I actually still think moral perfection is impossible, not because I disagree with anything in this article, but because I define moral perfection to consist not just in always making permissible choices, but in always choosing the best possible option, even among permissible ones. And there are possible scenarios where there is no best possible option, but instead an infinite succession of better and better possible options. You don't sin by choosing a sub-optimal option in such a scenario, since it's not possible to choose an optimal option, but you do fail to be morally perfect.
I'm not sure I've ever heard anyone say that they read Consciousness Explained and were convinced. *I* think it's fantastic, and I'm constantly assigning undergraduates to read Quining Qualia. Happy to hear someone else finds it as powerful as I do!
Interestingly, what you've described here is basically the Buddhist conception of enlightenment in a nutshell: removing the ignorance that makes you think there is such a thing as an overwhelming temptation that forces you to do anything, replacing that with a deep sense of full responsibility for all your choices, and then developing that habit so your actions are automatically in line with what you know to be right or wrong. The temptation combined with the ignorance of your ability to say no to it is basically what craving is, and craving is what suffering is, which is why Buddhist enlightenment is pretty much what you've described here: moral perfection and the elimination of suffering go hand in hand. The main difference would be what actions constitute moral perfection, which are defined as the actions that cause suffering.
The premise and conclusion only say of each choice, individually, that they have a non-wrong option. It doesn’t follow that in the case of the one very big action made up of the individual choices—namely, the singular choice to do one non-wrong thing in each scenario.
Deontic logic is definitely a thing though, and I’m sure someone has formulated a formal proof along these lines, but I imagine defending the premises requires informal arguments along the lines of what I gave.
Right, but if we assume that to act wrongly requires taking a wrong action in some single choice, then it still follows? (Maybe this assumption is too substantive for the purposes of the proof though.)
Suppose that someone were possibly unable to act non-wrongly for a whole life. That would then require that there be some point where they were forced to pick a wrong option in a choice, but that just contradicts the premise, as there are no such choices.
Another way to look at it: If some past choice were such that it made you unable to act non-wrongly in some future choice, then that would mean that there were now a future choice where you could not pick any non-wrong option. But again, that just contradicts the premise. Mutatis mutandis for any case where collections of choices rule out non-wrong options.
You can be obligated to do different things because obligations are (when safe & coherent) indexed to some obliging/expecting act. Steve can expect you to do this, Leroy can expect you to do that; voila, different obligations.
Now, Leroy's expectation is fair: He expects $5 from you today because you said you'd pay him back today. Steve's in unfair: He wants $5 from you for nothing and for no good reason, just a bad reason, Steve's greed. So although both have indebted you ("ought" means "owed"), you're not going to respect Steve's indebting/obligation.
You don't have $5 today, so you can't pay Leroy back today. You only have $1. Does this mean there is no fair obligation anymore? After all, you can't pay him back today. Well... you can at least give him $1, right? If you gave him zero, Leroy would probably hold you to account (not just regarding the owed $5 he already accounts, but now also your own faith & credit with him, that is, your reputation takes a hit if you don't do what you can). In other words, you'll be held culpable (by Leroy) to the extent that you are not doing what you *could* do under the antecedent of an acquiescent will.
(You then find out that he's out of town, and paying him anything is impossible; you breathe a sigh of relief because you're now "off the hook" for today, but per Leroy's probable expectations, the obligation is evolving to pay him back as soon as we can again make contact.)
So the chestnut "ought implies can" is quite a bit too facile and should not be used in serious moral reasoning. The more rigorous revision is a mouthful: "A minimum criterion for a *fair* ought is that it is accounted to the extend someone could contribute if they were cooperative (whether or not they are), and not more. This is just one criterion; there may also be more criteria for fairness depending on the myriad expectations & relationships at play in some given circumstance."
Perfection is "the quality or state of being perfect" and perfect is "being entirely without fault or defect". By definition, therefore, your new definition of "doing nothing immoral in a way one is culpable for" allows for someone to potentially have a state or quality without being entirely without moral fault or defect.
>Some people really, truly feel it’s wrong to commit one murder to save a billion people. That seems very unlikely to me.
Are you a Kantian who would pull the lever on the trolley problem? Or if you wouldn’t, what’s the difference? That surprises me!
>doing nothing immoral in a way one is culpable for.
This sounds a bit too close to slave morality, where the best thing to do is make yourself have as little power as possible so you can’t do anything immoral. There’s some differences in your argument, but how do you personally think your belief differs from that? https://www.astralcodexten.com/p/matt-yglesias-considered-as-the-nietzschean
If by perfect we mean perfect given constraints then sure, but it might be better labeled competent and blameless (as opposed to blameless because innocent) or something. Since perfect might also mean admits of no improvement by changing anything including the constraints. It seems unlikely you can be a being that admits of no moral improvement by relaxing or varying constraints whenever you want since you are presumably so constrained, but it seems important to acknowledge that you would be improved by relaxing of constraints because some constraints admit of relaxing (for example if you can't do something worthwhile because you are wearing a seatbelt, makes sense to recognize if you release that constraint more good things would be possible).
I think its possible to be morally competent and blameless, if that is what can means then sure. Yet it remains unclear that it is probable, statistically likely, a reasonable expectation or so on and in fact from my acquaintance with humanity it seems highly improbable that one would attain it. So can as in o"ne can easily be morally blameless" or "one can reasonable expect to be morally blameless", I'd need more than this argument.
I like to combine this view with necessitarianism, yielding the flattering conclusion that I have always been morally perfect all along.
I actually still think moral perfection is impossible, not because I disagree with anything in this article, but because I define moral perfection to consist not just in always making permissible choices, but in always choosing the best possible option, even among permissible ones. And there are possible scenarios where there is no best possible option, but instead an infinite succession of better and better possible options. You don't sin by choosing a sub-optimal option in such a scenario, since it's not possible to choose an optimal option, but you do fail to be morally perfect.
I'm not sure I've ever heard anyone say that they read Consciousness Explained and were convinced. *I* think it's fantastic, and I'm constantly assigning undergraduates to read Quining Qualia. Happy to hear someone else finds it as powerful as I do!
I love your writing style this is such a fun read. It’s as if I’m in the room with you.
☺️☺️☺️
Interestingly, what you've described here is basically the Buddhist conception of enlightenment in a nutshell: removing the ignorance that makes you think there is such a thing as an overwhelming temptation that forces you to do anything, replacing that with a deep sense of full responsibility for all your choices, and then developing that habit so your actions are automatically in line with what you know to be right or wrong. The temptation combined with the ignorance of your ability to say no to it is basically what craving is, and craving is what suffering is, which is why Buddhist enlightenment is pretty much what you've described here: moral perfection and the elimination of suffering go hand in hand. The main difference would be what actions constitute moral perfection, which are defined as the actions that cause suffering.
Can't you make the proof more simple?
Cx=x is a choice
Rxy=x is a non-wrong option in y
Premise: Ax(Cx->EyRyx)
Conclusion: ~Ex(Cx&Ay~Ryx)
A and E are quantifiers (I couldn't be bothered)
Maybe you'd still want a proof from "there are no choices with no right options" to "you can always act rightly."
Then you'd just need the premise that you cannot act wrongly unless you have some choice with a wrong option, which seems close to a conceptual truth.
Although all of this is also irrelevant as you already have a proof lol
The premise and conclusion only say of each choice, individually, that they have a non-wrong option. It doesn’t follow that in the case of the one very big action made up of the individual choices—namely, the singular choice to do one non-wrong thing in each scenario.
Deontic logic is definitely a thing though, and I’m sure someone has formulated a formal proof along these lines, but I imagine defending the premises requires informal arguments along the lines of what I gave.
Right, but if we assume that to act wrongly requires taking a wrong action in some single choice, then it still follows? (Maybe this assumption is too substantive for the purposes of the proof though.)
Suppose that someone were possibly unable to act non-wrongly for a whole life. That would then require that there be some point where they were forced to pick a wrong option in a choice, but that just contradicts the premise, as there are no such choices.
Another way to look at it: If some past choice were such that it made you unable to act non-wrongly in some future choice, then that would mean that there were now a future choice where you could not pick any non-wrong option. But again, that just contradicts the premise. Mutatis mutandis for any case where collections of choices rule out non-wrong options.
You can be obligated to do different things because obligations are (when safe & coherent) indexed to some obliging/expecting act. Steve can expect you to do this, Leroy can expect you to do that; voila, different obligations.
Now, Leroy's expectation is fair: He expects $5 from you today because you said you'd pay him back today. Steve's in unfair: He wants $5 from you for nothing and for no good reason, just a bad reason, Steve's greed. So although both have indebted you ("ought" means "owed"), you're not going to respect Steve's indebting/obligation.
You don't have $5 today, so you can't pay Leroy back today. You only have $1. Does this mean there is no fair obligation anymore? After all, you can't pay him back today. Well... you can at least give him $1, right? If you gave him zero, Leroy would probably hold you to account (not just regarding the owed $5 he already accounts, but now also your own faith & credit with him, that is, your reputation takes a hit if you don't do what you can). In other words, you'll be held culpable (by Leroy) to the extent that you are not doing what you *could* do under the antecedent of an acquiescent will.
(You then find out that he's out of town, and paying him anything is impossible; you breathe a sigh of relief because you're now "off the hook" for today, but per Leroy's probable expectations, the obligation is evolving to pay him back as soon as we can again make contact.)
So the chestnut "ought implies can" is quite a bit too facile and should not be used in serious moral reasoning. The more rigorous revision is a mouthful: "A minimum criterion for a *fair* ought is that it is accounted to the extend someone could contribute if they were cooperative (whether or not they are), and not more. This is just one criterion; there may also be more criteria for fairness depending on the myriad expectations & relationships at play in some given circumstance."
Respectfully, can’t this article be summarized as “you can achieve moral perfection if I define moral perfection to be something it is not?”
Respectfully, because that would be an inaccurate summary.
Perfection is "the quality or state of being perfect" and perfect is "being entirely without fault or defect". By definition, therefore, your new definition of "doing nothing immoral in a way one is culpable for" allows for someone to potentially have a state or quality without being entirely without moral fault or defect.
>Some people really, truly feel it’s wrong to commit one murder to save a billion people. That seems very unlikely to me.
Are you a Kantian who would pull the lever on the trolley problem? Or if you wouldn’t, what’s the difference? That surprises me!
>doing nothing immoral in a way one is culpable for.
This sounds a bit too close to slave morality, where the best thing to do is make yourself have as little power as possible so you can’t do anything immoral. There’s some differences in your argument, but how do you personally think your belief differs from that? https://www.astralcodexten.com/p/matt-yglesias-considered-as-the-nietzschean
If by perfect we mean perfect given constraints then sure, but it might be better labeled competent and blameless (as opposed to blameless because innocent) or something. Since perfect might also mean admits of no improvement by changing anything including the constraints. It seems unlikely you can be a being that admits of no moral improvement by relaxing or varying constraints whenever you want since you are presumably so constrained, but it seems important to acknowledge that you would be improved by relaxing of constraints because some constraints admit of relaxing (for example if you can't do something worthwhile because you are wearing a seatbelt, makes sense to recognize if you release that constraint more good things would be possible).
I think its possible to be morally competent and blameless, if that is what can means then sure. Yet it remains unclear that it is probable, statistically likely, a reasonable expectation or so on and in fact from my acquaintance with humanity it seems highly improbable that one would attain it. So can as in o"ne can easily be morally blameless" or "one can reasonable expect to be morally blameless", I'd need more than this argument.
Oh, so we're having the road to hell today? SICK! I'll get the bricks. 🧱
One of the most obviously ridiculous sayings that exist
I'm gonna tell Saint Bernard you said that.