Thursday, December 13, 2012

Why Jerry Coyne is Wrong (about Ethics)

Every few days I find something on the internet to blog about. I start typing, and then I get busy and give up on it. I will try to do better with today’s entry: Why Jerry Coyne is Wrong (about Ethics). Catchy, huh?

So, I like the stuff Coyne posts, and mostly I agree with it. I wish that I were in a position that I could make my colleagues take some responsibility for their friendly, no-strings-attached, totally-not-supporting-right-wing-crazies, and they-never-tell-me-what-to-conclude-even-though-strangely-all-the-many-opinions-I-have-that-directly-contradict-everything-they-believe-never-see-print Templeton money. Thus, I say with respect, Coyne is here totally wrong about objectivity and truth in non-scientific contexts.

Let’s begin.
Coyne is respectfully disagreeing about scientism, the view that science is the only way to acquire objective knowledge of external reality. His friend Eric McDonald disagrees and argues for the existence of objectivity or truth in other areas besides science.
Coyne writes about McDonald's post (forgive the long quotation, but I want to be fair),
His post, “On the strangely beguiling notion of scientism,” takes the stand that there are indeed ways to apprehend objective truth beyond the purview of science, and that those who claim otherwise are guilty of scientism.
We still disagree about this. I’m sorry to say that Eric’s piece, like nearly all pieces on scientism, fails to make a case for (or even give more than one example of) “truth” apprehended by other than scientific means—and I’m defining “science” as the combination of empirical observation, reason, and (usually) replicated observation and prediction that investigates what exists in the universe.
I’ll be brief here, as I’ve posted a lot on this topic lately, but I want to discuss what Eric sees as “objective knowledge” that goes beyond science.
It’s “moral knowledge”:
And though Jerry Coyne (this is one of the small number of areas where he and I differ significantly in our approach to things) may dismiss ideas concerning value as matters of opinion, it is very doubtful that girls in Afghanistan, who have acid thrown in their faces or see their schools being destroyed, share that view. It is not just a matter of opinion that their right to learn should be recognized and honoured; how we establish what can justly be considered objective moral understanding is something worthwhile considering.
. . .
Now I agree, of course, that throwing acid in the face of Afghan schoolgirls for trying to learn is wrong. But it is not an “objective” moral wrong—that is, you cannot deduce it from mere observation, not without adding some reasons why you think it’s wrong. And those reasons are based on opinions. In this case, the “opinion” is that it’s wrong to hurt anyone for trying to go to school. In other words, Eric claims that moral dicta are objective ones, on the par with the “knowledge” of science.
But such dicta are not “truths,” but “guides for living”. And some people, like the odious Taliban who perpetuate these crimes, do disagree. How do you prove, objectively, that they’re wrong? You need to bring in other subjective criteria.

The problem with “objective” moral truths is much clearer in less clear-cut cases. Is it objectively true that abortion is wrong, or that a moral society must give everyone health care? You can’t ascertain these “truths” by observation; you deduce them from some general principles of right and wrong that are, at bottom, opinion. (Of course, some opinions are more well-founded than others, and that’s what philosophy is good for.)

What precisely is Coyne’s position and reasoning? As far as I can tell, he agrees that certain actions (throwing acid in an Afghan girl’s face for going to school) and people (the Taliban) are morally wrong, but that his agreement or disagreement with McDonald is only a matter of their sharing certain subjective attitudes towards those actions and individuals. What reasons does Coyne offer for his opinion about morality?

Some people (e.g. the Taliban) disagree about the morality of these actions.
Therefore, there is no fact about the morality of these actions.

This is a strange argument coming from a leading voice in the debate with creationists. Clearly those odious individuals disagree about science, but we would not conclude that there is no objective truth in science but only opinion. Clearly, there is more going on. The second point:

“How do you prove, objectively, that they’re wrong?”

So, it is not the mere fact of disagreement that troubles Coyne, it is that he can find no common ground that one could use to prove to the satisfaction of the Taliban that their actions are wrong.

Still, I don’t think this helps much since Coyne himself has shown how difficult it is to find common ground with the creationists in order to convince them that they are wrong. We need not convince the Taliban that they are wrong, just as we need not convince the creationists that they are wrong, in order for there to be objective facts for them to be wrong about.

Coyne’s appeal to problem cases is a red herring. If one were to judge the objectivity of science only by appeal to its most problematic cases (e.g. string theory), one would similarly come to the conclusion that science also is not objective. The fact that there are cases in which we do not have answers is irrelevant as well since every living discipline has problem cases about which more research is needed.

Thus, what we need is a set of procedures or a method that can be used to show objectively that one is right. That is, we need something akin to the scientific method which relies not on the intuitive judgments or subjective opinions of the arguer in order to establish the claims. For example, we would not like to be caught in a debate on the ethics of torture only to have each side bellow at the other that their opinions are right and the other side’s are obviously wrong, as judged by some direct perception or intuition. In science, that independent arbiter is experiment, double-blinded and peer-reviewed, to show that the conclusion does not depend on anything purely internal to an agent.

Indeed, this is a challenge of moral epistemology, to find such an independently verifiable means of testing ethical claims. However, now Coyne has changed the subject from one of the reality or objectivity of ethical claims to our knowledge of those ethical claims. We may lack a clear method for adjudicating disputes about ethics, but it in no way follows that there is no objectivity or truth about ethics. We do not know, for example, whether there is intelligent extraterrestrial life elsewhere in the universe (Let’s hope so because “there’s bugger all down here on earth.”), but it does not follow from this lack of knowledge that there is no fact about the existence of such life. The fact that we do not know, that we may never be in a position to know, does not prove that there is no objectivity or truth.

Still, Coyne says, we need a method. The difference, he would no doubt claim, is that while we do not now know the answer to this question about extraterrestrial life, we do know how we could go about looking for it, and we would know what sorts of things would count as evidence for or against its existence. So, the scientist is apparently one up on the ethicist in having some idea what those methods are, but this is still irrelevant because the existence of standards for verifying or falsifying claims tells us nothing about reality. Epistemology and metaphysics, knowledge and reality, are distinct, and we cannot conclude from our inability (perhaps even a permanent inability) to formulate objective standards for knowing morality that there is no objective moral reality at all. The “How do you know?” question is a good one to ask, but it tells us nothing about the existence of objective reality.

But perhaps we can see what some methods in philosophy and ethics might be. Before that, one more quotation:

In other words, Eric is committing here the very sin he decried (as I recall) in Sam Harris’s book The Moral Landscape: he is saying that there are scientifically establishable truths about ethics. And if that’s true, then let Eric tell us what those truths are—without first defining, based on his taste, what is “moral” and “immoral.” Let him give us a list of all the behaviors he considers objectively immoral.

I don’t know McDonald’s critique of Harris, but I included this to be funny. Please tell us, Dr. Coyne what are all the scientifically establishable truths about science? If there are such truths, Coyne should be able to “give us a list of all” of the ones he considers to be true and all those he considers to be false. Allow me to reiterate, we do not need to know what the truths are in order to know that there are truths. If we did, then science (indeed investigation of any kind) would be pointless. Are there objective truths to discover about the world? We don’t know until we can first say what they are, but, having stated them, we now find ourselves with no need to attempt to discover them.

Wait, sorry. I think what Coyne was trying to say was that McDonald should be able to state in principle what sorts of claims are “scientifically establishable” in ethics. Obviously, if these claims were scientifically establishable, they would not be non-scientific claims, as McDonald says they are, so Coyne appears to be demanding that McDonald explain which non-scientifically establishable claims are scientifically establishable. I have rarely seen a question so begged.

Still, I cannot help feeling dirty, what with the quoting and all, so I will get back to the epistemic issue. Perhaps Coyne would be happy enough with moral skepticism. He might be better off saying that there are objective truths in ethics but no one is ever in a position to say what they are given the lack of intersubjective standards of verification. A reminder:

You can’t ascertain these “truths” by observation; you deduce them from some general principles of right and wrong that are, at bottom, opinion. (Of course, some opinions are more well-founded than others, and that’s what philosophy is good for.)

Should we be skeptical? It depends on what you mean by 'skeptical', of course. At any rate, we need not think that no ethical claim is any better justified than any other. We need not be extreme or Pyrrhonic skeptics. For example, as Coyne notes, philosophy is good for establishing ‘opinions’ as better founded than others. We should, of course, always be cautious and recognize the potential for fallibility in our method. And here’s where we get to the methods of philosophy and ethics. We use reason and argument based on some general principles, perhaps, but often these are based on intuitive judgments of individual cases.

Isn’t this just more subjectivity and opinion? No. How do we know that modus ponens is a valid form of argument? Modus ponens is the following argument form: If p, then q; p; therefore q. This argument form guarantees that true premises will result in a true conclusion, and our judgment of this is not based merely on popular opinion. It is based not on popularity, but on careful consideration of the concept of the conditional.

Psychologists have studied conditional reasoning using the Wason selection task and have shown that undergraduates are more likely to view the fallacy of affirming the consequent as valid than they are to view the valid argument form of modus tollens as valid. Affirming the consequent goes like this: If p then q; q; therefore p. Something like 70% of subjects apparently (given the structure of the selection task) believe that this is valid. Yet we can see that it is not. Here is a parallel argument. If you are driving legally, you must be 16 or older (in America without a learner’s permit etc.). Suppose you are 16 or older. Does it follow that you are driving legally? Obviously not, you might not have a driver’s license; you might be drunk; you might not be wearing your prescription eyewear.

On the other hand, modus tollens (considered by fewer than 10% of the subjects to be valid) is valid. Modus tollens goes like this: If p, then q; not q; therefore not-p. And we can see that this is valid by considering that the conditional indicates a necessary condition (that is, a condition without which the antecedent would not occur). So, using our previous example, we know that being 16 or older is a necessary condition for driving legally. It follows from this that if you are driving legally, you must be 16 or older. Now, consider what happens when the necessary condition is not met. In that case, the possible fact it is necessary for, the driving legally, cannot obtain. In short, supposing you are not over 16, you cannot drive legally.

This reasoning is not difficult to follow; it is familiar to everyone. I do not know how to characterize it. Is it deduction from general principles? Well, there clearly are general principles involved, but the role of the examples is essential in getting people to recognize those principles. In any event, it is not just opinion. Should I use all caps? If Coyne wants to tell us that these principles are just opinions, then he must be willing to agree that other people might have different opinions about basic logic, and so there is no possibility of arguing with, reasoning about, anything at all.

Moreover, consider what Coyne is claiming here. Ethics is merely a matter of opinion. What is the basis of this view? Is it also just opinion? If so, it is not my opinion, so I am perfectly within my rights to reject it. Is it society’s opinion? Not likely. We have consistent public opinions that rely on belief in objective moral fact. We don’t just outlaw murder because we do not prefer it; we outlaw murder because we think it is wrong (because of harm [another moral concept] to innocents [still yet another moral concept] etc.). So, it is not society’s opinion that ethics is mere opinion. Perhaps it is the opinion of the community of experts treated as knowledgeable (!) about such matters? If so, again ethics is not considered mere opinion; if one polls philosophers and ethicists one will find they believe ethics to be objective. So, if Coyne is correct that ethics is a matter of opinion, he must provide some argument for that conclusion that is based on something other than his own opinion.

But still, perhaps we can establish that science is the best way of discovering the truth about reality. If so, everything else would be second-rate. Such an argument would not show that ethics does not traffic in objectivity and truth, but it would puncture the ethicist’s pretensions. Actually, probably not. Most ethicists and philosophers are well aware, better than anyone else, of the limits of their methods. Still, can we establish that science is the best way of discovering objective truth about reality?

Coyne is aware of this argument since he quotes McDonald making exactly this point:

I want to differ with Eric on one other point: his claim that there’s no way to show a priori that science provides truth about reality. (Well, I agree with him in principle, but think it’s completely irrelevant as a criticism of science.):

[There follow quotations from various philosophers noting that the superiority of the scientific method is not itself a scientific question.]

Eric should be careful here, because he’s beginning to tread the road paved by people like Alvin Plantinga—theologians who try to drag science down to the level of faith because science can’t justify logically that it can finds truth.
My answer to this claim is this: “so fricking what?” While philosophers draw their pay by arguing interminably about such stuff (and achieving nothing by so doing), science goes ahead and accomplishes things: we find out what causes disease and then find cures; we put people on the Moon; we build computers and lasers. In other words, by assuming that there are external truths that are apprehended by science, we accomplish what we want to do, including alleviating suffering that no faith-healing could ever relieve. The tuberculosis bacterium is not an illusion. I don’t give a rat’s patootie for the philosophers who tell us that we can’t justify science’s ability to find truth by a priori lucubration. Let them squabble while science moves on. The success of science justifies its assumption of objective truths and its program for apprehending them.

Ahoy, mateys, it’s the dangerous and elusive (well, not really) slippery slope. If we think that science is not capable of justifying itself, then we will soon be like Alvin Plantinga denying evolution, rejecting science, and whatnot; it will be dogs and cats living together, the end of the world. But, as to his substantive answer, “so fricking what?” The what is that one cannot justify science (or induction, in Hume’s terms) in scientific terms on pain of circularity. We could appeal to the success of science on pragmatic grounds (and I would as well), but the point is that this appeal is itself not a kind of scientific knowledge. If it were, the justification would be circular and hence not really a justification at all. The point is that apologists for science must rely on some reason other than science to justify use of science. No one (except the aforementioned Plantingeans) wants to take your science away. At least no one around here does. I have no dispute with this pragmatic justification of science; I am as big a fan as you will find of science and the scientific method. This is not to say that we should not do science, or that science is not effective. But that’s not the issue. The point is, again, that even those who advocate for the superiority of science must base that argument on some non-scientific reasoning. That reasoning is, in this case, reasonable, but it is also absolutely essential if one is to justify science. Thus, there must, if science is to be trusted, be objective truth and knowledge of non-scientific matters (here, the epistemology of science).

One more attempt: Anyone who uses reason and argument to critique philosophy and philosophical method must, in the end, just be engaging in more philosophy. That’s not an attempt to reject science. Neither is it an apology for religion or superstition. It’s a simple recognition of the need for reason and argument that is not purely scientific or observational.

Let’s consider one last possibility. Coyne analogizes morality to religion, and, God forbid, I wouldn’t want to do that. Still, there might be an objection that the intuitive judgments about ethics differ from intuitive judgments about philosophy (and these self-referential arguments about morality and science) and logic. There are, of course, many poor arguments in ethics, just as there are poor papers in science journals, but at bottom there are principles of ethics that seem no less certain than the principles we rely on in these arguments for objective philosophical truth (such as, for example, the pragmatic argument for the value of science).

One such principle is the principle of equality: We should treat everyone in the same way provided there is no relevant difference between them.

I know that Coyne would love this example since it makes use of vague language and requires interpretation to apply. What is it to treat people the same way? Should I give everyone in the universe a size extra large Santa sweatshirt because I am buying one for my brother for Christmas [sorry, dude]? Clearly not. The ‘relevance’ clause also involves considerable potential for abuse. Are African-Americans or women relevantly different from white men so that I am justified in refusing to hire them? Presumably not. However, the vagueness and interpretation are necessary for the statement to have the status of truth. One should resist the temptation to critique the equality principle on the grounds of vagueness or interpretability because such a critique would be implicitly accepting the principle and only disagreeing about its application. There are not likely to be universally correct moral claims such as “Do not kill”. Correct principles will have to be abstract and general, such as the equality principle, in order for them to be true. Does 1 + 1 = 2? Obviously, but if you take one water droplet and add it to another water droplet, you don’t have two water droplets, you just have one bigger water droplet. Does this silly example show that our fundamental mathematical understanding is incorrect? No, we just have to recognize the limits of application of universally true principles. Readers are free to come up with counterexamples to the equality principle, but I would bet that all of them are questions of application, not questions of the truth of the claim.

So, in the end, philosophy is unavoidable no matter how we try. ("Oh, Lord, how I did try!") And ethics is as reliable a part of philosophy as any other (or approximately so). It’s not anti-scientific to believe this, and it’s not superstitious or religious either. There are, as they say, more things [objective truths, that is] in heaven and earth, [Dr. Coyne], than are dreamt of in your philosophy.

I apologize if I have run on at too great a length, but for one brief afternoon I have time to do what I want. So, here it is. I hope that it was at least worth the effort.


  1. Thank you for this (another perspective. I have been following the discussion between Coyne and MacDonald and there is another MacDonald post you may not have seen, "Picking up the gauntlet, taking to the field!": See especially the comments about poetry, literature and music.

  2. Hmm. I feel like one of those compatibilist discussions where there's a compatibility claim on one scale that does not hold up to scrutiny on another scale (where it might be hard determinism).

    I think with the principle of equality you run fast into compartmentalisation we're trying to (or should be trying to) avoid, from sexism, abelism, racism and other isms of much drama. We tend to not react to isms that are more benign to most people (like existentialism), but the problem of semantic (or ontological) scale (which I think is at the heart of the principle of equality) is still there, causing problems for us by passing under our radar of normal discourse.

    As to objective truths, well, the jury is still out on whether they exist or not. I'm inclined to speak of truths in terms of probability of them existing, making this sentence at least 70% true.


  3. Thanks for the link to MacDonald's response to Coyne. He tries the more constructive approach of the existence proof, showing cases of non-scientific knowledge. I tend not to do this since it's possible, for whatever case you give, to doubt that it's knowledge. Also, he's a lot nicer than I am.

    MacDonald gives a lot of plausible examples other than just the principle of equality. The relevance criterion makes the principle appear to be a priori, but we will need some empirical observation to determine what qualifies as a relevant difference. I suppose that a logical positivist might claim that I've just constructed a definitional truth. Maybe the best examples are logical inference patterns (as MacDonald mentions) necessary for scientific theory disconfirmation.