I think you are confusing different issues.

The moral law is situational in that it applies to all agents. So, if one is not an agent - that is, if one does not have a faculty of reason - then the moral law does not apply to you.

If you do possess a faculty of reason, then the moral law does apply to you.

All can agree to that, including utilitarians and particularists (particularists being those who think morality is very context sensitive to the point of rendering all attempts to distil a principle or principles misguided).

Kant also observed that the moral law does not depend for its rational authority upon one having any particular desire or end. So whereas imperatives of prudence make mention of our ends - if you are thirsty, then get yourself a drink - moral imperatives do not. For they are addressed to us all as agents, rather than to us as individuals with our own particular projects.

But even this is something that utilitarians and particularists can agree on. For these points are more metaethical than normative.

It is just that Kant thought we can from these observations derive substantial insight into the content of the moral law. That is, he thought that a rational imperative that applies to us all as agents would have to have a particular content. And it was from that thought that he arrived at his conclusion that lying is always wrong.

Your analogy with the colour of an object confuses the above issues with another one, namely the objectivity or subjectivity of morality. That issue applies to colour too. That is, whether colour is objective or subjective is a matter of debate. And for this very reason it is not useful in casting light on the moral issue. For those who think morality is subjective will likely also be persuaded that colour is, and vice versa.

Kant himself was horribly confused on this, because he held that moral imperatives are imperatives of reason - indeed that 'the'moral imperative is the supreme imperative of Reason - yet was never clear about who or what Reason was. He sometimes seems to confuse Reason with our faculty of reason, and sometimes seems to identify ourselves as Reason, and sometimes seems to think that if he can show that all faculties of reason will tell their bearers the same thing, then this serves somehow to show that what it tells them is objective (which is patently false).

But anyway, you will not gain insight into the nature of morality by thinking about colour, as whether colour is objective or subjective is, if anything, more debatable than it is in the moral case.

I say both perspectives have merit.

1. The sphere can be considered as white (as having the property of whiteness) in a stable, predictable manner. In most situations where there is enough light to spot colors, the ball will appear white. It's only in the rare cases of a colored light (or no light of course) that the ball will not appear white. But

2. There are indeed cases where the ball does not appear white, rare or trivial as they may be.

Likewise, most of the times it is not a moral course of action to kill a man. But if said man has only a few hours of a horrible agony to live, and if he asks you to shorten his suffering, it could appear as the moral thing to do.

Kant himself was horribly confused on this, because he held that moral imperatives are imperatives of reason - indeed that 'the'moral imperative is the supreme imperative of Reason - yet was never clear about who or what Reason was. — Bartricks

In the case of the murderer a the door he was, and this has been referenced on TPF numerous times. Best you look it up and satisfy yourself.

Kant himself was horribly confused on this — Bartricks

'Course he was, and the world has waited about 260 years for Bartricks to straighten him and us and all of it out. Thank you Bartricks.

In the case of the murderer a the door he was, and this has been referenced on TPF numerous times. Best you look it up and satisfy yourself. — tim wood

Once more: Kant is unclear on who or what Reason is. As you'd know if you'd read him.

'Course he was, and the world has waited about 260 years for Bartricks to straighten him and us and all of it out. Thank you Bartricks. — tim wood

You'd have adopted the same sarcastic tone towards Kant himself if you'd lived 260 years ago. Oo thank you Mr Kant for coming in and straightening morality out for us.

moral law does not depend for its rational authority upon one having any particular desire or end. — Bartricks

But even this is something that utilitarians and particularists can agree on. — Bartricks

Utilitarians agree that moral laws do not depend on our desires or the consequences of our actions? That’s a new one…

Yes, you don't have an obligation to maximize happiness because you happen to want to, but irrespective of whether you do.

But aside from that basis, they could never agree that moral laws do not depend on consequences or desires. You have an obligation to do X or Y precisely because of the consequences of X and Y.

We'll need to explore the matter of Kantian morality very carefully. What's relevant to the OP is how Kant's position on lying being immoral is non-negotiable. In layman's terms, lying is always bad. Period! The matter of why mendacity is bad seems to be a different issue from why being untrutfhul is always bad. In utilitarianism, for example, a case can be made that lying is bad but no utilitarian has been able to argue that lying is always unethical. In fact, lying isn't always bad for a utilitarian.

Let's look at how Kant deals with telling falsehoods. The categorical imperative states that if an action L can't be adopted universally i.e. made into a law applicable to all and at all times, L can't be good and, conversely, if L is amenable to such a treatment, L has to be good. I'm employing lying in my analysis but I'm certain that we can generalize the argument to all morally-apt actions.

In my analogy, Kant's categorical imperative would correspond to, should we consider the white object black in all cases (in the light and also in the dark)? No, of course not! However, there seems to be no issue with treating the object as white even in the dark. Can we make lying into a law? No we can't. Can we make being truthful a law? Yes, of course!

What does this mean at all?

If you ask me, there's a moral asymmetry - good actions retain their goodness so to speak even in situations that might not be appropriate for them. Bad actions, despite being good on occasion still feel wrong. For instance, even if the truth hurts, it doesn't lose its goodness but a lie, even if it saves the world, pricks the conscience. This moral asymmetry is, I'm afraid, the intuition that we've all had at some point in our lives and Kant's categorical imperative puts it on a firm rational foundation.

Not sure your analogy changes anything or clarifies anything. The general point he makes is that in a society where everyone follows Kantian ethics, we would not have any murders, because everyone follows this moral thinking. It's too absolute. The same goes for utilitarianism, it's too absolute to work in practice.

Morality is a tier list. You can take a moral act and divide it into fractions of more or less acceptable. A brutal sledgehammer murder of an infant can with some certainty be positioned as an absolute immoral killing. But kill someone who is just about to do it is not. This tier of different variations is easy to sense within a human's capability of empathy. We can absolutely figure out if something is considered justified or not justified and in turn where on the moral tier of killing it ends up. It also guides us to sense what is right or wrong.

The domino effect of this also informs how to process attempts at influencing someone to kill. We are then able to spot the justifications and methods used by someone to influence someone else to kill.

The problem in ethics rarely has to do with figuring out what is a moral or immoral act. More often than not it mostly attempts at inventing thought experiments that put moral thinking into paradoxes through thought experiments. But such paradoxes are not possible to apply to actions, and judging someone's moral act by how paradoxical the event was is not how we process ethics in real life.

The traditional Kantian thought experiment of the murderer asking you if your friend is in your house so he can murder him and the act of lying can make the murderer by chance find your friend is filled with so many variables to the act of lying and murder that it is impossible to apply to any real scenario.

What makes something immoral is if we act with destructive biases, extreme selfishness, lack of empathy, randomness in choice etc. It's how we process morality that defines if something is immoral or moral. How we process each situation before a choice. We cannot be judged for the consequences of a choice since it requires knowing the future, we can only act upon what we know.

This is why I am fond of the idea of epistemic responsibility. To rate morality based on how much thought has gone into a choice. A person who evaluates, to the limit of his ability, what the consequences of a choice are going to be. If epistemic responsibility has been taken and an unbiased, empathic choice with low selfishness has been taken, and the act is justified to cause as little harm as possible. That's the only act a human can accomplish that can be considered highly moral. All attempts at simplifying a moral choice down to either Kantian absolutes or utilitarian math fail because of this need to simplify ethics into an ideal.

What you're missing, that Kant does not miss, nor you either if you read him, is that Kant thinks it through. Especially in his "Murderer at the Door," (MD) referenced on TPF several times and easily findable online.

The categorical imperative (CI) against lying is easy enough to ounderstand. The question in MD is whether the circumstance in question outweighs the CI, perhaps justifying lying. Kant is pretty careful. If you tell the truth, you have acted in accordance with the CI and you have caused no harm. Which is not the same thing as saying that no harm may come. Come it may, but you are not an agent of the harm You have caused no harm becasue you do not know what the circumstance ot the moment is. If on the other hand, you lie, then you may become an agent of the harm. I invite you to read the original - it's not too long - for the details.

A different question is whether you may lie to the MD if you can know to a certainly that your lie will result in good. For this, Kant makes clear (elsewhere) that determining CIs is a bit of an art; and which is applicable when CIs conflict he resolves by saying that of those pertaining, one rules, and the one ruling, the others fall away.

Moral of the story:, lots of people criticize Kant when in fact they haven't even come close to understanding him, Don't be that person. Especially don't be the person who is wrong, doesn't know it, and insists he's right because, as one recently observed, "Kant himself was horribly confused."

The categorical imperative (CI) against lying is easy enough to ounderstand. — tim wood

:ok:

The categorical imperative (CI) against lying is easy enough to ounderstand. The question in MD is whether the circumstance in question outweighs the CI, perhaps justifying lying. — tim wood

:ok: I believe this is referred to as conflict of duties?!

Moral of the story:, lots of people criticize Kant when in fact they haven't even come close to understanding him, Don't be that person. Especially don't be the person who is wrong, doesn't know it, and insists he's right because, as one recently observed, "Kant himself was horribly confused." — tim wood

:lol: We're all confused in one way or another, right?

You've raised an important issue - old news, yes but, supposedly unresolved - which is the moral dilemma presented by the murderer at the door scenario (MD) to Kantian ethics.

Allow me to elaborate a little on the MD to test my own understanding and also to ensure we're on the same page. The MD is basically a situation a person who subscribes to Kantian ethics can encounter and one in which there's conflict of duties. Either this person lies to the would-be murderer or not. If fae lies then fae fails in faer duty to tell the truth and if fae tells the truth, fae fails faer duty to save a life. It's lose-lose for this person.

How do we approach this issue rationally?

Kant's formulation of the CI (Categorical Imperative) states that you are to "act only in accordance with that maxim through which you can at the same time will that it become a universal law — Stanford Encyclopedia Of Philosophy

Focus your attention on "...universal law...".

Every moral theory, Kantian or otherwise, aspires to two achiecve two aims:

1. Generate laws that are binding codes of conduct. Dos (moral) and don'ts (immoral) list.

2. Universal application: Applies to everyone, everywhere, everytime. The best-case scenario: no exceptions to moral laws, not even one!

To not make this post longer than necessary, let's discuss Kantian deontology in the context of the above two points (1. laws & 2. universal).

Kant's moral theory among other possible specific moral commandments contains the following two laws:

Law 1. Don't lie
Law 2. Save lives/don't participate in murder

The MD would have us think that there's something wrong with Kant's ethics but that's incorrect as I'll attempt to demonstrate in the following paragraphs.

First off, take note of a simple but very important fact. If a moral theory X entails laws L1, L2, L3,... (derived from the moral formula of X) then, it has to be that, if X is adopted, all laws L1, L2, L3,...are in effect simultaneously and universally. Bear this in mind.

Kant's moral theory (moral formula = CI) entails the following two laws among others of course:

Law 1. Don't lie
Law 2. Don't participate in murder

As I mentioned above, both laws must be "...in effect simultaneously and universally." This is essential to a moral theory. Lawws, being laws, must apply together, to all, everywhere, at all times - this is Kant's crucial insight into the nature of morality, it's all about "...universal laws..."

Notice now what happens or rather doesn't happen when both law 1. Don't lie and law 2. Don't participate in murder are being followed by a group, society. There will be no murderers and if there are no murderers the MD is an impossible scenario. People won't ever be in a situation in which they'd have to lie to a murderer because in the event Kant's moral theory is itself, in terms of its moral laws, universal, murderers won't exist, there'll never be a conflict of duties. The MD is a pesudoproblem - its possibility requires that only fragments/parts of Kant's moral theories are followed at any one time but then that contradicts the very essence ("...universal law...") of not only Kant's ethics but all other ethical theories.

Why then does the MD seem so plausibly problematic to Kant's ethics?, you might ask. I suppose it's because we feel Kant's moral theory should also work in a world where immoral people exist, another way of saying only few but not all of Kant's moral laws are adhered to by a society. However thinking this way contradicts the very idea of a moral theory as a "...universa law..." All or none! Tertium non datur (a third is not given).

Good.

Interesting. But if you're suggesting that there is no duty to evil people, that would be a mistake.

It's best to read Kant's words. From yours, above, "

act only in accordance with that maxim through which you can at the same time will that it become a universal law — Stanford Encyclopedia Of Philosophy

What is a maxim? 1. A general truth, fundamental principle, or rule of conduct. 2: a proverbial saying.

Not, that is, any kind of law at all. Where does law, or universal law, enter in? In your assessment that that which you would have govern your own conduct at that moment should be the rule for all, even as a universal law. And the test is whether your choice of conduct if adopted by all would destroy the maxim itself. If, for example, you willed that lying be a universal law, then whence truth? And closer to home you could not complain if someone lied to you.

So one part of the trick is to not suppose universal law already exists. (Although of course it does in various forms.) Rather the circumstance arises out of the, "What shall I do?" And off to the races, at some point asking, "What ought I to do." And Kant tells us how to go about figuring that out.

And it cannot be rocket science. Kant observed that an ethics had got to be readily grasped by all. Even if the mechanics of it as laid out for the academics and professional philosophers of his day seemed complicated.

It's too absolute. The same goes for utilitarianism, it's too absolute to work in practice. — Christoffer

I guess so. The impression that I get is all moral theories till date has the fatal flaw of being unable to cover all the bases i.e. there are multiple exceptions or, as some might say, many special cases in which the formula (categorical imperative, maximum happiness for the maximum number of people) fails to output a recommendation for an appropriate course of action, an action that sits well with our overall sense of right and wrong.

Nevertheless, Kant's theory is unique in one particularly significant sense - it makes a very crucial attribute common to all moral theories its foundation, that being their aspirations to a status of universal law explicit. Other moral theories seem to have it as a, how shall I put it?, implicit axiom. Perhaps, it's just so obvious that to state it would invite ridicule or scorn for it would be seen as superfluous. I'm not a 100% sure.

Where was I? Oh, as I was saying, all moral theories are created with the express purpose to become a universal law. The fact that they're considered inadequate/deficient when exceptions/special cases arise is proof of that. Kant seems to have grasped the full significance of this simple truth and I, suppose realized that the secret to a sound moral theory hinges on the essence of a universal law. That essence, that secret, is in the simplest of terms that immorality is existentially predicated on it being an exception. Immoral actions can't be, as per Kant, universalized for that immediately make the immoral action in question inconceivable, another name for contradiction. As you can see, I have a general idea of what Kant's moral theory is all about but, unfortunately, I'm still hazy on the details.

The long and short of it is Kant's deontological ethics zeroes in on the heart of the matter - we want to get our hands on a moral theory that is a universal law and this requirement is, Kant discovered, the crux of morality. There seems to be this quest for a perfect moral theory, one that has now occupied great minds for almost two millennia, and the received opinion on what it'll look like is that it should be able to handle exceptions/special cases as well as it manages to tackle the ordinary/usual problems. Can you spot any difference between this currently only hypothetical perfect moral theory and Kant's moral theory? There are none! To handle all exceptions is equivalent to having no exceptions. Put simply, the ultimate goal of moral theorists is to develop an absolute moral theory!

act only in accordance with that maxim through which you can at the same time will that it become a universal law — Stanford Encyclopedia Of Philosophy

A maxim: I'll steal
Universalization of the above maxim: Everyone steals

The maxim I'll steal implies there's private ownership
Universalizaton of the maxim, everybody can steal implies there's no private ownership.

(My maxim is I can steal + My maxim's universalization) -> (There's private ownership & there's no private ownership)

Ergo, my maxim I can steal can't be universalized. To steal is immoral.

That's as far as I could get. The sources I referred to are not as clear on this issue as I'd hoped.

Maybe you can shed some light on the matter, you know, clear up my confusion.

A maxim: I'll steal
Universalization of the above maxim: Everyone steals

The maxim I'll steal implies there's private ownership
Universalizaton of the maxim, everybody can steal implies there's no private ownership.

(My maxim is I can steal + My maxim's universalization) -> (There's private ownership & there's no private ownership)

Ergo, my maxim I can steal can't be universalized. To steal is immoral.

That's as far as I could get. The sources I referred to are not as clear on this issue as I'd hoped. — TheMadFool

What confusion? I think you've got as much as can be got. In your example the machinery seems to have clarity - and is instructive in itself. In other examples there might be nuance, but as long as you gain that clarity, what else?

And reading the rest of your post, the idea of a universal law, or a set of them, I do not think Kant had in mind. At least not so much as a what, but instead a way/method/understanding.

It's as if folks wanted the numeric solutions to all the problems, supposing there to be such, and Kant is teaching algebra. Sense? And of course there are such solutions, called numbers, but in just that sense not solutions at all. And no doubt - to a certainty - Kant himself had definite ideas about what folks should and should not be doing.

:ok: :up:

To handle all exceptions is equivalent to having no exceptions. Put simply, the ultimate goal of moral theorists is to develop an absolute moral theory! — TheMadFool

But an absolute moral theory is impossible with extreme variables. It's a desperate attempt at trivializing psychology, behavior, culture. The only true moral theory is one that needs to be a legion of interconnecting ideas. A web of actions and reactions that without breaking the line of rational thought, balance each other into harmony.

You can universalize that murder is wrong, but you can also universalize that killing someone to help another is morally justified. If everyone killed someone to save another, that is a complex grey area that isn't as easy to prove to be impossible to universalize. The same goes for stealing food to help a child. If everyone stole food to help children, isn't that also pretty universalizable? We could speculate that this would make other people go hungry, but then steal more food to help those children and such a universalized law would be systemic without hurting any more or less than a system where the child dies of starvation because you cannot steal food for them.

Absolute moral theories can only be applied if the moral act is simplistically basic, and no moral acts are in reality simple. Therefore we have all the different analogies and examples proving against most fundamental moral theories.

While I think there are merits to most fundamental moral theories I think that the key to something universal is to create an interconnected web that creates a synthesis of them or parts of them. A neural network of variables that work in tandem rather than trying to be absolute in simplicity.

I'm generalizing, but the reason we never find out anything universal is that morality is a human invention that is changing through culture, time, and psychology. We can only find true similarities between different moral ideals by looking at the things transcending culture and opinion. Empathy, pain, suffering, joy, happiness etc. We can universalize these emotions and feelings, the core nature of a human. But we cannot universalize concepts based on inventions and ideas that we create. Those shifts too rapidly over the course of history. If a world breaks down and society caves in on itself and there are no agreed-upon laws and rules anymore, then you could easily universalize stealing as that is the only form of feeding yourself and your loved ones and if everyone does it in this context it can be an agreed law of morality that is as justified as anything else. If something is going to be universalized, it needs to work in any form of existence for the human race, it cannot be universalized only for the established status quo of society. Otherwise, you would need to establish that you can only universalize a moral concept based on a specific society it is built upon.

A moral theory that survives all different variations of humanity's existence can only be formed upon the basics of the human condition. That is the only universal law for human morality. Examine what it is to be human, then form morality out of that. A web that incorporates all parts of the human condition, psychological, emotional, physical, combined with our ability to rationalize.

I'm not talking about Sam Harris nonsense where we use neurological data to pinpoint a "moral landscape", but instead a way to universalize moral around the only concepts that are actually universal throughout history and culture. It's also the only way I can see past the is-ought problem. But it also requires having an active rational mind rather than fall back on a spreadsheet of moral laws. Because it's not a one-sentence formula, it is a web, a multidimensional interconnected "algorithm" that shapes according to the situation based on fundamental human cornerstones. I think that most things in ethics philosophy are fundamentally broken as it is built upon an idea that the human concepts is in itself universal. The only universal thing about humanity is the fundamentals of our condition, nothing more or less. Anything further becomes an invention, anything less becomes indifference.

you can also universalize that killing someone to help another is morally justified — Christoffer

That's exactly what I meant - the point, if it is one, of morality is to cover all the bases i.e. any good moral theory must eventually pass the Kantian categorical imperative (CI) test. If a moral theory is only partially implemented - as is the current status quo - we'll never get anything done so to speak. Free will appears to be a crucial factor in that if we accept it's real, immoral people will have to be included in the moral equation and this is the stuff that exceptions/special cases are made of.

you could easily universalize stealing — Christoffer

It's not that immoral actions can't be universalized. They can but you would be guilty of a crime against logic, contradiction.

But it also requires having an active rational mind rather than fall back on a spreadsheet of moral laws. — Christoffer

Indeed, I worry about that. A moral theory will operate just like a mathematical function; you input the relevant information regarding a particular moral question and it'll output the right answer and by "the right answer" I mean you wouldn't have cause to doubt its goodness. A very formulaic approach but at least it's reliable insofar as the moral theory in question itself is.

t's not that immoral actions can't be universalized. They can but you would be guilty of a crime against logic, contradiction. — TheMadFool

But there you point out the action is immoral before the examination of whether or not it is. If it is universalized in a society where stealing is the way we feed ourselves and being viewed as a good act that helps people and that anyone can do it or protect against it. You cannot say it is immoral because it is in our society considered so. This is why both stealing or not stealing can't be universalized because that would demand the foundation of society is universal, which it isn't, it is an invention by us.

A moral theory will operate just like a mathematical function; you input the relevant information regarding a particular moral question and it'll output the right answer and by "the right answer" I mean you wouldn't have cause to doubt its goodness. — TheMadFool

This is what I mean by the algorithm. It's closer to a point system, where each point represents a calculation of emotional and physical values for people. If you are about to kill someone, the points can be derived from different aspects of that situation. The only universal laws to any of this is human suffering and pleasure, basic natural concepts of ourselves as biological beings. So you are about to kill someone; will this hurt them? The person does not want to die, and the person is in a rational position to judge that will = yes, you will hurt them by this = -1 points. Is this person about to hurt someone else? no, they are in no position to do so willingly = you will not prevent a bad act by this person as he is not willingly about to do any bad acts = -1 points. Are you gaining something by killing him? Yes, you will gain resources that will help you survive = +1 point. Will killing him prevent you from gaining resources in the future by instead collaborating with him? Yes, resources are limited to what you gain from the killing = -1 points.
Total = -2 points on the negative, meaning the act is immoral.

Now, this is a simple example, but the point is that the epistemically responsible way to calculate an act is to look at each possible consequence in accordance with our basic human conditions. If we start at 0 as a morally grey balanced position. The number of consequences included in your calculation will increase the quality of the outcome/conclusion. The more you rationalize around the act, the more accurate you can predict its moral nature. The more on the plus side you get, the more morally good the act is, the more on the negative, the more morally bad it gets. But nothing can be deductively good or bad by itself.

Such calculation incorporates the grey areas of morality and the only constants used are fundamentals of the human condition. If the outcome is close to 0, then we can only conclude that it might be good or bad, that it is an unknown moral value until more data is input. We can also input a bonus value for doing the calculation. If the moral act is calculated to the best of the ability of the agent of the act, then that is worth a point in itself. Since that agent takes the epistemic responsible choice of trying to calculate the morality of an act.

- Human basic conditions are the constants
- Individual consequences are the variables.
- Calculate the variables according to the constants to find out their positive or negative value.
- The more variables there are, the more accurate the final moral score is.
- The act of calculating is itself a positive value and more valuable the more accurate the calculation
- Getting an unknown at 0 is also a valid outcome that incorporates the unknown conclusion as a conclusion in itself and at that point the choice is the same as a coinflip without valid positive or negative judgment of the agent.

What moral choice to take is therefore always consolidated down to "do the calculation". You cannot act according to anything that points out an ought. What is, is the constants and collective of consequences, what oughts are the conclusion based on the variables calculated out of the constants.

So the way to act morally good is to act according to a positive stance in the algorithm and by always calculating acts with this algorithm. But nothing of this becomes a universal law, if not the algorithm itself being closest to be one for the human condition.

But there you point out the action is immoral before the examination of whether or not it is. If it is universalized in a society where stealing is the way we feed ourselves and being viewed as a good act that helps people and that anyone can do it or protect against it. You cannot say it is immoral because it is in our society considered so. This is why both stealing or not stealing can't be universalized because that would demand the foundation of society is universal, which it isn't, it is an invention by us. — Christoffer

If one considers adopting the personal maxim, "I shall steal", this maxim only provides benefits in a world in which stealing is a no-no! Now, if you want everyone to steal, then stealing is permissible. You then have the contradiction: Stealing is not ok (your maxim would be pointless without it) AND stealing is ok (stealing is universalized).

If one considers adopting the personal maxim, "I shall steal", this maxim only provides benefits in a world in which stealing is a no-no! — TheMadFool

How so? Stealing in a world where everyone steals and the foundation is built upon it and being good at protecting against it; could be the societal standard form of living in that type of society. We cannot judge it without putting into context what we already invented as moral law before judging it universalizable or not. So you already have judged it as morally negative or positive before thinking about it as being able to be universalized or not. You, therefore, need to have decided if it is morally good or not before applying any theory around it. You don't want everyone to steal, why? Why do you think that way? It's society that taught you this is bad so you think it is bad that everyone steals, but if that is part of the fundamental societal structure you wouldn't think about it as universally wrong and therefore it falls flat.

If something is supposed to be absolute, it cannot have flaws as soon as a society fundamentally change their invented values around different moral acts. What if killing is universally accepted in a society of warriors where killing or being killed is not considered negative? Then anyone can kill anyone at any time or defend against it and only the strongest survive. If that's the foundation of that society, their moral universal law is that killing is good. So then, is killing supposed to be good in the society we live in today? No, because our invented reality about "how to live" and what is important is fundamentally different, so for us, we don't see a universally positive moral choice of everyone being allowed to kill to be valid.

Kant's theories fall flat when society is fundamentally changed. We cannot have two universal maxims guiding us and we can't have two fundamentally different societies that come to different conclusions about what is universally accepted as good or bad moral acts. Then the entire idea of universalization falls flat.

If the theory requires pre-determined moral values, it isn't possible to act as a moral theory.

How so? — Christoffer

Indeed, it's true that Kant's categorical imperative is content independent. It's, let's just call it, a formula more general than, that's a big clue, than the utilitarian maxim, maximum happiness for the maximum number of people. Endorsing utilitarianism immediately commits you to happiness as the mainstay of morality.

Kant's categorical imperative (CI), on the other, hand prevents anyone from smuggling in any preconceptions about morality - whether it should be happiness-based or something else entirely is deliberately and wisely left out of the equation. Thus, even if the standards of society were to undergo radical shifts, even if these changes affect morality, they won't do anything to alter the fact that moral theories are essentially gunning for the status of a universal law. Given this, the CI being a simple test of whether or not a particular maxim can be universalized, it follows that the CI is like some sort of master key to morality. I'd bet even aliens, with completely novel ways of living, out-of-this-world (literally and metaphoricall) values, etc. will agree Kant's CI encapsulates in one single sentence the very heart of God's (perfect moral being) goodness.

Act only in accordance with that maxim through which you can at the same time will that it become a universal law — Stanford Encyclopedia Of Philosophy