32 Comments »

  1. “The idea of certain physicists, that the ultimate constituent of the physical universe is information. That can only be a metaphor.”

    “The idea of something being intrinsically information is nonsense.”

    I totally agree with this. However, some of the physicists don’t seem to think of this as metaphor. Some computer scientists as AI people may actually think in terms of intrinsic information.

    For that matter, I’m puzzled as to what Dennett means by “information” (in “From Bacteria to Bach and Back”).

    Liked by 2 people

    • “However, some of the physicists don’t seem to think of this as metaphor.”

      “Some physicists” is not the mainstream, so I wouldn’t worry too much about them, if I were Dan.

      On a side note: it’s pretty mysterious that the universe, a machine that produces so much information that is wrong (just check the internet), gets it perfectly right when it comes to physics (if “some physicists” are to be believed).

      Liked by 1 person

    • If one changes “the ultimate constituent” to “an ultimate constituent”, I don’t think there can be any dispute.You can read a few pretty mainstream papers that relate to “intrinsicality” that I have linked here here, if you are interested (you might have to search for “20160111”). These references do not include anything about “it from qubit” type cosmology and high energy physics, where there is much interest in black hole informational paradoxes, the holographic principle and its implications for entanglement entropy, quantum information, how space, time and gravity emerge from quantum information, eg
      https://www.nature.com/articles/s42254-019-0087-6

      Liked by 2 people

      • …, I don’t think there can be any dispute.

        No dispute at all. There just isn’t any such thing as intrinsic information.

        So we disagree about that. But there’s no point in further arguing it.

        Like

    • That’s because it’s really not the same concept. We have a common language concept where, if you tell me something I didn’t know, you’ve given me some information, and you of course still know what it is you told me, so there’s nothing “conserved”, nothing physical at all in this account – and also not at all what physicists mean when they say that, eg, information is conserved.
      In physics and engineering there are two different kinds of information that to some extent share the same formalism (the same math), but have very different interpretations.
      There is Shannon information, named after American engineer Claude Shannon, who published his “A Mathematical Theory of Communication” in 1948. As the book title implies, Shannon information is about the physical constraints on communication, and as such we would in this account disregard whatever semantic content the information has, and restrict ourselves to counting bits. It is in this sense that computers are information processors. In the common language sense, information without content is nonsense.
      Shannon information, like common language information is not conserved, and is not fundamental to anything in physics, but very important in computer hardware design, electrical engineering, and a number of other subjects.
      And there’s Gibbs information, developed in the late nineteenth century by the American engineer, physicist, and, some would have, information philosopher, Josiah Willard Gibbs, working in the context of statistical mechanics; and Gibbs information is a conserved property of the material universe, up to issues concerning black hole evaporation through Hawking radiation, but concerning all issues regarding all the physics that will ever have any meaning and relevance at the human scale, Gibbs information is conserved, and a fundamental property of the world. That’s not a matter of opinion, but a well-corroborated fact. For example; if we add two numbers together, often we’ll want to forget what went into that sum, and as the hot messes we are – classical thermodynamic systems – that’s easy, we just forget, and that (Gibbs) information is ultimately lost as heat to the environment. The same goes for a classical computer, but in a quantum computer we do not have that option, and that means there are classical algorithms that quantum computers simply cannot run. And for another example, for any given volume of space, there’s a maximum number of bits it can contain before collapsing into a black hole.
      Some hard drives can be encrypted at the root level. What that means is that the data is scrambled on the drive, constituting an unreadable mess, and if you lose your password there’s nothing anyone can do. That data is lost for good. At the start of this year, this happened to me. Thankfully, I found a way to re-remember the password, but for a month or so, I just had 4 terabytes worth of sod all, instead of all the nicely weird music that it was supposed to be. In the common language sense, that drive had no information, in the Shannon sense, the information was there, but it was inaccessible, but in the Gibbs sense nothing had fundamentally changed. And it is only in this latter sense that information is conserved and is fundamental.

      Liked by 1 person

      • I agree that what physicists mean by “information” is different from what we mean in ordinary discussion.

        However, what physicists mean seems to require a physical substrate. So it is far from clear how that could be the fundamental basis for the physical substrate that it depends on.

        Like

        • In my view the concept of information, as it applies to physics, is no more or less “intrinsic” than any other physical concept like “mass”. It is a useful concept created by humans to get better control on the reality we live in. Whenever and wherever it works, great. (As you can see, I come from an anti-realist position).

          Liked by 1 person

        • Gibbs information does not require a substrate, or at least not any particular substrate, since it derives from a discourse in statistical mechanics that is concerned with the constraints on any substrate, given certain very general assumptions, where it denotes a kind of secondary substance, or a substance-like property of the dynamical system. Or, it is the most general, most abstract formulation of a conservation law. It is first and last a technical term in a very difficult subject, and it is perhaps unfortunate that it got labeled ‘information’ that sounds a lot like that other word ‘information’ that everyone knows what is.
          Gibbs information is a crucially important concept in thermodynamics, in information theory, in complexity theory, in the study of black hole dynamics, and it is baked into quantum mechanics.
          We may interpret it as saying that from a certain point on, it really doesn’t matter what the substrate is, or even what it is doing in any fine detail. Or we may decide that it means that substrates don’t matter at all, that the stuff is irrelevant at all times, and only the dynamic matters. Or maybe we will say that it talks about what matter can “know” about itself. All of these are just manners of speaking about a formalism that seems to capture really remarkably well what nature is actually doing at certain scales, specifically the very large and the very small, both very far removed from the human scale, where it may not necessarily be of all that much use.
          All inquiries are framed, and in physics all frames have a scale attached like they were a kind of semantic tensors, and the relevant vocabularies do not generally carry over between scales. Thus, what “emerges” in emergence is, as far as I am concerned, first and foremost the need for us to deploy different kinds and different concepts specific to that scale; and all I can say then, is that where Gibbs information is deployed, it is usually operationally meaningful, and does give a good account of something that is really there. And if you still have misgivings, you are more than welcome to try to reinvent quantum mechanics, but as for me, I’d give up right there, and let the physicists have their toy.

          Like

          • I beg to differ about the idea that physicist and computer scientist and mathematician concepts of information are somehow nothing to do with the “real semantic information” that some people talk about. Right from the beginning, when Maxwell introduced his Demon, the question was about, at the very least, practical reason. That is, the everyday kind of thinking that allows knowers to get something they want out of the world, ideally for nothing. Ditto the quantum physicists, with their Alice and Bob observing the world, manipulating the world, communicating with each other, and proving to each other that they know things – they want to model how scientists know anything about a world that has non-classical physical foundations. And all these further think that the work on computational complexity and formal languages by logicians and theoretical computer scientists is directly contiguous with the physical concepts of information, else why the quantum computer industry, and the quantum computing derived cosmological models. As to whether they believe “It” literally comes from “naked” Qubits, it is a formalism that is equivalent to multiple other more “physical” models, but it is very suspicious that there is such equivalence at all. That is, the fact that computation is possible constrains what kind of physical theories can be right eg “gravity sets the rules for optimal quantum computation in conformal field theories”.

            Liked by 1 person

          • Yes, but I don’t see what it is we disagree on. Unless you wish to claim that because there is a viable quantum computational program for cosmology, then it must also be the case that Dan Kaufmans concept of information is in some sense the ‘same as’ Leonard Susskinds.
            A human being consists of around 10^30 quantum states, or qubits, or elementary particles, call them what you want. Whether you talk about it or bit, those are just manners of speaking. But they are manners of speaking that allow you to talk meaningfully about something that is really there, and while there are frames where the above description of a human being is on the money, there are also frames where it is completely missing the point, and translating between frames is a tricky business. Coming from my neck of the woods, the next obvious thing to say is that that business is functorial in kind, and that it almost certainly involves forgetful functors. And I don’t doubt that very few readers of the Agora would understand what I just said. Maybe you would. And I could press on, and try to talk about ėtalé spaces and étalé cohomology, and very rapidly approach a level of abstraction where I’m not sure I understand what the hell is going on. But the wider point is that this is somewhat more complicated than a simple ‘continuous with’.
            I never claimed that common language information is in any sense ‘more real’, nor that it is not related somehow to either Shannon or Gibbs. But, as it turns out, there seems to be a lot of things that can be talked about really well deploying Gibbs information as a kind of fundamental, that have nothing to do with common language information, where it would be indeed preposterous to claim ‘it from bit’.
            They are different concepts used in different frames, even though what is really there stays the same throughout. They are different operationally, in that Gibbs is (thought to be) a conserved property of the world, while Shannon and Common Language are manifestly not conserved; Common Language demands semantic content, which the others do not; Shannon divides the world into signal and noise, Gibbs does not; and so on.
            They also differ in scope; Common Language is a feature of human understanding; Shannon, of communication channels and processing power; and Gibbs, of the world as is.
            And finally, they differ in the space of reasons, in that they are deployed in order to talk (and calculate) about those very different things. In the end, all I wanted to say is that when Dan says he is sceptical about information being a fundamental property of the material universe, that’s mainly because he doesn’t understand what Susskind et al. are talking about, because he and they are in different frames, using different concepts that just happen to have the same name. And information in that other context is not a metaphor (or, I don’t think that’s a helpful way of thinking about it), but it is a name for a property of a mathematical formalism, and a conception that relates it to observables

            Like

          • The concepts are different. They are deployed in different frames for different reasons, and with different intensions and assumed kinds. And translating between frames is far from trivial, and certainly not contained in a simple ‘continuous with’. Other than that I don’t see that we disagree on anything here.

            Like

          • “As to whether they believe “It” literally comes from “naked” Qubits, it is a formalism that is equivalent to multiple other more “physical” models, but it is very suspicious that there is such equivalence at all.”

            The equivalence is at least very interesting, although “suspicious” is perhaps too strong. But I think Jesper makes a very valid point when he mentions that care is needed when the word information is used. He writes that a given volume of space can only contain a maximum number of bits before collapsing into a black hole. As far as I know, there’s nothing in Shannon-type information theory that suggests such a thing. You need more than that, perhaps Brillouin’s formula, that says changing a bit of information requires an energy of at least kT ln(2).

            This lower bound on the energy needed to manipulate information is, as far as I know, not provided by information theory on it’s own. Physics, and physical parameters like k and T are needed to provide it, and physics doesn’t measure mathematical formulas, it measures properties of real substrates.

            You point out (as Jesper does) that the formalism of information theory pops up in physics in many places and does so in a certain sense regardless of the exact properties of the substrate. I personally hesitate to see something very deep in this equivalence.

            The Laplace distribution in statistics is used for speech recognition, image compression, to model extreme events “such as annual maximum one-day rainfalls and river discharges” (Wikipedia) and so on, although river discharges and speech recognition are entirely different things. Formulas from fluid dynamics can be used the model the flow of water and other liquids, but in certain circumstance also the flow of sand etc., although water and sand are quite different. It just happens that many-particle systems (or systems with many degrees of freedom) sometimes are studied with statistical methods that cloud the underlying physical processes. Sometimes the fact that behavior is random, is more important than the behavior itself.

            One last word on quantum information theory. As far as I know, it functions entirely within the mathematical framework of quantum theory (Hilbert spaces, linear operators etc.). Now, for me the question is: where did this framework come from? What decided it was the correct framework? “Information theory”? I don’t think so. It was the study of spectra etc., i.e. the study of real substrates.

            Liked by 1 person

          • Jasper and Couvent2104, you both make excellent points. The synthesis between “information is physical” and “it comes from qubit” hopefully will be informative ;).

            Both of you might find something of interest in the first couple of chapters of
            this thesis, which does the category theory thing to
            “’general probabilistic theories’ (GPTs),…general enough to accommodate essentially every physical theory, admitting probabilistic processes…[F]our information-theoretic axioms [that] are satisfied by quantum theory and more exotic alternatives…allow for an information-theoretic derivation of Landauer’s principle.”

            Liked by 2 people

  2. It seems that Dr. Rasmussen is really bothered by the puzzle posed by the so called “hard problem of consciousness”. I think the puzzle originates from a physicalist view of reality that takes physics as the complete and exhaustive description of the physical world. Since physics describes reality in terms of spatial and temporal position, energy, momentum, etc. it is logically impossible to derive qualia from that kind of description, no matter how sophisticated we make the description itself.

    In my opinion the solution of the puzzle is to admit that current physics only represents a partial description of physical reality (i.e. the part dealing with the physical properties mentioned above). Coming from a physicalist perspective, the fact that matter organizes in ways that give rise to consciousness is in front of our eyes. So we need to add properties to matter that are not described purely in terms of spatial and temporal position, energy, momentum, etc.

    Having said that, I do not believe that panpsychism is the only possible solution. (Besides sounding kind of ridiculous in its attribution of conscious capabilities to elementary particles, it suffers from the so called “Combination Problem”). I personally think that emergence of qualia from particular configurations of matter, taken as a brute fact (the same way that the existence of quantum fields can be taken as a brute fact), could be an acceptable solution. The details of when and how these configurations of matter would give rise to qualia would of course have to be developed in details by a science of consciousness.

    Like

    • I think the solution to the puzzle is that there is no puzzle (or at least, that this is a clueless approach to it), and that physicalism by and large consists of profoundly misunderstanding the nature of the enterprise of physics.
      It is not matter that has qualia, or is-consciousness, as Chalmers would have, for example; that is essentially a frame violation, or conflating kinds that belong to different frames. What does have qualia is living things, at least according to biosemiotics, and a quale then is no more or less than the experience of a sign. That, plus language, plus higher cognitive function, plus social interactions that lead to social contracts and the kind of long memory we call culture, give rise to consciousness in humans, who’ll then decide that from now on, we’re people.
      Dan’s objective in the prolegomena was to show how the scientific image cannot stand alone, and that a culturally pluralistic approach to making sense of the world is not only allowed, but necessary, as we will need the Sellarsian stereoscopic vision to succeed at this. Yet even if we were to take the stance that “physics is everything”, the objective for the inquiry is still to make operational sense of the observed phenomena; it is not to weigh in on whether this or that prescientific, common language sensible term names a kind that exists, in whatever sense that might be, or even trying to say what that something might be. The observable here is not consciousness, however useful that word might otherwise be; it is people doing things for reasons, and that very clearly includes physicists and philosophers, and that’s what we will need an account for, even if we were daft enough to try do it with physics.
      The English word ‘mind’ is so robustly meaningful that it seems like there must be a thing that it maps to, and there is of course a huge literature on that subject. But in my native Danish there simply is no word for it, and yet there isn’t a thing that I could say in English that I could not say in Danish, or vice versa, with the sole exception being certain sentences in philosophy of mind. For any English sentence that uses ‘mind’, it may map into a Danish sentence word that may in some other sentence map into English as temper, or psyche, or desire, or sense, or reason…
      The point is that words aren’t things, and concepts aren’t kinds; but whatever kinds are assumed, and how they are conceptualized, is particular to an inquiry and to a language, and they have no meaning outside of that frame.

      Liked by 3 people

      • Your answer brings me back to the relation between the manifest image and the scientific image, and how to properly apply those different levels of description in the context of consciousness.

        There are very practical issues related to consciousness. If I undergo surgery I want my consciousness to be “turned off” momentarily. That is currently achieve with a purely physicalist approach (even though it cannot be said to be a science: we do not have a clear understanding of what happens when we apply anesthesia).

        In some cases patients “wake up” during anesthesia, but they are not able to communicate their condition to the medical team. It would be great if we could detect that “from the outside”. In a similar fashion, it would be a great achievement if we could tell when somebody who is apparently in a state of unrecoverable coma is actually still conscious.

        There are obviously many attempts currently operating within the scientific image and trying to address the kind of problems I just mentioned. Would a manifest image approach help in this context? Would a stereoscopic vision add anything to the equation? If not, wouldn’t that raise again Dan’s issue on the relevance of philosophy?

        Like

        • I don’t think a manifest image approach would help with something technical like that, yet you have already adopted such an approach by simply caring about these issues. You’re saying that people matter, and therefore the experiences of people matter also. And that’s just not something that can be established by any scientific experiment. But once you have that value down, that tells you something about the kinds of science you may want to do, i.e. you’ll do those kinds of science for those reasons.
          I don’t think it’s a matter of adopting or not adopting a manifest image approach. The manifest image is something you simply have as a person, just as, as a human being, you simply have a metabolism. Yet that does not preclude you from having a choice on what to eat. And likewise you may decide to try to be aware of your social self, what values you have, what attitudes you express, and so on, and seek to understand whether those are good values, good ways of interacting with other people, or even just good ways of treating yourself.
          It is clear that there is complex interplay between brain chemistry, and a social, personal self, but equally clear that the causal order runs both ways; that for example cognitive behavioral therapy can be a very effective long term treatment for various psychiatric disorders. This is entirely evidence based, and also something I can attest to from personal experience. But that requires a patient who is responsive and capable of common language communication.
          Someone once described Category Theory (which is a subject in either metamathematics or generalized abstract computer science) as “logic as if people matter”. Whether that’s true or not – I think it can be – I think that’s what we should strive for in all our disciplines. But the insight that people matter, that derives from ordinary lived life, it is accessible to anyone, regardless of background, and that’s manifest image through and through.
          But people don’t merely matter as lumps of quivering flesh, suffering or passionate. People also matter as people, that is, as authors of reason and purpose, and that is why the logic has to different, and maybe a part of why the sciences, and some of us who grapple with the sciences, with the inherently Aristotelian, or bivalent, logic that we inherit from mathematics, has such a hard time accommodating the stereoscopic vision in a coherent framework.

          Like

          • Thank you for your reply. Your last phrase, regarding the sciences having a hard time accommodating the stereoscopic vision, made me think about the psi-epistemic interpretations of quantum mechanics (as opposed to the psi-ontic interpretations) and made me wonder if they could be seen as an attempt to adopt that stereoscopic vision

            Like

          • I was thinking in terms of moving the focus of the experiment away from an objective perspective and toward a more subjective one, where what is important is the scientist and his/her subjective beliefs and expectations QBism in particular comes to my mind.

            Like

          • Giving up objectivity sounds to me like a terrible idea. There is a world out there, and we are trying our damnedest to invent manners of speaking about it, that will capture some of what is really there, in ways that have operational meaning.
            Sometimes, for sure, we also want to talk about our feelings, or discuss poetry, or traffic laws, or taxes; all of which are legitimate discourses. But conflating them with physics is precisely the wrong thing to do.
            As for Quantum Mechanics, there are many interpretations; most are harmless, few are useful. What we lack most of all in QM is a solid, dependable phenomenology (in physics, phenomenology is that subject that can tell us what an observation is. It has nothing to do with Husserl). In the absence of that, interpretation is a free-for-all, and it shows.

            Like

          • Subjectivity enters into physics as motivation. It comes from the space of reasons, and it is for example how I care a great deal about understanding nature at the smallest scale, regardless of its relevance at the human scale; and Dan just doesn’t care at all about that. Which is fine, and about on a par with how he cares about tennis, and I just don’t.

            Like

          • Would you then be more inclined toward one of the QM objective-collapse theories, for instance the Ghirardi-Rimini-Weber model?

            Like

          • Look, I can’t be bothered to keep track of that zoo of interpretations. For me it’s pragmatic Copenhagen, at least until somebody says something that has some operational meaning. But the point is that it doesn’t matter. You can go with whatever you like, whatever gets you up in the morning, thinking ‘this is worth doing’.

            Like

  3. I’d like to poke some more at Dan’s idea of a socio-cultural basis for personhood. Earlier, I asked about the guy in solitary confinement who goes on being a person to himself, knowing things, perhaps writing vast documents that he posts through the door for us to read, but never getting any feedback. The other questions are how big does a culture have to be for there to be persons? If it is a small community, surely we have a system that is small enough to be completely describable “from the outside”? And we have babies and small children, and some of us have cats and dogs who are part of the community, and have a personhood in a purely “external” social sense – we don’t expect the same things from every person. And many other animals have societies that we recognize, and meditate on the fact that our “space of reasons” based cultures have the same features. Do animals act for reasons? Audi says there are “rational societies, rational plans, rational views, rational reactions, and rational emotions”. Applying this to animal behaviour leads me to think we can talk about “rational” social arrangements even though the actors are not high-powered rational individuals – back to the example of bee language and group decision making about hive sites.

    Like

    • I’d like to poke some more at Dan’s idea of a socio-cultural basis for personhood.

      I’m inclined to say that there are different notions of personhood. There can be a socio-cultural basis. And when somebody is treated as a non-person (shunned by society, for example), that has to do with the social/cultural notion of personhood. But there is also a cognitive basis for personhood, which is what matters to a solitary person or a person who is not much into socializing.

      Like

  4. Dan,

    I’m really enjoying your series of posts and videos, it’s given me a lot to think about.

    I find the analogy of a stereoscopic view of manifest and scientific images interesting and productive, but I don’t think there’s a strong separation between the images. I guess I’m with Massimo when he says they’re more like on a continuum. I can see them as separate but I’m then not sure how they can serve as reasonably solid premises to build arguments against the things you mention like panpsychism.

    I think most extreme scientism views, like thinking thermostats or a clock spring’s intention, fail for simple reasons like considering as synonyms the meaning of the same terms used at different levels of explanation. But I don’t have a problem with the general idea of supervenience, I find the idea kind of natural and extends easily to cover externalism. I also don’t think it helps us much, if at all, to explain things at the psycho-social level (frame or perspective), fill in the holes between levels, or help us explain the relation between the levels we do know.

    You mention “I think there is one world … science endeavors to, the scientific image is, a picture of that world that attempts to adopt as perspective-less a perspective as …”

    A lot of science also refines a perspective, while holding it as fixed as possible, getting rid as much as possible of things like intentional biases (minus the choice of perspective and methodology). It seems to me science is (or isn’t) as much about a view from nowhere, as things at the level of philosophy for example.

    “there is one world … evolved to creatures capable of intentionality representation teleological ways of thinking it created a space that has it’s own ontology”

    If I’m following you correctly, you mean animals and early humans had a way of being, acting and sensing, and from there humans evolved together the particularities of human languages, more complex behaviors, and concepts like intentionality and representation. And this, please correct me if I’m wrong, is what you’re saying gave us the manifest image, and it distinguishes us from other animals who haven’t in a manner of speaking picked up the stereoscope.

    Like

  5. Dan,

    I’m only up to 45:00. Lots more to think about but it’s hard to formulate and say where I disagree, if I disagree and with what exactly, or if you’re going to find something relevant or you’ve already addressed a point I missed. Like I only just looked up hypostatic and I get now why you and other’s where talking religion. Finally, I don’t want to sound like I’m only disagreeing and I’m very sympathetic towards your wish to counteract “intellectually desperate theses” like Panpsychism or Illusionism.

    Just one thought on persons or people. They occupy space and have position, they’re brothers, lawyers, and friends, they’re also fragile, they bleed, and are mortal. Sometimes those kinds of descriptions overlap or come together and are more or less equally laden existentially.

    Like