed it is rather close to a weak we-intention in our present sense. His final analysis is this:[34]
We intend to J if and only if
(1) (a) I intend that we J and b) you intend that we J;
(2) (a) I intend that we J in accordance with and because of meshing subplans of (1)(a) and (1)(b); you intend that we J in accordance with and because of meshing subplans of (1)(a) and (1)(b);
(3) (1) and (2) are common knowledge between us.
Why is this an I-mode account? The most basic reason is that the analysans does not entail a group reason, viz. that the agents have their intentions of the form “I intend that we J” (for each participant) in part because of a group-reason such as a full-blown joint intention (and conceptually entailed proattitude) expressible by “We together intend that we J”. Hence the we-mode is not entailed, and thus there need not be collective acceptance of the joint action J as the group’s jointly intended content meant for the group to which the group members are assumed to be collectively committed. What may thus also be missing is social commitment that I take to be involved in collective commitment. (More precisely, the requirement that the participants be socially committed to each other in a quasi-normative way to participate is missing.) There need be no group-based intention and proattitude that “reasons” the agents’ personal intentions in Bratman’s account. The agents may have their personal intentions on the basis of their private reasons.
According to my we-mode account, the participants (“we”) are jointly lifting the table in accordance with and partly because of our joint intention to lift the table (or to do something closely related). Because of our joint intention we are jointly committed to the task. All this is in the we-mode – we have thus collectively accepted the joint intention (plan) for us considered as a group or as a unit. We are jointly intentionally performing the joint action of lifting the table and each of us is intentionally lifting his side of the table. However, we can also lift the table in the I-mode as can be seen from considering the satisfaction conditions of the shared intention in Bratman’s above account.[35]
The following thesis connecting joint intention and joint action seems defensible: It is necessarily the case that if there is joint intention (involving some agents jointly intending as a group) it can only be satisfied by joint action as a group (at least they must jointly see to it or bring it about that X, even if X is a neutral event or someone’s performance of an action). The converse also seems to hold in the sense that – and this my quasi-Davidsonian conjecture applied to joint action – if there is joint action (viz. if a piece of collective behavior is joint action) then it must be (jointly) intentional ”under some description” (there must be something relevant that the agents did jointly intentionally), and this in turn entails that there must be a relevant joint intention on the basis of which the agents performed the action jointly. In typical cases this joint intention has the content X, if X is the action jointly intentionally performed.
To start arguing for this thesis, consider a successful case of satisfaction of joint intention. You and I (”we”) have the joint intention to lift the table in ten seconds, …, in one second, …, now, and as result we are jointly lifting the table with the joint intention-in-action (or joint ”we-willing”) to lift the table. Thus we are jointly intentionally lifting the table, each of us intentionally lifting his side of the table. This is basically because we, the participants are jointly lifting the table in accordance with and partly because of our joint intention to lift the table (or to do something closely related). Because of our joint intention we are jointly committed to the task. All this is in the we-mode – we have thus collectively accepted the joint intention (plan) for us as a group or as a unit.
The joint action satisfying a joint intention cannot be (using the example just under discussion) just the aggregate action consisting of your and my private actions of my lifting my end and your lifting your end of the table. Rather, it is our joint action of lifting the table. That it is our joint action means that we take it to be an action that we have chosen to perform jointly, as our joint action. There is a group reason for the action and not (only) private reasons. We are jointly committed to this joint action, the joint commitment being based on the joint intention. All this entails that we are grounding our action (minimally) on our joint intention (a group reason, a we-mode reason) and that our action here is and must be in the we-mode and involve our acting together as group members.
As to Bratman’s account, the action J need not even be a joint action based on a shared intention.[36] We have here a second line of argument for showing that Bratman’s account is an I-mode one, if we assume that a joint intention can only be satisfied by a joint action based on that intention. The point is of course that in his account such a joint action is not required and that therefore there need be no joint intention of the full-blown, we-mode kind present.[37]
There are other we-mode features that seem at least generally to hold true. Thus, the participants in joint intention and action also satisfy a version of the three musketeers’ condition according to which necessarily “All for one and one for all”. This is a kind of collectivity principle that applies not only to semantic satisfaction of attitudes but to actions (e.g. part actions of joint actions, in contrast to the I-mode case), group reasons, and to full group responsibility. Accordingly, when the an attitude (or action) is satisfied for a member it is necessarily satisfied for all members. Furthermore, there is shared authority (and responsibility) in the we-mode case over the joint action X and its parts. This contrasts with the I-mode case where every participant rather has full (private) authority over his action in the aggregate action (but of course not over the aggregate action).[38]
My overall judgment concerning the above irreducibility argument is that (d) not only follows from the premises argued for but that the argument is about as “sound” as it can be in this kind of field of research. Furthermore, I would like to point out that my basic we-mode argument (the need of irreducible we-mode mental states) seems to be applicable, mutatis mutandis, to any other attitude or action involving jointness.
Space does not allow a proper discussion of what we-mode joint intentions and joint actions are functionally good for, but let me still make one point here related to collective action dilemmas. The argument goes as follows for a Prisoner´s Dilemma: Regarding a group as a collective agent, it is assumed that the group (“our group”, “we”), assumed to accept the dominance principle (“higher payoff dominates lower payoff”), intends to choose C (over D). Thus, here the group intends and acts only if the members correspondingly jointly intend, as a group, to cooperate and accordingly jointly cooperate, the participants form the joint intention to realize the joint cooperation outcome and perform their parts of the group’s achieving it, viz. they choose C. As the group members are assumed to act as group members, and thus as one agent, the joint outcomes CD and DC simply drop out. The Prisoner’s Dilemma simply disappears in principle. The agents act collectively rationally when they act for a we-reason as here, indeed there is no room for private individual action and private rationality at all. The upshot then is this: Strategically acting agents acting on the basis of the relevant I-mode reason involving maximization of private value (utility) can rationally choose only D (viz., mutual defection is a Nash equilibrium in a single-shot Prisoner’s Dilemma), but agents acting for the relevant we-mode reasons (e.g. “our group will maximize the value for the group by acting appropriately”) will rationally choose C. Of course, it must be added to all this that in actual practice groups may not act fully as units in the above committed way. There will be free-riders and there must be some sanctioning involved. Thus, from the backdoor, as it were, we may get the Prisoner´s Dilemma alternatives back. (I will not here discuss the control and other problems involved here.)
V CONCLUDING SUMMARY
I will now summarize my own account of joint intention
and we-intention. We may recall that an I-mode account of joint intention is
concerned with putting together individuals’ private intentions when they for
some reason want to act together or achieve something together. The basic model
here, in the two-person case, is: I regard you as a part of my environment
rather than as a fellow group member and we face a task to be solved together
by our actions. In contrast, in my we-mode account you and I form a “we”, a
group, which acts possibly in order to solve a jointly accepted task, or we
just act together for purely social reasons (“we value and enjoy each other’s
company, basically no matter what we are doing together”).
In my account the intention-dependencies go this way
in an unstructured group g with m members that has formed an intention for
itself:
(1) Group g, consisting of the agents A1,…,Am,
intends to bring about X (or see to it that X), where X may in principle be any
state or event (as long as no rationality constraints on g are imposed).
As groups are not literally agents, we must speak of
the group members’ relevant states and actions:[39]
(2) The members A1,…,Am of g
jointly as a group intend to bring about X; put in more linguistic terms, these
agents jointly accept that the intention expression “We will bring about X” is
true of them (collectively and also individually, as group members).
The next step is the following “distribution”
assumption:
(3) Each one of the members A1,…,Am
of g accepts, qua a member of g, in this situation that “We will bring about X”
is true of himself.
(4) Each member must reason or be disposed to reason
as follows:
(i) We will bring about X
Therefore,
(ii) I will participate in our (viz. A1,…,Am’s)
jointly bringing about X or jointly seeing to it that X.
As a paraphrase of (ii) we have
(5) I will perform my part of our jointly seeing to it
that X.
It is inessential here what exactly a part of X means
– it will have to come functionally from participation in X, whatever that
takes. Similarly, we can here ignore how exactly each participant’s part is
determined. Note that if there is at least minimal rationality involved the
participants must at least rationally hope that X will come about due to their
joint effort. This serves to give a partial indirect characterization of what
“a part” and “participation” involve.
Next, given that A1,…,Am have
the right presupposed mutual beliefs about the action opportunities for jointly
seeing to it that X and given that the analogy between single agents and group
agents (as “jointified” basically by (2) above) holds true, we can make (5)
more specific. This will still require that the analogy assumption is based on
the two ideas that (a) an intention to act can only be satisfied by means of
intentional action and that (b) intentional action must be analyzed as action
performed on the basis of a relevant intention – e.g. precisely the intention to
bring about X, and this is what I will assume here. So we get
(6) I intend to perform my part of our jointly seeing
to it that X as my part of X when acting as a member of g.
(6) presupposes that our jointly seeing to it that X
is our jointly intentional action performed as jointly intended (that is, as
based on our joint intention, however preanalytic, to jointly see to it that
X).
(6) gives the main individual-level content of what my
notion of we-intention involves. Its group-level content is given by (1) and
its jointness-level content is basically given by (2) while (3) – (6) concern
the individual personal level of acting as a group member. Note that any
collection of agents that satisfy (1)-(6) form a group in my sense and so does
a collection capable of forming a joint intention (and having mutual belief
about this), and indeed the latter, weaker requirement is necessary for their
forming a social group capable of action as a group.
Notice still that the concept of joint intention can
be taken to entail:
(7) If the members A1,…,Am of g
jointly intend to bring about X, then they are jointly committed (bound) to
intending and bringing about X.
We may leave it as a task of psychologists to find out
how people in ordinary life, as it were, simplify (6) when intending and acting
jointly so that the instrumentally and functionally right things comes about.
There is some circularity here: joint intention does come to depend on joint
intention, although perhaps only an unarticulated, preanalytic notion of joint
intention. But so what, we may say, as long as people are able to jointly
intend in terms of the above account and produce X in the right way. (Recall
Section IV.)
It is basically an empirical matter, a matter of upbringing, teaching, and education as well as, more generally, the surrounding culture whether people tend to function prevalently in the we-mode or in the I-mode or in which contexts they prevalently do which, or, more generally, how the people tend to weigh their we-mode and I-mode reasons relative to each other.*
* I wish to thank Raul Hakli, Frank Hindriks, Kaarlo Miller, Antti Saaristo, and Maj Tuomela for comments.
References
Balzer, W. and Tuomela, R. (1997), ‘A Fixed Point Approach to Collective Attitudes’, in Holmström-Hintikka, G. and Tuomela, R. (eds.), Contemporary Action Theory II: Social Action, Kluwer Academic Publishers, Dordrecht and Boston, pp. 115-142
Belnap, N., Perloff, M, and Xu, M. (2001),
Facing the Future: Agents and Choices in
Our Indeterministic World,
Bratman, M. (1999), Faces of Intention,
Gilbert, M. (1989), On Social Facts, Routledge,
Gilbert, M. (1990), ‘Walking Together:
A Paradigmatic Social Phenomenon’,
Kutz, C. (2000), Complicity: Ethics and Law for a Collective Age,
Mathiesen, K. (2002), ‘Searle, Collective Intentions, and Individualism’, in Meggle, G.. (ed.), Social Facts & Collective Intentionality, in German Library of Sciences, Philosophical Research, vol. 1, Dr. Hänsel-Hohenhausen AG, Frankfurt, pp. 187-204
Meijers, A. (2003), ‘Can Collective Intentionality Be Individualized?’, in Koepsell, D. and Moss, L. (eds.), John Searle’s Ideas About Social Reality, Blackwell Publishing, Malden, MA, pp. 167-183
Miller, S. (1995), ‘Intentions, Ends and Joint Action’, Philosophical Papers 24, 51-67
Miller, S. (2001), Social Action: A Teleological Account,
Sandu, G. and Tuomela, R. (1996), “Joint Action and Group Action Made Precise,” Synthese 105, 319-345
Searle, J. (1990), ‘Collective
Intentions and Actions’, in Cohen, P., Morgan, J., and Pollack, M. (eds.), Intentions in Communication, The MIT
Press,
Tuomela, R. (1984), A Theory of Social Action, Reidel,
Tuomela, R. (1989), ‘Actions by Collectives’, Philosophical Perspectives 3, 471-496
Tuomela, R. (1991), ‘We Will Do It: An Analysis of Group-Intentions’, Philosophy and Phenomenological Research LI, 249-277
Tuomela, R. (1995), The Importance of Us: A Philosophical Study
of Basic Social Notions, Stanford Series in Philosophy,
Tuomela, R. (2000), Cooperation: A Philosophical Study, Philosophical Studies Series 82, Kluwer
Academic Publishers,
Tuomela, R. (2002), The Philosophy of Social Practices: A
Collective Acceptance View,
Tuomela, R. (2003), ‘The We-mode and the I-mode’, in F. Schmitt (ed.), Socializing Metaphysics: The Nature of Social Reality, Rowman and Littlefield, Lanham, Md, pp. 93-127
Tuomela, R. (2005a), ‘We-Intentions Revisited’, Philosophical Studies 125, 327-369
Tuomela, R. (2005b), Philosophy of Sociality: The Shared Point of View, Oxford University Press, forthcoming
Tuomela, R. and Miller, K. (1988), ‘We-Intentions’, Philosophical Studies 53, 115-137
Tuomela, R. and Bonnevier-Tuomela, M.
(1997), ‘From Social Imitation to Teamwork’, in Holmström-Hintikka, G. and
Tuomela, R. (eds.), Contemporary Action
Theory, Vol. II: Social Action, Kluwer Academic Publishers,
Tuomela, R. and Tuomela, M. (2003), ‘Acting As a Group Member and Collective Commitment’, Protosociology 18, 7-65
Zaibert, L. (2003), ‘Collective Intentions and Collective Intentionality’, in Koepsell, D. and Moss, L. (eds.), John Searle’s Ideas About Social Reality, Blackwell Publishing Company, Malden, MA , and Oxford, pp. 209-232
[1] In the I-mode case there
may be a group reason involved but it is had privately. A we-mode
group reason, say R, can only be had when functioning as a group member. In contrast,
R can be a person’s I-mode reason when he is functioning merely as a private
person. Functioning in the we-mode is necessarily connected to a group reason
while functioning in the I-mode is only contingently connected to the group
reason (when there happens to be one involved). When R is a we-mode group
reason There are the following three possibilities for intending in the I-mode:
1) the person functions solely as a private person, 2) the person is a member
of an I-mode group, viz a collective in a broad sense in which the participants
are only privately committed to the group (see Chapter 1 of Tuomela, 2005b, for
I-mode and we-mode groups), or 3) the person is a member of a we-mode group but
only functions as a group member in a weak, I-mode sense of acting as rather
than in the we-mode sense. (Cf. the comments related to non-operative members
in Section IV as illustrating this possibility.)
[2] Cf. Tuomela (1984),
(1995), (2000) and Tuomela and Miller (1988).
[3] Jointly seeing to it that something is the
case can be used as umbrella notion to cover many kinds of activities – e.g.
jointly performing actions in a direct or in an indirect sense, jointly
bringing about states, jointly maintaining states, and so on (cf. Sandu and
Tuomela, 1996, and Belnap et al., 2001). Notice however, that in the context of
joint intention the actions in question can be of various kinds, as just
mentioned, and they can be non-equivalent (e.g. joint seeing to it that
something is the case does not entail direct performance, nor is the converse
true).
[4] In other words, there is
the activity in which the participants see to it that the (second part of the)
content comes about and there is the very activity that makes the content
obtain. Thus the participants might hire some other agents to build a house.
Here the first kind of activity is hiring the agents and monitoring their work;
and building the house (performed by the hired agents) is the second kind of
activity. In this paper I will, as said, mainly speak of the case of the
(direct) performance and acting together. (See Tuomela, 2000, Chapter 2, for
discussion.)
[5] A
linguistic formulation that typically seems apt for aim-intentions is
”intending that something be the case”, however, taken with the understanding
that an aim-intending agent is still required actively to contribute to the
coming about of the state in question (and this conceptual condition is in
general independent of whether realizing it is rational for him or not).
[6] See Bratman (1999), p.
125ff. and Miller (1995), p. 64, Gilbert
(1990), pp. 6-9.
[7] The collective acceptance
of an intention as the group’s intention entails the satisfaction of the
so-called Collectivity Condition. Applied to satisfaction, the Collectivity
Condition says, roughly, that necessarily, if the joint intention (goal) is
(semantically) satisfied for one of the participants, then it is satisfied for
all participants. (Note that here we are speaking of the participants personal
intentions in the we-mode, not in the I-mode.)
[8] Cf. the third clause in my analysis (WI) of we-intentions below in Section
III.
[9] This activity in general
requires public exchange of information between them if it is to lead to
mutually known (and not only mutually believed) unconditional participation
intentions. Another reason for publicity in a group context is that such
central social notions as the speech acts of agreement making, promising,
commanding, and informing – all relevant to joint-intention formation – are in
their core sense not only language-dependent but public.
[10] I have discussed we-intentions as shared
we-attitudes in the 1980’s and later more generally as applicable to all kinds
of attitudes (see e.g. Tuomela and Miller, 1988, Tuomela, 1995, Tuomela and
Bonnevier-Tuomela, 1997, Balzer and Tuomela, 1997, Tuomela, 2002, Tuomela, 2005b).
[11] This is the terminology
that Mathiesen (2002) uses in this connection.
[12] See Tuomela (1995),
Chapter 4, for an analysis of social groups along these lines.
[13] Cf. the treatments in Tuomela (1984, 1995,
2000, 2002, 2003, 2005a,b) and in Tuomela and Miller 1988).
[14]
From the point of view of the logical form of joint intentions, we-intentions,
and group intentions the following rather trivial point is central to keep in
mind. A joint intention can be expressed by an (m+1)-place predicate JI(A1,...,Am,X)
standing for “the agents
A1,...,Am jointly, as a group, intend to
perform the action
X jointly”. Here “as a group” can be symbolized by assuming that the
agents function as proper group members thus as members of a “we”, and we have
WE(A1,...,Am,X). The agents thus form a thick “we” at
least concerning X. This “togetherness-we” can be cashed out by saying that the
agents have collectively accepted joint action X as the agents’ “goal” or as
what they are to perform together for the group they form, being collectively
committed to performing X jointly. In this context, each member Ai has the we-intention to perform X (viz. each Ai
intends to perform X together with the others as
expressed by “We will do X”). This is symbolized by WI(Ai , X). The
account (WI) in the text gives an
analysis of what this involves. Finally, we can speak of a group’s intention to
perform an action, such as X, here. This is symbolized by I(g,X). In our simple
case g consists precisely of A1,...,Am. In this case JI(A1,...,Am,X)
and I(g,X) are truth equivalent, the two sides of one and the same coin, so to
speak. In more interesting cases, where the group involves the distinction
between authorized operative and non-operative members, the account (GINT) to be stated below, gives the
connection.
[14] Recall that I have characterized the we-mode in terms of three features: (1) the attitude or action must be collectively accepted as the group’s attitude or action, (2) it must be for the use and, typically, benefit of the group, and (3) the participants must be collectively committed to the attitude or action.
[15] In other contexts (e.g. Tuomela, 2002,
Tuomela and Tuomela, 2003) I have characterized the we-mode in terms of three
features: (1) the attitude or action must be collectively accepted as the
group’s attitude or action, (2) it must be for the use and, typically, benefit
of the group, and (3) the participants must be collectively committed to the
attitude or action.
[16] See the mentioned
references, especially Tuomela and Miller, 1988, for justification.
[17] See Tuomela (1991).
[18]
For two different logical formalizations of we-intentions in the sense of (WI)
and joint intentions in the sense of (JI), see Sandu and Tuomela (1996) and Balzer and Tuomela (1997).
[19] Cf. Tuomela, 1995, Chapter 9, on the
interrelationistic version of individualism, which has the feature that groups
as agents are viewed merely in an instrumental sense.
[20] This is what Searle,
1990, p. 405, says: “We are tempted to construe ‘doing his part’ to mean doing
his part toward achieving the collective goal. But if we adopt that move, then
we have included the notion of a collective intention in the notion of ‘doing
his part.’ We are thus faced with a dilemma: if we include the notion of
collective intention in the notion of ‘doing his part,’ the analysis fails
because of circularity; we would now be defining we-intentions in terms of
we-intentions. If we don’t so construe ‘doing his part’, then the analysis
fails because of inadequacy.” (Cf. also S. Miller, 2001, pp. 71-73.) In Tuomela (2005a) I respond to this charge
and claim that while there is some circularity, it is not vicious. On the
whole, Searle’s discussion of the Tuomela-Miller (1988) paper is based on a
massive misunderstanding probably based on his not having properly read the
paper, as I have shown elsewhere (e.g. in Tuomela, 1995).
[21] The discussion in Tuomela
(2002), Chapters 3 and 4, of collective pattern-governed behavior, generalizing
Sellars’ relevant notion to the collective case, may be consulted as giving
justification to the claims in the previous and this paragraph. Let me point
out that the present way of answering the charge of circularity is somewhat
different from the long answer I give in Tuomela (2003).
[22] Note that, at least typically, in unorganized groups all full-fledged members are operative
ones.
[23] See Tuomela (1995), Chapter 4.
[24] Similar formulations for intentions and other
attitudes including actions appear in Tuomela (1984), (1989), (1995), and (2005b).
The account given below is justified in these other works and cannot be
defended here. Note, however, that the present formulation concerns the case
where the groups are internally and externally free to form intentions. This
assumption denies the possibility of external and internal coercion (hence the
existence of a dictator is not a possibility covered here).
[25] See Kutz (2000).
[26] To go into some detail of
Kutz’s (2000) account, let me indicate where I think his analysis faces a
problem. His argument for the need of only participatory intentions goes in
terms of the following two conditionals where G is the end or “result” involved
in the joint action in question and P is the agent’s (“my”) part (p. 101):
(1) If I did not believe that P was a way of contributing to G in such
circumstances, I wouldn’t do P but would do P’.
(2) If I did not believe that P was a way of contributing to G’s
occurrence in such circumstances and I did not believe G could be realized in
these circumstances, I wouldn’t do P (but would do P’ or might call the whole
thing off).
According to Kutz, both conditionals have to be true in the case of
full-blown joint action, but in the case of minimal joint action (2) may be
false.
I claim that that these conditions are rationality conditions rather than mere conceptual conditions for joint action. (1) might not be satisfied in the case of an irrational agent. (1) appears to be compatible with all the agents believing or even mutually believing that g is not achievable. This seems problematic in most cases. Counterfactual (2) is too strong. Even a minimally rational agent might just hope that P together with the other participants’ contributions might lead to G but did not really believe it. But on conceptual grounds he could still intend the end G. This is at least what I have been claiming in my account, and this matter seems to me rather trivial. So (2) is not needed for intended ends. On the other hand, the participant in this kind of case, expressing minimal joint action, does not satisfy the requirement that he be (collectively or even privately) committed to G. This is basically because he may even believe that G might not be attainable.
However, there are also weaker,
I-mode kinds of collective actions in which not even a participatory intention
in the above sense is needed. These collective actions are actions based on shared we-attitudes,
where the shared we-attitude, e.g. a shared I-mode we-goal or we-belief or
we-fear, serves as the participants’ reason for action (in the typical case of
conformative we-attitudes for harmony-inducing action). See Tuomela and
Bonnevier-Tuomela (1997) and Tuomela (2002), Chapter 4.
[27] This notion in its full
sense is assumed to express the we-mode, which I have somewhat technically explicated
as a conjunction of collective acceptance (CA) and “forgroupness” (FG) and
collective commitment (Cocom) in the sense of involving collective acceptance of the “content” in
question as the group’s “content” to which the participants have bound themselves.
[28] See Miller (2001) and Bratman (1999).
[29] See Bratman (1999), p. 114, Schmitt (2003) and Tuomela (2005b), Chapter 3, for the fact that joint action is not required in Bratman’s account.
[30] Cf. Gilbert (1989) and
her later works. Gilbert is not specific about the jointly intending
participants’ attitudes. What happens in their personal minds when they
participate in a joint intention? As seen, my elaborate answer goes in terms of
we-intention and intention to perform one’s part.
[31] Searle claims that collective intentions are irreducible to individual intentions and mutual beliefs. However, when he actually tries to spell this out, he does not succeed in showing that collective intentions are different kinds of intentions, and indeed ends up with individual ”by-means-of ” intentions. This is argued by Zaibert (2003). Futhermore, as argued by Meijers (2003), Searle’s account is atomistic and does not account for interpersonal relations and in particular not for the social commitments that the participants are taken to have in relation to each other.
[32] The I-mode/ we-mode
distinction is different from the selfishness/altruism distinction. In the
present example I might as well have aimed to benefit you but contingently
ended up benefiting (also) myself.
[33] One can also list other
features of the we-mode which are different from the features of the I-mode.
One feature still worth emphasizing is
that the notion of a part of a joint action X is not conceptually
reducible to individual actions, as e.g. Seumas Miller’s theory assumes. One’s
part depends crucially on X and may not be antecedently fixable. Think of. the
practical inference schema (W2) which may generate new elements into one’s part
during the course of the performance of the joint action. This shows that there
are in principle indefinitely many factors that may enter the situation during
action. The part cannot be fully specified beforehand. However, the present
point seems not to apply to Bratman’s theory, as he only speaks about “our
J-ing” in his account.
[34] Cf. Bratman
(1999).
[35] A shared we-attitude
analysis of I-mode joint intention also yields something very similar. Here is
the account of I-mode joint action that I have recently proposed in Tuomela
(2005b):
You and I intentionally performed X together if and only if
(1) X is a collective action type, viz., an “achievement-whole” divisible
– either ex ante actu or ex post actu -- into A’s and B’s parts;
(2) We shared the (reason-based, see below) I-mode intention to perform X
together;
(3) We performed X together in accordance with and partly because of our
shared I-mode intention to perform X (or some “closely related” action) together.
(4) You and I mutually believed – or at least shared the belief – that
(1), (2), and (3).
(5) (2) in part because of (4).
In (2) we use the following analysis of I-mode shared we-attitude
analysis of intention:
You and I share (in the I-mode sense) the intention to jointly perform X
if and only if
1) I intend that we X in the reason-based we-attitude sense (viz. I
intend that we X in part because I believe that you intend in the reason-based
sense that we X and that we mutually believe that each of us so intends), and
2) you similarly intend that we X in the reason-based we-attitude sense.
(Space does not allow a discussion of the above I-mode account.)
[36] See Bratman
(1999), p. 114.
[37] Also Bratman’s case of ”unhelpful singers” indicates that his account is an I-mode account (see Bratman, 1999, p. 103): The singers are assumed to jointly intend to sing a duet without helping the other one if he, contrary to expectations, does something wrong. In contrast, a we-mode joint intention is necessarily (minimally) cooperative, as also the inference schema (W2) of the previous section shows. However, no actual inference is needed because we-intenders typically have internalized forgroupness and the substantive content of (W2) so that it has become a kind of skill.
[38] See Tuomela (2005b),
especially Chapter 1, for the points presented in this paragraph.
[39] See Tuomela (2005b) for
my reasons for not taking groups to be agents in an ontic sense.