Reply to Patrick O’Neil’s “Ayn Rand and the Is-Ought Problem”

This is a reply to this article criticizing Rand: https://www.mises.org/sites/default/files/7_1_4_0.pdf
Some big picture comments:

A big premise of the essay is that the Is-Ought Problem is a BIG DEAL and Rand fails to address it. And furthermore that she claims to have addressed it, and wrote post-Hume, so according to O’Neil, these facts make her failure to address Is-Ought is a bigger failing of hers when compared to lots of other philosophers who have failed to address Is-Ought.

And O’Neil thinks that every time a moral argument involves engaging with an actual reason existing in the mind of a real human trying to solve a real life problem, it’s okay to use this as the basis of calling the moral argument subjective/subjectivist, presumably because it fails to provide an ironclad, universally applicable justification in all situations. No matter how widely applicable the want, this approach treats such args (or piles of such args) as not mattering at all in terms of objective morality.

O’Neil states this explicitly. Page 14-15 of the PDF (94-95 on the internal pagination):

Any reason offered for why one should not act contrary to moral law must either be a fact or a value. If a fact, it will need to he made relevant by the acceptance (envaluation; not epistemological acceptance) of that fact by the individual’s will, which condition, in turn, renders it subjective.

Also minor annoyance is O’Neil talks about “Randian” ideas a lot. Sigh.

Before offering more detailed comments on the O’Neil article, I want to paste a short curi post in which he addresses the Is-Ought issue, along with some brief comments. He basically wrecks the bad approach laid out in the article and offers a better one, so it seemed very relevant:

http://www.curi.us/1343-the-isought-problem

The is/ought problem is the claim that theories should be supported by facts, but that moral theories apparently cannot be supported by factual statements about what exists. This leads to the problem: how can we justify our moral theories? Can we somehow bridge the gap and infer moral theories from facts? Can we derive moral theories in another way? Or are moral theories always to be mere assumptions or guesses without any sound basis?

A lot of the crit in the essay focuses around failures to justify/prove moral theories.

This is a bad problem, and we can avoid it.

We should start with the moral question: “How should I live?”

And we should start with the life we have now, not take a revolutionary view and try to discover morality starting from absolutely nothing.

O’Neil would likely criticize this as rendering morality subjective.

We should take our current life, and our ways of making decisions, and we should try to improve them. In particular we can criticize them and look for problems in our life, and then we can try to think of new ways of life that wouldn’t have those problems. Through this process of brainstorming and criticism we can improve on our life. Then we’ll have a better life. We’ll have made moral progress. We’ll have learned something about morality, which means to have created moral knowledge.

And thus the is/ought problem is circumvented. The is/ought problem is only important when you approach morality in the wrong way, e.g. by asking “What is good?” or by asking “How can we justify our moral theories?” If we are not essentialists or justificationists (ways of thinking that Karl Popper refuted) then we won’t care too much about those questions. If they were fruitful then they’d be fine, but if we find they are not (which is the thing the is/ought problem asserts: it says that these questions are very hard to answer) then that is not a serious problem, we are not required to answer them.

I like this approach. And I would add, the way to approach Rand’s moral thinking is not to try and see if she succeeded at justifying or proving her moral philosophy from the ground up (she didn’t), or whether we can do so with some modifications and tweaks (we can’t), but whether we can gain anything from Rand’s moral views in dealing with real life problems in our current life and ways of making decisions (we most certainly can!)

Ok so more specific comments on O’Neil.

O’Neil claims that Objectivism takes its name from claim to objective (as opposed to subjective) ethics. He’s trying to set up a frame of trashing Objectivism extra hard for failing to live up to its name. I seem to recall, though, that Objectivism name was more about the nature of reality — an objective external world — and not objective ethics so much. And that one alternative name which was taken but which Rand would have liked was Existentialism (as in, Existence Exists). Also Rand thought there was a philosophy hierarchy and metaphysics as at the base. All of which is consistent with emphasis on the metaphysics.

One thing I want to note is this: observe that curi above gives a succinct definition of the is-ought problem. This prestigious O’Neil guy in his essay, however, starts talking about is-ought gap and Hume’s Law without even a brief definition. He goes into some detail later, but still, seems like bad form.

O’Neil mentions that Rand talks about how Objectivism offers a “rational, scientific, objective code of ethics” and doesn’t like Hume cuz Hume’s a skeptic.

O’Neil quotes this crit of Hume by Rand, which O’Neil thinks is unfair:

Hume’s conclusions would be the conclusions of a consciousness limited to the perceptual level of awareness, passively reacting to the experience of immediate concretes, with no capacity to form abstractions, to integrate perceptions into concepts, waiting in vain for the appearance of an object labeled ‘causality.’ “

O’Neil says:

It is not the case, of course, that Hume’s consciousness does not draw the same conception of causality from witnessed events that virtually all humans do, but rather, that upon analysis, he cannot justify these conceptual products of his consciousness deductively. Intellectually, Hume would maintain that the concept of causality must rest upon faith, for all empirical evidence of its existence becomes evidence only by a prior assumption of the notion of causality.

So Hume wants to accept on faith something his consciousness uses because he can’t justify it deductively. Why is the alternative to “can’t justify deductively” faith? Why not “causality seems to work, pending crits”? Faith has lots of crits. It doesn’t help anything.

Page 5-6 PDF (85-86) O’Neil lays out his thoughts on is-ought problem. He says its an ethical problem not a logical problem. (I’m summing up quite a bit, he goes on in some more detail.)

He says that if you need a “prescriptive premise” in the assumptions of your ethics, then either you need an infinite regress of prescriptions (which he notes Aristotle says is illogical), or you need to have basic prescriptions or set of prescriptions. He says if you’re assuming something is a basic prescription, then either you can disobey that prescription or not. If you cannot disobey it, then it seems silly to call it an “ought” cuz there’s no choice involved. However, if you can disobey it, then you’ve got to offer some sort of reason to follow it, either a positive thing or the absence of a negative sanction. But by offering this reason, which can be rejected if someone lacks the personal desire to get the thing you are offering, you’ve made morality subjective, because you’re relying on the personal subjective desire of the person for your arg to work.

He follows up on this some page 14 PDF (94).

One can admit the intellectual correctness of a proposition of moral law (even of its ultimate value), and yet deliberately defy it.

If morality is knowledge about how to live well in the world, it’s not an issue of whim, and is an objective issue. And trying to defy the problems that choosing to live in a bad way causes will prove futile.

This being the case, we must question what moral law means: If one could not, by any means whatsoever, defy moral law, great difficulty would arise over the appropriateness of the use of “ought”-for we do not normally use that word to command actions which in any event cannot be avoided.

One can choose to live immorally but one cannot avoid the consequences of doing so.

If, on the other hand, one can and does defy the imperative of a moral law, how is one still seen as bound by it?

If you act as if being a thief is okay, other people may punch you in the face when you try and steal their stuff. Or you may be imprisoned. You can’t escape the consequences.

Your assent cannot be COMMANDED by moral truths — you are not a slave to moral truths —  but that doesn’t mean moral truth fails to have an effect on you. SLAVE or ZERO EFFECT aren’t the two alternatives. You can choose to live immorally, and suffer as a result. That is a possible outcome.

Punishment or other ill consequences do not suffice to make moral law objective, for if the person committing the violation of the maxim(s) prefers the object of his illicit action, even when coupled with punitive or other unfortunate consequences, that seems to mean that there is no apparent reason why he should not act contrary to that moral law.

So this is an appeal to the possibility of unspecified arbitrary preferences as an attack on objective morality, which I’ve noticed is very common amongst bad philosophers.

Let’s consider a concrete case: let’s say that someone steals someone’s PS4 because they wanted a PS4 to play Metal Gear Solid V: Phantom Pain. Then they are quickly caught and put in jail for 6 months. Pretty shitty plan for getting to play the game, wasn’t it?

Now an academic philosopher might say “but what if he GENUINELY PREFERS the moments of possession of the PS4 and values them higher than the months he’ll spend in jail.” But nobody has this preference. It’s ridiculous. Nobody steals PS4s for fleeting moments of possession which they know will be followed by lengthy incarcerations. They wanna play games on them or sell them or stuff like that.

Now, they may have very short sighted and unrealistic models of how likely they are to suffer negative consequences. But that just means they made a mistake. If a TIME WIZARD popped in their jail cell and offered them to rewind time back right before they stole the PS4, they’d take this option — they wouldn’t sit there, shake their head with a satisfied smile and say “No, I don’t think so. I’ve achieved what I set out to achieve. My work is done.”

On Page 6 PDF (86), O’Neil says that if people say they want Y, and you say that if you want Y you need to do X  and that they therefore ought to do X, and they seem to agree but then don’t do X, then the possibilities are: they did not actually want Y, or they did not believe you about needing to do X. And if you eliminate these possibilities, then the possibilities are that they are illogical/crazy/etc.

This ignores a case like:  someone wants to lose weight, and is convinced eating less would help, but doesn’t manage to do that. We do not assume they are illogical/crazy…this is common.

Page 7 PDF (87), O’Neil says:

Since Rand has admitted (often) that the entity can choose death over life, however, it is at this critical juncture that the subjectivity enters and dominates Rand’s entire ethical edifice. The whole Randian moral system rests upon the most basic moral command that one ought to do that which preserves one’s life (qua man) […]

I don’t think Rand would agree that her system of ethics is about moral commands. Like imagine describing Roark as someone who “followed” the moral command “do not be a second hander.”

This sounds kinda ridiculous to me. It’s obligation-kinda ethics. But Roark wasn’t following some externally imposed moral obligation which he needed to avoid the sin of second-handedness. He wasn’t tempted by second-handedness. It had no appeal for him.

Continuing:

It makes no difference, then, that the standard is not variable in response to the individual, subjective will, for the commitment to the command (“one ought to behave so as to survive qua man”) demands a more basic deontological moral imperative-setting up the pro- spect of an infinite regression of moral commands-for a rational creature has the option to choose nonsurvival as well as survival: “Metaphysically, the choice ‘to be conscious or not’ is the choice of life or death.” 35

Apparently, there is no reason not to select death over life.

Well if you want to do stuff then you should choose life.

Sometimes people choose death because they are e.g. suffering from some horrible illness that will kill them soon anyways. They value less suffering over continued existence where they can’t do much at all anyways. That seems reasonable.

But generally there’s plenty of reasons to choose life. Np.

Oh next he talks about Kant:

Despite Rand’s frequently expressed loathing for the German philosopher Kant-“the man who . . . closed the door of philosophy to reason, was Immanuel Kant” — she seems to have borrowed an ethical notion from Kant’s moral theory. She attempts to utilize one of the underlying principles of Kantian ethics, that immoral actions are self-contradictory.

He then quotes Rand:

Happiness is a state of non-contradictory joy-a joy without penalty or guilt, a joy that does not clash with any of your values and does not work for your own destruction.

I’m skeptical of his Kant-Rand connection here … like he doesn’t say exactly what Kant mean, or even give a Kant quote on the topic.

On page 8 PDF (88) he asks

Rand says (through Galt) that “rationality is a matter of choice.” This presents an immense problem for the Randian philosophy, for one is forced to ask: If ration- ality is a matter of choice, is the choice to maintain rationality or to abandon it a rational choice or a non-rational choice? This involves a great deal more than semantics, for if it is the case that either choice is non-rational (or even pre-rational), then the Randian system of morality depends, at root, on a happenstantial, unfree choice.

People do stuff instead of reason. Stuff like evasion, whim, authority, closing paths forwards, etc.

Our best understanding of complex human decision making is that it involves choice. Not always super directly — for instance, one cannot simply will one’s emotions to change. But one’s emotions are open to being changed by choosing to acquire certain knowledge and apply certain methods.

Likewise with the issue of choosing to be rational. One does not typically sit around and ask “Am I going to be rational, or a whim-worshipper?” But one makes choices which connect to and affect that choice. Like whether to read Rand or go to a getting-drunk get-together. Stuff like that matters.

So there’s choice involved in whether to be rational or not. Np.

The Randian might attempt to claim that there was a bifurcation, with the choice of rationality amounting to the rational choice, and with the choice of irra- tionality amounting to an irrational choice. This raises a double difficulty.

First, it is unclear how the selection of one of two alternatives can possibly alter the nature of the prior choice that produced the selection, but this point may be debatable. Secondly, the elimination of the possibility of a rationally chosen commitment to irrationality means that there can he no such thing as freely chosen evil.

While evasion involves an evil choice, it is a different kinda choice than what I think may be connoted by “freely chosen evil”, which sounds more like the “am I going to be rational, or a whim-worshipper?” kind of question I criticized above.

I think some philosophers make a rather big deal out of the idea that people do not consciously choose evil from the perspective of their current values, because if they thought something was evil in the sense of being wrong to do, they wouldn’t choose to do it. And they go so far as to attack the idea of evil itself on the grounds that if it requires choice, and no one can consciously choose evil, it can’t exist.

But that’s a bit like arguing that because people don’t consciously choose to wind up broke and penniless, there’s no blameworthy moral choices involved in someone choosing to blow his money at the track. It’s silly. Of course there’s choice involved, and of course some are good and bad.

Rand seems to reject this possibility, however, for she does appear to require the existence of freely chosen evil:

Man’s basic vice, the source of all his evils, is the act of unfocusing his mind, the suspension of his consciousness, which is not blindness, but the refusal to see, not ignorance, but the refusal to know.

For the choice to unfocus the consciousness to be evil-to he a refusal to see rather than blindness and a refusal to know rather than ignorance-it is necessary that it be a freely adopted choice by the reason, for if the individual did not freely and consciously choose it, it cannot be said that the individual acted as he ought not to have acted.

I think this is relying on a mistaken model of how minds work. Like there’s lots of somewhat autonomous ideas in a human mind. It’s not like one central executive freely making conscious choices all the time. We can say that maybe there is something like an executive or a judge or whatever when you consciously focus on some issue and consider it. But that’s not the only model that’s relevant. There’s like unconscious habits and emotions and stuff.

And in evasion, one way to look at what’s happening is, one part of your mind notices something, but another part of your mind refuses to hear/refuses to think about it. Slams the door shut. It is a bit like suppressing dissent.

So by refusing to consciously consider an issue, you’re doing evil. The refusal need not be made as a conscious decision — it can be made at like an emotional level and not really thought about. It can be made in a fog. I think it frequently is.

I think it’s a bit silly to basically say that for the choice to unfocus your consciousness to be evil, it has to be made consciously.

It’s a bit like saying that if a drunk driver didn’t choose to consciously mow people down on purpose, then they didn’t act evilly, because they didn’t really know what was going on at the time they killed the people, since they were so drunk. But everyone knows that nobody is drunk behind a wheel by accident … there is a long train of choices and decisions that put them there, and the driver is responsible for the outcome of that train of choices and decisions, whether they knew what was happening at the time or not.

The issue to focus on in moral evaluation of people’s decisions is more like: was it possible that you could have done better? If so, why didn’t you?’

PDF 9 / Page 89, O’Neil says:

Evil choices and evil actions cannot be seen as contradictory or impossible for the simple reason that they can be done. Since one can choose to will values which will lead to his demise, Rand is left with the problem of why one ought not so to will if one feels inclined in that direction.

There is more than one sense in which something can be contradictory.

Like, strict logical contradiction isn’t the only thing that counts here.

For instance if one wants to have great wealth but not expand any effort getting it, it’s possible that one can play the lotto and get lucky and have this happen. But is a very poor strategy for having this happen. It’s something like a contradiction.

Most of the remainder of the essay wasn’t engaging with Rand but with some would-be Rand defenders, which I found boring and didn’t have many comments on. One more thing though:

Rand and her defenders have done little to treat the problem of the effect which the inevitability of the loss of life has upon the role of life-as-the-ultimate-value. Since one must (will, if you prefer) die at time t+x, Rand and her defenders must elaborate why that loss of life at time t+ x is to be preferred (automatically) to the loss of life at time t+(x- 1) or at time t+(x- 1,000) or at any other time. This problem flows out of the lack of time differentiation in the Randian valuation of life, and it will not be pursued further here.

Life as the ultimate value does not make Oists lifespan-maximizing utilitarians, which seems to be the assumption here.

Nozick had a similar confusion.

If the pursuit of some value involved reduction in maximum lifespan it could be fine. Even in common usage, when people talk about having a good or great or successful life, they are not talking about absolute maximum lifespan.

NOTE: If you liked this post, FRIEND OF THE BLOG Alan Forrester posted various replies to Nozick’s criticism of Rand (including mine) here