Do we need go so far as to posit God - a maximally powerful and good entity - to explain moral knowledge under this argument? All we need is an entity that:
a) Has genuine moral knowledge
b) Wants to give that moral knowledge to humans
c) Has a way of transmitting that moral knowledge into humans' minds
To explain why our moral intuitions are suspiciously accurate. You could just as easily posit that we're all subjects in a huge social experiment run by aliens to see how humans would react if they had moral knowledge, and that our moral intuitions are being shaped by the aliens' telepathic ray guns to correctly give us moral knowledge. Of course, you might have other reasons to think God is more plausible than alien scientists, like fine tuning - but then you're relying on other arguments for God, and the moral knowledge argument in particular is not doing any work.
How do the aliens explain their own moral knowledge? Don't the aliens need to posit another entity that has genuine moral knowledge, wants to give that moral knowledge to aliens, and has a way of transmitting that moral knowledge into aliens' minds? Then we would have an infinite regress of entities unless we posit God.
No. One of the reasons we wouldn't expect humans to have accurate intuitions is that intuitions are generated by evolution, which obviously won't be correlated to transcendental moral truths. But we can imagine that these aliens developed a different way, or are higher dimensional beings, so it isn't surprising their intuitions line up with moral truths.
Yes, we can imagine aliens that are developed in such a way that their intuitions would line up with moral truths. But we can also equally imagine aliens with intuitions that don't line up with moral truths. If there are many more ways to be wrong than to be right about moral truths, then why alien's intuitions should line up with moral truths?
I think a naturalist could plausibly explain moral knowledge either in terms of rationalism, where our moral norms derive from some a priori discernible principle, such as the categorical imperative, so as long as our rational faculties are explainable in terms of naturalistic evolution (which they are, since being rational is evolutionarily advantageous), the naturalist could account for their moral knowledge. Or, they could be an Aristotelian, and think that moral facts correspond with natural facts, such as facts about human flourishing, or about pain and pleasure. Insofar as these are plausible moral accounts (which they are), there should be little problem for the naturalist in explaining their moral knowledge, no more than they would have in explaining the other forms of knowledge they might have, in which case moral knowledge isn’t particularly threatening to naturalism.
I don't think that the moral facts can be deduced from non-moral facts. The a priori discernible principle would have to be a distinctly moral principle, in which case the problem is just pushed back a level.
I think it's hard to see how the fact that "one flourishes when they do X," will explain why we believe that we flourish when we do X. But in any case, moral knowledge is just one example--the same point applies to inductive and modal knowledge.
Being rational is evolutionary advantageous until one uses reason to arrive at anti-natalism and convincing humanity collectively that the world would be better off without rational creatures and that it is immoral to bring rational creatures into the world.
Yeah, for example: "But on naturalism, one wouldn’t think we were any likelier to believe torture is wrong after learning that it really is wrong!" This isn't true if you believe that the "really wrongness" of torture is a natural fact relating to it producing intense amounts of pointless suffering. Upon learning that fact about torture, most people *would* come to believe it was wrong as a result. The mistake is assuming that the naturalist's moral facts are still non-natural features "layered over" the natural facts, which of course produces contradictions. Whereas when we reject that divide, no problems really arise. Not to promote myself here, but I wrote a similar piece about this a while ago: https://bothsidesbrigade.substack.com/p/you-cant-disprove-moral-realism-by
One very basic counter-argument to this is to say that either objective moral truths do not exist, or if they do exist, then they might be totally different from the moral beliefs that most humans have. In other words, your argument here seems to rest on the premise that most people do, in fact, have the correct moral beliefs, but I'm totally unconvinced that this is true.
Petition to linguistically engineer it to be "the argument from inductive knowledge" or something instead!
(Also, I feel like I should address what you say about my objection.
To the first point, I worry whether this doesn't just allow the naturalist an out as well. If I can directly see that pain is bad, then surely that doesn't require theism to be reliable, just like I don't need theism to know that I'm conscious. And if it's through some intuition that doesn't give you direct awareness of the badness, then I worry whether that would be subjectively indistinguishable from an unreliable intuition given by evolution.
Perhaps it's that you have this direct awareness, but that it's very unlikely that you would have it on atheism. Maybe so, but I have a hard time seeing what part of my experience is hard to account for on atheism. Like, atheistically evolved creatures would surely also have a feeling like pain was just obviously bad! But maybe it's just because I'm more inclined to think that we have intuitions about, rather than direct awareness of, morality.
On the second point: Yup, I agree, hence the petition above!
Though I think we should maybe be a little more careful in judging which knowledge is problematic and which isn't. I don't think some fact not being able to move atoms is sufficient.
For example logical knowledge is plausibly explained by evolution: 1) All bears are dangerous, 2) that is a bear, 3) therefore, that is dangerous. This seems like a very useful thing to know, even though the validity isn't able to move atoms. I think something similar is true for all types of deduction: It would make very good evolutionary sense for creatures to be able to reason deductively.
But I agree that abductive and inductive knowledge are plausibly problematic on naturalism, so the above pedantry is probably a bit unnecessary--I just want to flag that I think the case for some knowledge being problematic on theism is not quite as straightforward as it sounds.
I can see that pain is bad. But seeing something has a non-natural property is not the sort of thing that naturalism can account for. This seeing is a kind of direct intuition--but on standard naturalism, this won't be explained by the facts they're about.
Yeah, that would probably get you out of it. Then I guess I doubt that we really do have such a direct insight, as I don't really find anything when I introspect on my intuitions around pain that I can't make sense of on naturalism. But it's hard to argue about those kinds of things, so let's not let that divide us:)
I think abduction and induction are also evolutionarily advantageous. Creatures that don't learn that bears are dangerous after being attacked by bears a few times eventually get killed by one.
I agree that they're advantageous, just like how morality is advantageous. The problem is that their advantage (like moral beliefs) doesn't come from their accuracy. Ancestors who did believe that the future resembled the past were successful, since it actually turned out that the future resembled the past back then. But that has nothing to do with whether the future from *now* will resemble the past--to assume that is just to assume that induction works.
Correct me if I misunderstood your point, but doesn’t the evolutionary advantage of induction come exactly from its accuracy? E.g. ancestors who didn’t believe a certain food to be harmful even after it killed many others were less successful at reproducing
Well it was accurate in the past, but no one is questioning that. The question is whether we can continue trusting it in the future/unobserved instances, and whether or not it's reliable there cannot possibly have had any influence on evolutionary history, since it, well, happened in the past.
Thank you for clarifying. Maybe I’m still missing something but assuming that humans evolved a general “induction faculty” in the past, why would it stop working with similar cases to the ones above? If the question is how can we trust induction in principle things get more complicated, philosophers have talked about it since Hume afaik and many thought that you can’t have a rational justification for it.
We will definitely still think it works, but the question is the latter that you raise of how we can trust it. Now the problem here, I think, is a bit more difficult than the regular one as evolution (without theism) gives us an explanation for why we have a very strong propensity to believe induction, which doesn't make reference to its soundness. So our belief is actively debunked--somethinf that's surely unacceptable. I'm not sure that the standard view is that we just don't have rational justification for it, though that was certainly Hume's own view.
If there's a mystery here it's "why does induction continue to work" (which I actually find convincing as a theistic argument). Our belief that it will is completely explicible
Our belief is certainly explicable, and that's the starting point for the whole argument, since the explanation doesn't cite its reliability--just like our belief in moral facts is explicable on naturalism without reference to moral facts.
The problem is that we can't really deny induction without landing in unacceptable skepticism, so we must find some explanation that saves our knowledge, and theism is one such. So it actually sounds like you agree with the argument, maybe.
Surely it’s an odd choice to talk about how our moral intuition is accurate by using the example of slavery, which for the great majority of recorded history was viewed in many societies, including the most developed and religious societies of their time, as perfectly moral? I think the example of slavery absolutely should make us doubt our moral intuitions—it’s a much better example of moral intuition being socially constructed than it is of moral intuition being guided by a truly moral force.
The first problem with your argument is that your own explanation of moral knowledge is not derivative in nature. Your explanation of why you believe in moral knowledge relies on God, but you only believe in God because of an argument that starts from the assumption that your moral beliefs are true, so you don’t actually have independent justification for believing in moral knowledge, at least no more than someone who just declares that they believe in moral knowledge and that’s all the justification they need.
Your attempts to get around this also work for naturalists. if your experience that thinking on pain makes it obvious that it’s bad provides justification for moral knowledge that equally works for somebody who doesn’t believe in God.
Also at the end of the day, I think the problem of induction is similar to the brain in a vat thought experiment in that all arguments for the common sense position boil down to the fact that you don’t actually believe scepticism is true. At the end of the day, the only way to avoid scepticism is to assume that common sense and prior probability are reliable sources of information because you can’t actually make probability estimates without starting from a prior probability. At the end of the day, if you keep demanding justifications for beliefs, you’ll either have to take some beliefs on faith, accept that circular coherent it’s arguments are valid, or conclude that justifications form an infinite regress. But if you accept this, then arguments for induction that either work by assuming induction as a foundational axiom, or work through seemingly circular reasoning like arguing that we assume the world is simple because it worked in the past and an inductive world is simpler, no longer seem so unreasonable.
At the end of the day, I don’t think believing in God has actually placed you in a better position in resolving the problems you outline it’s just that your chain of reasoning, is longer and more complicated, but it still ultimately works by starting from the assumption that moral knowledge is true. In which case you can just assume that without adding God to the picture, because he doesn’t contribute anything to the argument beyond adding him to your picture.
Also, I don’t think God is actually that good of a justification for belief since he’s clearly not above misleading information as demonstrated by the fact that he has not made his existence as obvious as real world humans. if he’s not willing to demonstrate his existence, the way say real world, celebrities and heads of states do, and is willing to let lots of people persist in incorrect beliefs like his non-existence, what makes you think he won’t be willing to give you incorrect beliefs? And if he’s willing to give you incorrect beliefs, he is no longer such a solid justification for any kind of knowledge. At the your justification for moral knowledge should not be so good that it can’t explain people who have incorrect intuitions, as many demonstratably do. But if God is willing to give other people incorrect intuitions what makes you think he wouldn’t be willing to do it to you.
A good test for judging the validity of an argument for God is to ask whether it would be a better explanation to simply assume whatever needs explanation is a fundamental law of nature. After all, assuming that, for example, God wants life to exist is equivalent to assuming that the universe inevitably creates life minus the extra complications that you’re not assuming the laws of nature are an agent or have any of the other properties attributed to God. Applying this to your argument, how is your argument any better than somebody who asserts that moral knowledge must be justified so it must be a law of nature that it won’t deceive you by giving you incorrect intuitions about morality so moral knowledge is justified and hence we have solved the moral knowledge problem and proved that the universe will not fundamentally give you incorrect moral intuitions. As far as I can see, this is equivalent to your argument, except the role fulfilled by God in your argument is fulfilled by the laws of nature, and there’s nothing in your argument to favour God over this explanation.
>The fact that slavery is wrong does not directly shape selection pressures in ways that cause people to believe it.
This simply isn't true. Morality is the cultural DNA of a culture and represents generational wisdom. It directly impacts the survivability of the cultures which adopt it and therefore is itself subject to selective pressures. The reason that e.g. murder is "wrong" is that cultures that condone murder inevitably peter out in an endless series of blood feuds; you want your citizens working, not murdering. The reason slavery is "wrong" is because, in a post-industrial world, it represents both a misallocation of human capital and an affront to the individual liberties that unleash that capital. You'll note the qualification there. There's a reason it was considered moral for the entirety of human history right up until the Enlightenment, and that's because the industrial revolution radically changed the economic status of the average person. Before then it was economically reasonable and therefore morally acceptable. It's even condoned in the Bible.
Here's an interesting historical anecdote that illustrates the socioeconomic function of morality. When European explorers discovered Tahiti in the 17th century, they were amazed to discover that the Tahitians had a far higher standard of living than Europeans did. Meaning, specifically, that they had an abundance of food and plenty of leisure time - they got to lie around on the beach quite a bit instead of toiling endlessly in the fields. The reason for that is that Tahitians ritualistically murdered 4 out of every 5 babies that were born. Why did that make them rich? Pre-industrial societies live in a Malthusian equilibrium. Productive capacity was limited by agricultural output so the only way to raise living standards was to keep the population artificially low, otherwise it would equilibrate at a level just above starvation. Infanticide was Tahitians' method of birth control and that kept them rich. By contrast, Edo period Japan was *poorer* than contemporary Europe because of Japan's cultural rituals around cleanliness: that kept disease rates lower and therefore population densities were higher relative to agricultural capacity. I'm sure you've heard that the depopulation caused by the Plague led to a temporary boon for the common man in medieval Europe. It's the same principle: you had the same amount of food to feed a population that was suddenly 30% smaller.
Were the Tahitians immoral? No, they weren't. That's because moral rules aren't absolutes. They represent the cultural encoding of survivability-enhancing practices. Murdering babies made Tahitian society function better. Try explaining to them why that's immoral.
> it really is wrong to torture babies.
Uh, we routinely circumcise infants. That could certainly be construed as torture -- not by me, but the argument could easily be made -- so maybe be a little less cocksure in your references to moral absolutes.
Axioms such as modus ponens, as well as priors such as physics remaining the same over time, *must* be assumed instead of justified. I *cannot* be certain that those principles are correct, regardless of how intuitive they feel.
A simple counterexample is that I might be dreaming. I have memories of dreams during which I reasoned in ways that, in retrospect, blatantly violated basic logical principles: assumptions popping in and out of existence, my line of reasoning not so much a line as a web of disjointed thoughts. But I *thought* I was reasoning normally. My intuition of logical correctness was just as strong then as it is now.
For all I know, I might be doing the same right now as I write this comment. Modus ponens states that if p implies q, and p, then q. I believe that it’s true. But if I’m dreaming, then that definition might be wrong, in which case I might actually currently believe in some false axiom that I’ve mislabeled as modus ponens. Or perhaps there is no such principle as modus ponens and I’ve just hallucinated it.
Therefore, “insane skepticism” really is logically justified. The possibility that my most basic beliefs are false is not just some remote hypothetical but a common occurrence. And positing God doesn’t help with this: empirically, God does not mold my beliefs to be accurate while I’m dreaming.
The only way out is to assume. If my thinking is illogical, then my conclusions will be garbage no matter what I do, so I may as well assume it’s logical. If the laws of physics are about to get randomized, then there’s no way to predict what action will benefit me, so I may as well disregard the possibility when choosing how to act.
But once I admit my assumptions are just assumptions, then God doesn’t need to come into the picture.
As for morality, objective morality doesn’t exist. My sense of morality is ‘arbitrary’ in the sense that its origins can be explained by evolutionary principles rather than any deeper connection to the universe. But I’m still going to use my sense of morality to guide my actions, as well as to judge everyone else’s actions.
You say that having a natural faculty that allows us to grasp non-natural facts is mysterious, but what if both natural and non-natural truths are able to be intuited by the same cognitive system, such that evolution selecting for the ability to intuit evolutionarily useful natural truths entails as a necessary spandrel the additional capacity to intuit non-natural truths as well, because the same system is fundamentally able to grasp both of them? This is what I suspect to be the case, that intuiting the truth is selected for in brains because there are many cases in which it is evolutionarily useful, and even in cases where it is not evolutionarily useful, the fundamental truth-intuiting process we already selected for still gives us true beliefs as a nice built-in bonus.
I think this is a meaningful critique of ethical non-naturalists, but many of the morality-specific points being made don't really apply to ethical naturalists who hold that moral facts are a sort of non-reductive natural fact like facts about health. In that case, it seems like (properly interpreted, at least) it *is* the case that the wrongness of slavery plays a role in why we think it's wrong, in the same way that the unhealthiness of eating rotten meat plays a role in why we think it's unhealthy.
Otherwise, the expanded critique you're giving just seems to boil down into a general question of how priors are justified. And that's definitely a tricky question, but as far as I can tell, most of the popularly accepted accounts are not theistic anyway, right? I'm sure there is a good "you need theism to establish priors, period" argument out there, but I don't think the considerations here actually challenge those other accounts specifically.
I don't actually think that moral naturalism avoids the problem. The substantive claim of naturalism isn't that natural fact N plays causal roles X, but that N is identical to moral fact P. However if natural facts M=P instead, N and M would still play all the causal roles they actually do. And since these causal roles are what determine our beliefs, not which of N and M is identical to P, we still have a defeater.
I'm not sure I follow what you're saying here. Is the idea that we could fix all the non-moral natural facts and have it be that they were instead identical to a "swapped" set of moral natural facts? So we keep "torture causes tremendous amounts of pointless suffering" as the non-moral fact and then just swap in "Pointless suffering is good" as a moral fact? If that's what you mean to communicate, then I just don't think that's a concern that any naturalistic theory really needs to worry about - it would be like saying "Well, you could have the natural fact that arsenic causes your heart to stop and then the health-related fact that stopping your heart is healthy, but you'd still think it was unhealthy anyway." In that case, it seems obvious to me that naturalists can just say health-related facts aren't "layered over" other sorts of facts in that way but rather supervene on (or are grounded in or whatever) them by virtue of their meaning, such that the recognition of the higher-level property is legitimately influential. Why would it be different here?
Yeah that's basically the idea. I'm not sure it's unproblematic thought.
I take it that the claim of naturalism isn't just that "goodness" means "reduces pain" or whatever, but that there is a genuine property of goodness "out there" that's identical with some natural facts.
With the case of health, I would think that the reason it doesn't work is that "healthy" just means "promotes functioning of the body" or something along those lines. If the claim is that there is some genuine property of healthiness, then I just think this is a perfectly good debunking argument against there being such a property: Even if the property hadn't existed, or had been identical to some other phenomenon, we would both had written these exact sentences anyway.
Not to be "that guy" who's always linking to his own stuff, but I wrote a post about this a while ago: https://bothsidesbrigade.substack.com/p/you-cant-disprove-moral-realism-by In short, it seems to me that this sort of reasoning would allow you to reject the existence of pretty much any property, on the grounds that it's (trivially) true we would still act the way we do if it actually is the case that property doesn't exist. For example, if it's actually the case that there are no fish, then all our fish-related judgments have nothing to do with fish, and that means we'd all still act the same way without fish. But that's just another way of saying that we could be mistaken about fish existing - and that, by itself, doesn't give us any particular reason to doubt that fish *do* exist. On the other hand, someone who believes fish *do* exist and that they *do* explain our seemingly fish-related behaviors would be justified in rejecting the counterfactual claim that we would act the same way without them.
That's an interesting post, I'll have to read the paper too!
But I think that's just a reason to bite the bullet on fish, insofar as "fishiness" is a genuine property that we should at to the content of our theory, rather than a word. If the fish-theory being true or false doesn't make any difference to any of our evidence, then we have no evidence for or against it. The question is then just whether the fish-theory has a higher prior than the non-fish-theory. And seeing as the fish-theory adds an additional property to our inventory, it's surely just worse prior to looking at the evidence. Mutatis mutantis for moral realism.
In the post you make an analogy to an external world skeptic, but the situation is different here. Positing an external world allows us to explain our experiences in a much more simple way than not positing it, and so it makes up for the ontological cost. But I don't see moral properties doing any such thing.
Basically we should remove anything from our ontology that doesn't explain anything, and which complicates our worldview, and it looks like moral properties and fishiness are both in this boat.
Though I feel like I may not quite be getting your point, so please tell me what I'm missing:)
That's fair! I understand why people would have a vaguely eliminativist posture towards all sorts of second-level properties. I sometimes describe myself as a "relative realist" - I care less about the ultimate ontological status of moral facts and more about the notion that they're "as real as" the other sorts of facts that people consider uncontroversial. So if you don't believe that moral facts exist, but also that "fish facts" don't exist either, then I would say we just have a more fundamental disagreement, but one that doesn't really impact the relative structure of our ontological hierarchy. Whereas when someone is a realist about all sorts of things, *except* moral properties, that I think is more open to this sort of objection.
You might be right about moral naturalism. It's tricky to think about. Even on moral naturalism, it seems hard to see how the wrongness of torture would be responsible for us believing that it's wrong.
It's not just a "general question of how priors are justified." It's a reason to think that priors are not justified absent theism.
Why is that hard to see? It doesn't strike me as any weirder than saying (to use the same sort of analogy again) that the healthiness of exercise is responsible for us believing that exercise is healthy. In both cases, you've got a higher-level natural property fulfilling every sort of role a good explanation should, with explanatory power not necessarily possessed by any particular set of first-level natural facts it could be reduced to. So that, to me, makes it a perfectly reasonable (and actually indispensable) explanation for our beliefs. I lay out my case for this in detail here: https://bothsidesbrigade.substack.com/p/the-world-makes-more-sense-if-youre
And sure, I get the argument and I think it's interesting. I'm just saying it's not really an argument from moral knowledge in the traditional sense anymore, and more just a straightforward argument from the justification of prior probabilities.
Well, I don't think it's particularly problematic for the snakes to have different moral intuitions - snake flourishing is distinct from human flourishing when it comes to a bunch of little contingent details, so they probably would feel differently about some things. But if those snakes developed legitimate personhood, then I think the broad considerations we make would develop in them too. Of course, it's always possible that evolution would systematically mislead some groups and not others for various reasons. But that's just an expression of a general skeptical urge, not any particular issue with moral knowledge. And it's one a theist can't really avoid either, imo - at some point, everyone needs to accept that they could be wrong, generally.
It makes sense to say that every system takes something on faith (or intuition, as you frame it). Lewis Carroll’s Achilles and Tortoise paradox uses a similar (but shallower) approach to show that even simple logic requires acceptance on faith that logic means what we say it does. Similarly, a probabilistic approach to truth only works if one trusts probability. But I don’t quite follow the objection that it’s too great a coincidence in a naturalistic framing for us to have intuited moral facts that happen to be true. What does “true” mean in this context, with no theistic frame of reference? It seems like the more likely outcome of this system is just to slide into extreme moral relativism, in which no moral truths exists beyond what we intuit on an individual level, or can justify (imperfectly) by demonstrating its practical utility.
I'm not raising a general skeptical challenge. My point is that on naturalism, many of our beliefs are not explained by the facts they're about, which removes our justification for believing them. This applies not just to morality, so even if you're a moral skeptic, the same point applies.
Naturalists aren't committed to there being nonnaturalist moral facts. This isn't a problem for naturalists in the same way nonnatural fashion facts or demonic facts about possessions aren't a problem for naturalists, even though naturalists can say there are still true facts about circumstances in those domains.
Thanks for clarifying, and agreed on the broader point. All may not be as “vanity, vanity” as Ecclesiastes posits, but it gets pretty close without a core axiom or two to anchor things. With that in mind, though, why not raise a general skeptical challenge? You conclusion is “…therefore theism.” Why not “…therefore theism or epistemic sludge?”
Yeah, I don't think many philosophers would argue that non-natural moral facts exist *and* that evolution directly led us to contact with them. Usually the combo is "natural moral facts + evolutionary pressures" or "non-natural moral facts + cognitive faculties."
Moral intuitions are not the basis of moral knowledge because intuitions don't exist and the reporters of these intuitions are misled/deluded like people saying they have a divine sense or can talk in tongues or can see auras or read horoscopes or speak to the dead or astral project.
Morality did not evolve, most moral behavior is because of enculturation and agents seeking to construct systems of social behavior by which to survive and thrive among other agents.
I think you mishmash two quite different things together. One is moral knowledge. And the other is Münchhausen trilemma and overall problems of justification of beliefs.
On the other hand, when you jump to overall problem with justification of beliefs, the notion of "justification" that you use can't be satisfied by theism just as well. Let's use the example from your post:
> 1. Tomorrow physics will continue as before.
> 2. Tomorrow physics will break down wildly and become completely different. However, every day before tomorrow, physics will work consistently in the ways described by the laws of physics.
Why do we think that the first hypothesis is more likely on theism? Because God is good and therefore would like us to have stable physics and accurate beliefs. But how is this belief itself is justified? Consider two hypothesis:
1. God is good and wants us to have accurate beliefs and therefore gave us our intuitions and tomorrow physics will continue as before
2. God is evil and wants us to have inaccurate beliefs and therefore gave us our intuitions and tomorrow physics will break down wildly and become completely different. However, every day before tomorrow, physics will work consistently in the ways described by the laws of physics.
If we can't just claim that the second hypothesis has lower prior based on our intuitions about simplicity of hypothesises, which itself can be a delusion from an evil God, then our preference of first hypothesis is not justified on theism
The same principle can be applied to every step of reasoning, for instance:
1. God is good and will continue being good and so tomorrow physics will continue as before
2. God has been good all this time but will stop being good tomorrow and so tomorrow physics will break down wildly and become completely different.
Or, and probably my favorite:
1. The concept of goodness implies desire for us to have accurate beliefs and stable physics
2. The concept of goodness implies desire for us to have accurate beliefs and stable physics up untill tomorrow, and then it will imply desire for us to have inacurate beliefs and unstable physics.
And God is open to the same design objections as for induction. What explains God's wants for induction to keep holding staying the same in the future as in past cases? If it's induction separate from God, then God's existence is pointless. If it's something about God, then there's a circularity or unexplained brute fact, and we can make our theory simpler by just removing God from the picture and giving that property to something we already know exists like the universe.
Haven’t had the time to fully read section 6 but just a small point on modal knowledge: I don’t think this works for logical modality. The logical and mathematical facts plausibly stand in some certain intimate relation (not causal, but something. maybe something like constitution) to the natural facts such that it’s plausible on naturalism that the evolutionary explanation for our logical and mathematical faculties includes those facts themselves. But if we’ve got reliable mathematical and logical faculties we’re gonna get logical modality. That’ll also get us nomological modality. But yea metaphysical modality no, but that’s just in the bucket of metaphysical claims. I also think we should be kinda skeptical of metaphysical modality anw
I don't understand the inductive skepticism point. How does theism not face something exactly parallel? Just consider the hypothesis of a deity who changes all the laws of physics tomorrow and plug that into the comparison principle you spelled out: both the Skeptical Deity and the Non-Skeptical Deity's actions would explain our intuition that the sun will rise tomorrow equally well, so we have to be neutral between them. Surely at best both of us can equally wriggle out of countenancing the relevant skeptical explanations of our experiences by making use of (say) simplicity, and at worst we're both equally screwed.
It also seems like you're switching between two very different principles and treating them as the same. One is the one you explicitly articulated: intuitions can't justify belief in A over B if they're equally likely to be intuited under both hypotheses. This is clearly false, at least if you reject skepticism. (It's actually kind of true if we model intuitions as Bayesian evidence that's supposed to update our priors, but we shouldn't do that. Rather, we should just treat intuitions *as* our priors, essentially.) The other is a causal debunking principle: your intuition can justify your belief in hypothesis A only if the causal explanation of your experiencing the intuition has "something to do" with A being true. (Notice that this principle doesn't intrinsically involve a comparison with a different hypothesis B.) But what "something to do" means here is unclear, and in fact "explains" is unclear. Suppose the standard picture of naturalism is true. The sun rises or explodes according to a certain set of dynamical physical laws. My belief-forming processes in my brain also operate according to those same laws. Therefore, the causal explanation of my belief in the sunrise tomorrow has "something" to do with the process of the sunrise. If this is an insufficient causal connection, then sufficiency needs to be explicated in more depth.
There's an even deeper worry in the background about how God is supposed to know any of this stuff, as well. I don't think appealing to God's necessity is helpful - it wouldn't do for naturalists to appeal to necessitarianism (which I'm in fact sympathetic to, not just hypothetically). And if instead you want to go with the simplicity of omniscience as a property, I think that view faces serious problems also on vaguely Tarskian lines.
You don't have direct evidence from intuition for a Non-Skeptical Deity. Instead, you deduce a Skeptical Deity probably isn't real from the fact that it's more complex and violates plausible moral constraints. Those beliefs you have because they're true. This also addresses the second broader skepticism worry.
I think that there are different ways of formulating the principle--it can either be about relative probabilities of explanation. If it's about explanation, it still won't be that the sun rising tomorrow explains your beliefs about it, only that the things that make the sun rise tomorrow explain your beliefs abotu it.
God knows these things because of his constituent omniscience.
> Instead, you deduce a Skeptical Deity probably isn't real from the fact that it's more complex and violates plausible moral constraints.
As I mentioned, the complexity argument applies to the skeptical naturalistic hypothesis re: sunrise just as well as the skeptical theistic hypothesis. The moral constraints you consider plausible are only so because of your intuitions, but those are also victim to your comparison principle (they'd be equally expected if the skeptical deity is deliberately feeding you faulty moral intuitions, and if he is in fact "morally simpler"). For that matter, maybe we start questioning our intuitions about simplicity, which is, after all, a highly philosophically contentious concept...
>I think that there are different ways of formulating the principle--it can either be about relative probabilities of explanation.
But those are different principles, not reformulations of the same thing! Not that that's a big deal, but I think it's important to highlight that they're different, not least because one of them (the comparison one) almost certainly implies some sort of skepticism unless heavily elaborated upon and refined.
>If it's about explanation, it still won't be that the sun rising tomorrow explains your beliefs about it, only that the things that make the sun rise tomorrow explain your beliefs abotu it.
The things that make the sun rise tomorrow - the general physical regularities - explain my belief in those general regularities, and from those together with my belief in past sunrises (each explained by watching the sun rise, in fact!), I deduce that probably the sun will rise. Why isn't this a sufficient story? And how are you doing better here? On your view, it's not tomorrow's sunrise explaining your belief directly, it's only through a similar divine intermediary who is himself what explains the sun's rise (since he decided to make the sun do that).
>God knows these things because of his constituent omniscience.
Sure, but I don't think it's at all clear that omniscience should be counted as simple. In the worst case, it's not a hundred percent clear that it's even coherent to talk about "all truths" for Tarskian reasons, much less "knowing all truths," and probably even in the best case more modern attempts to evade the semantic paradoxes by revising our logic aren't going to end up furnishing you with something simple as opposed to hideously complex. (Or at least the ones I know something about, admittedly a far from exhaustive sample.)
Not going to continue the back and forth because I have too many commenters, and though you are among the most thoughtful, I can't spend all day on this. Just one brief comment.
//And how are you doing better here? On your view, it's not tomorrow's sunrise explaining your belief directly, it's only through a similar divine intermediary who is himself what explains the sun's rise//
God grants us the ability to grasp non-natural facts, including that simpler worlds are likelier than more complex worlds. From this, we infer the sun will probably rise.
>Not going to continue the back and forth because I have too many commenters, and though you are among the most thoughtful, I can't spend all day on this. Just one brief comment.
No problem, though I'll respond anyway just to provide food for thought.
>God grants us the ability to grasp non-natural facts, including that simpler worlds are likelier than more complex worlds. From this, we infer the sun will probably rise.
(I originally wrote a response here that invokes evolution, but I decided to edit this comment and revise that because I think there's a better and more direct answer.)
If it's true that this world is simple, and if it's true that simpler worlds are objectively more probable (whatever that means), then the latter fact partly explains why this specific world, with all of its inductive regularities, exists as opposed to some other, more complicated one. And this inductively regular world existing explains just about everything about my beliefs, since I'm a part of it.
Do we need go so far as to posit God - a maximally powerful and good entity - to explain moral knowledge under this argument? All we need is an entity that:
a) Has genuine moral knowledge
b) Wants to give that moral knowledge to humans
c) Has a way of transmitting that moral knowledge into humans' minds
To explain why our moral intuitions are suspiciously accurate. You could just as easily posit that we're all subjects in a huge social experiment run by aliens to see how humans would react if they had moral knowledge, and that our moral intuitions are being shaped by the aliens' telepathic ray guns to correctly give us moral knowledge. Of course, you might have other reasons to think God is more plausible than alien scientists, like fine tuning - but then you're relying on other arguments for God, and the moral knowledge argument in particular is not doing any work.
It won't entail God, but it will be evidence for God--insofar as it's likelier that there'd be moral knowledge if there was a God than not a God.
How do the aliens explain their own moral knowledge? Don't the aliens need to posit another entity that has genuine moral knowledge, wants to give that moral knowledge to aliens, and has a way of transmitting that moral knowledge into aliens' minds? Then we would have an infinite regress of entities unless we posit God.
No. One of the reasons we wouldn't expect humans to have accurate intuitions is that intuitions are generated by evolution, which obviously won't be correlated to transcendental moral truths. But we can imagine that these aliens developed a different way, or are higher dimensional beings, so it isn't surprising their intuitions line up with moral truths.
Yes, we can imagine aliens that are developed in such a way that their intuitions would line up with moral truths. But we can also equally imagine aliens with intuitions that don't line up with moral truths. If there are many more ways to be wrong than to be right about moral truths, then why alien's intuitions should line up with moral truths?
I think a naturalist could plausibly explain moral knowledge either in terms of rationalism, where our moral norms derive from some a priori discernible principle, such as the categorical imperative, so as long as our rational faculties are explainable in terms of naturalistic evolution (which they are, since being rational is evolutionarily advantageous), the naturalist could account for their moral knowledge. Or, they could be an Aristotelian, and think that moral facts correspond with natural facts, such as facts about human flourishing, or about pain and pleasure. Insofar as these are plausible moral accounts (which they are), there should be little problem for the naturalist in explaining their moral knowledge, no more than they would have in explaining the other forms of knowledge they might have, in which case moral knowledge isn’t particularly threatening to naturalism.
I don't think that the moral facts can be deduced from non-moral facts. The a priori discernible principle would have to be a distinctly moral principle, in which case the problem is just pushed back a level.
I think it's hard to see how the fact that "one flourishes when they do X," will explain why we believe that we flourish when we do X. But in any case, moral knowledge is just one example--the same point applies to inductive and modal knowledge.
Being rational is evolutionary advantageous until one uses reason to arrive at anti-natalism and convincing humanity collectively that the world would be better off without rational creatures and that it is immoral to bring rational creatures into the world.
Yeah, for example: "But on naturalism, one wouldn’t think we were any likelier to believe torture is wrong after learning that it really is wrong!" This isn't true if you believe that the "really wrongness" of torture is a natural fact relating to it producing intense amounts of pointless suffering. Upon learning that fact about torture, most people *would* come to believe it was wrong as a result. The mistake is assuming that the naturalist's moral facts are still non-natural features "layered over" the natural facts, which of course produces contradictions. Whereas when we reject that divide, no problems really arise. Not to promote myself here, but I wrote a similar piece about this a while ago: https://bothsidesbrigade.substack.com/p/you-cant-disprove-moral-realism-by
One very basic counter-argument to this is to say that either objective moral truths do not exist, or if they do exist, then they might be totally different from the moral beliefs that most humans have. In other words, your argument here seems to rest on the premise that most people do, in fact, have the correct moral beliefs, but I'm totally unconvinced that this is true.
But as I say in the article, the problem isn't specific to moral beliefs and generalizes to other e.g. inductive beliefs.
How could it be true? Most of us don't agree with each other about what's moral.
Petition to linguistically engineer it to be "the argument from inductive knowledge" or something instead!
(Also, I feel like I should address what you say about my objection.
To the first point, I worry whether this doesn't just allow the naturalist an out as well. If I can directly see that pain is bad, then surely that doesn't require theism to be reliable, just like I don't need theism to know that I'm conscious. And if it's through some intuition that doesn't give you direct awareness of the badness, then I worry whether that would be subjectively indistinguishable from an unreliable intuition given by evolution.
Perhaps it's that you have this direct awareness, but that it's very unlikely that you would have it on atheism. Maybe so, but I have a hard time seeing what part of my experience is hard to account for on atheism. Like, atheistically evolved creatures would surely also have a feeling like pain was just obviously bad! But maybe it's just because I'm more inclined to think that we have intuitions about, rather than direct awareness of, morality.
On the second point: Yup, I agree, hence the petition above!
Though I think we should maybe be a little more careful in judging which knowledge is problematic and which isn't. I don't think some fact not being able to move atoms is sufficient.
For example logical knowledge is plausibly explained by evolution: 1) All bears are dangerous, 2) that is a bear, 3) therefore, that is dangerous. This seems like a very useful thing to know, even though the validity isn't able to move atoms. I think something similar is true for all types of deduction: It would make very good evolutionary sense for creatures to be able to reason deductively.
But I agree that abductive and inductive knowledge are plausibly problematic on naturalism, so the above pedantry is probably a bit unnecessary--I just want to flag that I think the case for some knowledge being problematic on theism is not quite as straightforward as it sounds.
Anyways, a very good article!)
Thanks!
I can see that pain is bad. But seeing something has a non-natural property is not the sort of thing that naturalism can account for. This seeing is a kind of direct intuition--but on standard naturalism, this won't be explained by the facts they're about.
Yeah, that would probably get you out of it. Then I guess I doubt that we really do have such a direct insight, as I don't really find anything when I introspect on my intuitions around pain that I can't make sense of on naturalism. But it's hard to argue about those kinds of things, so let's not let that divide us:)
I think abduction and induction are also evolutionarily advantageous. Creatures that don't learn that bears are dangerous after being attacked by bears a few times eventually get killed by one.
I agree that they're advantageous, just like how morality is advantageous. The problem is that their advantage (like moral beliefs) doesn't come from their accuracy. Ancestors who did believe that the future resembled the past were successful, since it actually turned out that the future resembled the past back then. But that has nothing to do with whether the future from *now* will resemble the past--to assume that is just to assume that induction works.
Correct me if I misunderstood your point, but doesn’t the evolutionary advantage of induction come exactly from its accuracy? E.g. ancestors who didn’t believe a certain food to be harmful even after it killed many others were less successful at reproducing
Well it was accurate in the past, but no one is questioning that. The question is whether we can continue trusting it in the future/unobserved instances, and whether or not it's reliable there cannot possibly have had any influence on evolutionary history, since it, well, happened in the past.
Thank you for clarifying. Maybe I’m still missing something but assuming that humans evolved a general “induction faculty” in the past, why would it stop working with similar cases to the ones above? If the question is how can we trust induction in principle things get more complicated, philosophers have talked about it since Hume afaik and many thought that you can’t have a rational justification for it.
We will definitely still think it works, but the question is the latter that you raise of how we can trust it. Now the problem here, I think, is a bit more difficult than the regular one as evolution (without theism) gives us an explanation for why we have a very strong propensity to believe induction, which doesn't make reference to its soundness. So our belief is actively debunked--somethinf that's surely unacceptable. I'm not sure that the standard view is that we just don't have rational justification for it, though that was certainly Hume's own view.
If there's a mystery here it's "why does induction continue to work" (which I actually find convincing as a theistic argument). Our belief that it will is completely explicible
Our belief is certainly explicable, and that's the starting point for the whole argument, since the explanation doesn't cite its reliability--just like our belief in moral facts is explicable on naturalism without reference to moral facts.
The problem is that we can't really deny induction without landing in unacceptable skepticism, so we must find some explanation that saves our knowledge, and theism is one such. So it actually sounds like you agree with the argument, maybe.
Excellent work, Matthew! This argument makes a lot more sense for me now, and it seems to be extremely strong.
Surely it’s an odd choice to talk about how our moral intuition is accurate by using the example of slavery, which for the great majority of recorded history was viewed in many societies, including the most developed and religious societies of their time, as perfectly moral? I think the example of slavery absolutely should make us doubt our moral intuitions—it’s a much better example of moral intuition being socially constructed than it is of moral intuition being guided by a truly moral force.
The first problem with your argument is that your own explanation of moral knowledge is not derivative in nature. Your explanation of why you believe in moral knowledge relies on God, but you only believe in God because of an argument that starts from the assumption that your moral beliefs are true, so you don’t actually have independent justification for believing in moral knowledge, at least no more than someone who just declares that they believe in moral knowledge and that’s all the justification they need.
Your attempts to get around this also work for naturalists. if your experience that thinking on pain makes it obvious that it’s bad provides justification for moral knowledge that equally works for somebody who doesn’t believe in God.
Also at the end of the day, I think the problem of induction is similar to the brain in a vat thought experiment in that all arguments for the common sense position boil down to the fact that you don’t actually believe scepticism is true. At the end of the day, the only way to avoid scepticism is to assume that common sense and prior probability are reliable sources of information because you can’t actually make probability estimates without starting from a prior probability. At the end of the day, if you keep demanding justifications for beliefs, you’ll either have to take some beliefs on faith, accept that circular coherent it’s arguments are valid, or conclude that justifications form an infinite regress. But if you accept this, then arguments for induction that either work by assuming induction as a foundational axiom, or work through seemingly circular reasoning like arguing that we assume the world is simple because it worked in the past and an inductive world is simpler, no longer seem so unreasonable.
At the end of the day, I don’t think believing in God has actually placed you in a better position in resolving the problems you outline it’s just that your chain of reasoning, is longer and more complicated, but it still ultimately works by starting from the assumption that moral knowledge is true. In which case you can just assume that without adding God to the picture, because he doesn’t contribute anything to the argument beyond adding him to your picture.
Also, I don’t think God is actually that good of a justification for belief since he’s clearly not above misleading information as demonstrated by the fact that he has not made his existence as obvious as real world humans. if he’s not willing to demonstrate his existence, the way say real world, celebrities and heads of states do, and is willing to let lots of people persist in incorrect beliefs like his non-existence, what makes you think he won’t be willing to give you incorrect beliefs? And if he’s willing to give you incorrect beliefs, he is no longer such a solid justification for any kind of knowledge. At the your justification for moral knowledge should not be so good that it can’t explain people who have incorrect intuitions, as many demonstratably do. But if God is willing to give other people incorrect intuitions what makes you think he wouldn’t be willing to do it to you.
A good test for judging the validity of an argument for God is to ask whether it would be a better explanation to simply assume whatever needs explanation is a fundamental law of nature. After all, assuming that, for example, God wants life to exist is equivalent to assuming that the universe inevitably creates life minus the extra complications that you’re not assuming the laws of nature are an agent or have any of the other properties attributed to God. Applying this to your argument, how is your argument any better than somebody who asserts that moral knowledge must be justified so it must be a law of nature that it won’t deceive you by giving you incorrect intuitions about morality so moral knowledge is justified and hence we have solved the moral knowledge problem and proved that the universe will not fundamentally give you incorrect moral intuitions. As far as I can see, this is equivalent to your argument, except the role fulfilled by God in your argument is fulfilled by the laws of nature, and there’s nothing in your argument to favour God over this explanation.
>The fact that slavery is wrong does not directly shape selection pressures in ways that cause people to believe it.
This simply isn't true. Morality is the cultural DNA of a culture and represents generational wisdom. It directly impacts the survivability of the cultures which adopt it and therefore is itself subject to selective pressures. The reason that e.g. murder is "wrong" is that cultures that condone murder inevitably peter out in an endless series of blood feuds; you want your citizens working, not murdering. The reason slavery is "wrong" is because, in a post-industrial world, it represents both a misallocation of human capital and an affront to the individual liberties that unleash that capital. You'll note the qualification there. There's a reason it was considered moral for the entirety of human history right up until the Enlightenment, and that's because the industrial revolution radically changed the economic status of the average person. Before then it was economically reasonable and therefore morally acceptable. It's even condoned in the Bible.
Here's an interesting historical anecdote that illustrates the socioeconomic function of morality. When European explorers discovered Tahiti in the 17th century, they were amazed to discover that the Tahitians had a far higher standard of living than Europeans did. Meaning, specifically, that they had an abundance of food and plenty of leisure time - they got to lie around on the beach quite a bit instead of toiling endlessly in the fields. The reason for that is that Tahitians ritualistically murdered 4 out of every 5 babies that were born. Why did that make them rich? Pre-industrial societies live in a Malthusian equilibrium. Productive capacity was limited by agricultural output so the only way to raise living standards was to keep the population artificially low, otherwise it would equilibrate at a level just above starvation. Infanticide was Tahitians' method of birth control and that kept them rich. By contrast, Edo period Japan was *poorer* than contemporary Europe because of Japan's cultural rituals around cleanliness: that kept disease rates lower and therefore population densities were higher relative to agricultural capacity. I'm sure you've heard that the depopulation caused by the Plague led to a temporary boon for the common man in medieval Europe. It's the same principle: you had the same amount of food to feed a population that was suddenly 30% smaller.
Were the Tahitians immoral? No, they weren't. That's because moral rules aren't absolutes. They represent the cultural encoding of survivability-enhancing practices. Murdering babies made Tahitian society function better. Try explaining to them why that's immoral.
> it really is wrong to torture babies.
Uh, we routinely circumcise infants. That could certainly be construed as torture -- not by me, but the argument could easily be made -- so maybe be a little less cocksure in your references to moral absolutes.
Axioms such as modus ponens, as well as priors such as physics remaining the same over time, *must* be assumed instead of justified. I *cannot* be certain that those principles are correct, regardless of how intuitive they feel.
A simple counterexample is that I might be dreaming. I have memories of dreams during which I reasoned in ways that, in retrospect, blatantly violated basic logical principles: assumptions popping in and out of existence, my line of reasoning not so much a line as a web of disjointed thoughts. But I *thought* I was reasoning normally. My intuition of logical correctness was just as strong then as it is now.
For all I know, I might be doing the same right now as I write this comment. Modus ponens states that if p implies q, and p, then q. I believe that it’s true. But if I’m dreaming, then that definition might be wrong, in which case I might actually currently believe in some false axiom that I’ve mislabeled as modus ponens. Or perhaps there is no such principle as modus ponens and I’ve just hallucinated it.
Therefore, “insane skepticism” really is logically justified. The possibility that my most basic beliefs are false is not just some remote hypothetical but a common occurrence. And positing God doesn’t help with this: empirically, God does not mold my beliefs to be accurate while I’m dreaming.
The only way out is to assume. If my thinking is illogical, then my conclusions will be garbage no matter what I do, so I may as well assume it’s logical. If the laws of physics are about to get randomized, then there’s no way to predict what action will benefit me, so I may as well disregard the possibility when choosing how to act.
But once I admit my assumptions are just assumptions, then God doesn’t need to come into the picture.
As for morality, objective morality doesn’t exist. My sense of morality is ‘arbitrary’ in the sense that its origins can be explained by evolutionary principles rather than any deeper connection to the universe. But I’m still going to use my sense of morality to guide my actions, as well as to judge everyone else’s actions.
You say that having a natural faculty that allows us to grasp non-natural facts is mysterious, but what if both natural and non-natural truths are able to be intuited by the same cognitive system, such that evolution selecting for the ability to intuit evolutionarily useful natural truths entails as a necessary spandrel the additional capacity to intuit non-natural truths as well, because the same system is fundamentally able to grasp both of them? This is what I suspect to be the case, that intuiting the truth is selected for in brains because there are many cases in which it is evolutionarily useful, and even in cases where it is not evolutionarily useful, the fundamental truth-intuiting process we already selected for still gives us true beliefs as a nice built-in bonus.
I think this is a meaningful critique of ethical non-naturalists, but many of the morality-specific points being made don't really apply to ethical naturalists who hold that moral facts are a sort of non-reductive natural fact like facts about health. In that case, it seems like (properly interpreted, at least) it *is* the case that the wrongness of slavery plays a role in why we think it's wrong, in the same way that the unhealthiness of eating rotten meat plays a role in why we think it's unhealthy.
Otherwise, the expanded critique you're giving just seems to boil down into a general question of how priors are justified. And that's definitely a tricky question, but as far as I can tell, most of the popularly accepted accounts are not theistic anyway, right? I'm sure there is a good "you need theism to establish priors, period" argument out there, but I don't think the considerations here actually challenge those other accounts specifically.
I don't actually think that moral naturalism avoids the problem. The substantive claim of naturalism isn't that natural fact N plays causal roles X, but that N is identical to moral fact P. However if natural facts M=P instead, N and M would still play all the causal roles they actually do. And since these causal roles are what determine our beliefs, not which of N and M is identical to P, we still have a defeater.
I'm not sure I follow what you're saying here. Is the idea that we could fix all the non-moral natural facts and have it be that they were instead identical to a "swapped" set of moral natural facts? So we keep "torture causes tremendous amounts of pointless suffering" as the non-moral fact and then just swap in "Pointless suffering is good" as a moral fact? If that's what you mean to communicate, then I just don't think that's a concern that any naturalistic theory really needs to worry about - it would be like saying "Well, you could have the natural fact that arsenic causes your heart to stop and then the health-related fact that stopping your heart is healthy, but you'd still think it was unhealthy anyway." In that case, it seems obvious to me that naturalists can just say health-related facts aren't "layered over" other sorts of facts in that way but rather supervene on (or are grounded in or whatever) them by virtue of their meaning, such that the recognition of the higher-level property is legitimately influential. Why would it be different here?
Yeah that's basically the idea. I'm not sure it's unproblematic thought.
I take it that the claim of naturalism isn't just that "goodness" means "reduces pain" or whatever, but that there is a genuine property of goodness "out there" that's identical with some natural facts.
With the case of health, I would think that the reason it doesn't work is that "healthy" just means "promotes functioning of the body" or something along those lines. If the claim is that there is some genuine property of healthiness, then I just think this is a perfectly good debunking argument against there being such a property: Even if the property hadn't existed, or had been identical to some other phenomenon, we would both had written these exact sentences anyway.
Not to be "that guy" who's always linking to his own stuff, but I wrote a post about this a while ago: https://bothsidesbrigade.substack.com/p/you-cant-disprove-moral-realism-by In short, it seems to me that this sort of reasoning would allow you to reject the existence of pretty much any property, on the grounds that it's (trivially) true we would still act the way we do if it actually is the case that property doesn't exist. For example, if it's actually the case that there are no fish, then all our fish-related judgments have nothing to do with fish, and that means we'd all still act the same way without fish. But that's just another way of saying that we could be mistaken about fish existing - and that, by itself, doesn't give us any particular reason to doubt that fish *do* exist. On the other hand, someone who believes fish *do* exist and that they *do* explain our seemingly fish-related behaviors would be justified in rejecting the counterfactual claim that we would act the same way without them.
That's an interesting post, I'll have to read the paper too!
But I think that's just a reason to bite the bullet on fish, insofar as "fishiness" is a genuine property that we should at to the content of our theory, rather than a word. If the fish-theory being true or false doesn't make any difference to any of our evidence, then we have no evidence for or against it. The question is then just whether the fish-theory has a higher prior than the non-fish-theory. And seeing as the fish-theory adds an additional property to our inventory, it's surely just worse prior to looking at the evidence. Mutatis mutantis for moral realism.
In the post you make an analogy to an external world skeptic, but the situation is different here. Positing an external world allows us to explain our experiences in a much more simple way than not positing it, and so it makes up for the ontological cost. But I don't see moral properties doing any such thing.
Basically we should remove anything from our ontology that doesn't explain anything, and which complicates our worldview, and it looks like moral properties and fishiness are both in this boat.
Though I feel like I may not quite be getting your point, so please tell me what I'm missing:)
That's fair! I understand why people would have a vaguely eliminativist posture towards all sorts of second-level properties. I sometimes describe myself as a "relative realist" - I care less about the ultimate ontological status of moral facts and more about the notion that they're "as real as" the other sorts of facts that people consider uncontroversial. So if you don't believe that moral facts exist, but also that "fish facts" don't exist either, then I would say we just have a more fundamental disagreement, but one that doesn't really impact the relative structure of our ontological hierarchy. Whereas when someone is a realist about all sorts of things, *except* moral properties, that I think is more open to this sort of objection.
For what it's worth, I tackle the explanatory value of moral properties as a concept in the follow-up post: https://bothsidesbrigade.substack.com/p/the-world-makes-more-sense-if-youre
You might be right about moral naturalism. It's tricky to think about. Even on moral naturalism, it seems hard to see how the wrongness of torture would be responsible for us believing that it's wrong.
It's not just a "general question of how priors are justified." It's a reason to think that priors are not justified absent theism.
Why is that hard to see? It doesn't strike me as any weirder than saying (to use the same sort of analogy again) that the healthiness of exercise is responsible for us believing that exercise is healthy. In both cases, you've got a higher-level natural property fulfilling every sort of role a good explanation should, with explanatory power not necessarily possessed by any particular set of first-level natural facts it could be reduced to. So that, to me, makes it a perfectly reasonable (and actually indispensable) explanation for our beliefs. I lay out my case for this in detail here: https://bothsidesbrigade.substack.com/p/the-world-makes-more-sense-if-youre
And sure, I get the argument and I think it's interesting. I'm just saying it's not really an argument from moral knowledge in the traditional sense anymore, and more just a straightforward argument from the justification of prior probabilities.
Agree on the second point.
But why trust our moral intuitions over those of the hypothetical snakes that might have evolved and had defective moral judgments?
Well, I don't think it's particularly problematic for the snakes to have different moral intuitions - snake flourishing is distinct from human flourishing when it comes to a bunch of little contingent details, so they probably would feel differently about some things. But if those snakes developed legitimate personhood, then I think the broad considerations we make would develop in them too. Of course, it's always possible that evolution would systematically mislead some groups and not others for various reasons. But that's just an expression of a general skeptical urge, not any particular issue with moral knowledge. And it's one a theist can't really avoid either, imo - at some point, everyone needs to accept that they could be wrong, generally.
It makes sense to say that every system takes something on faith (or intuition, as you frame it). Lewis Carroll’s Achilles and Tortoise paradox uses a similar (but shallower) approach to show that even simple logic requires acceptance on faith that logic means what we say it does. Similarly, a probabilistic approach to truth only works if one trusts probability. But I don’t quite follow the objection that it’s too great a coincidence in a naturalistic framing for us to have intuited moral facts that happen to be true. What does “true” mean in this context, with no theistic frame of reference? It seems like the more likely outcome of this system is just to slide into extreme moral relativism, in which no moral truths exists beyond what we intuit on an individual level, or can justify (imperfectly) by demonstrating its practical utility.
I'm not raising a general skeptical challenge. My point is that on naturalism, many of our beliefs are not explained by the facts they're about, which removes our justification for believing them. This applies not just to morality, so even if you're a moral skeptic, the same point applies.
Naturalists aren't committed to there being nonnaturalist moral facts. This isn't a problem for naturalists in the same way nonnatural fashion facts or demonic facts about possessions aren't a problem for naturalists, even though naturalists can say there are still true facts about circumstances in those domains.
But you'll still have problems with e.g. induction.
Thanks for clarifying, and agreed on the broader point. All may not be as “vanity, vanity” as Ecclesiastes posits, but it gets pretty close without a core axiom or two to anchor things. With that in mind, though, why not raise a general skeptical challenge? You conclusion is “…therefore theism.” Why not “…therefore theism or epistemic sludge?”
Yeah, I don't think many philosophers would argue that non-natural moral facts exist *and* that evolution directly led us to contact with them. Usually the combo is "natural moral facts + evolutionary pressures" or "non-natural moral facts + cognitive faculties."
Moral intuitions are not the basis of moral knowledge because intuitions don't exist and the reporters of these intuitions are misled/deluded like people saying they have a divine sense or can talk in tongues or can see auras or read horoscopes or speak to the dead or astral project.
Morality did not evolve, most moral behavior is because of enculturation and agents seeking to construct systems of social behavior by which to survive and thrive among other agents.
I think you mishmash two quite different things together. One is moral knowledge. And the other is Münchhausen trilemma and overall problems of justification of beliefs.
Regarding moral knowledge in particular, Silas Abrahamsen's critique is completely on point. I talk about the similar thing in this comment: https://benthams.substack.com/p/reply-to-scott-alexander-once-more/comment/96040537
On the other hand, when you jump to overall problem with justification of beliefs, the notion of "justification" that you use can't be satisfied by theism just as well. Let's use the example from your post:
> 1. Tomorrow physics will continue as before.
> 2. Tomorrow physics will break down wildly and become completely different. However, every day before tomorrow, physics will work consistently in the ways described by the laws of physics.
Why do we think that the first hypothesis is more likely on theism? Because God is good and therefore would like us to have stable physics and accurate beliefs. But how is this belief itself is justified? Consider two hypothesis:
1. God is good and wants us to have accurate beliefs and therefore gave us our intuitions and tomorrow physics will continue as before
2. God is evil and wants us to have inaccurate beliefs and therefore gave us our intuitions and tomorrow physics will break down wildly and become completely different. However, every day before tomorrow, physics will work consistently in the ways described by the laws of physics.
If we can't just claim that the second hypothesis has lower prior based on our intuitions about simplicity of hypothesises, which itself can be a delusion from an evil God, then our preference of first hypothesis is not justified on theism
The same principle can be applied to every step of reasoning, for instance:
1. God is good and will continue being good and so tomorrow physics will continue as before
2. God has been good all this time but will stop being good tomorrow and so tomorrow physics will break down wildly and become completely different.
Or, and probably my favorite:
1. The concept of goodness implies desire for us to have accurate beliefs and stable physics
2. The concept of goodness implies desire for us to have accurate beliefs and stable physics up untill tomorrow, and then it will imply desire for us to have inacurate beliefs and unstable physics.
And God is open to the same design objections as for induction. What explains God's wants for induction to keep holding staying the same in the future as in past cases? If it's induction separate from God, then God's existence is pointless. If it's something about God, then there's a circularity or unexplained brute fact, and we can make our theory simpler by just removing God from the picture and giving that property to something we already know exists like the universe.
Haven’t had the time to fully read section 6 but just a small point on modal knowledge: I don’t think this works for logical modality. The logical and mathematical facts plausibly stand in some certain intimate relation (not causal, but something. maybe something like constitution) to the natural facts such that it’s plausible on naturalism that the evolutionary explanation for our logical and mathematical faculties includes those facts themselves. But if we’ve got reliable mathematical and logical faculties we’re gonna get logical modality. That’ll also get us nomological modality. But yea metaphysical modality no, but that’s just in the bucket of metaphysical claims. I also think we should be kinda skeptical of metaphysical modality anw
I don't understand the inductive skepticism point. How does theism not face something exactly parallel? Just consider the hypothesis of a deity who changes all the laws of physics tomorrow and plug that into the comparison principle you spelled out: both the Skeptical Deity and the Non-Skeptical Deity's actions would explain our intuition that the sun will rise tomorrow equally well, so we have to be neutral between them. Surely at best both of us can equally wriggle out of countenancing the relevant skeptical explanations of our experiences by making use of (say) simplicity, and at worst we're both equally screwed.
It also seems like you're switching between two very different principles and treating them as the same. One is the one you explicitly articulated: intuitions can't justify belief in A over B if they're equally likely to be intuited under both hypotheses. This is clearly false, at least if you reject skepticism. (It's actually kind of true if we model intuitions as Bayesian evidence that's supposed to update our priors, but we shouldn't do that. Rather, we should just treat intuitions *as* our priors, essentially.) The other is a causal debunking principle: your intuition can justify your belief in hypothesis A only if the causal explanation of your experiencing the intuition has "something to do" with A being true. (Notice that this principle doesn't intrinsically involve a comparison with a different hypothesis B.) But what "something to do" means here is unclear, and in fact "explains" is unclear. Suppose the standard picture of naturalism is true. The sun rises or explodes according to a certain set of dynamical physical laws. My belief-forming processes in my brain also operate according to those same laws. Therefore, the causal explanation of my belief in the sunrise tomorrow has "something" to do with the process of the sunrise. If this is an insufficient causal connection, then sufficiency needs to be explicated in more depth.
There's an even deeper worry in the background about how God is supposed to know any of this stuff, as well. I don't think appealing to God's necessity is helpful - it wouldn't do for naturalists to appeal to necessitarianism (which I'm in fact sympathetic to, not just hypothetically). And if instead you want to go with the simplicity of omniscience as a property, I think that view faces serious problems also on vaguely Tarskian lines.
Sorry for the late reply.
You don't have direct evidence from intuition for a Non-Skeptical Deity. Instead, you deduce a Skeptical Deity probably isn't real from the fact that it's more complex and violates plausible moral constraints. Those beliefs you have because they're true. This also addresses the second broader skepticism worry.
I think that there are different ways of formulating the principle--it can either be about relative probabilities of explanation. If it's about explanation, it still won't be that the sun rising tomorrow explains your beliefs about it, only that the things that make the sun rise tomorrow explain your beliefs abotu it.
God knows these things because of his constituent omniscience.
> Instead, you deduce a Skeptical Deity probably isn't real from the fact that it's more complex and violates plausible moral constraints.
As I mentioned, the complexity argument applies to the skeptical naturalistic hypothesis re: sunrise just as well as the skeptical theistic hypothesis. The moral constraints you consider plausible are only so because of your intuitions, but those are also victim to your comparison principle (they'd be equally expected if the skeptical deity is deliberately feeding you faulty moral intuitions, and if he is in fact "morally simpler"). For that matter, maybe we start questioning our intuitions about simplicity, which is, after all, a highly philosophically contentious concept...
>I think that there are different ways of formulating the principle--it can either be about relative probabilities of explanation.
But those are different principles, not reformulations of the same thing! Not that that's a big deal, but I think it's important to highlight that they're different, not least because one of them (the comparison one) almost certainly implies some sort of skepticism unless heavily elaborated upon and refined.
>If it's about explanation, it still won't be that the sun rising tomorrow explains your beliefs about it, only that the things that make the sun rise tomorrow explain your beliefs abotu it.
The things that make the sun rise tomorrow - the general physical regularities - explain my belief in those general regularities, and from those together with my belief in past sunrises (each explained by watching the sun rise, in fact!), I deduce that probably the sun will rise. Why isn't this a sufficient story? And how are you doing better here? On your view, it's not tomorrow's sunrise explaining your belief directly, it's only through a similar divine intermediary who is himself what explains the sun's rise (since he decided to make the sun do that).
>God knows these things because of his constituent omniscience.
Sure, but I don't think it's at all clear that omniscience should be counted as simple. In the worst case, it's not a hundred percent clear that it's even coherent to talk about "all truths" for Tarskian reasons, much less "knowing all truths," and probably even in the best case more modern attempts to evade the semantic paradoxes by revising our logic aren't going to end up furnishing you with something simple as opposed to hideously complex. (Or at least the ones I know something about, admittedly a far from exhaustive sample.)
Not going to continue the back and forth because I have too many commenters, and though you are among the most thoughtful, I can't spend all day on this. Just one brief comment.
//And how are you doing better here? On your view, it's not tomorrow's sunrise explaining your belief directly, it's only through a similar divine intermediary who is himself what explains the sun's rise//
God grants us the ability to grasp non-natural facts, including that simpler worlds are likelier than more complex worlds. From this, we infer the sun will probably rise.
>Not going to continue the back and forth because I have too many commenters, and though you are among the most thoughtful, I can't spend all day on this. Just one brief comment.
No problem, though I'll respond anyway just to provide food for thought.
>God grants us the ability to grasp non-natural facts, including that simpler worlds are likelier than more complex worlds. From this, we infer the sun will probably rise.
(I originally wrote a response here that invokes evolution, but I decided to edit this comment and revise that because I think there's a better and more direct answer.)
If it's true that this world is simple, and if it's true that simpler worlds are objectively more probable (whatever that means), then the latter fact partly explains why this specific world, with all of its inductive regularities, exists as opposed to some other, more complicated one. And this inductively regular world existing explains just about everything about my beliefs, since I'm a part of it.