Total Posts:24|Showing Posts:1-24
Jump to topic:

Ontological argument for Moral Realism

popculturepooka
Posts: 7,924
Add as Friend
Challenge to a Debate
Send a Message
7/14/2015 11:41:50 AM
Posted: 1 year ago
Seems like a pretty strong argument (but, hey, what do I know, I'm *already* a moral realist as every sensible person should be. ;)) Not gonna type it out so I'll just c/p from Reddit:

"First things first let"s be clear about what the argument aims to establish. Obviously it"s an argument for moral realism, but that term is famously ambiguous so it"ll help to clear it up. For the purposes of this thread "moral realism" will refer to any moral theory according to which moral reasons for action:

(A) Are non-selfish, so they concern our conduct with regard to the lives of others.

(B) Are categorical, meaning that they apply to us regardless of our desires or institutional roles.

Construed in this way Huemer"s argument makes no detailed ontological commitments by itself. If successful the argument could equally support Kantian constructivists, naturalists, or robust realists alike.

The Probabilistic Reasons Principle

Huemer"s argument hinges on what he calls the Probabilistic Reasons Principle (PRP). We can give a rough statement of the principle like this:

(PRP-rough) "[...] if some fact would (if you knew it) provide a reason for you to behave in a certain way, then your having some reason to believe that this fact obtains also provides you with a reason to behave in the same way." (Huemer 263)

Although it may seem obscure, the idea here is actually pretty straightforward. Suppose that you"re betting on the outcome of a sports match. The fact that Team A is going to win would, if you knew it, provide you with a reason to bet on Team A. Now suppose that you have some reason to believe that Team A is going to win. For example, they have a talented ball-moving-person who can score lots of points. According to the PRP, your reason to believe that Team A is going to win also gives you a reason to bet on Team A. That"s it. Also note that this doesn"t have to be an overriding reason. So it could be that you have more reason (or a greater reason) to spend your money on safer investments, as is almost always the case with gambling, but it"s still the case that your reason to believe that Team A will win gives you some reason, however small, to bet on Team A.

One might consider the following a counterexample to the PRP: you have some reason, however small, to believe that the spot you"re standing on right now will be hit by a meteor. So it seems like, by the PRP, you have a reason to move. But a meteor is just as likely to hit any other place you move to, so you don"t really have any reason to move. There"s an easy fix for this and it involves tightening up the PRP. Let"s have the following be our official statement of the principle:

(PRP) If the following conditions hold:

"(a) If S knew that P, this would provide a reason for S to ]4;,

"(b) if S knew that ~P, this would provide no reason for S not to ]4;,

"(c) S has some reason to believe that P

Then S thereby has a reason to ]4;. (Huemer 265)

Both (a) and (c) account for our rough statement of the principle in more precise terms and the addition of (b) helps to account for meteor cases.

The Argument

Huemer"s argument proceeds in two steps. First there is the Anti-Torture Argument, which aims to establish that we have a reason to avoid torturing babies, then there is the argument for realism itself, which aims to establish from the Anti-Torture Argument that moral realism (as understood for the purposes of this thread) is true. So let"s get into it:

(1) The PRP.

(2) If we knew that torturing babies was objectively wrong, this would provide a reason to avoid torturing babies.

(3) Even if we knew that torturing babies was not objectively wrong, this would provide no reason to torture babies.

(4) We have some reason to believe that torturing babies is objectively wrong. (So there is some non-zero probability that moral realism is true.)

(5) Therefore, we have a reason to avoid torturing babies.

This argument is rather straightforward. Premise 1 is just the PRP, which has already been discussed.

Premise 2 is just true by definition.

Premise 3 seems obviously correct; there"s nothing about the non-existence of objective reasons to not torture babies that, by itself, entails that you should torture babies.

Premise 4 seems uncontroversial; moral realism doesn"t involve any self-contradictory or absurd claims, so there"s some chance that moral realism is true. On top of that moral realism is an at least respectable position in contemporary moral philosophy, so there are at least some reasons to think it"s true even if those reasons aren"t convincing to everyone.

Finally the conclusion (premise 5) just follows from the PRP and premises 2-4, which each fill in a spot in the PRP.

Note that 5 alone does not mean that moral realism is true. Indeed, moral subjectivists, cultural relativists, and the like may all agree with 5 while maintaining that realism is false. Of course the Anti-Torture Argument isn"t aimed at establishing realism. That"s accomplished by the following:

(6) The premises of the Anti-Torture Argument are true independent of interests, desires, and attitudes (in the sense relevant to moral realism).

(7) The premises of the Anti-Torture argument logically entail its conclusion.

(8) If P is true independent of interests, desires, and attitudes (in the sense relevant to moral realism), and P entails C, then C is true independent of interests, desires, and attitudes (in the sense relevant to moral realism).

(9) Therefore, the conclusion of the Anti-Torture Argument is true independent of interests, desires, and attitudes (in the sense relevant to moral realism).

Premise 6 seems correct; premises 1-4 were true independent of interests, attitudes, and desires. The PRP, if true, is a necessary truth like other principles of rationality, premise 2 was true by definition, premise 3 was unproblematic, and premise 4 was just a fact about the state of metaethics in this day and age.

Of course premise 7 is just an observation about the Anti-Torture Argument, which has already been discussed.

Premise 8 can be shown to be true by entertaining a counterfactual. Suppose that P is objectively true, P entails C, and C is merely true in virtue of some desire. Well in that case the desire could be withdrawn while the objective facts that make P true remain and it wouldn"t really be the case that P entailed C.

Thus the notion of 6-9 is easily summed up as: the premises of the Anti-Torture Argument are all objectively true and they all entail that we have a reason to avoid torturing babies, so it is likewise objectively true that we have a reason to avoid torturing babies.

The form of Huemer"s argument follows the classical ontological argument for theism: the possibility of God entails the existence of God. However, like the classical ontological argument, it seems as though Huemer"s argument has tricked us somehow by leading us through premises which seemed fine by themselves only to drop us on the conclusion before we even knew what was happening. Indeed it may end up being the case that we have been tricked, but if the trick cannot be revealed it seems like the only sensible choice is to accept the argument"s conclusion."

http://www.reddit.com...
At 10/3/2016 11:49:13 PM, thett3 wrote:
BLACK LIVES MATTER!
Envisage
Posts: 3,646
Add as Friend
Challenge to a Debate
Send a Message
7/14/2015 11:56:25 AM
Posted: 1 year ago
At 7/14/2015 11:41:50 AM, popculturepooka wrote:
Ok.
Definitions:

"moral realism" will refer to any moral theory according to which moral reasons for action:
(A) Are non-selfish, so they concern our conduct with regard to the lives of others.

(B) Are categorical, meaning that they apply to us regardless of our desires or institutional roles."

I presume that moral realism by definition, entails what we ought to do. These two qualifiers seem fine though.

The Probabilistic Reasons Principle

Huemer"s argument hinges on what he calls the Probabilistic Reasons Principle (PRP). We can give a rough statement of the principle like this:

(PRP-rough) "[...] if some fact would (if you knew it) provide a reason for you to behave in a certain way, then your having some reason to believe that this fact obtains also provides you with a reason to behave in the same way." (Huemer 263)

Although it may seem obscure, the idea here is actually pretty straightforward. Suppose that you"re betting on the outcome of a sports match. The fact that Team A is going to win would, if you knew it, provide you with a reason to bet on Team A. Now suppose that you have some reason to believe that Team A is going to win. For example, they have a talented ball-moving-person who can score lots of points. According to the PRP, your reason to believe that Team A is going to win also gives you a reason to bet on Team A. That"s it. Also note that this doesn"t have to be an overriding reason. So it could be that you have more reason (or a greater reason) to spend your money on safer investments, as is almost always the case with gambling, but it"s still the case that your reason to believe that Team A will win gives you some reason, however small, to bet on Team A.

You are conflating epistemology with ontology here.... Believing that it is epistemologically possible for there to be a reason doesn't obtain that it is logically possible. So, thus while incomplete knowledge can entail one to act in a morally realistic manner, it doesn't obtain that such a reason is ontological. You are trying to affirm moral realism here, thus you are equivocating between two definitions. You are literally stating that "the epistemic possibility of there being a reason" is an ontological reason itself, which is just false.

(PRP) If the following conditions hold:

"(a) If S knew that P, this would provide a reason for S to ]4;,


"(b) if S knew that ~P, this would provide no reason for S not to ]4;,


"(c) S has some reason to believe that P


Then S thereby has a reason to ]4;. (Huemer 265)

Both (a) and (c) account for our rough statement of the principle in more precise terms and the addition of (b) helps to account for meteor cases.

The Argument

Huemer"s argument proceeds in two steps. First there is the Anti-Torture Argument, which aims to establish that we have a reason to avoid torturing babies, then there is the argument for realism itself, which aims to establish from the Anti-Torture Argument that moral realism (as understood for the purposes of this thread) is true. So let"s get into it:

(1) The PRP.

(2) If we knew that torturing babies was objectively wrong, this would provide a reason to avoid torturing babies.

(3) Even if we knew that torturing babies was not objectively wrong, this would provide no reason to torture babies.

(4) We have some reason to believe that torturing babies is objectively wrong. (So there is some non-zero probability that moral realism is true.)

(5) Therefore, we have a reason to avoid torturing babies.


This argument is rather straightforward. Premise 1 is just the PRP, which has already been discussed.

Premise 2 is just true by definition.

Premise 3 seems obviously correct; there"s nothing about the non-existence of objective reasons to not torture babies that, by itself, entails that you should torture babies.

Premise 4 seems uncontroversial; moral realism doesn"t involve any self-contradictory or absurd claims, so there"s some chance that moral realism is true.

1. You need to prove that (bolded)
2. See my previous paragraph, at best you can only establish an epistemological reason.
3. I outright reject this premise. Please provide this reason why torturing babies is objectively wrong.

On top of that moral realism is an at least respectable position in contemporary moral philosophy, so there are at least some reasons to think it"s true even if those reasons aren"t convincing to everyone.

Those reasons need to be true and coherentfor them to be reasons sufficient for moral realism to be true.

Why do none of these arguments for realism and god play by the rules?
n7
Posts: 1,355
Add as Friend
Challenge to a Debate
Send a Message
7/14/2015 1:21:03 PM
Posted: 1 year ago
Huemer is one of my favorite philosophers when it comes to philosophy of perception.

Anyway, I don't see how this proves moral realism. I believe there are moral reasons for not killing babies, but they are subjective reasons. Harming babies is wrong, just not objectively wrong. We have reason to reject P3, as there are subjective reasons why harming babies is wrong.
404 coherent debate topic not found. Please restart the debate with clear resolution.


Uphold Marxist-Leninist-Maoist-Sargonist-n7ism.
Fkkize
Posts: 2,147
Add as Friend
Challenge to a Debate
Send a Message
7/14/2015 2:58:09 PM
Posted: 1 year ago
Let me try to run this in reverse.

(1) The PRP.

(2) If we knew that torturing babies was objectively wrong, this would provide a reason to avoid torturing babies.

(3) Even if we knew that torturing babies was not objectively wrong, this would provide no reason to torture babies.

(4) We have some reason to believe that torturing babies is objectively wrong. (So there is some non-zero probability that moral realism is true.)

(5) Therefore, we have a reason to avoid torturing babies.

(1) The PRP.

(2) If we knew that torturing babies was objectively right, this would provide a reason to torture babies.
If we knew anything to be objectively right, then, according to him, it would provide us with a reason.

(3) Even if we knew that torturing babies was not objectively right, this would provide no reason to not torture babies.
This might seem rather apalling to people and rightfully so, but the original premise appeals to the common perception of it being wrong, not on any independent argument or fact. Therefore I am free to run it in reverse. After all, even if its not objectively right, but, say, morally neutral instead, then there is no reason to act either way.

(4) We have some reason to believe that torturing babies is objectively right. (So there is some non-zero probability that moral realism is true.)
Moral realism does not entail any specific facts/duties/whatever, such as toturing babies being objectively right or wrong, it states that moral statements can be objectively true or false and that at least one such statement is true. Whether 'torturing babies is reprehensible' is true or false is not a question for metaethics.
Since we are not certain about torture being right, we cannot be certain that it is wrong either.

(5) Therefore, we have a reason to torture babies.

I think it's a novel attempt, however, error-theoretic considerations apply to alleged categorical reasons and as such I would not have been very convinced even if the rest of the argument was rock solid.
: At 7/2/2016 3:05:07 PM, Rational_Thinker9119 wrote:
:
: space contradicts logic
popculturepooka
Posts: 7,924
Add as Friend
Challenge to a Debate
Send a Message
7/14/2015 7:27:57 PM
Posted: 1 year ago
At 7/14/2015 11:56:25 AM, Envisage wrote:
At 7/14/2015 11:41:50 AM, popculturepooka wrote:
Ok.
Definitions:

"moral realism" will refer to any moral theory according to which moral reasons for action:
(A) Are non-selfish, so they concern our conduct with regard to the lives of others.

(B) Are categorical, meaning that they apply to us regardless of our desires or institutional roles."

I presume that moral realism by definition, entails what we ought to do. These two qualifiers seem fine though.

The Probabilistic Reasons Principle

Huemer"s argument hinges on what he calls the Probabilistic Reasons Principle (PRP). We can give a rough statement of the principle like this:

(PRP-rough) "[...] if some fact would (if you knew it) provide a reason for you to behave in a certain way, then your having some reason to believe that this fact obtains also provides you with a reason to behave in the same way." (Huemer 263)

Although it may seem obscure, the idea here is actually pretty straightforward. Suppose that you"re betting on the outcome of a sports match. The fact that Team A is going to win would, if you knew it, provide you with a reason to bet on Team A. Now suppose that you have some reason to believe that Team A is going to win. For example, they have a talented ball-moving-person who can score lots of points. According to the PRP, your reason to believe that Team A is going to win also gives you a reason to bet on Team A. That"s it. Also note that this doesn"t have to be an overriding reason. So it could be that you have more reason (or a greater reason) to spend your money on safer investments, as is almost always the case with gambling, but it"s still the case that your reason to believe that Team A will win gives you some reason, however small, to bet on Team A.

You are conflating epistemology with ontology here.... Believing that it is epistemologically possible for there to be a reason doesn't obtain that it is logically possible.

But it gives *some* reason (however weak) to think that it is logically and metaphysically possible. And that's all that is needed for the argument to work.

Ex: We can't know whether Goldbach's conjecture is true, but by it's very nature it's either necessarily true or false. But from our perspective, it could be (epistemically speaking) be possibly true or false.

Now, you can assert that moral realism is logically or metaphysically impossible but that is a tough one. You need an argument for that.

So, thus while incomplete knowledge can entail one to act in a morally realistic manner, it doesn't obtain that such a reason is ontological. You are trying to affirm moral realism here, thus you are equivocating between two definitions.

Not really.

You are literally stating that "the epistemic possibility of there being a reason" is an ontological reason itself, which is just false.


It gives us some reason to think so. Again. That is all that is needed.

(PRP) If the following conditions hold:

"(a) If S knew that P, this would provide a reason for S to ]4;,


"(b) if S knew that ~P, this would provide no reason for S not to ]4;,


"(c) S has some reason to believe that P


Then S thereby has a reason to ]4;. (Huemer 265)

Both (a) and (c) account for our rough statement of the principle in more precise terms and the addition of (b) helps to account for meteor cases.

The Argument

Huemer"s argument proceeds in two steps. First there is the Anti-Torture Argument, which aims to establish that we have a reason to avoid torturing babies, then there is the argument for realism itself, which aims to establish from the Anti-Torture Argument that moral realism (as understood for the purposes of this thread) is true. So let"s get into it:

(1) The PRP.

(2) If we knew that torturing babies was objectively wrong, this would provide a reason to avoid torturing babies.

(3) Even if we knew that torturing babies was not objectively wrong, this would provide no reason to torture babies.

(4) We have some reason to believe that torturing babies is objectively wrong. (So there is some non-zero probability that moral realism is true.)

(5) Therefore, we have a reason to avoid torturing babies.


This argument is rather straightforward. Premise 1 is just the PRP, which has already been discussed.

Premise 2 is just true by definition.

Premise 3 seems obviously correct; there"s nothing about the non-existence of objective reasons to not torture babies that, by itself, entails that you should torture babies.

Premise 4 seems uncontroversial; moral realism doesn"t involve any self-contradictory or absurd claims, so there"s some chance that moral realism is true.

1. You need to prove that (bolded)

Not really. So far as I know in the literature no one has even argued that moral realism is a contradiction in terms or that it leads to philosophical absurdities.

2. See my previous paragraph, at best you can only establish an epistemological reason.

That's the whole point. And actually some epistemological claims entail ontological claims.

If I veridically perceive P, it follows that P exists.

(a) If you knew that somebody was in your line of fire, that would provide a moral reason not to shoot.
(b) If you knew that nobody was in your line of fire, that would provide no moral reason not to shoot.
(c) You have some epistemic reason to think that somebody is in your line fire.

Therefore you have some moral reason not to shoot.

Seems to be no issue there.

3. I outright reject this premise. Please provide this reason why torturing babies is objectively wrong.


Now, if you want to assert that moral realism isn't even possibly true for all you know (epistemic possiblity) or is flat out metaphysically impossible, you're entirely free to do so, but that's your hill to die on. And that seems like a wholly unreasonable to take, especially given the (epistemic) disagreement rife in philosophy of philosophical peers. I can wholly acknowledge that there are atheists out there as equally smart as me (or smarter), and even see the force of their arguments (problem of evil and argument from divine hiddenness - i.e. I could see how one so predisposed could think those arguments give overriding reasons to believe God doesn't exists) and still think that they actually don't provide overriding reasons. I could see how they could provide *some* reason to think atheism is true (just not ENOUGH reason to think so), and I can acknowledge that *for all I know* I could be wrong - i.e. atheism has some non-zero probabilistic chance of being correct and is epistemically possible.

That's the position most philosophers take in disagreement and it seems pretty reasonable to me. That's just the nature of probablistic reasoning. I don't hear many saying - in any domain in philosophy - that their opponents' position isn't even epistemtically possible. Much less in moral philosophy.

WRT to baby torture I think it's pretty easy to see how it *could* be objectively wrong, and how there is *some* reason (even if there aren't overriding reasons) to think that it *could* be objectively wrong.

On top of that moral realism is an at least respectable position in contemporary moral philosophy, so there are at least some reasons to think it"s true even if those reasons aren"t convincing to everyone.

Those reasons need to be true and coherentfor them to be reasons sufficient for moral realism to be true.


*Some* reasons is different than overriding reasoning or fully justifactory reasons...and what about the reasons are incoherent?

Why do none of these arguments for realism and god play by the
At 10/3/2016 11:49:13 PM, thett3 wrote:
BLACK LIVES MATTER!
popculturepooka
Posts: 7,924
Add as Friend
Challenge to a Debate
Send a Message
7/14/2015 7:33:38 PM
Posted: 1 year ago
At 7/14/2015 2:58:09 PM, Fkkize wrote:
Let me try to run this in reverse.

(1) The PRP.

(2) If we knew that torturing babies was objectively wrong, this would provide a reason to avoid torturing babies.

(3) Even if we knew that torturing babies was not objectively wrong, this would provide no reason to torture babies.

(4) We have some reason to believe that torturing babies is objectively wrong. (So there is some non-zero probability that moral realism is true.)

(5) Therefore, we have a reason to avoid torturing babies.

(1) The PRP.

(2) If we knew that torturing babies was objectively right, this would provide a reason to torture babies.
If we knew anything to be objectively right, then, according to him, it would provide us with a reason.

(3) Even if we knew that torturing babies was not objectively right, this would provide no reason to not torture babies.
This might seem rather apalling to people and rightfully so, but the original premise appeals to the common perception of it being wrong, not on any independent argument or fact. Therefore I am free to run it in reverse. After all, even if its not objectively right, but, say, morally neutral instead, then there is no reason to act either way.


We (Huemer and I) would simply deny (3). According to Huemer's (and mines) epistemology:

"If it seems to S that P, then, in the absence of defeaters, S thereby has at least some justification for believing that P (Huemer 2007, p. 30; compare Huemer 2001, p. 99)."

The perception itself is *some* evidence that torturing babies is wrong and we ought not do it, prima facie at least.

http://www.iep.utm.edu...

(4) We have some reason to believe that torturing babies is objectively right. (So there is some non-zero probability that moral realism is true.)
Moral realism does not entail any specific facts/duties/whatever, such as toturing babies being objectively right or wrong, it states that moral statements can be objectively true or false and that at least one such statement is true. Whether 'torturing babies is reprehensible' is true or false is not a question for metaethics.
Since we are not certain about torture being right, we cannot be certain that it is wrong either.

(5) Therefore, we have a reason to torture babies.

I think it's a novel attempt, however, error-theoretic considerations apply to alleged categorical reasons and as such I would not have been very convinced even if the rest of the argument was rock solid.
At 10/3/2016 11:49:13 PM, thett3 wrote:
BLACK LIVES MATTER!
popculturepooka
Posts: 7,924
Add as Friend
Challenge to a Debate
Send a Message
7/14/2015 7:36:11 PM
Posted: 1 year ago
At 7/14/2015 1:21:03 PM, n7 wrote:
Huemer is one of my favorite philosophers when it comes to philosophy of perception.

Anyway, I don't see how this proves moral realism. I believe there are moral reasons for not killing babies, but they are subjective reasons. Harming babies is wrong, just not objectively wrong. We have reason to reject P3, as there are subjective reasons why harming babies is wrong.

But it's *possible* there are *objective* moral reasons for torturing babies right?
At 10/3/2016 11:49:13 PM, thett3 wrote:
BLACK LIVES MATTER!
Rational_Thinker9119
Posts: 9,054
Add as Friend
Challenge to a Debate
Send a Message
7/14/2015 7:54:02 PM
Posted: 1 year ago
Moral realism being possible is predicated on there not being some unknown external reason making it impossible. So saying that moral realism is possible and not non-contradictory would be to say that you know that there no unknown reasons making it impossible... But you are not omniscient, so, you cannot justify moral realism being possible.
n7
Posts: 1,355
Add as Friend
Challenge to a Debate
Send a Message
7/14/2015 7:59:01 PM
Posted: 1 year ago
At 7/14/2015 7:36:11 PM, popculturepooka wrote:
At 7/14/2015 1:21:03 PM, n7 wrote:
Huemer is one of my favorite philosophers when it comes to philosophy of perception.

Anyway, I don't see how this proves moral realism. I believe there are moral reasons for not killing babies, but they are subjective reasons. Harming babies is wrong, just not objectively wrong. We have reason to reject P3, as there are subjective reasons why harming babies is wrong.

But it's *possible* there are *objective* moral reasons for torturing babies right?

Sure, but then it would follow that there are reasons to not murder babies. However, negating P3 means there's also reasons not to murder babies under moral relativism.
404 coherent debate topic not found. Please restart the debate with clear resolution.


Uphold Marxist-Leninist-Maoist-Sargonist-n7ism.
popculturepooka
Posts: 7,924
Add as Friend
Challenge to a Debate
Send a Message
7/14/2015 8:01:11 PM
Posted: 1 year ago
At 7/14/2015 7:54:02 PM, Rational_Thinker9119 wrote:
Moral realism being possible is predicated on there not being some unknown external reason making it impossible. So saying that moral realism is possible and not non-contradictory would be to say that you know that there no unknown reasons making it impossible... But you are not omniscient, so, you cannot justify moral realism being possible.

If there are *unknown* reasons to think moral realism is impossible then there's absolutely no reason we know of to think that it actually is. And given that we know of no existing reasons to think that it is, and positive reason think it is (i.e not self-contradictory or at odds with any necessary truths) then there is reason to think it is possible.

And if we were to run with your line of thought we could never justfiably say *any* propostion is possible and there could always be "unknown" reasons as to why it's impossible. But that's no good.
At 10/3/2016 11:49:13 PM, thett3 wrote:
BLACK LIVES MATTER!
popculturepooka
Posts: 7,924
Add as Friend
Challenge to a Debate
Send a Message
7/14/2015 8:02:10 PM
Posted: 1 year ago
At 7/14/2015 7:59:01 PM, n7 wrote:
At 7/14/2015 7:36:11 PM, popculturepooka wrote:
At 7/14/2015 1:21:03 PM, n7 wrote:
Huemer is one of my favorite philosophers when it comes to philosophy of perception.

Anyway, I don't see how this proves moral realism. I believe there are moral reasons for not killing babies, but they are subjective reasons. Harming babies is wrong, just not objectively wrong. We have reason to reject P3, as there are subjective reasons why harming babies is wrong.

But it's *possible* there are *objective* moral reasons for torturing babies right?

Sure, but then it would follow that there are reasons to not murder babies. However, negating P3 means there's also reasons not to murder babies under moral relativism.

I meant to add a "not" in there.
At 10/3/2016 11:49:13 PM, thett3 wrote:
BLACK LIVES MATTER!
popculturepooka
Posts: 7,924
Add as Friend
Challenge to a Debate
Send a Message
7/14/2015 8:03:27 PM
Posted: 1 year ago
Anyways, I need to get back to my language studies. I've already spent enough time (too much time) philosophizing in English today. See ya'll tommorrow.
At 10/3/2016 11:49:13 PM, thett3 wrote:
BLACK LIVES MATTER!
Rational_Thinker9119
Posts: 9,054
Add as Friend
Challenge to a Debate
Send a Message
7/14/2015 8:21:26 PM
Posted: 1 year ago
At 7/14/2015 8:01:11 PM, popculturepooka wrote:
At 7/14/2015 7:54:02 PM, Rational_Thinker9119 wrote:
Moral realism being possible is predicated on there not being some unknown external reason making it impossible. So saying that moral realism is possible and not non-contradictory would be to say that you know that there no unknown reasons making it impossible... But you are not omniscient, so, you cannot justify moral realism being possible


If there are *unknown* reasons to think moral realism is impossible then there's absolutely no reason we know of to think that it actually is.

Irrelevant. The premise "moral realism is possible" presupposes there is no reason at all, not that we don't know of any reason.

And given that we know of no existing reasons to think that it is, and positive reason think it is (i.e not self-contradictory or at odds with any necessary truths) then there is reason to think it is possible.

It's possible for all we know, but that is epistemic possibility not metaphysical possibility.


And if we were to run with your line of thought we could never justfiably say *any* propostion is possible and there could always be "unknown" reasons as to why it's impossible. But that's no good.

This is the problem with claiming anything is metaphysically possible. Epistemic possibility is fine though.
Fkkize
Posts: 2,147
Add as Friend
Challenge to a Debate
Send a Message
7/15/2015 4:40:45 AM
Posted: 1 year ago
At 7/14/2015 7:33:38 PM, popculturepooka wrote:
At 7/14/2015 2:58:09 PM, Fkkize wrote:
Let me try to run this in reverse.

(1) The PRP.

(2) If we knew that torturing babies was objectively wrong, this would provide a reason to avoid torturing babies.

(3) Even if we knew that torturing babies was not objectively wrong, this would provide no reason to torture babies.

(4) We have some reason to believe that torturing babies is objectively wrong. (So there is some non-zero probability that moral realism is true.)

(5) Therefore, we have a reason to avoid torturing babies.

(1) The PRP.

(2) If we knew that torturing babies was objectively right, this would provide a reason to torture babies.
If we knew anything to be objectively right, then, according to him, it would provide us with a reason.

(3) Even if we knew that torturing babies was not objectively right, this would provide no reason to not torture babies.
This might seem rather apalling to people and rightfully so, but the original premise appeals to the common perception of it being wrong, not on any independent argument or fact. Therefore I am free to run it in reverse. After all, even if its not objectively right, but, say, morally neutral instead, then there is no reason to act either way.


We (Huemer and I) would simply deny (3). According to Huemer's (and mines) epistemology:

"If it seems to S that P, then, in the absence of defeaters, S thereby has at least some justification for believing that P (Huemer 2007, p. 30; compare Huemer 2001, p. 99)."
Then the argument assumes there to be no arguments against moral realism.

The perception itself is *some* evidence that torturing babies is wrong and we ought not do it, prima facie at least.

http://www.iep.utm.edu...

(4) We have some reason to believe that torturing babies is objectively right. (So there is some non-zero probability that moral realism is true.)
Moral realism does not entail any specific facts/duties/whatever, such as toturing babies being objectively right or wrong, it states that moral statements can be objectively true or false and that at least one such statement is true. Whether 'torturing babies is reprehensible' is true or false is not a question for metaethics.
Since we are not certain about torture being right, we cannot be certain that it is wrong either.

(5) Therefore, we have a reason to torture babies.

I think it's a novel attempt, however, error-theoretic considerations apply to alleged categorical reasons and as such I would not have been very convinced even if the rest of the argument was rock solid.
: At 7/2/2016 3:05:07 PM, Rational_Thinker9119 wrote:
:
: space contradicts logic
kp98
Posts: 729
Add as Friend
Challenge to a Debate
Send a Message
7/15/2015 6:08:17 AM
Posted: 1 year ago
Imagine someone ignorant of magnetism. If you show such a person a piece of iron sticking to a magnet he might well be extremely puzzled. As he doesn't know about magnets he can't imagine how it is sticking, but he can't say it isn't sticking on the basis the he doesn't know how it is sticking.

But many people seem to do just than in another area. That it is bad to torture a baby is as obvious as a piece of metal stuck to a magnet, but because people find it hard to say exactly why torturing babies is bad they say 'well, maybe it isn't bad after all'.

Now I can prove metal sticks to a magnet without too much controversy, but what proof is needed to prove torturing an baby is bad? Would I have to show you a baby being tortured? I don't think I do, do I?

OK - that's emotive. But the point I think is valid. I don't think we can use our ignorance of how objective morality could work to deny even the possibility of objective morality. That would be an example of what Dennett called the 'philosophers curse' - mistaking a failure of imagination for a universal principle if impossibilty.
popculturepooka
Posts: 7,924
Add as Friend
Challenge to a Debate
Send a Message
7/15/2015 9:48:10 AM
Posted: 1 year ago
At 7/14/2015 8:21:26 PM, Rational_Thinker9119 wrote:
At 7/14/2015 8:01:11 PM, popculturepooka wrote:
At 7/14/2015 7:54:02 PM, Rational_Thinker9119 wrote:
Moral realism being possible is predicated on there not being some unknown external reason making it impossible. So saying that moral realism is possible and not non-contradictory would be to say that you know that there no unknown reasons making it impossible... But you are not omniscient, so, you cannot justify moral realism being possible



If there are *unknown* reasons to think moral realism is impossible then there's absolutely no reason we know of to think that it actually is.

Irrelevant. The premise "moral realism is possible" presupposes there is no reason at all, not that we don't know of any reason.


But if we don't know it there's no reason to think there is a reason why it is impossible. There's literally no reason why we should think so. And by the same token even if *appears* impossible there could be some unknown reason as to why it is possible.

And given that we know of no existing reasons to think that it is, and positive reason think it is (i.e not self-contradictory or at odds with any necessary truths) then there is reason to think it is possible.

It's possible for all we know, but that is epistemic possibility not metaphysical possibility.


And if we were to run with your line of thought we could never justfiably say *any* propostion is possible and there could always be "unknown" reasons as to why it's impossible. But that's no good.

This is the problem with claiming anything is metaphysically possible. Epistemic possibility is fine though.

Epistemic is a fine but fallible guide to metaphysical possibility.
At 10/3/2016 11:49:13 PM, thett3 wrote:
BLACK LIVES MATTER!
popculturepooka
Posts: 7,924
Add as Friend
Challenge to a Debate
Send a Message
7/15/2015 9:51:17 AM
Posted: 1 year ago
At 7/15/2015 4:40:45 AM, Fkkize wrote:
At 7/14/2015 7:33:38 PM, popculturepooka wrote:
At 7/14/2015 2:58:09 PM, Fkkize wrote:
Let me try to run this in reverse.

(1) The PRP.

(2) If we knew that torturing babies was objectively wrong, this would provide a reason to avoid torturing babies.

(3) Even if we knew that torturing babies was not objectively wrong, this would provide no reason to torture babies.

(4) We have some reason to believe that torturing babies is objectively wrong. (So there is some non-zero probability that moral realism is true.)

(5) Therefore, we have a reason to avoid torturing babies.

(1) The PRP.

(2) If we knew that torturing babies was objectively right, this would provide a reason to torture babies.
If we knew anything to be objectively right, then, according to him, it would provide us with a reason.

(3) Even if we knew that torturing babies was not objectively right, this would provide no reason to not torture babies.
This might seem rather apalling to people and rightfully so, but the original premise appeals to the common perception of it being wrong, not on any independent argument or fact. Therefore I am free to run it in reverse. After all, even if its not objectively right, but, say, morally neutral instead, then there is no reason to act either way.


We (Huemer and I) would simply deny (3). According to Huemer's (and mines) epistemology:

"If it seems to S that P, then, in the absence of defeaters, S thereby has at least some justification for believing that P (Huemer 2007, p. 30; compare Huemer 2001, p. 99)."
Then the argument assumes there to be no arguments against moral realism.


No we would just argue that those arguments don't work. Or work as full defeaters.
And besides, you said (3) doesn't appeal to any fact or argument. But it actually does on common-sense epistemology.

The perception itself is *some* evidence that torturing babies is wrong and we ought not do it, prima facie at least.

http://www.iep.utm.edu...

(4) We have some reason to believe that torturing babies is objectively right. (So there is some non-zero probability that moral realism is true.)
Moral realism does not entail any specific facts/duties/whatever, such as toturing babies being objectively right or wrong, it states that moral statements can be objectively true or false and that at least one such statement is true. Whether 'torturing babies is reprehensible' is true or false is not a question for metaethics.
Since we are not certain about torture being right, we cannot be certain that it is wrong either.

(5) Therefore, we have a reason to torture babies.

I think it's a novel attempt, however, error-theoretic considerations apply to alleged categorical reasons and as such I would not have been very convinced even if the rest of the argument was rock solid.
At 10/3/2016 11:49:13 PM, thett3 wrote:
BLACK LIVES MATTER!
Fkkize
Posts: 2,147
Add as Friend
Challenge to a Debate
Send a Message
7/15/2015 10:03:20 AM
Posted: 1 year ago
At 7/15/2015 9:51:17 AM, popculturepooka wrote:
At 7/15/2015 4:40:45 AM, Fkkize wrote:
At 7/14/2015 7:33:38 PM, popculturepooka wrote:
At 7/14/2015 2:58:09 PM, Fkkize wrote:
Let me try to run this in reverse.

(1) The PRP.

(2) If we knew that torturing babies was objectively wrong, this would provide a reason to avoid torturing babies.

(3) Even if we knew that torturing babies was not objectively wrong, this would provide no reason to torture babies.

(4) We have some reason to believe that torturing babies is objectively wrong. (So there is some non-zero probability that moral realism is true.)

(5) Therefore, we have a reason to avoid torturing babies.

(1) The PRP.

(2) If we knew that torturing babies was objectively right, this would provide a reason to torture babies.
If we knew anything to be objectively right, then, according to him, it would provide us with a reason.

(3) Even if we knew that torturing babies was not objectively right, this would provide no reason to not torture babies.
This might seem rather apalling to people and rightfully so, but the original premise appeals to the common perception of it being wrong, not on any independent argument or fact. Therefore I am free to run it in reverse. After all, even if its not objectively right, but, say, morally neutral instead, then there is no reason to act either way.


We (Huemer and I) would simply deny (3). According to Huemer's (and mines) epistemology:

"If it seems to S that P, then, in the absence of defeaters, S thereby has at least some justification for believing that P (Huemer 2007, p. 30; compare Huemer 2001, p. 99)."
Then the argument assumes there to be no arguments against moral realism.


No we would just argue that those arguments don't work. Or work as full defeaters.
That sounds like special pleading. What would something have to look like to count as a full defeater?
It seems to me like an antirealist, too, can quite simply claim this argument not to be a full defeater.

And besides, you said (3) doesn't appeal to any fact or argument. But it actually does on common-sense epistemology.
Yes, it appeals to an intuition, not any by way of reason established fact or argument.
: At 7/2/2016 3:05:07 PM, Rational_Thinker9119 wrote:
:
: space contradicts logic
popculturepooka
Posts: 7,924
Add as Friend
Challenge to a Debate
Send a Message
7/15/2015 10:53:35 AM
Posted: 1 year ago
At 7/15/2015 10:03:20 AM, Fkkize wrote:
At 7/15/2015 9:51:17 AM, popculturepooka wrote:
At 7/15/2015 4:40:45 AM, Fkkize wrote:
At 7/14/2015 7:33:38 PM, popculturepooka wrote:
At 7/14/2015 2:58:09 PM, Fkkize wrote:
Let me try to run this in reverse.

(1) The PRP.

(2) If we knew that torturing babies was objectively wrong, this would provide a reason to avoid torturing babies.

(3) Even if we knew that torturing babies was not objectively wrong, this would provide no reason to torture babies.

(4) We have some reason to believe that torturing babies is objectively wrong. (So there is some non-zero probability that moral realism is true.)

(5) Therefore, we have a reason to avoid torturing babies.

(1) The PRP.

(2) If we knew that torturing babies was objectively right, this would provide a reason to torture babies.
If we knew anything to be objectively right, then, according to him, it would provide us with a reason.

(3) Even if we knew that torturing babies was not objectively right, this would provide no reason to not torture babies.
This might seem rather apalling to people and rightfully so, but the original premise appeals to the common perception of it being wrong, not on any independent argument or fact. Therefore I am free to run it in reverse. After all, even if its not objectively right, but, say, morally neutral instead, then there is no reason to act either way.


We (Huemer and I) would simply deny (3). According to Huemer's (and mines) epistemology:

"If it seems to S that P, then, in the absence of defeaters, S thereby has at least some justification for believing that P (Huemer 2007, p. 30; compare Huemer 2001, p. 99)."
Then the argument assumes there to be no arguments against moral realism.


No we would just argue that those arguments don't work. Or work as full defeaters.
That sounds like special pleading. What would something have to look like to count as a full defeater?
It seems to me like an antirealist, too, can quite simply claim this argument not to be a full defeater.


As it undercuts or rebuts moral realism. Good thing there aren't any arguments that do such a thing. ;)

And besides, you said (3) doesn't appeal to any fact or argument. But it actually does on common-sense epistemology.
Yes, it appeals to an intuition, not any by way of reason established fact or argument.

Appears to be begging the question. Why exclude intuition from the domain of facts or arguments or evidence. Intuitions/appearances are prima facie evidence.
At 10/3/2016 11:49:13 PM, thett3 wrote:
BLACK LIVES MATTER!
Fkkize
Posts: 2,147
Add as Friend
Challenge to a Debate
Send a Message
7/15/2015 11:07:32 AM
Posted: 1 year ago
At 7/15/2015 10:53:35 AM, popculturepooka wrote:


We (Huemer and I) would simply deny (3). According to Huemer's (and mines) epistemology:

"If it seems to S that P, then, in the absence of defeaters, S thereby has at least some justification for believing that P (Huemer 2007, p. 30; compare Huemer 2001, p. 99)."
Then the argument assumes there to be no arguments against moral realism.


No we would just argue that those arguments don't work. Or work as full defeaters.
That sounds like special pleading. What would something have to look like to count as a full defeater?
It seems to me like an antirealist, too, can quite simply claim this argument not to be a full defeater.


As it undercuts or rebuts moral realism.
But my question is where the line is between partial and full defeat. Is an argument showing categorical reasons to be nonsensical a partial or full defeater?

Good thing there aren't any arguments that do such a thing. ;)
I beg to differ.


And besides, you said (3) doesn't appeal to any fact or argument. But it actually does on common-sense epistemology.
Yes, it appeals to an intuition, not any by way of reason established fact or argument.

Appears to be begging the question. Why exclude intuition from the domain of facts or arguments or evidence. Intuitions/appearances are prima facie evidence.
I'm not going to argue this point any further. My point was that picking anything, perhaps something like abortion where people's opinions differ at large, gives you at best subjectivism. We can single out everyone's opinion and construct a realist duty from it by means of this argument.
: At 7/2/2016 3:05:07 PM, Rational_Thinker9119 wrote:
:
: space contradicts logic
Fkkize
Posts: 2,147
Add as Friend
Challenge to a Debate
Send a Message
7/15/2015 11:23:10 AM
Posted: 1 year ago
At 7/15/2015 10:53:35 AM, popculturepooka wrote:
The argument takes moral realism like empirical facts we have yet to discover, without digging into any of its metaethical problems. The metaethical trilemma was not covered and as such it is not even clear whether or not realism is even a coherent view.
Metaethics is a highly controversial topic, only 56% of philosophers are realists. Running this argument for realism is no better than running it against it.
: At 7/2/2016 3:05:07 PM, Rational_Thinker9119 wrote:
:
: space contradicts logic
Nac
Posts: 326
Add as Friend
Challenge to a Debate
Send a Message
7/15/2015 5:40:09 PM
Posted: 1 year ago
At 7/14/2015 11:41:50 AM, popculturepooka wrote:
Seems like a pretty strong argument (but, hey, what do I know, I'm *already* a moral realist as every sensible person should be. ;)) Not gonna type it out so I'll just c/p from Reddit:

"First things first let"s be clear about what the argument aims to establish. Obviously it"s an argument for moral realism, but that term is famously ambiguous so it"ll help to clear it up. For the purposes of this thread "moral realism" will refer to any moral theory according to which moral reasons for action:

(A) Are non-selfish, so they concern our conduct with regard to the lives of others.

(B) Are categorical, meaning that they apply to us regardless of our desires or institutional roles.

Construed in this way Huemer"s argument makes no detailed ontological commitments by itself. If successful the argument could equally support Kantian constructivists, naturalists, or robust realists alike.

It should also be noted that, by definition, moral realism asserts that said statements are facts of objective reality.

Although it may seem obscure, the idea here is actually pretty straightforward. Suppose that you"re betting on the outcome of a sports match. The fact that Team A is going to win would, if you knew it, provide you with a reason to bet on Team A. Now suppose that you have some reason to believe that Team A is going to win. For example, they have a talented ball-moving-person who can score lots of points. According to the PRP, your reason to believe that Team A is going to win also gives you a reason to bet on Team A. That"s it. Also note that this doesn"t have to be an overriding reason. So it could be that you have more reason (or a greater reason) to spend your money on safer investments, as is almost always the case with gambling, but it"s still the case that your reason to believe that Team A will win gives you some reason, however small, to bet on Team A.

This reason, however, is not the same as the one gained by knowledge, with the relevant difference being that this reason can coexist with a reason to bet on Team B.

One might consider the following a counterexample to the PRP: you have some reason, however small, to believe that the spot you"re standing on right now will be hit by a meteor. So it seems like, by the PRP, you have a reason to move. But a meteor is just as likely to hit any other place you move to, so you don"t really have any reason to move. There"s an easy fix for this and it involves tightening up the PRP. Let"s have the following be our official statement of the principle:

(PRP) If the following conditions hold:

"(a) If S knew that P, this would provide a reason for S to ]4;,


"(b) if S knew that ~P, this would provide no reason for S not to ]4;,


"(c) S has some reason to believe that P


Then S thereby has a reason to ]4;. (Huemer 265)

Both (a) and (c) account for our rough statement of the principle in more precise terms and the addition of (b) helps to account for meteor cases.

Does it account for the meteor case?

It seems that this same argument could be applied to each and every location that the meteor could hit, leaving us in precisely the same state in which we began.

Additionally, it should be noted that the meteor case had only a possibility, meaning this is the only criteria for a reason. As such, if something is not logically impossible, it is validated. Would this not just lead to contradictions later, as well as an overly broad ontology?

The Argument

Huemer"s argument proceeds in two steps. First there is the Anti-Torture Argument, which aims to establish that we have a reason to avoid torturing babies, then there is the argument for realism itself, which aims to establish from the Anti-Torture Argument that moral realism (as understood for the purposes of this thread) is true. So let"s get into it:

(1) The PRP.

(2) If we knew that torturing babies was objectively wrong, this would provide a reason to avoid torturing babies.

(3) Even if we knew that torturing babies was not objectively wrong, this would provide no reason to torture babies.

(4) We have some reason to believe that torturing babies is objectively wrong. (So there is some non-zero probability that moral realism is true.)

(5) Therefore, we have a reason to avoid torturing babies.


This argument is rather straightforward. Premise 1 is just the PRP, which has already been discussed.

Premise 2 is just true by definition.

Premise 3 seems obviously correct; there"s nothing about the non-existence of objective reasons to not torture babies that, by itself, entails that you should torture babies.

Premise 4 seems uncontroversial; moral realism doesn"t involve any self-contradictory or absurd claims, so there"s some chance that moral realism is true. On top of that moral realism is an at least respectable position in contemporary moral philosophy, so there are at least some reasons to think it"s true even if those reasons aren"t convincing to everyone.

Finally the conclusion (premise 5) just follows from the PRP and premises 2-4, which each fill in a spot in the PRP.

This argument, as discue

Note that 5 alone does not mean that moral realism is true. Indeed, moral subjectivists, cultural relativists, and the like may all agree with 5 while maintaining that realism is false. Of course the Anti-Torture Argument isn"t aimed at establishing realism. That"s accomplished by the following:

(6) The premises of the Anti-Torture Argument are true independent of interests, desires, and attitudes (in the sense relevant to moral realism).

(7) The premises of the Anti-Torture argument logically entail its conclusion.

(8) If P is true independent of interests, desires, and attitudes (in the sense relevant to moral realism), and P entails C, then C is true independent of interests, desires, and attitudes (in the sense relevant to moral realism).

(9) Therefore, the conclusion of the Anti-Torture Argument is true independent of interests, desires, and attitudes (in the sense relevant to moral realism).


Premise 6 seems correct; premises 1-4 were true independent of interests, attitudes, and desires. The PRP, if true, is a necessary truth like other principles of rationality, premise 2 was true by definition, premise 3 was unproblematic, and premise 4 was just a fact about the state of metaethics in this day and age.

Of course premise 7 is just an observation about the Anti-Torture Argument, which has already been discussed.

Premise 8 can be shown to be true by entertaining a counterfactual. Suppose that P is objectively true, P entails C, and C is merely true in virtue of some desire. Well in that case the desire could be withdrawn while the objective facts that make P true remain and it wouldn"t really be the case that P entailed C.

Thus the notion of 6-9 is easily summed up as: the premises of the Anti-Torture Argument are all objectively true and they all entail that we have a reason to avoid torturing babies, so it is likewise objectively true that we have a reason to avoid torturing babies.

The form of Huemer"s argument follows the classical ontological argument for theism: the possibility of God entails the existence of God. However, like the classical ontological argument, it seems as though Huemer"s argument has tricked us somehow by leading us through premises which seemed fine by themselves only to drop us on the conclusion before we even knew what was happening. Indeed it may end up being the case that we have been tricked, but if the trick cannot be revealed it seems like the only sensible choice is to accept the argument"s conclusion."

http://www.reddit.com...

I will need to c
Nac
Posts: 326
Add as Friend
Challenge to a Debate
Send a Message
7/16/2015 7:55:45 AM
Posted: 1 year ago
At 7/14/2015 11:41:50 AM, popculturepooka wrote:
For the purposes of this thread "moral realism" will refer to any moral theory according to which moral reasons for action:

(A) Are non-selfish, so they concern our conduct with regard to the lives of others.

(B) Are categorical, meaning that they apply to us regardless of our desires or institutional roles.

Construed in this way Huemer"s argument makes no detailed ontological commitments by itself. If successful the argument could equally support Kantian constructivists, naturalists, or robust realists alike.

It should also be noted that, by definition, moral realism asserts that said statements are facts of objective reality.

Although it may seem obscure, the idea here is actually pretty straightforward. Suppose that you"re betting on the outcome of a sports match. The fact that Team A is going to win would, if you knew it, provide you with a reason to bet on Team A. Now suppose that you have some reason to believe that Team A is going to win. For example, they have a talented ball-moving-person who can score lots of points. According to the PRP, your reason to believe that Team A is going to win also gives you a reason to bet on Team A. That"s it. Also note that this doesn"t have to be an overriding reason. So it could be that you have more reason (or a greater reason) to spend your money on safer investments, as is almost always the case with gambling, but it"s still the case that your reason to believe that Team A will win gives you some reason, however small, to bet on Team A.

This reason, however, is not the same as the one gained by knowledge, with the relevant difference being that this reason can coexist with a reason to bet on Team B.

One might consider the following a counterexample to the PRP: you have some reason, however small, to believe that the spot you"re standing on right now will be hit by a meteor. So it seems like, by the PRP, you have a reason to move. But a meteor is just as likely to hit any other place you move to, so you don"t really have any reason to move. There"s an easy fix for this and it involves tightening up the PRP. Let"s have the following be our official statement of the principle:

(PRP) If the following conditions hold:

"(a) If S knew that P, this would provide a reason for S to ]4;,


"(b) if S knew that ~P, this would provide no reason for S not to ]4;,


"(c) S has some reason to believe that P


Then S thereby has a reason to ]4;. (Huemer 265)

Both (a) and (c) account for our rough statement of the principle in more precise terms and the addition of (b) helps to account for meteor cases.

Does it account for the meteor case?

It seems that this same argument could be applied to each and every location that the meteor could hit, leaving us in precisely the same state in which we began.

Additionally, it should be noted that the meteor case had only a possibility, meaning this is the only criteria for a reason. As such, if something is not logically impossible, it is validated. Would this not just lead to contradictions later, as well as an overly broad ontology?

The Argument

(1) The PRP.

(2) If we knew that torturing babies was objectively wrong, this would provide a reason to avoid torturing babies.

(3) Even if we knew that torturing babies was not objectively wrong, this would provide no reason to torture babies.

(4) We have some reason to believe that torturing babies is objectively wrong. (So there is some non-zero probability that moral realism is true.)

(5) Therefore, we have a reason to avoid torturing babies.

As previously noted, this reason does not imply certainty, so it can coexist with reasons to do the opposite.

Arguments can, therefore, be made for the opposite, as Fkkize pointed out.

The only contention held against it is perception, which works until we utilize the addendum I created earlier. If moral statements are objective facts, then our perception is ineffectual to the actual truth value. The use of perception in regards to facts is an assurance that, in an evidentialist sense, our understanding of a fact is accurate. These two ideas are not the same. If this were the case, then the earth started revolving around the sun when heliocentrism was discovered, gravity did not exist before Isaac Newton, etc.

This argument is rather straightforward. Premise 1 is just the PRP, which has already been discussed.

Premise 2 is just true by definition.

Premise 3 seems obviously correct; there"s nothing about the non-existence of objective reasons to not torture babies that, by itself, entails that you should torture babies.

Premise 4 seems uncontroversial; moral realism doesn"t involve any self-contradictory or absurd claims, so there"s some chance that moral realism is true. On top of that moral realism is an at least respectable position in contemporary moral philosophy, so there are at least some reasons to think it"s true even if those reasons aren"t convincing to everyone.

Finally the conclusion (premise 5) just follows from the PRP and premises 2-4, which each fill in a spot in the PRP.



Note that 5 alone does not mean that moral realism is true. Indeed, moral subjectivists, cultural relativists, and the like may all agree with 5 while maintaining that realism is false. Of course the Anti-Torture Argument isn"t aimed at establishing realism. That"s accomplished by the following:

(6) The premises of the Anti-Torture Argument are true independent of interests, desires, and attitudes (in the sense relevant to moral realism).

(7) The premises of the Anti-Torture argument logically entail its conclusion.

(8) If P is true independent of interests, desires, and attitudes (in the sense relevant to moral realism), and P entails C, then C is true independent of interests, desires, and attitudes (in the sense relevant to moral realism).

(9) Therefore, the conclusion of the Anti-Torture Argument is true independent of interests, desires, and attitudes (in the sense relevant to moral realism).

The existence of a reason which exists in conjunction with an opposing reason proves moral realism? In what way?

This all seems to be more of a conflation of Bayesian epistemology with ethics. It deals with decision matrices in a way which allows reasons to be made based upon probabilities.

However, judging between probabilities in this fashion does not imply moral realism. This requires reasons with certainty, and the argument you presented deals only in probabilities.

The form of Huemer"s argument follows the classical ontological argument for theism: the possibility of God entails the existence of God. However, like the classical ontological argument, it seems as though Huemer"s argument has tricked us somehow by leading us through premises which seemed fine by themselves only to drop us on the conclusion before we even knew what was happening. Indeed it may end up being the case that we have been tricked, but if the trick cannot be revealed it seems like the only sensible choice is to accept the argument"s conclusion."

The tricks seem to be a conflation of probabilities and certainties, a suppression of moral realism's claims to objective fact, and a invalid rebuttal of a noteworthy counterclaim.

This is an intriguing proposal, however. I do enjoy allowing bayesian epistemology to enter any conversation. It just needs to be grounded by an understanding of its probabilistic nature.

Thank you for providing this argument. It was actually difficult to find objections, since, as you stated, it utilizes numerous tricks which take immense effort to find.