Episode 68: David Chalmers Interview on the Scrutability of the World

On his book Constructing the World (2012).

How are all the various truths about the world related to each other? David Chalmers, famous for advocating a scientifically respectable form of brain-consciousness dualism, advocates a framework of scrutability: if one knew some set of base truths, then the rest would be knowable from them. What sort of base? Well, there may be many principles bases, and what's important for Chalmers is not the details of which is picked but of the scrutability framework as a whole. The base he discusses the most in the book is PQTI, for Physical, Qualia (mental), "That's all," and Indexical (like "I'm here now). Being able to derive the rest of reality from PQTI has implications, Chalmers thinks, for the philosophy of language, mind, and metaphysics.

Mark tries to draw Chalmers into speculating outside his areas of expertise. Dylan asks about physics, of course. Wes has technical issues and drops off half way through (listen for the "ping" as he texts Mark to that effect). Read more about the project and get the book.

End song: "What You Want" by New People, from Might Get It Right (2013).

Become a PEL Citizen to get bonus content, including a free e-copy of the album and a full discussion of Chalmers's book on consciousness. You can also participate there in Not School discussion groups Alternately, support us by making a donation.


  1. Ryan says

    For what it’s worth Dylan I found myself siding with you against the mighty Chalmers in that discussion toward the end there. The other case seems like an obvious concession to make, but an amusing one to me, because it is an absurdist liberal political demand that all perspectives necessarily be permitted to the table out of fair respect for all identities. Even those which are so marginalized that they can not hope to be brought into the discussion at all, let alone much more problematic surreal kinds of non-involved metaphysical entities. I guess that’s where Chalmers does betray his atheism though at times and delves into some sort of pantheistic phenomenological deism.

  2. says

    He’s saying that because from culture to culture, with their institutional facts which don’t align with one another, these facts must then be dependent upon some metaphysical/other realness which either is not revealed, or is not detectable. He makes the example of shortness and tallness, that being either 5 foot in z or 6 foot in z cannot themselves determine what is tall, and what is short. The glass is half-empty, half-full – the classic illustration which, rather than providing a method proving that the intricacy of the world is essentially unknowable, and is only open to interpretation, shows instead that from one person to another, perception and interpretation are not the only constants we should use in order to gain a holistic account of how things work. Science is, of course the current dominant theory, and Chalmers offers a kind of freedom-of-thought hypothesis – Let’s go with what feels right first, and work out the details later.

  3. says

    This was a fantastic episode, and I was very happy to hear Chalmers describe ways in which the project would be practically applied. (In some respects, his Construction of the World seems very much like a more systematized demonstration of how people would formulate their worldviews, anyway, which is not to speak ill of the project but to commend Chalmers’ efforts for trying to make use of concepts, categories, and the ‘scrutability hypothesis,’ in whatever form it might have to take, to show how a worldview could be constructed systematically.) I also now have an interest in buying the book and hopefully contributing to Chalmers becoming a wealthy man.

  4. Profile photo of Scott Anderson says

    Fantastic episode and congratulations to the team on getting access to such a big name! I listened to the episode twice, but I don’t believe that I heard much regarding the ‘I’ in PQTI, and I don’t see how the description’s example of “I’m here now” relates to what I would usually refer to as indexing, and the placement of ‘I’ after “That’s all” seems somewhat odd to me (you’d think that “That’s all” would be last), but perhaps this offers a clue that I failed to catch. A little help?

    • Profile photo of Mark Linsenmayer says

      That’s my #1 choice for a Not School reading in the philosophy of mind for February, so hopefully we’ll at least get a recording out of that for you. (or come join the group yourself and vote for it!) Who knows when PEL proper will return to this area… it’s not in our FY2013 Q1 plans, anyway. Thanks for listening!

  5. Frank Callo says

    Wow! This was SOOOOO much fun I actually listened to it twice to make sure I got as much as possible.

    There are so many points worth elaborating but I’ll pice only one after a brief aside;

    The Primitive Constituents of Reality would make a fantastic name for a band.

    OK, now for the serious thing.

    Toward the end Mark was talking to David about his (Davids”s ) ideas about AI. As I understand David, his contention is that a physical duplicate of me would have the same kinds of mental experience as I. This is the sort of thinking which leads many to believe that human consciousness is uploadable to some other kind of information processing system. I have always felt a strong objection to this idea but have never been able to articulate it beyond saying something like “artificial intelligence is like artificial fruit, ,you can make it look pretty good but you can’t eat it”. I admit that this might be more poerful as rhetoric than agrument. Never the less, I have always felt that there was more to this objection than “mere sentiment”.

    As I see it, you could train a robot to ACT just like me-that is to say, upload all of my biographical information and all the reactions I have to the world that are based in the nature of the sort of object my body is, as information that could be enacted as behavior, this wouldn’t be the same as BEING me.

    Taking the case of green as and example, you could design the robot to recognize green as a color and to associate the phenomenon of green with everything I associate it with for cultural and biological reasons, this information would not have the same emotional power. A robot can know that it is cold and act like it doesn’t like it in order to “fit in” with real people who are cold BUT it wouldn’t really CARE about being cold unless there were some reason that cold were detrimental to its normal function (which is why I don’t like to be too cold).

    In the case of a clone of me we might get closer. In that case you would still have all the physiological stuff in place but not the specific biographical stuff. Now suppose you could upload the biography to the “wet ware” as information? Then the person might act just like me but they would be doing so based on artifice.

    A lot of this bares on the continuity of all the facts about the universe. I actually DID spend the first year and a half of my elementary school years at a residential school for the blind. That fact is continuous with a lot of other things that led up to it that were simply of the unfolding of everything that ever happened. To program the brain of my clone to BELIEVE that he had the same experience does not correspond with all the vectors of contingency that led me to believe I had that experience. This is perhaps the fundamental difference between the idea of REAL and ARTIFICIAL.

    So I guess how the notion of the transferability of personality to different media cashes out in the broader context of your conversation with Chalmers as follows: I include something like “what really happened” in my scrutability code. This bares on a point that Dylan (I believe it was Dylan anyway) was making about whether it matters that you are in the matrix if there were no way of knowing you were. The whole drama of the film “the Matrix” centered on the idea that it is morally wrong to believe your life to be one thing when in fact it is another, expecially if this belief is placed in you by outside agencies. There is probably something to be said here about moral realism, at least as it applies to the real vs. the perceived conditions of one’s life.

    I’m not entirely sure of the status of “what really happened” in my personal scrutability code. It might be reducable to the disposition of some fundamental particles as they were propagated through space time. I am more than comfortable with property dualism here in saying that conscious states are not fully reducable to physical states, that they are simply a feature of physical states (kind of like the way three dots arranged so as to form a triangle doesn’t permit one to deduce the pattern of dots from anything about the dots themselves, pattern is just an embedded feature of reality). Truth requires an accurate account of the disposition of all the primitive constituents of reality. It might be that only the demon (or god) really knows what these are (we might all live in the matrix). But for me, scrutabliity must involve real (as opposed to artificial) representations of the disposition of the constituents. I simply take it as a matter of faith that there is is such a disposition even if we don’t know what it is.

    So, artificial intelligence, maybe, artificial consciousness no-not unless my digital doppleganger knows and says that it is in fact that and not the real me. Scrtability demands an accurate reckoning of all that has happened if it is not to founder on the possibility of artifice.

    I think this is one reason that most philosophy, when it gets around to ethics, emphasises the importance of honesty. Honesty is the enactment of reckoning in so far as one is capable of it. This might be a place where moral philosophy connects with deeper metaphysical issues. Moral behavior involves honesty with one’s self and others about what is actually the case. What is actually the case, in turn, is (or I argue should be) founded in the scrutability code. So to say what is the case, to the best of one’s ability to know, is to make information correspond with actuality. A thing is moral (has the quality of goodness) to the extent that it so corresponds. This is merely a minimal requirement as there is a lot more to what is good. But I think it is a necesary requirement that shows up in a lot of moral coding. For example, in Buddhist ethics m oral action is rooted in the dharma-it is unethical to cause harm to one’s self or others in the service of clinging to objects. This is because in Buddhist metaphysics objects are illusory. The impermanence of objects is implied in the Buddhist scrutability code and has a cash value in ethical action.

    All this to say that I believe that all traditional concerns of philosophy, including moral philosophy have a metaphysical base and the whole quesiton of scrutability might be employed profitably in making that case.


  6. Frank Callo says

    I’m sorry. Reading over what I wrote yesterday I see that it was WAY too long and scattered, kind of thinking rather than writing. More economical would been to have said.

    In my scrutibility code, pattern is one of the primitive constituents of reality. This is because pattern is not reducable (in any way I can see) to the components of the patern (whether these are dots arranged in a triangle, bicycle parts arranged into an actual bicycle, the neurons in the brain/CNS, the subatomic particles, etc.)

    Because of this, any AI which did not have as part of its “sense of self” the idea that it was constructed by an outside agent, and therefore discontinuous in its organization from the pattern that underlies the natural unfolding of the rest of the material universe, would loose scrutability for its part of what reality actually IS.

    Baring an “adequate idea” (to borrow a term from Spinoza) of what sort of thing it is, the AI would have to develope some mythological idea of its origen (like God made man of clay, etc.). To create such a being would violate what seems the fundamental basis of ethical action on our part.

  7. Profile photo of Frtiz Donaro says

    David Chalmers, quoted in a New York Magazine article about self-help books and “the self”:
    “But, in the spirit of being a better person, I should not be so hard on self-help. The fact is, selves are profoundly difficult to understand. “There is nothing that we know more intimately than conscious experience,” the contemporary philosopher David Chalmers observes, “but there is nothing that is harder to explain.””

    I always distrusted and had little respect for bullshitting, half-truth espousing, charlatan hucksters that make up the majority of writers/gurus of the tackiest, cheesiest genre of writing ever, namely “self-improvement” or “self-help”. It’s pure American horseshit, most of it. New York Magazine offers occasionally interesting pieces..solid middlebrow light reading that you needn’t be a Noo Yawker to get into. A long time ago I thought philosophy would make me a better person. I mostly ended up not knowing wtf a “better person” really means. Hah.

    • Stuart Parker says

      Self-help programs like Anthony Robins and helped me a lot. He specifically has a lot of programs that kind of re-program your brain and provide an easy, non-abstract way of providing motivation and understanding of how to succeed. They work if you are willing to follow through, most people aren’t and are lazy.

      Philosophy is will not make you a better person, (look at Peter Singer for God’s sake, he’d rather kill a newborn baby than a fully grown pig) it’s helpful at a basic level but one can quickly find oneself in an area where it’s more or less a word-puzzle game that considers emotions and actions in an abstract way, but has a hard time integrating the actual visceral emotive experiences that people have. I think this quote from Francis Bacon is applicable:

      “Philosophers make imaginary laws for imaginary commonwealths, and their discourses are as the stars which shed little light because they are so high”

  8. Frank Callo says

    Well, you have to have some sense of what a person is first. Took a class in existentialist philos in which that was the primary question. Like any philos class it did more to clarify the question than to give any answer.

    I have come to believe that “person” is a term we use to designate a recognizable yet unique pattern. This pattern is mostly “mental” or “psychological” supervening on a particular type of biological organism. There is also a social dimension as persons form part of a larger pattern called “society” or something like that.

    Most self help books seem to focus on how to become happy and effective. This can mean a lot of things depending on what one thinks happiness and effectiveness mean. What seems reasonably clear to me now (over a decade since that existentialist philos class) is that the person should, first and foremost, be constituted by a congruous pattern of ideas and projects in the world. I think this is where philosophy might (in an ideal world) help. Philosophy looks for patterns within distinct entities and the ways in which these entities, in turn, fit into a larger pattern.

    A pattern is is a funny thing because it is hard to say what one is without simply giving an example. But we all know, with greater or lesser degrees of clarity, when we are looking at one. Hopefully, when we look at what ever we conceive ourselves to be, we see a coherent pattern. If we don’t, this is where improvement takes place. “Know Thyself”. One could, and probably has to take a whole lifetime to follow this most basic advice. Se, self help can not help unless we come to some sense of what we mean by self.


Leave a Reply

Your email address will not be published. Required fields are marked *