Kant’s Categorical Imperative is Arbitrary
Kant argues that the morality of an action should be judged on the situation alone, and in such a way that if this action was was made universal: it should be considered good or bad in the light of everyone doing it regularly. The problem with this is that the boundary of what is and is not considered an action is not well defined, and depending on how you slice reality into actions, you will get different answers of whether an action was moral or not.
I am not an expert on Kant, but let me present a number of scenarios with progressively different scopes in order to show that there is no clear demarcation between what is and is not a discrete action. Without a clear demarcation, Kant’s Categorical imperative will provide different — and sometimes contradictory — assessments depending on how you define the action.
Please read the situations in order and consider how the additional information either changes the morality or not.
Situation 1
At a house in Holland, some people come to the door, they ask a question and the owner answers.
Is that moral? Most will say that we don’t have enough information yet about the situation, but bear with me. The point of this discussion is to talk about how the situation is affected by how much, and what kinds of information, are included in the definition of the situation and the action.
Kant says we need to judge the action above according to how it would be if this was universally applied. People coming to the door, asking questions, seems like a perfectly reasonable thing, and if this was happening all the time all over the world there would be no problem.
Situation 2
At a house in Holland, some people come to the door, they ask a question and the owner says “No”
Again, this would seem perfectly moral action. Again, most would say we don’t know enough about the situation to say whether the answer of “No” is moral or not. How much more do we need to know? How do we know when we have a “complete” description of the situation? What does complete mean? Is there any way to determine that your description of the action is completed.
Situation 3
Anne Frank lives in the attic of a house in Holland. Some people come to the door and ask whether Anne Frank is there. the owner of the house says “No”
On this telling of the story, Kant would say this is immoral because it is a lie, and lies are immoral on their own basis. If this action were performed universally, this falsehood would be a bad thing. The particulars around this particular situation don’t matter. The fact is that a falsehood was told, and telling falsehoods are universally morally repugnant.
However this is not clear. It might be that any time anyone asks for Anne Frank it is universally better to deny that. It might be that there are thousands of Anne Franks, and each one would experience significant harm or death from being exposed. Can we say that this is “just a lie” or must we also include that this is lie particularly about Anne Frank? Is this aspect of the situation relevant or not, and how do we determine whether it is relevant?
Situation 4
Anne Frank lives in the attic of a house in Holland. Some people from the Institute of Health come to the door. They are working on a treatment for rhino-carcinoma in humans, a disfiguring cancer of the nose which effects thousands of people. They are checking and treating anyone they can find. They ask whether Anne Frank is there. The owner of the house says “No”
On most accounts this now seems like a completely immoral thing to do. After all, the implied action is that they might identify a problem with Anne and the sooner this is found the better it will be for everyone. If we see this action repeated thousands of times it seems like a very bad thing, and so we might conclude that this action is immoral.
Situation 5
Anne Frank is a dove living in the attic of a house in Holland. Some people from the Institute of Health come to the door. They are working on a treatment for rhino-carcinoma in humans, a disfiguring cancer of the nose which effects thousands of people. They are checking and treating anyone they can find. They ask whether Anne Frank is available for a procedure. The owner of the house says “No”
Here we seem back in the territory of moral action. After all, the carcinoma only effects people, and as such, a dove really can’t benefit from this as doves don’t have noses like people. It seems that the people are confused. Saying “yes” would seemingly be a very bad thing to do if projected to the universal level. You might easily see this action done a thousand times, and realize that it is a good thing.
Situation 6
Anne Frank is a dove living in the attic of a house in Holland. Some people from the Institute of Health come to the door. They are working on a treatment for rhino-carcinoma in humans, a disfiguring cancer of the nose which effects thousands of people. They are checking and treating anyone they can find. It turns out that the treatment requires the collection of a serum from birds, particularly doves. The collection procedure is harmless to the bird, but a single collection might help hundreds of people. They ask whether Anne Frank is available for a procedure. The owner of the house says “No”
This action seems immoral. Again, project this to thousands of instances, and if the answer to this situation was played thousands of times, it would seem overall to be a very bad thing.
Situation 7
Anne Frank is a dove living in the attic of a house in Holland. Some people from the Institute of Health come to the door. They are working on a treatment for rhino-carcinoma in humans, a disfiguring cancer of the nose which effects thousands of people. They are checking and treating anyone they can find. It turns out that the treatment requires the collection of a serum from birds, particularly doves. The collection procedure is harmless to the bird, but a single collection might help hundreds of people. They ask whether Anne Frank is available for a procedure. The owner of the house says “No, not until I find the net for catching her.”
Now, this action seems perfectly moral.
Discussion
This example is quite contrived, like many in philosophy, however it is useful to illustrate a point: what you include in the description of an “action” makes a material difference as to whether the action is described is moral or not.
How can it be that leaving details out changes a moral action to an immoral action, or not?
It is important to note that when we talk about the morality of an action, all we have to refer to is a description of the action. A description constrains the action to the extent that it does, but leaves open the opportunity for many different variants of the action. A generic description allows for many variants, but a very detailed description has far fewer variants. The variants differ in their particulars. When we give a generic description, the implication is that the particulars simply don’t matter. But that is not always true. Consider each of the situations above: at what point can we say that the particulars don’t matter?
The big question here is: how do we determine when we have included enough details to precisely describe the action that we are judging?
It can never be acceptable to say to include ALL the details. That is simply not possible to say everything about a situation, including every small possible difference. Even if it was possible, that would make the situation unique in the history of the universe and any moral judgement would be similarly unique and unrepeatable. the space of all possible situations and actions must be sliced carefully into a group that includes all the variants of the action that agree with the moral assessment, and exclude all the variants that don’t.
Philosophers argue about “slicing at the joint” meaning that there are places in reality where things are naturally divided into distinct groupings. Is there any validity to that? Are there distinct ways to say that one thing is an action, and another is a part of an action, or a combination of actions? I don’t see any evidence of this. It is merely the mind’s model of the world that leads us to identify “actions” in the space of all the motion of the universe.
I say that the categorical imperative is arbitrary, because it depends on a definition of an action that is distinct. That definition must include enough to categorize the situation as something that is seen regularly enough to be useful, but it must also exclude information that makes it too specific and therefor unrepeatable. What you include and don’t include in the definition of the action is instrumental to whether the action can be judged moral or not. In my experience there are always exceptions, making the morality less than categorical.
Circularity
In fact it is worse. Often the decision of how to cut the action out of reality depends on an implicit moral judgement in the first place. The universe has motion without distinction, but it is our own model of the world, or perception of the world, that breaks this motion into actions. There is a certain circularity to this reasoning because the purpose that the mind does this breaking and categorizing of actions is in order to tell which actions are good and which are bad. We single out murder as a distinct action BECAUSE it is a pattern that we believe is universally bad. We identify tigers hiding in the underbrush because we understand they are dangerous. Leaves, which come in hundreds of shapes are just “leaves” because it really does not matter (except for those leaves that are dangerous and then we learn to recognize them).
It is not clear whether the action is defined and then judged as good or bad, or whether the goodness or badness defines the action.
Kant talks about a “lie” however we all know there are “white lies” which are certainly not immoral. What about telling the truth but leaving out critical details? Does it matter what is said, or must you include only what is heard? What if you state the lie in a language that the other person does not know? Can one judge the morality of saying the individual words of the lie, each of which might be perfectly moral to say in that situation, but when uttered together have a different meaning?
It is no good complaining that the description of the action is not the “real situation” because again the real situation is selected to fit the moral being discussed. the exercise is left to the performer to tell whether they have performed one action or another.
This is ultimately the problem with saying that actions themselves can have a moral valence. The universe blends smoothly from one end to the other, and from the beginning to the end. There are no clear discontinuities that break one action apart from all other actions. The action of breaking a window is considered a bad action. But breaking a window to rescue a person from a burning building is a good action. There is no clear guideline to say that the second is not a distinct action. If you say the description must be complete, then where do you stop? Is it necessary to mention that the person being rescued is a sniper in an active shooter situation? Is it necessary to mention the country that they are in, or the sitting leader of that country? Or that it is raining?
Ultimately, what you include in the description is arbitrary, which means there can be no basis for associating a moral valence to a particular “action”.
What Can Be Done?
Consequentialism says instead that you simply look forward an assess the outcome. You consider all aspects of a situation and action, no matter how big or how small. You can exclude things that make no difference to what you are considering, such as the price of eggs in China. Naturally, each situation might be completely unique. That is not really a problem in consequentialism, except of course application of this to decide whether to do something depends on being able to predict the effects of all the things that are going on — particularly any of the things that might be changed by the decision.
This approach requires consideration of the entire flow of the universe, which is impossible, but don’t dispair! There is hope. We can employ short cuts — common patters that almost always work out to be good or bad, and once we see that patters, we can use that as an approximation to the real situation.
Is it immoral to plunge this knife into the chest of a person? That usually leads to death, and actions that lead to death usually are worse for general well being. So we can use that as a short cut: stabbing a person is bad. Except for when you are a surgeon performing heart surgery. Then it is good. Except when the surgery is not needed and done for false purposes, and then it is bad again.
It is important to note that there is no moral valence to the stabbing action itself. Sometimes it is good, sometimes it is bad. You simply can not make a categorical judgement about stabbing a person in the chest with a knife. But you CAN use this as a convenient shortcut that works in most situations. Kant seems to have missed this when working out his moral framework.
Conclusion
Evaluating an action as good or bad is impossible, because it is impossible to precisely define an action such that it clearly includes all the cases that act the same. The particulars matter. The only way to determine whether an action is good or bad is to consider the entire present situation, and guess as best you can what the outcome will be. This is not easy. It is not always possible. If you can’t predict the outcome, then there is no way to know whether a particular action is good or bad. We want to know, but we simply can not know. Or at least, we can only to the extent that we can predict the outcome.