Sean Caroll’s keynote talk at LogiCal

January 15, 2017 • 1:30 pm

Last night, Sean Carroll (a cosmologist at CalTech and the Official Website Physicist™), gave the keynote address at the LogiCal meetings. As he told me at dinner beforehand, he was trying to condense all five of his Gifford Lectures (given in Glasgow last year) into a single hour. It was, like his Gifford Lectures, a summary of his excellent book The Big Picture.

Given his task, he did a good job, laying out the reasons why we completely understand all the physics of everyday life (he is of course a physical determinist), explaining why dualism isn’t possible, but also noting that we can talk about things, like meaning and purpose and value, that are “emergent phenomena”, consistent with but not possible to explain in the language of particle physics.

The only part of his talk that baffled me, as it did in his book, is his explanation of why entropy seems to be a violation of the symmetrical laws of physics, since it increases with time, but why (or so I thought), the passage of time from past to future is more or less an illusion. He asked a very good question: “Why do we remember what happened yesterday but don’t remember what hasn’t happened yet?” His explanation—that we don’t fully grasp the Second Law of Thermodynamics—didn’t satisfy me, and I’m still seeking an answer. Any reader who can explain this to me is encouraged to do so below.

Sean has developed into a very dynamic and engrossing speaker, with a lot of humor, and it was a very good after-dinner talk, but one with a lot of brain food.


Sean presenting “The Core Theory”: the equation that completely explains the physics of everyday life. It’s in his book.


I had the pleasure of dining with Sean before the talk, and asked him the perennial question, “Why is there something instead of nothing?” His answer was that the question doesn’t make sense—yet. “Why questions”, he said, always come embedded in a larger framework—often, in my view, a religious one—but in this case, said Sean, it’s possible that that question, meaningless now, might someday find an answer if we learn more about the circumstances that produced our universe and that may be producing other universes.

Anyway, I urge you to read his book, The Big Picture, which is deservedly popular. Although I don’t agree with some of it, most notably his compatibilism on free will, by and large it’s a rewarding read, and accessible to all educated people.

This afternoon, Dan Barker, co-President of the Freedom from Religion Foundation, is giving his own take on the issue: “Free Will, a Beautiful and Useful Illusion.” I’ve read the draft manuscript for him on this issue, and, as you can imagine, we had long and strong disagreements. I’ll see from his talk whether I had any influence on his thinking. Given his responses in our email correspondence, I doubt it. I’ll report back.

78 thoughts on “Sean Caroll’s keynote talk at LogiCal

  1. One possible response (rather than an answer) to ‘why is there something rather than nothing’ is ‘what’s so special about nothing that you’d expect it to be the default setting of the universe?’

    Of all the possible states of the universe, nothingness seems to be a rather special state, possibly more unlikely than the seemingly ‘fine tuned’ universe we find ourselves in.

    1. There is no evidence that there’s such a state as ‘nothing’. Which is to say that I agree entirely with your comment.

      1. It’s worse than that. The idea that nonexistence is a state of affairs that could have existed seems incoherent on its face.

    2. +1 to this. There are an infinite number of ways to have something, but only one way to have nothing. If one were to simply “count the options” to determine odds, the odds would seem to be very heavily against nothing.

      1. It’s not clear that there’s even one way to have nothing or for there to be nothing, since that assumes that “having” and “being” are properties that nothingness can have.

  2. I’m looking forward to seeing the filmed version of the event (I see a camera in the background behind Sean). I watched all his Gifford Lectures.

    I’ve read Sean’s book and enjoyed it very much.

  3. I hesitate to differ from Sean Carroll on physics, but as I see it he gets into problems with the 2nd Law of Thermo owing to adopting a fully deterministic interpretation of quantum mechanics (Everettian QM). From there the only way to explain the 2nd Law is by the boundary condition, specifying a highly ordered starting point of the universe, which doesn’t seem to me that satisfactory.

    Alternatively, if one just puts some quantum indeterminacy (dice throwing) into the low-level QM, such as at the collapse of the wavefunction, then 2nd Law behaviour arises automatically whatever the boundary condition.

    Either way, the issue isn’t fully resolved.

      1. EQM is not probabilistic, it is fully deterministic.

        [And this is about the 2nd law and the “arrow of time”, not about “free will’.]

        1. EQM is deterministic. But that does not mean much to us, since it does not tell us on what wave function we are on when we “make a choice”. That is still probabilistic.

          Sorry, I thought it was about Sean’s book, too. Guess that means I can’t ask about a related question. Thanks for correcting me…

          1. But that does not mean much to us, since it does not tell us on what wave function we are on when we “make a choice”.

            That is one of the problems with Everettian QM! Though their answer would be “there’s a you on all of the branches of the wavefunction”.

            That is still probabilistic.

            Well not really. The Everettians would declare that all of them are occurring! Thus there is no probabilistic element.

          2. It’s probabilistic in the sense that the Born Rule applies to observations made by any given observer.

            This is not at odds with determinism. The algorithm by which the phone company assigns phone numbers to new subscribers may be fully deterministic, but from the point of view of any given subscriber, the number they’re assigned is effectively a random string of digits.

    1. QM seems to be asymmetric in that it is time-directed.

      A wavefunction evolves gradually (according to the Schrodinger equations) but collapses instantaneously even for entangled objects (e.g. paired electrons) where the transition of information (e.g. direction of spin) would seamingly violate the speed of light.

      1. I don’t think that’s right. As I understand it (largely thanks to David Wallace’s The Emergent Multiverse), decoherence is the process by which the wave function appears to collapse, and propagates at subluminal speed through ordinary particle interactions. There is no spooky action at a distance. Entanglement effects are observable only where the light cones of the entangled particles overlap.

        1. As I understand it, if I test the spin of an electron right now and determine the spin is ‘up’ and my friend on Proxima Centauri tests its entangled partner at the same time he will find that his is ‘down’.

          It might be another four years before our timecones intersect and messages between us confirming the direction of spin can have been exchanged (assuming we both sent them now) but by that time my partner would already have known the direction of spin of his electron for four years.

          That suggests that whatever I do to my electron now determines the outcome of his experiment instantaneously, i.e. faster than light.

          If I recall correctly Einstein first used the phrase ‘spooky action at a distance’ as he had difficulty accepting this could be the case.

          1. What do you mean by “at the same time”? Einstein tells us there’s no such thing at such distances. Different observers will disagree about whose measurement came first. Did your measurement determine your friend’s result, or did his determine yours? There’s no objective fact of the matter about that, so the question must be ill-posed in some way.

            What the Schrödinger equation tells us is that when you do your measurement, your wave function forks into an “up” outcome and a “down” outcome. When your friend does his measurement, his wave function forks in a similar way. The effects of that forking propagate subluminally, and where the light cones overlap, those outcomes match up to produce the observed correlations.

            If you insist that one of those outcomes on each fork must have collapsed away before the light cones have time to intersect, then yes, you’re going to have a problem explaining how each apparatus knew which outcome to collapse so that the statistics would come out right four years later. To me, this is a powerful argument against the collapse hypothesis.

          2. I see this a bit differently. The wave function is a sum of terms, each representing a possible outcome of a measurement and measurer (in EQM). Of course, you must square it in order to get probabilities. When you measure, say, the spin of your electron, you pick out one of the members of that sum for your universe. Your buddy on Alpha Centauri, being in your universe and being included in the wave function you just chose, automatically is in the required state because of your choice of global wave function. No information is passed from your wave function to his because they are the same, again in EQM.

            I like the analogy where I have a blue marble and a red one. I put both into identical opaque cloth bags and give one to Ariadne and one to Bernard. Ariadne goes for a ride to, where else, Alpha Centauri. When Bernie then opens his bag, he knows what color marble Ariadne will find when she opens hers. It’s not complicated.

          3. I like the analogy where I have a blue marble and a red one. I put both into identical opaque cloth bags and give one to Ariadne and one to Bernard. Ariadne goes for a ride to, where else, Alpha Centauri. When Bernie then opens his bag, he knows what color marble Ariadne will find when she opens hers. It’s not complicated.

            I’ve seen this analogy before but with odd socks. It’s a ‘hidden variable’ explanation. It doesn’t work that way though:


          4. It’s a hidden variable because the outcome is decided when you hand out the marbles – you just don’t know who got which marble till they open the bags.

            Ariadne has always had either a red or blue marble and Bernard had the other; they just did not know which.

            The marble has never existed in a red and blue superposition.

            In QM the properties of the electrons is not determined till they are measured: how you decide to measure your electron will determine the outcome for your partner but this outcome is not determined till the measurement is taken. The electron’s spin is in a quantum superposition until the spin is measured.

          5. By ‘at the same time’ I mean anytime before a signal traveling at the speed of light can reach him letting him know in what direction my electron is spinning.

            I don’t but the Everett hypothesis. Inventing a whole new universe to ‘explain’ away the EPR paradox is about the least parsimonious explanation possible.

          6. And yet EQM is favored by many physicists today.

            You are right about Bell’s theorem, but I was not talking about that. As for my example, no, I did not make it up myself. I got it from Leonard Susskind. And it’s only an example. It does show that you do not need “spooky action at a distance” (non-locality), since everything was set up when I gave out the bags. I don’t see any hidden variables in there.

          7. Everett didn’t invent macroscopic superposition to solve EPR or for any other reason. It was already there in the Schrödinger equation from the beginning, and it takes additional theoretical machinery (the collapse postulate) to make it go away. So on parsimony grounds, EQM wins because it jettisons the superfluous extras and simply takes basic QM at face value, with no loss of predictive power. The fact that it also dissolves the EPR paradox is a bonus.

            Here’s Sean Carroll on why parsimony is a bad reason for rejecting EQM. For a more thorough treatment, see the Wallace book I referenced earlier.

      2. I don’t think its necessarily asymmetric with respect to time. IIRC erasing information about entangled particles can cause them to re-decohere. In practice that’s extremely rare, but so is spontaneous entropy decrease, so QM is still no different from standard thermodynamics in that the universe could theoretically ‘run the movie in reverse’ and everything we see would still be physically allowed.

    2. We are stuck with the low entropy boundary condition, i.e. the Big Bang, either way, so I don’t see how this is a mark against deterministic QM.

      This is a good moment to comment on, as Jerry puts it, “why entropy seems to be a violation of the symmetrical laws of physics, since it increases with time.” Just as you say, Coel: the low entropy Big Bang is a boundary condition. Boundary conditions don’t contradict laws of physics, they supplement them. Indeed, you can’t derive any particular events from laws of physics unless you specify boundary conditions to reason from.

      Entropy increases with time: dS/dt >= 0. But that’s because we define the “forward direction” of time by reference to processes like memory that depend on the increase of entropy. Entropy increases with time because that’s how we define “forward” in time.

  4. I also loved Sean’s book. And I really like his multi-level way of looking at things.

    So I still don’t get something. Let’s suppose, as Jerry and even Sean say, that everything is determined by the laws of physics — I suppose the Core Theory. Still, when I have to decide, say, what car to buy, I am going to read reviews of cars, talk to friends, test drive various models and generally study the situation before I engage in what we call “making up my mind” — making a choice. (Sorry, I know “choice” is a bad word on this forum.) I do not know how to avoid seeing it that way. Do you? Well, then, isn’t this what Sean refers to in his book, talking about levels of understanding? On the car-buying level, I think I am making a choice. And I am, even if I have none. I can not get around that use of words. What else can I do?

    As for the 2nd law. Isn’t it just that the laws of physics are symmetric under space reversal and even time reversal? But entropy can’t increase in both directions, we know that.

    1. Actually entropy can increase in both directions. A universe in thermodynamic equilibrium is at a uniform temperature throughout except for occasional random fluctuations. Such fluctuations are time-symmetric, with a (relatively) low-entropy peak, and entropy increasing in both directions away from that peak.

      If Carroll’s view of time is right, then such a universe is essentially timeless except in the vicinity of a fluctuation, where the arrow of time points in opposite directions away from the central peak.

        1. According to Carroll, entropy increases as time goes forward because the universe began as a state of low entropy. Put differently, it is the state of low entropy which defines that which we call “the past.”

    2. Excuse me, the laws of physics are not symmetric under space reversal. Lee and Yang won a Nobel Prize back in the 1950s for showing this to be the case (weak interactions violate parity). In addition, if the laws of physics are invariant under time reversal, then the TCP Theorem is violated because they are not invariant under CP (Cronin and Fitch won a Nobel Prize in the 1960s for pointing this out).

    3. I do not know how to avoid seeing it that way. Do you?

      I don’t think there is any need to see it another way. That’s part of the point of talking about emergent phenomena and allowing for different types of explanations for different scales of the universe.

      Talking about your choice to buy a car rather than going into a long-winded discussion on the deterministic chemical reactions occurring in your brain is a useful shorthand. Its the same useful shorthand as talking about natural selection working on cheetahs, rather than the deterministic chemical reactions occurring in the brains and bodies of the cheetah, its environment, and its prey. The movement and actions of organisms is the wrong scale on which to use an atom-by-atom explanation. The atom-by-atom description of such things might still be more correct, but in general its not as useful as the organism-level approximation.

    4. Just started “The Big Picture”.

      Excellent, so far.

      People underestimate often how difficult it is to explain scientific and philosophical concepts in understandable language.

  5. My understanding of memory and the arrow of time, and I think Carroll’s as well, starts with the idea that memory itself is in some sense an illusion. When we “remember” the past, what’s really happening is that we infer it from the traces it leaves in our brains.

    This may seem like hair-splitting, but consider a scenario in which you come upon a broken egg lying on the sidewalk. Without having seen the egg fall, you can reasonably infer, from the evidence you see in the present, that there was an unbroken egg nearby in the recent past. You do not infer that there will soon be an unbroken egg in the near future, because that would violate the Second Law.

    In the same way, we interpret the state of our brain in the present as evidence of the influences that acted on it in the recent past, not as premonitions of the future, and it’s the Second Law that enables us to do this reliably. That process of inference is what we call memory.

  6. I too struggle to comprehend Dr. Carroll’s explanations about entropy and the arrow of time. Recommend reading Dr. Lee Smolin’s book ‘Time Reborn’ for another point of view. Still struggling though even after reading the above book as well as Dr. Carroll’s book ‘From Eternity to Here’.

  7. “Why is there something instead of nothing?”
    My answer: because if there was nothing we wouldn’t be here to ask the question.
    Have I missed something? Please say.

    1. Maybe you could read ‘A Universe from Nothing’ by Lawrence M. Krauss. It basically states that ‘nothing’ is unstable and will produce something eventually.

      1. Thanks Gary. I’ve read Lawrence’s book to the best of my ability. If I got it right there is no nothing, all the nothing we can find out there contains something.I still ask my question though.

    2. Nothing is the uncreated and creates nothing at all. “Nothing” is seemingly irrelevant.

      Simple minded me would point to Concepts. Both Difference and Distance are concepts and each requires a minimum of two points of reference.

      Could anything exist if Difference did not exist? There is no requirement that points of reference must first exist in any way other than virtually. The concept of Difference requires at minimum two points of reference and also implies autonomy. Points of reference in any context = information.

      Higgs-Boson description: “….The real important thing for me is that fundamental particles are as far as we can tell zero-dimensional particles. They have no radius. You can’t think of fundamental particles as being glass marbles. They literally have no extension in space. They can never bump into anything else.”

      Transformation = the primary activity in the Universe.

    3. Frank – The answers to this question depends on the definition of “nothing” you’re using – after all a volume of space from which everything else has been forbidden [an imaginary shield that can block entry by gravity waves, neutrinos, photons, magnetic & electric fields, dark matter etc etc] still has the quantum vacuum & it still has spacetime properties [I think it has spacetime or does that require the presence of mass?]. If that’s you’re “nothing then Victor Stenger is a worthy short read here:

      If you’re asking about why does a physical reality exist with properties rather than a ‘philosophical nothing’ without properties then your answer is as good and bad as everyone else’s! When you think about the type of answers that might be possible we humans keep coming back to our limitations by perhaps asking malformed questions – trapped as we are in a cause/effect world view

      If the question has meaning [I don’t think it does really], then I suppose the answer is going to be very weird with a time paradox thrown in somewhere

      1. It appears that Stengler’s root assumption is that Difference exists and differences are certain to demonstrate something.

  8. As an ignoramus on this subject, my simplistic thought is that we think of aggregated atoms or molecules as “something”. “Nothing” is what we imagine them to be when they disaggregate. For example, I exist. I am “something”. When I die, I will disincorporate and my atoms/molecules will be released into the universe to possibly become “something” else. The atoms/molecules that used to be me do not become “nothing”. They are just no longer the “something” that used to be me.

    1. It’s a more basic nothing Rowena. No matter, no photons, no quantum fields, no time, no space. Nothing ever happens. Nothing.

  9. “Why is there something instead of nothing?”

    “…but in this case, said Sean, it’s possible that that question, meaningless now, might someday find an answer if we learn more about the circumstances that produced our universe and that may be producing other universes.”

    I like this(SC)answer basically because the question drives to the edge of a cliff and that, slams on the brakes.

  10. He asked a very good question: “Why do we remember what happened yesterday but don’t remember what hasn’t happened yet?” His explanation—that we don’t fully grasp the Second Law of Thermodynamics—didn’t satisfy me, and I’m still seeking an answer.

    Sean’s response is in line with his response to “something rather than nothing”. I don’t have any own opinion on the symmetry breaking of time, though I think the ‘nothing’ idea is meaningless. But there are a lot of interesting ideas on time and memory, some of them described here: .

    As you can see there, in Sean’s framework of emergent properties, a quantum field “everyday” physics does lead to your description of that the passage of time from past to future is more or less an illusion. And that is the prevalent view, I take it:

    “Many physicists argue that Einstein’s position is implied by the two pillars of modern physics: Einstein’s masterpiece, the general theory of relativity, and the Standard Model of particle physics. The laws that underlie these theories are time-symmetric — that is, the physics they describe is the same, regardless of whether the variable called “time” increases or decreases. Moreover, they say nothing at all about the point we call “now” — a special moment (or so it appears) for us, but seemingly undefined when we talk about the universe at large. The resulting timeless cosmos is sometimes called a “block universe” — a static block of space-time in which any flow of time, or passage through it, must presumably be a mental construct or other illusion.

    Many physicists have made peace with the idea of a block universe, arguing that the task of the physicist is to describe how the universe appears from the point of view of individual observers. To understand the distinction between past, present and future, you have to “plunge into this block universe and ask: ‘How is an observer perceiving time?’” said Andreas Albrecht, a physicist at the University of California, Davis, and one of the founders of the theory of cosmic inflation.”

    1. Excellent article! I would like to quote one physicist from it who gives a plausible answer to Jerry’s question about memory:

      Increasing complexity, Koslowski said, has one crucial side effect: It leads to the formation of certain arrangements of matter that maintain their structure over time. These structures can store information; Koslowski calls them “records.” Gravity is the first and primary force that makes record formation possible; other processes then give rise to everything from fossils and tree rings to written documents.

      (Koslwoski claims gravity always acts so that there is a minimum-complexity state, which can be regarded as a starting point. Time “flow forward” (in both directions, even!) from there.)

      Jenann Ismael argues that the “block universe” is compatible with, and explains without explaining away, our experience of time. She’s right, but then, she usually is.

      1. Addendum: In case it wasn’t obvious, “records” would include not only tree rings and written documents, but memories in your brain. Hence, why we remember the past but not the future.

  11. Jerry, if you read this and have the time, I would appreciate it if you would ask Sean what he thinks of “The Zero Energy Universe” theory. If he needs to. he can Google the above and look at The Astronomical Society of the Pacific’s description. Thanks.


  12. Nothing, like Everything is a feature of the human mind that arises out of the mental operation of removing elements from a set, until the set is empty. It’s a paradoxical idealization. Our idea of nothing, void, death etc depends on the imagination of a set, a container, that is meant to be filled by something. To the human mind, the idea of nothing is unsettling, and we grappled with it for ages, but Christian beliefs subdued and fought that quest, since Nothing has no place in the Christian worldview.

    Christianity had (undesired) deterministic features in “God’s Plan” because God knows everything, and he is just, therefore you being a beggar and suffering from a rotten teeth is your fault. People of the Renaissance could not understand the bizarre demands of christian Sophisticated Theology™ even then, and went with folk beliefs instead. The imagery of death that begins to emerge was perhaps an escape of the incomprehensible Christian phobocracy that wanted it both ways, Free Will, but yet a rigid “divine and just” social structure.

    Death already shows some features of entropy, as it was the great leveller, an even-making entity which gave comfort to our ancestors. A long history is visible in the traditions of the Danse Macabre or Totentanz, the Memento Mori (remember, you will die), or of course, the Grim Reaper. It always was a odd element, that doesn’t quit fit into Sophisticated Theology™ that accepts no nothing. God always existed, and afterlife doesn’t feature non-existence, either.

    To imagine the nothing, our ancestors solved this problem with a dancing pair. In a kind of table from king to beggar, as visualized by their attire, every member of society was dancing with another figure symbolizing Not-Life, and gave comfort that death will not spare anyone and treat everyone equally.

    At first, the image of “nothingness” was showing corpses with maggots and other visible signs that life wasn’t there anymore. Later, it was symbolized as a skeleton – probably with increased knowledge of anatomy – to a more recent invisible hooded grim-reaper-monk-like character. In short, we cannot escape the semiotic property of having at once at mental representation (which is “something”), which is however empty of referents. There’s always the container. We do it with language and words, and in visual arts, they threw a hood over an invisible man – the slot where our mind expects the “man”.

    Today, I learned from Sean Carroll the idea of “poetic naturalism” from the interview linked to in 9 (thanks Phil), which is described as “[t]here is only one world, but we have many ways of talking about that world”, which is a view, I embrace too, but know as “model-dependent realism”, which itself is just one name of an older idea.

    I’d combine that with experientialism. We apprehend reality as a creature that is a few feet tall, has a pair of eyes, hands an ears, and equipped with a mind that evolved to help us survive. We understand by storing the word in small models contained in our categories (ICMs, idealized cognitive model) which form the building blocks for bigger models, including the models of science, folk psychology, religion etc. (however, note, this isn’t a form of relativism, because there really is a territory and we can make more or less accurate maps of it, just no one map that features everything).

    By adding more elements into a category, a skeletal figure of properties remains as the concrete flesh is etched away. If you know just one moon, it’s of course pale, and orbits the earth. But once you know more moons, you remove that concrete property in favour of a theoretical range of only possible properties. The skeleton is quite a nice metaphor, since we imagine a generic skeleton that can be “fleshed out” to make any individual. The reverse process is adding more possibilities means that our ideas become abstract, skeletal, or structural. But if you can remove parts, you can also arrive at the skeleton itself, the container or set that has “no thing” in it, perhaps pure potential. But what is it in reality? It would require a space (as a container) where not even the laws of nature operate. It’s easy to say, because working with mental representations is easy, but impossible to imagine.

    1. Interesting analysis. It compliments Carroll’s views nicely. The set theoretic metaphor must definitely be the origin of the naive view of nothing. Hard to shake.

  13. I don’t think there is such a state as nothingness, there is always either Energy or Matter. so Space may look empty but it isn’t.

    1. The ‘nothingness’ of philosophy is also the absence of space.

      As others have noted above, Lawrence Krauss gives a good discussion of the various types of nothing and whether something can come from each of them in his book.

  14. In relation to the question “why is there something rather than nothing?”, I found Jim Holt’s book Why Does the World Exist? An Existential Detective Story to be entertaining and well-written.

Leave a Comment

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