I. RGC

RGC is a version of consequentialism that attempts to accommodate attachment, virtue and respect. What does Pettit understand by attachment, virtue and respect? Attachment is understood as an umbrella concept that covers love, friendship, solidarity, and other relationships that seem to be at odds with the comprehensive motivation, associated with traditional consequentialism, to promote the impartial good. Justice, honesty and fidelity are some of the virtues that appear to be hard to reconcile with consequentialism. Respect requires not interfering with the freedom and rights of others.

According to Pettit, attachment, virtue and respect should be understood as rich goods with a common structure. The analysis of the common structure is the core of RGC. What rich goods have in common is, according to Pettit, the robust provision of thin goods. Love requires the robust provision of care (12), friendship requires the robust provision of favour (34), fidelity requires robustly keeping promises (45), honesty is tied to robust truth-telling (46), and the thin good connected with respect is non-interference with someone's options (78).Footnote ¹ While rich goods are connected to the robust provision of thin goods, the value of rich goods goes beyond the value of the respective thin goods (111–37).

What does Pettit mean when he talks about the robust provision of thin goods? Pettit's idea is that, in order to provide a thin good robustly, an agent must be a reliable provider of that good. Reliability consists in having a disposition to provide the thin good not only under actual circumstances but also under relevant different circumstances. Which circumstances are relevant? Pettit's basic idea is that the disposition must be manifested under all and only those circumstances in which suitable priming and support are present. The notions of priming and support thus play crucial roles in Pettit's theory and we should have a closer look at them.Footnote ²

Primers trigger your action. Robustness with respect to suitable priming means that you must provide me with the respective thin good ‘only in variations on the actual scenario that preserve the personal primers or prompts, the stimuli or cues, that actually move you’ (16). Examples of primers are my need of consolation over the death of a pet (in the case of love/care) (15), my need for information (honesty/truth-telling) (45), your making of a promise (fidelity/promise-keeping) (45–6), and the availability of a choice on my side and of interference on your side (respect/non-interference) (81).

Suitable support, in Pettit's sense, means sufficient support by reasons. To provide a rich good, the provision of the respective thin good needs to be adequately supported by reasons associated with the rich good, i.e. reasons of the rich good to provide the thin good must outweigh the balance of competing reasons (15, 16, 50). For example, fidelity does not require that you keep your promise when you can do so only by letting a drowning person die. It does so, however, when your only reason for breaking it is that you are a bit lazy and don't feel like keeping it. The point is that the overall balance of reasons must support acting on the disposition to provide the thin good.

So far, the preceding summary might suggest that, according to Pettit, rich goods are dispositions for the robust provision of thin goods. This, however, is not the case. The rich goods Pettit is interested in are not the aforementioned dispositions themselves but the benefits you give to others by manifesting these dispositions (see, e.g., 43, 73, 140, 167–8). This also comes out clearly in Pettit's simplified general formula for rich goods, which he presents applied to love, honesty and respect as follows:

With these explanations at hand, we can now formulate the central building block of RGC:

This concludes my exposition of RGC.

II. The Problem

Claim_RG is the centrepiece of RGC. Unfortunately, Claim_RG has highly counterintuitive implications. The implications constitute what I call the Problem of Self-Imposed Priming and Support – or, in short, Problem _SIPS.

Problem_SIPS arises because Pettit conceives of attachment, virtue and respect as goods and because the associated ‘thin’ phenomena – such as the provision of care, the keeping of promises and truth-telling – are goods, too, on RGC. These assumptions imply implausible evaluative judgements. Here is a sample:

• • Rather than simply donating money to a good cause, it is better if you first promise to donate and then keep your promise.

• • Rather than prophylactically giving me your read-through newspaper before I need a particular piece of information, which, as you know, is contained in your newspaper, it is better if you wait until my need obtains and then provide me with what I need to know.Footnote ³

• • It is better if you are armed than unarmed and provide me with respect by not interfering with my choices (which you would not be able to do if you were unarmed).

Claim_RG implies that the states of affairs respectively judged to be better in these examples are better because they contain the provision of thin and rich goods: promise-keeping and fidelity in the first example, truth-telling and honesty in the second, non-interference and respect in the third. However, it is absurd to hold that these states of affairs are better than their respective counterparts.

In which cases does Claim_RG have counterintuitive implications? Relevant are situations in which you can bring it about that the priming and support conditions are met. Claim_RG seems to imply that it is good for you to see to it that the priming and support conditions are met and then provide the respective rich and thin goods. In fact, however, these actions seem pointless and without value in the envisioned examples.

We can formulate Problem_SIPS by generalizing the points just made and assuming that statements about which things have intrinsic value are necessary truths:

1. (1) On RGC, necessarily, it is good to provide a rich good.

2. (2) On RGC, necessarily, a person provides a beneficiary with a rich good to the extent that, while providing the beneficiary with the suitably primed and supported associated thin good, the person is disposed to maintain the provision of the thin good in all modest variations on those circumstances that preserve the priming and support.

3. (3) On consequentialism, and hence on RGC, necessarily, if a state of affairs p is good, then it is good to bring it about that p.

Therefore,

1. (4) On RGC, necessarily, it is good if, while providing a beneficiary with a suitably primed and supported thin good that is associated with a rich good, a person is disposed to maintain the provision of the thin good in all modest variations on those circumstances that preserve the priming and support. [From (1) and (2).]

Therefore,

1. (5) On RGC, necessarily, it is good to bring it about that, while providing a beneficiary with a suitably primed and supported thin good that is associated with a rich good, a person is disposed to maintain the provision of the thin good in all modest variations on those circumstances that preserve the priming and support. [From (3) and (4).]

However,

1. (6) It is not necessarily good to bring it about that, while providing a beneficiary with a suitably primed and supported thin good that is associated with a rich good, a person is disposed to maintain the provision of the thin good in all modest variations on those circumstances that preserve the priming and support.

The unacceptable evaluative judgements presented above justify (6). (1) and (2) are based on the cornerstones of RGC. To avoid Problem_SIPS, defenders of RGC must deny (3).

III. Wide and narrow scope

Defenders of RGC might want to attack (3) by attacking the principle that (3) ascribes to consequentialism:

1. (3*) If a state of affairs p is good, then it is good to bring it about that p.

If (3*) is false, so the idea goes, consequentialists shouldn't subscribe to it.

(3*) is, of course, a controversial principle. Some moral theories deny it. Defenders of RGC could try to deny (3*) by appealing to the concept of a rich good. To see how this works, it is helpful to distinguish between narrow-scope and wide-scope versions of Claim_RG, which can be abbreviated as follows:

1. (N) If suitable priming and support are given, then you provide someone with a rich good by acting out of the relevant disposition.

2. (W) You provide someone with a rich good by bringing it about that (i) if suitable priming and support are given, you act out of the relevant disposition and (ii) you act out of the relevant disposition.

It might be easier to see the difference if we formalize the principles in question. Letting G designate rich goods, P the presence of suitable priming and support, and A your acting out of the relevant disposition, we get:

$$\lpar {{\rm N}_{{\rm formal}}} \rpar \; P\to G(A{\rm \&} (P\to A) \; )$$

$$ \lpar {{\rm W}_{{\rm formal}}} \rpar \; G(P{\rm \&} A{\rm \&} (P\to A) \;)$$

How might the distinction between (N) and (W) be helpful for the defender of Claim_RG? On (W), it would be hard to deny (3*). But what if the defender of RGC insists that the relevant version of Claim_RG is (N)? (N) claims that if suitable priming and support are present, then one provides a rich good, and thus does good, by manifesting one's disposition to provide thin goods. And from this it does not follow by any uncontroversial axiological principle that one does good by bringing about suitable priming and support so as to manifest one's disposition and thus provide (what normally is) a rich good. The lesson is that acceptance of Claim_RG, understood as a narrow-scope principle, does not commit you to (3*).

Is RGC off the hook, then? The reason why I don't think so has to do with the consequentialist framework of RGC. I do not defend (3*) but (3) and argue that RGC qua consequentialist theory is committed to (3*). My argument has three steps. Notice, firstly, that RGC is officially silent on whether one way to bring about rich goods is to make it the case that suitable priming and support are present and trigger the relevant disposition: Pettit neither includes nor excludes cases of self-imposed priming and support. All that has been established above is that Claim_RG understood as (N) does not commit you to (3*), not that (N) is incompatible with (3*).

Second, assume with (N) that if suitable priming and support are present, then one provides a rich good by manifesting one's disposition to provide thin goods. On standard consequentialism, acts can be considered mere means for realizing good stuff. This instrumentalist understanding of acts suggests that on consequentialism you have reason to bring it about that you provide someone with a rich good. How can you bring this about? In a first step: by making it the case that suitable priming and support are present. Once this is done, you can, in conformity with (N), manifest your disposition to provide thin goods.

So far, I have argued that, if we assume (N) and a standard consequentialist framework, you have reason to bring it about that the priming and support conditions are met. This, however, does not yet bring us to (3*). We need a bridge between reasons and values.

This brings us to the third step of my argument, which is that we get this bridge for free on standard consequentialist assumptions. For acts share the values of their outcomes on the standard consequentialist picture: you do good by bringing about good states of affairs. The harmony between values on the one hand and reasons and duties on the other is one of the features that make consequentialism an attractive moral position in the first place.Footnote ⁴

One might want to resist the last two steps by arguing that, on Pettit's particular consequentialist theory, the agent's inner states – intentions, dispositions, motives – make a difference regarding the value of actions (chapter 5). This is, indeed, an important observation. But the crucial point here is that it is not a tenet of RGC that agents’ inner states preclude them from providing rich goods in cases where the agents bring about suitable priming and support themselves. Thus, given that the constellation under consideration is not subject to any particularities regarding reasons and values in RGC, it seems legitimate to assess it on the basis of default consequentialist assumptions. This verdict gains further support from the fact that there simply is no straightforward way to handle cases of self-imposed priming and support, as will become clearer in the next sections.

To sum up, it is true that, on a narrow-scope understanding of Claim_RG, we cannot justify (3*) by an uncontroversial axiological principle. In light of Pettit's silence regarding cases of self-imposed priming and support, however, it is appropriate to view these cases in the context of Pettit's more general assumptions and thus recede to the consequentialist framework of his theory. Within this framework, assuming that you have the relevant disposition to provide thin goods, you have instrumental reasons to bring about suitable priming and support and your doing so is good. The upshot is that RGC is committed to (3*) and the attack on (3) fails.

IV. The preceding behaviour

This section presents and critiques an attempt to solve Problem_SIPS by adding to RGC the claim that an agent's bringing about suitable priming and support is somehow bad so that the balance of this behaviour together with the provision of rich and thin goods comes out neutral in value. One might in this context even be tempted to think of something like the abuse of normative powers, e.g. when the only motive for making a promise is to keep it subsequently in order to do something good.

The idea can be made more precise by distinguishing between pro tanto and all-things-considered goodness. On RGC, it is, necessarily, pro tanto good to provide rich goods. This goodness is realized even in cases where priming and support are self-imposed. However, in such cases, so the idea goes, there is also counteracting badness, stemming from the self-imposition of priming and support. Therefore, on RGC, it isn't necessarily all things considered good to bring it about that priming and support conditions are met so that, out of relevant dispositions, a person provides a beneficiary with thin goods.

This idea does not work for two reasons. The first reason is that, upon closer examination, making the priming and support conditions met does not appear to be abusive or bad behaviour. There seems to be nothing bad about making a promise to donate money or with refraining from prophylactically offering a newspaper.

Second, the idea would not work even if we were to assume that it is somehow bad to bring it about that the priming and support conditions are met. For the badness would seem rather small in scale and at most outweigh the value resulting from thin goods, but not amount to counterbalancing the value of rich goods. To illustrate, withholding your newspaper from me might somehow ‘damage’ me in that I do not get the information that I will be looking for. But once you provide me with the thin good of truth-telling by informing me about what I need to know, any such damage would be undone. Thus, the counterintuitive implications of Claim_RG would still follow by virtue of the value of rich goods.

V. Sophisticated value theories

Value theorists have put forward several ideas that can be used in an effort to save RGC from Problem_SIPS.Footnote ⁵ Let us consider two of them. The first is the idea of conditional value. A defender of Claim_RG might suggest that a condition of your doing good by providing rich and thin goods is that you did not yourself bring it about that the priming and support conditions are met. A second important idea is that of organic unities. Here the proponent of Claim_RG might argue that, while the state of affairs of my intentionally bringing about suitable priming and support is by itself neutral and the state of affairs of my providing someone with a rich good is by itself good, the combination of these states of affairs is neutral in the problematic cases.

How could employing one of these ideas be justified? A starting point might be that the internal sides of agents – the intentions, motives, dispositions with which an agent performs an action – play a role for the assessment of the actions in Pettit's theory. Now, building on this, I see two promising arguments for the justification of conditionalization or the existence of organic unities. First, if an agent intentionally brings about priming and support just in order to provide rich goods, this intuitively bears witness to the agent's lack of ideal forms of attachment, virtue and respect. Second, from the perspective of the ‘beneficiary’ or patient, such intentionally brought about priming and support is not welcome. This second thought might be developed further based on the idea that the main purpose of rich goods is to reduce one's vulnerability (chapter 4). In so far as rich goods are protective measures, it is important that the person relying on such measures cannot plausibly be expected to appreciate the realization of situations that, but for those measures, would be dangerous just to be shown that the measures are effective.

Do these sophisticated manoeuvres solve Problem_SIPS? Let's assume that the existence of value conditions or organic unities explains why it is not better if you

• • promise that you will donate later and then keep your promise rather than donate now,

• • wait until my need of information obtains and then provide me with what I need rather than prophylactically give me the document that I will need,

• • are armed rather than unarmed and provide me with respect by not interfering with my choices.

This still doesn't solve Problem_SIPS, however, as Problem_SIPS might not only occur in cases with two persons but also in three-person cases. Consider the following situations with a manipulator (Morris), an agent (Anna) and a patient (Paula):

• • Anna is about to donate to a good cause, which would benefit Paula. Morris incentivizes Anna successfully not to make the donation now but to promise instead to Paula that she will donate next week. Out of the disposition of fidelity, Anna keeps her promise. The incentivizing is achieved in a morally harmless way; say, by offering a nice dinner so that Anna lacks the time for entering a bank transfer now. The net benefit resulting from Anna's donation is equally large, regardless of whether she donates this week or next week.

• • Morris knows that Paula will soon need some information that is contained in a document possessed by Anna. Anna is about to give Paula the document as Morris incentivizes Anna successfully and in a morally harmless way (say, by borrowing the document himself from Anna) not to hand over the document to Paula now. By the time Paula's need for the information obtains, Anna is again in possession of the document and, being an honest person, is happy to help Paula out.

• • Anna lives next door to Paula. Morris introduces Anna to the art of collecting arms. His efforts fall on fertile ground and Anna takes up the new hobby. Due to her arsenal, Anna would now be able to interfere with Paula's choices, but Anna doesn't interfere as she possesses the disposition of respect.

In each of these cases, Morris intentionally brings about priming and support for Anna who, in turn, provides rich and thin goods to Paula, whereas without the activity of Morris, Anna would only provide thin goods. On the revised version of RGC, it is good that Morris brings about suitable priming and support so that Anna does not provide only thin but also rich goods. In fact, however, what Morris does is not good (but pointless and thus neutral in value).Footnote ⁶

Let us now turn to three objections. One objection is based on the idea to reiterate the strategy considered at the outset of this section: conditionalization or organic unities, so the objection goes, preclude the provision of rich goods (or, depending on how you want to spell out the objection, that these goods have value).

While I was willing to accept this objection in two-person cases, it cannot be upheld in three-person cases. The reasons for adopting the ideas of value conditions or organic unities in two-person cases do not apply in three-person cases. This is because the agent providing rich goods in three-person cases is different from the person bringing about priming and support. Anna's intentions, motives and dispositions are perfectly fine and there are no grounds for holding that Anna fails to provide rich goods. Of course, there is something fishy going on in three-person cases. But this is entirely due to the behaviour of Morris, not of Anna. This finding gains further support from the perspective of the beneficiary. Paula will appreciate Anna's behaviour.

So far, I have argued that sophisticated value theories cannot successfully be used to argue that Anna does not provide rich goods. However, the critic might want to use these theories to explain why, even though what Morris does leads to good states of affairs, the actions of Morris do not inherit this goodness. This second objection is more promising. It does not attempt to show that what Anna does is not good but rather that Morris's actions don't participate in this goodness.

The problem with this idea is that it is not viable on standard consequentialism. Proponents of the idea must either admit that the values that accrue to an agent's actions can come apart from the values of the actions’ consequences or admit of agent-relative values. I fail to see how else they could maintain that the states of affairs of Anna providing rich goods are good consequences of Morris's bringing about suitable priming and support for Anna yet deny that these consequences are good-making features of Morris's actions.

The third objection grants that Anna does good and Morris's actions inherit this goodness but claims that, given that the character traits of Morris are problematic, what Morris does is bad and counterbalances the good done by Anna so that, overall, the states of affairs caused by Morris come out neutral and his actions bad. This objection does not work because it is subject to the considerations adduced in section IV. First, Morris's behaviour, though awkward, is not bad. It is not bad to make a person make and keep a promise to donate if the person would otherwise have donated without promising. Similar things can be said about borrowing a document before it is needed by somebody else and arguably about introducing somebody to the art of collecting arms.

Second, even if we assumed that Morris's behaviour is bad, the badness would not counterbalance the goodness of the rich goods. The amount of badness involved would at most suffice to cancel the goodness of the thin goods. For instance, the badness of seeing to it that Paula does not get the document before she needs it seems to be undone when Paula eventually gets hold of it.

The upshot is that sophisticated value ‘tricks’ will help the defender of RGC to avoid implausible verdicts in two-person cases of self-imposed priming and support. But these twists will not help with respect to three-person cases.