Locating Myself Metaethically
This is a new topic for me. I am just experiencing a spark of interest in metaethics. What follows will be a high-level tour of the most popular and most documented metaethical positions and my initial reactions to them. Many of the positions covered here will be entirely new to me and I will not have the space or time to completely make my mind up about many of the arguments. Given the sheer volume of fascinating work that exists on metaethics, it would be irrational to expect that I would walk away from this tour with my mind completely made up.
My goal here is to chart this territory for my own views. I hope that that goal can also serve the purpose, for the experienced reader, of a fresh perspective on familiar arguments and the purpose, for the reader also inexperienced in metaethics, of a general overview.
From a 30,000ft view, at least to my awareness, there are five major metaethical positions.
Moral Anti-Realist Positions
The first three fall under moral anti-realism: "There are no objective moral values"
- "Ethical statements are neither true nor false"
- Non-cognitivism
- "Ethical statements are always false"
- Error theory / nihilism (cognitivism)
- "Ethical statements are true or false but dependent on observers' attitudes"
- Subjectivism, and/or relativism
Moral Realist Positions
The last two fall under moral realism: "There are objective moral values"
- Naturalism: moral properties are natural properties. Ethical truths can be reduced to descriptive truths.
- Non-Naturalism:
Most often, ‘non-naturalism’ denotes the metaphysical thesis that moral properties exist and are not identical with or reducible to any natural property or properties in some interesting sense of ‘natural’.
(I will frequently quote the SEP pages for both naturalism and non-naturalism for exposition, links can be found at the end of this doc. I'm not being strict about citations)
I will iterate through each of these positions and try to turn a critical eye toward the common arguments for and against them, hoping to eventually find what I consider the most plausible view(s).
Quick Meta-Discussion First: Who Cares?
But first, the all-important "Who cares?" question!
You might think, or Skepticus might say:
![User Icon](/img/Skepticus.jpeg)
As I discuss with common criticisms of metaphysics, here, I think that is a dangerous attitude. Our beliefs, even the most abstract theoretical ones, will shape our thinking and acting and the longer they have influence over the discourse the more impact they will have. Presumably, ethics of all places should be an important focus for shaping how we view our actions and their relation to the world.
Moral Anti-Realism
1. Expressivism
Expressivism is a non-cognitivist moral anti-realist view that seems somewhat convincing at first pass. The claim that ethical statements are expressions and are thus neither true nor false (not truth-apt), has an intuitive thrust. It may be the case that moral statements are merely illocutionary acts and as such have no underlying truth. Without the "merely", the claim is almost trivially true. Of course, in my saying "You shouldn't kill animals" I am also expressing frustration or anger. It is a step further, though, to say that moral statements don't have any underlying truth value and are just that expression.
I'd like to approach considering expressivism by examining how we typically tell the difference between mere expression masquerading as a description and real description. No doubt, there are such sentences like:
: "I would rather die than eat pistachio ice cream"
Also, consider the following sentence:
: "It is wrong to kill children"
Now the expressivist wants to say that, in , we are just expressing that we "don't like the killing of children" or something. Just as the speaker of does not literally want to die.
Ultimately, I think the expressivist is making an empirical claim or a ridiculous one. Either they believe that some empirical facts should convince us that all moral statements are of the form "I have some disposition X" or that we should just believe an otherwise unintuitive conclusion, setting aside any empirical evidence, about moral statements. I could be missing something, but I fail to see what claim could be made in support of the expressivist claim that isn't an empirical one. Especially since the argument is typically motivated by evolutionary debunking or "you just feel that way" sort of claims. Either that's "what we are doing when we make moral claims" or it isn't and that fact of the matter will be an empirical one.
Now consider the following two factual sentences. I have added two types of factual language: naturalistic (based on empirical science) and non-naturalistic (based on intuition or some deeper truth).
: A neutral hydrogen atom has one electron.
: A quadrilateral has four sides
As I hope to show with these sentences, there are some differences between figurative, moral, and factual language that make the interpretation of moral language as pure expression seem improbable. If the expressivist is right, empirically speaking, our dispositions to expressive language should be more similar to that of moral language than that of factual language. I will now examine these two sentences along with and to see 1: if we hold others accountable for their truth, and 2: what risk we are willing to take on to defend them.
Figurative or expressive language
- For , we don't care if others like pistachio ice cream or not. We don't hold others accountable for mere preferences.
- For , We don't typically start altercations or go to great lengths to stop someone else from enjoying or disliking something.
Moral language
- For , we tend to hold others to the same standards as ourselves. We tend to find others inconsistent or irrational for not believing it.
- For , we will go to incredible lengths, perform heroic acts, etc. to uphold these beliefs.
Factual language (Naturalistic)
- For , we think someone is inconsistent or irrational for not believing it once they've been shown the evidence
- For , we will attempt to convince someone or persuade them otherwise, because it seems intuitively inadvisable not to.
Factual language (Non-naturalistic)
- For , we think someone is inconsistent or irrational for not believing it regardless of evidence.
- For , we will attempt to convince someone or persuade them otherwise, because it seems ridiculous not to.
I think these characterizations of four types of language should be pretty uncontroversial and fair. Based on these, it clearly seems to me that moral language like shares much more with factual language than it does expressive language. As an empirical matter, it seems common sense would say that the expressivist is wrong here.
Additionally, before moving on, I would like to point out that "Is true" is a linguistic modifier (of whatever kind) that if tacked on a true sentence, or a sentence intended to be true, strikes us as redundant, as in: "It is Tuesday and that is true". So consider the same for a moral claim: "It is wrong to kill infants and that is true". To me, that seems equally redundant in both cases. Of course, the expressivist can claim that its "seeming" is not sufficient. Maybe, since I think the only fight to be fought here is an empirical one, there is some science existing now or in the future that could tell us what module in the brain is doing what etc. but I take no more convincing at present.
Other Non-Cognitivisms
Non-cognitivism, another form of "neither true nor false" moral anti-realism, can come in many forms. All that is required to qualify as a non-cognitivist is that one believes that moral statements do not or cannot have substantive truth conditions. In this sense, the expressivist is a brand of non-cognitivist. For example, we could hold that "it is possible that there are some moral truths" but also hold that we could not have epistemic access to them and so moral statements cannot be meaningfully connected to such truths, even if they did exist, and therefore can be neither true nor false.
I am somewhat convinced by the idea that moral "facts" are just an evolutionary adaptation and that it is unclear what could ground them as "really true" in a more meaningful sense. I would say that of all of the anti-realist positions, a kind of non-cognitivism that is agnostic about the actual existence of moral truths is the most persuasive to me. This boils down to the claim that "X is good" is like saying "It is true that X is good", but just has no basis in any fact-of-the-matter real thing. This version of non-cognitivism could endorse some quasi-natural explanation but also say that we have no actual way of knowing currently.
2. Nihilism / Error Theory
"Ethical statements are always just false."
A Note On General Nihilism
The claim of Nihilism, generally, is that there is no reason to take one action over another whatsoever and therefore all moral claims are false. For the Nihilist, a statement like is just flat-out wrong.
If this includes belief and/or rationality, then I think the position is obviously self-defeating. Here it is simply in an argument, not because I think it is unobvious, but because I have had many a campfire discussion where this point fails to make it across:
There is no reason to believe one thing over another, no exceptions.
Nihilism is one thing we could believe.
There is no reason to believe nihilism.
Unless the general nihilist can explain why beliefs about nihilism should be exempt, this argument seems hopeless to me. If you'd like to bite the bullet and be a general nihilist Wisenheimer, then you cannot also be rational.
I have yet to find a coherent formulation of this kind of nihilism, once I do, I am all ears. Until then, this theory does not cut the cake.
Specifically Moral Nihilism
More specifically morals-focused nihilism make a more coherent claim:
The world has no inherent moral features.
If the world has no moral features, then moral statements cannot be true.
Moral statements like are not true.
This differs from the general nihilist claim in that it is specifically targeting only moral normative statements, not all normative statements of any kind. The proponent of this view is, then, not subject to the self-defeating nature of the general nihilist's argument.
Typically, the move is for these moral nihilists to be fictionalists or error theorists. Error theorists claim that all moral claims aim to be true, but fail for various reasons. Fictionalists claim that all moral claims are false, or not truth-apt (as in "The present King of France is Bald"), and fictionalists add that we can believe in these things as something like "useful fictions". This differs from the expressivist or other non-cognitivist views in that this view wants to say that these claims are just false. Full stop. These positions are cognitivist. They think that moral claims are capable of being true or false and they are false. Fictions, however useful, are still false.
It is worth noting that to the strict epistemic pragmatist, a useful fiction might be the best we get. While I am somewhat of an instrumentalist in the Quinean sense (which differs greatly from all-out truth-pragmatism), I think it is a bad idea to just start claiming anything we can't explain is a useful fiction. I think that sort of "useful fiction" thinking should be saved for the absolute furthest removed things from experience and all other available judgments if it is used at all.
The strongest argument I have encountered in favor of error theory is from J.L. Mackie's Ethics Inventing Right and Wrong in which he gives what is called the argument from queerness:
If there were objective values, then they would be entities…of a very strange sort, utterly different from anything else in the universe.
They would have to be so odd, Mackie argues, because they have a supposedly "reason-giving" property that nothing else in the world has. Mackie then claims that there are no such features and that such features are the only thing that could make a moral theory true. Therefore, all moral claims are false.
I disagree with this conclusion on a few points. My first objection is that it actually seems like an overly complex hypothesis. It seems more probable that truth-apt seeming language is truth-apt and that we just don't know what makes it so, than that we are walking around acting constantly on something obviously false.
The second issue is that I think it is hasty to claim that "no entity can fit the bill" in the case of morality. I am very skeptical that "weird metaphysical or natural things" are the only thing that can be what we quantify over when we make moral claims.
A third issue is that I think that making some claim which is so not born out in action is a hallmark of goofy philosophical theories. Here is one of my favorite quotes on the subject:
"Don't explain your philosophy. Embody it."
-- Seneca
How many moral nihilists or error theorists seek to "embody" their philosophy, metaethically speaking?
A fourth and more serious issue to me is that I think that there are features of the world (whether natural or not) that make at least some moral claims true. I am not sure what they are, but I am sure they are true (see the next section). If the world has undeniably true moral features, then claiming that "there are no such features" is obviously false. I hope the reader will permit a quick detour. Here is a thought experiment for such a case:
My main positive metaethical argument
Imagine two worlds:
whatever conditions would make each being's existence, respectively, the state which that being would most avoid out of any possible state that being could choose or be exposed to, optimize the nearness-to-that-state for every being in that universe.
whatever conditions would make each being's existence, respectively, the state which that being would most seek out of any possible state that being could choose or be exposed to, optimize the nearness-to-that-state for every being in that universe.
(However we count "Beings" will divide out as long as it is the same in both cases.)
We could also maybe add "most seek/avoid given full information", but I do not think that is necessary here. My goal is just to demonstrate that I think there is at least one objective moral truth.
It seems obviously true that World 2 is better. If World 2 is better and is otherwise identical to World 1, then there must exist some feature (whether a feature of the 'natural' world or not as long as it is common to both) of either world that can explain this being the case. I do not think it is possible to rationally prefer the World 1 for any metaphysically-possible being. Therefore, there is at least one moral truth.
It is also important to note that this does not yet secure any kind of normativity. One could agree with this thought experiment, yet reject that it gives us any reasons to act.
For now, all I am claiming is that "World 2 is better (has more good) than World 1" is an objective moral truth. I want to say that this is self-evidently true, but I will grapple with that later.
![User Icon](/img/Bloviatius.jpeg)
Let's not get carried away here. Maybe it's a bad sign if this guy likes the idea. There are doubtless issues with "most seek out of any possible state that being could choose or be exposed to", but I will expand on that later and focus on touring more of the metaethical positions out there.
Between my thought experiment above and my general objections above, I am so far unconvinced regarding error theory and moral nihilism.
3. Relativism
Perhaps the easiest refuted of the moral anti-realist theories listed here, moral relativism claims that morality is completely subjective and agent-dependent.
The main issue here is that I simply have not heard convincing argument for subjectivism and I have an already convincing reason to believe that it is false.
Here are some of the common arguments I have seen/heard. I will try to be somewhat charitable in adding additional as-plausible-as-possible premises:
Opinions differ?
People differ in opinions all over the world on moral truths.
If people differ in opinions on moral truths, there cannot be moral truths.
There are no moral truths.
Here, 2 is obviously false. We differ in opinions on all kinds of things. It simply does not follow that those things we differ on are true or false. Not even the softer version, that maybe they are "less probable to be true", flies.
I would grant that moral disagreement can be evidence against realist views that claim the something unified is causing moral beliefs, but that is a weaker rebuttal to a specific argument, not a general argument for relativism.
We should tolerate everyone?
It is wrong to call other culture's/people's ethical views wrong.
If we were all right we would all be wrong. (from 1, 2)
There can be no true moral truths.
I honestly struggled to get this one into anything like a valid or sound argument. I tried because I hear it a lot. It may be disrespectful to hastily judge those of differing opinions, but that says nothing about the moral status of their opinions, beliefs, etc.
It should be clear from the case I laid out in my positive position that I think that at least one (core) moral principle is objective.
Moral Realism
After a tour of the anti-realist positions, and having found none convincing, mostly because of my positive case for there being "at least one true moral principle" and the weakness of some of the views, it is time to turn to moral realist positions.
Moral realism asserts that there are moral truths and that therefore it is possible that some moral claims are true. Moral realists agree that moral statements do aim to be true about the world and that some of them are in fact true, but they disagree as to how these statements are true.
What is constitutive of morality? What makes it true metaphysically and, whatever that is, how can we come to know about it? The difficulty of this question no doubt has done much to fuel all of the above anti-realist positions.
There are two routes we can take in answering these questions. Naturalism and Non-naturalism. The Stanford Encyclopedia of Philosophy's Moral Naturalism article starts with the following:
There may be as much philosophical controversy about how to distinguish naturalism from non-naturalism as there is about which view is correct.
With that quote in mind, and the complexity of the issue, for my purposes here, I will differentiate very generally based on the following descriptions:
Naturalism: The view that morals are somehow reducible to some naturalistic property of the world (think science)
Non-naturalism: The view that moral philosophy is separate from natural science and that its methods are different. (think mathematics-type abstract reasoning)
Since I failed to find anything appealing in the moral anti-realist sphere, I will spend a bit more time here.
The blueprint for this part of the tour to "locate myself metaethically" is taken right from the SEP's pages on naturalism and non-naturalism. Without such a guide it might be a pretty confusing first pass to say the least.
Before diving in to all of the positions, I want to make a few things about my own views clear first:
I think that there is at least one moral truth in the world
- Based on the thought experiment I put forward, the result of which I hesitantly called self-evident.
These are the rough positive criteria that I think any moral realist theory has to satisfy:
- 1. The true metaethical theory explains how moral facts interact causally with our moral language.
- 2. The true metaethical theory explains moral/non-moral supervenience.
- 3. The true metaethical theory explains how we can know moral facts.
- 4. The true metaethical theory has room for normativity.
- 5. the true metaethical theory accounts for evolutionary debunking.
A Note on Supervenience
It is relatively uncontroversial in metaethics, and I certainly agree, that the moral supervenes on the non-moral. That is to say, that there can be no change in the moral without a change in the non-moral. If two possible worlds agree on all the non-moral facts, they must also agree on all of the moral facts, but not vice versa. This will become an important observation shortly.
Naturalism
Naturalism is the moral realist view that moral facts are natural facts. Typically "natural facts" here means facts discoverable by science.
Here is an excerpt from the SEP page on moral naturalism that does an excellent job of summing up the mission of moral naturalism.
Moral naturalists are committed to the idea that moral facts are a kind of natural fact—but which natural facts are the moral facts? If moral claims are synonymous with certain natural claims, we can know by conceptual analysis which natural facts are the moral facts. And if moral facts are the facts that causally regulate our use of moral terminology, that provides us with another way of investigating which facts are the moral facts. But if neither of those stories are available, it seems that we have no way of identifying which natural facts are the moral ones (Huemer 2005, chapter 4; Bedke 2012).
So, it appears, this strategy can go one of two ways. We can have moral-to-non-moral identity analytically, by synonymy, in which case we gain knowledge of moral facts by examining the meanings of words/concepts with relation to natural phenomena (analytic naturalism), or we can have moral-to-non-moral identity synthetically, by seeing states of affairs in the empirical world that causally give rise to our moral language (synthetic naturalism).
Analytic Vs Synthetic Naturalism
Analytic Naturalism
Analytic naturalism is the view that some moral properties are self-evidently equivalent to some natural properties. For example, "Pleasure is good". It seems my thought experiment from above falls into this camp and it would be something like "a state in which the least preferred or most avoided states maximized for all beings is maximally bad and its opposite is maximally good".
This approach faces a famous challenge called the "Open Question Argument" from G.E. Moore. The argument looks like this:
For moral property X, The question "Is it true that X is good?" is not meaningless (An "open question")
X is not analytically good.
This argument is, broadly applied, a challenge to the analyticity of any possible analytic definition about anything. It highlights the key problem with analytic naturalism well. If some non-moral property is to be analytically equal to some moral property, it has to do so unequivocally.
I see this argument as somewhat side-step-able for several reasons. Most important among these reasons to me is just that analytic truths can be non-obvious. See the Paradox of Analysis Consider:
K is a second cousin of L if and only if K is a relative who shares a great-grandparent with L
This sentence is analytically true, but to someone who has never heard of a second cousin it is non-obvious. Simpson's paradox refers to the phenomenon where a trend appears in different groups of data but reverses or disappears when the groups are combined. One can put together a very non-obvious exposition of this which is analytically true.
Second, I don't draw a hard line between analytic and synthetic and agree with Quine that there is just a gradient of connection to experience, rather than a hard cutoff. Maybe the second cousin case is a good example of that.
The most convincing argument, setting my own aside, for analytic naturalism to me is Frank Jackson's moral functionalism. Here is an excerpt from the SEP article on moral naturalism outlining his position:
Jackson believes that ethical properties are natural properties or, as he prefers to say in this context, descriptive properties. His argument for this appeals to the supervenience of the moral on the descriptive. This is the claim that no two completely specified situations that differ in their ethical properties can be exactly alike in their descriptive properties. Supervenience entails that no two worlds that are exactly like each other descriptively can differ ethically, which means that ethical properties and descriptive properties are necessarily coextensive.
This boils down to saying something like:
This is the most satisfying account of moral naturalism to me and I agree with Jackson that there is some set of natural properties and relations that is necessarily coextensive with some ethical property. However, there is a major problem with a strategy like this. When do we know we have the right description? There is one right description, I think Jackson's argument shows that convincingly, but how would we know we had it and furthermore what use would it be?
To illustrate these concerns, let's imagine we have a perfect computer simulation of events with some moral content that would fill out Jackson's network of properties and relations and we let it run for some amount of time. We stop the simulation and take a look at our what our model, which we based on Jackson's theory. There are bound to be some inconsistencies. It would be more surprising if there weren't. So how do we know which side to take when our model encounters an inconsistency? The proposed strategy is to let this observed morality be "matured" and analyzed and then decide on the right way to choose.
I actually think, considering that Jackson's argument is the best I have encountered so far, that this is a bit of a reductio ad absurdum of the entire position of analytic naturalism for me. Here are my reasons: Ultimately what Jackson's view generates is a model that predicts (by describing) what a moral property is based on current views, actions, relations, etc. We try to get to the "right set of relations and properties which must be coextensive with some ethical property", but I don't think we have a way to know we are there. If there is a discrepancy, ultimately we just appeal back to intuitions and it makes this theory seem barely predictive to me. I am not saying that predictiveness is Jackson's goal. He sets out to describe, but the end goal is a model of what some ethical property is. The only way a view like this can give us normativity is if we take these descriptions to be accurate definitions or at least fill out some more abstract notion of what the "out there" moral truth is. In this case the definitions are only as accurate as our agreement on them which makes the case seem hopelessly circular and not normatively useful.
My own views so far face a similar problem. Say we have these optimally "better" and optimally "worse" states of the world, what good could that do us in telling us how to act? Should we just act to maximize that state? How can we reasonably know what would serve to maximize that state? I think benefit in the potential-normativity department for my view over one like Jackson's is that at least in the case of my view we at least have some potential moral guidance that is significantly greater than the some of observations-to-date.
Synthetic Naturalism
Synthetic naturalism is the view that some moral properties are identical with some natural properties, but that this is knowable only empirically. For example, that water is H~2~0. It is important to acknowledge that this view is saying that the moral facts which are somehow in the world are what causally (or otherwise) drive the meanings of our moral terms, in much the same way that "H~2~0" in the world causally drives the meaning of "water".
Cornell Realism
The most convincing version of synthetic naturalism to me so far is Cornel Realism. Roughly, this is the view that moral properties are just empirical observable properties with complex causal profiles. An example of another such property might be "being in shape". Someone is in shape if their resting heart-rate is low, they are physically fit and capable, etc. In much the same way, this view says that "goodness" is just such a property in relation to an observed act. In other words, "goodness" just is the distributed complex properties that make up certain actions.
This view is appealing for a few reasons. Claiming that moral facts are merely complex empirical properties has us in keeping with a naturalistic methodology and also allows us to explain ethical supervenience easily.
The problem here, and indeed the problem for many synthetic naturalist accounts that I have seen so far, is how these terms get their meaning. Consider how "Water" gets its meaning. "Water" means what it means to us because there is some thing in the world, namely H~2~0, that causally gives rise to our reference of water.
In Hilary Putnam's Twin Earth thought experiment, he challenges the notion that brain states can be sufficient for describing a term's meaning, by asking us to imagine an otherwise identical Earth, save for the fact that on "Twin Earth" water is XYZ and not H~2~0. If meaning is in your head, then we have to say they are referring to the same thing, which seems obviously false. Critics of synthetic naturalist metaethical theories like Cornell Realism have deployed a similar criticism with respect to the meanings of ethical terms.
If Cornell Realism is correct, and the meanings of our moral terms are causally regulated, then the same things should be causing my meaning here on earth, as they would on "Moral Twin Earth". However, if "Moral Twin Earth" denizens disagree with us, the Cornell Realist here is committed to saying that ethical terms refer to different things on "Moral Twin Earth". This is absurd though because we were seeking to prove that the thing that causally regulates the ethical term is the meaning of said ethical term. So if it differs, it is not clear how this disagreement is to be explained along with the definition proposed. This SEP excerpt does a good job of summarizing the problem that persists even without "Moral Twin Earth" criticisms
Assume that causal reference theory is true for moral terms, and assume that different properties causally regulate the use of moral terms in different societies. If both of these assumptions are true, it follows that moral terms refer to different things in different societies. And it does seem, empirically, to be the case that different properties causally regulate the use of moral terms in different societies. Thus, if causal reference theory is true, cross-cultural moral disagreement should be impossible.
As the article also points out, things don't look so good if we move away from a causal theory of reference either, as we likely end up back in analytic land.
Summarizing Naturalism
The big issue I seem to have with naturalism is this: I can agree that moral facts must, in some important sense, be natural facts but if that is the case which natural facts are they and why? In my tentative view, they are the natural facts that (on an extremely high level of description) make this universe better than that one.
- Which natural facts?
- Analytic Naturalism: "These ones that are analytically moral ones."
- Synthetic Naturalism: "These ones that are causally related to moral language."
- Why?
- Analytic Naturalism: "Because they are self-evidently equal in meaning"
- Synthetic Naturalism: "Because they cause moral language."
So far, I am not knock-down-convinced of any of these arguments, even my own. So far in this adventure, though, I am persuaded by this line of reasoning:
Consider the following two hypotheses:
: moral properties are natural properties
: moral properties are not natural properties
If the moral supervenes on the non-moral, then ethical properties are coextensive with some non-ethical properties. (call the claim "ethical properties are coextensive with some non-ethical properties" )
The moral supervenes on the non-moral.
If ethical properties are coextensive with non-ethical properties, then (this hypothesis is strongly confirmed by explaining supervenience)
Ethical properties are coextensive with non-ethical properties, or " is true" (from 1 & 2)
Therefore, (from 3 & 4)
Again, the criteria are:
- 1. The true metaethical theory explains how moral facts interact causally with our moral language.
- 2. The true metaethical theory explains moral/non-moral supervenience.
- 3. The true metaethical theory explains how we can know moral facts.
- 4. The true metaethical theory has room for normativity.
- 5. the true metaethical theory accounts for evolutionary debunking.
So far the above argument is clearly doing well at 2, but poorly at the 4 and 3. 4 seems difficult because "Woopee" if I can explain supervenience if normativity seems hard to build on what I have. 3 seems difficult because it is not clear, even if moral facts are natural ones, how we come to know which are which, and the existing accounts I have seen get us into trouble. My own tentative view is that I came to my conclusion like I would come to any self-evident conclusion, by pondering the meanings of words.
I think the real damning thing to me about naturalism is that it missing the boat on normativity in general. The old "is-ought" gap? It is the idea that it seems hard for a description of the natural to lead a normative claim. That is not to mention the numerous potential problems with my thought experiment which I will hopefully get to explore elsewhere in greater detail.
Non-Naturalism
Non-Naturalism is the moral realist view that moral facts are facts, but they are not reducible to some natural facts. Non-naturalists usually appeal to irreducibility of some kind. These non-natural properties may be universals of some kind or the non-naturalist may claim that they are irreducible properties of some other non-natural sort. My own position could be shoehorned into this camp by claiming that it is not reducible to a feature of the natural world, it instead relies on some irreducibly normative concept.
Non-naturalism still faces a few of the challenges that naturalism faces. These are mainly that there is still an issue with getting normativity out of a descriptive relation between whatever non-natural property is supposed to underpin the ethical one and that any non-natural property still faces some version of Moore's Open Question Argument as well.
The benefits of non-naturalism are in avoiding many of the pitfalls of tying non-ethical natural phenomena to ethical properties, and mostly it provides something for moral truths to be true of.
The biggest issues for moral non-naturalism in my mind are the following, which I will go over in greater detail:
- Explanatory impotence. If the facts aren't natural, why does the natural explain them?
- Explaining supervenience. This is somewhat related to the previous bullet.
- Epistemic connection. How can we know them if they are not natural?
- Explaining evolutionary debunking.
- I will not say much on this last bullet other than it seems like something that, for most non-naturalist views, is yet another coincidence to explain.
There is a question that immediately bothers me, which falls under the "epistemic connection" concern. It is puzzling to me how we could have knowledge of any non-natural property if our very perceptual and intellectual machinery is itself naturalistic (I'm sure there are those that would argue that they are not). This also clearly depends on what we mean by "non-natural". Whatever we mean, it seems clear that non-naturalism has challenges of a sort that don't exist for naturalism regarding how we can have moral knowledge.
Explanatory Impotence and Intuitionism
It seems that many non-naturalists go the route of intuitionism to try to solve the issue of "how we can know", or because they are epistemic intuitionists to begin with. I may be convinced eventually, but so far it is not a position I am incredibly attracted to.
In its most straightforward form, intuitionism holds that we come to know about moral properties through direct observation of those properties. Just as we can learn that the cat is on the mat through direct observation we can also learn that kicking the cat on the mat is wrong through direct observation.
This raises many red flags for me. Posting these additional properties and our acquaintance with them comes with a cost. That cost is in both the inflation of our ontology and in that we now have a spectacular coincidence to explain. That coincidence is: how are these supposedly observed moral intuitions tracking with moral facts if they are non-natural? Maybe it is a crude comparison to some, but non-naturalism is beginning to seem like Cartesian dualism to me. The intuitionist is trying to claim that there are causally inert moral properties that explain how we have moral knowledge, but this seems impossible if we are committed to our epistemological equipment being naturalistic and gives us more to explain than we started with.
Supervenience?
Another big hurdle for non-naturalism in my view is that it prima facie has a very hard time explaining supervenience. As the above argument with and shows, the supervenience of the moral on the non-moral entails that the non-moral is in some way constitutive of the moral. Even supposing we found a problem with one of the premises of that argument and could reject its conclusion, we still have the problem of explaining why the moral supervenes on the non-moral.
We cannot reject supervenience as non-naturalists because then we have to accept that the physical world could be exactly the same, but the moral facts somehow different. That is an absurd conclusion. So the only two moves left are:
- to somehow give an account for the seeming coincidence that moral properties of a non-natural and not causal sort are dependent on natural facts or,
- reject the notion that this supervenience exists.
Maybe the non-naturalist can reject the notion that "for X to supervene on Y means that they are coextensive" (premise 1 above regarding and )? That is to deny the possibility of the reduction of to in principle. This seems more feasible for cartesian dualists that it does here, which is not a compliment. It seems almost analytically true (maybe it is) that a physically identical alternate world could not have different moral facts. This closes the case for me on 2.
Maybe the non-naturalist can give an account of how this supervenience is possible, 1? Realist and non-naturalist Russ Shafer-Landau attempts to explain it as follows
he argues that although each moral property is distinct from any natural property or properties, still each instantiation of a moral property is fully constituted (or, as he sometimes puts it, ‘realized’) by some concatenation of natural properties (see Shafer-Landau 2003: 76–78).
SEP article on non-naturalism for much greater detail
This seems much more promising to me and is, thus far, the most convincing this I have seem form the non-naturalist camp. However, it is yet unclear to me how a moral property can be a distinct entity from, yet instantiated by, some set of natural properties. This seems to require a metaphysical commitment of some kind to these property-conglomerations that keeps this thesis distinct form some kind of naturalism. This view also has the bonus of getting around the issue of evolutionary debunking. This will no doubt prompt further reading on my part.
That sums up the non-naturalist views on a very general level and it is time for some review.
Review
Forget about "An account of X". How does it actually happen?
A favorite philosophical strategy of mine. I think it is very worthwhile to stop for a second trying to "account for X" and just think simply about how it actually, literally happens in everyday life. This method is responsible for untangling a lot of philosophical knots but can run the risk of drastically oversimplifying things.
So how do we actually interact with morality? What is our workaday metaethical persuasion? Imagine a scenario where a moral judgment is questioned and your intuition screams "That's not how morals work!" Here I actually feel like intuitionism is persuasive if it is slimmed down a bit. Most of us would surely say that we consult our intuitions and think about them. That's what literally happens. However, I don't think that tells us a ton in this case. Many theories can equally explain what our intuition is doing psychologically and a great many of them are not disconfirmed by whatever is going on in our heads.
You might think that these real psychological facts, whether every-day or gained from neuroscientific experiments, not strongly confirming any of these metaethical theories is reason to think that this is all some pie-in-the-sky mumbo-jumbo. If my moral intuitions aren't related to the metaethical theory, what good is it? I disagree. As I have stated elsewhere here, I do not think that a topic being highly theoretical makes it highly useless. Thinking of the 'long game' of human inquiry, the theoretical presuppositions we take up have real and important value.
Another "how it actually happens" consideration might be related to moral constructivism, the idea that we "make up" morality but it can be nonetheless true. It does seem that this is what we actually do and seems more probable to be the case to be then us reading the facts off of nature or some platonic-object somehow.
My Sketched Out View
So far I think it is clear that my view is either an analytical-naturalist view or a very metaphysically slim non-naturalist one like Shafer-Landau's.
I will attempt to sketch the view I am putting together in a few bullets:
There is at least one moral truth in the world (namely that "a state in which the least sought or most avoided states maximized for all beings is maximally bad and its opposite is maximally good". )
That truth is analytically true (I'll just go for it for now 😃) about the natural world
We are not required to quantify over (we are not committed to the existence of) any non-natural entities to explain this truth
Subsequent moral truths are constructed from this self-evident one (this is how I imagine normativity can sneak in.)
There are doubtless many problems here but this is V0.0.0 of my metaethics!
The Big Five
I will refer to these four aforementioned bullets as "The Big Five". Again, these are the four criteria that I think any good realist metaethical theory needs to have.
- 1. The true metaethical theory explains how moral facts interact causally with our moral language.
- 2. The true metaethical theory explains moral/non-moral supervenience.
- 3. The true metaethical theory explains how we can know moral facts.
- 4. The true metaethical theory has room for normativity.
- 5. the true metaethical theory accounts for evolutionary debunking.
These three hypotheses in conjunction I will refer to as .
I will now consider the most likely metaethical hypotheses to me, thus far, as well as my own and their respective likelihoods given . This just makes things a bit clearer in my head. I hope it doesn't seem over-the-top to the reader.
- : A kind of naturalistic moral-truth-agnostic non-cognitivism is the true metaethical theory
- : Naturalism is the true metaethical theory
- : Non-naturalism (Shafer-Landau's type/token supervenience explanation) is the true metaethical theory
- Struggles the most with 1.: explaining why we have reason to act via morality and why moral claims behave like "truth-apt" language
- Satisfies 2
- Dismisses 3
- Struggles with 4. It is not clear how non-facts can be reasons to act.
- Satisfies 5, perhaps better than any others
- Satisfies 1
- Satisfies 2
- Satisfies 3 better than some other analytic-naturalist views
- Potential struggles with 4. without strong bridge for reasons to act.
- Satisfies 5 decently
- Struggles with 1. How can these non-natural things interact with the natural?
- Unless we accept something like Shafer-Landau's view
- Or unless we say that our moral-truth is self-evidently true of natural phenomena
- Struggles with 2.
- Unless we accept something like Shafer-Landau's view
- Or unless we say that our moral-truth is self-evidently true of natural phenomena
- Struggles with 3. There is still a problem of "how we know" here even under a more sophisticated view.
- Or, again, unless we say that our moral-truth is self-evidently true of natural phenomena
- Excels at 4. With the ceiling high for non-natural properties, there is lots of room here to make normativity work, as I see it.
- Struggles with 5
- Again, unless we accept something like Shafer-Landau's view
- Or, yet again, unless we say that our moral-truth is self-evidently true of natural phenomena
- Struggles with 1. How can these non-natural things interact with the natural?
So Where Am I?
The tour is over for now. At the end of this foray into all kinds of different metaethical arguments, and a bit of self-discovery, I certainly have more questions than I started with, but I also certainly have fewer theoretical options that I view as viable now.
In summation, my view is based on some irreducible non-natural truth so it is not a naturalist one. I believe that my view needs more work to secure normativity, explain moral knowledge more deeply, and explain what makes the states of 'seeking' and 'avoiding' count as such. Do they count for all beings? If not, which ones?
Based on the above likelihood sketch, I believe my view, or something like it, accounts best for these "Big Five" criteria that a good metaethical theory requires. I lean in the direction of a non-naturalism that is self-evidently true of the natural world in an analogous way to mathematical truths, but I am still learning.
I have lots of work to do before I can consider my mind fully made up, but I hope that this little tour has helped the reader to contend with their own views, even if only by learning from the mistakes made here.