Dubious Thought experiments
I find thought experiments useful. Some of them actually have a real life example. The Mary's Room Thought experiment is reflected in the case of Prof. Knut Nordby who was achromatic and studied colour vision and the psychology of perception among other things and he said he had no idea what colours were like despite access to extensive theory.
However some Thought Experiments are totally implausible and incoherent or so elaborate as to be totally confusing. Some of the premises are highly questionable.
A key "culprit" is the Twin earth experiment. Some people already think "Internalism" versus "Externalism" is a non issue which adds to its problems. But alot of the premises in the experiment are dubious such as someone on Twin Earth having identical mental states despite the human body consisting of a lot of water (H20) which means Twin humans body would consist largley of XYZ making any identical states implausible.
Also I have mentioned a problem with counterfactuals so in this case, and others, the existence of these planets (or aliens) is highly unlikely or non existent. Jesse Prinz is someone I believe who created a bizzare tentacled alien to make an arument for radical externalism.
However some Thought Experiments are totally implausible and incoherent or so elaborate as to be totally confusing. Some of the premises are highly questionable.
A key "culprit" is the Twin earth experiment. Some people already think "Internalism" versus "Externalism" is a non issue which adds to its problems. But alot of the premises in the experiment are dubious such as someone on Twin Earth having identical mental states despite the human body consisting of a lot of water (H20) which means Twin humans body would consist largley of XYZ making any identical states implausible.
Also I have mentioned a problem with counterfactuals so in this case, and others, the existence of these planets (or aliens) is highly unlikely or non existent. Jesse Prinz is someone I believe who created a bizzare tentacled alien to make an arument for radical externalism.
Comments (40)
It seems that the existence of a twin Earth (real water, not XYZ) is a certainty given that there is only a finite set of states of matter in a finite space (a Hubble Sphere for instance), and that space is infinite. Max Tegmark computed an upper limit to the distance to the nearest such twin, but elimination of unnatural states (aluminum cube planets, humans with memory of a different sky than what they see) puts the actual distance much closer.
Finally, what requirement is there for the plausibility of a thought-experiment if the implausible part does not interfere with the point? If there was no Prof Knut Nordby, would the Mary's Room thought-experiment be less valid? Such thought experiments seem the only way to explore subjectivity.
The premise of the thought experiment misrepresents language anyway. When people refer to water they are never referring to H20, as was pointed out most water contains much more than H20 so the referent of water is more vague and can refer to impure water in seas, lakes, streams and cups etc. There is no reason for language ever to refer directly and infallibly to essences.
The theory of H20 can be usurped so in which case language would never refer to anything (see changing models of atoms). I think language only refers to mental states, such as the appearance in consciousness of water, and not essences.
The Mary's room argument is plausible because nothing in it is impossible even if the conclusions or implications of the thought experiment are considered invalid. And people are born achromatic so there has always been the possibility of asking them about colour concepts. It was just convenient that Nordby existed and was an expert in the field so that he could prove that all current knowledge of colour science and perception did not get him close to knowing what red is like.
It is suspicious if in order to support a position you have to create a wild thought experiment invoking nothing that is immediately possible. Why can't a thought experiment for "externalism" just use our planet and our senses? The Prinz tentacles one is even more wild and confusing and I think invoking physics is pretenious because even phycists are uncertain about how to interpret phenomena so it is unlikely a philosopher without a physics background could invoke physics ideas validly.
Also, since we don't know what conscious states are assuming they are brain states that are identical to biochemical configurations is begging the question.
Other examples like Prof Nordby are human experts on bat echo-location or other animal senses that we either don't have or have in enormously lower amounts, like a shark sense of smell or its ability to detect electric currents.
But while Mary's Room is interesting to imagine, it doesn't actually signify anything because in the end it just comes down to discussing what certain words mean - in this case 'know'. David Papineau has a nice resolution of it. He says that when Mary sees the tomato she doesn't acquire knowledge of a new fact but rather she has learned new skills, which are to remember what it was like to see something red, and to recognise when something is red. I think it was in a Philosophy Bites podcast that he explained this resolution.
I think the argument is problematic because it seems to rely on the idea that scientific theories usurp experience. But scientific descriptions are only models of what might be behind experince. He makes the same mistake as Putnam in exagerating the scope of language.
I think the argument succeeds however in showing that there are features of experience that cannot be modelled. It undermines reductionism.
It puzzles me how little importance people place on consciousness when it is our only access to reality.
People's position on these thought experiments seems to be ideological in that they are prone to dislike a thought experiment that is opposing their position.
What concerns me though is why extravagant thought experiments get so much mileage. Swampman is another ludicrous one. I don't know any other field that would except these type of thought experiments.
I ask people to imagine a box with a triangular and circular holes cut in the bottom. (Easy to make at home with a cardboard box) then you put round and triangular objects in the box, shake it and it sorts them into two piles of similar objects.
This is to show how a structure can look like it is performing an intelligent process (categorising for example) when it is clearly it has no cognitive states. A computer is just a far more elaborate version of this but with no need to invoke mental type states.
Now what would it benefit, to place this thought experiment in the Year 3000 and invoke aliens with two brains who have mastered time travel.
(Parfitt is one of the worse for these kind of excrutiatingly elaborate fantasies.)
"Suppose Davidson goes hiking in the swamp and is struck and killed by a lightning bolt. At the same time, nearby in the swamp another lightning bolt spontaneously rearranges a bunch of molecules such that, entirely by coincidence, they take on exactly the same form that Davidson's body had at the moment of his untimely death."
What about entropy and the second law of thermodynamics?.....I mean come on
A thermal fluctuation is not out of the realm of possibility, at least from the point of view of classical thermodynamics (recall Boltzmann's Brain, for example). An updated version of this hypothetical involves a black hole emitting a fully-formed brain/human, and that too is supposed to be physically possible.
The thought experiment already concedes that the situation is extremely improbable. But why should it matter? Those who hold that there is a sharp, objective fact of the matter with regard to whatever question is being considered here (personal identity, mind, qualia) must accept any challenge, no matter how implausible. Plausibility is a red herring.
I disagree.
I said something similar in my counterfactuals thread. The more implausible a scenario gets the further it is removed from reality and the original premises. It is one thing for all the molecules in a gas to go into the corner of a container but what Davidson proposes goes way beyond that because matter would have to get itself into states of improbability that are supposed to have taken a whole life time and billions of years to reach.
And there is also the impossibility of a mental state being reformed that was derived from personal experience. For example say my boss at work calls me an idiot and that creates a nuanced mental state in me, then that mental state is inextricably linked to that event and can't be identically copied just by recreating a brain state. It is not the equivalent of making a square template and copying it to create an almost identical square, because experiences are not identical to each other or don't have this simplistic "copyability" structure.
In the end it just seems unclear what this thought experiment is saying. LikeThe Twin earth experiment I think it fails to defend the identity-identical claim.
Things are only superficially identical even things that appear identical are in a different space and their atoms are unlikely to be identical. Identical twins can easily be told apart straight away by a dog using scent so there can be very easy routes to proving things that appear identical aren't. For something to be truly identical they would have to be atom by atom identical and in the same space and time.
This strikes me as incorrect, though, of course, this is all just plausible speculation at this point. Perhaps some day we'll have super 3D printers which can print out identical copies of persons and can test some of these ideas in philosophy of mind, but til then, we are stuck in our armchairs.
I see no reason for supposing that, if you hold some particular memory or mental state, that an exact physical duplicate of yourself would not also have that memory (or a pseudomemory, if you like, the content of which exactly matches the content of your real memory) or realize the same mental state which you are realizing at the moment that your body is scanned for the copying process (such a scan would presumably have to be really, really fine-grained...).
Your argument seems to slip a bit when you move from saying that mental states cannot be identically realized in non-numerically identical minds because experiences are not identical to each other. But this seems to be a non-sequitur: even granting that experiences can't be identical to each other, it doesn't follow that mental states can't be identical to one another. I think you need something more for your argument to go through.
To scan a person exactly would violate Heisenberg's uncertainty principle. Matter can be (and has been) teleported, but not scanned for copying. On the other, the copy probably doesn't need to be exact to the quantum level. Any arrangement of molecules in the same places would suffice, so the molecular 3D printer at sufficient granularity to make an indistinguishable copy is not a violation of physics.
That said, everybody takes their own answers into the swampman story, so the thought experiment doesn't really illustrate anything. Arkady: You assume an exact physical copy would have all the memories, as do I, but somebody that argues for memory being part of immaterial state would disagree. So the thought experiment fails to resolve the issue for which it was posited.
That doesn't mean there is no use for thought-experiments. Some eventually suggest empirical tests, and others simply don't apply to humans, but do apply to other beings. What is it like to be cloned? Humans have no memory of it, and I can't ask the things that are. Our society and law is based heavily on the assumption that it cannot be done, or at least cannot be remembered. So I rely on thought experiments to work out the alternate rules, until I find consistent answers.
I don't think we really need dogs to tell apart identical twins: with one exception, I've never known a pair of identical twins which I've had much trouble telling apart.
This is an interesting topic. In reading about Leibniz's identity of indiscernibles thesis, I recall coming across the thought experiment (be forewarned: it is rather implausible :D ) of spheres of identical dimensions (and every other feature) in a symmetrical universe. Let us assume the following account of logical identity: A and B are logically identical iff anything which can be predicated of A can be predicated of B and vice-versa.
Now, with the spheres in the symmetrical universe, there is nothing which can be predicated of one sphere (call it "A") which cannot be predicated of the other sphere (call it "B"), and vice-versa, and yet any putative observer would clearly (I think) see that there are 2 spheres. If there are 2 spheres, then A and B are not logically (i.e. numerically) identical, and yet that conclusion contradicts our starting premise which defines logical identity.
There is much literature on the identity of indiscernibiles, and I'm not suggesting that there is no possible way out of this thicket of confusion, but I use this example to point out that the notion of identity is far from straightforward. We should perhaps take care in speaking of what is "identical" (or not) to what.
Is this true, though? A person may possibly believe that mental states, while themselves immaterial, nevertheless supervene on physical states (or are otherwise emergent from them). In that case, the physical duplicate would still possess the same mental states.
Perhaps a sort of thoroughgoing substance dualist might deny that there is any connection between the mental and the physical, but I don't see how that view can be plausibly maintained once we accept some basic metaphysical assumptions (e.g. that there are material bodies) and scientific observations (e.g. that memories are neurologically encoded in the brain in some fashion, by long-term potentiation or whatever the exact mechanism is, and mental states at the very least correlate in some fashion with the physical state of one's brain).
I don't understand the point you are trying to make. First, are you saying that the Swampman scenario is nomologically impossible, or just less probable than something else? Either way, the remove from reality may or may not matter, depending on the argument being made; you cannot just make this criticism in general. When using this thought experiment to probe certain crisp metaphysical stances on issues such as personhood, being improbable doesn't disqualify the scenario, though it may make it easier to bite the bullet.
Quoting Andrew4Handel
And that is why you should not offer your criticism without understanding the context in which it was proposed. The thought experiment doesn't say anything on its own - rather, it is reactions to it that matter. Your own reaction, which by the way is not unlike that of Davidson, who came up with the gedanken, reveals something about your metaphysical commitments. People with different commitments react to it differently. The experiment serves to highlight these differences.
I hear ya.
A putative observer introduces a point of reference, with respect to which some predicates will differ, e.g. left/right.
Sure. I wasn't claiming that any and all people who reject that mental states are themselves physical or material can accept the supervenience thesis, I was just responding to this point of yours:
"You assume an exact physical copy would have all the memories, as do I, but somebody that argues for memory being part of immaterial state would disagree. So the thought experiment fails to resolve the issue for which it was posited."
My point is just that someone could hold that mental states are immaterial and still accept that a person's physical duplicate would hold identical mental states as that person. One doesn't preclude the other.
Furthermore, even if one accepts the existence of an afterlife or OOB (which I as yet see no evidence to accept), it doesn't necessarily follow that physical duplicates wouldn't also be mental duplicates. If one adheres to a type identity of mind, then yes, such a thesis is probably incompatible with believing that a given mental state (a particular memory, say) could be realized both by physical brains and by spiritual ectoplasm, or whatever souls are purported to be made of.
However, any theory of mind which allows for multiple realizability seems to be perfectly compatible with the notion that particular mental states tightly correlate with particular physical states (by supervenience, emergence, or whatever), and that said mental states could also be realized in another medium. Presumably, at least some proponents of strong AI are physicalists with regard to mental states, and yet don't maintain that mental states are inextricably bound to the 3 pounds of oatmeal in our heads: there are other types of stuff which can realize those mental states (including, perhaps, as I said, spiritual ectoplasm).
I'm not sure the observer is actually necessary, though. We could talk about what would be the case in such a universe, even if no one were around to observe it.
However, I don't see that just any observer necessarily breaks the symmetry. One could appeal to a perfectly symmetrical observer, for instance, perhaps one who is himself spherical, and situated equidistant from each sphere.
There are other reasons to deny identity of a duplicate - take Davidson's view of the Swampman, for instance. According to Davidson, who is an externalist, it is not enough for two creatures to be instantaneously identical: diachronic differences matter. A swampman may believe that he is Davidson. In fact, he cannot help believing that, since he is an exact physical duplicate of Davidson and his mental state supervenes on his physical state. And yet, unlike the late Davidson before him, Swampman's belief is false, because the truth-value of a belief is contingent on its causal history and Swampman's causal history has nothing in common with Davidson's.
I am not endorsing Davidson's view here - just pointing out how views on the same thought experiment can differ. And that is really the point of such thought experiments.
Right, assume a spherical [s]cow[/s] observer in vacuum :)
I think what this thought experiment shows is that Leibniz's construal of identity cannot work with a view from nowhere.
Mmm, spherical cow.
Yea, like I said, notions of "identity" may seem relatively straightforward at 30,000 feet, but tend to become rather muddled when examined closely.
Being able to remember and recognize red sounds like knowledge. We do use "know" to mean experiential in addition to propositional knowledge.
Remembering and recognizing facts as well as forming propositions requires abilities as well.
Imagine I said to someone "I look just Like my brother" and the person wanted to know what "Like" meant. I could then point towards two sheep and explain how sheep A looked like sheep B.
I wouldn't need to say "Imagine a planet near Earth but without helium in it and then imagine a human like creature with two brains who couldn't speak....."
At what stage do you need to invoke these wild things. You would think they would be a last resort
I think an argument ad absurdum is a different matter. You might say "Utilitarianism leads to the conclusion that we should destroy all life" because this intends to show the extreme consequences of a position at its extreme not to argue that such a scenaio is possible.
I think this may speak to some of the confusions surrounding the notion of "identity" which I pointed out. Clearly, logical identity is more stringent than personal identity. I believe that two things A and B can be said to be logically identical iff whatever can be predicated of A can be predicated of B and vice-versa.
However, this is clearly too stringent a criterion for personal identity, which seems to be the relevant notion with regard to Swampman-style thought experiments (as far as I can tell). Persons psychologically and physically change a great deal throughout their lives, and yet they remain the same person. In the time it took me to type this post, I no doubt shed a few thousand or so skin cells, and thus am not now logically identical to the person I was when I began typing it, and yet I retain the same personhood or sense of self.
Yes, it can work with that definition of knowledge, as well as with a more restrictive definition.
If we count ability to recognise and remember as knowledge, then Mary did not have complete knowledge of colour vision prior to leaving the room. So she acquires new knowledge when she sees the tomato.
If we do not count ability to recognise and remember as knowledge, then Mary did have complete knowledge of colour vision prior to leaving the room, and she does not acquire any new knowledge upon seeing the tomato, because the new thing is not knowledge.
Either way, Papineau's approach seems to resolve the dilemma.
I like the thought experiment because it makes us reflect harder on what we mean by 'knowledge' in a way that is different from the tired old debates about Justified True Belief.
That's what all the criticisms of physicalism qua consciousness come down to. Why is there (in the case of human brains at the very least) an experiential aspect?
I think what the Thought Experiment inadvertently shows and what Jackson didn't seem to intend is our primary reliance on consciousness. The thought experiment seems to want to ask how you can represent experience without having experience. I don't see how a description of anything could usurp experience. The Experiment has a similar problem to the Twin earth one in pushing the idea that language must some how refer directly to essences.
Perhaps it can be recast to ask that. But I don't think it will help us to do so. Very few 'Why' questions have answers.
It occurs to me that another perspective on Mary's Room is that it highlights the limits of language. What Mary can learn while in the room is limited to what can be conveyed by language. It demonstrates rather neatly that one cannot convey the experience of colour by language alone. A critic might say that it can be highlighted more economically by simply observing that the experience of colour, or vision more generally, cannot be conveyed to somebody that was blind from birth.
Edit: I see the other Andrew also posted while I was writing this, and has also homed in on the limits of language issue.
I am saying that the scenario is both implausible and impossible. The implausibility comes from the violation of laws of entropy by this incredibly implausible arrangement of matter that would require massive coincidences. The impossibility comes from the idea of two things being identical.
It seems that things can only be identical with themselves.
I don't see how I could draw any conclusions from a thought experiment when I couldn't imagine the proposed scenario. I think the Chinese room argument suffers from this also to a lesser extent because complex task required of the interlocutor.
But I think the China brain argument is plausible because it is simply showing how we wouldn't expect a country to be conscious whatever physical state it got into.
So yes, I suppose I am advocating total simplicity in thought experiments. (But even then there are endless quibbles)
Other quasi thought experiments like The Trolley problem I think suffer from lack of ecological validity. I think you should assess real life reactions (Like Nordby's) over intuitions.
Sure, but the issue is that it seems like language can describe most of the world in scientific terms, so the question is what makes minds unique? Particularly given how the bodies those minds are part of are understood in scientific terms. Far as anyone can tell, there's nothing unique about the brain or body that would make it an experiencer.
If we ask why water is a liquid within a certain temperature range and pressure, we know that's because of it's chemical properties. If we ask why some pattern of brain activity is conscious, we have no clear answer without biting some philosophical bullet or other the many people will find objectionable.
By this definition, we cannot then say that Mary knew everything there was to know about red before she left the room, so the problem is resolved either way without giving us any insights other than clarifying the language.
The causal connection is what is missing, according to Davidson and other externalists like him. Obviously, this won't matter to those who don't construe consciousness in terms of representations and their causal connections to the represented objects.
I don't get the point about the causal connection being unverifiable. If even your causal connections to your earlier selves are unverifiable, then I suppose nothing is, in which case this is just a truism.
I think science is explaining and theorising about the world not describing it. We don't need a description for what is already a vivid experience.
I think the problem for science is explaining consciousness. When you have reasonable scientific and causal explanations/theories then you can manipulate experiences or entities.
I think that if we had an explanation or causal theory for consciousness then a lot of other things would be enlightened such as in psychology, the nature of colour and qualia in general and some issues in physics. A lack of conciousness is why Knut Nordby and Mary can't see Red. So what does consciousness reveal or add to phenomena? If we had a dogs sense of smell would we develop new theories?
In some sense theory is a tool for expanding consciousness.