Monday, April 06, 2009

BBC HARDtalk with Roger Penrose: before the Big Bang

We have already embedded and compared Hardtalks with Václav Klaus and Lee Smolin.

Here is one from 2006 with Roger Penrose, a very colorful and creative thinker, dealing with the pre-Big-Bang cosmology. See the full Real Video. In the remaining 10 minutes that are not available on YouTube, Penrose talks trash about quantum mechanics and defends the right of people to do pure science from the primitive journalist.

At some moment, they jump to artificial intelligence and consciousness of robots. Sackur can't distinguish "unconscious" objects from "man-made" things, so Penrose reminds him that we can make not only computers but children, too. ;-)

In my opinion, Mr Sackur confirms he is not fair. Klaus was asked where he took the "arrogance" to dispute the dangerous global warming theory promoted by those amazing "experts".

But when it comes to the Big Bang, suddenly it is "the most popular but far from unanimously" accepted a theory. Wow. So our robust state-of-the-art picture of the evolution of the Universe is just popular but it is "arrogant" to disagree with politicized pseudoscientific hacks who promote global warming fears - fears based on no non-trivial theory, no empirical evidence, and no high-quality scholars at all.

Sackur has tried to pump the global warming shit into Roger Penrose, too, convincing him that it matters more than the Universe. To be fair, he also spends 3 minutes trying to force Penrose to accept God. ;-) OK, let me stop complaining about Mr Sackur who is just an obnoxious BBC moron and let us return to the real cosmological issues.

A journey to ever more exotic corners of the Cosmos

One fact I find important here is that we are learning the story of the Universe essentially according to the scales of various types. In the case of the origin of the Universe, the age of the Universe is the crucial quantity.

The expansion of the Universe right now seems indisputable, and it seems obvious that we can extrapolate it billions year into the past when the Universe was smaller and denser.

The story almost certainly works up to the time when the Universe, starting from the conventional Big Bang, was 300,000 years old because this is when the microwave background - with features observed and correctly predicted by the theory - was created by "decoupling".

Our theory is arguably much more far-reaching and works up to the moment when the Universe was three minutes old and very dense: nucleosynthesis, the relevant part of the cosmic biography, is confirmed by the correct abundance of light elements we observe.

At this moment, after all those successes, it's sensible to trust quantum field theory coupled to general relativity. And because our effective quantum field theory, the Standard Model, was verified up to 100 GeV, we can say that we know the qualitative physics up to the ancient times when the Universe was a fraction of a second old.

Since that time, the Universe was expanding, the energy density was decreasing, but the total entropy had to be increasing because of the second law. The most obvious "minimal" hypothesis begins with a singularity. It is obvious that near this singularity, some additional laws of physics have to be added because every singularity in space is an ultraviolet effect that shows the incompleteness of your approximate (in this case even classical) theory.

But the next question must be what the new physics that governs the beginning is supposed to be.

Some people want to insert many cycles, proposing the cyclic models. I see zero theoretical or empirical evidence that this non-minimal model is superior. You know, if you have a minimal and simple model and a non-minimal one, you should expect the non-minimal model to give you better or more accurate predictions - or be consistent with a wider set of observations or insights. Otherwise the minimal model should be preferred.

The advocates of the cyclic or ekpyrotic models seem to think that they solve something by their non-minimal assumption but they have never explained what the something actually is. They surely don't remove singularities from the Universe. Because the entropy keeps on increasing, each cycle is longer than the previous one and the total geometric sequence can be summed, leading to a finite Universe, anyway.


Instead, proper physics as I understand it postulates a robust modification of the expansion of the Universe at the very beginning, the cosmic inflation. It solves the horizon problem and various monopole and other "exotic" problems.

Inflation makes the Universe at the end of this stage pretty much flat, large, and independent of the detailed state at the beginning of inflation itself. I think that all these things are good features of the paradigm and when people get self-confident about the logic in the future, they may say that some of them are inevitable for a consistent theory.

The cosmic inflation is testable through detailed observations of the cosmic microwave background (and perhaps the gravitational waves) and it is arguably successfully tested. So I think that inflation is "just another successful step" in the one-by-one process, explained at the beginning, to reveal increasingly more exotic and inaccessible secrets of the Universe by increasingly penetrating observations and logic.

Dmitry, I don't really understand why Roger Penrose doesn't believe inflation either.

Low entropy at the beginning

Thank God, Penrose accepts that the second law of thermodynamics means that things get more and more organized if we move towards the beginning. At any rate, he seems to be puzzled by the low entropy. I think that this puzzled attitude is irrational. When the Universe was small and close to its most fundamental laws, it's clear that the entropy had to be "of order one".

On the contrary, it would be very puzzling if some initial, fundamental quantity - such as the entropy - was a huge constant very different from one to start with. In that case, we would be asking why it was this number and not another number? A new type of a hierarchy problem would be born. It is very good that the existing theories, especially inflation, allow us to explain very large numbers in the present Universe - like the number of photons - by an exponential formula.

Now, Penrose says that "no theory explains why the entropy was low to start with". Well, yes and no. No theory about the evolution of the Universe explains this - and it is obvious why it cannot. If a theory only describes things that occur after "t=0", it can't say anything about "t=0" or the time before it if there is any.

To say something about the initial state, we need to look at a theory about the initial conditions. So the correct question is whether e.g. the Hartle-Hawking state implies a low initial entropy, and I think it does. This wave function doesn't have too many competitors, so it's not surprising that questions about the very origin are ill-defined in virtually all theories we know.

Before the Big Bang

If you summarize what I said about the hypothetical evolution before the Big Bang, or about the initial conditions - which are two frameworks to answer the very same question at the edge of our current knowledge - it's clearly important that such a theory solves something and gives you a more sensible or accurate explanation of something that doesn't follow from the general vague assumption that the "Big Bang began with some unknown initial conditions" or that the "Big Bang began after some unknown pre-history". If you say something more specific, it should be the right thing and you should have some arguments that it is more right than other possible answers.

These are deep and difficult questions but I am not aware of arguments of this kind offered by Roger Penrose to support his opinions. So, using Occam's words, entities should not be multiplied unless it is necessary.

Eternal inflation

The canonical picture assuming the evolution of a broader Universe is the eternal inflation. Our Universe is born from an inflating seed in some parent Universe that had a similar origin. It's important to notice that this picture doesn't try to modify the most recent 13.7 billion years of our cosmic history textbook. It tries to change our idea about what happened before that.

A complete theory can make such an evolution - the growth of bubbles within each Universe - inevitable and preliminary results in string theory may indicate that it does so. However, one must still carefully distinguish what is established, what is motivated, which concepts are justified, and which are not.

If the seed of our Universe, starting with inflation or the Big Bang, is born out of a random fluctuation in a parent Universe, you may lose any observable consequences of the pre-evolution - as Stephen Hawking correctly emphasizes - and the correct physicist's attitude should be to cut off the whole pre-history.

Eternal inflation as a tiny Boltzmann egg

In this context, it is useful to mention that such a birth of the small Universe is a modern, modest, and realistic variation of the Boltzmann brain theory. Boltzmann was the guy who really discovered the idea that because the matter is made out of atoms, a finite piece of it carries a finite amount of information. One really needs some quantum mechanics to regulate the UV degrees of freedom but Boltzmann correctly predicted what quantum mechanics would later do with such things.

If the piece of matter carries finite information, all of its possible configurations may be born out of chaos, after some time. The only question is what the time is. Boltzmann was a pioneer of this statistical understanding of matter and we often underestimate how revolutionary these ideas used to be. So it is not surprising that he coined the concept of the Boltzmann brain, too.

But that doesn't mean that Boltzmann actually believed that whole organized brains are actually created out of chaos. He strictly believed Darwin's evolution. Well, he may have believed some Lamarck's ideas, too. But it is historically absurd to suggest that Boltzmann was serious proposing the "Boltzmann brain" paradigm as an alternative to evolution.

The creation of a seed of the Big Bang from a fluctuation in a pre-existing larger Universe was much closer to what Boltzmann had in mind. Believe me, he was able to calculate that it is possible to create microscopic objects as random "fluctuations" but the probability becomes nonsensically tiny if you want to create macroscopic organized objects.

OK, what about this idea, of a seed of our Universe as a fluctuation? Is it useful for something? Can't we simply truncate time at the moment of the statistical fluctuation, as Hawking dictates?

The a priori physicist's answer must clearly be Yes. If the Universe and its internal logic dramatically changes, in some very unlikely way, then it is the best description of a "new beginning" you can ever get in physics. You should forget the "time before that". This "time before that" and the "new time" cannot be identified, anyway.

If we want to use one time throughout some history, physics must be at least qualitatively continuous as a function of this time. If it is brutally and qualitatively discontinuous, the time coordinates in the two phases are really different degrees of freedom, in the very same sense in which the momentum and the winding number in one Universe are different.

Time is a theoretical construct

The ability to use the same "time" throughout a long history is a simplification, a result of assumptions that hold when the Universe is smooth, but assumptions that can break down if it is not. People often don't realize this fact so let me state it again: "time" is a theoretical concept. By connecting our present world with any other world via a continuous time, we are explaining certain things.

But that doesn't mean that this "theory" (of time) must be valid in an arbitrarily extreme context. In the extreme contexts, time may be - and probably is - an emergent concept.

It is always possible to drop this assumption about time that always exists and reformulate our predictions about our present and future observations without the assumption of a time in the far past. The time is always just an auxiliary concept, if you wish, and it becomes increasingly "abstract" or "unreal" if you move to increasingly exotic regions of spacetime.

When you get to the very beginning of the Big Bang expansion, you may be forced to drop the time as a concept. So the more general question you should be asking is what are the probabilities of different values of quantities when our Universe was very small. Can these probabilities be calculated from a pre-existing Universe?

Incidentally, Penrose agrees that time can "fail to exist". Except that his particular context in which time should disappear - the disappearance of massive particles - seems to be a wrong context to show this disappearance.

Looking at the DNA of the parent Universe

Now, if you need a very unlikely fluctuation, it is likely that "anything can happen". You may still argue that even among the unlikely events, some of them are more likely than others. And these likelier events could be derived from the pre-existing Universe. It is conceivable that our Universe was born from a piece of essentially empty space in the parent Universe. Fine.

I just happen to think that we will always be unlikely to derive any detailed information about the initial conditions from this sort of reasoning. If your probabilities about the current Universe depend on probabilities of different matter distributions in its parent, you are reducing a question about our current Universe to an equally difficult question about the parent Universe.

It just doesn't look like you are making a progress. Where's the beef? You are simply postponing the question, much like if you use God to explain the beauty of Nature, without asking who created the beautiful God. And every level of gods or parent Universes makes your theory more inaccessible and abstract, so unless you show me otherwise, it is sensible for me to expect that the situation is getting more hopeless, not more promising, as you are adding the layers.

And because I think that it is pure crap to link the probabilities of different options to the volume in a multiverse, there is probably no way to convince me that this construction of new layers in a Universe is scientifically valuable.

Also, if you try to imagine the initial configurations of matter in our Universe that were likely to emerge from a different one, I think that you won't be able to derive any "microscopic details" about the seed of our Cosmos. In the most optimistic case, you will be able to say that it was small and its scalar fields were pretty closed to to one of "selected" stationary points in the configuration space (landscape).

That would be great and it could be enough to solve the vacuum selection problem. But the whole long pre-history in the parent Universe would almost certainly be useless for such a hypothetical victory. In this sense, we should still say that our Universe began with the ordinary Big Bang.

New theories must bring new results

The text above defines what kinds of questions, answers, and explanations can a priori exist. You may view them as tautologies because I am not really assuming anything that should be controversial. On the other hand, if someone wants to say something about the initial conditions for our Big Bang, he must actually have some results that make sense - some predictions or some derivations that are made inevitable by other established pieces of theory.

I am afraid that many people in this area, including some very well-known people, are just bullshitting around. They have no results, they have no logic that puts things together, they have no new mathematical objects that can be shown to be relevant for quantum cosmology: they are just trying to be interesting by taking sides. And that's not interesting for me.

Via Dmitry.


  1. Gravity Limits Link Ultracold And Superhot,
    Our Inability To Create Singularity

    A. From "Strings Link the Ultracold with the Superhot"
    Perfect liquids suggest theory’s math mirrors something real

    "When the universe was very young, and still superhot from the aftermath of the Big Bang, plasma should have been the only state of matter around. And that’s what scientists at Brookhaven expected to see when they smashed gold ions together at 99.99 percent of the speed of light using a machine called RHIC (for Relativistic Heavy Ion Collider). RHIC physicists thought the ion collisions would melt the gold’s protons and neutrons into a hot plasma of quarks and gluons at a temperature of a trillion kelvins, replicating conditions similar to those a microsecond after the birth of the universe. But instead of a gaslike plasma, the physicists reported in 2005, RHIC served up a hot quark soup, behaving more like a liquid than a plasma or gas."

    B. The expectation of Brookhaven scientists was a bit unrealistic

    The "aftermath of the Big Bang" lasted much less than 10^-35 seconds. This is evidenced by the fact that "Gravity Is THE Manifestation Of The Onset Of Cosmic Inflation Cataclysm" :

    With all respect due to the scientists at Brookhaven it is very difficult to expect that they can recreate the state of pre big-bang energy-mass singularity.

    Commonsense is still the best scientific approach.

    Respectfully suggesting,

    Dov Henis
    (Comments From The 22nd Century)
    EVOLUTION Beyond Darwin 200


  2. Origin And Nature Of Earth Life, An Update…

    tags: life genesis, natural selection, life mass format

    Liberate your mind from concepts dictated by religious
    trade-union AAAS.

    Life is just another mass format + re-comprehend natural
    selection + natural selection is ubiquitous.

    Life Evolves by Naturally Selected Organic Matter


    Homegrown Organic Matter Found on Mars, But No Life

    II. EarthLife
    Genesis From Aromaticity/H-Bonding

    September 30, 2011


    Purines and pyrimidines are two of the building blocks of
    nucleic acids. Only two purines and three pyrimidines occur widely in nucleic


    Pyrimidine is a heterocyclic aromatic organic compound
    similar to benzene and pyridine, containing two nitrogen atoms at positions 1
    and 3 of the six-member ring.

    A purine is a heterocyclic aromatic organic compound,
    consisting of a pyrimidine ring fused to an imidazole ring. Purines, including
    substituted purines and their tautomers, are the most widely distributed kind
    of nitrogen-containing heterocycle in nature.

    Aromaticity ( Kekule, Loschmidt, Thiele) is essential for
    the Krebs Cycle for energy production.


    Natural selection is E (energy) temporarily constrained
    in an m (mass) format.

    Natural selection is a universal ubiquitous trait of
    ALL mass spin formats, inanimate and animate.

    Life began/evolved on Earth with the natural selection of
    inanimate RNA, then of some RNA nucleotides, then arriving at the ultimate mode
    of natural selection – self replication.

    Aromaticity enables good constraining of energy and good
    propensity to hydrogen bonding. The address of Earth Life Genesis, of phasing
    from inanimate to animate natural selection, is Aromaticity.Hydrogen Bonding.

    Dov Henis (comments from 22nd century)