Chapter 15: J.L. Mackie’s Error Theory
I. A Seminar on Moral Realism
On the first day of a seminar on Moral Realism, Dr. Patricia Greenspan turned on an overhead projector and began writing down a list of categories of moral theories. Within each category, she wrote the name of a contemporary moral philosopher who represented that view.
Near the middle of her list, she identified a branch of metaethics called ‘cognitivism’. Cognitivists say that moral statements have a truth value — they are capable of being true or false and we can know (at least in many cases) what this truth value happens to be.
Under ‘cognitivism’ she drew a branch called ‘error theory’. An error theorist is a cognitivist — he or she believes that all moral statements have a truth value. However, the error theorist believes that the ‘truth value’ for all moral claims is ‘false’. This is because moral claims themselves are built on a false assumption. There is, according to the error theorist, a systematic error in everything we say and do in the field of ethics.
Next to ‘error theory’, she wrote the name “J.L. Mackie.”
She had also placed Mackie’s book, Ethics: Inventing Right and Wrong on the reading list for this course. At the time, it was just one book among many. When I started to read it, I made one of the most horrible discoveries a graduate student could make. Here was somebody who was already saying many of the things that I had been thinking about. Ideas that I had been struggling with had already been written up and published.
It was enough to make a grown graduate student cry.
But I had a ray of hope. J.L. Mackie was dead, and people were still debating his work. In that debate, it seemed that people were getting a few of the details wrong. Simon Blackburn, for example, wrote an essay in which he contrasted two different types of anti-realism; a type that makes a difference in practice, and a type that does not.
He argued that Mackie defended a type of moral anti-realism that made no difference in practice. I thought Blackburn was wrong. So, I still had a paper topic — defending Mackie from Blackburn.
I had to admit that I felt a twinge of irony as I settled on this topic in my Moral Realism class. Mackie was considered the paradigm example of an anti-realist. His error theory effectively said, “Moral statements do not refer to anything real.” Yet, I was going to write a paper defending Mackie as a realist, against Blackburn’s charge that Mackie was an anti-realist whose theories had no practical consequences. I was going to do defend a realist interpretation of a book that started off with the sentence, “There are no objective values.”
II. There are no objective values.
“There are no objective values.” This is the first sentence in the first chapter of Mackie’s book.
I agree with this, in a sense. I accept that there are no hard objective values — what I called in the previous chapter ‘objective(3)’ values. So, what looks on the surface to be a major disagreement between Mackie and myself is not a disagreement at all.
The difference is that I assert the objectivity of ‘soft objective’ values. The question remains as to whether Mackie goes along with this distinction.
A. False Dichotomy
In Chapter 14, I discussed how debate on moral philosophy is confused by a false distinction — a distinction that says that moral values are either objective in the absolutist sense, or subjective in the “each person gets to decide for himself what is right and wrong” sense. Reading Mackie’s first sentence with this dichotomy in mind means that Mackie must be an “each person gets to decide for himself what is right and wrong” subjectivist.
I hoped that Mackie would make this mistake, because that would give me an opportunity to present my own ideas. However, I was sorely disappointed. Mackie almost immediately shows that he recognizes this false dichotomy, and that he is not going to make that mistake himself.
“The denial of objective values does not commit one to any particular view about what moral statements mean, and particularly not the view that they are equivalent to subjective reports. No doubt if moral values are not objective they are in some very broad sense subjective, and for this reason I would accept ‘moral subjectivism’ as an alternative name to ‘moral skepticism’. But subjectivism in this sense must be distinguished from the specific meaning referred to above. Neither name is altogether satisfactory: we simply have to guard against the (different) misinterpretations which each may suggest.” (p. 18)
Translating this into the language I offered in Chapter 13 of this book, it says, “Hey, when I say that there are no objective values, don’t go thinking that I believe values are subjective in the common subjectivist sense. It may mean that — I am not ruling it out (not yet). However, it doesn’t have to. All I am saying here is that there are no hard objective (or objective(3)) values. Nothing more.”
It was as if Mackie turned his attention to me and said, “Alonzo, others may make that mistake. However, I do not.”
The jerk.
B. Common Subjectivism
Well, there was still the chance that, even though he knew that the denial of hard objectivism does not imply common subjectivism, he would still end up defending common subjectivism anyway, leaving me free to argue for soft objectivism.
But, no. Instead, Mackie says, “The subjectivist about value does not deny that there are objective evaluations relative to standards.” (p. 26). Then, after some work relating standards to requirements, adds, “‘Good’, I think, always imports some reference to something like interests or wants, and I intend ‘requirements’ to be read in this sense.” (p. 58).
Okay, this is not exactly the same as saying that ‘good’ means ‘is such as to fulfill the desires in question’, but it is not far off either.
Mackie then claims, “The subjectivist may try to make his point by insisting that there is no objective validity with respect to standards. Yet he would clearly be wrong if he said that the choice of even the most basic standards in any field was completely arbitrary.” (p.27)
So, we can’t pick the relevant desires willy nilly out of thin air.
Then, he went so far as to say, “Something may be called good simply in so far as it satisfies or is such as to satisfy a certain desire; but the objectivity of such relations of satisfaction does not constitute in our sense an objective value.” (p. 27).
I had to read that sentence over and over again. “The objectivity of such relations of satisfaction does not constitute our sense of objective value.”
That did it. Mackie now explicitly wrote that in denying that values are objective, he is not denying the objectivity of value in the sense of ‘being such as to fulfill the desires in question’. He not only admitted that ‘being such as to fulfill the desires in question’ was a real property, but that they yielded a type of objective value — just not the type he was arguing against.
Okay, he uses the word ‘satisfy’. In Chapter 14, I made an explicit distinction between desire fulfillment and desire satisfaction. A desire is fulfilled if the proposition that is the object of the desire is made or kept true, and satisfied if one gets the sensation of pleasure or joy if a desire has been fulfilled and one knows this.
However, Mackie never read my essay, and we cannot assume that he would have disagreed with the distinction.
Still, Mackie has now stated that he believes that objective values do exist, in the sense of ‘being such as to fulfill the desires in question’, and that we cannot arbitrarily select the ‘desires in question’. However, this is not the type of objectivity he is disputing when he says, ‘There are no objective values.’
I hate him.
He was a graduate student’s worst nightmare. Well, he was THIS graduate student’s worst nightmare.
C. Defending an Objective ‘Ought’
Okay, I still had the thesis that moral values in specific are soft objective values that refer to desires other than those of the person making the moral claim. At this point, Mackie had largely only presented options and distinctions. He did not defend any of them, except for a few passive remarks, and his obvious attack on hard-objective values.
Then, Mackie wrote, “‘Ought’ is never purely egocentric; it always points to reasons of some kind other than the speaker’s attitude, though it can in part indicate that the speaker gives that reason his backing.”
AAAAARRRRGGGGGGHHHHHH! Stab! Stab! Stab! Stab! Stab!
Mackie had pretty much stated most of the major elements of what I thought was true about value. There are no hard objective values (values independent of desire) — all values are at most softly objective (dependent on desire). This is not a question of the meaning of words. It is a question of what exists. Hard objective values do not exist. ‘Being such as to fulfill the desires in question’ does exist.
There are still different ways in which an object can be said to “be such as to fulfill the desires in question,” each depending on different desires in question. Of these, there are hard subjective values (those dependent on my desires — what Mackie calls egocentric values) and soft objective values dependent on desires but not necessarily those of the speaker (not entirely egocentric).
Saying that hard objective values do not allow one to argue that common subjectivism has won the day, or even that it has an advantage over soft objectivism. That victory has to be earned. It does not matter how many people make the mistake of arguing, “There are no hard objective values; therefore, we must accept common subjectivism.” It is still a mistake, and one that is frequently found in the claims made by those who defemd common subjectivism.
So, morality is not objective in the hard sense, but is it still objective in the soft sense? Mackie argues, as I do, that this question is best answered by looking at the way people use moral terms. This tells us that moral claims are at least softly objective — they employ reasons of some kind other than the speaker’s attitude.
Mackie still said that moral terms contained a hidden assumption of hard-objectivism. I was not so sure that this was true. It appeared to me that soft-objectivism could handle moral discourse quite effectively. Here was something that I could disagree with Mackie on.
Then I found a chapter in which Mackie made the claim that this dispute was of little significance.
III. Error Theory
Mackie claimed that his thesis, that there are no hard objective values, was not a thesis about the meaning of moral terms. According to Mackie, if one looked at the meaning of moral terms, this ‘standard usage’ contained a false assumption. Specifically, common usage carried the assumption that objects of evaluation had intrinsic value properties. The statement “X is good” (in the moral sense) literally means “X is intrinsically good”. Because nothing is intrinsically good, all statements of the form “X is good” are in error. Thus, Mackie’s theory is an error theory.
Okay, not ALL value claims are false. Claims of the form “X is not wrong” are not false, because claims of the form “X is not wrong” mean “X does not possess hard objective intrinsic wrongness.” Which is true, since nothing possesses hard objective wrongness.
This implication is one that drives those who defend morality up a wall. “Are you saying that slavery, rape, and murder are not wrong?”
No. I’m saying that any claim that they have intrinsic badness is false.
A. The Arguments
Mackie claims that we can find evidence of this false assumption of hard objective value in the way that people use moral terms. All too frequently, people make moral arguments as if certain types of acts have an intrinsic ought-to-be-doneness or ought-not-to-be-doneness written into their vary nature — that they are, as I have described earlier, to be thought of as goodon emitters or badon emitters.
Moral reasoning is done, then, in about the same way that Luke Skywalker uses the force to read the future. You first clear your mind of all distractions, close your eyes, and concentrate on the act in question. Then, The Force will deliver the answer to you, as to whether the act is right or wrong. If it says ‘wrong’ this doesn’t mean that you don’t like it. This means that it has an intrinsic ought-not-to-be-doneness that applies to all beings.
It is all a bunch of hocus pocus, simple tricks, and nonsense, if you ask me.
What one ‘discovers’ using this method is not whether the act is, in fact, good or bad, but only whether one likes or dislikes the idea of having the act done. The sadist can easily use this same method to demonstrate how good it is to torture people. He clears his mind of all distractions, imagines the act of torturing his victim, and ‘the force’ reveals the answer to him that “Hey, this can be fun!”
It’s not a very reliable way of reading moral value.
This probably explains the popularity of this ritual. It allows each person to elevate their own likes and dislikes to the status of supreme goodness, thus giving them permission to disregard anybody who would stand between them and their getting what they want.
I will come back to saying all sorts of mean and nasty things about this ritual a little later in this chapter. It is sufficient to note at this point that people do practice it, and that it is a sign that people take moral claims, at least, to be claims that intrinsic ought-to-be-doneness and ought-not-to-be-doneness exist, and that they think of themselves as having special goodon and badon detectors that they can read when their mind is calm and at peace.
Hogwash.
B. Dworkin’s ‘morons’
Ronald Dworkin uses a similar criticism of moral realism, only his term for this special moral radiation is ‘morons’ — where ‘mor’ comes from the word ‘morality’. I do not like the insulting implications of his word. I have few good things to say about the practice itself, but I can easily see how a person with too many other things to think about, however bright, can fall into this trap. They are not ‘morons’.
On the other hand, if an individual continues to insist on policies detrimental to the interests of others on the grounds the basis of this imagined ‘sense of goodon emissions,” then it becomes appropriate to say, “Listen, buster, the only thing you are detecting is your own bigotry. Civil society means more than granting such people unfettered access to the victims of their preferred form of prejudice.”
C. Atomic Error Theory
Mackie points out that we have had to deal with ‘error theories’ before, such as in science. We regularly make assumptions about the world. Sometimes, we build assumptions into the meanings of our words and then say that the words now refer to something or other. Then, as our knowledge increases, we discover that those original assumptions were, well, less than totally accurate. So, we have to make adjustments, not only to our ways of thinking, but to our language.
The word ‘atom’ comes from ancient Greek, and literally means ‘without parts’. That is to say, atoms were originally taken to be indivisible particles — things that could not be cut into even smaller particles.
So, the 19th century scientist rants, “You say we can split the atom! That is the most absurd hypothesis I have ever heard. The word ‘atom’ means indivisible particle. By the very definition of the word, it is something that cannot be split, you moron. Splitting an atom is like squaring a circle. Perhaps you should go back to undergraduate school and take some remedial chemistry.”
However, the fact that the word ‘atom’ means ‘without parts’ does not, in fact, force us to view the universe as one in which these smallest pieces of gold, carbon, lead, and oxygen are not made up of parts. Whether or not these smallest pieces of each element can be further split into parts is not a matter for the linguist to decide. And when it is discovered that these smallest pieces can be divided into parts, we simply need to revise our language.
D. How to Improve a Language
Faced with the prospect that items that were being called ‘atoms’ could be split, chemists faced three options.
(1) They could have continued the practice of referring to these smallest pieces of an element (e.g., gold, carbon) using the word ‘atom’. If they did this, then the sense that atoms could not be divided up into parts would have to be abandoned.
(2) They could have continued the practice of using the word ‘atom’ to refer to the smallest particles of matter that could not be split. However, they would have had to quit talking about oxygen, carbon, and iron atoms, because such things did not exist.
(3) They could have selected both options at the same time. That is to say, they could have continued to use the word to refer to the smallest bits of an element, and at the same time held that these elements were indivisible. However, faced with the evidence that the smallest bits of an element could be split, they would have been forced to conclude, “Atoms do not really exist.” They would have had to abandon the the word ‘atom’ entirely, and invented two new words. One word would refer to elements that cannot be split, and the other word would refer to the tiniest bits of an element.
We know how this story ends; chemists selected option (1).
If Mackie is right about moral terms, we speak of the things we call ‘wrong’ as if they have intrinsic prescriptivity. No such entity exists. I want to make an important point about this choice. There is no hard-objective answer to this question. There is no experiment to be conducted, no set of observations to be made, that will dictate that any of these is correct and the other two is incorrect. We could role a die to make the decision, if we want, just as long as we make a decision.
This is a fact that I think common subjectivists often grab onto (with respect to ethics) to say that ethics is subjective. They recognize that there are no objective criteria dictating the meanings of moral terms. Therefore, the common subjectivist argues, morality is subjective.
However, there are no objective criteria dictating the meanings of terms in chemistry. These meanings are assigned, sometimes humorously, to the entities the chemist is talking about. Elements are named by the location where they were first discovered. If the scientists who discovered Pennsylvanium had been working in a different state, the same element could have been called Masachusettium instead.
Yet, this fact about chemistry — that the meanings of terms in chemistry are subjective — does not support the conclusion that chemistry is subjective (in the common subjectivist sense). Just as this argument fails to prove chemical subjectivism, it also fails to prove moral subjectivism.
Our tendency is to do what is easiest. For the chemists, the option that generated the least confusion and the least need to reinterpret prior works in the field of chemistry was option (1). They dropped ‘without parts’ from the meaning of the word ‘atom’, then continued to use the word ‘atom’ to refer to the smallest piece of gold, copper, or oxygen.
Yet, it is not difficult to imagine an instance in which this issue came under dispute — a case where chemists took sides, some finding themselves attached to one usage and others to a different usage, and entering into a huge debate about which option is best (simplest).
We see a similar debate in astronomy today over whether or not to call Pluto a ‘planet’. Back when we knew only about big chunks of material in space called “planets”, other we called “asteroids” and “strange things called “comets”, it seemed most fitting to call Pluto a planet. However, we now have a group of discoveries at our disposal that includes the discovery of the Kupuyer Objects. Now, it seems to make more sense to call Pluto a Kupuyer Belt Object then a Planet. But there are people who are emotionally attached to the idea that Pluto is a planet, who protest this change. No objective argument can be provided to prove that those who want to keep Pluto in the realm of planets is objectively correct or incurred.
The important point is that this subjective element — the fact that there is no hard objective answer to the question of what ‘atom’ (or ‘planet’) means, did not, in any way, imply that chemistry (or astronomy) itself is subjective. The linguistic dispute in no way impairs the objectivity of the subject in which the terms are used.
E. Malaria
As an additional exercise, one can run through the same sort of analysis using the word “malaria”. Literally, this word means “bad air” and comes from a time when bad air was thought to cause disease. Why would they think that? Because certain diseases were more common around places that stink, such as swamps and bogs. To prevent disease, physicians tried to keep out the bad air using full-bodied robes with long snouts and a dose of perfume at the tip. Whatever kept out the bad smell, kept out the disease.
As the bacteria theory of disease started to come into view, we can well imagine somebody arguing, “Malaria cannot possibly come from bacteria. Malaria means ‘bad-air.’ To say that malaria comes from a bacteria is to contradict the meaning of the term.”
However, again, the meaning of the term contains a false assumption. Again, we have a choice of preserving the term (and arguing that malaria does not exist, but this disease that has the same symptoms that is caused by bacteria does exist); altering the term (malaria no longer means ‘bad air’ but, instead, ‘that set of symptoms that we originally and incorrectly thought was caused by bad air’. No strong objective answer can be given to this question, but neither the question nor the answer proves that medicine is subjective.
F. Ethical Error Theory
According to Mackie, we have the same type of linguistic problem in the field of ethics. We have the word ‘immoral’, which means, in part, ‘containing intrinsic badness’. At the same time we use the word to refer to such things as lying, rape, murder, and talking out loud in the theater. We discover that ‘intrinsic badness’ is not real. So, we face the error theorist’s problem.
(1) We can continue to use the word ‘immoral’ to refer to such things as rape and murder, while recognizing that it is not in virtue of containing an intrinsic badness that they are immoral. Instead, it is ‘something else’.
(2) We can continue to use the word ‘immoral’ to mean ‘containing an intrinsic badness’ and say that nothing is immoral. However, this does not mean that everything is permissible. There is still a badness which is not an intrinsic badness in rape and murder. We are simply not going to use the word ‘immoral’ to refer to this badness. We will find a different word instead.
The only argument that can be given for preferring (1) over (2) is that (1) is the easiest option, involving the least drastic change in our conventional ways of speaking, the least confusion, and generating the smallest amount of error. This is true with ‘atom’ and ‘malaria’ as well. The only argument that the chemist or the physician could offer for keeping the term’s reference and dropping the theory-laden assumptions was due to the ease of use. But this was not a problem. We make the adjustment, and we move on in these purely objective fields of study.
I believe that the argument is equally strong in all three cases; ‘atom’, ‘malaria’, and ‘immoral’. Our interest has been primarily in the miniscule particles, the symptoms, and the rape, respectively. Discovering that atoms can be split, are malaria is not caused by bad air, and immorality contains no intrinsic demerit, was of significantly less concern. Therefore, we preserve the reference for these terms, but false the flawed
Nowhere in this can anybody find any justification for saying that the absence of a fixed meaning of the word ‘atom’ or ‘malaria’ — bound by God or nature to have one and only one meaning forever, for all people, in all languages, from the beginning until the end of time — implies that chemistry and medicine are subjective. Nowhere in this can anybody find any justification for saying that the absence of a fixed meaning of the word ‘immoral’ — bound by God or nature to have one and only one meaning forever, for all people, in all languages, from the beginning until the end of time — implies that morality is subjective.
G. Moral Good
So, chemists replaced the thesis that atoms are without parts with the thesis that they are made up of electrons, neutrons, and protons. Doctors replaced the thesis that malaria was caused by bad air by the thesis that it is caused by disease-laden mosquitoes and fleas.
In ethics, we can replace the thesis that moral badness and goodness has anything to do with intrinsic prescriptivity and claim, instead, that it has to do with good and bad desires — those being desires that tended to fulfill or thwart the desires of others.
None of this threatens objectivity.
The best defense available for a proposed revision of the meaning of moral terms is to look for the alternative that it makes the most error-free sense — and contributes to the fewest possibilities for misunderstanding — of what we are interested in when we make moral claims.
Yet, ultimately — and this is important, though mistakes here are common — it does not matter which option one picks. Whether the chemist had selected any of the three available options did not affect the objectivity of chemistry. Researchers in the field of medicine could have picked any of the available options and it would have made no substantive difference in the field of medicine. And the existence of choices in the field of ethics makes no difference to the objectivity of ethical claims.
All that follows from the fact that there are options is that, if different people select different options, we need a way to translate phrases from one language to another. Everybody would still be saying the same thing, they would just be using different languages to say it in.
H. Something Else
There is another invalid form of reasoning that is often used to in support of subjectivism that is relevant here, concerning the meaning and reference of moral terms.
Let us take two people using the term ‘moral’ in different ways. One wants to use the term ‘immoral’ to refer to things having an intrinsic property of ought-not-to-be-doneness and says that nothing immoral exists. The other states that the term ‘immoral’ refers to things that a person with good desires would not do, where ‘good desires’ or those desires that tend to fulfill other desires, and argues that immoral things do exist.
It is important to recognize that these two people would not disagree with each other. In fact, these positions can be held by the same person without suffering the least amount of conflict. In fact, I am one such person.
There is only a conflict if one works on the assumption that the word ‘immoral’ must always contain, as a part of its meaning, a ‘something else’, and calling something immoral means attributing this “something else ness” to that which is being called immoral..
If this is what people are doing, then to say that nothing immoral exists is to say that nothing has this ‘something else’ that would make it immoral. While, at the same time, to hold that immoral things (understood as things that a person with good desires would not do) exists is to say that these actions contain ‘something else’.
It is, indeed, a contradiction to hold that nothing has this ‘something else’ and that actions that a good person would not perform has this ‘something else’.
This is the thought process behind the attitude that these views must be in conflict — that something in fact matters depending on which of these three options one selects.
The correct process recognizes that there is nothing at stake in the decision between asserting that moral terms refer to intrinsic prescriptivity and nothing immoral exists, and asserting that moral terms refer to acts according to whether a person with good desires would perform them and immoral actions do exist.
There is nothing more at stake here than there is in the decision between asserting that ‘atom’ refers to things without parts and saying that these smallest pieces of gold, oxygen, and carbon are not atoms, and asserting that ‘atom’ refers to the smallest pieces of gold, oxygen, and carbon.
It is a difficult but important distinction to grasp.
IV. Universalizability
When a person makes a moral statement, like “rape is wrong” or “capital punishment is murder” or “we ought to keep church separate from state”, or, “It is wrong to coerce children in public schools to pledge allegiance to the Christian God”, that person is typically understood to mean “nobody ought to commit rape” or “all capital punishment is wrong” or “nobody ought to try to unite church and state”.
Some people say superficially “capital punishment can be wrong for me and okay for you.” What does that mean? Moral attitudes do not fit comfortably into the model of ordering dinner from a buffet. “I’ll take a prohibition on abortion and an order of the virtue of selfishness on the side. My friend will have the right to choose with love thy neighbor as thyself.” When you order from the moral menu it is assumed that you are ordering the same thing for everybody. Whatever justification lies at the root of that order has to be something that can support the conclusion, “X is wrong for everybody.”
In other words, moral statements are presumed to be universal.
In his book, Mackie described three types of universalizability. To illustrate the difference, he uses an example of driver, on a stormy night, who encounters a stranded motorist. He thinks about driving on and leaving the motorist out in the rain.
A. Type 1 Universalizability
First-level universalizability involves universalizing the actions of the driver, looking only at the situation from the driver’s perspective. Assume that the driver goes past the stranded motorist. If this driver would not condemn any other driver who passed by a stranded motorist in similar circumstances, then that driver’s moral prescription passes the test for Type 1 Universalizability.
B. Type 2 Universalizability
Type 2 universalizability follows closer to the model of, “Do unto others as you would have them do unto you.” Do not just imagine yourself as the potential driver in a car and permit driving on if, as a driver, you are comfortable with driving on. But also look at the situation from the point of view of the stranded motorist, and ask if you, as a stranded motorist, would want the driver to stop.
The answer, according to Mackie, still depends on the attitudes of the agent. A proudly self-reliant agent who loathes asking for help and would not accept it even if it was offered is free to apply that standard to the motorist on the side of the road. “Because I would not be whining for help if I were stranded, I have no obligation to help you.”
C. Type 3 Universalizability
At this level, the agent looks not only at his attitudes as a driver (option 1), or at the attitudes he would have if he were stranded (option 2), but considers the attitudes of all of the different types of people who may be drivers or stranded motorists. This option follows the model, “Do unto others as you would have them do unto you if you were them.”
Just to use an extreme example to illustrate the point, a masochist, under Type 2 Universalization, may liked to be whipped, flogged, and abused. He is permitted, under Option 2, to flog and abuse others as he would have others flog and abuse him. Type 3 universalization, on the other hand, says that flogging and abusing others depends on the preferences of the person being flogged, not the preferences of the person doing the flogging. If the agent would not want to be flogged if he had the preferences of the victim, then flogging the victim would fail the test for Type 3 Universalization.
Here, it does not matter whether the self-reliant driver would not want people to help him if he were stranded. That gives him no more of a moral permission to leave others stranded than a preference for being flogged gives one permission to flog others.
D. Universalizability with Intrinsic Value
Mackie argued that Type 2 universalizability is most compatible with the error of hard objective values. Somebody who takes his or her attitude toward some state of affairs as signs of intrinsic prescriptivity is likely to dismiss as wrong (or perverse, or depraved, or just plain evil) anybody who has a different attitude.
In the case of our hypothetical motorist, if he takes his attitude toward self-reliance as identifying an ‘ought-to-be-doneness’ intrinsic to self-reliance, then he will conclude that all other attitudes toward self reliance are mistaken, and refuse to consider their alternative interests.
In contrast, eliminating hard objective values removes this popular excuse for elevating one’s own preferences to the status of universal, intrinsic, ought-to-be-doneness. The justification from “I value X” to “You ought to do X” vanishes.
E. Mackie and I Part
Here, at last, I can identify an area where Mackie and I seem to have a difference of opinion. Mackie seems to argue that these three types of universalizability are mutually exclusive — that one practices either Type 2 or Type 3 universalizability, but not both (at least not consistently). I want to suggest that morality is concerned with a little of both.
We can easily imagine the driver arguing that, if people were a bit more self-reliant, then we would all be better off. Individuals who do not learn how to take care of themselves are a burden on others. Not only do they end up thwarting their own desires (it would be difficult to imagine the stranded individual wanting to be stranded), they impede on the fulfillment of the desires of others (who must stop and provide assistance). All things considered, the driver would argue, we would be better off with more and stronger self-reliance. One way to accomplish this is to condemn and punish those who are a burden on others, and to praise and reward those who show an ability and willingness to take care of themselves.
Please note that I am not in any way endorsing the conclusion of this argument. It may contain a number of false premises. However, the form of argument is understandable. The types of reasons being brought up are the types of reasons that make sense.
The argument is easier to understand when applied to examples where we do not need to wonder as much about the correctness of the conclusion. If the person standing at the side of the road was a hit man or drunk driver, the claim that ‘their point of view is as valid as mine’ is more easily questioned.
In discussing Type 2 universalizability, Mackie discusses one way of determining whether ‘their point of view’ has merit. This method is consistent with objective(3) (hard objective) values, which Mackie and I agree do not exist. Using this method, the driver notices in himself a certain dislike for being dependent on others, and argues straight from this that others should hold the same attitude. The conclusion would follow from the premises if we took the agent’s disapproval as a sign of intrinsic prescriptivity — a built-in ‘ought-to-be-ness’ that is a part of the very nature of self-reliance. Of course, the argument is flawed. The only legitimate inference an agent can draw from his own immediate emotional reaction to the idea of self-reliance is his own likes and dislikes.
However, this does not lead us directly to Type 3 universalizability, which immediately concludes that all points of view have equal merit.
The second major way to determine if ‘their point of view has merit’ is to ask whether it contributes to the greater fulfillment of desires generally. This is fully consistent with the idea that nothing has value except in virtue of how it stands in relationship to desires — there is no intrinsic merit.
In my description of the reasoning that would be involved, I provided some examples of arguments that can be used to show that self-reliance is good. For example, self-reliant people get more of what they want and are less of a burden on others. Yet, counter-points can be raised. There are benefits to be gained through specialization and trade — by focusing on some task that one can do better than others, and offering those skills in exchange for what others can do better. Indeed, specialization and trade is responsible for much of the great wealth and comfort that people enjoy today.
It is not my point here to argue that a high value for ‘self reliance’ is a defensible. In fact, if asked for an opinion, I would argue against it. Specialization and trade is far more efficient at fulfilling desires generally than self-reliance. Taken to extreme, the self-reliant person lives alone, making his own clothes, raising and canning his own food, building his own house, performing his own medical care, generating his own electricity from a generator he built himself from copper wire that he made himself from copper that he mined and smelted himself. Yes, he is not a burden of others. However, universalizing the principle of self-sufficiency is promoting a short and sure road to universal poverty and suffering.
Ultimately, I cannot say for certain that Mackie would disagree with this. At some point, he suggests that the driver should pick up the stranded motorist. However, he does not provide much of an argument for this view. He might think that no argument can be provided.
The above considerations suggest that there may be a way to rationally defend Mackie’s conclusion in this matter — one that is consistent with his opening statement that there are no hard objective values.
An important corollary that deserves a brief mention is to note that these types of questions are not settled by looking inside oneself. “Let your conscience be your guide” would be great if certain acts in fact emitted goodons and badons and we came equipped with a ‘conscience’ that was, in fact, a reliable goodon detector. But it is a very poor way of determining whether a trait generally thwarts or fulfills desires. To answer these types of questions, an agent has to look outside himself, out at the world, and be willing to discover that he may like things that tend to thwart the desires of others, and may dislike things that will fulfill the desires of others.
V. Reasons
In discussions that I have had on the objectivity of morality, there is another type of argument that pops up quite frequently that I would like to address.
To ‘prove’ subjectivism, some opponents start with a situation where an Agent is in a position to do something horrendous to a Victim. A typical case would be that of an adult seeking to rape and murder a child. The ‘proof’ of subjectivism comes from recognizing that there is nothing that can be said to this Agent that, through reason alone, will cause him to decide against his preferred course of action. The argument states that the fact that a “full appreciation” of the wrongness of rape and murder – determined by whether the Agent changes his mind on what he will do — cannot be conveyed by words (reason) alone, that therefore common subjectivism must be true.
Some objectivists accept the assumptions build into this argument. They then try to show how it could not possibly be in Agent’s interest to rape and murder Victim. If we present these objectivists with a case where Agent has no chance of getting caught, these objectivists will often still answer that the risk of harm is too great. Propose that the adult is somebody who values rape more than any other possible outcome, and the objectivist will assert that this is impossible or irrational, without any argument to defend this claim.
However, it is easy to show that there are possible interests where the rape and murder of the Victim is rational. On a personal desire-fulfillment level, the decision to perform the action (or not) hinges on the result of an equation. In its simplest form, the equation is:
(Benefit × probability) > (Cost × probability).
So, the rape and murder should be performed if the benefit of the rape (its value to Agent) times the possibility of performing a successful rape (virtually certain in these examples), exceeds the cost of the rape (including the cost of punishment if caught) times the chance of one’s being caught.
To argue that such a rape and murder is always wrong requires an argument that it is never the case that the left hand side of this equation could exceed the right hand side of this equation.
Mathematically, this is clearly not the case.
(100 × 99%) > (1,000 × 1%)
Under the assumption that morality consists in what the agent would do if he was aware of all of the facts, raping and murdering the child in these circumstances becomes morally permissible — perhaps even obligatory.
At this point, those who link morality to what the agent would do has already lost. This view leads to the absurd conclusion that the wrongness of raping and murdering a child hinges on the fact that the Agent does not enjoy the rape and murder of a child quite enough. If he can simply muster a stronger craving for this lust-murder, it would become permissible, or obligatory.
The defeat turns into a route when we look at the contrived absurdity of the claim that it is not causally possible for circumstances to occur where rape is rational — that it is not causally possible for an agent to want the rape so much, or care about the consequences so little, or reduce the possibility of suffering the consequences to the degree that would make rape rational.
Just imagine, a cancer patient who has only one week more to live who has obsessed his entire life about forcing himself on some girl. Where does one get the idea that a rational rape is not causally possible, except by wishful thinking and clutching at straws in order to save an absurd theory?
Furthermore, we cannot treat as a mere afterthought that the rationalist moral equation ignores entirely the effect on the victim. Look at the rationalist equation again:
(Benefit × probability) > (Cost × probability).
There is no variable, no place in the equation at all, for the cost to Victim. Victim’s interests do not matter in the slightest to the rationalist, only Agent’s interests matter. According to the rationalist’s conception of moral value, we can alter the effects of the victim from saving her life to a lifetime of torture without affecting the moral value of the act in the slightest.
A. Forms of Persuasion
I have found that it is usually the case that all widely held views have some truth about them. In the case of moral claims, the rationalist has the formula right, but is applying it to the wrong people They look at the rape and murder from the point of view of the Agent. What they should be doing is looking at the rape and murder from the point of view of those of us looking at the Agent.
The question to be answered is not whether the Agent has a reason to refrain from raping and murdering this child. The question to be answered is whether we, society, “the people”, generally have reasons to give people in society a reason to refrain from such actions. The question to be answered is not whether we, by words alone, can convince the Agent not to commit the crime. The question to be answered is whether we can justify giving the agent a reason — by whatever means possible — not to commit the crime.
‘By whatever means possible’ includes much more than getting the Agent to accept a propositional argument that makes no reference to any desires other than the Agent’s own. Propositional arguments are not the only tools we have to cause the Agent to decide not to commit the crime.
Another set of tools that we have involves altering Agent’s desires through the tools of praise, condemnation, reward, and punishment. Through judicious use of these tools, particularly during the Agent’s formative childhood years, we can create a child who will grow up with an aversion to rape and murder. In this case the Agent can be made not to commit the crime because he does not want to. Even if the Agent is in a situation where he knows he could rape and murder a child without getting caught, or where there is not enough time for others to make him suffer for his crimes even if they do catch him (as with the cancer patient near death or the suicide bomber who will die while committing the crime), he will not commit rape and murder. In this type of situation, the child is perfectly safe.
Another set of tools is more brutal — actual threats. Let us assume that our use of the first set of tools has failed. We have sought, since the Agent’s childhood, to give the agent an aversion to rape and murder of children, and failed. We cannot talk the lust-murderer out of his crime, since words alone are impotent when it comes to persuading a person against doing that which they desire.
In this case, we use our second line of defense. We give the Agent (or any who are like them) another type of reason not to rape. That reason focuses on what we will do to him if we discover he has violated the rule (law) against such actions. Using this tactic, we do not change the Agent’s desires. We use the desires he already has, promising to put him in a state which we can reasonably expect him to be adverse to, if he should act to fulfill those other (bad) desires.
Just to be safe, we act to put him in a state he will find undesirable even before he commits the crime, if we can show that he was planning such a crime.
This is the distinction between law and morality. Morality is concerned with using social tools, particularly the tools of praise, condemnation, reward, and punishment, to promote good desires and inhibit bad desires. Its aim is to create internal reasons against (aversions to) doing things that harm others, and creating internal reasons for (desires for) doing things that help others. To the degree that these tools are used successfully, Agents will refrain from harming others even when they can get away with it, and will help others for no other reason than that they want to.
Law is concerned with the second type of reason. Law says, “If the moral reasons fail to take, and you have desires to do things that harm others or aversions to things that protect others from harm, we will inflict consequences that we are sure you will like even less.”
There is no sharp dividing line between law and morality. We still need to be concerned with the concept of ‘just law’, where we apply moral concepts to the institution of law. And, of course, criminal punishment is not the only type of sanction that we have in our arsenal that we can use to threaten people. However, it provides a useful place to start in looking at this distinction.
In Chapter 17, I will be looking more closely at this distinction when I look at the subject of “legislating morality”.
B. Internalism vs. Externalism
Part of the defense of the rationalist view of ethics is tied to a view called ‘internalism’.
Internalists hold that it is not sensible to tell somebody that something is ‘good’ unless you can recommend it to the person that you are talking to. To tell somebody that some particular state is good, while at the same time saying that the person you are talking to has no reason to pursue it is thought to be nonsense by those who hold this view.
If anybody uses the word ‘good’ outside of this context, then they are using the term in what philosophers call the ‘inverted commas sense’. Inverted commas are used when you use a word the way somebody else would use it, though you do not think the word is appropriate. So, a father, on a dark and stormy night, might hold up a bag he had just fetched out of a tree and say to his children, “Here is your ‘ghost’.”. He uses the word ‘ghost’ in an inverted commas sense as if to say, “This is what you idiots thought was a ghost.”
Internalism fits comfortably in with the rationalist view discussed above. The rationalist also says that what it is good or bad for a person to do is tied intimately in with what the agent has a reason to do or to avoid. Where there is no such reason, there is no such goodness or badness. To tell the potential rapist and murderer of the young child that it is wrong only makes sense if one is saying that the potential rapist and murderer already has a reason not to perform the act.
Consequently, all of the problems with rationalism are also problems for internalism. For example, it treats the welfare of the Victim as contingent upon the interests of the Agent. Before we can say that there is any wrongness in what the Agent intends to do to the Victim, we must look to see if the Agent has a reason to be concerned with the well-being of the Victim. If not, then whatever the Agent wants to do to the Victim is morally permissible – perhaps obligatory. Whatever harm comes to the Victim is morally irrelevant – irrelevant to what we tell the Agent he ought and ought not to do..
Internalism says that the less interested the Agent becomes in the Victim’s welfare, the more morally acceptable his rape of the Victim will become. This is in direct conflict with the sense that the less interest the the Agent has in this Victim’s welfare, the more evil he becomes.
Internalism fails to pay attention to the fact that more reasons exist for the Agent not to rape the Victim than the Agent may have. The other reasons that exist are those of the Victim not to be raped, and the reasons that the rest of us have to promote an aversion to rape. It is appropriate to bring to bear also the reasons that potential victims have not to live in fear, family and friends have in the well-being of potential victims, and all of the other reasons that exist against the rape and murder other than the reasons the agent may have against the rape and murder.
Indeed, there is a very subtle distinction to be had that no doubt confuses the issue. It is a distinction between the reasons for the Agent not to rape the Victim, and the reasons-for-the-Agent not to rape the Victim. The former phrase can be interpreted as “the reasons that exist for Agent not to rape Victim,” while the latter is to be understood as closer to “the reasons that Agent has not to rape Victim.” The agent does not have all of the reasons that exist.
I am not saying that these external reasons — these reasons that exist that are tied to the desires of others — will somehow mystically motivate the Agent not to commit the rape if the Agent was aware of them. Reasons that exist are only motivational for those who have them. But nothing yet argues against their existence. Nothing prevents us from talking about them. No law of nature or man says that we may not mention in our discussion of the badness of rape and murder its relationship to the desires of the child, or the child’s parents, you, me, and all those around us who care about the wellbeing of children. We don’t have to talk only about what the Agent who is interested in this rape and murder cares about, as if he were the only being in the universe.
In discussing the moral quality of the rape and murder, why would we even want to assess the rape purely according to its merits and demerits from the Agent’s point of view? We still have our own reasons not to be raped and murdered, and not to have others we care about raped and murdered, and it is perfectly legitimate to make our assessment based on those reasons — based on the reasons that exist, and not just the reasons that Agent has.
C. Externalism
I am an externalist. I do not think that the wrongness of this rape and murder has anything at all to do with Agent’s reasons for or against performing these actions. They have a lot to do with the Victim’s interests in not being raped and murdered. It has to do with our reasons for employing the tools of praise, condemnation, reward, and punishment to create in the Agent a reason not to commit rape and murder. The impact of rape on those interests is in no way contingent upon whether the Agent is concerned with those interests. They are a part of the raw data from which the assessment of wrongness directly springs.
If the Agent understands the meaning of the term ‘wrong’ in this case, he needs to understand that the test for wrongness does not rest in whether or not he gets off on such activity. The test is whether we have reason to try to stop him. He might not care, but his not caring is not proof that the action is not, in fact, wrong.
D. Why be moral?
A good person does not need to ask the question, “Why be moral?” He will do the right action because he wants to. If the Agent is not such a good person, and is genuinely asking the question, there is only one reason left to give him. “If you can’t find a reason in yourself to be moral, then I shall give you one.” I will do so using the tools of praise, condemnation, reward, and punishment. If those have no effect on the desires (reasons) that the Agent ends up having, then I will give yet another set of reasons to be moral – “because of what we will do to you if you are not.”
VI. Simon Blackburn
At the start of this essay, I mentioned that Simon Blackburn noted a problem with Mackie’s presentation.
Mackie began his essay by noting that all moral claims are false — they contain an error of assuming that some type of hard objective intrinsic prescriptivity is built into the very things called “good” in the moral sense of the term. Because no such prescriptivity exists, all moral claims are false.
Yet, Mackie went on to argue for a specific set of moral claims, to defend a form of utilitarian ethics. This caused Blackburn to wonder, “Since Mackie said that all moral claims are false, and yet he goes on to make moral claims, isn’t Mackie being a bit inconsistent?”
Trying to defend Mackie from this charge of inconsistency, Blackburn assessed the situation as one in which realism, or anti-realism, about moral properties makes no difference.
To illustrate his point, he noted that there are numbers-realists who hold that numbers exist in the real world, and numbers-antirealists who hold that numbers are human inventions. Yet, both of them balance their checkbooks the same way. Realism about numbers makes no difference in the ways realists and antirealists use numbers in their day-to-day lives.
Applying this to Mackie, we can understand Blackburn as saying that being a value anti-realist makes no difference in the way a person actually uses moral terms.
Blackburn’s attempt to save Mackie from inconsistency misinterprets Mackie’s work. There is a difference to be found in the moral practices of somebody such as Mackie and those who believe in hard objective values.
We see this difference in the two main types of universalization. Realists about hard objective values tend to engage in Type 2 universalization. Whatever they happen to like is what everybody else ought to pursue. Mackie argues for Type 3 universalizability (or something close to it). We should consider all desires, not just those of the person making the assessment.
There is a limited number of cases in which Type 2 Universalizability would actually work. A good person has the moral freedom to do what he likes. It just so happens that what he likes is that which is compatible with the fulfillment of the desires of others. So, a good person can turn his attention inward, can let his conscience be his guide, go with his gut feeling, and not worry the least about duty and obligation to others, and still come up with what is moral.
The problem comes when a person is not good — when, for example, a person has been raised with a prejudice against blacks or gays or atheists, or a negative emotional reaction to a woman holding a position of authority.
It is easy to see why Type 2 universalization would be tempting to the individual. Using it, no person can ever do wrong. The difference between right and wrong is written on his soul and, by a wonderful piece of coincidence, the individual also happens to find this irrefutable indicator of universal rightness and duty to point to exactly what he likes and dislikes.
Sarcasm aside; ultimately, moral knowledge is not to be found by looking inside oneself — focusing one’s attention inward. This method is not granting access to some font of infallible moral truth written onto the soul by an almighty God or Evolution or The Cartesian Deceiver or any other type of force. It only tells you of your prejudices. Making them the foundation for one’s moral beliefs is to elevate prejudice to the rank of godhood.
There is a clear and distinct different between the Type 2 Universalizability recommended by the false belief in intrinsic values, and the Type 3 Universalizability that Mackie talks about. Contrary to what Simon Blackburn argues, Mackie’s error theory is not a distinction without a difference. The error makes a great deal of difference in practice.