NationStates Jolt Archive

Greetings NSG! I've been a long time lurker on these forums and have posted a few things before under different names. Today I'm posting something several orders of magnitude more nerdy than the usual conversations on this forum, namely, a definition of morality that invokes basic set theory. Allow me to explain.

The way I see it, there are three different types of moral systems: arbitrary guideline systems, principled maximization systems, and means to an end (no not the troll) maximization systems.

Arbitrary guideline systems are just sets of rules with no real underlying philosophy. There is no grand unifying logic in their prohibitions, even if most or all of their rules make sense.

Both principled and means to an end maximization systems have an underlying philosophy, namely, to maximize certain qualities, which shall henceforth be referred to as maximizables. The libertarian political philosophy, for example, is based on the idea of maximizing both personal and economic freedom. The difference between principled and means to an end systems is that principled systems do not permit all actions to raise the magnitude of the maximizables but means to an end systems do.

Let S denote the set of all physically possible scenarios.
Let M denote the set of all maximizables and m be a subset of M.
Let P denote the set of all actions and p be a subset of P.

A moral framework, q, is the Cartesian product m x p and denotes a given moral system. q is a principled maximization iff p is a proper subset of P.
Let Q denote the set of all moral frameworks.

The maximization function, upsilon, is a mapping Q x S -> R^n. R^n is an n dimensional Euclidean space whose component axes are represented by m1 through mn where mi represents the value of the change in a given maximizable in m. Thus upsilon(q,s) for a particular value of q and s produces a vector-valued curve in n dimensional space, where n is the number of maximizables in m, whose component along the the mi axis denotes the positive or negative change in the mi maximizable for a given action p. Thus, for a specific value of q and s, the upsilon function returns n functions of the parameter, p. The value (m1^2 + m2^2 + ... mn^2)^1/2 for a given p is the magnitude of the vector, and thus represents, in some sense, the total net change in all maximizables. It is usually a good idea to maximize the magnitude, unless, of course, the vector points mostly along -mi axes.

I feel that, in this sense, all maximization systems can thus be objective moralities. This does not demonstrate the existence of an absolute morality, which I don't even believe in to begin with, but it does generalize the notion of morality into specific classes. I feel that maximization systems are, in general, superior to arbitrary guideline systems. I personally follow a principled maximization system wherein my maximizables are knowledge (defined as justified true belief) and personal freedom.

The way I see it, there are three different types of moral systems: arbitrary guideline systems, principled maximization systems, and means to an end (no not the troll) maximization systems.


Where would something like Kantian ethics, or Virtue ethics, or natural law come into this?


I feel that, in this sense, all maximization systems can thus be objective moralities.

I don't see it, a function that works out how to best maximise an action, subject to a constraint, is only objective if you rely on the assumption that maximising something is good in itself, and that you can differentiate objectively between something good that ought to be maximised, and something bad. Both of these require subjectivism.


This does not demonstrate the existence of an absolute morality, which I don't even believe in to begin with, but it does generalize the notion of morality into specific classes.

Ah, ok then.

I personally follow a principled maximization system wherein my maximizables are knowledge (defined as justified true belief) and personal freedom.

So basically a form of utilitarianism?

If you doze off while watching a football game with your friends, you risk getting awakened via a strike to the groin with a rubber mallet. *nod*

And yeah, this is super nerdy, but that's a good thing! Right? Riight!? :D

Arbitrary guideline systems are just sets of rules with no real underlying philosophy. There is no grand unifying logic in their prohibitions, even if most or all of their rules make sense.

Both principled and means to an end maximization systems have an underlying philosophy, namely, to maximize certain qualities, which shall henceforth be referred to as maximizables. The libertarian political philosophy, for example, is based on the idea of maximizing both personal and economic freedom. The difference between principled and means to an end systems is that principled systems do not permit all actions to raise the magnitude of the maximizables but means to an end systems do.
Are there really any moral systems that are arbitrary? It seems all have an underlying goal, something to be maximized, even if that something is "satisfaction brought to God" or another similarly unclear goal.

Well, first of all, I'm not a big fan of Kantian ethics. Ultimately, I feel that all meaningful moral considerations must take the consequences of an action into consideration. Besides, from whence does Kant's assumption that lying is always wrong (which is highly debatable and even morally reprehensible in certain situations) come? Lying is wrong because one wouldn't want to live in a world where everyone lies. Well duh, but that seems, at its base, a consideration of consequences in a universal rather than local sense. He also made references to duty, which is rather ill=defined if you ask me. Granted, I don't believe that consequences should be the only consideration, and that's why I consider the possibility of principled maximization systems.

As for your statement that the maximization functions rely on the assumption that maximizing things is inherently good, well yes, you're absolutely right. However, M refers to the set of ALL maximizables, so it is thus an uncountable set. If you think that minimizing suffering is good, that is the same as maximizing the absence of pain.

Hmm, well I'm a little too tired to deal with this now. I'll explain why I think you're wrong tomorrow. :cool:

Are there really any moral systems that are arbitrary? It seems all have an underlying goal, something to be maximized, even if that something is "satisfaction brought to God" or another similarly unclear goal.

Well, the scientist in my says "probably not, as there is no evidence to suggest their existence" but the mathematician says "who cares, let's put it there so that we consider every possible case." As for where Kantian ethics fit in, I don't know, perhaps they could be classified as a very tightly restricted principled maximization system wherein the maximization function does not consider the change in the maximizables due to one course of action but rather the global change in the maximizables due to the course of action being applied universally?

Where is the place in this framework for the distinction between "strong" and "weak" positive duties?

Most "principled" moral theories rest ultimately on a distinction between bringing something about and merely letting it happen; this is the only way they can maintain that, say, murdering one person to save ten others from murder is unethical. But associated with that distinction is the argument that while a negative duty not to murder is a duty of necessity, something we must do, a positive duty to prevent the murders of others is a duty of goodness or virtue, a decent thing to do but not a strict, necessary obligation in any particular concrete circumstance.

That means, however, that the "maximizable" you refer to is a very different thing in a utilitarian theory (which makes no such distinction) than it is in at least many "principled" ones that reject ends-justify-the-means logic.

Well, the scientist in my says "probably not, as there is no evidence to suggest their existence" but the mathematician says "who cares, let's put it there so that we consider every possible case."
Right, I don't all underlying goals actually make sense, but they're there.
As for where Kantian ethics fit in, I don't know, perhaps they could be classified as a very tightly restricted principled maximization system wherein the maximization function does not consider the change in the maximizables due to one course of action but rather the global change in the maximizables due to the course of action being applied universally?
That could be, or something as simple as maximizing one's personal virtue.

If you doze off while watching a football game with your friends, you risk getting awakened via a strike to the groin with a rubber mallet. *nod*

No, if I doze off while watching a football game with your friends, I risk getting awakened via a strike to the groin with a rubber mallet. :o

I usually like using math to do things, as they seem to reflect a more thought out answer. But I do have a few questions.

Explain to me how you quantify variable like "true beliefs" or "personal freedom"?

Also, does this equation account for investment cost in the sense that each resource used to maximize one field are resources takes away from maximizing another? In that sense you will probably need to create a equation to compare the quantified true believe with personal freedom so that both side is equal (and so on for all necessary variables).

Greetings NSG! I've been a long time lurker on these forums and have posted a few things before under different names. Today I'm posting something several orders of magnitude more nerdy than the usual conversations on this forum, namely, a definition of morality that invokes basic set theory. Allow me to explain.

The way I see it, there are three different types of moral systems: arbitrary guideline systems, principled maximization systems, and means to an end (no not the troll) maximization systems.

Arbitrary guideline systems are just sets of rules with no real underlying philosophy. There is no grand unifying logic in their prohibitions, even if most or all of their rules make sense.

Both principled and means to an end maximization systems have an underlying philosophy, namely, to maximize certain qualities, which shall henceforth be referred to as maximizables. The libertarian political philosophy, for example, is based on the idea of maximizing both personal and economic freedom. The difference between principled and means to an end systems is that principled systems do not permit all actions to raise the magnitude of the maximizables but means to an end systems do.

Let S denote the set of all physically possible scenarios.
Let M denote the set of all maximizables and m be a subset of M.
Let P denote the set of all actions and p be a subset of P.

A moral framework, q, is the Cartesian product m x p and denotes a given moral system. q is a principled maximization iff p is a proper subset of P.
Let Q denote the set of all moral frameworks.

The maximization function, upsilon, is a mapping Q x S -> R^n. R^n is an n dimensional Euclidean space whose component axes are represented by m1 through mn where mi represents the value of the change in a given maximizable in m. Thus upsilon(q,s) for a particular value of q and s produces a vector-valued curve in n dimensional space, where n is the number of maximizables in m, whose component along the the mi axis denotes the positive or negative change in the mi maximizable for a given action p. Thus, for a specific value of q and s, the upsilon function returns n functions of the parameter, p. The value (m1^2 + m2^2 + ... mn^2)^1/2 for a given p is the magnitude of the vector, and thus represents, in some sense, the total net change in all maximizables. It is usually a good idea to maximize the magnitude, unless, of course, the vector points mostly along -mi axes.

I feel that, in this sense, all maximization systems can thus be objective moralities. This does not demonstrate the existence of an absolute morality, which I don't even believe in to begin with, but it does generalize the notion of morality into specific classes. I feel that maximization systems are, in general, superior to arbitrary guideline systems. I personally follow a principled maximization system wherein my maximizables are knowledge (defined as justified true belief) and personal freedom.

Right now, I'm seeing this as being an 'anti-sausage'. That is, (the geek in me means) I'm happy to see how it is produced, but I don't see any way to use it...?

No, if I doze off while watching a football game with your friends, I risk getting awakened via a strike to the groin with a rubber mallet. :o

Semantics(some antics?).

Semantics(some antics?).

It's more the 'splitting hairs' I'm worried about. :o

Right now, I'm seeing this as being an 'anti-sausage'. That is, (the geek in me means) I'm happy to see how it is produced, but I don't see any way to use it...?

This raises some odd questions, like... does this mean you know how anti-sausage is produced?

And obviously anti-sausage would be combined with sausage to create an energetic reaction that goes well with ketchup. Or to destroy the world. The possibilities are endless, the only way you 'don't see any way to use it' is if you're lying because you control and are illegally hoarding the world's remaining sausage supply!

I usually like using math to do things, as they seem to reflect a more thought out answer. But I do have a few questions.

Explain to me how you quantify variable like "true beliefs" or "personal freedom"?

Also, does this equation account for investment cost in the sense that each resource used to maximize one field are resources takes away from maximizing another? In that sense you will probably need to create a equation to compare the quantified true believe with personal freedom so that both side is equal (and so on for all necessary variables).

The fact that maximizing one could take away from another would be reflected by the signs of the various components of the vector quantity. Ultimately, each maximizable would be linearly independent of the other, a basic property of vectors, but the fact that each maximizable is directly dependent upon the parameter of action takes into account consequences of going "too far" in maximizing just one.

As for quantifying personal freedom and knowledge, yeah I agree, that's rather difficult. However, even if I can't directly measure either, we do have many ways of measuring both. The knowledge of a society can be measured partially through the number of academic papers published in a given length of time or the rate of graduation of higher education, and many other things. Doesn't the UN also do statistical analysis on the civil liberties of various nations? If I could locate other similar studies by other organizations and perform a meta-analysis, it would provide a somewhat statistically significant basis with which to construct a basic metric.

It's more the 'splitting hairs' I'm worried about. :o

It only hurts for a couple days. :)

Ultimately, I feel that all meaningful moral considerations must take the consequences of an action into consideration.

Sometimes? Yes. Kant's theory doesn't dispute this. I certainly cannot will maxims allowing potentially catastrophic actions as universal law.

But is considerations always the determinative factor? That stronger claim, Kant rejects. Then again, so do you.

Lying is wrong because one wouldn't want to live in a world where everyone lies.

That is actually not his argument, but I won't get into this because I think that his stance on lying does not (clearly or indisputably) follow from his "universal law" formulation of the categorical imperative.

The more compelling, more important argument against lying comes from the notion of treating people as ends-in-themselves: to lie, to present false reasons to someone for a manipulative end, is to deprive her of the freedom to decide for herself, according to her own judgment of the actual circumstances.

Well duh, but that seems, at its base, a consideration of consequences in a universal rather than local sense.

So?

He also made references to duty, which is rather ill=defined if you ask me.

Kant's main point with regard to duty is simply that one should do what is right because it is right, not for other reasons. People have objected to this on numerous (generally poor) grounds, but I'm not sure what's "ill-defined" about it.

This raises some odd questions, like... does this mean you know how anti-sausage is produced?

And obviously anti-sausage would be combined with sausage to create an energetic reaction that goes well with ketchup. Or to destroy the world. The possibilities are endless, the only way you 'don't see any way to use it' is if you're lying because you control and are illegally hoarding the world's remaining sausage supply!

Anti-sausage supply, surely.

I mean... nuh uh? What?

The fact that maximizing one could take away from another would be reflected by the signs of the various components of the vector quantity. Ultimately, each maximizable would be linearly independent of the other, a basic property of vectors, but the fact that each maximizable is directly dependent upon the parameter of action takes into account consequences of going "too far" in maximizing just one.

Actually I was thinking about something along the line of limited resources, such that if I got 100 buck and want to do something good with it, where do I spend it on. Even if we assume that variable doesn't affect each others directly I am sure they don't always come with the same price tag. Not sure if I am reading your formula correctly, but somewhere in there it got to account for cost efficiency don't you think?

It only hurts for a couple days. :)

It's not the couple of days I'm worried about. It's the darkness of the back of my eyelids followed by the simultaneous squishy noise and compressive feeling between the legs that signals an 'authentic LG wakeup(TM)'. It's the two seconds of naive bliss before, and the writhing rolling couple of minutes after, that worry me.

My old mama always used to say "never take a nap watching football with the kinds of bastards who would whack you in the goolies with hammers"... and it was good advice*.


(* in the interests of full disclosure, what she actually used to say was "there's many a mickle as makes a muckle", but I know what she meant.)

Asimov's three laws, applied to humans, is the best morality we'll ever do.

It's not the couple of days I'm worried about. It's the darkness of the back of my eyelids followed by the simultaneous squishy noise and compressive feeling between the legs that signals an 'authentic LG wakeup(TM)'. It's the two seconds of naive bliss before, and the writhing rolling couple of minutes after, that worry me.

My old mama always used to say "never take a nap watching football with the kinds of bastards who would whack you in the goolies with hammers"... and it was good advice*.


(* in the interests of full disclosure, what she actually used to say was "there's many a mickle as makes a muckle", but I know what she meant.)

Who the hell listens to their parents anyway? :p

Imll be honest....About half way through that I went lolwut?

*snip

Ergh, Kantian ethics always got me confused, mainly because of the clashes brought on by 'means to an ends' and 'universal law'. Which formulation of the C.I. has more weight?!

Which formulation of the C.I. has more weight?!

Both? Neither?

My own view is that they coexist independently as two separate manifestations of the respect for autonomy that is at the heart of the whole thing. It requires some creative interpretation to make this work well with the text, but in my view it makes the most logical sense. (Others disagree.)

Outright logical contradictions are avoided by the nature of the "universal law" formulation: by the "ends-in-themselves" formulation, you cannot legitimately will that someone be treated merely as a means, so you also cannot will as universal law a maxim that would involve such treatment.

Both? Neither?

My own view is that they coexist independently as two separate manifestations of the respect for autonomy that is at the heart of the whole thing. It requires some creative interpretation to make this work well with the text, but in my view it makes the most logical sense. (Others disagree.)

Outright logical contradictions are avoided by the nature of the "universal law" formulation: by the "ends-in-themselves" formulation, you cannot legitimately will that someone be treated merely as a means, so you also cannot will as universal law a maxim that would involve such treatment.

I see what you mean.. But the 'universal laws' principle is easily exploited. Framing of maxims leads to an infinite number of rules. I mean, you can create a maxim which applies solely to ones situation and by willing it become universal there is no problem with that.. So, you can effectively say that 'it is permissible for Joanna Black who is 50 years old to cheat on her husband and steal on a Monday', and 'universal laws' has no problem with that..

But if we go back to the humanity formula, these actions would clearly be immoral as she is using other people merely as a means to her own ends. Doesn't this, then, create a contradiction within the C.I.?

Besides, from whence does Kant's assumption that lying is always wrong (which is highly debatable and even morally reprehensible in certain situations) come?It's not an assumption, it's a conclusion. Lying is wrong because one wouldn't want to live in a world where everyone lies.Actually, lying is wrong because it treats the lied-to as a means-to-an-end. The universalizability rule you have in mind serves, for Kant, more as a heuristic device than a justification in itself.

Ergh, Kantian ethics always got me confused, mainly because of the clashes brought on by 'means to an ends' and 'universal law'. Which formulation of the C.I. has more weight?!While Soheran's "both-neither" suggestion has some truth to it, I'm going to say that the means-to-an-end formulation works more nearly as an expression of the actual values underlying the ethic. And I actually don't think it takes much "creative" interpretation to see that the real purpose of the universal-law formulation is to suggest a variation of the classic "do unto others" maxim: the point is to ask yourself whether YOU would want people lying to you--asking what happens if EVERYONE does so gives it a particularly sharp point (and tends to highlight certain "contradictions in the will"), but the compatibility with the means-to-an-end formulation lies essentially in that it presses the doer to try on the shoes of the done-to.

I see what you mean.. But the 'universal laws' principle is easily exploited.Perhaps that's one of the reasons Kant moves on from that principle to better ones.

Framing of maxims leads to an infinite number of rules. I mean, you can create a maxim which applies solely to ones situation and by willing it become universal there is no problem with that.. So, you can effectively say that 'it is permissible for Joanna Black who is 50 years old to cheat on her husband and steal on a Monday', and 'universal laws' has no problem with that.Actually, Kant specifically explains that this is not what he means by universalizing a rule. To universalize is to do an open-ended substitution: "for all x, y, it is permissible for x to do P to y."

So, you can effectively say that 'it is permissible for Joanna Black who is 50 years old to cheat on her husband and steal on a Monday', and 'universal laws' has no problem with that..

No, you couldn't. You could (presumably) not will this as applying to the circumstance of you being the husband.

More broadly, universalizing a maxim requires generalizing those aspects of the maxim that have merely personal significance. For instance, if my maxim is "To make Soheran happy, I may use any available means", my willing of the maxim is predicated on the relation between "Soheran" and "I" (the only reason it's "Soheran" and not "Peisandros" is because I am Soheran). Thus, to truly universalize my maxim, I must generalize its form to "To make myself (whoever I am) happy, I may use any available means"--and I clearly cannot will this as universal law.

But if we go back to the humanity formula, these actions would clearly be immoral as she is using other people merely as a means to her own ends. Doesn't this, then, create a contradiction within the C.I.?

In my view, no. Kant certainly invested much effort in the Groundwork attempting to show that the formulations have equivalent results, but I am not convinced, and more importantly, it is not logically necessary. I am bound by both formulations: if my course of action is permissible under one but forbidden under the other, it is morally prohibited. No contradiction there.

Perhaps that's one of the reasons Kant moves on from that principle to better ones.

Actually, Kant specifically explains that this is not what he means by universalizing a rule. To universalize is to do an open-ended substitution: "for all x, y, it is permissible for x to do P to y."

Oh really? Hmm.. I was basing what I said on Anscombe and not Kant directly, so that makes a bit more sense.

No, you couldn't. You could (presumably) not will this as applying to the circumstance of you being the husband.

More broadly, universalizing a maxim requires generalizing those aspects of the maxim that have merely personal significance. For instance, if my maxim is "To make Soheran happy, I may use any available means", my willing of the maxim is predicated on the relation between "Soheran" and "I" (the only reason it's "Soheran" and not "Peisandros" is because I am Soheran). Thus, to truly universalize my maxim, I must generalize its form to "To make myself (whoever I am) happy, I may use any available means"--and I clearly cannot will this as universal law.

Yeah it does make more sense to generalise. I could never find good Kant stuff in our library so had to base lots of my essays and stuff on other philosophers -- probably a bit biased.

In my view, no. Kant certainly invested much effort in the Groundwork attempting to show that the formulations have equivalent results, but I am not convinced, and more importantly, it is not logically necessary. I am bound by both formulations: if my course of action is permissible under one but forbidden under the other, it is morally prohibited. No contradiction there.

Interesting, I hadn't thought of that way. I mean I too am not convinced about the formulations having equivalent results.. I had always thought that because of that, the C.I. becomes somewhat redundant.. But that (bolded part) is a good way of looking at it.

but the compatibility with the means-to-an-end formulation lies essentially in that it presses the doer to try on the shoes of the done-to.

Even when the two coincide in the norms of treatment they suggest, though, they do so for different reasons.

"Universal law" certainly demands the sort of empathy you point to, but it is perfectly compatible with a kind of paternalism: "I will consider your interests, but I will do so on my terms (according to what I can will)." It essentially demands that we act responsibly in the full sense of the word, but since it is the autonomy of our own will that is key,

"Ends-in-themselves" demands consideration not so much of "interests", but autonomy: "I have my own valued ends and my own moral judgments, but I may not impose them on you."

I find the clearest demonstration to be the political applications: as I think I've argued to you before, "universal law" is compatible with a benevolent dictatorship ruled for the public good (but without public consent or participation), while "ends-in-themselves" is compatible with a democratic politics focused on identity and interest rather than universal right.

I see no particular reason to subordinate one formulation to the other; because they say very different things about moral action, they are best conceived of as independent, equal, and complementary.

The way I see it, there are three different types of moral systems: arbitrary guideline systems, principled maximization systems, and means to an end maximization systems.

Arbitrary guideline systems are just sets of rules with no real underlying philosophy. There is no grand unifying logic in their prohibitions, even if most or all of their rules make sense.

Both principled and means to an end maximization systems have an underlying philosophy, namely, to maximize certain qualities, which shall henceforth be referred to as maximizables. The libertarian political philosophy, for example, is based on the idea of maximizing both personal and economic freedom. The difference between principled and means to an end systems is that principled systems do not permit all actions to raise the magnitude of the maximizables but means to an end systems do.
Are you saying that both 'principled' and 'means to an end' maximization systems have, as their goal maximising certain qualities?

If you are, I don't quite see how would you could fit, say, Aristotelian virtue ethics into this framework, with its (non-quantifiable?) conception of eudaimonia. The human life of flourishing/happiness/well-being is a goal above maximising of certain qualities; a goal which they (the virtues) make possible.

I find the clearest demonstration to be the political applications: as I think I've argued to you before, "universal law" is compatible with a benevolent dictatorship ruled for the public good (but without public consent or participation)From a purely analytic perspective this may be true, but from an interpretive perspective I think it is clear that this is not how Kant intended it to be taken.

I see no particular reason to subordinate one formulation to the other; because they say very different things about moral action, they are best conceived of as independent, equal, and complementary.Whereas I think they only say "very different things" on a disingenuous reading of Kant--one which sees an analytic distinction and tries to resolve it in the neatest possible way, as opposed to one which recognizes a tension in Kant's presentation and attempts to resolve it by reflecting on what he might have said had he seen the need to do so.

If you doze off while watching a football game with your friends, you risk getting awakened via a strike to the groin with a rubber mallet. *nod*

People fall asleep during football games?! :eek:

well as i've put it previously and repeatedly, quite simply, the avoidance of knowingly causing avoidable suffering. that's the only REAL morality.

nothing against honest, sincere and nonfanatical beliefs, but they cannot, it seems to me, singly nor severally, offer such a completely universal standard.

i would also point out, that any attempt to legislate morality is itself an immoral act, both on the basis of my prefered definician, AND under the context of each and every major system of organized belief.

People fall asleep during football games?! :eek:

It's difficult, especially among my circle of friends, so when it happens it must be dealt with in a quick and decisive manner. *nod*

It's difficult, especially among my circle of friends, so when it happens it must be dealt with in a quick and decisive manner. *nod*

You have to be lying :p

You have to be lying :p

You realize that this is while watching games on television and not while actually playing them, right?

From a purely analytic perspective this may be true, but from an interpretive perspective I think it is clear that this is not how Kant intended it to be taken.

Maybe, but I'm not sure what point you're trying to prove: Kant may have believed that the "universal law" formulation does produce, in the final analysis, equivalent results to the other formulations, but this neither means that he was right nor (whether Kant is right or not) that I am wrong in noting that they proceed on different grounds. Simply adopting the perspective of the "done-to" may bring us to some notion of the worth of freedom, but it simply does not do so on the basis of the subjecthood of persons that underlies the humanity formulation.

While Kant does suggest the heuristic use of the "universal law" formulation, I don't think he ever suggests that it is an incomplete or lesser version of the other formulations. Most in accordance with common sense, perhaps, but that is hardly the same thing.

Whereas I think they only say "very different things" on a disingenuous reading of Kant

Well, I'm curious about the the kind of similarity you'd maintain, then.

You point to an equivalence in terms of the concern for others they both demand, but as I've said the sort of concern they demand is different. It's worth noting further that Kant's use of "Golden Rule" reasoning in the context of the universal law formulation is quite limited--if I recall correctly, only mutual aid proceeds from such reasoning. The "perfect" duty against lying he derives comes instead from straightforward universalization: it does not come from putting oneself in the shoes of the person lied to, but rather from imagining a world where the practice of lying for convenience is universal. This is even more distant from the "ends-in-themselves" logic of respect for autonomy.

as opposed to one which recognizes a tension in Kant's presentation and attempts to resolve it by reflecting on what he might have said had he seen the need to do so.

Alright; I'm just not sure how this interpretation actually solves the problem, insofar as there is one.

If you want to lay a groundwork for morality, make sure it's not tainted by religion.

I'm not sure I understood everything but I'm trying to.

maximizables are properties like alive? and m is a set of these properties. I'll assume three values for these for now (-1,0,1)

as in 'alive'

and actions in p are things which can happen to affect a change in aliveness.

like 'kill'

q is the set of all ordered pairs of elements in m and p

as in (kill, alive)

killing affects a change on our maximizable property alive.

by my interpretation we can look at q and a matrix.

S is a set of situations ... this is the part I really dont understand. How are we to define situations. I can imagine S more like elements in the world ... like mike and jane but S might be more like action sentences containing elements from p

as in mike killed jane.

I see S as something like a third dimension.

I know this is not quite how you put it but Am I too far off base here?

this is like: mi x pj x sk -> ri

this is like for the maximizable property mi and an action pj in the situation sk there is an outcome ri

Then you want to find the set of maximal q for all situations in S?

Rereading this I think I'm way off but I think my interpretation is also interesting ... but it wouldn't really represent morality... or anything like it. More like actions and outcomes.

Maybe, but I'm not sure what point you're trying to prove: Kant may have believed that the "universal law" formulation does produce, in the final analysis, equivalent results to the other formulations, but this neither means that he was right nor (whether Kant is right or not) that I am wrong in noting that they proceed on different grounds.

1) I think it's clear from the text that Kant thought the various formulations of the categorical imperative should (properly interpreted) produce equivalent results.

2) It is a well-accepted rule of interpretation, at least so far as interpretation is an historical project, that given alternative interpretations we should favor those that best accord with the intentions of the author. In this case, if we want to know how Kant meant us to take the universalizing test, we should favor interpretations that tend to produce results respecting mutual autonomy--after all, that was the purpose for which he advanced the test.

3) We actually agree that these formulations proceed on different grounds--and I think Kant would agree as well. If they were merely analytic reproductions of precisely the same understanding, this interpretation would render too much of his discussion redundant--and an important corollary to the above-stated interpretive rule is that we should never "interpret away" an author's rhetorical moves simply because some alternative offers a simpler or neater resolution. We should not assume that he intended the analytically simple or neat.

3a) This is, of course, one of the reasons that analytic philosophers detest Kant. What really gets injected into the several formulations of the categorical imperative is, from the analytic perspective, too much "psychology" that undermines logical rigor. Admittedly, Kant himself seems too resist psychological ethics as too "phenomenal" for his purposes, but I think this misses the point: what Kant actually manages to develop is the basis for thinking of a "psychology of reason" in terms of the abstracted subject--a move which Hegel takes much further, and with which he very rightly raises important criticisms against Kant, but a point to which he would not have arrived without Kant's lead.

3b) Fundamentally, if we are to remain faithful to Kant's intention we must reject any interpretation of universal law that allows for the scenario you suggest--that is, the possibility that a benevolent dictatorship can meet Kant's ethical criteria. In a certain sense, this should be obvious from the fact that the "ultimate" universal is that I should not treat another as a means-to-an-end, even when that end is her own happiness. If I cannot will this as a universal law, then neither can a benevolent dictator.

Simply adopting the perspective of the "done-to" may bring us to some notion of the worth of freedom, but it simply does not do so on the basis of the subjecthood of persons that underlies the humanity formulation.I think it does, but the "psychology" is different. The humanity formulation abstracts the subject's rebellion against being treated as an object to the level of a duty--but the subject rebels before it ever attains to the abstraction. Hegel put a much finer point on this, but it's clearly at the heart of Kant's ethics.

While Kant does suggest the heuristic use of the "universal law" formulation, I don't think he ever suggests that it is an incomplete or lesser version of the other formulations. Most in accordance with common sense, perhaps, but that is hardly the same thing.I beg to differ. The entire movement is one in which Kant develops a psychology that connects "common sense" to an abstracted duty. The progression of his formulations of the categorical imperative reflects this.

Well, I'm curious about the the kind of similarity you'd maintain, then.I maintain only the similarity that Kant proposes himself: that choice inevitably confers a duty, and that this duty must be identical for all choosing subjects.

You point to an equivalence in terms of the concern for others they both demand, but as I've said the sort of concern they demand is different. It's worth noting further that Kant's use of "Golden Rule" reasoning in the context of the universal law formulation is quite limited--if I recall correctly, only mutual aid proceeds from such reasoning.On the contrary, the Golden Rule remains operative throughout--and let us not forget that Kant was, unlike many of his contemporaries, a very religious man. On a very sensible, quite transparent reading, mutual respect between agents as ends-in-themselves is the ideal formulation of the Golden Rule: I must treat others as I would be treated, and the one universal covering how-I-would-not-be-treated is as an object rather than a subject.

The "perfect" duty against lying he derives comes instead from straightforward universalization: it does not come from putting oneself in the shoes of the person lied to, but rather from imagining a world where the practice of lying for convenience is universal. This is even more distant from the "ends-in-themselves" logic of respect for autonomy.Actually, he does neither. The movement of his logic in "On a supposed right to lie" actually tends to supercede the argument from autonomy in favor of an even more formal duty to the ideal itself. He arrives here: "And although by telling a certain lie I in fact do not wrong anyone [Note here that he distinguishes carefully between "harm" and "wrong"--his position at this point seems to suggest that a lie might, perhaps, not even "wrong" a person in the more abstract sense of violating his autonomy], I nevertheless violate the principle of right in regard to all unavoidably necessary statements generally (i.e., the principle of right is thereby wronged formally, though not materially)." He admits, in fact, that a lie may not even wrong the person lied-to.

(On this interpretation, the rule admits no exception even in the supposedly marginal cases of lying to children or other people deemed to lack full autonomy. Indeed, a lie is wrong even if no one hears it--a supposition which helps to make sense of his assertion that a person is a "liar" as soon as she so much as thinks about the possibility of exceptions to the rule, whether she actually "tells" a lie or not.)

Soheran: Apologies in advance if this is the last substantial word I have on the subject, at least for a while. I greatly enjoy our discussions, but I'm so overworked right now I can barely breathe. Certainly sleep is out of the quesion...

Fundamentally, if we are to remain faithful to Kant's intention we must reject any interpretation of universal law that allows for the scenario you suggest--that is, the possibility that a benevolent dictatorship can meet Kant's ethical criteria.

I find it hard to avoid that conclusion. It's certainly possible for me to think, "In certain circumstances (due to any number of problems with my own judgment), I would rather be ruled by a benevolent dictator than by my own will", or even, "Such and such a person deserves to rule over me whatever my view of the subject." If I can will subordination, then even under "universal law" I can also will being a master.

On what interpretation of the "universal law" formulation does this change?

In a certain sense, this should be obvious from the fact that the "ultimate" universal is that I should not treat another as a means-to-an-end, even when that end is her own happiness.

Right, and the moment we introduce the idea of humanity as an end-in-itself, it's clear that my earlier reasoning cannot be held by: I have no legitimate authority to impose my will upon others, even if my personal values allow for such a thing. But this only emphasizes the distinction I have drawn between the two.

I think it does, but the "psychology" is different. The humanity formulation abstracts the subject's rebellion against being treated as an object to the level of a duty--but the subject rebels before it ever attains to the abstraction.

The subject "rebel[ling]", though, is not strictly relevant under "universal law": I am not forbidden to will maxims that would be objectionable to other people. What I cannot do is will a maxim that I could not accept in the place of those other people, which both means that it is my values that decide, not theirs, and that the person who is judging the facts of the situation is me (in good faith, on objective grounds), not them.

This is, of course, different entirely from the approach of treating persons as ends-in-themselves, the entire point of which is to, in certain contexts, defer to another person's values and judgments regardless of our opinion of them.

I maintain only the similarity that Kant proposes himself: that choice inevitably confers a duty, and that this duty must be identical for all choosing subjects.

I don't see how maintaining that the formulations are fundamentally different changes this. They are both equally binding on "all choosing subjects", so the duty of each remains identical.

You realize that this is while watching games on television and not while actually playing them, right?

Games on TV? What?

(Yeah I did know, I still haven't ever seen people fall asleep, but then my family is full of rabid sports fans.

I find it hard to avoid that conclusion. It's certainly possible for me to think, "In certain circumstances (due to any number of problems with my own judgment), I would rather be ruled by a benevolent dictator than by my own will", or even, "Such and such a person deserves to rule over me whatever my view of the subject." If I can will subordination, then even under "universal law" I can also will being a master.If I will that the will of someone wiser should substitute for my own, I am still being ruled by my own will. Indeed, I can will that everyone should submit themselves (willingly) to the rule of the wisest among us.

What I cannot will is that the wisest among us, finding some unwilling to submit, should force his rule upon them--which is something like what you must mean by a benevolent "dictator." If we all submit willingly, this is really a kind of democracy.

Supposing myself to be the wisest person, I can will the following universal: "Everyone should submit to the will of the wisest among us (which happens to be me)." What I CANNOT will is, "Being the wisest, I should force others to submit to my will." While I might willingly submit to the wisest person, I would not will that I should be forced against my will--if ever there were a contradiction, surely this is it!!

[This universal is worsened further if we make the reasonable assumption that no one KNOWS herself, in fact, to be the wisest--and that there should be some disagreement as to this fact. In this case, the universal becomes, "If I believe myself to be the wisest, I should force everyone else to submit to my rule." This leads to a very clear contradiction in the will, in that I must will also that others believing themselves to be wisest should compel submission, which contradicts quite literally my own will to rule.]

I have no legitimate authority to impose my will upon others, even if my personal values allow for such a thing.You never had legitimate authority to impose your will on others. You only ever had authority to will that they should submit themselves to you--and that's different.

While I might willingly submit to the wisest person, I would not will that I should be forced against my will--if ever there were a contradiction, surely this is it!!

How is that contradictory?

"I don't trust myself to decide who is wisest, so if there is such a wise person, I will that I be forced against my will (founded on my poor judgment) to submit to that person."

Or, even more problematically: "I think Person X has a right to rule, one that has nothing to do with my consent." (Obviously, if this is my will, I do consent. But my understanding of the basis of that person's right to rule has nothing to do with my consent or anyone else's: I will that my subordination to that person be maintained regardless of my belief or my opinion.)

In universalizing a maxim, I do not actually adopt the will of the person in whose place I imagine myself: if that were the case, it would be impossible for me to ever say, "I understand that that person objects to this behavior, but I do not think his or her objection is compelling." I merely adopt the person's position, which allows me to say, "Even if it were me in those circumstances, I still think it would be right to use force." It's still my will that's the deciding factor, not the person's: it's my circumstances, not my volitional judgment, from which I have abstracted.

How is that contradictory?It's pretty simple. You can't will that you should do that which you do not will. Hence, you cannot will that you should be forced against your will.

"I don't trust myself to decide who is wisest, so if there is such a wise person, I will that I be forced against my will (founded on my poor judgment) to submit to that person."You're equivocating on the meaning of "will." It's possible you could will that you should be forced against your judgment, but you cannot will that you should be forced against your will.

Or, even more problematically: "I think Person X has a right to rule, one that has nothing to do with my consent."Actually, that's significantly less problematic, because it relies on my judgment that Person X has such a right. Hence, I can only universalize across people who also think that Person X has such a right: "Anyone who thinks Person X has a right to rule, one that has nothing to do with her/his consent, should submit to Person X." But that does absolutely nothing to justify X's tyranny against those who disagree.

In universalizing a maxim, I do not actually adopt the will of the person in whose place I imagine myself.Of course not. The principle of universalization, like all of Kant's ethics, is premised on the possibility of asserting will as an abstraction independent of the individual who wills. If we abandon this--if we believe that will is always particular--then we abandon the entire ethic. Indeed, this is essentially the postmodern critique of Kantian ethics.

I'm curious: do you adopt this perspective as a rhetorical device to win the argument, or have you actually shifted to the postmodern position that all universalization is, inevitably, a generalization from the particular?

...if that were the case, it would be impossible for me to ever say, "I understand that that person objects to this behavior, but I do not think his or her objection is compelling."How so? This is only true if I admit that there is no independent standard by which to judge objections, which is the whole point of the categorical imperative.

Certainly others may will things that are not consistent with our standard. When I take up the other's will, I judge it by the same standard I judge my own: empathy does not reduce to relativism, and I see no reason to suppose that it does. But that, again, is the postmodern position; it seems again that you are tending in that direction.

I merely adopt the person's position, which allows me to say, "Even if it were me in those circumstances, I still think it would be right to use force."Indeed, but in this case the "circumstance" MUST include the belief as to X's right, since that is the premise of your own maxim.

It does not make sense to say, "I know X has a right to rule, but even if I didn't know that I would still submit to X."

It's still my will that's the deciding factor, not the person's: it's my circumstances, not my volitional judgment, from which I have abstracted.If you believe that, then there's nothing left of Kant for you.

You're equivocating on the meaning of "will." It's possible you could will that you should be forced against your judgment, but you cannot will that you should be forced against your will.

Okay, but what's the difference in terms of force? At first brush, this distinction suggests to me that paternalism could be justified this way ("I will force you, against your wrong judgment, to fulfill the notion of 'good' you will"), but not a stronger kind of mastery ("I will force you to do what I will, whatever that may be.") I don't think that's what you mean, though.

Actually, that's significantly less problematic, because it relies on my judgment that Person X has such a right. Hence, I can only universalize across people who also think that Person X has such a right: "Anyone who thinks Person X has a right to rule, one that has nothing to do with her/his consent, should submit to Person X."

Compare: "I have (and everyone else has) a right to put up billboards promoting atheism, regardless of anyone else's objection or offense."

Is this similarly universalizable only across people who believe in such a right? Am I morally forbidden to put up such billboards if in a single case a person wills otherwise?

Of course not. The principle of universalization, like all of Kant's ethics, is premised on the possibility of asserting will as an abstraction independent of the individual who wills.

Yes, but this is coupled with the fact that this abstraction, in and of itself, has no content: it gives us the formal rule "Act only upon the maxim that you can will as universal law", but it doesn't actually directly provide us with any maxims. I exercise my own judgment as to which standards are applicable to will as such--and because, beyond the universalization formulation itself, there are no hard-and-fast rules here, differences among people in the standards adopted are inevitable. It is "my" will only in that sense.

I'm curious: do you adopt this perspective as a rhetorical device to win the argument, or have you actually shifted to the postmodern position that all universalization is, inevitably, a generalization from the particular?

Neither.

Certainly others may will things that are not consistent with our standard. When I take up the other's will, I judge it by the same standard I judge my own: empathy does not reduce to relativism, and I see no reason to suppose that it does.

I don't think it does either. That's precisely why I have maintained that to act against someone's particular will is in no sense clearly prohibited by the "universalization" formulation: as I said before, I may adopt the person's circumstances (including subjective traits like emotions and attitudes), but I need not adopt his or her moral judgments (more broadly, his or her judgments of objective truth and right). This distinction, which is presumed by the fact that empathy need not result in relativism, inherently requires that in adopting another person's position, I can will what that person materially does not: I can say, "This person wills x, but in this person's place I can will y." ("I understand that that person objects to this behavior, but I do not think his or her objection is compelling.")

Again: no contradiction. (But this is the essence of the difference between "universal law" and "ends-in-themselves": the latter requires me to sometimes defer to another's will whatever I think of it.)

Indeed, but in this case the "circumstance" MUST include the belief as to X's right, since that is the premise of your own maxim.

It does not make sense to say, "I know X has a right to rule, but even if I didn't know that I would still submit to X."

The hypothetical claim here is more like, "I know the categorical imperative is binding on all rational beings, but even if I didn't know that I would still be bound by it."

It is true that, materially, my act of submission (or my willing of submission) is predicated on the belief that submission is morally required. But the basis of my belief that submission is morally required need not be that belief itself.

Okay, but what's the difference in terms of force? At first brush, this distinction suggests to me that paternalism could be justified this way ("I will force you, against your wrong judgment, to fulfill the notion of 'good' you will"), but not a stronger kind of mastery ("I will force you to do what I will, whatever that may be.") I don't think that's what you mean, though.That's not what I mean.

The short answer (a cop-out of sorts): this is that terrible line at which ethics becomes politics. The difficulty in giving it any kind of sensible answer lies behind all the twists and turns by which first Rousseau and then Kant attempt to establish the sense in which I must regard the general will as "my" will even when my particular "will" disagrees. I tried to simplify it earlier by reducing this distinction to one of "will" versus "judgment," which is to a certain extent unfair to the language they use... but then, I think they too equivocate on the meaning of "will."

Ultimately the question comes down to this: under what circumstances can I suppose you to "will" a good that you do not expressly advocate? I don't think any politics perfectly resolves the problem, but I'm sure that a person supposing himself to be wise cannot make that presumption on the part of others; and I'm sure that if there IS a politics that resolves the issue, democratic/anarchist politics comes the closest by attempting, so far as possible, to dissolve the distinction between ruler and ruled.

Compare: "I have (and everyone else has) a right to put up billboards promoting atheism, regardless of anyone else's objection or offense."

Is this similarly universalizable only across people who believe in such a right? Am I morally forbidden to put up such billboards if in a single case a person wills otherwise?No, but you're missing an important analytic difference. Here you are testing the right itself--that is, you are asserting a right and attempting to universalize it: if it passes the test, you know you have such a right.

In your supposition about another person's right to rule, however, you do not universalize the right (you do not ask, "what if everyone had the right to rule")--rather, you make the right particular to this person, the wisest person. By holding it constant, you put it beyond your test: what you wind up testing is the question whether, given a person with such a right, other people should obey her.

Even if it survives in that form (as a maxim for submission), it utterly fails in the even more particularized formulation required to instruct the ruler to use violence against the disobedience: "Being that X [a particular person] has a right to rule, he should force everyone to do as he says." Here you haven't universalized across willing agents at all: you've only universalized across objects.

You might try to save this by suggesting we should universalized in the following way: "Anyone who happens to have the characteristics of X should force everyone to do as he says." But Kant tells us again and again that this is not at all what he means by universalizing a maxim. I cannot "sneak in the particulars" by universalizing over characteristics applying only to myself.

Games on TV? What?

(Yeah I did know, I still haven't ever seen people fall asleep, but then my family is full of rabid sports fans.

Well, when one is exhausted from work, bruised and battered from football or rugby for real and/or mildly intoxicated, it is a bit easier to doze off if one isn't careful.

You might try to save this by suggesting we should universalized in the following way: "Anyone who happens to have the characteristics of X should force everyone to do as he says." But Kant tells us again and again that this is not at all what he means by universalizing a maxim. I cannot "sneak in the particulars" by universalizing over characteristics applying only to myself.

But this restricting rule requires a standard to distinguish between arbitrary and non-arbitrary particulars: I don't think you or Kant would maintain that any distinction between persons is illegitimate when universalizing a maxim, and in any case I see no justification for such a blanket rule. (What about, to use a trivial example, "As an engineer, I may speak with authority about engineering"?)

The standard most clearly inherent in the formulation seems to be something like the universalization test itself. Part of universalizing across agents involves considering what it would be like for others to behave toward me according to the universalized maxim: you've emphasized this yourself. Similarly, we could, while universalizing across the relevant class of agents, also universalize across possible circumstances: "Can I, whoever I might be, will that this person/this limited set of people consistently have this right?"

This clearly rules out the most morally absurd and self-serving cases, like "As the great Soheran, to achieve my ends, I may treat others as I please." "Soheran" is only significant as long as I am Soheran: if I were to imagine myself as instead being one of Soheran's victims, I would not find the mere fact that my victimizer is Soheran to be significant.

But if I have some reason independent of my self-love (or similarly morally arbitrary circumstantial factors) to grant "Soheran"--or any status--particular importance, importance that I can continue to will even when I imagine myself in the place of others, I have not truly compromised universality. This sort of distinction between persons is hardly any logically different from the distinctions between actions that are necessary in any moral theory.