Tuesday, January 20, 2015

Morality and ontological grounding. Some comments on Craig's metaethical argument.

In the context of his metaethical argument[a], William Lane Craig contends that only theism can provide an ontological foundation or grounding of morality – or of “objective moral values and duties”.

As Morriston explains[1], when Craig talks about “grounding” of moral values, etc., he means an “informative” identification. That would be similar to the way in which heat is identified with molecular motion, and water with H2O.

Yet, Craig does not give any good reasons to suspect that lacking a plausible non-theistic theory of informative identification would raise any problems for non-theists, let alone establish that theism is true.

On that note, let's consider some other cases of informative identification:

1. A posteriori informative identification. Water, heat, and some other cases.

Assuming the identification of water with H2O is correct[b], that is a case of a posteriori identification – as Morriston also points out. [1]

Moreover, for most of history, the identification was not known. In fact, philosophers managed to figure that out only after the development of the relevant physics, and of new terminology – i. e., hydrogen atom, electron, etc.

Without those developments, the identification would not have been discovered.

The same applies to the case of heat and molecular motion.

In the case of moral obligations, moral goodness, etc., no such science has been developed.

In fact, moral psychology, and psychology in general, are in their infancy compared to physics, and there is no known a posteriori informative identification informed by scientific discoveries of, say, kindness, cruelty, hatred, illness, etc. [c]

Given the previous considerations, the fact that there is no known a posteriori informative identification informed by science in the cases of moral goodness, immorality, etc., is unsurprising.

Whether one should expect that philosophers will eventually come up with an a posteriori identification of moral goodness, etc., informed by future science, is another matter, but at this point, one should not expect such identification to be forthcoming.

So, if the non-theist is in no position to offer one such identification, that is not a problem for her. She has no such burden.

What about some sort of a posteriori identification that is not informed by science?

It's hard to see what else such an identification might be informed by without science, but in any case, one may point out the lack of such a posteriori identification not informed by science in the cases of water, or heat – there are identifications informed by science -, and also in the cases of cruelty, kindness, illness, etc. [c]

Given all of the above, the non-theist may properly say that while she does not know of any a posteriori informative identification in the cases of moral obligations, or moral goodness, etc., she has no burden to come up with them. In fact, not having an a posteriori informative identifications is a very common occurrence, and there seems to be no particular reason why the non-theist would have any burden to come up with such theories in the case of moral duties, etc.

2. A priori non-analytical informative identification.[d] Greenness, kindness, cruelty, and illness.[c]

In the cases of water, or heat, there is no known a priori non-analytical identification. In the cases of cruelty, hatred, kindness, and greenness, there is no known a priori non-analytical informative identification, either, just as there is no known a posteriori informative identification.

In fact, cases in which there is no known a priori non-analytical informative identification are by far the most common. [e]

It seems this is not just because philosophers have not attempted to come up with that. Rather, it seems that coming up with such identifications, if possible at all, is notoriously difficult. [e]

As before, given all of the above, a non-theist may properly say that while she does not know of any a priori non-analytical informative identification in the cases of moral obligations, or moral goodness, etc., she has no particular burden to come up with one. Even if she's a metaethicist, lacking such a theory should not put any significant particular pressure on her.

3. Analytical informative identification.

We're only left with analytical cases. Also, in order to be informative, the identification would have to be based on a non-transparent analytical statement.

Does a non-theist have a burden to come up with an analytical informative identification?

There seems to be no good reason to think so.

Moreover, there seems to be no good reason to believe that there is an identification of that sort between moral obligations, moral goodness, etc., and properties or entities we usually describe in terms that appear to be non-moral terms.

As for analytical identifications involving terms that are clearly moral ones, such as moral obligation, immorality, etc., some identifications of that sort may not be so difficult to find. But in those cases, one might question how informative they are. Moreover, to the extent that some identifications of that sort are available, there seems to be no good reason to think the non-theist would have a particular problem with them.

4. Is it there a possible non-theistic informative identification of moral goodness? What about cruelty?

Perhaps, a theist might argue that given non-theism, no informative identification of moral goodness (or moral obligation, etc.) will ever be forthcoming , and indeed it is not possible it's allegedly an ontological problem for the non-theism -, and that that is the problem for the non-theist. But no good reason to think so has been given by Craig. [a]

Moreover, for that matter, one might suggest that without a necessarily existent maximally cruel agent, there can be no informative identification of cruelty, or even that that would be an ontological problem for non-necessary-cruel-agent-ism. The claim that without God there is no moral goodness seems to be relevantly similar.

Incidentally, Craig holds that if God did not exist, rape would still be cruel[3][f]. But let's define C-god to be an essentially maximally cruel, omnipotent agent. One might suggest the parallel:

Q1.1: If C-god did not exist, objective cruelty would not exist.

Q2.1: Objective cruelty does exist.

C1.1: Therefore, C-god exists.

If the omnipotence condition is an issue, one may remove it and define a c-god as an essentially maximally cruel agent, and then suggest the parallel:

Q1.2: If no c-god existed, objective cruelty would not exist.

Q2.2: Objective cruelty does exist.

C1.2: Therefore, at least one c-god exists.

Granted, defenders of Craig's metaethical argument might try to get around this on different grounds, perhaps suggesting a difference in terms of analyticity of certain statements, or supervenience, or something like that. But the burden would be on them.[g]

5. Theistic accounts.

It might be suggested that even though the non-theist wouldn't have the burden to offer a theory informatively identifying moral obligations, goodness, etc., with something described by other terms, there is pressure because the theist has a plausible account, like – allegedly – Craig's DCT.
But this is not so, either, since the non-theist may reply that:

5.1. Any theistic account is committed to the existence of God. That alone makes the account's [prior and final] probability negligible. Of course, a theist would deny that, but disagreement is to be expected as usual. It's beyond the scope of this post to argue that point, but that is no particular pressure on the non-theist.

5.2. Any theistic attempt to provide an informative identification will have to include a reference to God – in one way or another – among the properties or entities moral obligation, moral goodness, etc., are identified with – else, it wouldn't be theistic, even if it were logically compatible with theism.

But furthermore, given that “God” is defined in terms of moral terms, there will be moral properties that will not be informatively identified with anything else. But then – the non-theist might ask -, why should one think that there is an informative identification of moral obligations, or moral goodness?

5.3. As Morriston explains[1], according to the account defended by Craig, moral goodness is - allegedly - informatively identified with resemblance to [some aspects of] God. But according to Craig, God is to be understood as the “greatest conceivable being”, and that's a definition.[4][5][6]

Yet, moral goodness is a great-making property[4], so it seems that greatness is a complex property, a conjunction of moral goodness and some other properties, apparently involving knowledge and power, and perhaps lovingness and some other properties. So, a maximally great being would be maximally morally good, it seems.

But moral goodness is identified with resemblance to God, and God is a being that has some properties or traits to a maximal degree, including power, knowledge, and moral goodness. How is that identification informative?

If the identification is not between moral goodness and resemblance to God, but between moral goodness and some properties of God, the problem remains given the definition of “God”.

There are ways for the theist to try to get around this, like saying that the ontological grounding condition is a matter of ontology not epistemology and allegedly it does not matter whether it's informative – but then, why would Craig say he was talking about informative identification? Would informative identification not matter in the ontological sense? -, or saying that one can pick the right entity – namely, God -, without giving a definition, but otherwise pointing to him, etc. In my assessment, they all fail, but in any case, the burden would be on the defender of the metaethical argument.

6. Conclusion.

Craig[a] claims – among other things - that non-theists have no grounding of moral obligations and moral values, where “grounding” is understood in terms of informative identification. But non-theists have no burden to come up with such an identification at this point, and Craig has not provided any good reason to think that theists are or will ever be better equipped to come up such a theory.


[a] I posted a much more thorough reply to Craig's metaethical argument in another post, and elsewhere. Some parts of it are outdated, in the sense that I would write them somewhat differently if I did it now, but for the most part, I would make the same arguments today.

[b] The metaphysical claim of identify might be questioned without challenging the science behind it, but that is not required in this context.

[c] It might be suggested that cruelty is merely the absence of kindness, and that that would be relevant in this context. But it's not the case that cruelty is merely the absence of kindness.

For example, two agents A and B could both have no kindness whatsoever, and yet A could be more cruel than B. For instance, A may be completely indifferent to the suffering of others – or, indeed, to anything that happens to others, good or bad -, whereas B revels in torturing others for pleasure. Neither A nor B has any kindness at all, but B is more cruel than A. Yet, if cruelty were merely the absence of kindness, two agents that are equal with respect to kindness would be also equal with respect to cruelty.

Also, it might be suggested that cruelty is not objective in the relevant sense. But it is. One may mirror Craig's arguments on the matter.[2] For example, the Holocaust was a cruel act. And it would have been cruel even if the Nazis had won the war and convinced everyone that it wasn't cruel, and even if some of the Nazis themselves failed to see it was cruel.

Moreover, whether an act is cruel is a matter of fact, not a matter of opinion or taste.

So, cruelty is objective in the relevant sense.

[d] By “non-analytical” informative identification I mean an informative identification between properties or entities that can't be established on the basis of the meaning of the words that pick those properties or entities alone, plus logic.

Conversely, by “analytical identification”, I mean one that is not non-analytical.

[e] There is no need here to address here the issues of whether there is at least one known non-analytical, a priori identification, or even whether a non-analytical, a priori identification is possible. It's sufficient to point out that at least, in most cases, there is no known a priori, non-analytical identification.

[f] Craig also holds that it's impossible that God does not exist, so I guess he considers that “If God did not exist, rape would be cruel” is a non-trivially true counterpossible, or something like that. But in any case, I don't intend to object in this post to Craig's use of what he believes are counterpossible scenarios.

[g] It's beyond the scope of this post to assess potential replies and show they fail – though I hold they do fail -, but the burden to defeat the parallel is, in any case, on the defender of the metaethical argument.


[1] Morriston, Wesley, “God and the Ontological Foundation of Morality”, Religious Studies (2012) 48, 15–34 f Cambridge University Press 2011

link: http://spot.colorado.edu/~morristo/DoesGodGround.html

[2] http://www.reasonablefaith.org/defenders-2-podcast/transcript/s4-19

[3] http://www.reasonablefaith.org/the-euthyphro-dilemma-once-more

[4] http://www.reasonablefaith.org/defining-god

[5] http://www.reasonablefaith.org/defenders-2-podcast/transcript/s3-3

[6] http://www.reasonablefaith.org/personal-god


Narciso said...

Hi Angra,

I'd like to say how much I like your writings on this blog and others such as ex apologist's and the Secular Outpost.

Anyhow, I've done lots of thinking on the moral argument and Morriston's objections to it, and my thoughts are that defining God as the maximally great being cuts Morriston's "moral properties are simpler" argument.

However, I'm not sure how to support this. I haven't done much reading on the ontological argument but my thoughts are that such a being would be "maximally independent" as one could be. What I mean by this is that God would- and therefore could not be- dependent on such properties to be considered morally good. In other words, God, as defined, would necessarily depend only on his moral character to be good (that would entail that, if God exists, DCT is true, and, if an objective morality must be applicable logically to all possible beings, then premise 1 of the argument is correct)

I'm sure there is an error somewhere. Perhaps Morriston's argument entails that a maximally great being, if exists, would need such properties to be good, but I'm not sure how to support that either.

Perhaps an analogy could be that I could also say that such being would, as part of its greatness, win any game in any condition anyway it wants, and thus it would entail that it can win odd or even on the mirror with even. Of course that is impossible, so what this being can be depends on what we know of the world and not the other way around.

I'm not a philosopher, by all means (else I'd probably know how to answer this) but these are my thoughts. Any replies would be appreciated.

Angra Mainyu said...

Hi, Felippe,


With regard to your objection to Morriston's arguments, I'm just writing a new reply to Craig's metaethical argument that addresses that worry – if I understood your point right - and others, by arguing that defining "God" as a maximally great being is not a move available to the DCTist, other than to pick the right entity, so to speak.

It's still a draft (in a non-traditional format; I set up a debate between a defender of each position), but briefly, a problem I raise is that it's a transparent conceptual truth (as Craig uses the words) that maximal greatness entails maximal moral goodness, so it would be like saying that to be morally good is to resemble some qualities of a being that is maximally morally good. But that leaves us in the dark as to what moral goodness is.

An alternative reply - one that does not rely on the conceptual connection between maximal greatness and maximal moral goodness -, is to ask the DCTist what the ontological foundation of greatness is.

Any reply like 'greatness is resemblance to God in some relevant respects' - or any other involving God - would have a circularity problem if 'God' is defined as "the greatest conceivable being", because the account would be saying that greatness is resemblance to a being that is maximally great, in some relevant respects. As an analogy, it would be improper to say that to be a meter long is to be as long as the distance between two lines that are at a distance of a meter.

There is a way out of this for the DCTist, but it has a cost: the DCTist would have to say that the definition of the word “God” is just a means of pointing at the right entity – since God is not a physical object, one can't just point one's finger at him.
So, for example, one might as well use some other property to pick the right object - in this case, God -, and the identification holds. One might just say that moral goodness is resemblance to the only omniscient, omnipotent, maximally kind and loving being that exists, in some relevant respects, or resemblance to some qualities of that entity, etc.

In that way, circularity is avoided, but the cost is that the definition of 'God' as the GCB is lost.

For example, let's say that there is a being that is omnipotent, omniscient, and maximally loving and kind, for the sake of the argument. Let's name that being “Luke”. Now, that just picks an - assumed – actual being by naming some of its properties - enough to identify him uniquely -, and then uses a proper name – Luke – to name that being.

Then, one might still ask: "Why is Luke good?"
It would be improper for the DCTist to say that by definition, Luke is morally good.

The DCTist may still say that Luke is the paradigm of goodness, that he is the GCB, and/or that it's a necessary brute fact that he's morally good, but that's not like saying that it's by definition. The DCTist still needs some brute fact like that.

A non-theist might respond that, perhaps, to be morally good is to have certain qualities like being disposed to be kind in such-and-such situations, having such-and-such feelings in such-and-such situations, etc., and then propose that the alternative theory (which might not be true, but it's a suggestion) is at least no less plausible than DCT (which the DCTist may deny, but then it's a debate on who one has a more plausible suggestion; as you may have guess, in my assessment the non-theist wins hands down – and the non-theist may also insist that she does not actually need to present any accounts, though that's a different argument).

I'm not sure that addresses your specific worry (please let me know if I missed your point) and it's clear enough, though. The new reply has a more elaborate explanation, but
I can post a briefer reply to the metaethical argument before the draft is ready, only containing the relevant part of the hypothetical debate, if you like.

Narciso said...

Hello again Angra,

I think you got my point correctly, most of it. I was primarly concerned if defining God as the maximally great being would entail that DCT is true, but I think that if your approach is correct there is no such option for the DCTist.

This can be a little off-topic, but are you following the Feser-Parsons debate on morality? That is a great piece of exchange.

Angra Mainyu said...

Hi again Felippe,

Okay, so while the longer post I was writing is not ready yet (I keep adding details, so it's getting longer), I'm going to post a shorter version which deals specifically with the worry you raise, and defends my claim that that's not an option for the DCTist. Additionally, it suggests another problem for the DCTist, in terms of potential non-theistic accounts that any defender of the metaethical argument (and any DCTist) would have to defeat.

That shorter post is ready, so it should be up in a few minutes. If you read it, please let me know if you think that some parts are not clear, so that I clarify them.

With regard to your question, I haven't been following the exchange. But I'll take a look; thanks for the tip.