There are certain formal logical relationships which we know hold true independently of our knowledge of particular facts about the world. For example, we do not need to know any of the particular facts of the world in order to know the law of non-contradiction is necessary. The law of non-contradiction is the a priori recognizable truth that that for any A, A =/= ~A (meaning that for any A, A cannot be ~A in the same respect in which it is A).
When we assess particular claims and arguments for claims about what the world is like, we apply this and the rest of the rules of logic, all of which are known formally and independently of particular experiences. In this way logic is neutral with respect to particular belief systems and paradigms. You and I may share a number of different specific views about what the world is like but the rules of logic are neutral in our disputes and can help us arbitrate between our conflicting viewpoints.
Of course there may be many times at which we both make plausible, logically consistent inferences and in those cases principles of formal deductive logic cannot themselves settle our dispute. But whenever a logical contradiction is involved, formal logic can be the most unbiased arbiter we have.
R.M. Hare wants to argue that there are logical principles of morality which can serve as the same sort of netural, rationally arbitrating tool for settling some moral disagreements as logic is for settling claims about the world generally. His chief example of such a logical principle of morality is that “There is something wrong in doing what we ought not to do.” He argues that this is formally, logically, a priori true in the sense that it does not presuppose anything particular about what we ought or ought not to do but it refers to the more basic intuition that whatever we ought not to do, it is wrong to do it.
Regardless of whether we wind up judging that parents ought to send their daughters to school or not send them to school, ought to torture people or ought not to torture them, ought to give to charity or ought not, the formal entailment of the word ought is that it will be wrong not to do whatever we ought to do and it will be right to do whatever we ought to do.
And as a result of this formal principle the statement, “There is nothing wrong in doing what we ought not to do” is necessarily false and capable of rejection on logical grounds. This means that while you and I on many occasions may have a disagreement about what we ought to do, nonetheless whenever we agree on what ought to be done, you are being logically contradictory if you insist that it is not wrong for you to do as you think you ought not.
John Searle makes a similar argument that anyone who makes a promise formally obligates herself to keep her promise. It is a contradiction for someone to say, “I promised to do x but that did not obligate me to do x.” The act of promising is the act of obligating oneself. If you make a promise it is now a fact that you have obligated yourself and you are obligated.
Of course in most cases there is no force of nature that will force you to keep your promise. But if you do not keep your promise you cannot say that you were not obligated and you cannot say that you did not contradict yourself or that you did not fail to meet an objective obligation. You did obligate yourself, even if you failed to fulfill the responsibility you undertook when you bound yourself to that obligation with the speech act of promising. And so you are committing a practical contradiction by breaking your promise and you are objectively failing to meet an actual obligation.
Returning to our discussion of Hare, he goes on to make explicit two logical features of the word “ought”. The first feature is what he calls the prescriptivity principle and the other is what he calls the universalizability principle. The prescriptivity of an ought statement is the feature of it that indicates to us that when we determine that an action is what we ought to do, that we shall do it, that our actions shall be determined by this ought judgment.
The universalizability principle is one that specifies that not just I but anyone in my exact matrix of circumstances should be doing exactly as I do. He explains (on pgs. 480-481 of Ethical Theory: classical and contemporary readings, edited by Louis Pojman. Wadsworth: California, 2007).)
Universalizability means that, by saying “I ought,” he commits himself to agreeing that anybody ought who is in just those circumstances. If I say “I ought, but there is someone else in exactly the same circumstances, doing it to someone who is just like the person I should be doing it to, but he ought not to do it,” then logical eyebrows will be raised; it is logically inconsistent to say, of two exactly similar people in exactly similar situations, that the first ought to do something and the second ought not. I must explain that the similarity of the situations extends to the personal characteristics, and in particular to the likes and dislikes, of the people in them. If, for example, the person I was flogging actually liked being flogged (some people do) that would mean the situation was not exactly similar to the normal case, and the difference might be relevant.
So, putting together these two features of prescriptivity and universalizability, we see that if I say, “I ought to do it to him,” I commit myself to say, not just that I should do it to him (and accordingly doing it), but that he should do it to me were our roles precisely reversed. That is, as we shall see, how moral argument gets its grip.
Hare’s point about the two formally identical situations requiring formally identical responses is a crucial one. Often when people want to claim that there are no binding moral rules, they will offer a counter-example scenario that shows how a particular principle does not always apply, which they take to show that it is not a principle at all. Their assumption is that if they can vary the circumstances and reveal that we would judge the new situation differently, that the principle is not really binding. And those trying to undermine all claims to moral objectivity whatsoever try to draw the inference from the limitation of a principle in applying to some specific case that moral judgments in general are impossible to formulate with any universally binding character.
Hare’s argument though stresses how situationalism which takes into account the particular variations of distinct scenarios is not incompatible with moral universalizability. If I talk about the importance of principle A and you counter with example situation y, wherein applying principle A would be clearly unjust and instead following principle B would be clearly preferable, you do not refute the entire value of principle A or its general universalizability.
Rather all you have done is give a specific situation in which the universalizable thing to do is not follow principle A but instead to follow principle B. As long as in most cases the action most worthy of universalizing is the one prescribed by principle A, it is still valuable to promulgate principle A as one’s primary rule for the general type of circumstances to which it applies. Pointing out a case in which it universally (for anyone in the same exact matrix of circumstances) does not apply does not undermine its validity in those matrices of circumstances in which it does apply to everyone universally.
The fact that exception situation y exists does not mean that we cannot universalize principle A, it means that we should universalize principle A for most scenarios but in situation y, and others relevantly like it, principle B is the rule we should universalize and expect all to follow instead.
The importance of this is that when we say that ethics “depends on situations” we do not mean that there are no abilities to judge universally correct actions within situations. Every situation may be radically unique and express a distinct and original matrix of factors, anyone of which may make it a case ethically distinct in important ways from most similar cases. That does not mean though that there is no objectivity to our choices. As long as we can universalize our action for anyone in just our combination of circumstances then our action is still universalizable, even though it is informed and justified by considerations of various highly specific situational factors.
If I think about someone else in a situation with all the same formal characteristics as my situation and I determine what she ought to do given all the relevant factors, then I have guidance about what I ought to do given all those same relevant factors. Attention to contextual specifics requires that I not just assume that for any two cases of general similarity that all the morally relevant features are necessarily identical. In every situation I must investigate whether there are morally relevant formal differences that make a principle I would normally assume was primary actually secondary, or vice versa.
But wherever all the morally relevant factors are formally similar it is logically inconsistent to say that person 1 should act contrary to the way person 2 should act. And when similar cases are not identical in morally relevant factors and we do say that person 1 should act differently than person 2, this is not relativism or an instance of non-universalizability but a case of rationally assessing the relevantly different formal characteristics of the situation and judging differently according to the reasons which those differences give us.
It is because of these considerations that I am entirely wary of anit-universalist forms of situationalist ethics. I agree wholeheartedly that we must assess particular situations extremely carefully, scrutinizing them to make sure that we highlight and take into account all the novel peculiarities that might distinguish them from other highly similar cases that are nonetheless importantly different.
When we talk, for example, about “not judging” people or their actions, usually it is not because there is no right and fair judgment to make about them or their actions. The issue, rather, is that we are usually at a severe epistemic disadvantage in assessing all the relevant factors of a situation in order that we may take into account all of the morally germane aspects of the situation.
Only with relatively comprehensive understanding of people’s psychologies, their perceptions and misperceptions, their habits, their social conditioning, the extent of their knowledge and the extent to which they possibly could have known more, etc. can we properly assess the character of their will when they perform a given action. Only were we to know extensively their intentions, the consequences they expected, their emotions and the degree to which these were in their control, etc. can we judge their character and its praiseworthiness or blameworthiness.
In many cases, we may not need exhaustive information to make effective moral judgments. But to the greater extent that our disapproval of them has consequences for their well being, the burden raises for us to adequately inform ourselves about all the variables that they were weighing and all their reasons for weighing them as they did, so that we may adequately put ourselves in their shoes in our imagination and decide what agents in their exact shoes should do as the universalizable rule (or might excusably be misled into doing wrong).
Finally, let me quickly address how this universalizability test works in Hare’s paradigmatic example of the decision to flog someone and thereby inflict great suffering on them.
Hare argues that knowing what it is to suffer is to know that suffering, by definition, entails the intense desire to have an irritation removed. If I am suffering from being flogged, this suffering is a strong desire to not be flogged. If I am suffering from a painful stomach ailment, the suffering is the strong desire not to have this ailment. When I am suffering I have a clear overriding ought in my mind—I ought to do whatever I can to end this suffering. I can universalize this awareness of the ought which suffering gives me to say that whenever anyone suffers they too will that it ought to end, even at the expense of other goods which are lesser.
So, the ought, or “the rule”, requiring suffering to end that comes with the presence of suffering is universalizable, it is what everyone should do when they are suffering.
Now in normal cases, when all things are equal, I ought to do what brings me pleasure. But in cases in which I flog someone and it causes them to suffer (as opposed to strange hypothetical situations in which they welcome the experience rather than desire it should end), I have to weigh whether this is universalizable against the sufferer’s universalizable wish for suffering to end.
In this case clearly a choice between competing oughts must be made—do I universalize the action of inflicting suffering for pleasure, which is the choice preferred by sadists, or do I universalize the action of relieving suffering from pain, which is the choice non-masochistic (and, so, genuine) sufferers prefer?
In order to decide I must in my mind imaginatively vary the scenario and ask, “Just how much pleasure does the sadist derive from inflicting suffering?” and then ask, “Just how much anguish does the sufferer derive from incurring suffering?” And in comparing the two experiences Hare thinks I will realize that the suffering creates a far more intense impression and creates a far stronger imperative to make suffering stop than pleasure in hurting creates pleasure and an imperative to hurt.
Since the imperative to stop suffering is stronger for the sufferer than the imperative to impose suffering is for the sadist, the universalizable action is to go with the stronger imperative and stop causing suffering. I should stop (or never initiate) flogging. To do otherwise would be to act in a way that was not formally consistent, not one I could recommend for everyone in formally similar circumstances. And it is irrational and immoral to act inconsistently in such a manner.