Philosophy Discussion Forums | A Humans-Only Club for Open-Minded Discussion & Debate

[Gertie]

GE

I'm running out of steam on this.

So the fact that we humans create a model of the world which includes a model of our self within it, has no apparent bearing on how experience arises. Far less complex experiencing animals probably don't create such a model. It doesn't look like a necessary condition for mental experience. And if it's not, copying the creation of that 'model maker within the model' function won't make any difference to whether an AI can experience.

Well, sure it has a bearing. I think there is pretty widespread agreement among modern philosophers (hardcore naive realists excepted) that the phenomenal world, the world we experience, is a conceptual model of a hypothetical external, "noumenal" world which we can never experience directly. That experienced world is constructed of impressions --- sensations, concepts, feelings, etc. --- that are intangible, subjective, and intrinsically private, but which somehow represent, and are elicited by, states of affairs in that presumed external world ( which includes one's --- presumed --- physical body). Hence a creature which can create such a model will be conscious, by definition.

And I disagree that "less complex animals don't create such a model." I think we should assume that any animal with a nervous system complex enough to support one does create such a model. Amoebae? No. Vertebrates and even some insects? Yes --- probably. Honeybees' brains consist of about 1 million neurons --- more than enough to construct at least a rough conceptual model of their environment. And they exhibit behaviors and capabilities that not long ago were thought to be restricted to primates.

Read back, you've missed my original point. I'll repeat it. There's nothing special about a model which includes the model maker which is likely to be a necessary condition for experience. There's no reason to think an AI copying that model-maker-within-the-model feature will help enable it to experience.

Yeah could be. It leaves you with the problem of not knowing if AI is the right type of wire.

Well, that is the central issue here --- how will we ever know, other than by observing the system's behavior? Do you really want a theory that leaves that question permanently open --- that is empirically unconfirmable and unfalsifiable?

A question which isn't answered is an open question. A theory which empirically unconfirmable and unfalsifiable is called a hypothesis, it's necessarily speculative. It's a What If. Do you really want to pretend it isn't?

Maybe. But to assume the observable behaviour resulting from biological stuff and processes is less likely to be coincidental/superficial than the biological stuff and processes itself would be ****-backwards imo.

Well, that is not what I'm suggesting. I think that biological stuff, of a certain kind and arranged in certain ways, will produce consciousness. But also that non-biological stuff, or non-natural biological stuff will also produce consciousness, when arranged in analogous ways.

Maybe.

And again, the only means we have, or will ever have (given what we do know about the problem) for deciding whether the biology is critical is by observing the system's behavior. You seem to be holding out for some future "transcendental" insight into this issue. But for now, and for the foreseeable future, behavior is all we have.

Just don't say behavioural tests are reliable.

Pragmatically perhaps, but that doesn't make it reliable.

What would?

A Theory of Consciousness which explained the necessary and sufficient conditions, which we could then test for.

Look at this way - why do we assume other humans have experiences like us?
- They are physically almost identical, and brain scans show similar responses to similar stimuli, which match similar verbal reports to ours.
- Their observable behaviour is experientally understandable to us, in that we can imagine behaving similarly in similar circs.
It's all about similarity. That's why the hope is that if we create an AI sufficiently similar to a human, it will somehow capture the necessary and sufficient conditions for experience.

As pointed out before, your first similarity there is insufficient, and may be irrelevant.

It might be insufficient and irrelevant, you don't know.

The brain-dead person is also physically similar to us, but not conscious --- a judgment we make based on the lack of conscious behavior.

We make that judgement because experience as we embodied humans experience it is obviously dynamic, changing moment to moment. Like a steam train in motion, not like a bee which makes honey then goes off again about its bee business. The brain stops working when we die, all those biological electrochemical processes cease. The point is AI don't have the same biological electrochemical processes.

And we can correlate brain scan information with perceptual phenomena only if it results in observable behavior. That is the only means we have of knowing --- inferring --- what perceptual phenomena is occurring (in anyone other than ourselves).

And our self reports. What scans confirm is that some types of specific biological, electro-chemical activity correlate to consistent self-reports of specific types of experience by biological humans. We then reasonably assume that certain types of biological electrochemical interactions possess the necessary and sufficient conditions for experience.

Not stubbornness. Just because it's the best we can do doesn't mean it's reliable. We might be forced to act as if it's reliable, but we should realise that's what we're doing.

Still holding out for that transcendental insight, eh?

Not my point. My point, which I'm repeating over and over now, is that just because observed behaviour is the only available way of testing AI, doesn't mean it's reliable. Because we don't know if the AI's substrate will capture the nec and sufficient conditions.

Anyway, I'm done with just repeating this same obvious point.


Why is it so hard to just say you don't know - nobody does?

[GE Morton]

Physical, on my account, as I've probably written at least 20 times or so here over the years refers to materials, relations of materials and processes (dynamic relations) of materials. Those three things do not seem to be separable in reality, just conceptually. They all amount to properties, too. Or in other words, properties are just another way of talking about materials, relations and processes.

Circular and uninformative. "Material" is merely a synonym for "physical," "material thing" for "physical thing."

As I've suggested before, you need to abandon these hokey, spurious definitions of common terms and stick with the dictionary:

Physical (adjective):

"1a: of or relating to natural science
b(1): of or relating to physics
(2): characterized or produced by the forces and operations of physics
2a: having material existence : perceptible especially through the senses and subject to the laws of nature."

https://www.merriam-webster.com/dictionary/physical

Qualia are not going to be merely "produced" by physical things, where qualia are not identical to physical things.

Oh? Why not? Are you suggesting that X cannot produce Y unless X and Y are identical? Is cotton fabric identical to the textile mill that produced it? A musical note identical to the flute that produced it? And, of course, physical things produce all manner of non-physical things. Humans (physical things) produce non-physical ideas, laws, theories, religions, moralities, etc. Hurricanes (physical things) produce worry, fear, grief. The world is full of non-physical things produced by physical things.

"Physical" in philosophy, is obviously not going to amount to " analyzable via and predictable from the laws of physics as they're presently instantiated in the science of physics" . . .

Well, that depends upon whose philosophy you have in mind.

. . . because it's not as if we're wondering if qualia is something that's covered or at all near being covered in physics textbooks.

It is not covered in physics textbooks because physicists don't consider qualia to fall within their purview. Neither do most philosophers. To make qualia "physical" you either need some hokey definition of "physical" or to claim they are "identical" with something physical, per some hokey definition of "identical."

[GE Morton]

GE

I'm running out of steam on this.

Well, we both seem to think the other is missing, or misunderstanding, the other's points and hence not addressing them.

Mine is this: Behavior is the only criterion we have, or will ever have, for determining whether a system other than ourselves is conscious. That is because the "stuff" of consciousness, thoughts, ideas, qualia, desires, moods, even dreams, are necessarily and impenetrably private, inaccessible to any third-party observer. I'll try to make that case by going through the other points in your comment:

Read back, you've missed my original point. I'll repeat it. There's nothing special about a model which includes the model maker which is likely to be a necessary condition for experience. There's no reason to think an AI copying that model-maker-within-the-model feature will help enable it to experience.

We can have no idea whether any particular property of a system is "special," in the sense of being necessary for consciousness, unless we construct a system with a candidate property and observe its behavior. If an AI's behavior, over as wide a variety of situations you wish to obseve, is indistinguishable from that of humans (in relevant ways) in similar situations, then that certainly is evidence that it is consicous, whether it is "copying" human behavior or not. It is the only kind of evidence we'll ever have.

A question which isn't answered is an open question. A theory which empirically unconfirmable and unfalsifiable is called a hypothesis, it's necessarily speculative. It's a What If. Do you really want to pretend it isn't?

Not correct re: hypothesis. An hypothesis is a cognitive proposition whose truth value is unknown, but can be determined by experiment ("hypothesis testing"). Theories are not hypotheses; they are neither true nor false, but sound or unsound --- they either generate testable propositions and predictions, or they don't. A theory which yields no testable propositions, is unconfirmable or unfalsifiable, is unsound; vacuous. A theory which suggests that things or systems whose behavior is not indicative of consciousness may nonetheless be conscious is vacuous, since there is no other way, in principle, to confirm/disconfirm such a claim.

Just don't say behavioural tests are reliable.

They are sometimes not reliable in the short run (e.g., a wide-awake person may be feigning sleep, or unconsciousness). But they are quite reliable over an extended period of observation. But speaking strictly, we can't even assess their reliability, because we can only assess the reliability of some chosen method by comparing it with another method --- and we have no other method. That makes behavior the decisive criterion for consciousness.

What would?

A Theory of Consciousness which explained the necessary and sufficient conditions, which we could then test for.

We have no means of knowing what conditions are necessary or sufficient, i.e., whether that theory is sound, other than by implementing those conditions and observing the resulting behavior.

The brain stops working when we die, all those biological electrochemical processes cease. The point is AI don't have the same biological electrochemical processes.

Yes, some functioning, physical substrate is necessary for consciousness (per all of the evidence we have). But whether that particular substrate is necessary can only be determined by experimenting with other substrates and observing the system's behavior. If that behavior is unquestionably affirmative for consciousness, then there is no room for further doubt about whether the system is "really" conscious. "Really" has no meaning there; it refers to nothing testable or observable.

And our self reports. What scans confirm is that some types of specific biological, electro-chemical activity correlate to consistent self-reports of specific types of experience by biological humans. We then reasonably assume that certain types of biological electrochemical interactions possess the necessary and sufficient conditions for experience.

Of course they do. But that is not to say that is the only type of system which can implement those conditions. Also, keep in mind that self-reports are themselves behaviors.

Not my point. My point, which I'm repeating over and over now, is that just because observed behaviour is the only available way of testing AI, doesn't mean it's reliable. Because we don't know if the AI's substrate will capture the nec and sufficient conditions.

If behavior is deemed an unreliable indicator of consciousness then we can never know whether any system, including other humans, are "really" conscious, or what are the necessary and sufficient conditions for it (this is, of course, the topic of the voluminous "zombie" literature). That is because phenomenal experience is intractably private, and forever inaccessible to third party observers.

[Steve3007]

Physical, on my account, as I've probably written at least 20 times or so here over the years refers to materials, relations of materials and processes (dynamic relations) of materials. Those three things do not seem to be separable in reality, just conceptually. They all amount to properties, too. Or in other words, properties are just another way of talking about materials, relations and processes.

Simply saying "physical = material" doesn't advance the cause of providing a useful definition of "physical". It just makes it a task of providing a useful definition of "material".

"Physical" in philosophy, is obviously not going to amount to " analyzable via and predictable from the laws of physics as they're presently instantiated in the science of physics" because it's not as if we're wondering if qualia is something that's covered or at all near being covered in physics textbooks. We could just look at a physics textbook and check, obviously. Likewise, we're not wondering if anatomy is at all covered or near being covered in physics textbooks, but there's no doubt that anatomy is physical. Furthermore, one does not need to be a realist on physical laws to be a physicalist.

I don't think many people would suggest that "physical" means "relating to physics as it currently happens to be". As I've said a few times myself, I think the only useful (as opposed to empty/circular) definition of "physical" is something like "the things we propose to be the common causes of, or patterns in, diverse potential and actual sensations.". Since physics is a fundamentally empirical subject, I think a reasonable shorthand is therefore to say that "physical" means "the kinds of things that physics studies".

And "physical" is obviously not going to refer to some colloquial nonsense of whether we can "touch" something, or see it with our naked eyes, etc.

Talk about whether something can potentially be touched or seen, in defining "physical" informally, is not a bad route to take in my view.

Circular and uninformative. "Material" is merely a synonym for "physical," "material thing" for "physical thing."

As I've suggested before, you need to abandon these hokey, spurious definitions of common terms and stick with the dictionary:

Physical (adjective):

"1a: of or relating to natural science
b(1): of or relating to physics
(2): characterized or produced by the forces and operations of physics
2a: having material existence : perceptible especially through the senses and subject to the laws of nature."

On this point, I agree.

[Pattern-chaser]

I just followed that link, and Malwarebytes said "Website blocked due to reputation". I decided not to proceed, but posted this because I thought we should be aware of a possible problem?

Don't know what you did, here it says the site is clean according to 66 out of 66 engines.

https://www.virustotal.com/gui/url/d6fc ... /detection

That's good to hear.

My bank - would you believe it? - arranged a free subscription for its customers to Malwarebytes Premium, so I took them up on it. I was stunned! A bank doing something useful for its customers! Anyway, Malwarebytes has been around forever, and has a sound reputation based on performance and use. So when it advised me to avoid that website, I did. It's good to hear this is probably a false positive.

[Terrapin Station]

Circular and uninformative.

Can you give me an example of a definition that's not circular?

[Terrapin Station]

Physical, on my account, as I've probably written at least 20 times or so here over the years refers to materials, relations of materials and processes (dynamic relations) of materials. Those three things do not seem to be separable in reality, just conceptually. They all amount to properties, too. Or in other words, properties are just another way of talking about materials, relations and processes.

Simply saying "physical = material"

I would have pointed this out to General Electric Morton above, too, but I don't want to give him anything else that might distract him.

First, I didn't say that physical = material (period), did I? I mean, you're quoting what I said right there. It doesn't stop with the word "material(s)."

Aside from that, is the idea here that we're dealing with someone who has no grasp at all re what "physical" might refer to, so we need to find a synonymous phrase that they might have a grasp of, where we are dealing with someone who also has no grasp of what "material," "relations" etc. refers to? If we're dealing with such a person, who would have to be a very odd person, maybe from another planet or some kind of robot or something, then we'd need to proceed by trying to figure out some terms that they do have a grasp on, because otherwise we might exhaust hundreds where the person would say, "I have no idea what that is, either." That could be endless if they're odd enough.

I didn't think the idea was supposed to be that we were supposed to bootstrap, or pretend to bootstrap, someone who has no idea of what any term at all might refer to.

[Terrapin Station]

But I just did: Just answer in relation to whatever analysis of propositional knowledge you personally use--whoever you agree with, let's say.

I do NOT have an analysis of propositional knowledge that I personally use.

It's it not something you're interested in? You're not (philosophically) curious what propositional knowledge is?

[GE Morton]

1. Mental phenomena are not reducible to physical phenomena, though there is a causal relation between them.

2. Mental phenomena are not identical with physical phenomena.

3. Dualism is false, i.e., there is no "mental" (or "spiritual," "non-physical,") substance, or "stuff," of which qualia and other mental phenomena are constituted.

4. Though mental phenomena are not reducible to or derivable from the laws of physics, those laws are adequate to explain them to the extent they are explicable.

My concern is that the 4 are not obviously compatible. If we accept 1 and 2, the physical causes the mental but is not identical to it then what is it you have caused? Are we not forced into dualism?

The assumption that we are so forced rests on another assumption, namely, that whatever exists is, or is constituted from, some sort of "substance," and therefore that if X is not a substance of a given type, then it must be or be constituted from a substance of another type. But many more things exist than can be fairly characterized as substances.

The concept of substance has been around since the inception of philosophy, a matter of central concern since the extensive discussions of the subject by Aristrotle. There is nothing like a consensus on what "substance" is, on how that term should be understood. In addition to the several analytical definitions that have been suggested, the term also has many connotations, which come to the fore in different contexts.

The Stanford Encyclopedia has an extensive review article on the topic:

https://plato.stanford.edu/entries/subs ... onstructed.

The author lists several features, or qualities, that various philosophers have taken to be descriptive, if not definitive, of "substance:"

i. being ontologically basic—substances are the things from which everything else is made or by which it is metaphysically sustained;
ii. being, at least compared to other things, relatively independent and durable, and, perhaps, absolutely so;
iii. being the paradigm subjects of predication and bearers of properties;
iv. being, at least for the more ordinary kinds of substance, the subjects of change;
v. being typified by those things we normally classify as objects, or kinds of objects;
vi. being typified by kinds of stuff.
vii. (Kant) those enduring particulars that give unity to our spatio-temporal framework, and the individuation and re-identification of which enables us to locate ourselves in that framework.

The first three are probably the most widely shared, and closest to what the "common man" understands by the word, especially # iii. That conception is embodied in the usual way we speak about things, via declarative sentences in which we attribute a predicate, denoting some property, to a subject. The subject "thing" is substance, or composed of some more fundamental substance, and the properties --- universals --- though they exist, are not substances (nominalists deny the existence of universals altogether).

So everything consists of some sort of substance, to which some sorts of properties attach. Different sets of properties may apply to substances of different categories (so that trying to apply a property to substance of the wrong type for that category of properties is a "category mistake").

But this entire ontology of substances which take on properties is derived from contemplation of public things, and serves us more-or-less well for that purpose. But it has no room for existents that are neither substances nor universals --- such as qualia. So we try to force them into that framework.

I think the evidence is consistent with identity, (depending on your criteria for identity as per), and this is simpler, one mystery, rather than two (i.e. what is this separate mentality as well as how is it caused).

Yes, it simplifies things. Unfortunately, the two things in question are not identical per the ordinary criteria for declaring two things to be identical.

Point 4., that mentality is not derivable from the laws of physics though these laws are adequate to explain it is interesting but needs unpacking. Could we argue that a frog is not directly derivable from the laws of physics but physics can explain a frog in the context of evolution?

A frog is derivable from the laws of physics, but not predictable by them. Qualia are predictable by the laws of physics (per the cognitive model theory), but not derivable from them.

No, if qualia are not physical we are not forced into dualism. What they force us to do is re-examine our ontological assumptions.

[Atla]

My concern is that the 4 are not obviously compatible. If we accept 1 and 2, the physical causes the mental but is not identical to it then what is it you have caused? Are we not forced into dualism?

The assumption that we are so forced rests on another assumption, namely, that whatever exists is, or is constituted from, some sort of "substance," and therefore that if X is not a substance of a given type, then it must be or be constituted from a substance of another type. But many more things exist than can be fairly characterized as substances.

The concept of substance has been around since the inception of philosophy, a matter of central concern since the extensive discussions of the subject by Aristrotle. There is nothing like a consensus on what "substance" is, on how that term should be understood. In addition to the several analytical definitions that have been suggested, the term also has many connotations, which come to the fore in different contexts.

The Stanford Encyclopedia has an extensive review article on the topic:

https://plato.stanford.edu/entries/subs ... onstructed.

The author lists several features, or qualities, that various philosophers have taken to be descriptive, if not definitive, of "substance:"

i. being ontologically basic—substances are the things from which everything else is made or by which it is metaphysically sustained;
ii. being, at least compared to other things, relatively independent and durable, and, perhaps, absolutely so;
iii. being the paradigm subjects of predication and bearers of properties;
iv. being, at least for the more ordinary kinds of substance, the subjects of change;
v. being typified by those things we normally classify as objects, or kinds of objects;
vi. being typified by kinds of stuff.
vii. (Kant) those enduring particulars that give unity to our spatio-temporal framework, and the individuation and re-identification of which enables us to locate ourselves in that framework.

The first three are probably the most widely shared, and closest to what the "common man" understands by the word, especially # iii. That conception is embodied in the usual way we speak about things, via declarative sentences in which we attribute a predicate, denoting some property, to a subject. The subject "thing" is substance, or composed of some more fundamental substance, and the properties --- universals --- though they exist, are not substances (nominalists deny the existence of universals altogether).

So everything consists of some sort of substance, to which some sorts of properties attach. Different sets of properties may apply to substances of different categories (so that trying to apply a property to substance of the wrong type for that category of properties is a "category mistake").

But this entire ontology of substances which take on properties is derived from contemplation of public things, and serves us more-or-less well for that purpose. But it has no room for existents that are neither substances nor universals --- such as qualia. So we try to force them into that framework.

I think the evidence is consistent with identity, (depending on your criteria for identity as per), and this is simpler, one mystery, rather than two (i.e. what is this separate mentality as well as how is it caused).

Yes, it simplifies things. Unfortunately, the two things in question are not identical per the ordinary criteria for declaring two things to be identical.

Point 4., that mentality is not derivable from the laws of physics though these laws are adequate to explain it is interesting but needs unpacking. Could we argue that a frog is not directly derivable from the laws of physics but physics can explain a frog in the context of evolution?

A frog is derivable from the laws of physics, but not predictable by them. Qualia are predictable by the laws of physics (per the cognitive model theory), but not derivable from them.

No, if qualia are not physical we are not forced into dualism. What they force us to do is re-examine our ontological assumptions.

So your 'ephemeral qualia' can't be detected so far, and its causal relation to physical stuff can't be explained either. Its identity with physical stuff is rejected, because of semantics about 'identity', even though all the known correlations point towards their identity. Yet somehow, none of this is supposed to be a 'physical stuff - qualia' dualism either, because of substance theory, which isn't even the issue here.

[evolution]

I do NOT have an analysis of propositional knowledge that I personally use.

It's it not something you're interested in?

Once again, you pose a statement, and again about me, but add a question mark at the end of your statement.

You're not (philosophically) curious what propositional knowledge is?

And again, ANOTHER proposed statement, with ANOTHER question mark at the end of it.

If I recall correctly, I have ALREADY asked you what is 'propositional knowledge', to you? But you do have a tendency to NOT clarify or NOT answer the actual questions, which I pose to you.

[Wossname]

GE Morton » Today, 4:49 am

But this entire ontology of substances which take on properties is derived from contemplation of public things, and serves us more-or-less well for that purpose. But it has no room for existents that are neither substances nor universals --- such as qualia. So we try to force them into that framework.
I think the evidence is consistent with identity, (depending on your criteria for identity as per), and this is simpler, one mystery, rather than two (i.e. what is this separate mentality as well as how is it caused).
Yes, it simplifies things. Unfortunately, the two things in question are not identical per the ordinary criteria for declaring two things to be identical.
Point 4., that mentality is not derivable from the laws of physics though these laws are adequate to explain it is interesting but needs unpacking. Could we argue that a frog is not directly derivable from the laws of physics but physics can explain a frog in the context of evolution?
A frog is derivable from the laws of physics, but not predictable by them. Qualia are predictable by the laws of physics (per the cognitive model theory), but not derivable from them.

No, if qualia are not physical we are not forced into dualism. What they force us to do is re-examine our ontological assumptions.

Not sure about this GEM.

If we have physical and non-physical events we seem to have two sorts of events and that is dualism as I understand it. And the problem has always been to marry these two things back together once you have separated them.

The cognitive model you suggest seems to avoid the problem of how they interact by suggesting that they are somehow both the same thing and a different thing, and I am struggling to understand you. I am in a muddle with the view that there is this second thing, mentality, separate from the physical yet able to interact with it, something new and different (since it is not physical but caused by physical processes, and if I have you, can also cause them), but still it should not be considered new and different? It sounds quite close to identity theory but I clearly haven’t got it, and I accept it may be my fault. I do agree that the normal criteria for identity don’t work once we introduce both a public and private perspective.

I think of the perception of qualia as a physical event. We can identify physical events and brain processing associated with seeing red say. You seem to be suggesting that the perception of qualia is not a physical event and not a non-physical event either? And you say I would understand this if I just re-examined my ontological assumptions? Identity theory, the notion that physical things can be mental things is hard enough. It is the Hard Problem that Gertie rightly to points to and a mystery. And it is what I think your ontological reframing is seeking to crack. But I am not clear how you have cracked it. I need help from you to make sense of this reframing, because at the moment I still prefer to stick with a problem of perspective than a problem of ontology, mainly because I can’t grasp this reframing that you are proposing.

If it helps, my reasoning is that either complex physical processes may be able to produce a new thing that is non-physical (mentality) or that physical processes can also be mental ones. We don’t know how this is possible either way, the first is dualism, the second identity theory, but dualism faces the additional problem of getting the two separate things to interact, so I prefer identity. (If I understand you Atla, and I may not, so do shout Atla, your view is that consciousness is pervasive in all matter and not just in matter of sufficient complex organisation). You are seeking, I think GEM, a third option, something that is neither dualism nor identity, (we have ignored idealism from shared prejudice perhaps). I am not unwilling to re-examine my ontological assumptions, and I have tried, but I haven’t grasped your reframing yet.

[Terrapin Station]

Once again, you pose a statement, and again about me, but add a question mark at the end of your statement.

Aside from the typo, it was a question. Here it is without the typo:

Is it not something you're interested in? You're not (philosophically) curious what propositional knowledge is?

Can you answer those questions? I'll answer yours after we're through with this part. Tit for tat.

[Gertie]

GE

I'm just going to agree to differ on the AI and testing stuff unless there's something you think I haven't addressed.

More interesting to me is this -

1. Mental phenomena are not reducible to physical phenomena, though there is a causal relation between them.

How do you account for this?

If a brain causally produces something which isn't reducible to the brain, then we'd think it is acting causally on something else which is not part of the brain. But I don't think you're claiming that?

[Faustus5]

If a brain causally produces something which isn't reducible to the brain, then we'd think it is acting causally on something else which is not part of the brain. But I don't think you're claiming that?

I think GE is adhering to a very strict definition of what "reductionism" requires, given his favorable response to the definition I supplied a week or two ago in this thread. I too deny that mental states can be reduced to physical states for the same reason, but do not think of mental states as something different than brain states.