10 Comments

Thanks very much for this. There’s a lot in your reply; you raise a number of interesting and important challenges. But let me just focus on right-making properties.

Suppose my act both:

(i) benefits Lucy

and, as a result

(ii) maximizes the good.

Which of these things makes my act right? You suggest that we focus on ‘ground level’ explanations of the rightness of acts and so argue that (i) makes my act right.

Yet it seems to me that it is a necessary condition of a right-making feature (ground level or otherwise) that it makes actions right. I assume that ‘makes’ is modal. So I assume that:

If F is a right-making feature, then necessarily, if x is F, then x is right.

But (i) doesn’t satisfy this constraint. That is, it doesn’t make actions right, since an action could have this feature (on the consequentialist view) and still be wrong. And this isn’t peculiar to (i); I don’t see how to select any ‘ground level’ feature that is going to satisfy this constraint. (This is, of course, just the familiar problem raised by Russell that the individual facts don’t entail the general facts).

Later in your paper you seem to concede this point and propose that your view isn’t about right-making features at all but about pro-tanto right-making features or decisive reasons.

But now I worry about redundancy. (i) is a decisive reason (let us assume). But so is (ii), surely. If some feature is right-making simpliciter it is also right making pro tanto.

Now there’s a competition between (i) and (ii); double counting suggests we have to pick. Otherwise we would need to say there are two reason to perform my action—because it benefits Lucy and because it maximizes the good. But this seems to me one reason too many, (especially since (ii) seems to in some sense include (i)).

How should we resolve the competition? I propose the competition should be resolved counterfactually. Would (i) make my action right without (ii)? No. Would (ii) make my action right without (i)? Yes. So the competition should be resolved in favor of (ii).

I wonder where you think this reasoning goes wrong. I do not mean to assume that all actions must be made right by the same feature—I’m not yet clear on which part of my argument relies upon this assumption. Perhaps monism about right-making features/decisive reasons is a consequence of my argument. But I don’t think it’s a premise, as it were.

Thanks again for thinking about all this with me!

Expand full comment

Hi Miles, thanks for your reply!

I think your modal constraint is too strong (and would, for example, imply that individual objects are never causes). Water makes plants grow, but only given suitable background conditions (soil, sunlight, etc.). Similarly, benefiting Lucy made the action right, given suitable background enabling conditions (it didn't harm others more, wasn't outweighed by opportunity costs, etc.). The fact that it wouldn't have this effect in the absence of the enabling conditions doesn't show that it isn't in fact a right-making feature of the act; it just shows that there are enabling conditions that must be met in order for the feature to have the normative effect that it in fact does.

In general, benefiting any given individual is a pro tanto right-making consideration, and when not outweighed by competing considerations it becomes an actual (at least partial) right-making feature or ground-level explanation of why the act was right. (The full normative explanation will involve further details regarding who else was harmed and benefited.)

Re: redundancy, this is why I distinguish "ground-level" reasons vs. "summary / criterial explanations".

(ii) [that an act maximizes the good] is not a fundamental reason at all. It's a summary fact that entails that there are (other, more specific) reasons that decisively favor that action.

It's comparable to:

(iii) An infallible oracle would say that you ought to do this act.

As we know from the Euthyphro dilemma, (iii) is not a right-making feature or fundamental reason for performing the act. But it entails that the act is right and supported by decisive reasons. (ii) is like this.

> "How should we resolve the competition? I propose the competition should be resolved counterfactually. Would (i) make my action right without (ii)? No. Would (ii) make my action right without (i)? Yes. So the competition should be resolved in favor of (ii). I wonder where you think this reasoning goes wrong."

I think the counterfactual test is no good. For one thing: it's incompatible with the idea of pro tanto reasons that turn out to be outweighed. But *obviously* there can be good reasons that are outweighed. So not all reasons are modally guaranteed to succeed in making an action right. Still, they may so succeed when not outweighed. You shouldn't be ruling out this possibility in advance.

A better test is the apt motivation test: it's fitting to be motivated by good reasons. Is it more fitting to be motivated by concern for each individual separately, or by concern for their aggregate well-being? Obviously the former. (See 'Value Receptacles': https://philpapers.org/rec/CHAVR )

Or just direct normative assessment: Does individual welfare matter because it contributes to the aggregate, or does aggregate welfare matter because it includes each individual? Obviously the latter.

Ultimately, the question of which of (i) or (ii) is the "real", fundamental reason is a *normative* question, and should be answered on normative rather than metaphysical grounds.

Expand full comment

Thanks Richard. I have a couple of thoughts but let me keep focusing on right-making properties for the moment.

First, I am not so worried about the analogy you mention. I’m not at all sure that the causal ‘makes’ and the non-causal ‘makes’ play by the same rules. Further, I don’t think individual objects do cause things, except in a loose and popular sense; the relata of the causation relation are facts, in my view.

I think that right-making features play a unique conceptual role. In particular, they (or facts that involve their instantiation) completely explain the rightness of actions. I am thinking that the ‘explanation’ here is full metaphysical grounding. The ground level features you have identified don’t fully ground the rightness of an action, of course, since they neither necessitate nor completely explain the rightness of an action.

You mention in a footnote that you’re interested in normative grounding and that normative grounding is different from metaphysical grounding. So maybe you would say that *maximizing the good* metaphysically grounds the rightness of my action but *benefiting Lucy* normatively grounds the rightness of my action.

I must admit I find the distinction between metaphysical and normative grounding obscure; pace Fine (and in alignment with Berker) I’m inclined to think there’s just one relation here—grounding.

But suppose I’m wrong. I’m not sure how this changes the dialectic. We can still distinguish between full and partial normative grounding. As I understand you in section 2.4, you don’t want to say that the relation between the ground-level right-making features and rightness is partial normative grounding. But I can’t see how we’re getting full normative grounding—full normative explanation. I assume that you accept this when you say above:

“(The full normative explanation will involve further details regarding who else was harmed and benefited.)”

So I guess my reply is this: you suggest above that selecting fundamental right-making features is a normative, not metaphysical question. But the selection of fundamental right-making features *for consequentialism* must somehow be constrained by consequentialism itself. And the relation between the features you mention and the rightness of an action under the consequentialist principle seems mysterious.

The features you identify don’t entail the rightness of an act. They don’t (metaphysically) explain the rightness of an act. The don’t fully normatively ground the rightness of act. And they also shouldn’t be thought of as partial grounds of the rightness of an act. But they are consequentialist right-making features, at least in some sense.

So: what is the explanatory relation between these features and the rightness of acts on the consequentialist picture? And why care about these features, given their seeming disconnect from what matters for the consequentialist, which is that outcomes be as good as possible?

My concern is that you are introducing *something else that matters*—namely that we care about and be motivated by certain things that are connected to the underlying axiology of the theory (when understood in the token pluralistic way that you prefer). But I don’t see why this isn’t simply a separate moral concern, tacked on to but not grounded in the fundamental consequentialist insight that what matters is that outcomes be as good as possible. This worry seems to me especially pressing when e.g. it would make the outcome worse to be motivated in the ways that you recommend.

Expand full comment

The complete grounding explanation of an act's deontic status just requires listing all the specific costs and benefits (including opportunity costs) together with a "that's all" clause. (Note that this complete explanation then satisfies your modal constraint.)

There's no great mystery about how the partial and complete grounds interact: the former are just a subset of the latter. Specific costs are pro tanto wrong-makers. Specific benefits are pro tanto right-makers. Whether the act is overall right or wrong depends on how the two balance out, and you can't say which side wins until you specify that the respective listings are complete. But it's a mistake to say that the higher-level fact about balancing is what does the right- or wrong-making *work*. The normative work is done by the pro tanto right- and wrong-makers, and the final balancing is just a piece of accounting that tells you which side won out.

> "So: what is the explanatory relation between these features and the rightness of acts on the consequentialist picture?"

These features are the reasons that enter into a weighing explanation of the act's deontic status. They are the considerations that have normative weight, do normative work, and contribute towards making an act right. When they win out, they individually partially explain, and collectively (together with the "that's all" clause) completely explain, the rightness of acts.

> "And why care about these features, given their seeming disconnect from what matters for the consequentialist, which is that outcomes be as good as possible?"

I'd say just the opposite. Individual well-being is precisely, on my view, what fundamentally matters. (That was the point of the normative tests I suggested in my previous comment.)

> "My concern is that you are introducing *something else that matters*—namely that we care about and be motivated by certain things that are connected to the underlying axiology of the theory (when understood in the token pluralistic way that you prefer)."

Not at all; I haven't claimed that fitting attitudes "matter" (in the sense of being desirable - see my response to Stephen, where I explicitly repudiate the *Oughts are Practical Endorsements* assumption). Rather, they illuminate our normative commitments, including our views of what matters.

> "This worry seems to me especially pressing when e.g. it would make the outcome worse to be motivated in the ways that you recommend."

But that's irrelevant to whether individual welfare or aggregate good is what fundamentally matters. (After all, it's also possible for it to make the outcome worse to be motivated by a fundamental concern for the aggregate good.)

Going back to your counterfactual test: let (i) refer to the complete accounting of costs and benefits, including opportunity costs and the "that's all" clause; and let (ii) refer to the abstract property of maximizing value.

In a case where the action actually maximizes value, (i) entails (ii). So your counterfactual test can't count against (i) on those grounds. But you also ask: "Would (ii) make my action right without (i)? Yes. So the competition should be resolved in favor of (ii)."

That's where you mistakenly assume that all acts must be right for the same reason. (ii) could be satisfied, and so an action could *be* right, for a different reason from why *this* action is right. It could be that instead of the complete grounding explanation in (i), a different act's rightness is grounded in complete details (i*). The plausibility of diverse normative explanations is a further reason to favor (i) over (ii) as playing the normative explanatory role.

Expand full comment

This is really interesting! Just for clarity, I take you and Tucker both agree that 1) a movitivation/attitude being fitting doesn't answer the question of whether or not one morally ought to have it, since fitting motivations can sometimes be instrumentally bad, and 2) the motivations we morally ought to have are the ones that make the world go best (which seems undeniable for a consequentialist!).

So it seems the disagreement is on what to do with fittingness. I'm sympathetic to your point that talk of truth and falsity doesn't get to the heart of the criticism we want to make of someone with perverse motivations. But I'm not sure that fittingness does either. Fittingness alone doesn't carry heavy moral baggage--it'd be unfitting for me to prefer to watch Desperate Housewives over Game of Thrones, but that isn't a serious criticism of my moral character. We need moral stakes for a criticism to carry moral weight; when someone desires that others suffer that seems like a motivation that could really harm others and make the world worse.

Anyhoo, if we want to say that fittingness doesn't convey serious criticism on its own but does when there are moral stakes at play, I don't see why Tucker couldn't say the same of truth/falsity. Having false beliefs alone isn't a serious criticism, but it can be when those false beliefs have stakes.

At any rate, if it's the stakes at hand that are doing the work in our criticisms of others, I don't see what difference it makes whether we talk of a motivation being false or unfitting (and I agree they may be the same thing!).

Expand full comment

I think "ought" is prima facie ambiguous when applied to judgment-sensitive attitudes like beliefs and desires (though as we'll see, I actually think there's something conceptually defective about your preferred usage!). It can either pick out:

* fitting / rationally warranted attitudes -- e.g. epistemically justified beliefs, and desires whose objects are indeed desirable, the latter being a kind of "moral ought". Now, just as consequentialists needn't feel the slightest pressure to deny epistemologists' claims about what people ought (or have reason) to believe, so we needn't deny the corresponding sorts of claims about what people ought (or have reason) to desire.

* (instrumentally) good/desirable attitudes, i.e. the ones we ought *to choose* if given the choice.

Because the latter concerns our reasons for action, rather than our reasons for the attitudes in question, I think it's actually misleading to say that we ought to desire [objective bads that it would be instrumentally valuable to desire]. We ought to [try to bring about such a desire], but that's not the same thing. The rationality of choosing a state does not entail the rationality of the state itself, as per Parfit's "rational irrationality".

As I explain in my Beyond Right & Wrong draft:

> I personally find it clearer to think of all reasons as reasons of fittingness. For example, an apparent "practical reason for belief" can be better understood as a higher-order fitting reason to *desire* to have that belief. After all, what is made rational by *the fact that believing p would benefit you* is not the epistemic process of believing *p*, but the practical process of wanting and pursuing *belief in p*. This practical consideration is no more a "reason for belief" than my gustatory reasons for wanting to eat chocolate are "reasons for chocolate". (Chocolate doesn't respond to reasons of this sort; my desires do. Likewise, reasons for desiring a belief are not reasons for beliefs. The mere fact that beliefs, as judgment-sensitive attitudes, are capable of responding to *their own* types of reasons shouldn't lead us to mistake practical reasons as ever being directed *to* them---even when a belief is the goal-state that a reason-for-desire directs us *to pursue*.)"

* * *

So, I think it's conceptually confused to hold that one "morally ought to *have* instrumentally good attitudes." Ought entails reasons, and reasons don't work like that. We morally ought to *choose* instrumentally good attitudes (and desire them, be glad to have them, and so on). But we morally ought to *want* the objects that it's morally fitting to want.

There's no conflict here, because we shouldn't necessarily want (or choose) to have fitting attitudes. It's perfectly coherent to hold that we ought to bring it about that we have desires whose objects are not the things we ought to desire. It's just another form of rational irrationality (or virtuous viciousness, or fitting unfittingness).

I think the underlying mistake of consequentialists who resist this line of thought is that they are implicitly assuming something like:

*(Oughts are Practical Endorsements):* If S sincerely believes that one ought to believe that p, then S whole-heartedly endorses the state of believing that p, and would discourage one from taking a magical pill that induces the opposite belief.

But this principle is silly and nobody ought to believe it. Once you explicitly reject it, you can more easily appreciate how we can say that you ought to desire [good things] even though we don't *practically endorse or recommend choosing* that desire (if it turns out to be instrumentally bad).

Expand full comment

Okay, it's helpful to distinguish between the different oughts that are at play. I wonder if there's anything to be said for some oughts mattering more than others. I'm not sure quite how to cash out this thought, but I'm inclined to think that the sort of moral ought that guides choices is authoritative, and has a kind of must-be-done-ness built into it that (say) the epistemic ought doesn't.

One thing we could say is that the we as moral agents are concerned with what choices to make, and so the ought that ranges over choices is action-guiding in a way that oughts ranging over beliefs or attitudes won't be. But I think there's more going on; the criticisms we make in light of someone acting wrongly seem much more serious than criticisms of someone's irrational beliefs or unfitting attitudes.

I remember us talking about having different conceptions of this 'authority' idea, so this may just be a point of disagreement. But I do wonder if, on the fittingness-first approach, there's any way to say that some oughts matter more than others (or if that's just something you want to deny). Thanks for the response!

Expand full comment

Yes, I think the normative cluster associated with value and choiceworthiness has much greater normative significance than, say, irrational beliefs. It's more important -- which is to say that we should *care more* about it. Or, as I set out in an old post: (https://www.philosophyetc.net/2014/05/fittingness-and-normativity.html )

> The difference in normative "quality" comes not from the raw status of "being a fittingness relation", but rather from the significance of the attitudes which are thereby warranted. Warranting belief is perhaps a fairly weak kind of normative significance, whereas warranting intense moral outrage is another matter altogether. One misses these important variations in normative significance if all one considers is the relation of warrant, glossing over the crucial question of what response is thereby being warranted.

Expand full comment

I think the problem that I have with fittingness is that I don’t think there is something to fit here. I think something is good for someone because it is desired by them, while fittingness seems to need some source of goodness for someone that is prior to their desires. One could say that it is fitting for others to desire things because they are desired by the one that it affects, but I suspect this will run into problems when there are deep conflicts of interest.

In any case, I suspect it’s better to make ones behaviors track all desires, rather than making them only track one’s own desires and make one’s desires track the desires of others.

Expand full comment

How could one's behavior track all desires without one's motivations doing the same? What is motivating the behavior in question?

I think it's fitting to desire that people's lives go well. If the latter is a matter of desire satisfaction, then I'd say that it's fitting to desire that people's desires (in general) are satisfied. That's why helping people to achieve their (harmless) goals is virtuous, and gratuitously obstructing them is vicious.

This would risk vacuity if everyone had *only* this fitting desire, but human nature furnishes us with plenty of other desires whether they're warranted or not. (I take it that most desires, on this account, are not "unwarranted" in the sense of being normatively criticizable, but just non-warranted: we might say the normative facts don't commit one either way regarding the question of what first-order desires to have.)

Expand full comment