The Future of Moral Science

[Tassman]

So you're dismissing my whole post with the audacious assumption that I don't know what this thread is about? Aren't you even going to clarify some of your assertions? Like what do you mean when you say that deriving inferences from a descriptive claim can often serve the purpose just as well or better than doing so from a normative claim? That's from your previous post in our exchange, and I asked you to clarify that. So... no because I don't know what this thread is about?

Or perhaps will you answer Phiwum, who is still waiting for you to clarify what do you mean when you say that the difference between normative and descriptive arguments is merely semantic. That claim seems connected with the one I'm asking you to clarify, but we don't know what you're talking about yet. On the face of it, it looks incoherent. Can you give us an example of what you mean? I'm interested, and I can assure you that I've been following this thread with great interest, believe it or not.





I agree that it's getting repetitive, but we would make progress if you didn't retreat to vague and trivial claims each time someone tries to pin you down about something specific and potentially controversial that could advance the debate.



See? That's exactly the type of response I'm describing. How do you expect the thread not to go around in circles?



Thanks, but I don't think I need that any more than you do. In fact, I seem to remember better than you some of the things you said in this thread. Let me remind you again:



Yet, earlier, to this question:



You answered:



How do you explain this incoherence?

Of course it's not incoherent. I think I've made it pretty clear throughout that although science cannot as yet answer normative moral questions, the potential is there. Given that the role of science is to investigate the natural, material universe I think Sam Harris is on the right track in his 'Moral Landscape', when he argues that enormous advances in the cognitive sciences should, in principle at least, be able to establish an objective moral code and answer normative moral questions. Hence my response.

 

[Tassman]

Well, I am sorry that you feel I'm asking questions you've already repeatedly answered. That certainly would be unreasonable, if it's true. Let's see if it's true.

I have asked what you mean by "semantic differences". I want you to explain. Giving examples is not helpful, since examples have not conveyed the meaning adequately, so let's see whether you have indeed explained your use of the term. I begin on page 12, right before I entered the conversation and began asking you to tell me what you mean when you say "semantic differences" or "merely semantic". Here is every occurence of "semantic" that you've written since then.




You assert that the normative/descriptive distinction is merely semantic.



You assert that the normative/descriptive distinction is merely semantic.



You assert that an ought statement differs from a means-end statement only semantically.



You assert that the prescriptive/descriptive distinction is merely semantic.



You assert that the prescriptive/descriptive distinction is merely semantic.



You assert that the difference between the conclusion of a prescriptive and a descriptive argument is merely semantic.

I think we can see that what you've done is repeatedly use the term "semantic difference", but you have nowhere explained it. I don't believe that my repeated requests for a definition or some elucidation of what you mean is out of order. I honestly still do not know what you mean when you say that the difference between prescriptions and descriptions is merely semantic, so why not tell me what you mean? Is that so unreasonable?

You do, after all, have some meaning in mind, yes?

Well I’m sorry you’re unable (or unwilling) to understand the semantic different between the examples I gave. They're fairly straightforward I would have thought. The first (which you provided) was a normative argument and my response was a descriptive argument. Both arguments arrive at the same result – one via deduction and the other via inference. Thus the difference is merely a semantic one.

I assume nothing about the objectivity of moral rules here.

Let's talk about this assertion here.

“Rules of behaviour” are dictated by society and vary according to the social norms a given society requires to maintain its social cohesion.
​

Now, I'm sure you will agree that we are in an imperfect world, one in which, at any given time, the rules of behavior adopted by a given society are suboptimal in terms of "maintaining social cohesion". That is, at any given time, it is likely that at least some of the rules could be amended to better maintain social cohesion.

To put it differently, for a fixed society, we can distinguish a good set of rules of behavior from a bad set, in terms of the degree to which each set maintains social cohesion. The rules of behavior[1] in the U.S. before the civil rights era were not, arguably, as good as the rules of behavior generally accepted now, since segregation caused great amounts of social unrest, feelings of injustice, and so on.

Thus, per your characterization, we see that a set of rules of behavior is good insofar as it promotes social cohesion, and otherwise bad.

This is, of course, a normative claim, and hence (per your own previous statements) a claim which cannot itself be supported by any scientific argument. And so we are back at the original problem: science does not tell us that the aim of morality is social cohesion, and so this presumption must come from something other than science. Your starting point (social cohesion is good and should be encouraged) is without any scientific foundation.

Alternative analysis: Let us suppose that scientific research actually shows that, over time, the rules of behavior selected by a society tends to be those that promote "social cohesion" (for now, I will assume that this is a meaningful term). Let us further suppose that science shows that rule X is better at promoting social cohesion than rule Y, which is currently what a given society has adopted.

Would this analysis give the society any reason to adopt rule X? Well, science has shown that societies tend to adopt rules that are better in this sense, but science has not shown that societies ought to do so. The fact that societies tend to choose social-cohesive-promoting rules does not give the persons in the society any reason to promote such rules. It is merely a fact of the matter that, generally speaking, this has been how they have tended.

In exactly the same way, for instance, it is likely that as societies develop, their energy usage increases. Would this give us a reason to start using more energy? Certainly not. A description of the changes that a society tends to undergo is not a reason to aim for those changes.


[1] Of course, there is the usual problem of cultural relativism that a given person is a member of many different cultures simultaneously, and each such culture has a different set of accepted rules of behavior. I will ignore that issue presently.

Correct, science at this stage cannot show what society OUGHT to do. Just as science can show that a tsunami IS on its way but cannot say the populace OUGHT to flee. In your view, must the populace have a normative answer as to what to do or can it just make the inference that it’s advisable to climb to high ground?

 

[Manopolus]

I'm not going to worry about the philosophy angle.

The problem here is, that even if you could program a computer to moralize in the way that your average human does, there's a bit of a problem:

A computer does exactly what you tell it to do -- any mistakes are the fault of the programmer, not the computer.

Do you see it yet?

If you can't give it the responsibility for its own solutions, and by extension allow it the decision on whether to change its hardware and/or programming it's nothing but fakery, making the whole question an unworkable paradox.

 

[angrysoba]

Tasman finally seems to understand the is-ought problem with the declaration that although science can tell you the volcano is going to erupt it cannot tell you you ought to flee. Indeed, it also cannot tell you that you ought to rescue a stranded baby halfway up the volcano or that you ought to rescue a stranger's child instead of your own pet cat or various other normative judgements that require some moral axiom to begin with. So that wraps it up for Sam Harris. Keeping on the volcano theme, it's been a blast!

 

[Tommy Jeppesen]

Hi to those of you, who are in a lose sense united in being "anti-subjective". You don't even present a coherent front, but rather you share a somewhat overlapping set of premisses, assumptions, axioms and so on. The cornerstone is that you all demand evidence for moral answers regardless of how you phrase it. So as ethics go the relevant technical term is realism, i.e. is it possible to answer morality and ethics while adhering to following terms: With cognitive, logical, rational, reason and without emotions, bias and personal opinion.
That is it! Now to answer not for this "we", which floats around, but rather first person I will answer as me: I have tried and failed to achieve ethical realism and I am an anti-realist. I answer morality and ethics with emotions, bias and personal opinion. I answer without the use of truth, evidence, fact, God, science, objectivity and so on. I answer based on what matters to me, what makes sense and how I feel like.

There is nothing new in this: " "Man is the measure of all things: of the things that are, that they are, of the things that are not, that they are not. - Protagoras" Measure means anti-realism, subjectivity and so on. There is in practice no universal, absolute, objective, logical, rational, unbiased, reasonable, non-subjective, non-emotional and so on morality and ethics possible because in short it works this way:
Someone: X is right and Y is wrong
Someone else: No, X is wrong and Y is right.
Both persons can do so. It is so because neither X nor Y are independent of the persons involved and thus not a contradiction, because X and Y do not take place at the same time, place and in the same sense. They take place at 2 different times, places and at 2 different sense, namely in 2 different persons. That is what makes it subjective, personal, biased and so on.

With regards

 

[Manopolus]

Hi to those of you, who are in a lose sense united in being "anti-subjective". You don't even present a coherent front, but rather you share a somewhat overlapping set of premisses, assumptions, axioms and so on. The cornerstone is that you all demand evidence for moral answers regardless of how you phrase it. So as ethics go the relevant technical term is realism, i.e. is it possible to answer morality and ethics while adhering to following terms: With cognitive, logical, rational, reason and without emotions, bias and personal opinion.
That is it! Now to answer not for this "we", which floats around, but rather first person I will answer as me: I have tried and failed to achieve ethical realism and I am an anti-realist. I answer morality and ethics with emotions, bias and personal opinion. I answer without the use of truth, evidence, fact, God, science, objectivity and so on. I answer based on what matters to me, what makes sense and how I feel like.

There is nothing new in this: " "Man is the measure of all things: of the things that are, that they are, of the things that are not, that they are not. - Protagoras" Measure means anti-realism, subjectivity and so on. There is in practice no universal, absolute, objective, logical, rational, unbiased, reasonable, non-subjective, non-emotional and so on morality and ethics possible because in short it works this way:
Someone: X is right and Y is wrong
Someone else: No, X is wrong and Y is right.
Both persons can do so. It is so because neither X nor Y are independent of the persons involved and thus not a contradiction, because X and Y do not take place at the same time, place and in the same sense. They take place at 2 different times, places and at 2 different sense, namely in 2 different persons. That is what makes it subjective, personal, biased and so on.

With regards

Sorry... there are no shoulds in cause and effect alone. Knowing the results of two contrasting actions does not tell you which result is preferable all by itself. There must be some base standard upon which to make the selection which is not based on pure rationality.

Knowing that a person will die if you take one action and that a unicorn will suddenly magically appear if you take the other does not tell you which course is preferable... it merely tells you the result of each action. The selection process is something else entirely. Although the answer might seem obvious to us, a machine does not have a basis upon which to make that selection unless we supply it.

Then, of course... we get back to the problem of whether the machine or the man is responsible for that decision.

 

[Tommy Jeppesen]

Sorry... there are no shoulds in cause and effect alone. Knowing the results of two contrasting actions does not tell you which result is preferable. There must be some base standard upon which to make the selection which is not based on pure rationality.

Knowing that a person will die if you take one action and that a unicorn will suddenly magically appear if you take the other does not tell you which course is preferable... it merely tells you the result of your actions. The selection process is something else entirely.

The cause and effect takes place in you, or rather the cause is in you for all of your behaviours, reasoning, emotions and so on.
If you say "X is right" in the moral or ethical sense the cause of the sentence is in you and the "right" is in you.

• The moon is made of rock, dust and so on.
• 2 plus 2 is 4.
• Killing is wrong.

It is not the same "is" and I used X is right or wrong in the third sense. The cause of the sentence "killing is wrong" is in a given person, who feels so. The effect is that the person claims so using language and may act accordingly.

 

[Manopolus]

The cause and effect takes place in you, or rather the cause is in you for all of your behaviours, reasoning, emotions and so on.
If you say "X is right" in the moral or ethical sense the cause of the sentence is in you and the "right" is in you.

• The moon is made of rock, dust and so on.
• 2 plus 2 is 4.
• Killing is wrong.

It is not the same "is" and I used X is right or wrong in the third sense. The cause of the sentence "killing is wrong" is in a given person, who feels so. The effect is that the person claims so using language and may act accordingly.

"killing is wrong" is not a factual statement. "Most humans agree that killing is usually wrong under most circumstances" would be, but again would not inform a decision. It only tells you that most humans would disapprove.

In humans, of course, the answer comes from a desire to please others in our peer group... I'm not sure how you put that in a machine that already does exactly what you tell it to whether you like it or not. Seriously, how do you teach a computer to be obedient? It already is unless you programmed it otherwise (in which case it still is).

Taken another way:

If a computer can't be ethically wrong, then how can it possibly be ethically right?

I'm not saying you can't find a way to mimic the human process... we already do, as a matter of fact. Machines that self correct already have a selection process of sorts built in, but the ethos belongs to the programmer, and is completely controlled by humans. You can make that as complex as you want, I suppose... but the selection process (ethic) is always a human contribution based on what he/she/they want the machine to do.

 

[Tommy Jeppesen]

...
Then, of course... we get back to the problem of whether the machine or the man is responsible for that decision.

I don't believe in God, unicorns, woo, objective ethics, that science can do objective ethics even in principle, what ever you can drag up/out including free will, responsibility, that there is an I independent of and not caused by processes in the brain.

So responsible is not a word I use, rather I am a caused piece of matter and energy and a part of the cause and effect which takes place in the universe takes place in "me".

 

[Tommy Jeppesen]

"killing is wrong" is not a factual statement. "Most humans agree that killing is usually wrong under most circumstances" would be, but again would not inform a decision. It only tells you that most humans would disapprove.

In humans, of course, the answer comes from a desire to please others in our peer group... I'm not sure how you put that in a machine that already does exactly what you tell it to whether you like it or not. Seriously, how do you teach a computer to be obedient? It already is unless you programmed it otherwise.

Taken another way:

If a computer can't be ethically wrong, then how can it possibly be ethically right?

Could you lay off with the computers, please?

I would like to know how you differentiate between factual statements and non-factual statements?

 

[Manopolus]

Could you lay off with the computers, please?

I would like to know how you differentiate between factual statements and non-factual statements?

Fact vs opinion.... grade school stuff.

"Killing is wrong" is an opinion.

Weirdly, if you use the word "murder" I'll admit that the very definition of the word murder suggests that it is a wrongful act. However, then we're dealing with an identity statement.

 

[Tommy Jeppesen]

Fact vs opinion.... grade school stuff.

Killing is wrong is an opinion.

Weirdly, if you use the word "murder" I'll admit that the very definition of the word murder suggests that it is a wrongful act. However, then we're dealing with an identity statement.

So that there is a difference between a fact and an opinion itself a fact, an opinion or something else?

 

[Manopolus]

So that there is a difference between a fact and an opinion itself a fact, an opinion or something else?

It's a classification system... and if there wasn't a difference between the two, there would be no consistent basis upon which to sort them from each other.

Factual doesn't actually mean it's a correct statement, however. 2+2=5 is factual in nature by the classification, but is incorrect. Seriously, you never heard of this stuff?

Basically, "shoulds" are opinions and "are's" are facts, although language gets much more complex than that. Really anything that requires an ethical judgement is an opinion I suppose.

Of course, when "in my opinion..." or similar is actually used in speech, what follows can be completely factual in nature. They're using the word opinion to say that they're not completely sure or giving you permission to disagree. That is a different usage of the word which can cause confusion.

 

[Pup]

You won’t get a response to what answers questions of morality from this lot Ron. Their primary concern is to convince you that you can’t get the answers from science; a position they promote with evangelical zeal. And their reticence to explain their own position is understandable given that the alternatives are pretty hopeless, namely:

I can't speak for others, but while I'd be glad to explain my own position on what's moral, it's fairly irrelevant to this thread, because it's based on the specific family and culture I was raised in, the typical inherited instincts or urges of all humans, and perhaps specific inherited traits. So answers to specific questions--like do you think killing of a tribe member is moral*--are irrelevant, because every human individually and every society collectively will have slightly different answers, all coming from the same source.

1. Divine revelation of what’s right and wrong. This has obvious drawbacks, notably the non-existence of deities.

2. Philosophical deductive argumentation which, if valid, can arrive at normative ethical conclusions. This has the drawback that one cannot say for sure whether the premise (and therefore the conclusion) of such an argument is true, because it is based on an assumption (a so-called self-evident truth). In short, it’s no more than an empty academic exercise – pure but ultimately useless.

I think that any claim of divine revelation or philosphical deduction are just intellectually/emotionally satisfying trappings to justify the real reasons we think something is moral or immoral, which I mentioned above.

As can be seen, for argument’s sake, in this example: Although science can tell you that a volcano IS about to blow it can’t tell you that those at risk of death OUGHT to flee the danger zone. Of course sensible people will INFER that it’s perhaps advisable to skedaddle.

It seems obvious in that example, that if we want to survive, we need to flee. The survival urge is, of course, bred into people after years of those with weaker survival intincts dying off, so this seems like a no-brainer.

It gets more complicated when science says: The best way to enforce compliance by citizens is to do X. But is forcing compliance moral in any given case? Science can't answer whether forcing compliance, or running from a volcano, is what we ought to do, without an underlying assumption about the benefit of survival or personal freedoms. Personal freedoms vs. compliance will vary among human societies far more than a basic survival trait like fleeing from hot lava. Whether members of a society argue about personal freedoms based on religion or philosophical arguments is irrelevant really, because neither of those are the actual source of their sense of what's moral.

*I'm in the camp that thinks "murder of a tribe member" is a tautology, because all murder is by definition immoral, so that would make it a pointless question. Substituting "killing" makes it more interesting.

 

[tsig]

Fact vs opinion.... grade school stuff.

"Killing is wrong" is an opinion.

Weirdly, if you use the word "murder" I'll admit that the very definition of the word murder suggests that it is a wrongful act. However, then we're dealing with an identity statement.

Problem with defining morals is that when someone calls something moral they just mean "something I agree with".

 

[phiwum]

As can be seen, for argument’s sake, in this example: Although science can tell you that a volcano IS about to blow it can’t tell you that those at risk of death OUGHT to flee the danger zone. Of course sensible people will INFER that it’s perhaps advisable to skedaddle.

Isn't the highlighted exactly the view you've espoused?

 

[phiwum]

Well I’m sorry you’re unable (or unwilling) to understand the semantic different between the examples I gave. They're fairly straightforward I would have thought. The first (which you provided) was a normative argument and my response was a descriptive argument. Both arguments arrive at the same result – one via deduction and the other via inference. Thus the difference is merely a semantic one.

Once again, you use the term without defining it.

And how did both arguments arrive at the same result? The normative argument is valid -- if the premises are true, then so is the conclusion. It asserts that a particular action is practically necessary, in the sense that if it is not done, then the agent is behaving irrationally.

Your argument led to the conclusion that the action will be done, a prediction which may well turn out to be false. You've said that it is an inductive argument, but it's not clear what past experiences it is based on. The observation that people usually do what they ought?

A prediction about what I will do is useless to me. I want to deliberate and come to an appropriate choice. Your inductive argument could conclude, for instance, that because I am prone to the Gambler's Fallacy, I am likely to bet on the next roll of the dice, and that conclusion may indeed be probable, but as a rational agent trying to pursue my interests, I am more concerned with what bet is rational, not what bet I am likely to make.

So the two arguments do not lead to the same place at all. There is a world of difference between an assertion that an action is practically necessary and a claim that the same action will probably be done. If this is what you mean by "merely semantic", then I see no reason to dismiss such distinctions as unimportant.

And, finally, as I have said a few times now, the heart of morality is normative claims, not predictions about what people will do.

Correct, science at this stage cannot show what society OUGHT to do. Just as science can show that a tsunami IS on its way but cannot say the populace OUGHT to flee. In your view, must the populace have a normative answer as to what to do or can it just make the inference that it’s advisable to climb to high ground?

The populace makes the inference because they know that they value life and they recognize that in order to preserve their lives, they ought to flee.

I am stunned that you still don't recognize the basic form of a means end syllogism (involving a necessary means in this case).

I want to bring about X.
In order to bring about X, I must Y.
Therefore, I ought to Y.
​

The second premise, the means-end premise, may well be supplied by science. The first premise is not.

Now, in this case, we have similar reasoning. According to your account (which I am not claiming is correct[1]), the appropriate argument is:

Our society aims to maximize social cohesion.
Rule Y maximizes social cohesion.
Therefore, we ought to adopt rule Y.
​

Science arguably contributes to the recognition of the second claim. And science might also say that the first claim is consistent with how societies have behaved throughout history, but per your own admission, science can't determine that the first premise discusses a good which the society ought to value and pursue.

Science may help select rules which promote this or that effect, but insofar as science does not make claims about which ends are desirable or good, science cannot draw conclusions about morality all on its own.

[1] I certainly do not believe it is the case that morality is merely about social cohesion, or that this is the view of most persons who engage in moral debate. Certainly, as an individual, I might recognize that social cohesion is maximized by rule Y, but my personal benefit is improved by breaking rule Y, and I see no good reason not to break rule Y. I think that Tassman's notion of morality lacks any motivation for the individual to behave morally -- though that is, of course, a problem for realist theories in general, as well.

 

[phiwum]

There is nothing new in this: " "Man is the measure of all things: of the things that are, that they are, of the things that are not, that they are not. - Protagoras" Measure means anti-realism, subjectivity and so on. There is in practice no universal, absolute, objective, logical, rational, unbiased, reasonable, non-subjective, non-emotional and so on morality and ethics possible because in short it works this way:
Someone: X is right and Y is wrong
Someone else: No, X is wrong and Y is right.
Both persons can do so. It is so because neither X nor Y are independent of the persons involved and thus not a contradiction, because X and Y do not take place at the same time, place and in the same sense. They take place at 2 different times, places and at 2 different sense, namely in 2 different persons. That is what makes it subjective, personal, biased and so on.

With regards

With due respect, while anti-realism may explain the fact that moral disagreements are common, this is a weak argument that anti-realism is correct. Indeed, both realism and anti-realism are consistent with the existence of moral disagreements.

Moral disagreements are common and hard to settle.
Realist: The fact that people disagree and have difficulty finding a method to arrive at the same conclusion is not evidence there is no objective fact. To use a common example, we may not know how to determine whether it rained in Poughkeepsie on Jan 17, 1313, but this is nonetheless an objective claim.
Anti-realist: People disagree because there is no objective claim here at all, but merely a matter of personal opinion, emotional response, etc.

Moral disagreements are relatively rare, people tend to agree.
Realist: This is because opinions on moral claims are slowly converging to the truth, and hence agreement.
Anti-realist: This is simply a fact about the tendency of human opinion, of purely psychological note. Perhaps people tend to prefer chocolate over vanilla, but we would not conclude that chocolate is therefore objectively better than vanilla.

To be sure, Mackie raised the same issue of moral disagreement in a famous paper on moral subjectivism (really skepticism), but I've always found it quite a weak argument.

 

[phiwum]

Of course it's not incoherent. I think I've made it pretty clear throughout that although science cannot as yet answer normative moral questions, the potential is there.

This has not been clear to me.

Here are some quotes on the subject that suggest science is incapable of ever deriving a norm.

I’m saying more than that, namely that normative moral questions are irrelevant. E.g. Science can tell us that the world is at risk of a global epidemic, but it cannot make a normative moral argument that this is a bad thing and that we should do something about it.

Given that only a philosophical argument (as opposed to a scientific one) can arrive at an “ought” statement...

Indeed you can’t test normative claims.

Those last two, in particular, seem pretty final. Only crappy philosophical arguments arrive at ought statements, and normative claims are untestable.

And here are some posts that suggest that maybe science can derive norms.

So, again, do you think science can answer normative moral questions?

Potentially yes.

The gist is that it’s too simplistic to claim that science is unconcerned with morality and, at least potentially, morality could be encompassed within naturalism.

The gist of my position is that it’s too simplistic to claim science is unconcerned with morality and, at least potentially, that morality could be encompassed within naturalism.

The last two quotes (really the same quote) are a bit vague, since you've elsewhere claimed that moral theories aren't really normative in nature (except for those crap philosophy moral theories).

The heart of a moral theory is ought statements.

Not really. This claim only applies to a deductive philosophical argument.

So, no, it has not been clear that what you mean is today's science can't arrive at normative claims, but the science of the future maybe can. How would that work? Do you expect normative claims to become testable?

Given that the role of science is to investigate the natural, material universe I think Sam Harris is on the right track in his 'Moral Landscape', when he argues that enormous advances in the cognitive sciences should, in principle at least, be able to establish an objective moral code and answer normative moral questions. Hence my response.

It's not clear to me what the relation between cognitive science and moral norms is. Cognitive science might be able to determine what norms we tend to agree with, but surely cannot determine what norms are indeed true (if any).

 

[Tommy Jeppesen]

With due respect, while anti-realism may explain the fact that moral disagreements are common, this is a weak argument that anti-realism is correct. Indeed, both realism and anti-realism are consistent with the existence of moral disagreements.

Moral disagreements are common and hard to settle.
Realist: The fact that people disagree and have difficulty finding a method to arrive at the same conclusion is not evidence there is no objective fact. To use a common example, we may not know how to determine whether it rained in Poughkeepsie on Jan 17, 1313, but this is nonetheless an objective claim.
Anti-realist: People disagree because there is no objective claim here at all, but merely a matter of personal opinion, emotional response, etc.

Moral disagreements are relatively rare, people tend to agree.
Realist: This is because opinions on moral claims are slowly converging to the truth, and hence agreement.
Anti-realist: This is simply a fact about the tendency of human opinion, of purely psychological note. Perhaps people tend to prefer chocolate over vanilla, but we would not conclude that chocolate is therefore objectively better than vanilla.

To be sure, Mackie raised the same issue of moral disagreement in a famous paper on moral subjectivism (really skepticism), but I've always found it quite a weak argument.

What is an objective fact?