Why Johnny Can't Proselytize:
A Primer on Meta-Ethics
Peter A. Taylor
April 30, 2014
Deontology is rationalised emotion.
Moral arguments that have to cross political or theological boundaries are often exercises in frustration. Libertarians' moral arguments aren't cogent for Progressives and vice versa. Christians' moral arguments aren't generally cogent for atheists and vice versa, etc. The criticisms I make here are not unique to libertarians, but libertarians are my primary intended audience. My purpose here is, first, to convince my fellow libertarians that they need to raise their game, and second, to suggest how to do it.
Why is it so hard to construct a moral argument that is cogent for someone who doesn't already agree with you?
1. Your listeners' motives are perverse.
Consider Machiavelli's prince. Machiavelli was writing for a decision maker whose choices were decisive. If the prince made a bad political decision, that bad decision had a 100% chance of being put into effect, and he stood a good chance of paying for it with his life. Machiavelli didn't need to make moral arguments. If the prince was rational and self-interested, practical arguments would be sufficient.
Now consider your audience if you are talking to ordinary voters in a large democracy. Here the chance of any one voter's decision changing the outcome in a major election is closer to one in a million. We are faced with "the Paradox of Voting". The question is why anyone bothers to vote at all at those odds.
The way Jonathan Haidt explains it is that voting is expressive. The odds that your vote will alter the amount of money in your pocket may be one in a million, but the odds that "your" side will win are more like 50-50, and the odds that you will get emotional gratification for your vote are 100%. Practical arguments are therefore of very limited value. Practical arguments are important only in so far as they establish in which direction the greater emotional gratification lies.
When a politician promises to put more money in the pockets of some dispersed group (e.g. the poor), what he's really doing is stroking their egos by telling them that they deserve more money. (Promises to major campaign money bundlers are another story.) As long as the politician continues to make occasional speeches telling them how worthy they are, he will remain popular with this dispersed group and for the most part, they won't care whether he keeps his promises.
A moral argument may be defined loosely as one that entails emotional gratification. If you are going to persuade ordinary people to vote your way in an election, you pretty much have to use moral arguments, even if those arguments are only implicit. Otherwise, why does my audience care what I have to say? I may be asking them to do something that they do not individually have the power to do, and they may lose face by even listening to me.
2. Your opponents' motives are perverse.
It is not just listeners who have perverse incentives. The person making an argument may be trying not to overcome his listener's biases but to manipulate them. Persuasive speakers may find themselves in several very different situations:
A good scientist should be trying to overcome biases and searching for truth, even if it means personally getting egg on his face.
A lawyer may be trying to persuade, manipulating biases, and may be indifferent to the truth, but doing so in the face of opposition that he can't get rid of.
A preacher may not be trying to persuade the infidels at all; he may be trying to clear the sanctuary of them in order to create an illusion of consensus within his own flock. Alternately, a heckler may be trying to shatter the illusion of consensus in someone else's church.
A bully may be trying to intimidate the opposition into silence rather than to persuade. This could be an attempt to create an illusion of consensus or to interfere with his political opponents' attempts to organize. Or it could just be that the bully is trying to demonstrate his social dominance.
Your opponents may stand to lose a great deal of face if they are seen to lose a debate with you. They may also be in a position to threaten your listeners with ostracism or other social sanctions if they give your argument a fair hearing. If this is the case, you are facing an implicit ad baculum argument (threats or bribes). Ad hominem (to the person) attacks are often implicit ad baculum arguments, threats of ostracism aimed at one's audience.
This is, in fact, the usual situation libertarians face in arguing against progressives who have the support of the mainstream media. This is true regardless of whether media bias is the result of "attractive power distortion" or "coercive power distortion" (see Open Letter part 4).
Ad baculum arguments work. That's why people use them. Pascal's wager (see Research on Religion podcast) may sound stupid as it is commonly Bowdlerized, but consider it again. Belief is not a yes/no proposition; there are degrees of uncertainty, and degrees of committment to a proposition. If I go skydiving, I may commit myself to the proposition that the parachute was packed correctly, even though my confidence in it is something less than absolute. I may not be able to flip a switch from absolute disbelief to absolute belief, but I can commit myself to a course of action despite some degree of doubt. I can invest resources, reputation and "face" in a belief (e.g. by publicly endorsing it). Threats influence the expected payoffs to Pascal's wager. Ad baculum arguments may also exploit a natural tendency for people to kid themselves about uncertainty.
One consequence of the desire to save face is that arguments about political legitimacy tend to get hijacked by attitudes towards the status quo. Will Wilkinson (via Rand Simberg) complains about this tendency in Liberalism, Libertarianism, and the Illiberal Security State.
It's nice that libertarians have kept liberalism alive, but it would be even nicer if it were possible for liberals to espouse liberalism without therefore being confused for libertarians.
3. Your motives are perverse.
You, like your listeners and your opponents, are a human being. You are probably motivated largely by social status, just like they probably are.
If you're a libertarian, you're probably not in a position to threaten anyone with ostracism, but you may have a bruised ego and some cravings for acceptance, reassurance, or simply escape, any of which will be hard to satisfy in a normal social environment. These cravings can get you in trouble. You may think you're trying to persuade a progressive that libertarianism is respectable, but unconsciously, you may be trying to reassure yourself. Or perhaps you just want to escape, to stop interacting with someone without being seen as fleeing a battle. Either of these motives could drive you to try to delegitimize the other person's feelings rather than trying to engage them.
Worse, you could be tempted to make unprincipled compromises in an attempt to gain social acceptance, essentially begging for scraps at the Progressives' dinner table. This is usually done either by embracing the latest Progressive fetish or by making exaggerated criticism of potential allies such as social conservatives. This results in the phenomenon known as "Beltway libertarians".
Freedom from the desire for an answer is essential to the understanding of a problem.
— J. Krishnamurty
What I mean by overmotivation in this context is the burning desire for a shortcut. Sometimes it looks like laziness. Sometimes it leads to "noble lies". Often it means jumping to a conclusion. It's hard to tell whether someone reached a silly conclusion because of overmotivation or because he was simply confused, but if he doesn't correct the error when it's pointed out to him, I blame overmotivation.
4. There is a temptation to misrepresent opinions and tastes as facts.
First of all, if you're at all attracted to Objectivism, please spend five minutes with Madsen Pirie, Economics is Fun, Part 1: Value. Value is subjective. If value were not subjective, there would be no reason for people to trade with one another. But beware of possible conflicting usages of the word, "value".
Consider The Prisoners' Dilemma. Given a certain payoff matrix, it is an objective fact that the game has a certain "Nash equilibrium" solution. The lengths of the prison sentences are objective facts. But there is a disconnect between the sentences, measured objectively in years, and the values of the various outcomes to the prisoners, measured subjectively, perhaps in "utils". The payoffs that matter are subjective. If the prisoners care about each other enough (e.g. a married couple who are madly in love), the prosecutor may find that he has correctly solved the wrong problem. We have to assume something about subjective values before we can derive our objective Nash equilibrium. Our objective superstructure is built on a subjective foundation. There is no escape from Hume's Guillotine.
There are a number of different schools of thought in moral philosophy, and one of the ways in which they differ is in the claims they make regarding objectivity. I'm a fan of Adam Smith's The Theory of Moral Sentiments, which puts me in the "moral sympathy" school. There is also a contractarian school, virtue ethics (e.g. ch. 8 of The Happiness Hypothesis), and a consequentialist school (which includes utilitarianism). But what concerns me here is mostly the rule-based deontological school, specifically the moral absolutism faction. I don't want to dismiss the deontological school entirely; we need rules, and there is a place for extrapolation. But in so far as deontology turns moral intuitions into axioms to be manipulated with mathematical logic, and asks people to treat the results as matters of fact, it is cheating. Deontology in general, and Objectivism in particular, tends toward begging the question (petitio principii). The best one can hope for through this kind of reasoning is to point out non-obvious connections between different rules.
I will return to the distinction between objective and subjective at length later.
5. There is a temptation to deny the amount of uncertainty in consequentialist arguments.
|Astronomers quail at the three-body problem and throw up their hands in surrender before the four-body problem.|
|Any given moment in history is a problem of at least two billion bodies.|
|— C. M. Kornbluth's fictional character, F. W. Taylor, from Syndic|
I'm much more sympathetic to the consequentialist school than the deontological school, but consequentialists can also get into trouble by trying to take a shortcut. This often takes the form of vague tendencies being misrepresented as laws of nature.
This problem is partly inherent in inductive reasoning, the process by which people with finite amounts of information draw inferences about general principles. You may have a principle that seems to work 95% of the time, and not to work the other 5%, but the evidence is subject to interpretation. When you see a limited amount of questionable evidence against a theory that generally works, do you say that the theory is wrong, or that you have some bad data? I'll cut you some slack when you guess wrong, but you have to admit to some uncertainty.
Utilitarianism comes in at least two flavors, act-utilitarianism and rule-utilitarianism. The former judges each act independently according to its anticipated consequences, and the latter lumps acts into categories, and judges them by category. Act-utilitarianism strikes me as insanely hubristic. A sane person can't expect to predict the future in detail. The best we can do is try to guess which way the statistics are likely to trend in the long term.
The trolley car scenarios one sometimes encounters in discussions of ethics similarly overlook the problem of dealing with uncertainty. Harry Emerson Fosdick wrote, "He who chooses the beginning of a road chooses the place it leads to. It is the means that determines the end." Fosdick is half-right. I knowingly choose the beginning of a road, and this determines the place it leads to, but whether I can be said to have knowingly chosen the place it leads to depends on whether or not I have a reliable map. The distinction between ends and means is a reflection of uncertainty. One of the topics I studied as an Industrial Engineering student was "Decision-Making under Uncertainty," and it strikes me that this topic could just as easily have been called "Moral Philosophy for Engineers". Many of the moral questions we are interested in are things like, "What are the long term consequences of no-fault divorce?" These sorts of practical sociology problems are too complicated for intuition to work reliably, and our analysis tools are pitifully inadequate. We have to rely on experiment. But it takes 500 years to run an experiment, and there are no controls.
Libertarian and Objectivist moral arguments are heavily influenced by the natural law tradition. The versions of natural law I am most familiar with are emphatically consequentialist; they depend for their plausibility on the belief that human behavior can be classified fairly easily and objectively into behavior that is "natural" in some desirable sense and that which is not. There are two problems here. One problem is that what is desirable is inherently subjective, even when there is widespread agreement about its desirability. Even if everyone in the world likes vanilla ice cream, the claim that vanilla ice cream tastes good is an opinion, not a fact. And when we start talking about a desireable distribution of wealth, there is not widespread agreement. So strictly speaking, natural law is not objective. The other problem is that natural law depends on different people drawing similar conclusions when relating usefully broad categories of behavior to long term consequences. For example, what are the consequences of minimum wage laws? Not only do lay people not believe what expert economists say about the effects of the minimum wage, the experts don't even reliably agree with one another. As a general rule, initiating coercion is a bad thing, but we recognize exceptions for things like taxes to pay for various programs. But there is very little agreement on what the effects of these programs are and on whether a hypothetical "state of nature" is even relevant in the modern world. I discuss natural law in more detail in section 10.
6. There is a temptation to deny the context in which your arguments make sense.
Bernard Crick warns about this in In Defence of Politics. Similarly, if I recall correctly, David Friedman explained that the difference between a scholar and an ideologue is that a scholar remembers the assumptions he had to make in deriving his conclusions, and is worried about whether they're still true.
This is seen in the immigration debate. Libertarians like Bryan Caplan take political stability for granted. Steve Sailer, in contrast, points out that in practice, property rights come out of the barrel of a gun. The numbers of people on your side and on the other side matter, and the beliefs and loyalties of prospective immigrants matter. Other people have ethnic solidarity even if you don't. Good government depends on culture, especially in a democracy.
Mencius Moldbug's Sick journey from Mises to Carlyle has a good discussion of the context within which libertarianism can work. He has another good discussion of the context in which democracy can work in Democracy, cis and trans.
7. Neither you nor your listeners are entirely rational.
I've been trying to couch my explanations in terms of game theory or rational choice theory, with comprehensible conflicts of interest. But the human mind isn't all that rational, even in the limited sense of being dishonest in understandable ways.
Also, game theory depends on having stable, or at least predictable, preferences. Part of the magic with moral persuasion is that, to some extent, the listener's reward system is being rewired.
More confusion between objective and subjective
Reading through Jonathan Haidt's The Righteous Mind should be a good replacement for years and years of degrees on ethics and moral philosophy, which I now visualize as an industrial sweatshop of rationalization hamsters run by an evil medieval bishop.
8. Meta-ethics is qualitatively different for atheists vs. Christians.
Confusion between objective and subjective came up recently in an exchange with Handle on Nick Land's Outside In website. It began with a discussion of "demotism" as a theory of governmental legitimacy here (the idea that majority consent trumps natural law), then started up again in Chaos Patch #8:
No legal thinker represents this trend more clearly than Justice Oliver Wendell Holmes, Jr. Holmes took delight in ridiculing the notion of individual rights. "All my life," he said, "I have sneered at the natural rights of man." And he meant it. That idea rested on the notion that there are such things as knowable principles of justice, when in reality, Holmes believed, there was only physical violence and pre-rational emotional preferences. The idea of, say, the right not to be raped or murdered, has no more validity than a person's preference for one type of beer over another. Legal and constitutional protections for rights are simply an aggregation of arbitrary personal preferences that the majority are willing to kill or die for. But that means that one can have no rights valid against the state: "there can be no legal right as against the authority that make the law on which the right depends," he wrote. The idea of a right valid against the state was like "shaking one's fist at the sky, when the sky furnishes the energy that enables one to raise the fist."
Comment 33054: Handle describes two streams of progressive thought, "eliminative materialist" (or "nihilist") vs. "hollow-universalism". The former (e.g. Holmes) tend to regard morality as too subjective to take seriously. The latter (e.g. Dewey and most modern commentators) pretend that it is grounded in objective fact, but tend to become incoherent when pressed for details.
Comment 33070: I complain about "people pretending for propaganda purposes that their sentiments are objective facts," and link to "Book Review: Knowledge and Politics", by Arthur Leff (via instapundit), AKA "Memorandum from the Devil". Is your moral system grounded from within human society (i.e. human preferences), or is it grounded from outside human society (i.e. God's preferences)?
Traditional Western moral teachings are based on Christian theology. God's opinion may be technically an opinion, but it lies outside of human society. As far as human beings are concerned, if they can figure out what God's opinion is, it can be treated as being an objective fact. Christians worry about whether God has been misquoted, whereas atheists worry about how to aggregate human opinions. The atheists' meta-ethics problem is very different from the one the Christians face. Atheists can't use the Christian God to serve as their ethical standard.
It is somewhat misleading to talk about "the death of God". Most Americans still believe in some sort of God, but the God most of them seem to believe in is more like the God of Deism than the God of the Bible. Deists have the same problem atheists have. "The vast indifference of heaven" that Warren Zevon sang of is not a suitable basis for a system of ethics. Leff drives home this point by making his Devil into a sort of Deist, who writes,
My own opinion is that the Hand that holds you suspended over my fiery pit doesn't abhor you, but has forgotten completely that It has anything in It.
Comment 33138: Handle links to:
Unaussprechlichen Kulten, by Friedrich von Junzt (via H. P. Lovecraft)
"Unspeakable Ethics, Unnatural Law", by Arthur Leff: More on the Gödel problem: If you want to present morality as an exercise in logic-chopping (deontology), you are either faced with the problem of validating your system from within the system or finding a moral reference external to the system.
9. There is a temptation to moral scientism.
The Outside In thread continued with a discussion of some philosophical inconsistencies in attacks by atheists Jerry A. Coyne and Noah Millman on an article by Ross Douthat.
References (via Handle):
Douthat: Ideas From a Manger
Coyne responds to Douthat
Douthat responds to Coyne
Millman responds to Douthat
Douthat responds to Millman
Comment 33086: Handle describes two levels of confusion involving sentiment and fact. Edited quotation:
Intentionally misrepresenting one's actual position for the purposes of social influence and manipulation (noble or otherwise)
Negligently fooling themselves into biased cognitive dissonance.
Now, since there is such a huge reward for winning (getting people to act and believe the way you want them to which they wouldn't otherwise do, i.e. 'power'), there is always a huge amount of perpetual competition for Deutungshoheit, to establish 'social norm moral-narrative air dominance'. And, yes, as usual, the best liars are the folks who can fool themselves into really believing their own lies. It helps in suppressing one's tells....
The result is a kind of moral scientism. And I see this style of thinking all the time: "Moral progress is just like Scientific progress. Man can discover and reveal the moral natural laws by a process of observation and reason, and make moral progress by displacing the erroneous, superstitious, mythological moral theories of the past...."
Anyway, the point, I think, that Ross Douthat is trying to make is that, if you care about being coherent and consistent, you should take one of the two following approaches as axiomatic priors / premises:
Morality is in principle groundable outside 'preference'. And we can debate whether it has been grounded, or what that ground should be, or how to rank systems, or what process we should use to find that ground, the proper jurisprudence of hermeneutics, interpretation, derivation, etc.
Morality is not groundable beyond preference (Like that Devil's Memo you linked). So I'll stop pretending it is and that I'm in possession of the best enlightened form of it.
Douthat says people like Coyne use general 'strong not-groundable' arguments against traditional religious morality — that is to say, arguments that are not specifically only anti-religious and are of the nature of (B) — but then immediately start lecturing us about their awesome, enlightened, universal form of secular ethics as the best substitute alternative as if they really believed in (A) after all.
Whenever supporters of secular ethics hear this argument they go nuts, get very emotional and insulted, etc.... If they were in possession of a solid argument then perhaps they could just make it and be done with it. But that doesn't happen. Possibilities (not necessarily mutually exclusive)
They have a solid argument, but it is unpersuasive or intellectually inaccessible to us stubborn, biased, and stupid types, and banging their head against us just makes them frustrated and irate.
They don't have a solid argument, and:
As with case (1) above, they realize they need to act 'insulted' and 'as if they do' in order to defend the 'noble lie' that is central to maintaining the Deutungshoheit essential to their propaganda efforts.
As with case (2) above, they think they have a strong, sufficient argument, but are both biased and mistaken on this point and unable or unwilling to realize their error.
I think Douthat's position [explaining Coyne's behavior] is II.B. And that when you tell a II.B person that they are a II.B. person, they scoff and act insulted and go nuts for two reasons.
First, there is a ideological-inertia bias, and a personal, emotional, ego-preservation, psychological-defense mechanism reaction whenever someone tells you that what you sincerely believe is bunk you've been fooled into believing. Again, I'm sure we've both witnessed, and even personally experienced, reactions like this.
Second (and regardless of whether or not you have a strong argument, or whether or not you believe in it), there is the social implications of the prospect of other people coming to believe you're a II.B. You'll lose credibility, trust, status, and influence. The stakes are huge, and so you'll deny it vociferously until your dying day, call the charge 'plainly stupid' and try to mock, ridicule, and otherwise downgrade the status of anyone making the charge.
And this, in my view, is precisely what Noah Millman tried to do in his response to Douthat....
Handle has another article up, The Atheist Evangelist, that discusses Sam Harris, wrestling with similar issues to those discussed above; having subjective preferences and wanting to treat them as objective facts.
Claire Berlinski makes the same complaint in a review of Chantal Delsol's Icarus Fallen: The Search for Meaning in an Uncertain World:
Delsol's is certainly not the first baleful assessment of our ambient culture of moral relativism — perhaps quasi-relativism is more apt because, as she rightly notes, its practitioners unquestionably accept moral absolutes ("one must be tolerant") while insisting that they indignantly reject them.
See also Divine command theory. There is an old joke that Unitarian Universalists are a bunch of atheists who can't break the habit of going to church on Sunday. Similarly, it seems like there are a lot of atheists who can't break the habit of relying on Divine command theory.
Update: Theo Hobson writes (Hat tip: Nick Land):
On one hand [Dawkins] believes that morality, being natural, is a constant thing, stable throughout history. On the other hand, he believes in moral progress. To square the circle he plunges out of his depth, explaining that different ages have different ideas of morality, and that in recent times there has happily been a major advance in our moral conventions: above all, the principle of equality has triumphed. Such changes 'certainly have not come from religion', he snaps. He instead points to better education about our 'common humanity with members of other races and with the other sex — both deeply unbiblical ideas that come from biological science, especially evolution'. But biological science, especially evolution, can be used to authorise eugenics and racism. The real issue is the triumph of an ideology of equality, of humanism. Instead of asking what this tradition is, and where it comes from, he treats it as axiomatic. This is just the natural human morality, he wants us to think, and in our times we are fortunate to see a particularly full expression of it.
10. Natural law doesn't solve the problem.
As Arthur Leff suggested with his title, "Unspeakable Ethics, Unnatural Law", the natural law tradition is an attempt to ground moral philosophy in objective fact, but it fails to do so.
The best argument I've seen in favor of natural law is James A. Donald's essay, Natural Law and Natural Rights. Donald tries to justify natural law in terms of game theory (evolutionary stable strategy) in a state of nature, the environment of evolutionary adaptedness (EEA). He writes,
Natural law is, or follows from, an ESS [evolutionary stable strategy] for the use of force: Conduct which violates natural law is conduct such that, if a man were to use individual unorganized violence to prevent such conduct, or, in the absence of orderly society, use individual unorganized violence to punish such conduct, then such violence would not indicate that the person using such violence, (violence in accord with natural law) is a danger to a reasonable man. This definition is equivalent to the definition that comes from the game theory of iterated three or more player non zero sum games, applied to evolutionary theory. The idea of law, of actions being lawful or unlawful, has the emotional significance that it does have, because this ESS for the use of force is part of our nature.
An act is a violation of natural law if, were a man to commit such an act in a state of nature, (that is to say, in the absence of an orderly and widely accepted method of resolving disputes), a second man, knowing the facts and being a reasonable man, would reasonably conclude that the first man constituted a threat or danger to the second man, his family, or his property, and if a third man, knowing the facts and being a reasonable man, were to observe the second man getting rid of the first man, the third man would not reasonably conclude that the second man constituted a threat or danger to third man, his family, or his property.
I need to put this into my own words in order to clarify my response to it:
You can find my full response here.
There are two sets of problems with this defense of natural law: normative and positive. The positive problems are that we can't be sure that our intuitions are a reliable guide to practical consequences of our and others' behavior. The normative problems are that we don't agree on the desirability of the various possible outcomes. The positive problems involve a denial of the amount of uncertainty in consequentialist arguments. The normative problems involve opinions and tastes being portrayed as objective facts about "reasonable" men.
On the positive side, Donald presents (1) the practical consequences of actions that violate natural law, (2) social norms, and (3) human psychology as being tightly coupled, which in my opinion they are not. Earlier, I mentioned Adam Smith's The Theory of Moral Sentiments (TMS). My introduction to this book was through a series of Russ Roberts' EconTalk podcasts with Daniel Klein. There was a particular phrase from the book that kept coming up, which is apropos here. Yes, there are connections between practical consequences, social norms, and psychology, but as Smith put it, these relationships are "loose, vague, and indeterminate". If the purpose of natural law is to enable libertarians to argue that their feelings about the use of force are a reliable guide to the long term interests of society, then the libertarians are in for a frustrating time when they try to present this argument to the progressives. It puts libertarians in the position of saying, "My feelings are a trustworthy guide to the long term interests of society, but your feelings are not."
Also on the positive side, in terms of game theory, Donald's framing of the problem is underdetermined. It allows too many potential solutions.
On the normative side, people don't have innate morals so much as we have an innate capacity for building moral systems. Mother Nature didn't provide us with One True Path through the psychological wilderness. She gave us psychological map-making tools.
Also on the normative side, there are the conflicts of interest, deceit, and self-deception. Jonathan Haidt is very good here. Social norms and moral sentiments are largely mechanisms for managing conflicts of interest. Haidt emphasizes our capacity for "motivated reasoning". The human moral sense, the "righteous mind" that Haidt writes of, is in large part a weapon rather than a truth-seeking instrument. I view moral education as being analogous to a gun safety class.
"Right" behavior is the sort of behavior that would make me happy to have you as a member of my club, i.e. allow you to wander my neighborhood unsupervised, have the right to vote, or marry my daughter. It's highly subjective, and culture plays a huge role in influencing these sorts of judgments.
Margaret Thatcher spoke of working "with the grain of human nature", but it's more complicated than that. As Adam Smith explained in TMS, we don't literally feel one another's pain; we have to imagine it. Moral education consists largely of training people to use their imaginations in certain ways. This is affected by social norms and by various encounters we have in our day-to-day lives.
So "the grain of human nature" is only part of the story of how people's feelings are determined. It depends on how we train our imaginations. What we call "natural law" is better understood as English culture from 200 years ago.
Robert Anton Wilson also has a humorous but long-winded critique of natural law (via David Throop) called Don't Put a Rubber on Your Willy. This addresses similar issues to those in the Leff articles and Handle's comment-33086. Wilson makes a point of distinguishing between real-world horseshit, which is subject to scientific investigation, and "ideal Platonic horseshit", which is not.
I'm going to give Arthur Leff's Devil the final word on attempts to derive normative principles from purely objective facts about human nature, having given up on Deism:
But while you try to live as best you can until His revelation, perhaps you will accept some practical advice from Me. Look around you at your species, throughout time and all over the world, and see what men seem to be like. Okay? Now take this hint from what you have seen: If He exists, Me too.
Things people care about
11. "Morality" is an overloaded word.
Earlier, I defined a "moral argument" as one that entails emotional gratification. This is an oversimplification.
Here's how Jonathan Haidt defines a "moral system" in (The Righteous Mind: Why Good People Are Divided by Politics and Religion, p. 270):
Moral systems are interlocking sets of values, virtues, norms, practices, identities, institutions, technologies, and evolved psychological mechanisms that work together to suppress or regulate self-interest and make cooperative societies possible.
Words like "moral" and "immoral" can be extremely vague and are used in many different ways. For example, I can think of 19 different ways to interpret the statement, "Polygamy is immoral." But here I'm only interested in the art of persuasion, applied to voters in a democracy, so we can narrow things down quite a bit.
A better definition for present purposes might be that a moral argument is one that tells the listener that if he behaves in accordance with my doctrine, then he deserves to be respected for promoting a better society. But respected by whom? By reasonable people in general, but mainly by himself.
As Adam Smith wrote regarding human motivation in The Theory of Moral Sentiments (pt. 3, ch. 3), the most important relationship one has is one's relationship with one's self:
It is not the love of our neighbour, it is not the love of mankind, which upon many occasions prompts us to the practice of those divine virtues. It is a stronger love, a more powerful affection, which generally takes place upon such occasions; the love of what is honourable and noble, of the grandeur, and dignity, and superiority of our own characters.
Again, there's both a positive and a normative side to this. I want to convince my listeners
that my methods produce the results I claim for them (positive claims), and
that the results are something that they like, i.e. something that they will feel that they deserve to be respected for (normative claims).
The positive claims should never be taken for granted, but here I'm focusing on the normative arguments. What would make me feel like I deserved respect for embracing your doctrine? There are three answers:
I personally have strong feelings about some process or outcome. I don't care what other people say, I think what you're doing is cool. The key question is, "How do I feel?"
I don't necessarily personally feel all that strongly about it, but I respect the beliefs and opinions of the other people in my community. I will obey my community's social norms. The key question is, "What are the current social norms in my community?"
This can get complicated because I may be a member of many different overlapping communities. For example, I may be a member of a small group of libertarians living in a conservative region in a socialist country. Which of these conflicting sets of social norms has jurisdiction?
I have thought about the prevailing social norms in my community, and I would like to see them changed. This crosses the boundary between normative and positive. It is both. I could want the social norms changed because I have strong feelings about a particular subject, which differ from most other people's, or because I believe that other people are mistaken about the long term consequences of the prevailing social norms, even though our feelings about what long term consequences are desirable are similar. The key question is, "What social norms do I think are desirable?"
Regarding the third question, you may well ask, "Desirable for whom?" My answer is, for me, the people whose respect I care about (my community), and our descendents for the forseeable future. I recommend David Friedman's paper, "What Does Optimal Population Mean?" This may not be exactly utilitarianism, but it is close enough to be confusing. The future orientation also implies a Rawlsian "veil of ignorance".
While there are more ways than these to interpret the word, "morality", I think these are the big three. If you ask me whether I think something is morally permissible, I have to think about the context of the question, and then ask myself one or more of these three questions. "How do I feel about it?", "What are the social norms in my community?" and "What do I think is desirable for 'my people' in the long term?"
12. Most moral arguments beg the question (petitio principii).
A quick look at Wikipedia makes me think I may be using the word, "deontology", incorrectly. "Moral absolutism" is closer to what I mean. From the Wikipedia article: "Some deontologists are moral absolutists, believing that certain actions are absolutely right or wrong, regardless of the intentions behind them as well as the consequences." What I'm trying to get at is that there is a style of moral philosophy that approaches morality the way a mathematician approaches Euclidian geometry. It asks the reader to accept a set of axioms, and proceeds to prove a series of theorems and corollaries which the reader is then expected to accept. For lack of an adequate understanding of the philosophical terms of art, I'm going to call this the "logic-chopping style". It doesn't seem to be peculiar to deontology.
Most libertarian moral arguments seem to fall into this logic-chopping category. The good news: stating axioms up front makes it fast and easy for people to figure out that your arguments aren't cogent for them. The bad news: your arguments aren't cogent.
Consider the three interpretations of "morality" in section 11, the three types of appeal. Basically, I have to plan on making all three of these appeals, because I don't know which one is going to be most effective, and every listener is different. Libertarian moral arguments typically don't do this.
Earlier I wrote that "deontology" tends toward begging the question. It's possible that you can point out to your listener certain connections or contraditions between certain rules that he was not aware of. You might be able to persuade him that the rule you want him to embrace is a non-obvious special case of a more general rule that he already embraces. But it's far more likely that he is aware of the connection, but sees the rules as working in parallel in different domains, rather than one being subordinate to the other. As the saying goes, "Different tools for different jobs." Even if he didn't see the connection before, he is likely to reach that conclusion now.
With libertarians, the general rule is typically the non-aggression principle (where "aggression" includes taxation). Radical libertarians often try to present the non-aggression principle as sacrosanct, the root principle of all morality, to be followed as if everyone were operating under Divine command theory. Everyone else regards non-aggression as a heuristic, a tentative rule that is usually good enough for our daily interactions with neighbors, but by no means a reliable ultimate principle. Even gradualist anarchists like David Friedman and Eric Raymond acknowledge that the non-aggression principle is unworkable as a moral absolute under present circumstances.
I am reminded of what Bernard Crick wrote about what he calls "student's politics" (In Defence of Politics, 4th ed., p.125-6):
There is almost nothing that can do less harm or good to man or beast, or which has less political power, than students' politics....
The voice of student politics in Britain, for instance, says 'Ban the Bomb', with fantastic disregard of a level of scientific and technical knowledge in nuclear physics which makes 'banning' about as sensible as urging the abolition of all hemp, steel, and electricity in order to make capital punishment impossible. The problem of the military use of atomic power is a problem of control, not all or nothing. But control is a complicated political problem — so abandon politics and 'Ban the Bomb', indeed unilaterally, just to make no mistake that one is not interested in politics or diplomacy in any form.
If I demand that my listener adhere rigidly to a principle that we both know is often unworkable, that seems like a pretty good indication that I'm not really interested in being taken seriously. At least, I'm not interested in the intellectual content being taken seriously. It's possible that my purpose is to administer an emotional "carrot and stick", and that what I intended was for the intellectual content to be tossed aside and for the emotional content to be taken to heart. I call this the "nice doggy" style of moral philosophy. But as far as the intellectual content is concerned, an appeal to the non-aggression principle as a moral absolute is almost certainly begging the question.
I associate this "nice doggy" style of moral pleading more with Christianity than with libertarianism. Christians are much more concerned than libertarians with getting people to be generous with one another; libertarians and especially Objectivists tend to be focused on the passive aspects of "right relationship" (stopping abuse), where Christians focus on the active parts (generosity). These calls for generosity tend to be open-ended. It seems to me that there is a tendency for preachers to ask for 1000 times what they think is reasonable, and hope for 0.1% of what they ask for. This might make sense in terms of a menu of options for an exchange: for a $10 donation you get 1 social status point, for $100 you get 10 points, and for $1000 you get 100 points.
But in order for this exchange to work, the preacher has to have the moral authority to award status points. That could mean several things. It could simply mean that the target of the appeal likes the preacher. It could mean the target considers the preacher an expert on that community's social norms. It also could mean that the target thinks that other people in his community respect the preacher enough to accept the preacher's judgments about who deserves high social status.
In so far as these "nice doggy" appeals can be said to be arguments, as opposed to solicitations, they are appeals to authority. As such, they have their place. For example, a Catholic bishop is a recognized authority among Catholics. But this has no relevance to a libertarian trying to persuade a progressive.
13. Moral persuasion is more about building emotional bridges than destroying intellectual bunkers.
Again, I can't take positive arguments for granted, but I'm focusing on normative arguments here. The question is, how do I pull on someone's heart strings? The way Nicholas Phillipson put it in a discussion of The Theory of Moral Sentiments (TMS),
Exchange when it's discussed in TMS is about trading -- we trade sentiments with each other, looking for a sort of psychological deal; it's nice to have a discussion with people and to feel that what we are saying is regarded with sympathy. We relish the process of trading our ideas. It's what happens in any tutorial at any university, any conversation in a park. The process he's describing in TMS and that regulates our social lives is exactly the same process which he is discussing when we trade our goods and services. He says: We spend all our lives trying to persuade people. That's why rhetoric matters.
Thich Nhat Hanh writes, in Living Buddha, Living Christ,
In a true dialogue, both sides are willing to change. We have to appreciate that truth can be received from outside of — not only within — our own group. If we do not believe that, entering into dialogue would be a waste of time. If we think we monopolize the truth and we still organize a dialogue, it is not authentic. We have to believe that by engaging in dialogue with the other person, we have the possibility of making a change within ourselves, that we can become deeper.
The way to get someone to sympathize with your feelings is to offer him your sympathy for his. You have to understand his concerns and address them. If, instead, you try to tell him that his feelings are invalid, you're likely to harden his position rather than soften it.
Again, I don't want to dismiss deontological reasoning entirely. The nature of social norms is that they are rules. They have to be limited in number so that ordinary people can know all the important ones. But the number of circumstances in which we have to make decisions is at least astronomical, if not uncountably infinite. So (1) we have to adapt our rules on the fly to fit our circumstances, and (2) if we get in trouble, we have to be able to justify our actions. Deontology makes sense in this context. It is a tool for interpolating and extrapolating existing rules into novel or murky domains.
It can also be used to invite exchanges of sympathy, by pointing out similarities between things we care about and things our listeners care about. But it is dishonest when this pretends to be more than an invitation, when people use it to try to delegitimize their listeners' or their opponents' concerns. It is also inappropriate when people use it to try to force other people to agree to changes in social norms, insisting that the same tool be used for different jobs.
Consequentialism is similarly necessary, but limited. It's necessary because, as Bruce Ramsey put it, "anyone who severs all connection between his political theory and worldly consequences may be dismissed as a fool." Most of the reason people care about rules and behavior is because of their consequences. But you're also a fool if you think you can predict the consequences of your and others' behavior in enough detail to navigate through life without concern for rules and moral intuitions. These intuitions are formed largely by how we use our imaginations, and a good consequential argument binds imagination to reality. But consequential arguments, at their best, rely on experience bases that can be made ambiguously irrelevant by changing circumstances. When arguments about human behavior depend heavily on analysis, with little data to support them, they are extremely unreliable — just-so stories. At their worst, they are little better than fantasy.
Consequentialism is a tool for inviting an exchange of sympathy and for binding imagination to reality. But it's unreliable at doing the latter. Your listeners are rightfully distrustful of it. (It's really easy to cherry-pick social statistics even when you're trying not to.) So when you use a consequential argument, you have to be careful to be honest about uncertainty. You should also bear in mind that if you know more about a topic that your listeners, your listeners' uncertainty ought to be higher than yours (but beware of the Dunning-Kruger effect). Above all, beware of motivated reasoning, both on your listeners' part and on yours.
And even the best consequential argument doesn't relieve you of the need to couch your argument in terms of things your listeners care about. You may have an airtight argument that the policy you advocate is good for Irish Catholics, but you have no right to insist that your Irish Protestant listeners be moved by it.
One important piece of advice that Smith gives in TMS is not to be a drama queen. You may feel strongly about something, but your listeners probably don't, and you have take that into account in asking them to sympathize with you. If you bark your shin, you can ask your listener to imagine how you feel, but his imagination is never as vivid as your nerve endings. You need to moderate how you express your feelings in order to match other people's willingness to sympathize with them.
Another aspect of being mindful of your listeners' feelings is taking into account their desire to save face. If someone has publicly taken a position in a debate, and you are trying to get him to walk it back, you're probably going to have to give him some time. It's not just that he needs time in order to think through the consequences of a novel idea. He also needs time to deal with the loss of face, to quietly disinvest his emotions and reputation from his previous position. You can embarrass an opposing extremist or two as an example to the others, but try not to embarrass the people whose minds you're trying to change.
That last point introduces an element of triage. It may help to divide your opponents into three classes: those who can be persuaded by cold, hard logic, those who can be persuaded by moral arguments, and those who are too invested in their position to be worth trying to persuade at all.
People can be invested in their beliefs in many ways, and it can be hard to talk them away from their investments. You probably can't get a sympathetic hearing from someone whose social life depends on convincing his peers that he hates you. (Faking it on the job is usually easier than faking it in ones' social life.) In order to get a sympathetic hearing, you may have to offer him an entirely new social environment (i.e. a new church, and his interest in a new church probably depends on him already being unhappy with his old church for reasons which have nothing to do with you).
But your emotional investments are probably just as big a problem as your listeners'. Trying to have a meeting of minds (and souls) is hard work. If your real motive is to reassure yourself of your moral or intellectual superiority, you're not going to want to have a genuine meeting of minds. You're more likely to want to delegitimize other people's feelings to satisfy your own emotional needs.
Appendix A: How the pieces of moral philosophy fit together
Someone suggested that representatives of the different schools of moral philosophy were like the men in the parable of the blind men and the elephant. Update: For example, Alrenous writes,
Virtue ethics can be justified consequentially – humans are bad at working out true consequences in real time, and thus need heuristics, which we call virtues.
Here's how I think the different schools fit together:
There is a biological level, described by Edward O. Wilson in a chapter in Conscilience. He has a sidebar about what a State of the Union address would sound like if it were delivered by the queen of a colony of giant, intelligent termites. Cannibalism, yes. Individualism, no. Wilson's morality is biologically based. We have at least some innate predispositions and biases. This is the "evolutionary psychology" I described in The Baby and the Bathwater. Steven Pinker writes on this level, too, in The Blank Slate. Wilson divides moral philosophers into "transcendentalists, who think that moral guidelines exist outside the human mind, and empiricists, who think them contrivances of the mind." He made a comment about Kant (rough quote), "Sometimes an idea is baffling not because it is profound, but because it is wrong." hbd chick deserves mention here, too.
I like Adam Smith's description of the psychological and small-scale social level in The Theory of Moral Sentiments. He doesn't say enough about religion, though. Jonathan Haidt has a broader scope, and scientific credibility, but Smith does detailed introspection better. F. A. Hayek has a section in The Fatal Conceit (chapter 1) that also fits in here. He explicitly describes morality as an emergent phenomenon, the result of often unconscious interactions among large numbers of people.
There is also a game theory and economics level, which overlaps quite a lot with the psychological, but is more consciously purposeful. There has to be a relationship between morality and enlightened self-interest. If you have a moral system that isn't related to anything that I'm interested in, how is it even logically possible for me to be interested in it? A coherent moral theory has to make sense in terms of game theory. But this can be subtle.
Then there is a political/pseudo-scientific/religious level. There is sincere pseudo-science, in which people try to explain hideously complicated emergent phenomena in terms of simple principles of what Robert Anton Wilson called "ideal Platonic horseshit". On the political side, it often manifests itself as Deutungshoheit. Some of this "politics" consists of perfectly reasonable attempts to change or carve out exceptions to social norms that are causing problems. Some of it is status-whoring. But the different strains of sophistry and honest confusion interact with one another, and it often seems hopeless to try to pull them apart.
The position I want to argue is that Smith is the guy who set the overall problem up correctly, who sees into the elephant's soul, and the other guys are providing inputs to him, describing the elephant's limbs. There are a number of problems I see with my position.
One problem is that Smith understates the role of religion. For example, hbd chick argues that the fact that Europe is less tribal than the Middle East is because of the Catholic church's historic policies on consanguinuity. That is, morality may be an emergent phenomenon to a large extent, but it is also driven to a considerable extent by the deliberate policies of religious authorities. Guenter Lewy also reluctantly concluded that religion has a lot to do with the preservation of "moral capital" in America.
Another problem is that Smith understates the role of coercion (i.e. the State). As the saying goes, "When you've got 'em by the balls, their hearts and minds will follow."
Conformism is a third problem: people may be afraid to express their sentiments honestly even when they're not under any significant threat.
But with those qualifications, I think Smith is the guy who got it right.
Let's go back to the question of how to pull apart the strains of sophistry and honest confusion. What's happening is that a bunch of people, including yours truly, are practicing politics and/or religion. Some of this is in bad faith, but most of it is in good faith, however ill-considered. Sometimes social norms change as a side effect of a successful bid for social status or political power (see Alrenous again). Sometimes changes to social norms are the primary objective. But the process of making moral arguments that social norms should be changed means that these political/religious arguments are arguments about what social norms ought to be, not what they are or what people's immediate emotional reactions are. As before, "deontological" axiom-based arguments are generally out of place here. They're almost always blarney, apparently intended not to persuade skeptics but to make choir members feel morally superior to the skeptics. We're left with a mixture of subjective emotional appeals about what is desirable (normative), and consequentialist arguments about how best to acheive some agreed-upon goal (positive). On the normative side, this brings me back to the Memorandum from the Devil, and people pretending for propaganda purposes that their sentiments are objective facts. On the positive side, it brings me back to C. M. Kornbluth's 2 billion-body problem, and lying about uncertainty.
Between the subjectivity of the normative side and the uncertainty about the positive side, these arguments about the desirability of various social norms are necessarily going to be difficult or impossible to verify. Thus morality, understood as conformance with desirable social norms, is a credence good: a good whose quality is difficult or impossible to verify. Arguments about what social norms are desirable will almost never be cogent for skeptical audiences.
But the preeminent institutions for the promotion of credence goods are churches. As Laurence Iannaccone wrote in "Introduction to the Economics of Religion", eternal salvation is 'the ultimate "credence good"'. Promoting credence goods seems to be central to what religion is about. Moral indoctrination is therefore inherently at least quasi-religious. This agrees with observations by other social scientists. As Jonathan Haidt writes in The Righteous Mind, "Religions are moral exoskeletons." But in our "enlightened" modern world, instead of invoking God to huck their credence goods, people with modern ideas invoke "science". Thus we end up with people like Ross Douthat complaining about moral scientism.
Appendix B: Objectivism and moral scientism
I took an Artificial Intelligence class years ago that mentioned three types of reasoning: deductive, inductive, and abductive. The text (pp. 60-63 of The Engineering of Knowledge-Based Systems: Theory and Practice, by Avelino J. Gonzalez and Douglas D. Dankel) described deduction as reasoning (reliably) from the general to the specific. Socrates is a man, all men are mortal, therefore Socrates is mortal. Induction is extrapolating from the specific to the general, from the known into the unknown, which is inherently unreliable. I test one sample of bolts to failure, then I install a hopefully similar bolt in my airplane, but it isn't one of the ones I tested. Abduction is (unreliable) diagnostic reasoning, reasoning backwards from effects to causes. My lawnmower won't start. A clogged carburetor could cause that. Maybe I should try spraying some carburetor cleaner into it.
I had a conversation with an Objectivist friend a while ago on the nature of Bad Philosophy, specifically political thinking. I find Objectivists hard to understand, so maybe I am getting their position wrong, but they seem to be crediting their opponents with honest errors in deductive reasoning. In this view, political philosophy is like an egg. Good philosophy produces true beliefs. Bad philosophy produces honest mistakes, mainly by making false assumptions but still using valid deductive logic. Replace a bad person's damaged egg with a good egg, and you get a good person.
My view is that most of the Bad Philosophy I encounter looks like abduction. I start out with the conclusion I want to reach, which 90% of the time is that my tribe is morally superior to my opponent's tribe. Then I work backwards to find rationalizations to justify it. So I view Bad Philosophy not like an egg, but like a throw rug. A cat drops a turd on the floor, and pushes a small rug over it to cover it up. Arguing with people over their supposed political premises is thus pointless. All you can hope to do is get the cat to choose a different covering. The turd doesn't go away.
I think there is some fundamental dishonesty here. Even if the person counting coup over his moral superiority is merely being sloppy in his reasoning rather than consciously dishonest, he is almost always claiming a higher level of confidence in his beliefs than is consistent with honest sloppiness. So Henry Frankfurt claims in On Bullshit that "bullshit" is distinct from lying (reckless disregard of the truth rather than deliberate falsehood), but I disagree. I regard people who "bullshit" this way as being dishonest. At a minimum, they are lying about uncertainty.
See also Robin Hanson, "Politics isn't about policy."
Appendix C: Scattered related thoughts
This essay started out mostly as a bunch of links to, and short descriptions of, some of my previous writings that had gone off on meta-ethical tangents. Some of this other material may still be of interest.
In The Dog Ate My Manifesto: Reflections on Moderate Libertarianism, I mostly tried to keep it practical, and avoid making moral arguments. In so far as I talked about morality at all, it was in consequentialist terms. I also took a hard-line consequentialist position in The Market for Sanctimony, or why we need Yet Another Space Alien Cult (YASAC).
I first floated the idea of morality as a credence good (a good whose quality is difficult or impossible to verify) in an appendix to Is Rational Religion Possible?. The related question of what a moral doctrine is supposed to accomplish also came up briefly and somewhat more coherently in my review of Guenter Lewy's Why America Needs Religion.
My essay, What does 'morality' mean?, began as a book review of Claire Berlinski's Why Margaret Thatcher Matters, but it was heavily influenced by Adam Smith's The Theory of Moral Sentiments (TMS). There are a bunch of links related to TMS at the beginning of part 1. In part 2, I tried to list all the senses I could think of in which someone could use the word "immoral". I am up to 19. I offer 5 explanations of how morality might relate to enlightened self-interest. Berlinski convinced me that my attempt in The Dog Ate My Manifesto to limit myself to practical arguments was a bad idea.
Much of the material here appeared before in
The Resurrection of Classical Liberalism, mainly in the section on "Making moral arguments cogent". This section was largely a response to James A. Donald's essay on Natural Law. The later sections on "Universalism vs. particularism" and "Moral-practical dichotomies" also partly relate to meta-ethics.
An earlier essay, The Baby and the Bathwater, tries to explain the relationship between morality, religion, and evolutionary psychology. It thus covers some of the same subject matter as Donald's essay, but from a "loose, vague, and indeterminate" viewpoint. Mother Nature seems to work both by group selection and individual selection. She gave us tools to enable us both to cooperate with one another and to selectively refuse to cooperate. We have to balance these tendencies. "God" is a tool for social cooperation. The "rationalization hamster" is a tool for resisting social control.
With the death of God, the Holy Hamster is out of control. We need a new religion. With apologies to Jonathan Haidt, I propose The Church of Glaucon.
I attempt to illustrate the relationships among Mother Nature, God, and the Rationalization Hamster with
with apologies to the makers of openclipart-0.18
(and envious glances at Sunshine Mary)
Sorry, no ferrets.
I suggest in Designing the Church of Glaucon that church polity is important, and tentatively that moral philosophy should be left mainly to professionals.
I argue in my review of Michael Strong's book, Be The Solution, that there are "good" and "evil" religions, which differ largely because of conflicts of interest over the suppression of self-deception.
In defense of heaving dead cats is about the art of persuasion in general. Part of this has to do with constructing cogent moral arguments, but mostly it is about Deutungshoheit and the nature of verbal abuse in political arguments (especially sections 4, 5, 6, and 8).