2,657 thoughts on “Elon Musk Thinks Evolution is Bullshit.

  1. keiths: Any knowledge claim based on the veridicality of our senses is illegitimate, because we can’t know that our senses are veridical.

    Just an aside. The etymology of “veridical” is a combination of “verus” or “veritas” (Latin – true, truth) and “dicere” (to say or speak). “Veridical” really only applies to speech. Regarding sensory information, “accurate” makes sense, though “reliable” gets across the idea that sensory information is mapping, not reality.

  2. BruceS:
    . . .
    ETA:Actually, I will add one more consideration:Physics requires that information on the past is always recoverable from the current physical state of the universe.(Hence the Black Hole Wars between Hawking and Susskind).So in principle the causal history is reflected in the current state of the universe.Perhaps that might help the most austere physicalists out there.
    . . . .

    I don’t think that’s entirely accurate. Some information on the past may be recoverable from the current state, but I don’t see that as a generally reliable principle. A particular state may be reachable from multiple pasts. There are many roads to Rome.

  3. petrushka:

    Since the current states are identical, there is no physical difference upon which to hang a difference in meaning.To salvage your theory, you would need to posit a non-physical “metaphysical tether”, subject to its own rules, that is dependent on the precise causal history but exists in the present.

    There really is no logical argument against Last Thursdayism.

    For the record (sorry if that gives anyone KairosFocus flashbacks), the words you quoted are from keiths not me, although I agree with them.*

    I also agree with your statement. I simply see no reason to say that one brain in a particular state understands English (for example) while an identical brain does not. I’ll keep following the discussion to see if someone comes up with a convincing argument for the alternative view.

    * I’ve seen this happen with nested quotations a couple of times. Tomorrow I should have time to determine if it’s a WordPress bug.

  4. Alan,

    Just an aside. The etymology of “veridical” is a combination of “verus” or “veritas” (Latin – true, truth) and “dicere” (to say or speak). “Veridical” really only applies to speech.

    Meaning is determined by usage, not etymology. We covered the meaning of ‘veridical’ earlier in the thread:

    KN:

    Personally, I think that it is deeply mistaken to ask, “are the senses veridical?” First, because it is never clear what “the senses” means here; second, because “veridical” is the wrong kind of concept to use in talking about perception.

    keiths:

    To the contrary, it’s exactly the right kind of concept to use. Check out definition #2 from Merriam-Webster and note the usage example (plus its source):

    Definition of veridical

    1
    : truthful, veracious <tried … to supply … a veridical background to the events and people portrayed — Laura Krey>

    2
    : not illusory : genuine <it is assumed that … perception is veridical — George Lakoff>
    [emphasis added]

  5. Alan Fox: I think the core problem for me and I think several others is what are the consequences of Keith’s point (for me there is also the additional “what is the point”). Keiths has said it is “interesting” and “important” but I’ve not seen him explain that. He seems to think it goes without saying.

    The question is:

    What are the consequences?

    Consequences? From philosophy? 😉

    (I kid, Kantian Naturalist, I kid.)

  6. Alan Fox: Just an aside. The etymology of “veridical” is a combination of “verus” or “veritas” (Latin – true, truth) and “dicere” (to say or speak). “Veridical” really only applies to speech. Regarding sensory information, “accurate” makes sense, though “reliable” gets across the idea that sensory information is mapping, not reality.

    Nice point. Saying that “the senses are veridical” vs. saying “the senses are reliable” pushes the metaphor in slightly different directions. There are indefinitely many ways of cashing out that in precise terms, on a continuum from the obviously absurd to the manifestly true. I tried pointing out that earlier but keiths didn’t seem to think it was worth responding to.

  7. KN,

    Saying that “the senses are veridical” vs. saying “the senses are reliable” pushes the metaphor in slightly different directions.

    First, it’s not a metaphor. We are really talking about whether the senses are veridical.

    Second, could you explain precisely how “the senses are veridical” differs from “the senses are reliable”, in your view, and how that difference is relevant to our discussion of Cartesian skepticism?

    There are indefinitely many ways of cashing out that in precise terms, on a continuum from the obviously absurd to the manifestly true. I tried pointing out that earlier but keiths didn’t seem to think it was worth responding to.

    Could you provide a link and a quote?

  8. keiths:

    I could be wrong about that, though, and maybe even knowledge of something as basic as “a thing is identical to itself” is impossible.If so, no crisis. I already attach an implicit “if my senses are veridical” to the word “know” in cases that require it, so I can easily do the same with “if my reasoning is correct”.

    KN:

    But that just means, “I’m right, unless I’m wrong.” That’s trivially true.

    “I know X, unless my senses are non-veridical or my reasoning is faulty” is not the same as “I’m right, unless I’m wrong.”

  9. KN,

    You have suggested that it [Cartesian skepticism] has inspired some entertaining works of fiction. You have not show that it has any consequences for our conduct or has any bearing at all on how we live our lives.

    Be careful when playing with sharp objects, KN. You might cut yourself.

    Let’s consider the book you wrote. Has it had “any consequences for our conduct” or “any bearing at all on how we live our lives”? Should we reject it as worthless on that basis?

    On another thread, Sal has been arguing that money spent on phylogenetic research is wasted if it doesn’t advance medical science. What you and Alan are saying here is equally stupid.

    Fortunately, the world is full of people who — unlike you, Alan, and Sal — actually value knowledge and ideas for their own sake, independent of their practical consequences. People who think it’s worth studying epistemology, or cosmology, or phylogenetics, even if their discoveries don’t have “any consequences for our conduct” or “any bearing at all on how we live our lives”.

  10. walto, to Bruce:

    I think he [keiths] also balls up numerical and qualitative identity.

    How so?

  11. Patrick,

    As an aside, my intuition, which could definitely turn out to be wrong, is that exactly identical physical states are not necessary. If consciousness can arise on non-biological substrates like silicon chips, it is the patterns, relationships, flows, etc. that are essential. Perhaps I read too much speculative fiction, though.

    That’s my intuition, too. I’d be quite surprised if consciousness turned out to depend on a biology-like substrate. We lack a satisfactory theory of consciousness, though, and it’s hard to back up that intuition without one.

    Far easier to justify the claim that the understanding of language is a capability, not a history. Otherwise you get absurdities like those I mentioned here:

    Yes, accepting the Turing test would be inconsistent. By your lights, Swampman doesn’t understand English even if he’s capable of passing the most demanding fluency test. Understanding English is not an ability, it’s a particular kind of history.

    Swamp Shakespeare can bowl you over with his plays, but he doesn’t understand English. Swamp Jeff Gordon can win at Talladega, but he doesn’t know how to drive a car, or even what a car is. A swamp surgeon can save your life, but he doesn’t know what a human body is or how to operate on one.

    To which one can only respond, “Seriously, Bruce?”

  12. Fixed.

    keiths: Shakespeare can bowl you over with his plays, but he doesn’t know that he’s ever written anything. Jeff Gordon can win at Talladega, but he doesn’t know whether what he’s in is actually a car or a fish. A surgeon can save your life, but he doesn’t know whether he’s ever even seen a human body, let alone operated on one.

  13. walto,

    To whom are you addressing the “fixed” version of my comment, and what are you trying to say?

  14. Bruce,

    ETA: Actually, I will add one more consideration: Physics requires that information on the past is always recoverable from the current physical state of the universe. (Hence the Black Hole Wars between Hawking and Susskind). So in principle the causal history is reflected in the current state of the universe. Perhaps that might help the most austere physicalists out there.

    ETA 2: I understand that one concern would be that if we cannot recover the information in practice, then it does not answer the issue. That is why I mentioned the relevance of the exchange with Neil.

    What you’re saying above does not fit with what you said earlier:

    I am not saying that meaning does not supervene on the physical (whatever that is). For Burge scenarios, it supervenes on the brain states of the linguistic community; for Putnam and twin earth scenarios, it supervenes on those brain states and on their physical context.

    You can’t make the Burge scheme work, because “the brain states of the linguistic community” doesn’t refer to the same thing as “the current physical state of the universe”.

    You can’t make the Putnam scheme work, either, because the “brain states and their physical context” also doesn’t refer to the same thing as “the current physical state of the universe”.

    Also, you are putting yourself in the extremely odd position of implying the following: Exactly what you and Twin Bruce mean by ‘Anaximander Rodriguez’ does not depend on your brain states, or on the brain states of the linguistic community, but on something else entirely! After all, Earth and Twin Earth are in identical physical states, so any difference in meaning must depend on something else — something outside.

    Furthermore, whatever that “something else” is, its causal influence (if any) on the Earth and Twin Earth systems must be exactly the same because their states remain synchronized. The true meaning of ‘Anaximander Rodriguez’ has no causal impact whatsoever on the people using the name!

    So in my scheme, meanings (which are really only “sorta” meanings) have causal power, while in yours they might as well not exist at all.

    Your theory just doesn’t work, Bruce.

  15. walto, to Patrick:

    As Bruce said, it’s precisely the same distinction we would make between a painting and a perfect forgery. If one defines “physicalism” in such a way that one can’t make that distinction, and then accepts physicalism because of religious preferences, one will lose a lot of money in the art market.

    Physicalism doesn’t preclude the ability to track things, walto. If someone watches the forgery being created and then tracks it carefully and continuously, he or she will always know which is the forgery and which is the original.

    Tracking isn’t essential, either. If we’re talking about the Mona Lisa versus a perfect forgery, for example, then the fact that security is so tight at the Louvre would be something that we would take into account. If we see one Mona Lisa outside the Louvre and another Mona Lisa inside the Louvre, it’s a safe bet that the one outside is a forgery. In other words, we take advantage of the physical state of the world, which includes the state of security at the Louvre, to decide which of the two identical paintings is a forgery.

    ETA: I’ll just add that what we count as a “physical property” makes the difference here. If “emerging from a swamp five minutes ago” is deemed a “relational physical property” the two items can’t be “physically identical.” So whether the difference “supervenes on the physical” is a matter of how we define our terms.

    Luckily, the laws of physics don’t “care” about such ostensibly physical properties. They only care about the current physical state of Swampman plus any interactions currently taking place between him and his environment.

  16. Patrick, to Bruce:

    I don’t think that’s entirely accurate. Some information on the past may be recoverable from the current state, but I don’t see that as a generally reliable principle. A particular state may be reachable from multiple pasts. There are many roads to Rome.

    Patrick,

    Bruce is correct if we are talking about the entire universe, while you are correct if we are talking about a subset, such as a planet or a linguistic community. 🙂

    The difference is that a planet interacts with its surroundings, while the universe has no surroundings with which to interact.

  17. Bruce,

    I’d still be interested in hearing your response to my question below.

    Bruce:

    I am taking a functionalist view of understanding as long as the causal relations captured in the functionalism have the right causal history. Blind luck won’t do. I realize that hurts one’s intuition. But I think it is needed to explain how understanding and meaning can work for real-life cases.

    keiths:

    Could you give a specific example of such a real-life case, and how a non-history-based approach fails to deal with it adequately?

  18. keiths:
    Patrick, to Bruce:

    I don’t think that’s entirely accurate. Some information on the past may be recoverable from the current state, but I don’t see that as a generally reliable principle. A particular state may be reachable from multiple pasts. There are many roads to Rome.

    Patrick,

    Bruce is correct if we are talking about the entire universe, while you are correct if we are talking about a subset, such as a planet or a linguistic community. 🙂

    The difference is that a planet interacts with its surroundings, while the universe has no surroundings with which to interact.

    Actually, I think it’s the other way around. The closer in time an event is to a particular state, the more constrained the possible number of states are. Unless I’m misunderstanding you.

  19. Patrick,

    Actually, I think it’s the other way around. The closer in time an event is to a particular state, the more constrained the possible number of states are. Unless I’m misunderstanding you.

    I think you’re misunderstanding me. I’m saying that the next state of a system is a function of the current state and any “inputs” from the environment.

    If the system is self-contained, like the universe as a whole, then there are no “inputs” because there is no environment with which to interact. In that case there is only one possible next state for any given current state, assuming the laws are deterministic.

    If we are talking about a subset of the universe, by contrast, then there are (or at least can be) inputs from the surroundings, and thus there is more than one possible next state for a given state.

    If the laws are both deterministic and time-reversible, then you can (in principle) take the state of the universe and run time backwards, recreating the sequence of states that led to the current state.

    You can’t do that with a smaller system that isn’t causally isolated, because it isn’t just the state that matters in that case — the inputs do, also. In such a system, a given current state can have more than one possible next state and more than one possible previous state.

  20. keiths:
    KN,

    First, it’s not a metaphor.We are really talking about whether the senses are veridical.

    Not “we” — you are talking about whether the senses are veridical. I’ve given my reasons several times for why I don’t think that’s even a helpful question.

    Second, could you explain precisely how “the senses are veridical” differs from “the senses are reliable”, in your view, and how that difference is relevant to our discussion of Cartesian skepticism?

    Well, “veridical” suggests that the function of the senses is to accurately represent the world in some way — think of the correspondence theory of truth transposed into sensory receptors. As Alan and I have both pointed out, the idea of the senses as “veridical” suggests a picture in which the senses have a language-like function, since it is propositions (thoughts, etc.) which can correspond (or fail to correspond) to states of affairs. By contrast, “reliable” invites the question, “reliable for what?” — to which one natural response is, “reliable for guiding actions that contribute to the satisfaction of the organism’s goals”.

    So one term invites a picture of the senses in which the senses have a language-like structure, and the other term does not invite that picture. That’s a substantive difference.

    Could you provide a link and a quote?

    From here:

    What does “perfectly veridical” mean here? Does it mean that sensory information always conveys to the brain a 1:1 map of microphysical structure of the objects causing that sensory information? (Obviously not — I’m being silly, to mark out one extreme sort of view.) Or does it mean that we can, under most conditions, reliably distinguish perceptions from hallucinations, illusions, and dreams? (That seems mere sanity, to me.)

    The claim I am defending is that our sensorimotor systems (not “the senses”) are generally reliable (not “perfectly veridical”), to the point that we can under most conditions, distinguish perceptions from hallucinations, illusions, and dreams. To me this is mere sanity, to be kept safe from the extravagances of unrestrained imagination.

    There is a world of difference between taking “the senses are veridical” to mean

    “the constituents of sensory consciousness are perfect representations of the world at all times” —

    which is manifestly false, since there are hallucinations, dreams, etc. — and taking “the senses are reliable” to mean

    “our sensorimotor systems are sufficiently stable that, under most conditions, we can distinguish perceptions from hallucinations, illusions, and dreams.”

    I understand a Cartesian skeptic about the senses to be someone who denies that we have the capacity to distinguish perceptions from hallucinations, illusions, and dreams. And I say that based on how Descartes runs the argument in the First Meditation. Descartes aims to refute precisely this skepticism by way of vindicating the reliability of the intellect, and then arguing that a reliable intellect gives us the capacity to distinguish perceptions from hallucinations, illusions, and dreams.

    Given this wide range of possible interpretations of “the sense are not veridical”, I don’t even know what you’re saying. If the claim is that we sometimes have perceptual errors, like hallucinations and illusions, that’s one thing. If the claim is that we can never reliably distinguish perceptions from hallucinations and illusions, that’s quite something else.

    keiths:

    Let’s consider the book you wrote. Has it had “any consequences for our conduct” or “any bearing at all on how we live our lives”? Should we reject it as worthless on that basis?

    In fact, if my position is right, it decisively undermines a core argument of anti-naturalism that is widely used by the Religious Right as to why Darwinism is dangerous. That’s not nothing. Even though my book does not address topics of immediate concern, it has implications for real-world issues.

    By contrast, I cannot see how your version of Cartesian skepticism has any implications at all. Nothing follows from it. It is inferentially sterile.

    Fortunately, the world is full of people who — unlike you, Alan, and Sal — actually value knowledge and ideas for their own sake, independent of their practical consequences. People who think it’s worth studying epistemology, or cosmology, or phylogenetics, even if their discoveries don’t have “any consequences for our conduct” or “any bearing at all on how we live our lives”.

    Actually your skepticism is incompatible with all knowledge. The difficulty is your inability to recognize this. You have said yourself that all knowledge based on the senses is illegitimate. If you think that it is consistent to say that science is both useful and illegitimate, that only goes to show how deeply confused you are.

    Even worse, the problem with your position is far worse than you want to realize. Here’s why.

    The Cartesian skeptic wants to run the following argument

    1. A cognitive capacity is reliable if and only if there is a deductively valid proof establishing that the representations produced by that capacity always correspond to the world as it really is.
    2. The reliability of a capacity cannot be grounded in the exercise of that capacity.
    3. There are two cognitive capacities: the senses and the intellect.

    From (2) and (3), she then concludes that

    4. The reliability of the senses cannot be grounded in the exercise of the senses.

    Now, the Cartesian — in contrast to Hume!! — wants to refute skepticism. She wants to say

    6. The almost-reliability of the senses can be grounded in the exercise of the intellect.*

    (* I say ‘almost reliability’ here because Descartes does not think that there is a deductively valid proof establishing that the representations produced the senses always correspond to the world as it really is, although he does think that there is a deductively valid proof establishing that the representations produced by the intellect always correspond to the world as it is really is. Descartes does think that the exercise of the reliable intellect allows us to distinguish perceptions from hallucinations, illusions, etc. — even though, he also maintains, the proper function of the senses is not to tell us what is true and false about the world but to tell us what is harmful and beneficial to the body (Sixth Meditation).)

    But in order to do that, she must first show

    5. The reliability of the intellect can be grounded in the exercise of the intellect.

    which is to say, she needs a way of getting around (2). And the way Descartes himself actually does so is by arguing, from the fact that we have an idea of God, that God must actually exist. The circularity in that proof has been well-noted since Arnauld.

    In contrast, Hume rightly observes

    5′ The reliability of the intellect cannot be grounded in the exercise of the intellect.

    hence

    6′ The reliability of the senses cannot be grounded in the exercise of the intellect.

    hence

    7′ Neither the senses nor the intellect can be shown to be reliable.

    This means that Cartesian skepticism in your sense — skepticism about a posteriori knowledge but not a priori knowledge — is not a stable position. Once the reliability of the senses is on the table, so too is the reliability of the intellect. And as Hume nicely shows, such skepticism is irrefutable. Knowledge is impossible.

    It is well-known that Hume’s way of dealing with this pessimistic conclusion is by pairing with with an ‘optimism’ about human nature. Roughly, he says that although philosophical reflection establishes the impossibility of knowledge, the non-rational side of human nature — our sentiments, customs, and habits — will always prevail over philosophical reflection, thereby making it impossible for us to live as skeptics.

  21. keiths:
    Patrick,

    I think you’re misunderstanding me. I’m saying that the next state of a system is a function of the current state and any “inputs” from the environment.

    If the system is self-contained, like the universe as a whole, then there are no “inputs” because there is no environment with which to interact.In that case there is only one possible next state for any given current state, assuming the laws are deterministic.

    Agreed, and I’ll leave aside quantum indeterminacy as adding nothing to this aspect of the discussion.

    If we are talking about a subset of the universe, by contrast, then there are (or at least can be) inputs from the surroundings, and thus there is more than one possible next state for a given state.

    Not if the larger system providing the inputs is deterministic.

    If the laws are both deterministic and time-reversible, then you can (in principle) take the state of the universe and run time backwards, recreating the sequence of states that led to the current state.

    You can’t do that with a smaller system that isn’t causally isolated, because it isn’t just the state that matters in that case — the inputs do, also. In such a system, a given current state can have more than one possible next state and more than one possible previous state.

    I see what you’re saying. I was thinking more in terms of our practical ability to determine the past from the present state. Without nearly infinite detail, the further back we go the more uncertainty there is.

    ETA: Thinking on it a bit more, it still seems like the same state could come from multiple sets of previous states and inputs. There is no reason to assume there is a unique history leading to a given state.

  22. Patrick:

    ETA: Thinking on it a bit more, it still seems like the same state could come from multiple sets of previous states and inputs.There is no reason to assume there is a unique history leading to a given state.

    Consider the time reversibility of physics equations.

  23. Patrick,

    Agreed, and I’ll leave aside quantum indeterminacy as adding nothing to this aspect of the discussion.

    Quantum indeterminacy only comes into play when a wavefunction collapses. The universe’s wavefunction never collapses, so indeterminacy isn’t an issue.

    keiths:

    If we are talking about a subset of the universe, by contrast, then there are (or at least can be) inputs from the surroundings, and thus there is more than one possible next state for a given state.

    Patrick:

    Not if the larger system providing the inputs is deterministic.

    Yes, even then, because the larger system may have its own inputs. To reduce the set of possible next states to one, you need to ascend to a level where there are no inputs. That’s why I wrote:

    Bruce is correct if we are talking about the entire universe, while you are correct if we are talking about a subset, such as a planet or a linguistic community. 🙂

    The difference is that a planet interacts with its surroundings, while the universe has no surroundings with which to interact.

    keiths:

    If the laws are both deterministic and time-reversible, then you can (in principle) take the state of the universe and run time backwards, recreating the sequence of states that led to the current state.

    You can’t do that with a smaller system that isn’t causally isolated, because it isn’t just the state that matters in that case — the inputs do, also. In such a system, a given current state can have more than one possible next state and more than one possible previous state.

    Patrick:

    I see what you’re saying. I was thinking more in terms of our practical ability to determine the past from the present state. Without nearly infinite detail, the further back we go the more uncertainty there is.

    That’s right, but it doesn’t really matter for Bruce’s purposes. His theory of meaning doesn’t require us to determine the past — it just requires the causal history to be implicit in the present, so that meaning has something physical to supervene upon. For the universe as a whole, the causal history is implicit in the present, according to current theory.

    Bruce’s problem is that in his theory, meaning supervenes on the wrong parts of the universe’s physical state. He wants it to supervene on the brain states of the “meaner” and his or her linguistic community, plus their physical context. But the only available state differences upon which to hang meaning exist outside the linguistic community and the causally relevant context. That’s what I was getting at in this comment.

    ETA: Thinking on it a bit more, it still seems like the same state could come from multiple sets of previous states and inputs. There is no reason to assume there is a unique history leading to a given state.

    Yes, if there are inputs. For a system having inputs, the current state does not fix either the past or the future. For a system with no inputs, such as the universe, the current state fixes both the past and the future (assuming the laws are deterministic and time-reversible).

  24. KN,

    Not “we” — you are talking about whether the senses are veridical.

    How soon you forget:

    KN:

    Thus far we have several different versions of the claim “the senses are not veridical”:

    1. The senses are not veridical because we sometimes have non-veridical sensory episodes, such as dreams, illusions, and hallucinations.

    2. The senses are not veridical because we sometimes have non-veridical sensory episodes, such as dreams, illusions, and hallucinations AND we cannot distinguish between veridical and non-veridical sensory episodes on the basis of sensory evidence (on pain of circularity).

    3. The senses are not veridical because we sometimes have non-veridical sensory episodes, such as dreams, illusions, and hallucinations AND we cannot distinguish between veridical and non-veridical sensory episodes at all.

    4. The senses are not veridical because we sometimes have non-veridical sensory episodes, such as dreams, illusions, and hallucinations AND although we can and do we distinguish between veridical and non-veridical sensory episodes at all, we can never be absolutely certain that we have done so correctly, because of conceivable scenarios in which none of our perceptions are veridical.

    If keiths is arguing for (4), I don’t see why that’s an interesting-enough claim to be worth holding. It looks trivially true and without consequence.

  25. KN,

    As Alan and I have both pointed out, the idea of the senses as “veridical” suggests a picture in which the senses have a language-like function, since it is propositions (thoughts, etc.) which can correspond (or fail to correspond) to states of affairs.

    Alan’s claim was that the meaning of ‘veridical’ is determined by its etymology, which is obviously wrong, as I explained here.

    Your claim is that perceptions can’t be veridical unless they are language-like, which is also obviously wrong. A fox can veridically perceive a rabbit runnning for cover without expressing that perception in language or proto-language.

    By contrast, “reliable” invites the question, “reliable for what?” — to which one natural response is, “reliable for guiding actions that contribute to the satisfaction of the organism’s goals”.

    You’re undermining your position. That’s actually an argument for using “veridical” instead of “reliable”! Cartesian skepticism is an epistemological stance, after all, so we are concerned about what is true, not what is useful.

  26. KN,

    There is a world of difference between taking “the senses are veridical” to mean

    “the constituents of sensory consciousness are perfect representations of the world at all times” —

    which is manifestly false, since there are hallucinations, dreams, etc. — and taking “the senses are reliable” to mean

    “our sensorimotor systems are sufficiently stable that, under most conditions, we can distinguish perceptions from hallucinations, illusions, and dreams.”

    There are two confusions here. One is that you are taking “veridical” to mean “perfectly veridical”. That’s a red herring — I have never asserted that perfect veridicality is at issue. It obviously isn’t.

    The second confusion is that you are implicitly assuming that a “stable” sensorimotor system will generally be right, when there is no justification for that assumption.

    You need to distinguish between two types of non-veridical perception:

    1) Non-veridical perception due to shortcomings or malfunctions in the perceptual apparatus; and

    2) Non-veridical perception due to the non-veridicality of the sensory information arriving at the perceptual apparatus.

    An optimally functioning perceptual apparatus can still be fooled if the sensory information it’s operating upon is non-veridical.

  27. keiths,

    Do you want to talk about how animals perceive their environments? Or do you want to talk about how brains manipulate information? I’ll happily discuss those issues with you and everyone else here, since I find them fascinating in their own right and with profound philosophical implications that I’m working through in my own research. Or do you want to argue entirely a priori about a posteriori knowledge? That seems decisively less interesting to me.

  28. KN:

    You have suggested that it [Cartesian skepticism] has inspired some entertaining works of fiction. You have not show that it has any consequences for our conduct or has any bearing at all on how we live our lives.

    keiths:

    Be careful when playing with sharp objects, KN. You might cut yourself.

    Let’s consider the book you wrote. Has it had “any consequences for our conduct” or “any bearing at all on how we live our lives”? Should we reject it as worthless on that basis?

    KN:

    In fact, if my position is right, it decisively undermines a core argument of anti-naturalism that is widely used by the Religious Right as to why Darwinism is dangerous. That’s not nothing.

    You’re forgetting your criteria for usefulness. You dismiss Cartesian skepticism on the grounds that it hasn’t had “any consequences for our conduct” or “any bearing at all on how we live our lives”. Neither has your book. By your own standards, your book can be dismissed as worthless, as can cosmology, phylogenetics, and much of pure mathematics.

    You, Alan, and Sal are taking a profoundly anti-intellectual position. Alan and Sal don’t know any better, but you certainly should.

    By contrast, I cannot see how your version of Cartesian skepticism has any implications at all. Nothing follows from it. It is inferentially sterile.

    For the third time, the implications of Cartesian skepticism are huge. It establishes that the world might not be anything like what we take it to be. That’s not just huge — it’s enormous.

    Actually your skepticism is incompatible with all knowledge. [Followed by a long argument for this assertion.]

    I’ve already addressed that:

    That might well be true. I already rule out absolute certainty, even of “self-evident” truths, on those grounds. The only reason I don’t rule out knowledge of them, as well, is that the apparent consistency of our reasoning makes it likely to be largely correct, in my estimation. (Consistency has different implications for reasoning than it does for the senses. I can elaborate on that if needed.)

    I could be wrong about that, though, and maybe even knowledge of something as basic as “a thing is identical to itself” is impossible. If so, no crisis. I already attach an implicit “if my senses are veridical” to the word “know” in cases that require it, so I can easily do the same with “if my reasoning is correct”.

    KN:

    It is well-known that Hume’s way of dealing with this pessimistic conclusion is by pairing with with an ‘optimism’ about human nature. Roughly, he says that although philosophical reflection establishes the impossibility of knowledge, the non-rational side of human nature — our sentiments, customs, and habits — will always prevail over philosophical reflection, thereby making it impossible for us to live as skeptics.

    It’s not impossible to live as a skeptic. Give it a try!

    You can even go on using the word “know”. Just remember that there’s an implicit asterisk next to it.

  29. KN,

    I wrote:

    Your claim is that perceptions can’t be veridical unless they are language-like, which is also obviously wrong. A fox can veridically perceive a rabbit runnning for cover without expressing that perception in language or proto-language.

    Do you have a counterargument to offer, or do you concede the point?

  30. keiths: Do you have a counterargument to offer, or do you concede the point?

    If you want to talk seriously about animal cognition and animal perception, then I do not concede the point.

  31. keiths:
    Do you have a counterargument to go along with your non-concession?

    That depends. Previously you claimed that all empirical science was off-limits since the veridicality of the senses was up for grabs. I would rather use empirical science to contest your conception of “the senses” and whether their “veridicality” is the best way of thinking about them.

    Briefly, you want to do epistemology in terms of Descartes, and I want to do epistemology in terms of Darwin.

  32. KN,

    That depends. Previously you claimed that all empirical science was off-limits since the veridicality of the senses was up for grabs

    Using the empirical sciences to argue for the veridicality of the senses is circular, given that the empirical sciences depend on the veridicality of the senses. You concede this, but still you want to make that (circular) argument. Why?

    Briefly, you want to do epistemology in terms of Descartes, and I want to do epistemology in terms of Darwin.

    More accurately, I want to do epistemology using valid, non-circular arguments, while you are okay with circularity and invalidity.

    You keep telling us that Bad Things Will Happen if we become Cartesian skeptics, and you apparently think that this danger justifies the embrace of an irrational, circular counterskeptical argument. I’ve shown you that those Bad Things won’t (and don’t) happen.

    So again, what is your reason for preferring an invalid argument to a valid one?

  33. keiths: Using the empirical sciences to argue for the veridicality of the senses is circular, given that the empirical sciences depend on the veridicality of the senses. You concede this, but still you want to make that (circular) argument. Why?

    First of all, I don’t believe it is as viciously circular as you believe. I think that the threat of circularity is based on a misbegotten search for foundationalism in epistemic justification. There is no “foundation” for knowledge, either empirical or rational. Nor does knowledge need any such foundation in order to count as genuine knowledge.

    Second of all, the circularity objection is just as serious for the intellect as it is for the senses — as I tried showing several times above. So if the circularity objection takes science off the board, then it also takes all reasoning off the board at the same time.

    Thirdly, the distinction you urge between knowledge and knowledge with an asterisk seems roughly analogous to a distinction I would urge between knowledge with a mythical foundation attached to it and genuine knowledge without the attempt at “grounding” it in anything else.

  34. KN,

    First of all, I don’t believe it is as viciously circular as you believe.

    The circularity is quite vicious:

    1. Assume that the senses are veridical.
    2. Develop a body of scientific conclusions based on that assumption.
    3. Use those conclusions to demonstrate that the senses are veridical.

    I think that the threat of circularity is based on a misbegotten search for foundationalism in epistemic justification. There is no “foundation” for knowledge, either empirical or rational. Nor does knowledge need any such foundation in order to count as genuine knowledge.

    Foundationalists would obviously disagree with you, but in any case the rejection of foundationalism does not entail an embrace of the claim that the senses are veridical.

    And why resort to a circular argument against Cartesian skepticism when a non-circular argument for it is available? Whence the skeptophobia?

    Second of all, the circularity objection is just as serious for the intellect as it is for the senses — as I tried showing several times above. So if the circularity objection takes science off the board, then it also takes all reasoning off the board at the same time.

    As I keep telling you, apparently in vain:

    That might well be true. I already rule out absolute certainty, even of “self-evident” truths, on those grounds. The only reason I don’t rule out knowledge of them, as well, is that the apparent consistency of our reasoning makes it likely to be largely correct, in my estimation. (Consistency has different implications for reasoning than it does for the senses. I can elaborate on that if needed.)

    I could be wrong about that, though, and maybe even knowledge of something as basic as “a thing is identical to itself” is impossible. If so, no crisis. I already attach an implicit “if my senses are veridical” to the word “know” in cases that require it, so I can easily do the same with “if my reasoning is correct”.

    KN:

    Thirdly, the distinction you urge between knowledge and knowledge with an asterisk seems roughly analogous to a distinction I would urge between knowledge with a mythical foundation attached to it and genuine knowledge without the attempt at “grounding” it in anything else.

    No, because the asterisk doesn’t amount to a “mythical foundation”. It’s exactly the opposite: an explicit acknowledgement that the knowledge claim in question depends on some assumptions that aren’t known to be true. To attach the implicit asterisk is simply to be honest with ourselves.

    If we can deploy a non-circular argument while being honest with ourselves, why should we try to fool ourselves with your blatantly circular argument?

  35. BruceS:

    ETA: Thinking on it a bit more, it still seems like the same state could come from multiple sets of previous states and inputs.There is no reason to assume there is a unique history leading to a given state.

    Consider the time reversibility of physics equations.

    I see the point you’re trying to make, but consider walking from Grand Central to Central Park in Manhattan. The streets in midtown are (mostly) a grid. As long as you take the correct number of lefts and rights, you’ll end up in the same spot as someone who followed a different route.

    Yes, you’ll each have a slightly different experience. If that’s essential to your argument, assume that you’re being transported blindfolded in a water tank that dampens the inertia enough so that you can’t distinguish between turns.

  36. keiths:

    ETA: Thinking on it a bit more, it still seems like the same state could come from multiple sets of previous states and inputs. There is no reason to assume there is a unique history leading to a given state.

    Yes, if there are inputs. For a system having inputs, the current state does not fix either the past or the future.For a system with no inputs, such as the universe, the current state fixes both the past and the future (assuming the laws are deterministic and time-reversible).

    I don’t think that follows. There can still be multiple paths to the same end state in a deterministic system.

  37. Patrick:

    I see the point you’re trying to make, but consider walking from Grand Central to Central Park in Manhattan.The streets in midtown are (mostly) a grid.As long as you take the correct number of lefts and rights, you’ll end up in the same spot as someone who followed a different route.

    Yes, you’ll each have a slightly different experience.If that’s essential to your argument, assume that you’re being transported blindfolded in a water tank that dampens the inertia enough so that you can’t distinguish between turns.

    That seems right to me.

  38. walto: That seems right to me.

    You have to express the situation at the lowest level of physics to get time reversibility and preservation of information.

    So for example, you express the state of a gas in terms of the exact energies, positions, and momentums of the molecules (some redundancy there). You cannot use macrostates, like temperature.

    As soon as you summarize to macrostates like temperature, you lose information, and so you cannot expect info to be conserved when you explain transitions at that higher level.

    That is not to say that such macrostate explanations are not better or more useful in some cases, only to say that they are not reversible because they do not preserve information.

    ETA: Because of the time reversibility of physics at the lowest level, you cannot get the same state resulting from different previous configurations (for if you went backwards, you would get indeterminancy).

    You can get cycles however if there are a finite number of states in the universe. Current theories of unlimited expansion may preclude that.

  39. Kantian Naturalist: If you want to talk seriously about animal cognition and animal perception, then I do not concede the point.

    If we take the intentional stance towards foxes, then they have beliefs, and beliefs can be taken to have language-like contents. For a representationalist, experiences can be understood as being structured like the corresponding belief, without committing to the exact nature of relationship between beliefs and experiences.

    So if we take the intentional stance towards animals, we can ascribe language contents to their experiences and beliefs at the level of explanation of that stance.

    But the real patterns which underly the effectiveness of this stance are explained by biology, neuroscience, and ethology. Even if we think these patterns involve representations which have content, that content need not have a language structure. The structure could be map-like. Or under PP, it would be a representation of probablities and errors.

    Within the context of one science, one might be able to rely only on local current state and that science’s causal explanations to predict next state. But that does not help with norms.

    To explain why one might call some language meaning correct but not others, some perception veridical but not others, or some action successful but not others, one needs norms. AFAIK the best ways for naturalizing the justification of those norms requires causal history.

    The nature of the norms and the associated causal history can differ depending on the level of explanation. We can explain everyday items like forgeries in terms of social norms (ETA: that is, social constructions). In ethology, action and perception norms might involve the evolutionary history of the organism and its resulting niche. At the neuroscience level, if we are using PP, those norms might be formulated in terms of minimization of some type of average statistical error where what constitutes an error for an organism would involve its evolution and niche.

  40. keiths:

    Yes, if there are inputs. For a system having inputs, the current state does not fix either the past or the future.For a system with no inputs, such as the universe, the current state fixes both the past and the future (assuming the laws are deterministic and time-reversible).

    Patrick:

    I don’t think that follows. There can still be multiple paths to the same end state in a deterministic system.

    Not if the laws are time-reversible.

    If there are two paths to the same end state, it means that the paths join somewhere. If you run time backwards, they fork at that same point, which means the time-reversed system isn’t deterministic.

  41. BruceS,

    I think you have to be a little careful there, Bruce. As i’ve mentioned before, there are various ways to define both physical properties and ‘supervenience’ on micro-particles. On some of them supervenience isn’t plausible. Second, as both you and keiths note, you require as a premise that time-backwards movement is deterministic.

  42. keiths: If you run time backwards, they fork at that same point, which means the time-reversed system isn’t deterministic.

    Time reversibility in theory is not reversibility in fact.

    Indeterminacy implies that we have no way of predicting events at the quantum level, and having backed a chain of events, we have no way of knowing if it will play the same way forward.

    Patrick’s map traversal metaphor at the level of quantum phenomena, amplified by complexity.

  43. Was it Neil who said that things that are identical are in fact, the same thing? I don’t know if there’s any theory in physics saying that, but it seems to be an unavoidable.

    Objects having exactly the same quantum state must necessarily have the same past and same future. Absent outside input.

  44. BruceS,

    One can take an intentional stance towards non-human animals, sure. But I think the limitations of doing so will eventually become apparent, because predictions that are licensed from within that stance will be thwarted on a regular basis. It is not a reliably efficacious strategy for coping with non-human animals. There are too many real patterns of non-human animal behavior that aren’t adequately described by the resources of the intentional stance, and the resulting anthropomorphism will lead to too many false positives and false negatives. There’s a reason why anthropomorphism is a constant danger in cognitive ethology. (Though it is also true that a dogmatic anti-anthropomorphism and knee-jerk behaviorism held back the development of animal cognition studies for many years.) In many ways the study of animal cognition is just now coming into its own (see Are We Smart Enough to Know How Smart Animals Are?, with many future surprises ahead.

    BruceS: To explain why one might call some language meaning correct but not others, some perception veridical but not others, or some action successful but not others, one needs norms. AFAIK the best ways for naturalizing the justification of those norms requires causal history.

    The nature of the norms and the associated causal history can differ depending on the level of explanation. We can explain everyday items like forgeries in terms of social norms. In ethology, action and perception norms might involve the evolutionary history of the organism and its resulting niche. At the neuroscience level, if we are using PP, those norms might be formulated in terms of minimization of some type of average statistical error where what constitutes an error for an organism would involve its evolution and niche.

    On the view I currently favor, there are basically two different kinds of norms relevant for determining linguistic meaning: the discursive norms of a linguistic community and the habitual norms of perceptuo-practical engagement, where the former are best understood as a distinctively hominid modification of the latter.

    (Hence my rejection of the Descartes-Kant-Lewis picture of “the senses” and “the intellect” as separate-but-equal cognitive powers.)

    We can explain those norms in terms of distal evolutionary history (as to how the norms came into being) and proximal cognitive mechanism (as to how the norms are causally implemented).

    I agree that PP is an attractive model for understanding the causal implementation of perceptuo-practical norms, and perhaps discursive norms as well, but I don’t want to make my naturalization of normativity depend too heavily on a theory that might turn out to be a fad.

  45. petrushka: Was it Neil who said that things that are identical are in fact, the same thing?

    I don’t think so. That is what “identical” means in mathematics, but not what it means in ordinary life.

    I do think that is the correct meaning of “identical” for use in logic. And since philosophers mostly try to limit themselves to logic, perhaps it is what they should use. Or, better still, they should broaden their horizons to include more than logic.

    In ordinary life, “identical” means about the same as “indistinguishable”. However what we can distinguish varies from person to person, so I don’t think “indistinguishable” works well with logic.

    I’m inclined to be suspicious of “swampman” and “twin earth” arguments (or “star trek transport malfunction” arguments). They can be useful to stimulate the intuition, but one should be careful about drawing conclusions from them.

Leave a Reply