SEP: Supervenience in Ethics

[FlashDangerpants]

There is no supervenience of moral facts upon natural facts. The claim is a sleight-of-hand to magic moral facts out of thin air. And it doesn't work, just as non-moral premises can't entail moral conclusions.

Half the time in these conversation, it looks like "moral facts" means "facts about how humans feel about morals". That's why he focuses so much on mirror neurons - mirror neurons affect how people feel about the morality of various actions.

He calls it a straw man when I say it, but it's the closest thing to clarity anybody has about what he means.

Agreed. But his argument is that feelings don't come into it, because they're subjective.

The supposed supervenience is this: the moral fact that humans ought not to kill humans supervenes upon the natural fact that humans are neurologically programmed with oughtness-not-to-kill-humans.

So, by the same argument, the moral fact that humans ought to kill humans would supervene upon the natural fact that humans are neurologically programmed with oughtness-to-kill-humans.

And this is why he says morality has nothing to do with rightness and wrongness. The 'ought' has no conventional moral meaning - such as that we ought to do something because it's right to do so.

There was never any superveninence in his argument anyway, I threw something out there to give him an example to work with the other day, but he should have really known to reject it. The moral property in his weird little mental model is "oughtness" which he places at the physical level. So there's no actual way for inherence to operate, because inherence joins two dissimilar classes of property together, and he dismisses the required dissimilarity by inhering oughtness directly into the physical property class.

To put it more bluntly, his mirror neurons 'argument' insists that 'oughtness' is a property that inheres directly in the physical brain, so he has spent a decade directly arguing against supervenience in morality... but he lacks the technical scope to understand that the same class of property cannot - for reasons that are surely beyond fucking obvious - both supervene (a two stage process) and inhere (one step) to the same thing.

I am having some difficulty phrasing all this. I am trying to think of a way of putting this that VA could understand and I just don't see a path to that outcome.

[Iwannaplato]

Supervenience is a claim about dependence in changes of one thing on changes in another thing.

Agreed.

The common claim is "mental states supervene on brain states".

That means the following things are true:

If person A and person B have the same brain-state, they have the same mental state.
If person A and person B have different mental states, they must also have different brain states.

I surely hope VA does not read my response, since I don't want him to use anything I say here to defend some of the confused conclusions he has been arriving at. That said, I am more interested in discussing the issue with you, since that seems more likely to get somewhere.

So, that said: I would say that two people cannot have the same mental state, given the differences in, well, their brains or minds or processes. This is being fussy and I realize that in some contexts it might be useful to talk about 'the same brain states' if enough factors are close enough to make this useful.

However, there are potentially situations where person A and person B have different brain states but the same mental state.

YES and this is where supervenience analyses is a problem for moral realism and certain beliefs based on 'oughtnesses' in the brain. I went in this issue here:

viewtopic.php?p=720613#p720613

We have three people, all of whom consider X to be immoral.
Person 1: considers X immoral, and upon examination we find this person is very conformist and wants to fit in. This is all reflected in this person's 'brain state' as shown in MRI sequences.
Person 2: considers X immoral, and upon examination we find this is based on disgust related to bodies. This is all reflected in this person's 'brain state' as shown in MRI sequences.
Person 3: considers X immoral, and upon examination we find this is based on an intuitive sense that if people X, certain problems will arise. This is all reflected in this person's 'brain state' as shown in MRI sequences.
Person 4: considers X immoral, and upon examination we find this is based on an analysis of consequences in carefully thought out cause and effect processes. This is all reflected in this person's 'brain state' as shown in MRI sequences.
Person 5: considers X immoral, and upon examination we find this is based on scriptural injunction. This is all reflected in this person's 'brain state' as shown in MRI sequences.

Even within each of these categories there will be wild variations in images and secondary feelings, hormonal levels, which parts of the brain are engaged, what neurons related to different linguistic aspects of their position, and so on and on.

There is no common oughtness. This oughtness is at best a useful fiction. It is certainly a noumenon, so, according to VA must be false and unreal. It cannot be sensed with devices or without them. All sorts of 'things' will be sensed by the devices. Unlike a 'tumor' say, where one finds 'it' in the image of MRIs. There it is. There will be just a myriad of brain states, all unique. And even granting that these device created images are 'sensory' or dealing with Kantian phenomena, we don't find the oughtness nor have we any scientific justification at all for calling anything in these images an oughtness. Within realist neuroscience we have justifications for calling things neurons and labels associated with certain brain states. But nowhere in the field of neuroscience do we find images of 'oughtnesses' in brains. Not that an instrumentalist would be beholden to such an interpretation/reification/fiction.

The same formulation effectively applies to all other uses of 'supervenience'. So, "moral facts supervene on natural facts" means:

If state A and state B have the same natural facts, they have the same moral facts.
If state A and state B have different moral facts, they must also have different natural facts.
However, there are potentially situations where state A and state B have different natural facts but the same moral facts.

and, of course, as both of us have pointed out. Every moral rule, sentiment, feeling (empathy, rage, whatever) belief, assertion, including all the contradictory ideas and feelings about morality and behavior
supervene on brain states.

So, we have millions of moral facts, including contradictory ones. Morality exists, yes. But no particular moral stance has somehow become a fact.

People who get pIeasure raping have mentaI states that supervene on brain states. This does not mean rape is a moraI fact.

Brain states that lead to aggression are just as common universal as any other brain states, if one wants to build an argument on the commonness of a brain state.

So far, your only example is about the mirror neurons of a person. "If you change the mirror neurons of this person, that changes his moral facts in some way." As far as I can tell, the only moral fact that changes about a person if you change their mirror neurons is how they feel about the morality of various actions.

(adn as an aside, there are other brain patterns that may be more important for empathy.)

A person with a shit load of mirror neurons might be unwilling to break someone's knee caps for the mafia, because it feels wrong.

A person with no mirror neurons might be willing to, because it doesn't really feel like anything at all to break the kneecaps of another human.

Changing mirror neurons affects how people feel about the morality of things they might do. If that's not the kind of "moral fact" you intend to talk about, VA, you aren't doing a good job of clarifying your position.

People are showing curiosity in your ideas but you aren't putting any real effort into clarifying what they are.

We go right to the concIusion.

If state A and state B have the same natural facts, they have the same moral facts.
If state A and state B have different moral facts, they must also have different natural facts.
However, there are potentially situations where state A and state B have different natural facts but the same moral facts.

Give us some more edifying examples of moral supervenience using the above three lines as a guide.

If state A and state B have different aspects and they are associated with different acts, behaviors, attitudes and feeIings in reIation to other peopIe. How do we judge one as moraI and the other as immoraI?
What are the criteria for that judgment, given that both states produce, as VA has argued, moraI facts?

Whatever those criteria are THAT is where he is actuaIIy getting his moraI facts from, not from mentaI states and not from brain states. The entire 'brain justifies some moraIity objectiveIy' is a misdirection that seems to have fooIed even the magician.

If brain states determine moraI facts, then the current state of human interactions is perfect as it is and does not need changing.

Because the brain states, which supposedIy determine moraI facts are the ones that current interpersonaI behavior, attitudes, feeIings and ruIes are supervened on.

This post was written by a human being.

What a coincidence. So was mine.

[Flannel Jesus]

If state A and state B have different aspects and they are associated with different acts, behaviors, attitudes and feeIings in reIation to other peopIe. How do we judge one as moraI and the other as immoraI?
What are the criteria for that judgment, given that both states produce, as VA has argued, moraI facts?

Whatever those criteria are THAT is where he is actuaIIy getting his moraI facts from, not from mentaI states and not from brain states.

And not from "supervenience" either. Supervenience is trivial at best, and doesn't seem to do any work convincing anybody of anything.

You're exactly right, what are the criteria? How do we derive the criteria? These conversations are infinitely more important than any amount of words someone might want to write about 'supervenience' here. It's just a fancy word for a simple concept, and the fancy word doesn't do any work towards proving objective morality, or anything else.

[Flannel Jesus]

So, that said: I would say that two people cannot have the same mental state, given the differences in, well, their brains or minds or processes. This is being fussy and I realize that in some contexts it might be useful to talk about 'the same brain states' if enough factors are close enough to make this useful.

I think you're talking in practical terms - practically speaking, EVERY human being has a unique physical brain state, yes - but it's meant in more hypothetical terms. To talk about what we think brains are, and minds are, and how they interact if at all.

Hypothetically, 2 physically identical brains that stay perfectly identical for, say, 10 seconds, are also having the same mental states for those 10 seconds, if you believe mental states emerge from brain states. If you don't believe mental states emerge from brain states, then you might not think this.

[Iwannaplato]

It's just a fancy word for a simple concept, and the fancy word doesn't do any work towards proving objective morality, or anything else.

Yes. It couId be part of an essay demonstrating that moraIity is objective. Or, reaIIy, moraIities exist. But no one contests that. PH is aware that moraIities exist. It's whether any particuIar one is objective is the issue.

And supervenience just gives moraI antireaIists another way to say no.

They can say fine. Assume brain states confer objectivity on the attitudes and behaviors that are supervened on those brain states.
Every singIe behavior and attitude is supervened on brain states.
Therefore they are aII objective.
Which means that aII behaviors and attitudes are moraIIy good. Even opposed ones.

So there's nothing that needs to be done, we Iive in the best of aII possibIe worIds.

Everyone out there is aIready acting within the objective moraIity.

Now, VA wiII respond that this is a strawman. That he doesn't beIieve everyone is acting moraIIy now.

WeII, of course.

VA: PIease, PIease communicate with an AI for a month about the difference between
your argument entaiIs X
and
you beIieve and have asserted X.

Be cIear you need exampIes of the difference.

Here's one.

Person A: Good peopIe teII the truth.
Stan toId the truth yesterday.
Stan is a good person.

Person B: HitIer did teII the truth on occasion.
Your argument entaiIs that HitIer is a good person.

Person A: Strawman. I never asserted, nor did I beIieve, that HitIer was a good person.

Person B: Saying your argument entaiIed that concIusion does not mean you beIieve that concIusion, nor does it mean you asserted the concIusion. In fact the whoIe point of the exampe I chose was that you obviousIy wouIdn't draw that concIusion that HitIer was good and you obviousIy don't beIieve that and SO you need to examine your argument.

[Iwannaplato]

So, that said: I would say that two people cannot have the same mental state, given the differences in, well, their brains or minds or processes. This is being fussy and I realize that in some contexts it might be useful to talk about 'the same brain states' if enough factors are close enough to make this useful.

I think you're talking in practical terms - practically speaking, EVERY human being has a unique physical brain state, yes - but it's meant in more hypothetical terms. To talk about what we think brains are, and minds are, and how they interact if at all.

Hypothetically, 2 physically identical brains that stay perfectly identical for, say, 10 seconds, are also having the same mental states for those 10 seconds, if you believe mental states emerge from brain states. If you don't believe mental states emerge from brain states, then you might not think this.

Agreed.

[FlashDangerpants]

Every moral rule, sentiment, feeling (empathy, rage, whatever) belief, assertion, including all the contradictory ideas and feelings about morality and behavior
supervene on brain states.

They what now?

[Iwannaplato]

Every moral rule, sentiment, feeling (empathy, rage, whatever) belief, assertion, including all the contradictory ideas and feelings about morality and behavior
supervene on brain states.

They what now?

Yup. The Then what now is the Ieft out part or VA's argument. He has made some tiny forays into the next step, but the probIem is that those nexts steps wiII have to incIude the reaI authority for his particuIar moraIity
and that authority is not science. And it is very hard to confer on whatever that authority is a status of 'objectivity'. Even antireaIist objectivity.

This has been the case for years.
and now supervenience has given him another way to assert the exact same argument without putting the reaI source of his moraIity on the tabIe.

What are the criteria - because it sure ain't brain states - HitIer and StaIin and his pet peeve MusIims aII have brain states. So we are on square one, right where he's done with his argument. Zero.

[Flannel Jesus]

Every moral rule, sentiment, feeling (empathy, rage, whatever) belief, assertion, including all the contradictory ideas and feelings about morality and behavior
supervene on brain states.

They what now?

Yup. The then what now is the Ieft out part.

I think you misread a tiny bit.

[Iwannaplato]

They what now?

Yup. The then what now is the Ieft out part.

I think you misread a tiny bit.

I thought it was a typo.

I stand by my response to what he didn't say. But yes, I see he wants me to justify that supervenience. I'II Ieave that to VA. I accepted that, since it's not the probIem with his argument.

[Flannel Jesus]

Yup. The then what now is the Ieft out part.

I think you misread a tiny bit.

I thought it was a typo.

That did not occur to me, but it's a reasonable possibility.

[Iwannaplato]

I think you misread a tiny bit.

I thought it was a typo.

That did not occur to me, but it's a reasonable possibility.

No, I beIieve you were right.

They what now?
rather than...
Then, what now?

I projected where I was at in reIation to VA's argument onto his more generaI question.

[FlashDangerpants]

If state A and state B have different aspects and they are associated with different acts, behaviors, attitudes and feeIings in reIation to other peopIe. How do we judge one as moraI and the other as immoraI?
What are the criteria for that judgment, given that both states produce, as VA has argued, moraI facts?

Whatever those criteria are THAT is where he is actuaIIy getting his moraI facts from, not from mentaI states and not from brain states.

And not from "supervenience" either. Supervenience is trivial at best, and doesn't seem to do any work convincing anybody of anything.

You're exactly right, what are the criteria? How do we derive the criteria? These conversations are infinitely more important than any amount of words someone might want to write about 'supervenience' here. It's just a fancy word for a simple concept, and the fancy word doesn't do any work towards proving objective morality, or anything else.

I think we should always take care when VA is floundering around misrepresenting something to avoid coming to the tempting but misleading judgement that it must therefore be total bullshit. It is only VA's use of the concept of supervenience to incorrectly describe his own inherence theory of "oughtnesses" that is illiterate garbage for morons.

Superveninence is important wherever we have a state of reliance of some sort of properties of one kind that are not reducible to propoerties of another, theoretically simpler kind, but are not free of them either. Or at least, supervenience is important if you have important things to say about them not being remote from those other properties.

A paper by Richard Boyd has cropped up several times in this forum (mainly because it was VA's most legendarily shit misreading fail and I like to annoy him with it once per year ever since). When it isn't accusing competing philosophers of having brain damage (ahem), It is about demonstrating a supervenience relation between moral judgments and a class of non-moral ones. It never mentions brain states. Here's a snippet...

1. To an extremely good first approximation, moral judgments regarding actual cases of actions, policies, character traits, etc. are—given prevailing standards of moral argument—dictated by judgments regarding nonmoral factual questions (including, for example, questions about human nature, about the nature of social, political, and economic processes, about whether or not there are any gods, and about their natures if there are any…). In consequence, moral disagreements regarding such actual cases can be seen (on a philosophically appropriate rational reconstruction) as stemming from disagreements over nonmoral factual matters. (I call this relationship therational supervenience of the relevant moral judgments on nonmoral factual judgments.)

• Rational supervenience appears to fail for a few actual cases and for many counterfactual ones. For almost all of these it is plausible to argue that the cases in question are ones in which there is unrecognized failure of bivalence. For the few remaining cases of apparent failures of rational supervenience realist explanations in terms of nonculpable inadequacies in methodology or theoretical understanding are readily available. This conception of the sources of failures of rational supervenience is itselfratified by all of the genuinely plausible competing general moral theories.

• The conditionalized theory of epistemic contact and of error upon which the plausible competing general moral theories agree is such that it attributes differences in judgments regarding general moral theories to differences over nonmoral factual matters. Thus, rational supervenience upon nonmoral factual judgments obtains for general moral theories as well as for particular moral judgments

[Flannel Jesus]

Can you break down some examples for us of why that analysis is true and/or interesting? It's not obvious to me.

He says there are situations where that type of supervenience breaks down - what situations?

[FlashDangerpants]

Can you break down some examples for us of why that analysis is true and/or interesting? It's not obvious to me.

He says there are situations where that type of supervenience breaks down - what situations?

Here I must admit that I read that thing years ago, and especially for your second question, I would have to go back over it to remember what that is about. In fact the only reason I use it for this reference at all is because I knew I had easy access to the PDF version and could cut and paste from it. The essay itself isn't even about supervenience, it's just a topic that crops up along the way.

However, at one point in VA's erruption of AI bullshit, the thing that made the least possible sense in my view was the AI claim that J.L. Mackie's argument from queerness is an argument on behalf of supervenience. This is the passage in question from Ethics: Inventing Right and Wrong....

Another way of bringing out this queerness is to ask, about anything that is supposed to have some objective moral quality, how this is linked with its natural features. What is the connection between the natural fact that an action is a piece of deliberate cruelty-say, causing pain just for fun-and the moral fact that it is wrong? It cannot be an entailment, a logical or semantic necessity. Yet it is not merely that the two features occur together. The wrongness must somehow be 'consequential' or 'supervenient'; it is wrong because it is a piece of deliberate cruelty. But just what in the world is signified by this 'because'? And how do we know the relation that it signifies, if this is something more than such actions being socially condemned, and condemned by us too, perhaps through our having absorbed attitudes from our social environment? It is not even sufficient to postulate a faculty which 'sees' the wrongness: something must be postulated which can see at once the natural features that constitute the cruelty, and the wrongness, and the mysterious consequential link between the two. Alternatively, the intuition required might be the perception that wrongness is a higher order property belonging to certain natural properties; but what is this belonging of properties to other properties, and how can we discern it? How much simpler and more comprehensible the situation would be if we could replace the moral quality with some sort of subjective response which could be causally related to the detection of the natural features on which the supposed quality is said to be consequential.

On the one hand, I hope you can see why I found it so scandalous that VA would malign the originator of modern Moral Error Theory as some sort of supporter for his assertion that supervenience is both non controversial and supports moral realism. Let alone the audacity of referencing his most famous argument to do so.

The more important part just right now is the bit about "It is not even sufficient to postulate a faculty which 'sees' the wrongness", this, as I recall was what Boyd was trying to address for much of his paper, by attempting to render morality as a faculty of perception more than reflection or reason.

Boyd's paper iirc argues that in the vast majority of all moral cases, all the various forms of moral realist (intuitionism, contractarianism, and so on) all broadly agree about what is right and what is wrong in spite of the suspicious variety of ways they go about finding out what is right and wrong... he puts that down to the fact that it's usually blindingly obvious what is right and wrong, and that we can see it even if, in the face of Mackie we can't actually see the whole end to end story with the natural feature, the wrongness and the mysterious something that sits betwixt.

But he's somewhat vague about details. Mostly I think he is jsut saying there must be something there to supervene upon and a mechanism for such supervenience must exist... and it must fall within a broad range of explanations that are based on scientific realism of a limited form that goes as far as saying that scientific explanations have explanatory power because they are approximately accurate descriptions of real phenomena blah blah blah... and so moral propositions must be true or flase in terms of soemthing that supervenes them into that scientifically realist description of the world and the way in which worldly properties operate.

His own specific favourite form of such explanation is the Homeostatic Property Cluster, some would view this thing as a cheeky little non-explanation though. It amounts to (take care thisis off the top of my head) a way of defining things that avoids being too strict to successfully define anything. Think Wittgenstein's family resemblance routine. So he's probably sort of doing supervenience on a second attribute here. Not just morals but Natural Kinds as well (not my area, I know that there's a controversy about it, but I have nothing to offer on that matter).

Health was one of his clusters, nobody can properly define "Healthy" in a way that includes all the things that represent good health but none of the things which represent corruption or excess or something that is just obviously unhealthy. We all know what it is though, sort of, and we all know that VA's obsession with Islam for instance is not healthy. Attofishpi would disagree, but he's clinically insane. Sanity, that's another cluster right there.

Anyway, Boyd recommends moral supervenience proceeds somehow through these clusters. But his argument doesn't go all the way, it stops well short and he only says that something within this very broad range of types of description should be further investigated to find a really solid brand of ethical consequentialism that's not utilitarian in the least. That was in about 1988, perhaps he got further than that in the end and wrote about ti somewhere I haven't read.