lpetrich

I've thought a bit about that, and I consider that argument partially bogus. However, there is a related problem that is more serious, as I shall explain.

One version of the argument states that if some physical constants were only a tiny bit different, then life would be impossible because of such things as chemical bonding and reactions not working properly.

However, if one takes a closer look at the physics behind chemical bonding, it turns out to be less fine-tuned than some people seem to think. This comes from work in quantum chemistry, attempts to predict the properties of chemical bonds using from-scratch quantum mechanics. The only physics that enters is that electrons are spin-1/2, electrostatically interacting, and much lighter than nuclei -- meaning that chemical bonds would keep their angles, relative lengths, and relative energies if the Fine Structure Constant ((electric charge)^2/(4*pi); dimensionless in hbar = c = 1 units) got changed or the electron got lighter or heavier relative to the nuclei.

There are interesting effects in other areas. If the FSC was smaller, then nuclei could extend up to higher masses, since protons would repel each other less. Likewise, if the FSC was larger, then nuclei would extend up to lower masses than they do, because of protons' greater repulsion of each other.

However, there would still be long-lived radioisotopes at the upper end of the mass scale, like the uranium and thorium in our Universe, because instability due to proton repulsion is not a sharp cliff. And stellar-nucleosynthesis processes that produce heavy elements would produce them all the way up to where they become too unstable to last.

It may be possible to derive nuclear binding energies directly from quarks, gluons, and Quantum Chromodynamics in the way that one would do quantum chemistry; one nuclear physicist once joked that particle physicists have turned them into chemists.

But there is a serious difficulty. The electromagnetic FSC has a value of 1/137, making it easy to expand in powers of it, while the QCD equivalent has a value near 1 at the energy scales typical of nucleons -- making it much more difficult. To do nucleon structure requires an enormous amount of computer time, and that's with picturing space-time as a very coarse grid, something like 10*10*10*10. And that barely covers nucleon-nucleon interactions. However, one may be able to hand-wave one's way through nuclear structure by using the fact that the up and down quarks (those in nucleons) are nearly massless (a few MeV) compared to QCD's built-in energy scale of a few hundred MeV.

There are other interesting problems, such as the maximum sizes of planets and the luminosities and lifetimes of stars that can be treated in this way, but I'll skip on that.

There are some features that seem very convenient for us, but it is always possible that there are even better possibilities than those of our Universe.

One curious oddity is that neutrons are more massive than protons. From this, one infers that down quarks are more massive than up ones, which is contrary to the trend shown by their heavier relatives (strange less than charm, bottom less than top). This circumstance has allowed hydrogen to form in the Big Bang; if protons were the unstable one, then the Big Bang would have formed a surfeit of neutrons instead of protons, meaning that hydrogen would be a rare element. Stars would behave differently, since they'd be mostly helium and heavier elements, meaning that they'd burn out more quickly than the stars of our Universe do.

It is interesting that QCD becomes superstrong at energies of a few hundred MeV; this is why nucleons are much more massive than electrons -- the strong quark-gluon and gluon-gluon interactions make the quarks relativistic, with kinetic energies at the QCD energy scale.

If QCD got strong at much less energy, then the rest masses of the up and down quarks would dominate the nucleons' masses, making them not much more massive than electrons. This would have fun consequences in chemical bonds, but I don't think that that would be fatal. There is also the question of nuclei forming with a weak QCD interaction; at energies higher than its natural energy scale, its FSC equivalent gets smaller and smaller in reciprocal-of-logarithm-of-energy fashion.

But if QCD got superstrong at much higher energy scales, then the Universe would still be much like our Universe, but with much more massive nucleons.

Now to the question of weak interactions. These are weak because of a "symmetry breaking" of a combined electromagnetic-weak force that happens at energies of a few hundred GeV. If this symmetry breaking did not happen, then beta decays would happen much faster, and there would be an "extra" electromagnetic-like force. But it happens, and it is likely connected to the masses of the electrons (e, mu, tau) and the quarks. The top quark has a mass of 150 GeV, which is almost the right size, but the other quarks and all the electrons are much less massive, with the "true" electron being 300,000 times less massive! This may be due to some quantum-mechanical spillover, but the details are unclear.

Also, if supersymmetry is real, it is expected to be broken, and the energy scale of this breaking is expected to be a little above the electroweak-interaction symmetry breaking -- and may be related to that breaking.

What makes this symmetry breaking happen is obscure, however; but one thing less obscure is that the FSC equivalents of the electroweak and QCD forces change reciprocal-logarithmically with interaction energy, converging on a single value at about 10^15 GeV -- implying that they were parts of a single interaction that was split up by symmetry breaking. This is, of course, GUT territory. But it's not clear why there is a 10^12 ratio of energies between GUT symmetry breaking and SUSY/electroweak symmetry breaking. It may be connected with the logarithmic rate of change of various interaction constants with energy (huge difference in energy to produce a significant change).

But the GUT energy scale is close to the Planck energy scale of 10^19 GeV (10^19 that of a nucleon, 10^22 that of an electron), that of quantum gravity. This suggests some connection, though exactly what connection is obscure.

But one interesting consequence of this great difference in energies between the familiar elementary-particle world and gravity is that we can live in a very complicated Universe, owing to the resulting weakness of gravity. For example, this enables the largest planet (approx. Jupiter) to have an enormous number of elementary particles, while if the familiar elementary particles had GUT-scale masses, then such an object could not have many particles in it.

However, there does not seem to be much fine tuning here; the gravity-elementary-particle energy-scale discrepancy can be somewhat larger or smaller without producing a drastically-different Universe.

But the final question is: why the particular GUT that had led to the particles of our Universe? What other possibilities could there be?

One attempt to answer this question has been to explore superstring theory, but although that has no free parameters, it does have a large number of possible ground states -- which correspond to different GUT's. What makes a superstring "choose" one and not another is, however, an unsolved problem. Another oddity that must be explained is that superstrings prefer to live in 10 space-time dimensions, while we directly observe only 4. It is expected that the other 6 will curl up into a tiny ball somewhere from Planck-sized to GUT-sized. But why this 4+6 split? Why not some other?

Actually, 4 space-time dimensions are convenient for us, since that allows for complicated structures while allowing objects to orbit each other, producing many of the structures of our familiar Universe. The inverse-square law of gravity and electromagnetism becomes inverse-(D-1) for D space dimensions; if D = 4, then orbits are borderline unstable, and if D is greater than 4, then orbits are definitely unstable. Thus, our familiar Universe has to have 3 space dimensions to allow us to exist.

But superstrings offer an intriguing hypothesis; our Universe could be a supercooled bubble in some "superstring soup" that has several other such bubbles in different ground states, producing Universes that were usually sterile -- and sometimes inhabited when they could allow inhabitants to come into existence, as ours does.

liquid

Nice piece.

I've always had a more basic problem with the argument, which in some ways is what you have expanded on.

It's putting the cart before the horse, so to speak.

It seems amazing the universe is so suited to us... but it is incredibly mundane when you realise that the real position is that we are so suited to the universe, because we are part of it.

Friar Bellows

My main objections with "fine tuning" arguments are the following:

1) Are the fundamental constants more likely, equally likely or less likely to have taken on one set of values compared to another set of values? We have no idea how to estimate the probability of the fundamental constants taking on the values that they have. So how can we conclude that the tuning was "fine" or evidence of "design"?

2) But OK, let's accept that the universe appears to be fine-tuned to allow our existence. And what is so special about that? If I give you a certain set of values for the fundamental constants to take on, can you predict with any certainty and describe with any accuracy what sort of complex structures may (or may not) arise in such a universe? I don't think so.

We simply lack the imagination and the knowledge of how nature works to be able to make such predictions and descriptions. For example, every time we build a bigger and better telescope and discover something new, it always surprises us. I think it's possible that there are forms of intelligent life in the universe which are beyond our wildest dreams and certainly beyond our predictive capabilities.

So different values of the fundamental constants may produce different complex structures beyond our imagining. And one particular set of values may not only produce one type of complex structure (e.g. carbon-based), it may produce several types.

3) The universe as defined by our cosmological models and our fundamental constants may be one among many. Perhaps countless others don't have complex structures. We would be in a universe that did allow it, but it would be an acceptable fluke. If zillions of universes exist then we're just a statistical fluke.

OK, maybe you prefer Bayesian arguments:

<a href="http://quasar.as.utexas.edu/anthropic.html" target="_blank">Proposed Anthropic Principle FAQ</a>



[ April 30, 2002: Message edited by: Friar Bellows ]</p>

lpetrich

GUT's or superstring theory might have an answer to that question; but such theories are still incompletely developed. One can get some of the Standard Model out of some GUT"s and superstring states, but there are still plenty of loose ends, like the masses of the elementary particles.

A good indicator of that is the physics of materials and chemstry. One could calculate the physical and chemical properties of familiar materials entirely from scratch, using only quantum mechanics and a big-enough computer.

However, that is an extremely difficult task, as can be seen from the state of the art in quantum chemistry. So in practice, one usually cheats and measures various quantities experimentally, using those values as the theoretically-expected ones.

And if quantum chemistry seems difficult, one should consider the difficulty of determining hadron structure -- an enormous amount of computation is required to get the behavior of only two or three quarks interacting with gluons. Which is why nuclear and particle physicists end up doing the same sort of cheating.

But that sort of cheating is generally unavailable when one considers some hypothetical Universe; only if one can show that some situation is mathematically equivalent to a comparable situation in our Universe can one use our big library of cheats. That will likely happen in some cases, but by no means all.

Which is one of my speculations. There could be some superstring-soup super-Universe -- a super-Universe from which bubbles would occasionally condense, with us living in one of those bubbles -- a bubble that had condensed into a state that would produce our familiar laws of physics. Many of these bubbles, however, would condense into Universes that are completely sterile, thus having no observers that wonder why the Universe is the way it is.

[ April 30, 2002: Message edited by: lpetrich ]</p>

Bill

In <a href="http://www.secweb.org/bookstore/bookdetail.asp?BookID=186" target="_blank">The Elegant Universe</a>, Brian Greene explains that one of the goals for superstring theory is to derive a model of the universe which has no input variables whatsoever, and which produces all of the known constants (like all those discussed as "fine tuning" candidates) as natural outputs from the equations of superstring theory. This is, of course, an ambitious goal, but there are some very exciting hints of possible success from just the little bit of computation that has been done to date.

For so long as this scenario remains a viable candidate for a future Nobel Prize, then I'm willing to suspend my disbelief in the upcoming discovery of the "Holy Grail" of Physics and patiently await further advancements.

Needless to say, if Greene's prediction is met, the whole "Fine Tuning" argument goes out the window (along with any idea that God had anything at all to do with the underlying "machinery" of the universe).

== Bill

echidna

But Bill, he only attempts half the problem. To be truly explanatory, Superstring Theory not only needs to derive the universal constants, but also to be able to bootstrap itself into existence ex nihilo.

“Buckley’s” is an Australian expression for no chance whatsoever.

The Fine Tuning argument remains.

lpetrich

As I had mentioned, one ought to be able to derive hadronic/nuclear structure and atomic/molecular/condensed-matter structure from the Standard Model of particle physics, but in practice, that is very difficult, which can make it difficult to evaluate alternatives to it. The next questions are how easy would it be to derive something like the Standard Model from some Grand Unified Theory, and how easy would it be to work out what GUT's can possibly occur.

The Standard Model features this ensemble of particles:

* A set of photonlike "gauge" particles, with a set of symmetries called SU(3)*SU(2)*U(1). The SU(3) part makes up the 8 gluons, which produce QCD. The SU(2)*U(1) part is the electroweak part, which makes up 4 particles, two of which mix to produce the (massless) photon and the (massive) Z; the other two produce the two charges of the (massive) W.

Each set of gauge fields has its own coupling constant; the SU(3) ones have one, the SU(2) ones have one, and the U(1) ones have one.

The mechanism that makes the W's and the Z massive is called "symmetry breaking". Here is how it works. Imagine a bowl. It is symmetric around its central axis; rotate the bowl around that axis, and it will look the same. Put a marble in it, and it will come to rest at the bowl's center. The bowl + marble keeps that symmetry. Now imagine a bowl with a hump in the middle -- a hump that keeps that rotational symmetry. Put a marble in it, and it will come to rest in the valley around the hump. But its presence there will break the rotation symmetry of the bowl + marble system -- rotating it will make the marble have a different position relative to you.

The favorite mechanism for symmetry breaking is a spin-0 "Higgs field" that works much like the marble in the central-hump bowl, acquiring a nonzero ground-state value. This nonzero value will give masses to all the elementary particles that interact with it, and this is thought to be the origin of most elementary-particle mass.

* A set of electronlike "elementary fermions", which come in three generations, each with a highly electronlike particle (e, mu, tau), a neutrino, a downlike quark (down, strange, bottom), and an uplike quark (up, charm, top). The quarks have three QCD states or "colors", while the leptons (electrons, neutrinos) have only the neutral QCD state. Each one of these has its own mass, and weak interactions turn a mass state of one of them into a mixture of mass states of the related particle. Thus, an up quark maps onto a down, a strange, and a bottom, in differing proportions, of course.

All this means that the Standard Model has lots and lots of free parameters, a proliferation that has happened before: the chemical elements (electrons + nuclei), nuclei (nucleons), and hadrons (quarks + gluons). Which has led to a lot of investigation of GUT's, which picture a gauge field following some high-dimension symmetry that somehow gets broken to the Standard Model's symmetry. Imagine rotating a multi-dimensional hypersphere and all its symmetries. This symmetry breaking would also split some GUT elementary-fermion fields into the multitude of fields that we observe.

Working out the pattern of symmetry breakings of a GUT is straightforward, at least for a particle physicist, turning the problem into one of why one GUT and not some other.

Another symmetry seriously considered is supersymmetry. It states that every elementary particle has a counterpart with a spin different by 1/2. Thus, the gauge-field particles have counterparts with spin 1/2, and the elementary fermions have counterparts with spin 0. But one important prediction of supersymmetry is that related particles must have the same mass. But that is not observed, so supersymmetry must somehow be broken; this breaking will give a mass to at least one of some superpartner set.

But why a mass to one and not another superpartner? Why the elementary fermions and not their scalar (spin-0) partners? Why the superpartners of the gauge fields and not those fields themselves?

It must be said that this is convenient for us, because electrons being fermions allows them to form complicated structures by electrons being forced into states different from those of other electrons (the Pauli Exclusion Principle), instead of trying to get into the same state and form a Bose-Einstein Condensate, as their scalar counterparts would try to do.

Also, nucleons do that because they also have spin 1/2, just like their component quarks; if the QCD interaction featured 2 or 4 colors instead of 3, then nucleons would have integer spin, meaning that they'd also try to form a B-E condensate and not the familiar nuclear structures.

The answer to this is in the structure of the original supersymmetric GUT; as before, one must find a SUSY GUT that produces the appropriate supersymmetry breaking.

But now that we have pushed the question back to GUT structure, we must consider what possible GUT's there can be.

One possible answer that has been emerging is superstring theory. This posits an infinite set of interrelated elementary particles that are actually the vibration modes of a stringlike entity. All but a few of them are massive, with the squares of the masses being multiples of the string tension. That tension is thought to be somewhere around the square of the Planck mass (quantum-gravity scale), meaning that superstringiness will be hard to test.

However, superstrings will produce a spectrum of zero-mass particles, including supersymmetric gravity ("supergravity") and a supersymmetric gauge field with a big symmetry (something like the rotations of a 32-dimensional sphere) -- all in 10 space-time dimensions.

This may seem like what we are looking for -- a fully-constrained GUT. But superstrings have a multitude of possible ground states, many with various space-time symmetry breakings. It's expected that our Universe has such a symmetry breaking, from 10 dimensions to 4 big dimensions and 6 dimensions curled up into a Planck-sized ball ("compactification"). Unlike the case with earlier elementary-particle theories, it has been difficult to work out which of these symmetry breakings is from the "true" ground state -- if there is one at all.

Such a space-time symmetry breaking turns the gauge field into a smaller-symmetry gauge field plus some elementary fermions and their superpartners -- the usual contents of a GUT. Thus, different space-time symmetry breakings yield different GUT's, meaning that the allowed STSB's control the spectrum of allowed GUT's.

So an answer might be in sight, especially if the STSB's allowed by superstrings turn out to be a very small set -- or only one.

eh

Well it be great to see the fine tuning argument snuffed out by a TOE. But who here doubts that deep thinkers like Metacrock will still be posting it as a proof for god on their websites?

Say echidna,

Why would string theory need to explain an ex-nihilo creation event? We don't need to explain how a universe could have come from nothing if there was never a time when 'nothing' existed.

lpetrich

I agree, though I think it's bogus for other reasons, as I have explained.

I think that echidna was making a joke. Actually, creation from nothing is a rather unusual doctrine; many creation stories picture emergence from some sort of primordial chaos. Even the Bible seems to suggest that its God often needs some sort of pre-existing material to work with. This is rather obvious from the Genesis 2 story, and somewhat less so in the Genesis 1 story.

And creation from some primordial chaos is what my superstring-soup scenario implies.

[ May 04, 2002: Message edited by: lpetrich ]</p>

echidna

My loose understanding is such that even beyond a simple Big Bang / Creation, existence still reduces down to a problem of contingency, necessary-ness. And theologists have already been considering for this many centuries, long before Big Bangs were speculated. I’m weak in philosophy but I think in the thirteenth century St Thomas Aquinas was exploring the possibility of a contingent essence, something which was necessary for existence to exist (shades of Metacrock I know). Fortunately I differ that I don’t believe his work resulted in proofs of the existence of God, but I find the concepts intriguing, even moreso now science begins to acknowledge the possibility of existence beyond the Big Bang.

Unfortunately infinite existence still answers nothing with respect to our origins. It might rule out a Big Bang or Biblical Creation as the start of our universe, however goes no further towards explanation, unless the TOE can demonstrate that it is inherently contingent, self-requisite.