Misinterpreting J.L. Mackie

This is quite uncharacteristic of me.

In the last week, I have taken one of my philosophical ideas, written it up as best as I could, and emailed them to certain academic philosophers who have written on the subject asking them if the argument makes sense.

It is nothing earth-shattering. It is the thesis that, for 40 years, philosophers have been misinterpreting the error theory of J.L. Mackie. It is an argument that can be expressed without going off into volumes of text (which would be a huge imposition on the recipient). Yet, it does touch upon a number of ideas that I present in this blog.

What I am attempting to demonstrate is that a standard interpretation of J.L. Mackie's error theory is mistaken.

That mistake probably originates with Simon Blackburn in the article, "Errors and the Phenomonology of Values" (in Morality and Objectivity: A Tribute to J.L. Mackie, Ted Honderich (ed.)).

Blackburn was puzzled by the fact that Mackie spent the first part of his book, Ethics: Inventing Right and Wrong arguing that all moral claims are false. Then, in the second half of his book, Mackie went on to make a bunch of moral claims as if they were true. In other words, even though Mackie argued that all moral claims are false, he went on to conduct his moral business as usual in the second half of his book.

Blackburn, in pointing out this problem, explained what he thought Mackie should have done.

If a vocabulary embodies an error in some use it would be better if either it, or a replacement vocabulary, were used differently. We could better describe this by saying that our old, infected moral concepts or ways of thought should be replaced by ones which serve our legitimate needs, but avoid the mistake.

Blackburn, however, denies that Mackie did this.

Yet Mackie does not say what such a way of thought would look like, and how it would differ in order to show its innocence of the old error. On the contrary, in the second part of the book he is quite happy to go on to express a large number of straightforward moral views… All of these are expressed in the old, supposedly infected vocabulary.

I want to argue that Mackie did exactly what Blackburn said he should have done, and not what Blackburn said Mackie did.

To begin with, Blackburn gave Mackie a choice to either change the existing vocabulary or to replace it. To illustrate the second option, Blackburn suggested replacing the term “moral” and its cognates with “shmoral” and its cognates.

However, Mackie selected the first option. He opted to use the same words, but to use them in different ways.

He illustrated the way that he was going to reform moral concepts by using the concept of an “atom” as an example.

The fact that the word 'atom', as used in nineteenth-century physics, had as a part of its meaning 'indivisible particle of matter' did not in itself, even in the nineteenth century, compel anyone to believe that there are indivisible material particles. One could either refrain from using the term 'atom' in affirmative statement or, as physicists have subsequently done, use the term with other parts of its meaning only, dropping the requirement of indivisibility. (Mackie: p. 100).

Specifically, the term "atom" originally meant "without parts" or "uncuttable". Physicists were studying helium, hydrogen, oxygen, and carbon atoms under this definition – assuming that these smallest bits of each element could not be further divided into parts.

However, as physics progressed, physicists began to suspect that the smallest pieces of an element actually had parts – that they were made up of electrons, neutrons, and protons. It was at least theoretically possible to “split” what people had been calling atoms. However, that meant splitting what, by definition, is that which could not be split.

Ultimately, physicists dropped “without parts” from the meaning of the term “atom” and simply used the term to refer to the smallest pieces of an element – the smallest bits of matter that still could be identified as hydrogen, helium, gold, carbon, iron, and the like.

This did not produce any great crisis for physicists. With this slight change, they were able to continue to go along studying atoms. However, this was not “business as usual”. From that point on it was no longer the case that atoms, by definition, could not be broken up into parts.

Similarly, Mackie is going to continue to use the same moral terms. However, he is going to alter the meaning of those terms. Specifically, he will replace a meaning that contains an false claim of intrinsic prescriptivity and put in its place a term that lacks this error.

This brings us to the next question: How do we distinguish between error-filled “moralizing” and error-free “shmoralizing”

I think we can find that distinction in Mackie's chapter on universalizability.

In Chapter 4 of Ethics, Mackie distinguishes three stages of universalizability for moral claims. The argument here is that what distinguishes "moralizing" from "shmoralizing" is that the former uses Stage 2 universalizability, while the latter employs Stage 3 universalizability.

Mackie described Stage 2 universalizability as follows:

In this second stage of universalization, we look for prescriptive maxims that we are prepared not only to apply to all persons (groups of persons, nations, and so on) alike as things are, but also to go on applying no matter how individuals change their mental and physical qualities and resources and social status. And we must allow not only for changes which may, as a matter of practical, causal, possibility come about, starting from where we are, but also for differences of condition and inversions of role that could not possibly occur, and which it may take a considerable effort even to imagine. (Mackie: p. 90)

This is to be distinguished from Stage 3 universalizability, which he described as follows:

In this third stage we are taking some account of all actual desires, tastes, preferences, ideals, and values, including ones which are radically different from and hostile to our own, and consequently taking some account of all the actual interests that anyone has, including those that arise from his having preferences and values that we do not share . . . We must . . . look not for principles which can be wholeheartedly endorsed from every point of view, but for ones which represent an acceptable compromise between the different actual points of view. (Mackie: p. 93).

He then explains the way that Stage 2 universalizability relates to the practice of taking moral values as intrinsically prescriptive properties, and the way that Stage 3 universalizability avoids this error.

This second sort of universalizability is linked with the fact . . . that moral judgments commonly include a claim of objectivity. The claim that some difference is objectively morally relevant in a certain context is not easy to reconcile with the admission that, while it appears relevant from one interested point of view, it does not appear relevant from the point of view of someone whose situation and qualities are different. By contrast, the claim to objectivity has no tendency to support the third stage of universalization. Quite the reverse. It is all too easy to believe that the objective validity of one's own ideals provides an overwhelmingly strong reason for taking no account at all ideals that conflict with them, or of interests associated with including rival ideals (Mackie: pp. 96-7).

In other words, “shmorality” is not “business as usual”. “Business as usual” would involve continuing with the practice of engaging in Stage 2 universalizability. However, this level of universalizability embeds the false assumption that “objective values” (a.k.a. intrinsic prescriptivity) exists. We avoid this error by dropping intrinsic prescriptivity from the meaning of our moral terms and adopting the new practice of Stage 3 universalizability.

So, when it comes to error-free moralizing, Blackburn’s claim that “Mackie does not say what such a way of thought would look like, and how it would differ in order to show its innocence of the old error," is false. It would look like Stage 3 universalizability, and it denies that objects of evaluation had intrinsic prescriptivity.

Blackburn wrote that, "...in the second part of the book he is quite happy to go on to express a large number of straightforward moral views… All of these are expressed in the old, supposedly infected vocabulary." This is false. Those moral views are expressed in the error-free vocabulary of Stage 3 universalizability.

If I may beg your indulgence, I think I can further support this conclusion by approaching it from another direction.

I think that there might be some confusion generated by the fact that Mackie talks about two types of objectivity in his book.

There is a type of objectivity that involves a claim of intrinsic prescriptivity. This is the type of "objective value" that Mackie says does not exist.

However, there is - at the same time - a type of objectivity that can be captured by the phrase, "Is such as to satisfy requirements of the kind in question," (Mackie: p. 55).

"Requirements", in turn, is cashed out in terms of desires or interests. From Mackie:

'Good', I think, always imports some reference to something like interests or wants, and I intend 'requirements' to be read in this sense, not so colourlessly as to be almost equivalent to 'criteria'. (Mackie: p. 58)

This type of objectivity is real.

However, this type of objectivity is not the type that Mackie is talking about when he says that there are no objective values. From Mackie:

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. (Mackie: p. 24)

Mackie writes about the “objectivity of such relations” as if they are real things. However, these real things are not currently built into the meanings of moral terms.

But he does not say that they could not be.

I understand Mackie as wanting to get rid of objective values in the sense of getting rid of intrinsic prescriptivity (Blackburn’s "morality") and replacing it with objective values in the sense of "being such as to satisfy the desires and interests in question" (Blackburn’s "shmorality").

The switch, then, from "moralizing" to "schmoralizing" is a switch from a morality that defends its moral evaluations of character traits or principles in terms of intrinsic prescriptivity, to a morality that defends its moral evaluations of character traits or principles in terms of "being such as to satisfy the desires and interests in question".

As I see it, David Hume's moral theory provides a model by which we may understand Mackie's "schmorality". A character trait, in Hume's system, is to be evaluated according to whether it is agreeable or useful to the agent (giving the agent reasons to promote it), and whether it is agreeable or useful to others (giving others reason to promote it). It has no intrinsic goodness or badness.

Similarly, we find in Mackie's account of stage-3 universalizability:

In this third stage we are taking some account of all actual desires, tastes, preferences, ideals, and values, including ones which are radically different from and hostile to our own, and consequently taking some account of all the actual interests that anyone has, including those that arise from having preferences and values that we do not share. (Mackie: p. 95)

Once again, “shmorality” is not “business as usual”. “Business as usual” is objectivity in the form of intrinsic prescriptivity. “Shmorality”, in contrast, is objectivity in the form of “being such as to fulfill the desires or interests in question.”

I am not saying that Mackie is right in all of this. I am only saying that this seems to be a better interpretation of Mackie than the "business as usual" interpretation. Once we understand what Mackie actually wanted to say, then we can go on and ask whether or not he was right when he said it.