Tuesday, September 10, 2019

Ethics and complexity

Here is a picture of ethics. We are designed to operate with a specific algorithm A for generating imperatives from circumstances. Unfortunately, we are broken in two ways: we don’t always follow the generated imperatives and we don’t always operate by means of A. We thus need to reverse engineer algorithm A on the basis of our broken functioning.

In general, reverse-engineering has to be based on a presumption of relative simplicity of the algorithm. However, Kantian, utilitarian ethics and divine command ethics go beyond that and hold that A is at base very simple. But should we think that the algorithm describing the normative operation of a human being is very simple? The official USA Fencing rule book is over 200 pages long. Human life is more complex than a fencing competition. Why should we think that there are fundamental rules for human life that can be encompassed briefly, from which all other rules can be derived without further normative input? It would be nice to find such brief rules. Many have a hope of finding analogous brief rules in physics.

We haven’t done well in ethics in our attempts to find such brief rules: the Kantian and utilitarian projects make (I would argue) incorrect normative claims, while the divine command project seems to give the wrong grounds for moral obligations.

It seems not unlikely to me that the correct full set of norms for human behavior will actually be very complex.

But there is still a hope for a unification. While I am dubious whether one can find a simple and elegant set of rules such that all ethical truths can be derived from them with no further normative input, there may be elegant unifying ethical principles that nonetheless require further normative input to generate the complex rules governing human life. Here are two such options:

  • Natural Law: Live in accordance with your nature! But to generate the rules governing human life requires the further information as to what your nature requires, and that is normative information.

  • Agapic ethics: Love everyone! But one of the things that are a part of love is adapting the form of one’s love to fit the the persons and circumstances (fraternal love for siblings, collegial love for colleagues, etc.), and the rules of “fit” are extremely complex and require further normative input.


Heath White said...

Do the right thing!

Angra Mainyu said...

Hi Alex,

From a very different perspective, I agree the rules would likely be very complex: our minds resulted from a messy evolutionary process, so we're likely to get a very complex set of rules.

But that aside, I would challenge the "Natural Law" characterization, at least assuming that it gives testable outputs, like the Kantian and utilitarian projects (which, I agree, make false normative claims, but we know that precisely because we have a sense of right and wrong we can properly use to test such claims).

If "live in accordance with your nature" is or includes a claim that we have a moral obligation not to use organs, functions, etc., against their proper function (or "purpose", but I think the latter entails creation, so I'd avoid it), and the usual claims about the proper function(s) are warranted, then I would argue we have counter examples, as we do in the case of Kantian or utilitarian ethics in its several varieties.

Now, granted, you or others might disagree with all of the examples offered as counter examples of the Natural Law claims. But then, Kantians or utilitarians might do the same with their respective counter examples. If using my own sense of right and wrong is warranted, then I rationally reject all of those theories due to counter examples. If it is not, how should I test them?

Also, granted, in some social contexts, most people would reject some of the counterexamples to Natural Law (as usually constructed). However, I think there are counter examples that would be accepted no less frequently than the counter examples to, say, Kantian ethics or utilitarianism.

On the other hand, if the usual claims about our nature are not warranted, then how are we to test the theory that we have an obligation to live in accordance to our nature?

Finally, if the claim is not testable (counting testing it against one's intuitions, of course), then how could one get evidence for it? And if one can't get evidence for it (not even intuition-based evidence), then why should one believe it?

Alexander R Pruss said...

I don't think it is an essential part of natural law ethics that organs and biological functions are not to be used against their proper function. What is essential to natural law ethics is that the *will* is to be used only in accordance with its proper function.

A lot of natural law theorists add that other aspects of us, besides the will, should never be used against their proper function, but I think this is a separate thesis, a thesis I have argued against in general (though it may be correct in special cases, including the controversial sexual ones). I also think a lot of the apparent counterexamples stop working if this separate thesis is changed to read something like: "It is wrong to use an organ contrary to its immediate natural function except in pursuit of a higher function that the organ's immediate natural function exists for the sake of."

Angra Mainyu said...

I guess it might depend on how you construe the proper function of the will in the first case. Are you familiar with any Natural Law theory of that sort that makes moral predictions in specific cases, so that I can test them?

With regard to the counterexamples, I don't think the immediate function generally exists "for the sake of a higher one", in general. But for example:

An experiment is designed with the purpose of testing a vomit-inducing drug. The idea is to reliably and safely induce vomit in case of medical necessity, but to do that, they need to test the drug. Now, make a parallel between the testing of the drug (on healthy patients) and the use of, say, emergency contraception, or generally a contraceptive pill. In the first case, the drug does not prevent the act (consisting in eating food) from putting food in the stomach, but results in vomit afterwards. In the second, the act does not prevent a sexual act from placing semen into a vagina, but prevents fertilization afterwards. Natural Law in its usual formulations condemns the latter. I think both are pretty good counterexamples, but leaving the sexual case aside, I would offer the vomit-inducing drug as a counterexample (it's a variant of one of Grisez's objections, but simpler and without the medical necessity, though I'd argue that Grisez's objection succeeds as well, even if most of his other objections don't), since the arguments given against the latter (on the basis of Natural Law) seem to apply to the vomit drug as well.

Also, there seems to be no higher function that the eating function seems to exist for the sake of, at least not one that is involved here. If, however, you think there is such a higher function (which one?), then I would need more information in order to test this version of the Natural Law theory (though in that case, arguably a contraceptive pill could be used to fulfill some higher function as well; I don't see a way around that one).

On the other hand, if the theory makes no testable predictions (i.e., no claims in specific cases about what's permissible, impermissible, etc.), it would be difficult to warrant belief, I think. Do you think there is an argument for it, lacking predictions?

Red said...

Trying to understand the post and comments here,I was wondering what it would mean for ethics to be complex. Does it mean that for every single true moral imperatives and the circumstances under which they are true, there is an independent, intrinsically authoritative, irreducible moral truth or something like that?

Angra Mainyu said...

I don't know what you mean by that, Red. But I understand it in terms of algorithms. There seems to be an algorithm by which we make moral assessments on the basis of some information, even if individual implementation is often (or always, but that's debatable) flawed. The question seems to be whether the algorithm is complex (like a law book, with all sorts of details, exceptions, etc., or even more complex), or is simple, like maximize happiness (though it's also difficult to ascertain what maximizes it, so there is that), or some other simple maxim. I think it's very probably very complex.

Red said...

OK, so ethics being complex means that there are a lot of different things that need to be considered before we can act morally rather than simply thinking whether that action promotes some simple principle like "maximize happiness"?

Angra Mainyu said...

As I understand it, yes, I'm I'm reading your point right.

However, the rules are not transparent to us: we have a sense of right and wrong that is applying the rules, very probably considering many variables, but we do not know what the variables are, in general: if we knew, we would know the algorithm, and we would know whether ethics is complex or not, or more precisely, how complex it is.

So, in short, we need to consider a lot of stuff, but we probably are already doing so, intuitively. But it's extremely difficult to reverse engineer the algorithm.

Alexander R Pruss said...


In this context, I am interested in Natural Law as a *meta*ethical theory. And metaethical theories may not make much in the way of predictions.

Angra Mainyu said...


Fair enough, so there are no first-order ethical predictions. That still leaves open the question of justification, though: even though it makes no first-oder predictions, there should at least be supporting arguments, or something to that effect, at least something that would allow us to compare it with other metaethical theories and to test them - not against our sense of right and wrong if it makes no such predictions, but against something - arguments, background knowledge, some sort of intuition, etc.

In a general sense, I do think different things have different moral obligations - or have no moral obligations at all, depending on the case. Also, I think ethics is (very) probably (very) complex. What you've described of Natural Law seems compatible with that. But does it make other claims, even if not first-order ethical ones? For example, is there a commitment to Aristotelian or Aristotelian-Thomistic metaphysics?