2,657 thoughts on “Elon Musk Thinks Evolution is Bullshit.”
keiths:
Only their histories differ.
Alan:
As Petrushka points out, you have no way to know this.
That isn’t what petrushka was saying, and it isn’t true anyway. You can know the history if you witnessed it, of course.
The salient point is that you can’t infer Swampman’s history — or non-Swampman’s history — solely by examining their current physical state.
My argument, as a physicalist, is that the capabilities of both Swampman and non-Swampman — including their ability to understand English — are functions of their (identical) physical states.
Bruce is a physicalist, but he somehow thinks that the ability to understand English depends not just on their physical states, but also on their histories.
As I put it earlier:
It’s not like there’s some non-physical residue of causal history that attaches to one but not the other, somehow granting true understanding to the attachee.
Which comes first:mental or language norms?Good question.My intuition based on continuity of cognition in evolution is thatis Dennett is right (and not eg Brandom) thatnorms for mental representation are first, both explanatorily and metaphysically.
It just occurred to me there is a inconsistency in saying that norms for individual representation preceded those for language and at the same time saying that Swampman’s words will eventually acquire meaning after it interacts with a community.
If mental representation comes first, and the creature lacks the biological history to provide norms for its mechanism for mental representation, then that seems to be a problem for me.
Oh well, tomorrow is another day.
Also, perhaps we should say “Swamp-person”? Then again, lacking biological history, I suppose that the creature does not have a claim on being either a man or a woman.
Read my ETA and simply answer my questions about the meanings of ‘Keith’ ‘people’ and ‘hearts’. Thanks.
Alan Fox: How do I decide which is the real Captain Kirk?
To restate it differently: there is no way to distinguish between two protons or two hydrogen atoms. There is no way to establish their histories. Ditto for our Captain Kirks, though they might subsequently have different histories, following their materialisation. And this matters in any way that impinges on reality?
Alan,
But you said these entities are identical to the last molecule. How do I tell one from the other?
As I said, you could know the history if you witnessed it. If you saw the dead tree being struck by lightning, transforming it into Swampman, then you would know you were looking at Swampman, not non-Swampman.
But why do you care? They’re physically identical. That’s my point. Whether Swampman understands English is a question about his current capabilities, not his history. If non-Swampman is capable of understanding English, then in my view the physically identical Swampman understands English too. I’m a physicalist, after all.
Bruce disagrees, thinking that history is the decisive factor. In his view, Swampman doesn’t understand English, even though the physically identical non-Swampman does. Hence my comment:
If two physically identical beings are conversing identically in English and having the same subjective experiences, how does it make sense to say that one of them understands English and the other doesn’t?
It’s not like there’s some non-physical residue of causal history that attaches to one but not the other, somehow granting true understanding to the attachee.
These questions affect the questions of whether it makes sense for some people to have inverted spectra, whether something that’s not H2O could nevertheless be water, and whether we could really be brains in vats. There’s not much of practical value–unless questions about survival of death, reincarnation, transplanted brains and whether Star Trek scenarios make sense matter to you.
That’s the nature of philosophy, for good or ill.
I’ve never known what to make of Swampman or Twin Earth or any other thought-experiments. Generally speaking I find thought-experiments of limited value. They can help explicate our “intuitions,” but they can’t help us think about what “intuitions” we ought to have, if any at all.
BruceS: The first issue is that there are two types of norms involved: (1) norms for mental representation and (2) norms for meaning and language. For mental representation, causal history is needed for the biological explanations which I think are the best bet for explaining norms scientifically. (An alternative which also depends on history, I believe, is KN’s view from Okrent involving goals instead of function). As for norms of meaning, I’ve made it clear from the start that my argument depends on assuming content externality as per Putnam, eg, so again meaning depends on context and causal history in that context and hence initially Swampman can have no meaning. I see your twater example as making a similar point about how its subsequent interactions in a language community will set the meaning of its words.
I would say that there are teleological norms, some of which are cognitive terms (if cognition is a biological function). I would say that some cognitive norms are also semantic norms, if they are involved not just in information processing about features of the world but in more complex activities, such as inferring and referring. And I would say that some semantic norms are discursive norms if the norms are shared such that behaviors can be reinforced or sanctioned in order for more consistency across individuals.
And I think that one can say all that while still insisting that teleology itself bottoms out in terms of the complex dynamics of self-organizing (autopoeitic) systems are shaped by their causal histories, including (but of course not limited to) natural selection.
Which comes first: mental or language norms? Good question. My intuition based on continuity of cognition in evolution is that is Dennett is right (and not e.g. Brandom) that norms for mental representation are first, both explanatorily and metaphysically.
I happily agree that the norms of cognition have ontological priority over the norms of discourse, and that we do get a better understanding of the latter in light of the former. So I do not agree with Brandom that our understanding of the norms of discourse can be completely autonomous from cognitive science or evolutionary theory.
However, I think that we can get at a provisional and tentative understanding of discourse norms independent of scientific inquiry, just in the sense that we need to describe the phenomenon to be explained (the explanandum) independent of the competing explanations. (If the explanandum could not be specified independent of the competing explanations, then we couldn’t ever evaluate the explanations to see which one is best.) And that’s true even if the explanations also motivate some light revision to our description of the phenomenon being explained.
The ideal, then, would be a story that starts off with very minimal forms of cognition (perhaps representational, perhaps not even that) and show how naturalistic processes can take us, step-by-step, all the way to Brandomian or quasi-Brandomian discursive communities of rational & moral agents.
I’m quite painfully aware of my inability to tell that story with sufficient detail and precision. But every day more is discovered and published and refined — such that I am actually oddly optimistic about our ability to achieve that ideal, or a good approximation of it.
Just idly googling (too bleeding hot to do anything else at the moment), I came across “heterophenomenology” proposed by Dennett as an alternative to Cartesianism and a short article by Dennett:
He writes:
Rorty’s linguistic convention is a close kin (a heretofore unacknowledged ancestor) of the ploy I attribute to “heterophenomenologists” (1991): deliberately permitting the subject’s word to constitute the subject’s “heterophenomenological world,” creating by fiat a subjective or first-person perspective whose details then become the explicanda for a materialist, third-person theory of consciousness. I took the existence of a wide-spread belief in the primacy of the first-person point of view as given, and characterized heterophenomenology as the neutral method science could–and does–use to investigate the relations between the subjective and objective. Rorty’s papers suggest that the emergence of a first-person point of view is itself an effect of a similar burden-shifting move.
Bruce,
I think the core reason Keith and I (and KN in other threads) talk past one another on this issue is my concern with how we can justify norms using scientific language. As I read him, Keith goes with the intuition that Swampman is supposed to motivate: that functionally identical means identical with respect to norms.
No, that’s not what I’m saying. I’m saying that norms are irrelevant to physics, just as meanings are. When the two-bitser recognizes a quarter, it does so syntactically — via norm-insensitive physics. Ditto for Swampman recognizing a word. The syntactic engine is approximating an impossible semantic engine that is sensitive to meanings and norms.
After its [Swampman’s] creation, when it starts to interact when language users (your example), the situation changes. How soon one can say it understands a language is an open issue in my mind.
Actually, you told us this:
According to my intuitions, understanding a language requires
1. Causal interaction with some of the objects that language references.* If my memory of the movie of Helen Keller’s life is accurate, my point is illustrated by the scene where Helen “gets it” after the teaches dips her hand in water and signs “water” on her palm and then repeats that with other objects.
That’s why I was careful to set up my scenario this way:
Bruce:
On Swamp Man: at its instant* of creation, yes “Conversing” takes things beyond that instant and would meet my criteria.
keiths:
No, it doesn’t meet your criterion. Suppose Swamp Man materializes next to me and I immediately ask him to name the last four US presidents. He replies, “Bush, Clinton, Bush, Obama.”
My question is the first utterance he’s ever heard. He hasn’t causally interacted, either directly or indirectly, with those four men or with the US political system. By your criterion, he doesn’t understand the question, he doesn’t understand his own answer, and he doesn’t understand the English language.
walto: These questions affect the questions of whether it makes sense for some people to have inverted spectra, whether something that’s not H2O could nevertheless be water, and whether we could really be brains in vats. There’s not much of practical value–unless questions about survival of death, reincarnation, transplanted brains and whether Star Trek scenarios make sense matter to you.
Perhaps it’s the heat or perhaps I’m not very imaginative. In particular I’m irritated by fiction that does not pay close enough attention to reality – a criticism that I do not level at Asimov. On the other hand, Star Trek did almost invent the mobile phone.
That’s the nature of philosophy, for good or ill.
I’m sure I get a distorted impression from some exchanges in these threads.
keiths: But why do you care? They’re physically identical. That’s my point. Whether Swampman understands English is a question about his current capabilities, not his history. If non-Swampman is capable of understanding English, then in my view the physically identical Swampman understands English too. I’m a physicalist, after all.
On the narrow point that identical objects cannot be distinguished, I agree.
Bruce disagrees,
Well, I hope he speaks for himself.
I started reading Gibson’s Ecological Approach to Visual Perception last night. At the very start he points out that meanings are irrelevant to physics but not to ecology. I think that this is the right way to proceed. We don’t need to eliminate meanings, or deny that there are genuine semantic engines. We only need to point out that the theory of semantic engines is ecological and neuroscientific. The fact that “meaning” is not a term that appears in any of our theories of fundamental physics is just as true and also as irrelevant as the fact that “anticline” and “trophic level” do not appears in our fundamental physics, either.
BruceS: After its creation, when it starts to interact when language users (your example), the situation changes.
Like the identical Captains!
walto:
Read my ETA and simply answer my questions about the meanings of ‘Keith’ ‘people’ and ‘hearts’. Thanks.
Your ETA:
ETA: (if you like you can substitute ‘Swamp Walto’ for ‘Swamp Bruce’ so there will be no concern about ‘Keith recognizers’–whatever they might be.)
Your question, modified to take the ETA into account:
I move to Swampville and at the same moment, Swamp walto and I say “Hi Keith!” to your double. We don’t mean the same thing by “Keith,” do we?
You do mean the same thing. Or more carefully, you do sorta mean the same thing, using “sorta” in the Dennettian sense. You’re physically identical, after all.
Ditto for ‘people’ and ‘hearts’.
keiths: In his view, Swampman doesn’t understand English
Not all words in English are of the same type. Whether somebody ‘understands English’ depends on what the minimum standard is for ‘understanding English.’
Your question, modified to take the ETA into account:
You do mean the same thing.Or more carefully, you do sorta mean the same thing, using “sorta” in the Dennettian sense.You’re physically identical, after all.
Ditto for ‘people’ and ‘hearts’.
If i’m going to shoot one of the 2 Keiths, do you care which one it is?
KN,
We don’t need to eliminate meanings, or deny that there are genuine semantic engines. We only need to point out that the theory of semantic engines is ecological and neuroscientific.
A genuine semantic engine would be one whose operation was causally influenced by actual meanings, not mere syntax. How is that possible if the fundamental laws of physics are meaning-insensitive?
Or if meaning emerges from physics, as I believe you were suggesting earlier in the thread, then how does it exert any causal power that isn’t already accounted for by the physics?
walto,
If i’m going to shoot one of the 2 Keiths, do you care which one it is?
But my two Captain Kirks cease to be identical the moment they materialise. From that moment, they have different new sensations, experiences, and interactions.
Not all words in English are of the same type. Whether somebody ‘understands English’ depends on what the minimum standard is for ‘understanding English.’
If competence, not history, is the criterion, then Swampman understands English just as well as non-Swampman does. They’re identical, after all.
Alan,
But my two Captain Kirks cease to be identical the moment they materialise. From that moment, they have different new sensations, experiences, and interactions.
Yes, which is why I specified the conditions more carefully than you did:
At the moment of his creation, Swampman is physically identical to his non-swamp counterpart in every respect. Assuming you are not a dualist, would you agree that every behavior exhibited by Swampman is identical (modulo quantum indeterminism) to the behavior that would have been exhibited by non-Swampman under identical conditions?
As a physicalist, I think their behaviors would be identical. They start in the same physical state, their environmental “inputs” are identical, and so they will proceed through the same sequence of physical states (again, modulo quantum indeterminism). That means that not only will their behaviors be identical, but also their subjective experiences.
walto,
So why can’t my ‘keith’ refer to that one only?
It can. You can, for example, refer to me as ‘Keith’, and to Swamp Keith as ‘Swamp Keith’.
keiths: They start in the same physical state, their environmental “inputs” are identical, and so they will proceed through the same sequence of physical states (again, modulo quantum indeterminism). That means that not only will their behaviors be identical, but also their subjective experiences.
This is impossible. If two entities, capable of having experiences, occupy different spaces, as they would have to do, their experiences will not be identical. Ergo they will cease to be identical.
ETA careless spelling error!
Alan,
Pay attention to the bolded words:
At the moment of his creation, Swampman is physically identical to his non-swamp counterpart in every respect. Assuming you are not a dualist, would you agree that every behavior exhibited by Swampman is identical (modulo quantum indeterminism) to the behavior that would have been exhibited by non-Swampman under identical conditions?
It can.You can, for example, refer to me as ‘Keith’, and to Swamp Keith as ‘Swamp Keith’.
So you agree with me that when I say ‘keith’ and swamp walto says ‘keith’ we are referring to two different entities–me to the guy who you care about and hime to the guy who swamp keith cares about.
Different referents.
KN,
I’ve never known what to make of Swampman or Twin Earth or any other thought-experiments. Generally speaking I find thought-experiments of limited value. They can help explicate our “intuitions,” but they can’t help us think about what “intuitions” we ought to have, if any at all.
In this discussion the Swampman thought experiment has been invaluable, because it has exposed the fact that for Bruce, “understanding English” is not a capability, it’s a history.
That is, of two physically identical individuals with identical capabilities, Bruce thinks it is legitimate to say that one understands English and the other does not — solely because of their different histories.
Without the thought experiment it would have been harder to tease apart the contributions of history vs physical state, because every real person is unique in both respects.
walto,
So you agree with me that when I say ‘keith’ and swamp walto says ‘keith’ we are referring to two different entities–me to the guy who you care about and hime to the guy who swamp keith cares about.
No, I don’t agree. How did you get that idea?
keiths: A genuine semantic engine would be one whose operation was causally influenced by actual meanings, not mere syntax. How is that possible if the fundamental laws of physics are meaning-insensitive?
The principles of fundamental physics (e.g. quantum mechanics and genera relativity) are meaning-insensitive. Ecology is not.
“The world of physical reality does not consist of meaningful things. The world of ecological reality, as I have been trying to describe it, does. If what we perceived were the entities of physics and mathematics, meanings would have to be imposed on them. But if what we perceive are the entities of environmental science, their meanings can be discovered” (Gibson, Ecological Approach to Visual Perception, p. 28).
Or if meaning emerges from physics, as I believe you were suggesting earlier in the thread, then how does it exert any causal power that isn’t already accounted for by the physics?
The causal power of perceptual structures as meaningful for sentient animals can be “accounted for by the physics” only if we could explain ecology in terms of fundamental physics. That’s part of what “accounted for” means.
But we need to put metaphysical claims on the gold standard of epistemology, since we shouldn’t make claims about reality independent of specifying how we can know that reality has the feature we claim it to have. (More provocatively: metaphysics without epistemology is methodologically identical with theology, even if it is anti-theistic in content.)
On that basis, we would be entitled to assert that everything in ecology can be accounted for by physics only if we could show that ecology can be successfully reduced to physics. And that is a much stronger claim than merely stipulating that nothing in ecology can violate a principle of fundamental physics. And it has to be a real reduction, actually demonstrating how it can be carried out. Simply saying that ecology is reducible to physics “in principle” is just hand-waving.
I am not a physicalist because physicalism must be reductive if it is to have any content, and there’s no evidence that any of the sciences that don’t belong to fundamental physics can be reduced to fundamental physics — not in a contentful, non-hand-waving sense of “reduced to”, anyway.
I do I regard scientific techniques of knowledge production as having greater epistemic authority with regard to the actual world than non-scientific ways of knowing. For various historically contingent reasons, “naturalism” is the best term available for that view. (I would be OK with “weak scientism” as well.)
keiths: In this discussion the Swampman thought experiment has been invaluable, because it has exposed the fact that for Bruce, “understanding English” is not a capability, it’s a history.
That is, of two physically identical individuals with identical capabilities, Bruce thinks it is legitimate to say that one understands English and the other does not — solely because of their different histories.
Without the thought experiment it would have been harder to tease apart the contributions of history vs physical state, because every real person is unique in both respects.
I guess my “intuition” here — much I hate the term — is that Swampman does understand a language.
If I understand the thought-experiment correctly, the synaptic wiring that causally implements the Swampman’s semantic capabilities is identical to the synaptic wiring that actually resulted from the causal history of the non-Swampman’s interactions with its physical and social environments. The fact that the Swampman differs in its actual causal history doesn’t make a difference — not one that I can see, anyway.
I’m not sure if Bruce would deny that, if indeed he would.
I got it from when you said just acouple posts up that I could refer to you only with my use of ‘Keith’. Swamp walto would be referring to your double when he says ‘Keith’.
Note that it doesn’t follow from that alone that swamp walto is not speaking English.
In this discussion the Swampman thought experiment has been invaluable, because it has exposed the fact that for Bruce, “understanding English” is not a capability, it’s a history.
That is, of two physically identical individuals with identical capabilities, Bruce thinks it is legitimate to say that one understands English and the other does not — solely because of their different histories.
Without the thought experiment it would have been harder to tease apart the contributions of history vs physical state, because every real person is unique in both respects.
I take it Bruce would hold that understanding English is a capability that has resulted from a particular history. (As indicated earlier, in my own case, I’m not sure what the criteria for “understanding English” involves.) But the “solely” there implies only that the history is necessary, not that it is sufficient, as you suggest.
Re: the “teasing apart,” I’m afraid you haven’t quite managed that yourself yet.
Alan Fox:
Imagine thé Enterprise transporter malfunctions and beams two Captain Kirks. Who is thé imposter? One, both or neither? And we care because…?
Note that there are causal links in beaming people around, at least assuming it is a scientifically developed and tested technology and not (say) some scientifically-impossible plot device invented to save budget money on special effects for rocket landings or to speed up the narrative.
Whether they are the right causal links for norms I leave to the philosophy of the 23rd century.
. The fact that the Swampman differs in its actual causal history doesn’t make a difference — not one that I can see, anyway.
I’m not sure if Bruce would deny that, if indeed he would.
Yes, I do, although I won’t deny it is a pain intuitively.
The randomness means that is no causal history on which to hang (say) teleosemantic norms for concepts/mental representations, so you have the disjunctivism issue for them. For linguistic norms, there are no causal links to a linguistic community on which to base norms of meaning (I am assuming semantic externalism). So in effect you have void earth, so to speak: a twin which is molecularly the same as me but which has no context to give meaning to its verbal behavior until it joins a linguistic community.
On the teleosemanatic stuff: I’m not sure if biological history affects Okrent’s accounts. I understand his approach to replace selected-for function with the goals of a living organism. I think history still matters, since these goals seem to me to be tied to the organism’s niche in nature and hence its evolutionary history.
keiths: . When the two-bitser recognizes a quarter, it does so syntactically
“Recognition” requires norms. The two bitser syntactically outputs “it is a quarter”. Its designers then decide whether that output is right or wrong. Until they do, it is wrong to say it has recognized anything.
Perhaps this analogy will help: when or not the output from a program is correct depends the user requirements, not the code itself.
No, it doesn’t meet your criterion. Suppose Swamp Man materializes next to me and I immediately ask him to name the last four US presidents. He replies, “Bush, Clinton, Bush, Obama.”
My question is the first utterance he’s ever heard. He hasn’t causally interacted, either directly or indirectly, with those four men or with the US political system. By your criterion, he doesn’t understand the question, he doesn’t understand his own answer, and he doesn’t understand the English language.
Yes, exactly.
(Ignoring my proviso that understanding is not yes/no).
ETA:
Suppose the materialization was not a random event but instead the being had been beamed from twin earth. In that being’s actual world, those names happen to refer to different people even though they still became president. So its English would mean something different from ours due to a different causal history linked to a different linguistic community.
If different causal histories imply the possibility of different meanings, what can we conclude from no causal history? I say no meaning.
BruceS: The randomness means that is no causal history on which to hang (say) teleosemantic norms for concepts/mental representations, so you have the disjunctivism issue for them.
I’m inclined to disagree with that.
I take history to be an interpretation of the present. So perhaps the history is there.
I am not a physicalist because physicalism must be reductive if it is to have any content, and there’s no evidence that any of the sciences that don’t belong to fundamental physics can be reduced to fundamental physics — not in a contentful, non-hand-waving sense of “reduced to”, anyway.
ell.)
Interesting post, KN. I agree that intentionality could never be reduced to the language of physics.
In case you missed it, I wanted to draw you attention to this post I came across this blog post by a neursoscientist who was mentored by Gallagher. He discusses the same issues we have been discussing in various exchanges:
– how enactivism could fit with neuroscience
– the relation of enactivism and PP
– when DST is explanatory and when it is only predictive
I take history to be an interpretation of the present.So perhaps the history is there.
That brand of anti-realism about the past is a bridge too far for me. Way too far. Off the deep end, in fact.
BruceS: That brand of anti-realism about the past is a bridge too far for me.
Actually, I think I am being very realistic.
When I pick up a history book, I don’t travel backward in time to see what the historian actually wrote. Rather, I read what is in the history book at present.
When an historian studies ancient manuscripts, he does not time travel to see what was originally in the manuscripts. He goes by what is there at the present. I don’t see how history could be anything other than an interpretation of the present.
Neil Rickert: I don’t see how history could be anything other than an interpretation of the present.
….or how anything outside my visual field could be anything other than an interpretation of stuff inside my visual field!
walto: ….or how anything outside my visual field could be anything other than an interpretation of stuff inside my visual field!
Not the same thing at all. You can move around, which changes the visual field. When time machines are readily available, you’ll be able to do that with time, too. But until that unlikely event, …
You have it exactly backwards. The default (and provisional) position should be that the special sciences, including ecology, are reducible to physics. Why? Because despite searching for hundreds of years, no one has yet identified a single phenomenon that can be shown not to be reducible to physics.
Can you think of one? If not, then why on earth should irreducibility be the default assumption?
KN,
I guess my “intuition” here — much I hate the term — is that Swampman does understand a language.
If I understand the thought-experiment correctly, the synaptic wiring that causally implements the Swampman’s semantic capabilities is identical to the synaptic wiring that actually resulted from the causal history of the non-Swampman’s interactions with its physical and social environments.
So you agree with me that when I say ‘keith’ and swamp walto says ‘keith’ we are referring to two different entities–me to the guy who you care about and hime to the guy who swamp keith cares about.
keiths:
No, I don’t agree. How did you get that idea?
walto:
I got it from when you said just acouple posts up that I could refer to you only with my use of ‘Keith’. Swamp walto would be referring to your double when he says ‘Keith’.
That doesn’t follow. You seem to be assuming that if Swamp X refers to ‘Y’, he must mean Swamp Y. What’s the basis for that assumption?
It’s for me to decide how much effort to expend in trying to see any merit in “swampman” as a thought experiment.
I’ll rephrase my trivial point, and the only point that has so far caught my attention. You can’t have two identical Captain Kirks. From the moment of their creation they begin to have different experiences and are therefore not identical.
keiths, to KN:
In this discussion the Swampman thought experiment has been invaluable, because it has exposed the fact that for Bruce, “understanding English” is not a capability, it’s a history.
That is, of two physically identical individuals with identical capabilities, Bruce thinks it is legitimate to say that one understands English and the other does not — solely because of their different histories.
walto:
I take it Bruce would hold that understanding English is a capability that has resulted from a particular history. (As indicated earlier, in my own case, I’m not sure what the criteria for “understanding English” involves.) But the “solely” there implies only that the history is necessary, not that it is sufficient, as you suggest.
The word “solely” emphasizes the fact that since Swampman and non-Swampman are physically identical individuals with identical capabilities, it is only the difference in causal histories that justifies (in Bruce’s view) the claim that Swampman doesn’t understand English, while non-Swampman does.
In any case, the history is sufficient, because that particular history leads inexorably to a being that understands English. Once you fix the causal history, you fix the capabilities.
keiths:
Without the thought experiment it would have been harder to tease apart the contributions of history vs physical state, because every real person is unique in both respects.
walto:
Re: the “teasing apart,” I’m afraid you haven’t quite managed that yourself yet.
keiths:
Alan:
That isn’t what petrushka was saying, and it isn’t true anyway. You can know the history if you witnessed it, of course.
The salient point is that you can’t infer Swampman’s history — or non-Swampman’s history — solely by examining their current physical state.
My argument, as a physicalist, is that the capabilities of both Swampman and non-Swampman — including their ability to understand English — are functions of their (identical) physical states.
Bruce is a physicalist, but he somehow thinks that the ability to understand English depends not just on their physical states, but also on their histories.
As I put it earlier:
It just occurred to me there is a inconsistency in saying that norms for individual representation preceded those for language and at the same time saying that Swampman’s words will eventually acquire meaning after it interacts with a community.
If mental representation comes first, and the creature lacks the biological history to provide norms for its mechanism for mental representation, then that seems to be a problem for me.
Oh well, tomorrow is another day.
Also, perhaps we should say “Swamp-person”? Then again, lacking biological history, I suppose that the creature does not have a claim on being either a man or a woman.
walto,
Verbatim restatements are unhelpful.
I thought you were agreeing with Bruce that meaning is history-dependent, even when there is no difference in physical state.
If so, then I already responded to your point.
If not, then please restate your point with more clarity.
But you said these entities are identical to the last molecule. How do I tell one from the other? How do I decide which is the real Captain Kirk?
They certainly are!
keiths,
Read my ETA and simply answer my questions about the meanings of ‘Keith’ ‘people’ and ‘hearts’. Thanks.
To restate it differently: there is no way to distinguish between two protons or two hydrogen atoms. There is no way to establish their histories. Ditto for our Captain Kirks, though they might subsequently have different histories, following their materialisation. And this matters in any way that impinges on reality?
Alan,
As I said, you could know the history if you witnessed it. If you saw the dead tree being struck by lightning, transforming it into Swampman, then you would know you were looking at Swampman, not non-Swampman.
But why do you care? They’re physically identical. That’s my point. Whether Swampman understands English is a question about his current capabilities, not his history. If non-Swampman is capable of understanding English, then in my view the physically identical Swampman understands English too. I’m a physicalist, after all.
Bruce disagrees, thinking that history is the decisive factor. In his view, Swampman doesn’t understand English, even though the physically identical non-Swampman does. Hence my comment:
Alan Fox,
These questions affect the questions of whether it makes sense for some people to have inverted spectra, whether something that’s not H2O could nevertheless be water, and whether we could really be brains in vats. There’s not much of practical value–unless questions about survival of death, reincarnation, transplanted brains and whether Star Trek scenarios make sense matter to you.
That’s the nature of philosophy, for good or ill.
I’ve never known what to make of Swampman or Twin Earth or any other thought-experiments. Generally speaking I find thought-experiments of limited value. They can help explicate our “intuitions,” but they can’t help us think about what “intuitions” we ought to have, if any at all.
I would say that there are teleological norms, some of which are cognitive terms (if cognition is a biological function). I would say that some cognitive norms are also semantic norms, if they are involved not just in information processing about features of the world but in more complex activities, such as inferring and referring. And I would say that some semantic norms are discursive norms if the norms are shared such that behaviors can be reinforced or sanctioned in order for more consistency across individuals.
And I think that one can say all that while still insisting that teleology itself bottoms out in terms of the complex dynamics of self-organizing (autopoeitic) systems are shaped by their causal histories, including (but of course not limited to) natural selection.
I happily agree that the norms of cognition have ontological priority over the norms of discourse, and that we do get a better understanding of the latter in light of the former. So I do not agree with Brandom that our understanding of the norms of discourse can be completely autonomous from cognitive science or evolutionary theory.
However, I think that we can get at a provisional and tentative understanding of discourse norms independent of scientific inquiry, just in the sense that we need to describe the phenomenon to be explained (the explanandum) independent of the competing explanations. (If the explanandum could not be specified independent of the competing explanations, then we couldn’t ever evaluate the explanations to see which one is best.) And that’s true even if the explanations also motivate some light revision to our description of the phenomenon being explained.
The ideal, then, would be a story that starts off with very minimal forms of cognition (perhaps representational, perhaps not even that) and show how naturalistic processes can take us, step-by-step, all the way to Brandomian or quasi-Brandomian discursive communities of rational & moral agents.
I’m quite painfully aware of my inability to tell that story with sufficient detail and precision. But every day more is discovered and published and refined — such that I am actually oddly optimistic about our ability to achieve that ideal, or a good approximation of it.
Just idly googling (too bleeding hot to do anything else at the moment), I came across “heterophenomenology” proposed by Dennett as an alternative to Cartesianism and a short article by Dennett:
He writes:
Rorty’s linguistic convention is a close kin (a heretofore unacknowledged ancestor) of the ploy I attribute to “heterophenomenologists” (1991): deliberately permitting the subject’s word to constitute the subject’s “heterophenomenological world,” creating by fiat a subjective or first-person perspective whose details then become the explicanda for a materialist, third-person theory of consciousness. I took the existence of a wide-spread belief in the primacy of the first-person point of view as given, and characterized heterophenomenology as the neutral method science could–and does–use to investigate the relations between the subjective and objective. Rorty’s papers suggest that the emergence of a first-person point of view is itself an effect of a similar burden-shifting move.
Bruce,
No, that’s not what I’m saying. I’m saying that norms are irrelevant to physics, just as meanings are. When the two-bitser recognizes a quarter, it does so syntactically — via norm-insensitive physics. Ditto for Swampman recognizing a word. The syntactic engine is approximating an impossible semantic engine that is sensitive to meanings and norms.
Actually, you told us this:
That’s why I was careful to set up my scenario this way:
Bruce:
keiths:
Perhaps it’s the heat or perhaps I’m not very imaginative. In particular I’m irritated by fiction that does not pay close enough attention to reality – a criticism that I do not level at Asimov. On the other hand, Star Trek did almost invent the mobile phone.
I’m sure I get a distorted impression from some exchanges in these threads.
On the narrow point that identical objects cannot be distinguished, I agree.
Well, I hope he speaks for himself.
I started reading Gibson’s Ecological Approach to Visual Perception last night. At the very start he points out that meanings are irrelevant to physics but not to ecology. I think that this is the right way to proceed. We don’t need to eliminate meanings, or deny that there are genuine semantic engines. We only need to point out that the theory of semantic engines is ecological and neuroscientific. The fact that “meaning” is not a term that appears in any of our theories of fundamental physics is just as true and also as irrelevant as the fact that “anticline” and “trophic level” do not appears in our fundamental physics, either.
Like the identical Captains!
walto:
Your ETA:
Your question, modified to take the ETA into account:
You do mean the same thing. Or more carefully, you do sorta mean the same thing, using “sorta” in the Dennettian sense. You’re physically identical, after all.
Ditto for ‘people’ and ‘hearts’.
Not all words in English are of the same type. Whether somebody ‘understands English’ depends on what the minimum standard is for ‘understanding English.’
If i’m going to shoot one of the 2 Keiths, do you care which one it is?
KN,
A genuine semantic engine would be one whose operation was causally influenced by actual meanings, not mere syntax. How is that possible if the fundamental laws of physics are meaning-insensitive?
Or if meaning emerges from physics, as I believe you were suggesting earlier in the thread, then how does it exert any causal power that isn’t already accounted for by the physics?
walto,
Of course. Your point is?
But my two Captain Kirks cease to be identical the moment they materialise. From that moment, they have different new sensations, experiences, and interactions.
keiths,
So why can’t my ‘keith’ refer to that one only?
walto,
If competence, not history, is the criterion, then Swampman understands English just as well as non-Swampman does. They’re identical, after all.
Alan,
Yes, which is why I specified the conditions more carefully than you did:
walto,
It can. You can, for example, refer to me as ‘Keith’, and to Swamp Keith as ‘Swamp Keith’.
This is impossible. If two entities, capable of having experiences, occupy different spaces, as they would have to do, their experiences will not be identical. Ergo they will cease to be identical.
ETA careless spelling error!
Alan,
Pay attention to the bolded words:
So you agree with me that when I say ‘keith’ and swamp walto says ‘keith’ we are referring to two different entities–me to the guy who you care about and hime to the guy who swamp keith cares about.
Different referents.
KN,
In this discussion the Swampman thought experiment has been invaluable, because it has exposed the fact that for Bruce, “understanding English” is not a capability, it’s a history.
That is, of two physically identical individuals with identical capabilities, Bruce thinks it is legitimate to say that one understands English and the other does not — solely because of their different histories.
Without the thought experiment it would have been harder to tease apart the contributions of history vs physical state, because every real person is unique in both respects.
walto,
No, I don’t agree. How did you get that idea?
The principles of fundamental physics (e.g. quantum mechanics and genera relativity) are meaning-insensitive. Ecology is not.
“The world of physical reality does not consist of meaningful things. The world of ecological reality, as I have been trying to describe it, does. If what we perceived were the entities of physics and mathematics, meanings would have to be imposed on them. But if what we perceive are the entities of environmental science, their meanings can be discovered” (Gibson, Ecological Approach to Visual Perception, p. 28).
The causal power of perceptual structures as meaningful for sentient animals can be “accounted for by the physics” only if we could explain ecology in terms of fundamental physics. That’s part of what “accounted for” means.
But we need to put metaphysical claims on the gold standard of epistemology, since we shouldn’t make claims about reality independent of specifying how we can know that reality has the feature we claim it to have. (More provocatively: metaphysics without epistemology is methodologically identical with theology, even if it is anti-theistic in content.)
On that basis, we would be entitled to assert that everything in ecology can be accounted for by physics only if we could show that ecology can be successfully reduced to physics. And that is a much stronger claim than merely stipulating that nothing in ecology can violate a principle of fundamental physics. And it has to be a real reduction, actually demonstrating how it can be carried out. Simply saying that ecology is reducible to physics “in principle” is just hand-waving.
I am not a physicalist because physicalism must be reductive if it is to have any content, and there’s no evidence that any of the sciences that don’t belong to fundamental physics can be reduced to fundamental physics — not in a contentful, non-hand-waving sense of “reduced to”, anyway.
I do I regard scientific techniques of knowledge production as having greater epistemic authority with regard to the actual world than non-scientific ways of knowing. For various historically contingent reasons, “naturalism” is the best term available for that view. (I would be OK with “weak scientism” as well.)
I guess my “intuition” here — much I hate the term — is that Swampman does understand a language.
If I understand the thought-experiment correctly, the synaptic wiring that causally implements the Swampman’s semantic capabilities is identical to the synaptic wiring that actually resulted from the causal history of the non-Swampman’s interactions with its physical and social environments. The fact that the Swampman differs in its actual causal history doesn’t make a difference — not one that I can see, anyway.
I’m not sure if Bruce would deny that, if indeed he would.
I got it from when you said just acouple posts up that I could refer to you only with my use of ‘Keith’. Swamp walto would be referring to your double when he says ‘Keith’.
Note that it doesn’t follow from that alone that swamp walto is not speaking English.
I take it Bruce would hold that understanding English is a capability that has resulted from a particular history. (As indicated earlier, in my own case, I’m not sure what the criteria for “understanding English” involves.) But the “solely” there implies only that the history is necessary, not that it is sufficient, as you suggest.
Re: the “teasing apart,” I’m afraid you haven’t quite managed that yourself yet.
Note that there are causal links in beaming people around, at least assuming it is a scientifically developed and tested technology and not (say) some scientifically-impossible plot device invented to save budget money on special effects for rocket landings or to speed up the narrative.
Whether they are the right causal links for norms I leave to the philosophy of the 23rd century.
Yes, I do, although I won’t deny it is a pain intuitively.
The randomness means that is no causal history on which to hang (say) teleosemantic norms for concepts/mental representations, so you have the disjunctivism issue for them. For linguistic norms, there are no causal links to a linguistic community on which to base norms of meaning (I am assuming semantic externalism). So in effect you have void earth, so to speak: a twin which is molecularly the same as me but which has no context to give meaning to its verbal behavior until it joins a linguistic community.
On the teleosemanatic stuff: I’m not sure if biological history affects Okrent’s accounts. I understand his approach to replace selected-for function with the goals of a living organism. I think history still matters, since these goals seem to me to be tied to the organism’s niche in nature and hence its evolutionary history.
There’s a good SEP review with more details.
“Recognition” requires norms. The two bitser syntactically outputs “it is a quarter”. Its designers then decide whether that output is right or wrong. Until they do, it is wrong to say it has recognized anything.
Perhaps this analogy will help: when or not the output from a program is correct depends the user requirements, not the code itself.
Yes, exactly.
(Ignoring my proviso that understanding is not yes/no).
ETA:
Suppose the materialization was not a random event but instead the being had been beamed from twin earth. In that being’s actual world, those names happen to refer to different people even though they still became president. So its English would mean something different from ours due to a different causal history linked to a different linguistic community.
If different causal histories imply the possibility of different meanings, what can we conclude from no causal history? I say no meaning.
I’m inclined to disagree with that.
I take history to be an interpretation of the present. So perhaps the history is there.
Interesting post, KN. I agree that intentionality could never be reduced to the language of physics.
In case you missed it, I wanted to draw you attention to this post
I came across this blog post by a neursoscientist who was mentored by Gallagher. He discusses the same issues we have been discussing in various exchanges:
– how enactivism could fit with neuroscience
– the relation of enactivism and PP
– when DST is explanatory and when it is only predictive
Gallagher shows up in the comments.
enactive-bayesians-response-to-the-brain-as-an-enactive-system-by-gallagher-et-al
That brand of anti-realism about the past is a bridge too far for me. Way too far. Off the deep end, in fact.
Actually, I think I am being very realistic.
When I pick up a history book, I don’t travel backward in time to see what the historian actually wrote. Rather, I read what is in the history book at present.
When an historian studies ancient manuscripts, he does not time travel to see what was originally in the manuscripts. He goes by what is there at the present. I don’t see how history could be anything other than an interpretation of the present.
….or how anything outside my visual field could be anything other than an interpretation of stuff inside my visual field!
Not the same thing at all. You can move around, which changes the visual field. When time machines are readily available, you’ll be able to do that with time, too. But until that unlikely event, …
KN,
You have it exactly backwards. The default (and provisional) position should be that the special sciences, including ecology, are reducible to physics. Why? Because despite searching for hundreds of years, no one has yet identified a single phenomenon that can be shown not to be reducible to physics.
Can you think of one? If not, then why on earth should irreducibility be the default assumption?
KN,
That’s right. They’re physically identical, atom-for-atom.
Yet he does. It’s quite odd.
walto:
keiths:
walto:
That doesn’t follow. You seem to be assuming that if Swamp X refers to ‘Y’, he must mean Swamp Y. What’s the basis for that assumption?
There might not even be a Swamp Y.
It’s for me to decide how much effort to expend in trying to see any merit in “swampman” as a thought experiment.
I’ll rephrase my trivial point, and the only point that has so far caught my attention. You can’t have two identical Captain Kirks. From the moment of their creation they begin to have different experiences and are therefore not identical.
keiths, to KN:
walto:
The word “solely” emphasizes the fact that since Swampman and non-Swampman are physically identical individuals with identical capabilities, it is only the difference in causal histories that justifies (in Bruce’s view) the claim that Swampman doesn’t understand English, while non-Swampman does.
In any case, the history is sufficient, because that particular history leads inexorably to a being that understands English. Once you fix the causal history, you fix the capabilities.
keiths:
walto:
To the contrary, Bruce indicates that I got his position right.
The thought experiment did its job.
Alan,
If your attitude were better, I might be inclined to help you catch up with the rest of us. But it isn’t, so I’m not.