3.1. The motivation argument

The Discontinuity Thesis for praiseworthiness can be formulated as follows:

According to (DT_P), discontinuities between corporate praiseworthiness and member praiseworthiness are possible. As in the case of blameworthiness, this entails that the former cannot be reduced to the latter. Our argument for this thesis has the same structure as the one that supports (DT_B).

The motivation argument

1. 1. A collective moral agent is morally praiseworthy for a corporate action.

2. 2. The members are not morally praiseworthy for contributing to the corporate action, because they did not perform their action for the right reason.

3. 3. Hence, the collective moral agent is morally praiseworthy for its action without its members being praiseworthy for contributing to it.

The motivation argument revolves around the idea that, to be praiseworthy, an agent must be appropriately motivated. This does not hold for blameworthiness. Even an indifferent agent can be blameworthy. In this respect, there is an important asymmetry between praise and blame, known as “the Praise-Blame Asymmetry” (Hindriks, Reference Hindriks2008). Lacking appropriate motivation is a defeater for praiseworthiness, but lacking inappropriate motivation is not a defeater for blameworthiness. Due to the Praise-Blame Asymmetry, the Motivation Argument is unique to praiseworthiness. Hence, it differs from the arguments that have and can be offered for (DT_B).

Our core claim is that corporate praiseworthiness is irreducibly collective because corporate motivation is. So, we need to establish that corporate moral motivation can be discontinuous from member motivation (DT_M):

In other words, it must not be the case that, necessarily, an appropriately motivated collective agent has appropriately motivated members. If (DT_M) is true, it is possible for Premise 1 and Premise 2 to be true simultaneously. In this way, (DT_M) supports the Motivation Argument.

To defend (DT_P), we pursue a dual strategy. In Section 4, we defend the two premises of the Motivation Argument directly. To this end, we present an account of moral concern. This supports the idea that collective moral agents can have the motivation needed for being praiseworthy. In the remainder of this section, we offer an indirect argument for the irreducibility of corporate moral motivation (DT_M). This ‘Self-Interest Argument’ provides indirect support for Premises 1 and 2 along the lines just discussed. We complement it with an account of corporate reasons.

We are not the first to defend (DT_P). Jessica Brown (Reference Brown2024, 120) supports it by means of the following example. A charity does the right thing. Yet, its members contributed to this corporate action for the wrong reasons. For instance, one of them did so to impress his girlfriend. As charities cannot have girlfriends, this reason does not transpose to the group agent. Now, suppose that the charity did the right thing for the right reasons. If so, it is praiseworthy, even though its members are not. Brown takes this to imply that corporate reasons are also irreducibly collective (DT_M).

Our strategy has several advantages over Brown’s relatively short discussion of praiseworthiness. First, we do not just provide examples, but we support (DT_P) by presenting an account of corporate moral concern. Furthermore, we present a separate argument for (DT_M). Brown (Reference Brown2024, 120) defends the irreducibility of corporate reasons partly via considerations about the praiseworthiness of collective agents. We believe it is important to establish the irreducibility of corporate reasons independent from considerations of praiseworthiness.

To simultaneously defend Premise 1 and Premise 2, we rely on the claim that corporate motivation is irreducibly collective. But we cannot then rely on considerations about the praiseworthiness of collective agents to establish the irreducibility of motivating corporate reasons, because this would make our argument circular. Finally, Brown assumes “the Right Reasons View,” according to which an agent is morally praiseworthy for doing the right thing exactly if they did it for the right reason. In contrast, our proposal is not tied to a particular conception of praiseworthiness. As we discuss in Section 4, it is robust across several different conceptions.

3.2. The Self-Interest Argument

According to neoclassical economic theory, the goal of a firm is to maximize its profits (Tirole, Reference Tirole1988). Yet, employees maximize their own utility. Their motivation increases along with their pay and decreases with the effort they have to exert. Employees are not motivated by the profit the firm makes. That is what the shareholders care about. It follows that a firm and its employees are motivated by different things. The former by profit; the latter by their paycheck. Because of this divergence, employees need to be incentivized such that their motivation comes to align with the goal of the firm. This idea has been developed in terms of the principal-agent theory. The firm needs to hire a manager who monitors the employees and makes sure they do not shirk or slack off. Consequently, members do what is in the interest of the firm, but for reasons of their own.Footnote ¹¹

This Self-Interest Argument supports (DT_M). It can be combined with a widely accepted account of motivating reasons. They are psychological states that explain why the agent intended or performed an action. And they help us see what motivated the agent: they rationalize the intention or action from the perspective of the agent (Davidson, Reference Davidson1963). Motivating reasons consist of a pro-attitude towards actions with a certain property paired with a belief that the specific action (under some description) has that property. Under pro-attitudes we may include desires, goals, moral views, economic principles, public or private goals, and values insofar as we can understand these as attitudes directed towards actions of a certain kind (Davidson, Reference Davidson1963, 686). In virtue of having a motivating reason to Phi, the agent is in a state that is potentially explanatory of his Phi-ing (Smith, Reference Smith1987, 38). We further understand intentions as sui generis mental states, not reducible to beliefs or desires (Bratman, Reference Bratman1987). We form intentions as responses to reasons, meaning intentions are interposed between motivation and actions.

We think the same is true for collective agents. The core idea is that nothing stands in the way of collective agents to adopt a pair of such psychological states via its decision-making procedures. These psychological states can lead to collective intentions and collective actions. When a collective agent adopts a pro-attitude towards actions of a certain kind and a related belief that identifies certain actions with such properties, then this is sufficient for attributing a motivating reason to the collective agent.Footnote ¹² Its actions can be explained in terms of such motivational states.

But what does this mean in practice? Imagine that an investment firm performs a hostile takeover. Suppose that the investment firm’s long-standing corporate policy specifies this as an effective means for pursuing profit maximization. The firm may have created internal roles for members to scout potential targets and a corporate policy that specifies the relevant criteria for the targets. If so, the collective agent has adopted various means-end beliefs about act-types that are embedded in its corporate policy, and it authorizes members to identify and perform act-tokens that fall under these act-types. By hypothesis, the employees perform their contributory tasks purely out of self-interest. They just want their paycheck. But while the members’ agent-relative motivating reasons rationalize their behavior, they do not rationalize the firm’s behavior. The reason why the investment firm performs the hostile takeover is that it will be conducive to maximizing its profits. In sum, both a collective agent and its members can be motivated by self-interest. If our argument is correct, this implies that the motivation of the former comes apart from that of the latter.

This confirms the conclusion of the Self-Interest Argument: corporate motivation is irreducibly collective (DT_M). Furthermore, it reveals that the notion of a motivating corporate reason is needed to give substance to this idea (French, Reference French1984; Tuomela, Reference Tuomela2012). The function of such reasons is to pick out the relevant corporate motivational states that help explain corporate action. In view of this function, we highlight two differences between our account and Brown’s (Reference Brown2024). First, in contrast to that of Brown, our account allows for motivating reasons that do not result in action. Second, we take motivating reasons to be less closely connected to normative reasons than she does. Let us explain.

According to Brown: “A group acts for a reason r if and only if r causes the group’s action where that manifests the group’s disposition to act for normative reasons of that kind” (2024, 134). This account is restricted to motivating reasons that lead to action. However, they often do not. The investment firm may have intended to undertake a hostile takeover, but found themselves in a state of illiquidity at the time of action. Our account allows for motivating reasons that are not strong enough to cause an action.

Furthermore, Brown assumes that motivating reasons are propositions. Yet, she wants to remain neutral on whether they are factive (2024, 117). At the same time, she claims that having a motivating reason implies a disposition to act for normative reasons. But normative reasons are factive. Hence, Brown links motivating corporate reasons to factivity. This is problematic insofar as agents can act on false beliefs.Footnote ¹³ Suppose the target of the investment firm is not actually a profitable choice for a takeover, and so the takeover is not an effective means for profit maximization. There is no agent-relative normative reason for the firm to undertake this specific takeover, and yet the corporation is disposed to act exactly the same as when the proposition would be true.Footnote ¹⁴ Surely, when the investment firm completes the takeover, this is still a corporate action done for a corporate reason.Footnote ¹⁵ Thus, linking motivating reasons too closely to normative reasons impedes their explanatory function.

3.3. Objections and replies

At this point, the individualist might object that there must be a tighter connection between the motivation of the collective agent and that of its members. The underlying intuition might be that there can be no free-floating corporate motivation. This objection can be developed in several ways.

The first concerns supervenience. Plausibly, the social supervenes on the individual. Once the individual facts are fixed, so are the social facts. Furthermore, any change at the social level is accompanied by a change at the individual level. Yet, the social level can remain the same even if the individual level changes. The attitudes of a collective agent, including its beliefs and desires, supervene on those of its members. If so, how could corporate motivation possibly come apart from member motivation?

However, corporate attitudes do not supervene on member attitudes only. The supervenience base of corporate attitudes also includes the collective decision procedure. In fact, it has been argued that, if they are to be rational, the mental states of collective agents are bound to diverge from those of their members (List & Pettit, Reference List and Pettit2011). Hence, the supervenience base of corporate attitudes must extend beyond member attitudes. Now, perhaps there are other ways in which a collective agent can acquire attitudes. But this only means that the supervenience base is more complex yet, and the objection fails.

The second objection concerns collective acceptance. For a collective decision procedure to be in force, its members must accept it. Furthermore, they are thereby committed to its outputs, the attitudes formed by the collective agent (Tuomela, Reference Tuomela2013). This implies that if a collective agent makes a decision, its intention should somehow be reflected in the attitudes of the members. This seems to be in tension with a discontinuity between corporate and member motivation.

Yet, even if corporate intentions should be reflected in member attitudes, this does not mean they actually are. This is perhaps a requirement of rationality, but such requirements can be violated. Imagine that a collective agent decides to regard profit considerations as decisive for what it is going to do next. This would be reflected in the attitudes of the members if they were moved to do their part by those very same considerations. However, suppose that they in fact do so for reasons of their own. At most, it follows that members are not always fully rational and may fail to honor their commitments.

The third objection concerns the reason for which the collective agent decides to do something. As Michael Bratman (Reference Bratman2022) points out, the members might disagree about this. Even though they agree about what to do, their disagreement about why can be so widespread that there is in fact no reason for which it is adopted. So, on his view, a collective agent can perform an action for no reason at all. It is far from obvious that this is possible. But if it is, the collective agent will not be praiseworthy for such actions (Hindriks, Reference Hindriks2023).

So, if a collective agent is to act for a reason, there must be at least limited agreement among its members. This implies that a collective agent has a motivating reason only if at least some members support it. However, members can support a consideration as the reason of the collective agent without themselves being motivated by it. They might acknowledge a consideration as the reason for which the collective agent acts but perform their parts for different reasons.

As Bratman (Reference Bratman2022) recognizes that collective agents often act on the basis of policies. These are standing intentions to perform an action whenever a particular situation arises. Whenever a collective agent acts on the basis of such an intention, it does so for the reasons why it was adopted. However, again, members who contribute to the action need not do so for this reason. Members may not have been members when the policy was adopted. Or even if they agreed with the policy when it was adopted, they might have changed their mind. Or they might be alienated from the collective agent and do their part only to collect their paycheck (Bratman, Reference Bratman2022; Shapiro, Reference Shapiro2011).

Besides procedures, policies and organizational structure, one might think that the culture or ethos of a collective agent might play a substantial role in its decision-making process and generating corporate action. Seumas Miller proposes that the culture of an institutional organization “comprises the informal attitudes, values, norms, and ethos or ‘spirit’ that pervades an institution; importantly, it includes the extent and degree of institutional loyalty.’ (Miller, Reference Miller2010, p. 26) In light of this, a fourth objection is that collective agents can form attitudes in other ways that do not allow for discontinuities.Footnote ¹⁶ For instance, Bratman (Reference Bratman2022) considers the idea that agreement about a reason for action among members induces a reason for the collective agent. Furthermore, Kendy Hess (Reference Hess2014) proposes that collective agents can form mental states by means of distributed decision-making and by means of a “culture shift.” Some might worry that, in these cases, member attitudes do correspond to those of the collective agent. And they might take this to mean that these corporate attitudes are reducible after all.

For the sake of clarity, we distinguish between the ethos of a collective agent and the culture within it. The latter is more plausibly attributed to the members. It encompasses the informal attitudes that prevail among the members as well as the loyalty they have towards the collective agent. Although they can influence corporate decision-making, they do not constitute corporate attitudes. As they exist only at the member level, the question of discontinuity does not even arise.

The corporate ethos does belong to the collective agent. It encompasses its constitutive norms and values.Footnote ¹⁷ Like policies, they are general. Some maintain that collective agents have a rational point of view (RPV) that encompasses their ethos along with membership rules, policies and other outputs of decision-making procedures (Hess, Reference Hess2025; Rovane, Reference Rovane1997). This RPV enables the collective agent to (largely) preserve the coherence of its attitudes. However, this corporate RPV is not to be found in any member, nor does it rely on any single member to be instantiated (Hess, Reference Hess2025). Members have a RPV of their own. Because of this, they face different constraints and pressures. Their RPV’s are almost bound to diverge from that of the collective agent.Footnote ¹⁸ Moreover, members can acknowledge the corporate RPV and perform their parts of a corporate act without being motivated by features within the corporate RPV. Hence, the corporate RPV simply involves additional action-generating mechanisms by which a collective agent may come to have commitments about fact and value that diverge from its members (Hess, Reference Hess2025).

Thus, the Self-Interest Argument survives the challenges from supervenience, acceptance, agreement, and ethos. Both the structure and ethos of a collective agent can give rise to discontinuities between corporate and member motivation. This completes our argument for (DT_M). Next, we present our direct argument for (DT_P). This requires us to say more about praiseworthiness and the function of praise.

4.1. The Part-whole objection

To be praiseworthy, the collective agent must have had the appropriate moral motivation. Now, suppose that Premise 2 of the Motivation Argument is true. None of the members acted on morally appropriate motivation. And yet the collective agent is somehow praiseworthy according to Premise 1. But how, someone might object, can the whole be morally better than its parts? The intuition is that the fact that the motivation of the collective agent is appropriate must be due to the moral quality of the motivation of the members.

This “part-whole objection” can also be formulated in slightly different terms. For instance, an agent with morally appropriate motivation has a good quality of will. This raises the question how a collective agent can have a good quality of will if its members do not. Similarly, an agent with morally appropriate motivation has enough moral concern. But how can a collective agent have enough moral concern if its members have none? The first question reflects the part-whole objection on the assumption that a corporate will is a whole that has the wills of the members as parts. The second does so assuming that the degree of corporate moral concern is the sum, or more generally a function, of the moral concern of the members.Footnote ¹⁹

Brown (Reference Brown2022, 502) seems to have the part-whole objection in mind when she considers two versions of an example concerning a company that decides to buy safety equipment, which is the right thing to do. In the first version, the members support this decision for the right reasons, including that the equipment is cost-effective. The second version differs from this only in that the members do not believe that this is the case, but vote in favor of this premise because they have been bribed to do so. One might think that in this second version, the collective agent does not act for the right reason. The idea must be that the fact that the members have been bribed somehow taints the motivation of the collective agent.

If the part-whole objection hits its target, then a collective agent cannot really act for the right reason if its members do not. If members are not suitably motivated, any appearance that the collective agent is praiseworthy must be illusory. When formulated in these terms, it becomes apparent that the objection resembles the supervenience objection. And this we have answered already. Apart from the fact that the part-whole objection is formulated in terms of moral motivation, it pushes the issue one step further, at least on a charitable interpretation. An objection that has the form ‘How can this be?’ can be answered in terms of the process, the structure or the mechanism that explains the phenomenon at issue. To answer this, we develop an account of corporate moral concern. Corporate moral concern helps us see how collective moral agents can indeed be irreducibly praiseworthy.

4.2. Moral concern and normative policies

To prepare for our account, we first say more about moral competence and collective agency. A morally competent agent can think, decide, and act on the basis of moral reasons. Following John Martin Fischer and Mark Ravizza (Reference Fischer and Ravizza1999), we distinguish two aspects of moral competence: receptivity and reactivity. Both concern the extent to which an agent is sensitive to moral reasons. But receptivity concerns the decision-making process, whereas reactivity pertains to the process of enacting the intentions of the agent. An agent who is both receptive and reactive to moral reasons is responsive to them.

According to procedural collectivism, collective decision procedures are crucial for collective decision-making and collective action. For a collective agent to possess moral competence, these procedures must allow for moral considerations to enter the decision-making process and evidence-gathering (see de Haan, Reference de Haan2023a; Hindriks, Reference Hindriks2024; Pettit, Reference Pettit2007). This, in turn, requires that some of the members are moral agents who are themselves responsive to moral reasons. This makes it at least possible for such moral considerations to be translated into relevant collective decisions and actions.

However, these conditions do not yet ensure that the collective agent is systematically sensitive to moral reasons. This raises the question of whether they suffice for collective moral agency.Footnote ²⁰ Here, we focus on the more specific question of whether it suffices for corporate moral concern. We propose that three features are crucial for understanding how corporate moral concern functions: (1) procedural policies, (2) roles and membership rules, and (3) operational policies. Each of these features can be normatively informed.

First, for a collective agent to be adequately motivated by moral reasons, it must have adopted normative policies of valuing.Footnote ²¹ An agent with moral concern has other-regarding values. A collective agent can acquire such values by freely adopting normative procedural policies to attribute weight to particular values when deliberating. These procedural policies serve to systematically bring normative considerations to bear on the decisions of the collective agent (Hindriks, Reference Hindriks2018, Reference Hindriks2024). During the deliberative process, members apply such policies by attributing a particular weight to relevant considerations corresponding with the values. The dispositions of members are crucial for the extent to which such policies bear on the decision-making process of the collective agent. If the members collectively accept such policies, the collective agent is disposed to attribute weight to the corresponding other-regarding reasons. It is important that such policies are adopted not only regarding the centralized decision-making processes, higher up in the collective agent, but also with respect to lower-level decision-making processes in sub-groups.

To give an example, consider a chocolate company. Let us assume that the chocolate company morally ought to ensure that its production chain does not involve child labor, resource depletion, or exploitation. For example, the fact that deforestation is a significant cause of global warming is a moral reason to refrain from buying products from suppliers that are involved in significant amounts of deforestation. Or the fact that laborers have a right to a fair wage is a moral reason to avoid cooperation with suppliers that engage in exploitative practices. To support these ends, the company can adopt procedural policies to ensure that values such as justice and sustainability will be considered when it deliberates about new courses of action. Crucially, those policies should function properly even if the members do not favor the relevant values. Even if they do not, the members should be required, if not incentivized, to take the relevant considerations into account. Such procedural policies should be installed at different levels of decision-making. The way the collective agent is organized matters a great deal for how well it functions morally, for how robustly it responds to moral reasons.

However, this is only part of the story. Moral agents are not only receptive to moral reasons (when making decisions) but also reactive (when enacting them). This reveals that an account of collective agency purely in terms of collective decision procedures is incomplete. Inspired by French (Reference French1984), we propose that it also includes a system of roles and a membership rule. Collective agents make decisions by means of collective decision procedures. Their members enact those decisions primarily because of the roles they have. Each role comes with various tasks and a set of recognized powers. Who the members are and what these powers are is settled by the membership rules. For those rules to be in force, the members must generally accept or at least acknowledge them. Together, these rules form what we call “the corporate structure.” Thus, collective agents are constituted by individuals who support a structure of rules. In short, a collective agent is a realized structure of rules (Hindriks, Reference Hindriks2025b).

Importantly, these membership rules themselves can be normatively informed. The relevant role-bearers can be allocated various powers that are conducive to securing the relevant other-regarding values. For example, the Corporate Responsibility Officer has the role to systematically check for the presence of exploitation and the Chief Operating Officer has the power to unilaterally halt production when exploitation is part of the production chain. The way the collective agent has delineated roles and powers can reflect some degree of corporate moral concern, even if members fail to do their job properly. Even if the Corporate Responsibility Officer fails to recognize the exploitation and the Chief Operating Officer does not halt production, the fact that the collective agent has designated a member to detect exploitation and empowers another member to halt production, potentially at the expense of profit, does suggest that the collective agent is genuinely concerned with avoiding exploitation in its production process. It just failed to do so this time. Hence, the way the corporate structure is realized can reflect various degrees of moral concern.

The notion of a corporate structure in general and that of a system of roles in particular can further be used to shed light on how corporate moral concern affects action. If a collective moral agent is to function well, it also needs to have normative policies that bear on its actions. We call these operational policies. Such operational policies specify various act-types that the group is to perform to secure its goals and values. They guide the members when enacting a corporate decision. The operational policies can constrain the options members have or require them to initiate certain actions on behalf of the collective agent. When applying them, the members pick out act-tokens that fall under one of the targeted act-types. Those actions are targeted at realizing the goals and values of the collective agent. Remember the policies of the investment firm that seeks out hostile takeovers to maximize profit. Here, the members enact such operational policies that are geared towards profit-maximization. But when such operational policies are appropriately normatively informed, they direct agents to perform act-tokens that secure the relevant moral values.

Suppose the chocolate company appropriately values sustainability and justice. It can implement normative operational policies that require it to stop buying cocoa beans from any given producer that engages in deforestation or exploitation. Such policies should be made specific such that the members know what to do under which conditions. For example, if a member receives evidence that a supplier engages in deforestation or exploitation, the member must file an internal report, which, if confirmed, results in a termination of the contract by the Chief Operating Officer. Likewise, the company could install policies that ensure there are unexpected inspections for both prospective and current suppliers. Due to policies such as these, the company is set to perform act-types that together secure the values of sustainability and justice.

Thus, a collective agent with moral concern will have normative policies that bear on both its decision-making and action-taking. They enable it to be receptive as well as reactive to moral reasons. Moral concern comes in degrees. It is crucial that the organization’s procedural and operational policies and role division are appropriately integrated. When these do not work in tandem, this may inhibit moral concern. For example, if the Chief Operating Officer does not have the power to unilaterally halt production, and the decision must be approved by the board of directors, then the company may be unable to respond in time to prevent being complicit in human rights violations. While the policies reflect some degree of moral concern, due to the poor integration, the company lacks an adequate degree of moral concern.

4.3. Corporate praiseworthiness

The proposed account of corporate moral concern provides support for the claim that a collective moral agent can be praiseworthy for an action. This is Premise 1 of the Motivation Argument (Section 3.1). To explain how, we consider various conceptions of praiseworthiness. Notably, some argue that praiseworthiness does not require the kind of control that moral blameworthiness requires (Nelkin, Reference Nelkin2011; Wolf, Reference Wolf1990; cf. McKenna, Reference McKenna2012). If so, there is an important asymmetry regarding blameworthiness and praiseworthiness (see also Stout, Reference Stout2020). For readers who think praiseworthiness requires control, we encourage them to plug in their favored account of group control.Footnote ²² In the remainder, we focus on other possible conditions for praiseworthiness.

According to the Right Reason View, which we briefly discussed in Section 3.2, an agent is praiseworthy exactly if they did the right thing for the right reason (Arpaly, Reference Arpaly2002Arpaly, Reference Arpaly2003Markovits, Reference Markovits2010Markovits, Reference Markovits2012). An agent with appropriate moral concern can act for the right reason. Nomy Arpaly (Reference Arpaly2003, 84) explicates moral concern in terms of an intrinsic desire to act morally. To do the right thing is to act on the basis of this desire. Yet, Arpaly does not insist that the agent conceives of the action as morally right.Footnote ²³ All that is required is that the desire motivates the performance of an action that is in fact morally right. According to Julia Markovits (Reference Markovits2010, Reference Markovits2012), an agent acts for the right reasons exactly if her motivating reasons coincide with the justifying reasons. Our account of corporate moral concern is neutral between these two conceptions. Given that collective moral agents can have moral concern, it follows that a collective moral agent can indeed be praiseworthy.

But what does this mean in practice? Suppose that our chocolate company stops buying cacao beans. It does so in order to avoid facilitating cacao bean suppliers who exploit the farmers. What is the reason for which the chocolate company does so? The company has adopted a pro-attitude towards fair trade and several policies that identify various actions that secure that value. Hence, the company’s reason is that the producers engage in unfair trade. In fact, they violate human rights by exploiting its workers. Furthermore, it recognizes that this reason is sufficiently strong such that it would be morally wrong to continue buying beans. Thus, the chocolate company performs the morally right action for the right reason. Hence, on the Right Reason View, it is praiseworthy to discontinue the trade in cacao beans.

In contrast, suppose that another chocolate company has none of these normatively informed policies and membership rules. The company does not have any concern for justice or sustainability but is first and foremost concerned with profit-making. Suppose that this company too stops buying cocoa beans from the very same supplier, which is the right thing to do. However, this collective agent has adopted the belief that the supplier’s cocoa beans are too pricey given the market price. The reason that rationalizes the corporate act is profit oriented. This company is not praiseworthy for halting to buy beans from the producer, precisely because this collective agent does not act for the right reason.

Although having the appropriate motivation is necessary, not everybody regards it as sufficient for being praiseworthy. We consider three alternatives: the Exceptionality View, the Knowledge View, and the Robustness View. We argue that collective agents can satisfy the conditions of each of them. This provides further support for Premise 1 of the Motivation Argument.

According to the Exceptionality View, a praiseworthy agent does not merely have appropriate motivation. There must be something exceptional about the action or the agent’s motivation (Darwall, Reference Darwall2006; Eshleman, Reference Eshleman, Shoemaker and Tognazzini2014; McKenna, Reference McKenna2012; Stout, Reference Stout2020; Wallace, Reference Wallace1994).Footnote ²⁴ The chocolate company might, for instance, pay above-market prices to its suppliers to ensure their employees have good working conditions. This makes it more difficult for them to survive the competitive process. Because of this, we assume that this qualifies as exceptional motivation. So, the conditions of the Exceptionality View can be met.

Furthermore, the Knowledge View requires that the praiseworthy agent knows that they acted for the right reason. For example, Paulina Sliwa (Reference Sliwa2016), 394) defends the Rightness Condition: a morally right action has moral worth if and only if it is motivated by concern for doing what’s right and by knowledge that it is the right thing to do. Michael McKenna (Reference McKenna2012) requires that the agent know that the action is exceptional. Hence, his hybrid account combines the Knowledge View and the Exceptionality View.

To show that collective agents can meet these conditions, suppose that the chocolate company faces an instance of exploitation in its supply chain. The chocolate company has adopted a normative policy that forbids engaging in any actions that support exploitation, because the corporation believes that this is morally wrong. The corporation may further have operational policies that specify when and what is to be done their supply chain involves exploitation. This means the chocolate company has adopted several appropriate truth-tracking beliefs that constitute group knowledge (List, Reference List2005). The corporation knows that it is morally wrong to exploit workers and it knows what kind of actions would fall under exploitation. The company recognizes the fact that the supplier severely underpaying the farmers is an instance of exploitation. Hence, the collective agent knows that this action is morally wrong because it would result in exploitation. And because of this, the collective agent decides to halt trade. The collective agent does the right thing for the right reason and it knows this is the right thing to do. So, the conditions of the Knowledge View can be met as well.

Finally, the Robustness View requires that the agent’s motivating reason is sufficiently robust. If it is to be praiseworthy, it must do the right thing for the right reason, not only in the actual world, but in a range of nearby possible worlds (Fearnley, Reference Fearnley2022). This will be the case if the policies due to which the collective agent has the appropriate motivation are sufficiently coherent and resilient. The fair-trade policy of the chocolate company must be robust to increases in the price of cacao beans or to decisions of its shareholders to divest. Although market pressures can be high, it is possible for collective agents to exhibit this kind of commitment.

However, in Section 3.3, we observed that corporate moral motivation might be fragile if insufficiently supported by the members. Suppose our chocolate company sets out to fight the exploitation of cacao bean farmers. But, due to a culture shift, the members no longer favor this. The new culture within the company is one of indifference towards the issue within the company, perhaps fueled by racist attitudes. This makes it difficult for the company to stay on course. Its moral motivation becomes increasingly fragile until it collapses. At this point, the chocolate company may drop its fair-trade policy. But even if it retains its policy, its moral motivation is likely insufficiently robust for it to be praiseworthy. At the same time, if an individual agent were to be similarly conflicted, it would not meet this standard either. So, the question is whether corporate moral motivation might be sufficiently robust if the culture that prevails among its members is less antagonistic to begin with. Suppose employees are loyal to the company for reasons that have little to do with the moral quality of its decisions. Because of this, they are disinclined to challenge its policies. In such a situation, the moral motivation of the company might be sufficiently robust for it to be praiseworthy.

But what if employees have a lot of autonomy? They can interpret their roles in a flexible manner. And they might have discretion over how to apply the procedures and policies that are in force, especially when procedural and operational policies are not overly specific. If so, it seems that praiseworthy corporate decisions will require admirable choices on the part of individuals exercising this discretion. In particular, in leader roles such as the CEO and CFO. Now, it is not a given that the collective agent is actually praiseworthy in such a case, this depends on the degree of moral concern. Moreover, though admirable choices of employees may well enhance the moral quality of a corporate action, they are not needed for the company to be praiseworthy. Strikingly, this is supported by the Self-interest Argument. It reveals that employees are compensated so as to align their motivation with that of the company. This holds not only if its goal is to maximize profit, but also if it is a laudable goal such as reducing exploitation.

On our account, members play a crucial role in realizing corporate moral concern. Yet, the collective agent can have moral concern with respect to a particular issue even if the members do not. For instance, a collective agent can subscribe to the value of justice while its members are indifferent to it. What is required instead is that they are sufficiently motivated to enact those policies. And they might be motivated to fulfill their roles by other considerations, such as money, esteem or status. This reveals that corporate moral concern is not a whole that has the moral concern of the members as parts and it can be sufficiently robust for corporate praiseworthiness.

Thus, collective moral agents can be morally praiseworthy on all four conceptions of praiseworthiness. This puts Premise 1 of our Motivation Argument on firmer ground. In Section 3.2, we argued that employees often do their work for self-interested reasons. This supports Premise 2. Hence, this completes our argument for (DT_P).

4.4. The function of praise

Thus far, we have argued that collective agents can be moral agents and have moral concern, and they can meet the prerequisites of praiseworthiness. One might nonetheless wonder whether the nature of praise makes any difference to the fittingness of praise directed toward collective agents.

For most theories of praise, we think it is evident that collective agents are fitting targets as soon as one accepts that they are distinct moral agents with their own beliefs, desires, intentions, and actions.Footnote ²⁵ For example, according to judgment views, to praise someone is to judge the agent praiseworthy (Zimmerman, Reference Zimmerman1988). Action views identify praise with treating the agent in some beneficial manner (Nowell-Smith, Reference Nowell-Smith1954). And emotion views identify praise with reactive attitudes or interpersonal emotions such as approbation and gratitude (Macnamara, Reference Macnamara2011; Strawson, Reference Strawson1962). Of course, the overt dyadic praise interaction may look somewhat different, because one typically will have to engage with a representative of the collective agent, and the audience is not identical to the target of praise. But other than this, if the target qualifies as a moral agent, we think the praise interactions themselves will not be wholly different from those with individuals. We can judge that the chocolate company did something praiseworthy. We can perform various actions, say writing a positive opinion piece in the newspaper, to express our approbation towards the company. There does not seem to be any problem with collective moral agents being the target of such praise.

However, certain functional theories of praise may seem to pose some difficulties. The idea that the function of praise is communicative is widespread: praise aims to communicate something to its target, which might simply be that the praisee is praiseworthy, and it is successful if uptake is secured (Darwall, Reference Darwall2006; McKenna, Reference McKenna2012; Watson, Reference Watson and Schoeman1987). In addition to this communicative purpose, praise may well serve a broader function within a moral community. For instance, Daniel Telech (Reference Telech and Shoemaker2021) proposes that praise functions as an invitation to engage in shared valuing. And Jules Holroyd (Reference Holroyd2024) proposes that its function is to entrench moral norms and values and exert pressure in relation to those norms by signaling the commitments of the praiser.Footnote ²⁶ In view of such communicative and social functions, do collective agents qualify as suitable targets of praise?

Can the communicative function of praise be fulfilled with collective agents as targets? Michael McKenna (Reference McKenna2006) argues that it is inappropriate to direct blame at them, because they lack emotions.Footnote ²⁷ The underlying idea is that blaming an agent is a matter of expressing a reactive attitude, such as resentment or indignation, and communicating a moral demand. But this makes sense only if the target can understand the meaning of this emotional communication. To be a morally responsible agent, the agent must be a capable interlocutor in these “conversational exchanges.” Blaming the collective agent will be unsuccessful as a communicative act, because there can be no uptake. They are not capable of understanding this emotional communication. They lack sensitivity to the reasons concerning the shared interpretive framework in which members of the moral community assign salience to types of action (McKenna, Reference McKenna2006, 29). If successful, this argument generalizes from blame to praise. A further implication would be that collective agents cannot be moral agents after all, because they are incapable of having emotions.

Now, we grant that collective agents cannot have feelings (List, Reference List2018). But this does not mean that they cannot appreciate the significance of reactive attitudes. Collective agents can have functional equivalents of emotions, including the reactive attitudes (Björnsson & Hess, Reference Björnsson and Hess2017). To mark that they do not have an affective component, we call them “quasi-emotions” (Hindriks, Reference Hindriks2018). And quasi-emotions suffice for being a suitable target of praise. A more combative response is to deny that emotions are required at all for collective moral agency (2). Collective agents can have moral understanding without having emotions, because this capacity supervenes on member abilities, who do have such emotions. And most of our theories of praise and blame are overly anthropomorphized, because their proponents have not paid enough attention to other kinds of agents, including collective agents.Footnote ²⁸

So, praise can serve its communicative purpose with respect to collective agents. Can praise also fulfill its social function? Let us consider Holroyd’s (Reference Holroyd2024) view in more detail. Praise is directed at an action that complies with some norm or value. By praising someone, the praiser signals that she is committed to that norm or value. Furthermore, a praiser exerts pressure on others to comply with the moral demand or standard. In this way, praise serves to entrench norms and values.

This account readily transposes to collective moral agents. As just discussed, they can be the target of praise. They can be committed to norms and values. And they can increase their commitment to them in response to praise. Consider the news article again, such a public expression of praise affirms the importance of sustainability and justice in the corporate environment, and it may exert pressure on individual and collective agents to affirm these values too. Furthermore, collective agents can praise others. They can signal their own commitment to norms and values. In doing so, they can exert pressure on others to do likewise.

Similarly, they can, as Telech (Reference Telech and Shoemaker2021) proposes, engage in shared valuing. To be sure, they will do so by means of some of their members, such as CEO’s and spokespersons. But these need and should not push their own view in the process. Instead, they have a role responsibility to represent the view of the collective agent. They can experience pressure to adjust their own view if it diverges from that of the praiser. But this still does not guarantee that their motivation coincides with that of the collective agent.