A Universal Moral Code?

Might there be a universal moral code? When we look around, we everywhere find bitter and seemingly interminable moral disagreements about abortion, or euthanasia, or animal rights, or social justice, and many other issues, not to mention the vast gulfs that separate the moral outlooks of different cultures. The idea that there is a universal moral code can thus sound farfetched. Yet the Harvard psychologists Marc Hauser, and several other scientists, have recently claimed that, contrary to appearances, there really is a universal moral code, and that this scientific discovery should change the way we think about ethics (see here, and here for a longer piece by John Mikhail. Hauser’s views are spelled out at length in his book Moral Minds.). Is this really so?

Let me just very quickly review the evidence for this claim. It is basically that when people fill out an internet questionnaire about a range of moral dilemmas about killing others, their patterns of response are very similar, regardless of nationality, religion, age or gender. Moreover, people are often not aware of the principles that guide their responses to these dilemmas, suggesting that these are not principles they have explicitly learned. This interesting finding has been claimed to show that there is an innate, universal moral code that is shared by all humans. Without going into the details, I think it’s clear enough that the current evidence is at best suggestive (the evidence that we are born with natural moral motivations is much stronger — see here). But let us assume that Hauser is right about this. What would follow?

Clearly, when Hauser speaks about a universal moral code, he doesn’t mean that all humans accept and follow the same set of moral norms. That’s patently false. He means that we are all born with the disposition to respond to certain moral situation in structurally similar ways, even though different cultures and epochs fill out the actual details very differently (the idea is based on Chomsky’s view that all human languages share an underlying grammatical structure). That’s not in itself very helpful. It doesn’t make our moral disagreements with others less deep—and it doesn’t tell us which way (if any) of filling out the details is the correct one. (Perhaps all it tells us is that if we encountered aliens who do not share this innate human structure, we would really have a hard time agreeing with them on any moral question.)

Hauser sometimes writes as if there really is a specific moral code we all born with, a moral code that is completely ‘impartial’ and ‘dispassionate’, but which our emotions and various pernicious social influences distort, leading to our current sorry situation. If we could only overcome these pernicious influences, we would be well on the way to having a universal moral code in the literal sense—a set of moral norms and values we all share.

This is an odd idea. Of course if we were all designed by a wise and benevolent deity, then it would make sense to find out what moral code is inscribed in our brains, and to then follow it. But we are products of blind natural selection. Why on earth should we think that the moral code that was selected by evolution is in any valid? (Hauser admits that is innate code might be a little out of date, but that’s not really the issue. See this paper of mine for more on evolution and morality.)

A universal moral code might be a set of underlying dispositions we are all born with. Or it might be a set of explicit norms and values humans might one day universally accept. But a more important sense of ‘universal moral code’ is of a set of moral values that is universally valid, whether or not it is inscribed in our brains, or accepted by people. Of course that is a very controversial idea. If there is such a universal moral code, then we have an imperative to try to discover it, and to make it universally accepted (to make it a moral code in the descriptive sense). But this requires thinking hard about ethics, not looking for some code that might or might not be written into our brains.

  • Facebook
  • Twitter
  • Reddit

5 Responses to A Universal Moral Code?

  • Peter Wicks says:

    I would like to suggest a different take on this issue.

    My starting point is that the project of finding a set of moral values that is universally valid, in the sense of being based on pure logic, is futile. My basic reason for this is that logical systems proceed from premises to conclusions. Without premise, there can be no conclusion.

    By contrast, if it could be demonstrated that (by whatever fluke of evolution) people do actually share certain dispositions to react to moral situations in certain ways, then this could provide the input required to construct a complete, or at least almost complete, and logically consistent system of ethics. (I say “almost” because of Gödel’s theorem.) This could be tremendously useful as a tool to help build consensus on ethical issues.

    Indeed, this is more or less what ethical and indeed legal systems try to do: draw deductive conclusions on the details based on accepted general principles set out in documents such as national constitutions or the Universal Declaration of Human Rights. It seems to me to be extremely valuable to study the extent to which the principles enshrined in such documents (or treatises such as Rawl’s Theory of Justice) actually correspond to empirically measured dispositions, and to which such dispositions do or don’t vary across cultures, genders etc.

    We would still need to decide whether or not to commit to such a system (see my previous comments on morality as choice), but this seems to me to be a far more meaningful philosophical project than searching for the holy grail of universal and irrefutable morals.

  • Dennis Tuchler says:

    If there is a hard-wired moral code (or ethics) it is probably made of a set of broad principles about killing, stealing and so forth, relating to survival in an economy founded on exchange. More likely than not, you can find broad ideas about what counts as justified killing or taking of property, what counts as property, etc. all over the world. It is not hard to expect some sort of social advantage in having such broad ideas wired-in and thus a reason to expect that the ones with it survived more than the ones without. But there is an objection that the moral code or ethics might not be “valid” (“Why on earth should we think that the moral code that was selected by evolution is in any valid?”)

    In what way is a moral code “(in)valid”? A working moral code (or ethics) is the product of evolution — of social systems. If you look at ethics as a way of determining right conduct in all social situations, then it seems there should be more than one possible acceptable ethics (or moral code). The one that is in effect is probably a result of the climate, population size, economy, information flows, etc. and changes as those factors change. I don’t understand how any one of all possible ethics can be valid, unless there is a perfect moral code or a proper way of acquiring an ethics. I suppose a coerced ethics or moral code is “invalid”, but I doubt that such an ethics can survive actual practice in the affected social system for long.

  • Peter Wicks says:

    For me the interesting question is not whether a moral code is “(in)valid” but whether it is something that humanity may at some point in the reasonably near future (i.e. before we all kill ourselves through strife, irresponsibility or neglect) be able to agree on.

    Which is not to say that I don’t have preferences of my own about what might (should) emerge in such a code, and even red lines beyond which I would consider no code to be preferable, but this is “academic” (in the colloquial sense of “irrelevant in practice”) to the extent that any code that everyone (or let’s say a workable majority) could agree on would almost certainly be one that I myself would consider acceptable and an improvement on the current situation (of confusion and dangerous polarisations).

    Which is why I think, unlike Guy, that studies such as the one referred to in this post are immensely important.

  • Guy Kahane says:

    Thanks for these interesting comments.

    My main point was that we must be careful to distinguish three distinct senses of ‘universal moral code’: (1) a set of moral norms that is universally agreed, (2) a set of formal moral structures we are innately disposed to follow, and (3) a set of moral norms that is universally valid.

    I anticipated that the idea of universally valid norms would generate some scepticism, but I could not say more about it in what is already a long post.

    Peter Wicks suggests we forget about (3), and focus on (1), on getting all (or even most) people to agree on basic moral questions. I agree that this will be a good thing: it would greatly reduce conflict and violence. Of course to say that this would be a good thing is already implicitly to appeal to some moral values which are not themselves dependent on universal agreement. And many forms of conceivable universal agreement would be horrendous; a world governed by a racist totalitarian regime might get most people to subscribe to horrific things, and these would remain horrific even if most people were conditioned to think they are good.

    So to ask whether abortion is wrong is one thing, whether most people think it is another. You can accept this distinction even if you refuse to think about morality in terms of absolute objective values that we know a priori. But this opens up big questions I do not have space to properly discuss here. In any case, in writing this post I was really interested in relating (1) and (3) to (2), the idea of innate moral structures shared by all humans. What we are all innately disposed to accept is not thereby made true, or valid, or good. But even you just want to focus on (1), on universal moral code as something we agree on, it is not at all clear how (2) helps. It obviously doesn’t mean we can’t disagree in deep ways that lead to conflict. Nor does it give us reasons to endorse the innate code scientists have claimed to discover—which in any case is just a formal structure compatible with numerous possible moralities. So even if you want to drop all talk of ‘validity’, it’s not yet clear what, in practical terms, we would get even if this scientific hypotheses turned out to be true.

  • Peter Wicks says:

    Thanks, this is very clear… although I still think that (2) can help, in particular by providing guidance on how to formulate (1) in a way that has a good chance of success, given that people are more likely to agree on something they are innately predisposed to agree to. Not only can help, but is “immensely important” as I claimed above, in the sense that forging such agreement is an important and urgent challenge for humanity.

    By the way: “These woud remain horrific even if more people were conditioned to think they are good.” I’m not sure how this statement can be proved logically. You and I find the idea horrific, but this reflects our own values, I don’t think it relies on any claim that these values are universally valid.

Authors

Subscribe Via Email

Affiliations