A Challenge for the Scaffolding View of Responsibility

Abstract

According to Victoria McGeer’s “scaffolding view” (SV) (McGeer 2019), responsibility is a matter of moral reasons-sensitivity (MRS) which, in turn, requires only a “susceptibility to the scaffolding power of the reactive attitudes, experienced as a form of moral address” (2019: 315). This claim prompts a prima facie challenge: doesn’t this susceptibility lead to doing the right things for the wrong reasons? Although the SV offers a nuanced and sophisticated answer to this challenge, one that moreover respects the social nature of moral knowledge and the fragility of moral motivation, it does not succeed. It redefines MRS to fit our responsibility practices in a way that overlooks our (fragile) capacity for “genuine MRS.” The first and primary objective of this paper is to contrast SV-MRS with genuine MRS. The second objective is to suggest that rather than redefining MRS (which is both unwarranted and costly), we should accept that there is a gap between our practices (and thus responsible agency) and genuine MRS.

1 Introduction

Recent instrumentalist¹ and communicative accounts² of responsibility alike take moral reasons-sensitivity (MRS) to be the hallmark of responsible agency. They also share a view of the responsibility practices³ as playing a vital role in establishing and affirming this sensitivity. Crucially, whereas proponents of the communicative account take MRS to be a precondition of our practices, understood as moral address, instrumentalists claim that our practices bring about MRS. Instrumentalists begin with the assertion that MRS is the end point of our practices and develop a model to explain this outcome. This paper rejects that strategy in favor of building on an investigation of our practices. The result of this investigation is that our practices do not necessarily track or bring about MRS. Establishing this result is the principal aim of this paper. If correct, it challenges the core premise of both accounts – that MRS is a necessary condition of responsible agency.⁴

I focus on a specific instrumentalist account, Victoria McGeer’s recent moral influence account, the “scaffolding view” (SV) (McGeer 2019). The SV incorporates the model at work in the communicative account by way of an idealized moral address, “the ideal reactive exchange.” I proceed on the assumption that this model offers the strongest possible defense of the relationship between our practices and MRS. The SV begins from the premise that responsibility is a matter of MRS. It then identifies the agential feature required for MRS as follows.

SV. Responsibility (MRS) requires only a “susceptibility to the scaffolding power of the reactive attitudes, experienced as a form of moral address” (McGeer 2019: 315).

This conception is both realistic and modest. It admits of the dynamic, fragile, and “socially negotiated” nature (or “ecological character”) of moral agency (McGeer and Pettit 2015; Vargas 2013) and aims at a “modest picture of what it takes to be a responsible agent” (McGeer 2019: 314–315). However, it also prompts a prima facie challenge: doesn’t this susceptibility lead to doing the right things for the wrong reasons?

To overcome this challenge, the SV must redefine MRS in such a way that it cannot make the important difference between social-prudential and moral reasons or between heteronomous and autonomous moral agency. Dependent on social feedback (qua the reactive attitudes) for both moral knowledge and moral motivation, SV-responsible agents need never act with moral worth⁵ or morally autonomously.⁶, ⁷ SV-responsibility is thus in conflict with what I will refer to as “genuine MRS”: an autonomous sensitivity to the right-making features or rightness of actions.

Distinguishing SV-responsibility (or SV-MRS) from genuine MRS occupies most of the space on these pages. With the remainder, I argue that rather than redefining MRS (which is unwarranted and costly, as I will show), we should accept that there is a gap between our practices and (genuine) MRS.

The paper proceeds as follows. §2 elaborates the prima facie challenge by offering an influence account according to which our practices require only prudential reasons-sensitivity. §3 presents the SV and identifies those claims which support the relationship between the reactive attitudes and MRS. §§4 through 6 assess them. §4 argues that reactive attitudes only clearly target sensitivity to a limited set of moral reasons. §5 investigates whether and how reactive attitudes could sensitize agents to a broader class of moral reasons. It shows that although they may be necessary to this end, they are not sufficient. Moreover, it remains unclear whether the object of sensitization is the kind of reason that would suffice for genuine MRS. §6 discusses the conditions of the ideal reactive exchange and determines that whether one accepts that enhanced MRS obtains is ultimately a matter of what counts as moral reasons. The SV’s conception of moral reasons as tied to the reactive attitudes is supported by reference to Bernard Williams’ view. §7 challenges that view and contrasts SV-responsibility with genuine MRS. §8 concludes and suggests how to proceed in light of these findings.

2 The Prima Facie Challenge

The original moral influence account (Schlick 1939; Nowell-Smith 1948),⁸ that of the “optimist” in Freedom and Resentment (Strawson 2008), has been characterized by Strawson and virtually every responsibility theorist since Strawson as – to put it mildly – having missed the whole point of our practices. By focusing on their effects, the optimist avoids the threat to responsibility posed by the thesis of determinism, but he also neglects their interpersonal, attitudinal nature and their commitments to desert, condemnation, and justice. The SV’s claim (McGeer 2019; see also Vargas 2013; Jefferson 2019; Milam 2021) addresses precisely that criticism. By showing the vital role that the reactive attitudes specifically play in “calling forth” responsible agency qua MRS, the moral influence account can accommodate the interpersonal-attitudinal dimension of our practices. What’s more, this revision can overcome the “economy of threats” characterization leveled at the original account (Wallace 1996), according to which our practices regulate only behavior via the equivalents of carrots and sticks.

The driving question of this paper is whether the SV’s claim is warranted. Do our practices really track or cultivate MRS? An affirmative answer cannot simply be inferred on the basis that they cultivate responsible behavior (behavior that aligns with our normative expectations), because it is possible that people only comply with normative demands for non-moral reasons. According to my version of the moral influence account, the “prudential reasons view” (PRV) – a view which I do not set out to defend here, but which I present as a useful foil to the SV – our practices are responses to the fulfilling, exceeding, and, more often, transgressing of normative demands and expectations, and they track and develop an agent’s capacity for meeting them. They do this by way of conditioning – pairing negative and positive social feedback with norm-relevant behavior (Gogoshin 2021a) – and by providing social-prudential (henceforth just “prudential”) reasons for norm-compliance. We might fulfill our responsibility-relevant obligations⁹ by acting to avoid actual or hypothetical blame, censure, resentment, or sanction, or to garner praise, approval, admiration, gratitude, or reward.

PRV. Responsibility (the capacity for meeting certain normative demands and expectations) requires only a susceptibility to (the prudential reasons provided by) our practices.

One need not be sensitive to the moral reasons (the right-making features or the rightness of actions) behind these obligations in order to be motivated to abide by them, but rather to (the reasons provided by) the practices themselves. Like the original influence account, the PRV holds that the responsibility-relevant capacity is susceptibility to our practices, only it, unlike the SV, takes this to be a matter of (a certain class and range of) prudential rather than moral reasons-sensitivity. For the SV to trump the potentially more straightforward PRV (and defeat the prima facie challenge), it would have to show that reactive attitudes amount to more than prudential incentives to comply with norms, e.g., or that we could not fulfill our responsibility-relevant obligations without MRS.

Consider the moral obligation to follow the speed limit. The moral reason behind this obligation is that not following the speed limit increases the likelihood of causing severe bodily harm and (less significantly) property damage. We may hold speedy drivers accountable by way of a legal and/or financial sanction. Drivers might well choose to follow the speed limit only in order to avoid this sanction rather than to avoid the harms that not doing so is likelier to bring about. Legal and financial sanctions potentially present different reasons than do reactive attitudes, however, and it is specifically the latter with which the SV is concerned.¹⁰ Only mature, responsible agents are thought to possess the capacity to be fitting targets of our full-blown reactive attitudes. According to the SV, this capacity is MRS and our reactive attitudes target, reflect, develop, and sustain it.

So consider instead the obligation to uphold promises. One might fulfill this obligation in order to avoid the resentment or other negative interpersonal consequences incurred by promise-breaking rather than because of what makes promise-breaking wrong (which I take to be other than because someone, even a hypothetical, idealized someone, will be resentful). On this interpretation, resentment is a mere substitute for legal/financial sanctions, but is this interpretation warranted? (1) Does (e.g.) resentment really present the same kind of reason as financial and legal sanctions do? (2) Does resentment stand a better chance than legal/financial sanctions at sensitizing us to specifically moral reasons irrespective of its status as a moral reason? I answer (1) below. I will come back to (2) in §5 after considering the SV’s positive answer to it.

Although sensitivity to reactive attitudes like resentment and sensitivity to legal, financial, and physical sanctions are different in important respects, and the first sensitivity seems clearly more morally relevant (especially to moral development) than the second, the reasons they concern are nonetheless prudential. Sensitivity to reactive attitudes is a markedly social sensitivity and concerns others’ well-being. Sociality and morality are interlinked. Even granting this linkage, however, it is possible to distinguish social from moral reasons.¹¹ Consider the previous case of promise-keeping. This obligation, at least on the (non-response-dependent) view of morality adopted here, is not justified by whether the promisee expresses or even feels slighted by the promise-breaking. We ought to refrain from promise-breaking not only because it will make someone feel bad (though sometimes this is what makes something wrong) but rather because it’s wrong.¹² Hence, although sensitivity to others’ expressions of disappointment or disapproval are surely necessary for moral agency – in that it enables us to apprehend the full, morally relevant impact of our actions (especially when it is is the way an action makes someone feel that determines its rightness or wrongness) – and for social success and individual well-being, it is, in principle, not necessary for upholding the bulk of our moral duties. Moreover, although this particular sensitivity may well serve to scaffold our morally relevant motivational structures in a way that the other sensitivity cannot, this kind of motivational scaffolding is not distinctly moral.

That social reasons-sensitivity is distinct from MRS is perhaps most obvious in what I call “going-against-the-grain” cases (Gogoshin 2021b) – where what is socially condemned or approved is at odds with what is morally wrong or right.¹³ During the era of racial segregation in the United States, a white American inviting a Black American to share her seat in the front of the bus would have been met with condemnation by the former’s peers. Assuming the White American acted for the moral reason, her MRS would have had to trump her sensitivity to social reasons in order to act rightly.¹⁴

How do we promote engaging in right-doing for the moral reasons rather than to avoid the social sanctions or garner the social rewards of our practices? How do agents come to effectively care about the (potential) harm or good they (might) bring about – the right-making features or rightness of actions – directly? It’s far from obvious at this point that our responsibility practices are the package-deal. The SV must show that they are.

3 The Scaffolding View

The SV takes a “metaphysically constructivist” or skill-based approach to responsible agency, according to which responsible agency is dependent on a capacity to recognize and respond to moral reasons. MRS is itself a kind of intelligent capacity (following Ryle 2002) which comes in degrees and which, like all skills, requires work, practice, and environmental feedback to develop and maintain. As a distinctly social capacity, MRS requires specifically social feedback.

[...T]he capacity to recognize and respond to moral reasons is an essentially social skill, requiring social feedback to develop and maintain. For moral reasons are themselves concerned with our interpersonal relationships—specifically, with how we ought to treat and regard one another in the context of those relationships (though of course moral reasons may extend beyond this interpersonal domain as well). But how we are to treat and regard one another is something we work out together, developing in community with others a shared set of norms that are in their nature never static, but continuously subject to renegotiation in light of changing epistemic and material circumstances. To recognize and respond to moral reasons is thus to be sensitive to such norms: to understand what they demand of us and to govern our actions accordingly—or, of course, to challenge such norms if we think they are normatively objectionable. But this is an on‐going project requiring continual social feedback, both because the norms themselves are subject to socially negotiated challenge and change but also because the norms themselves are invariably complex and open to socially negotiated interpretation in the demands they make of us. (McGeer 2019: 313)

The requisite feedback for MRS is provided in our reactive exchanges. Reactive attitudes are in this light proleptic practices in that they help to call forth the capacity on which they depend (McGeer 2019: 313). They are a bidirectional form of moral address in that they express normative demands and expectations and call for a specific response from the addressee – e.g., to explain or justify, to acknowledge wrongdoing, to apologize, etc. (McGeer 2019: 314).

The ideal reactive exchange results in moral growth – in an increased capacity to respond to the moral reasons that there are. Blame, for instance, demands that the wrongdoer go beyond mere recognition of the wrongdoing; it demands regret, remorse, and a commitment to do better in future. Thereupon, the blamer’s angry resentment is replaced by forgiveness which, in turn, reinforces the wrongdoer’s commitments. The wrongdoer grows more confident in his/her “own enlarged capacity to track and respond to moral reasons” (McGeer 2019: 314). There are other trajectories that the ideal reactive exchange can take – e.g., normative disagreement and discussion – but they all result in the enhanced MRS of at least one of the respondents.

What then, is the objective underlying feature agents must possess to be genuinely responsible in the sense that they are fitting targets of our reactive attitudes? What, i.e., is the objective underlying feature agents must possess to be sensitive to moral reasons? “On the view I am sketching, the answer is simply this: What they need to possess—indeed, all they need to possess—is a susceptibility to the scaffolding power of reactive attitudes, experienced as a form of moral address” (McGeer 2019: 315). Having established the necessary and sufficient condition for MRS, the SV offers a specific definition of MRS: “an intertemporal and (essentially) interpersonal skill‐based capacity for responding to the reasons—that is, a capacity that involves having whatever it takes to be sensitizable to the kind of reasons present at the time of their action, in part by way of the exhortatory effects of (ex post) reactive scaffolding” (McGeer 2019: 315).

According to this formulation (see also McGeer and Pettit 2015), MRS is not strictly a matter of responding to moral reasons per se, it is also a matter of responding to (others’ judgments expressed via) the reactive attitudes. The negative reactive attitudes convey normative conflicts (between moral reasons and actions) and are thus informative. Due to their emotional and stinging nature, these attitudes can also be said – about those who have the right sensitivities – to sensitize agents to moral reasons and to exhort or motivate agents to abide by them. Note that in order to trump the PRV, the SV would have to adopt a notion of sensitization which is greater than mere conditioning (pairing positive or negative social feedback with norm-relevant behavior). To this end (in §5), I will explore a notion of sensitization according to which agents come to (more) effectively care (more) about moral reasons.

In the following two sections, I assess the SV’s claims regarding the epistemic and formative roles of the reactive attitudes in cultivating moral agency. For the purposes of this assessment, I exclude their exhortatory dimension, which I take to be a function of their status as prudential incentives. This would support the PRV rather than the SV. In §§6 and 7, however, I consider how sensitivity to the reactive attitudes might be construed as MRS.

4 Reactive Attitudes and the Normative Landscape

Recall that on the SV, “how we are to treat and regard one another is something we work out together, developing in community with others a shared set of norms that are in their nature never static, but continuously subject to renegotiation in light of changing epistemic and material circumstances” (McGeer 2019: 313). It is through our practices that we construct, learn, and negotiate the normative landscape. For present purposes, the normative landscape can be divided into two kinds of moral obligations. I contend, contra the SV, that the first kind is a fitting object of discovery rather than of construction or negotiation, for which the second kind is a more fitting object. This gives moral reasons (and our access to them) some independence from the reactive attitudes.

The first kind concerns the standing (moral) reasons we have to act one way vs. another (“type 1 reasons”). These can be thought of as the familiar demands of ethics – demands to, e.g., keep promises, avoid unnecessary harm, uphold human dignity and autonomy, etc. The second kind, “type 2 reasons,” concerns the reasons we have to fulfill the obligations incurred by our transgressions of type 1 reasons. If we engage in wrongdoing, we thereby incur additional obligations and have (a type 2) reason to, e.g., feel remorse, apologize, repent, ask for forgiveness, commit to doing better in future, etc. On the view of moral reasons adopted here, type 1 reasons stand apart from our practices – from how or whether anyone might respond to our actions. They justify the obligations at stake in our practices. Type 1 reasons are an object of discovery. Indeed, we may discover them by way of the practices, but in principle, they are discoverable by multiple means (e.g., via the imagination and conversation, literature and history books, religious, political, and ethics texts, classroom lectures and church sermons, etc.). Type 1 reasons are also an object of discussion and debate, but not necessarily an object of construction or negotiation (depending on one’s metaethics, of course¹⁵).

Type 2 moral reasons concern what we owe to one another in virtue of how we have responded to type 1 reasons (whether, i.e., we have acted in accordance with those reasons – e.g., whether we’ve kept our promise).¹⁶ Since our practices are responses precisely to whether/how we respond to type 1 reasons, they concern type 2 reasons. It is plausible that our practices are the means by which we construct and negotiate the obligations tied to type 2 reasons (e.g., when an apology is necessary and what form it should take). So although our practices may target type 2 reasons, and so responding to them is to respond to moral reasons (and thus perhaps also to reinforce/enhance MRS), type 2 reasons are nonetheless distinct from an entire class of moral reasons (type 1 reasons). Sensitivity to type 1 reasons, as per the PRV, is not necessary for fulfilling the obligations which our practices target.

SV-responsible agents will, minimally, respond to type 2 reasons (i.e., SV-responsible agents possess type 2 MRS). Is this responsiveness sufficient for type 1 MRS? It is an empirical question whether agents can feel and express genuine guilt and genuinely apologize (i.e., fulfill their type 2 obligations) without caring about the type 1 reason at stake in the exchange. It is nonetheless conceivable that being the target of someone’s resentment, e.g., (especially someone whose respect we desire) is sufficient to inspire the kinds of responses that allow one to fulfill type 2 obligations.¹⁷ Accordingly, SV-responsibility does not necessarily suffice for type 1 MRS. This will be considered in more detail in the following section.

5 Sensitization to Moral Reasons

According to the SV, our reactive attitudes enhance MRS by way of our responsiveness to them, I take it, in a more substantive way than via mere conditioning or the provision of prudential incentives to comply with norms. It must therefore be the case either (1) that the reactive attitudes can affect our perception of moral reasons or our responsiveness to them or (2) that reactive attitudes are, in some sense, the “right kind of reason.” This section explores the first possibility; §6 explores the second.

Surely our practices can draw our attention to (type 1) moral reasons – when someone whose opinion we care about or respect resents us for some perceived slight, we have reason to attend to the basis of this resentment. On its own, however, this fact says nothing about whether our practices prioritize moral reasons over prudential reasons, whether they can sensitize us to moral reasons (vs. merely drawing our attention to them), or whether they are necessary for either drawing our attention to moral reasons or sensitizing us to them.

Consider as a metaphor the practice of highlighting text in academic textbooks. One may highlight important passages, passages which merit further consideration or attention. It is not the highlighting that renders these passages important; rather it is the importance that we attach to them that prompts us to highlight them. We can think of our practices in a similar manner. The importance of our moral obligations merits drawing special attention to them. When we blame an agent for transgressing them, we reveal their importance. This presents a global, external justification of blame – blame educates us about the importance of our moral obligations. However, we can separate the moral reasons behind the obligations targeted by our practices from the (prudential) reasons with which these practices furnish us. Accordingly, it might be that blame educates us only about the importance of fulfilling our obligations rather than the moral reasons behind them.

Let us recall the earlier speed limit example. We can fulfill our obligation – driving at the recommended speed – because we want to avoid sanction or because we want to reduce the likelihood of harm. It is far from obvious that, via the threat of sanction, we are able to draw agents’ attention to the moral reason over the prudential reason. In fact, depending on the agent’s sensitivity or severity of the response, it may do the opposite.¹⁸ The threat of being pulled over by a police officer and/or slapped with a hefty fine might easily dominate any consideration of the harms incurred by a hypothetical car accident. If the agent finds herself in an area where her speeding is unlikely to be witnessed or sanctioned, she may lose sight of a reason to temper her desire to speed.

This effect may be compounded for immature or developing agents. To take a non-moral, hypothetical example, consider a child who loves drawing. Imagine the parent who praises the child every time the child draws. It is conceivable that the child’s internal (intrinsic) motivation could give way, over time, to a desire for praise, such that the child loses interest in drawing should the parent cease praising him when he draws. The possibility that our practices can reinforce prudential reasons-responsiveness and extrinsic motivation over (and perhaps at the cost of) MRS reduces the force of the SV’s claim (or increases its justificatory burden).¹⁹

This analysis doesn’t entirely do justice to the thesis of the SV, however, which is concerned specifically with our interpersonal attitudinal responses. Suppose therefore that our partner, seated in the vehicle with us, expresses resentment at what he labels our “reckless driving.” He reminds us of the great harm to ourselves and others we are making more likely to occur by speeding. This seems to be the right place to look for whether our practices can make a difference to MRS. I think it’s possible that, certain conditions obtaining, we could be moved by our partner’s reaction to take another look at the relevant moral consideration and, in so doing, be moved by it. We could additionally imagine losing control of the vehicle and horrible injuries resulting from the crash. In this case, resentment appears to serve as a prompt to (re)consider the relevant moral reasons. It’s not yet clear that prompting is enough to enhance MRS or what form the prompting has to take or content it has to take on, even if it is. Moreover, this is not the only trajectory a reactive exchange could take.²⁰ It’s possible, for instance, stung by our partner’s resentment, in a conciliatory move, that we modify our future driving behavior in accordance with his judgment only to appease him and/or restore harmony between us. We might do so without attending further to the moral reason. We may be embarrassed by our impulsive and selfish appearance in his eyes and henceforth repress it in his presence. In a retaliatory or defensive move, we might dismiss our partner’s opinion and rationalize our behavior.

Even if our practices were limited to interpersonal exchanges (which is the focus of the SV), rather than also being tied to larger societal institutions as I think they are, they are highly varied and their influence on MRS is unlikely all (or even mostly) positive.²¹ Putting this challenge to the big picture instrumental justification of our practices to the side, the SV still faces the challenge of whether the ideal reactive exchange can enhance MRS.

Let us consider a possible blame exchange in greater detail. My partner’s blame, according to the position he occupies in my life, his moral standing, etc., gives me reason to attend to the basis of his blame. His blame is justified if my behavior has transgressed either a moral reason or the balance of moral reasons. His blame (potentially) signifies, communicates, and punishes my wrongdoing. His blame (the highlighter) is not what determines whether I have erred (it does not constitute the moral reason; it is not what merits the highlighting); it is a reaction to that fact. Supposing my behavior has transgressed only the balance of moral reasons – if I have recognized all the relevant moral reasons but have failed to put them together in the right way – then my partner’s blame still functions (potentially) to signify, communicate, and punish my mistake. If all wrongdoing were a matter of transgressing the balance of moral reasons, though, it would be easier to see how our practices lead to substantive moral reformation – sensitizing us to where the balance of moral reasons lies (something which may presuppose type 1 MRS) rather than to the moral reasons themselves. But I think that wrongdoing is at least very often a matter of failing to recognize or respond to moral reasons, and to be successful, the SV will need to show how these failures can be remedied by our practices.

Supposing I hadn’t realized that my action would amount to a transgression, then I acted in ignorance, in which case the blame is only fitting if I oughtn’t have been ignorant of the relevant moral reason. Blame could nonetheless serve to inform me of the moral reasons, though it may suffice for my partner to simply state the moral reason. Since it isn’t prima facie evident that blame is necessary to convey moral reasons, blame isn’t necessarily always a fitting moral educational tool. It is in cases where we act with full awareness of the moral reason at stake – where we act immorally – that we should look for whether a reactive attitude like blame has the power to enhance MRS (specifically, to bring about moral reformation) and/or whether it is necessary for this end.

If it is my partner’s positive regard that prompts me to reform my behavior (if it is this reason to which I respond in future), then I will not have become more responsive to the moral reason. So either McGeer holds that others’ positive regard is in fact the right kind of reason because it is itself a moral reason, or she holds that it is a reason – perhaps the only reason available to those of us who have not been sensitized to a given moral reason – to consider the moral reason more deeply. It is possible that in so doing, the moral reason will itself begin to matter to us and to motivate us going forward.

It may be helpful at this point to draw a distinction between considering reasons and sensitizing (agents) to reasons. Perhaps all we can ever do toward the aim of sensitizing agents to moral reasons is to present the (type 1) moral reasons for consideration. After all, we can lead a horse to water, but we can’t make him drink. In this respect, the instrumentalist need only show that our practices successfully present moral reasons by, e.g., drawing our attention to them. Once we do this, either the agent comes to care about those reasons or she doesn’t; nothing more can be done.

The SV suggests that there is a right way of drawing agents’ attention to the moral reasons – via the moral address exemplified in the ideal reactive exchange described earlier. As stated in §4, SV-responsible agents will, minimally, respond to type 2 but not necessarily to type 1 reasons. This notwithstanding, in cases of immoral action, blame may be a necessary condition of our attending to the type 1 moral reason. Having failed to care about it at the time of wrongdoing, we need something more – an additional reason – to attend to it. My partner’s blame can then provide that something more. But in the end, can it do more than prompt me to reconsider the type 1 moral reason or give me an extra-moral reason to behave in accordance with it?

6 The Decisive Issue

This section has two objectives. The first is to discuss the conditions of the ideal reactive exchange and the second, granting that they are met, is to show that whether MRS is enhanced in this exchange ultimately comes down to one’s view of moral reasons. I consider how the SV might connect the reactive attitudes to moral reasons by way of Bernard Williams’ account of proleptic blame.

The ideal reactive exchange requires that the blamer be a member of the right audience and that the blamee be adequately sensitive to blame. This sensitivity will involve being prone to those moral emotions – shame, guilt, remorse – which enable the blamee to appreciate the moral significance of her action. Even if we could reliably guarantee these conditions, it remains an open question whether what the blamee comes to appreciate is constitutive of the right-making features or rightness of the action in question.

Recalling the highlighter metaphor, during the learning process, we surely lack the capacity to identify all highlighter-worthy information. We naturally depend on our teachers – (ideally) experts – to help us with this task. We learn not only what the content of the important passages are but what makes them important; we learn the method. Borrowing a familiar parable, our teachers provide us with a fishing rod and teach us how to use it.

There are two kinds of concern that arise on this picture. (1) The first concerns expertise and learning well. (a) What constitutes expertise and how do we come to possess it or recognize it in others? (b) What constitutes the right disposition for learning and how do we cultivate it? And (2), even assuming that our teachers are genuine (though inevitably imperfect/limited) experts and that we learn well, how do we get beyond mere knowledge transmission to acquire new knowledge? Question (2) is important due to what I take to be the limitations of our practices on the acquisition of moral knowledge and MRS and thus on our developmental potential. I return to (2) in §7.

Right audience. (1) (a) concerns the first condition from above – that the blamer be a member of the right audience. (1) (b) concerns the second – whether the blamee possesses the constitution to be moved by blame in the right way. I take it that this constitution is simply a precondition (rather than a matter of cultivation) of (the cultivation of) responsible agency on the SV.²² As to (1) (a), a lack of expertise can result, as it were, in the blind leading the blind.

This is a problem that McGeer and Pettit (2015) consider in their account of reasons-responsiveness, which they describe as a two-factor phenomenon composed of sensitivity to reasons and sensitivity to audience. Not just any audience, of course – it must be the right audience (the audience who has the right moral standing and moral views), such that (moral) reasons and social feedback coincide. “And so, inevitably, you must let your sensitivity to others express itself in a heightened sensitivity to the reasons that you and they both take to be relevant. Or at least you must do so in your dealings with others who, by your lights, are worthy of being authorized as advisors” (2015: 172). McGeer and Pettit do not proffer specific criteria for discerning the right from the wrong advisors (but neither does Aristotle). We must be content, for the sake of argument, with the notion that we have an intuitive sense about who are the virtuous and thus who ought to be our advisors.

Right reasons. Assuming thus that our audience is the right audience and that we can be correctly guided by their judgments, why should we view acting on others’ judgments as acting for moral reasons (or acting with moral worth)? Bernard Williams offers an answer in his proleptic account of blame (Williams 1995).²³ In defense of reasons internalism, Williams claims that someone only has a moral obligation to φ if he can legitimately believe that φ-ing will satisfy one of his desires. John Skorupski provides a helpful interpretation.

Williams ingeniously attempts to mitigate the force of this point by invoking what he calls a “proleptic” theory of blame. It appeals to the “desire to be respected by people whom, in turn, one respects.” Blaming a person who has that desire but otherwise has no motivation to avoid some particular moral wrong is “as it were, a proleptic invocation of a reason.” It makes it true that he has a reason to avoid it, in virtue of his desire to avoid blame when it comes from people he respects (Skorupski 2007: 96).

So whereas I may not have a reason to help someone in need, my desire to have the respect of others (worthy others, specifically – “respectable” others), gives me a reason to do so. The obvious objection to this is that doing the right thing to garner respect is acting for the wrong kind of reason. For Williams and, I suggest, McGeer and Pettit too, this is not the case. Williams (per Skorupski) considers that respect for (respectable) others’ judgments is the right kind of reason (itself a moral reason or at least practically indistinguishable from a moral reason) and suggests that to claim otherwise reflects an erroneous, distinctly modern view of morality. The following passage in Skorupski (2007: 101) casts Williams’ view in the best possible light.

Responsibility – responsiveness to moral reasons – is a short word for a multiform capacity which comes in degrees. It is not a package-deal. Developing the various kinds of responsiveness it requires should not mainly be a matter of communing in solitude with one’s private conscience (though the degree to which it has to be is a matter of the society in which one lives). It should be dialogical: I come to appreciate reasons that I wouldn’t have come to see on my own by listening to what people I respect think. I am willingly recruited into this deliberative community, seeing myself as a genuine member, not a follower. This, surely, is the element of truth in Williams’ proleptic theory of blame. A certain ruggedly conscience-driven and egalitarian attitude would emphasise that doing the right thing to earn respect is doing it for the wrong reason and thus cannot earn respect. The element of truth in this makes it difficult to see what is limiting and ungenerous in it. A more forgiving and worldly wisdom says that motives can’t be so finely discriminated, that the desire for respect shades into the desire to do those things that command respect for the very reasons for which they command it, and that the desire for respect, or even honour and glory, is in any case in no way an ignoble desire.

The core idea here, and in McGeer and Pettit (2015) and McGeer (2019), is that both moral motivation and the acquisition of moral knowledge, as constitutively social, are wholly dependent on others in some form or fashion. Acceptable as this may be from a developmental or reformative point of view (and from the perspective of responsible agency, on my view), it does not do justice to our moral aspirations or (albeit fragile) capacities (vis-à-vis notions like moral worth and moral autonomy).

7 SV-Responsibility vs. Genuine MRS

My objection to the SV is not that it defends a notion of responsibility that does not require moral worth or moral autonomy. My objection is that it presents SV-responsibility as full-blown, genuinely moral reasons-sensitive, autonomous agency. This section draws out the differences between them.

I submit that in the context of moral development (which, like McGeer, I take to be an ongoing process of skill building and skill refinement), the desire for others’ respect, e.g., might sometimes be the right kind of reason, or at least not the wrong kind of reason, for right action. Norm internalization – a crucial aspect of moral development (see e.g., Tomasello 2020) – is a matter of coming to care effectively about the moral reasons themselves rather than others’ judgments, even if the latter serves as a necessary intermediary of the former. With respect to the aim of norm internalization, the desire for others’ respect can at best possess only instrumental value. At worst, it is simply the wrong kind of reason.²⁴

It feels intuitively wrong (a “one thought too many” kind of wrong) for it to matter to me at all, nevertheless be necessary for motivation, when faced with the chance to save someone from a burning house, whether it is a respectable thing to do or whether someone will respect me for it. Perhaps this is an extreme example, but we can scale it down and still come up with the same result for all but perhaps highly interpersonal situations. Even in these situations, however, I would argue that others’ opinions are morally relevant only insofar as they make the fulfilling of a universal moral obligation possible. It may be possible, for example, to respect our partners’ dignity and well-being (type 1 reasons) only when we know their distinct personalities and preferences. Learning how to do this may require the reactive exchanges McGeer (2019) identifies. Still, type 1 moral reasons are themselves independent of the reactive attitudes. Genuine MRS requires being motivated by moral reasons directly.

McGeer and Pettit (2015: 175) anticipate my worry about a dependence on others’ judgments as a necessary condition of responsible agency.

For all that our view supposes, it is quite possible for someone to become so sensitive to reasons that, while they remain sensitive to audience, while it continues to matter to them that they act well in the eyes of the others they authorize, this concern does not play an essential role in making them responsive to reasons in any range of choice; their standing sensitivity to reasons—their excellence or aretaia in that regard—is sufficient on its own to keep them on the path of virtue.

However, they point out that this picture of responsible agency (as virtuous agency) does not preclude the picture of responsible agency according to which concern for others’ judgments is the dominant motive. “It may be […] that your standing capacity to respond to reasons, however autonomous in its operation, is dependent for its survival on your sensitivity to others” (2015:175). As previously discussed, the others in question must be the right others, but provided this is so, McGeer and Pettit allow for a combination in any ratio of one or the other sensitivities.

This result is counterintuitive. After all, autonomy and external scaffolding seem conceptually at odds with one another. How could it be that a fully autonomous agent requires the ongoing (and possibly increasing; McGeer 2015: 272) scaffolding of others? McGeer and Pettit have responded that their view accommodates agents (virtuous agents) who do not in fact depend on social feedback to respond to reasons directly and responsible agency can result from any combination – in any ratio – of sensitivity to reasons and to audience. However, this does not fully resolve the challenge – for our reactive attitudes are, from a cultivation standpoint, unfitting for agents who do not require their scaffolding. Hence, McGeer and Pettit still must identify a dependence for such agents on the reactive attitudes. Their further response can be found here:

[...A]ll power corrupts and [...] the only hope of keeping the powerful virtuous is eternal vigilance: that is, a sustained interrogation and examination of their performance (Pettit 1997: ch. 7). It may be that while you now have Aristotelian virtue, your political power is such that, if we ceased to be vigilant, if we enabled you to serve your own interests with invisibility and impunity, that would change your character and destroy your virtue. It may be, in other words, that your standing capacity to respond to reasons, however autonomous in its operation, is dependent for its survival on your sensitivity to others (McGeer and Pettit 2015: 175).

Still, virtuous agents – although fitting targets of our reactive attitudes in principle (since they have the right capacities) – behave virtuously, not in blameworthy ways; thus they are rarely if ever the targets of our actual (blaming) reactive attitudes. It would then seem that their responsibility (qua virtuous behavior) does not depend on actual (negative) reactive attitudes, and thus that the SV cannot properly account for them.

To this challenge, I expect that McGeer and Pettit would respond as follows. Autonomous agents have internalized the social feedback provided by the reactive attitudes and are in this way “internally scaffolded.” They have, in effect, internalized the dialogue of their moral community and can be moved by its hypothetical (or via the memories of actual) feedback. Their awareness (conscious or not) of a standing demand for answerability is then the vigilance to which McGeer and Pettit were referring, and which scaffolds their moral motivation.

However, this solution still falls short of the standards for genuine MRS. Agents who behave responsibly by way of a sensitivity to actual or internalized feedback or to an awareness of their answerability to others do not act for the right-making features or rightness of actions. And, insofar as agents depend on sensitivity to social feedback for right action (i.e., insofar as they require the highlighter to perceive moral significance), their moral agency is heteronomous rather than autonomous. ^{25, 26}

On the other hand, this picture meshes well with the Scanlonian view (Scanlon 1998, 2008), where morality is a matter of what we owe to one another and moral motivation is a matter of being concerned with the (in principle) justifiability of one’s actions to others. If this is right, then the SV-responsible agent is morally autonomous. In response, rather than rejecting Scanlon’s conception of the foundations of morality (as does, e.g., Zimmerman 2016), I contest his conception of moral motivation. More precisely, I suggest (contra the SV²⁷), that our sources of moral motivation evolve – or ought to evolve – over time.²⁸ Sensitivity to others’ judgments is surely necessary for moral development. When we lack sufficient MRS, we must rely on such a proleptic reason. But I take it that proleptic reasons (see Callard 2018) stand in for the moral reasons until we come to care about them intrinsically.

I am at last able to make good on the promissory note in §6 regarding the following questions. Even assuming that our teachers are genuine experts and that we learn well, how do we get beyond mere knowledge transmission to acquire new knowledge and hence to make progress? How do we acquire the skill that would enable us to highlight passages that our teachers have overlooked or ignore passages which do not (or no longer) merit highlighting or, even better, to add new highlight-worthy text? The answer – an autonomous capacity for identifying and responding to type 1 moral reasons. Though I have not yet provided an obvious means for cultivating this capacity, I have argued that our practices do not guarantee its development. I have even suggested that they may sometimes obstruct its development when they draw an agent’s attention away from the moral reasons.²⁹

8 Conclusion

In this paper, I have challenged the SV’s construal of the relationship between MRS and our practices. According to the SV, MRS requires only a susceptibility to the scaffolding power of the reactive attitudes. Whatever that scaffolding exactly amounts to – which is both formative and motivational in nature – it does not require sensitivity to what I have termed type 1 moral reasons. Moreover, SV-responsible agency is mediated through others, both in terms of moral knowledge and motivation. Consequently, SV-responsibility falls short of genuinely MRS agency.

This notwithstanding, the SV presents a persuasive picture of our practices – one which reveals their ecological character and, in my view, their limitations. Moreover, it identifies a potential avenue for moral development and reformation via the ideal reactive exchange. But (1), ideal reactive exchanges are unlikely to comprise a significant portion of our practices³⁰ and (2), even in the ideal reactive exchange, type 1 MRS remains an uncertain outcome. In fact, I rather suspect that type 1 MRS is a necessary condition of the ideal exchange. The instrumentalist is thus, at this point, unwarranted in identifying MRS as necessary for responsibility. This, unless one is willing to confine their view of moral reasons to type 2 reasons (which may themselves be a function of type 1 MRS), or stretch it to include, e.g., social acceptance or respect.

I would resist this move, however, in light of the fact that we do sometimes (perhaps more often) autonomously act for (type 1) moral reasons strictly speaking (sometimes in spite of our practices) and we have good reason to pursue why and how this is so. Defining MRS in the way the SV has eliminates the impetus to do so. But genuine MRS, given its tenuous status among our practices, is too demanding to be a fitting object of accountability. We should regard it instead as an object of aspiration. This allows us to uphold the virtues of the SV – a realistic conception of moral agency and a modest (and fair) account of responsibility – while yet leaving room for moral growth.