Planck regime - GR and QM incompatibility?

[Spaceman Spiff]

We've drifted off the OP - but that's OK, this is a very interesting topic.

There's one angle which we've only just touched on - changes in the nature of science itself.

We can look back and identify - imperfectly - what elements of 'modern science' we see in the work of the ancients; we can look to the more recent past and see how 'modern science' is continuing to change ... but can we project into the future? With what degree of confidence can we say that what we are doing today, as physics, as cosmology, will be regarded as solid 500 years from now? 5 million?? Or the converse - what will a future Robin Dunbar tease out of the dusty records of the 20th/21st century to show correspondence with the then consensus view of 'modern science' (and point out just a few examples of where it falls way short)?

While I am no historian of science (I am an astronomer by profession), let me start this off by looking back.

While it is tempting to compare the cosmological (in the broadest sense of that word) models of Eudoxis, Aristotle, and Ptolemy to that of Copernicus, Galileo and Newton to that of Einstein, Hubble, FRW & modern cosmology (and it appears that some have suggested doing so, above), I hesitate in taking these comparisons too far. While these different major groups of investigators of the natural world had many similar ways in approaching the problem (e.g., the appreciation of simplicity in unifying many observational phenomena, the appreciation of symmetry, amongst others), science as a methodolgy of investigating, predicting and modeling the behavior of nature developed along the way (e.g., the importance of data and experiment in support of a hypothesis, amongst many others). As this methodology matured, the rapidity in scientific discoveries grew enormously - indeed the two likely fed back on one another.

This rapid progression in our ability to predict the behavior of nature (identifying its "laws", deriving general theories, etc) is telling us that we must be doing something right, and not just randomly wapping around in the dark. In fact when measured across all the sciences, one might argue this growth is similar to an exponential function with a shortening time constant. That doesn't mean I believe that the "Theory of Everything" lies just around the corner (if it even exists at all).

That's enough for a start...somebody want to run with (or against) that or project forward into the future?

[Ken G]

Projecting the evolution of science into the future is a fascinating endeavor, and although it is just guesswork, it might bring into better perspective the other issues we have been addressing. I think that if someone does come up with a theory that unifies graviton exchange with spacetime curvature, the temptation will be to say-- "aha! We've unified all the forces, we have plumbed the full depths of how reality works by identifying a single process that beats at its core, and governs the behavior of all its particles in all energy regimes." That's what they'll say, and it might be just right around the corner. But the next thing that will happen is a big letdown-- because people will soon realize that this "theory of everything" is indeed of purely academic interest, an intellectual game like Fermat's theorem that maximizes our sense of power and importance but alters our true power and importance only minimally. Yes, it is a story of tremendous beauty, so it was worthwhile-- symmetry and unification, unification and symmetry. The universe as a pearl, and it will all be our oyster to master. But... we'll still have all the same deep questions we had before-- the why is it this way questions, the why are we here questions, the is it better to be lucky than good questions, etc. And don't expect your dentist or your lawyer or your barber to quickly begin utilizing the theory of everything in their array of services, either.

So this would be my prediction of how it will go. Let's start back with Plato and the Greeks. As I understand it, Plato basically felt that the "perfect" world (and perhaps to him, the "true" reality) was the one we conceive in our minds. It was the world of perfect mathemetical concepts, and philosophical ideals (I'm guessing a bit here, true Plato experts are free to correct me). He then might have seen the world of our measurements and experiments as the imperfect projection of that perfect ideal onto the messy sloppy everyday world we have access to. Using that view, the Greeks conceived of a universe of pure beauty, symmetry, and unification of all things as four fundamental elements embedded in a greater whole.

Fast forward to Galileo, and the reversal of the Platonic ideal. Now it is the measurements themselves, the telescopic view, that has access to real truth, and the mathematical constructs like velocity and straight lines are the slave to what we observe. We no longer impose onto reality that it must follow what we imagine, instead we endeavor to imagine what reality has imposed. The measurement is everything, the theory follows. Call it the early-modern science approach.

Fast forward to Heisenberg and company. The Galilean approach takes an even more profound turn-- not only is the measurement the key to understanding reality, but the measurement can determine the reality. We've come completely 180 degrees from Plato-- reality awaits the messy sloppy apparatuses of the world of men and planets to tell it what to do. It still obeys beautiful mathematical principles and symmetries, but they serve only to tell us probabilities of what can happen-- reality still awaits our signal, or the signal of that sloppy material realm, to achieve self-actualization. I think that would have come of a bit of a surprise to Plato, who must have thought he had a handle on at least a key element of the essence of reality. Call it the intermediate-modern science approach.

Keep moving forward to Einstein. Now we have a hybrid view-- reality must expose itself to us via experiment, yet we can still successfully anticipate its glories if we adopt an axiomatic approach. Identify key unified principles, attach some symmetries that seem plausible, and voila-- a theory of reality that precedes the result of experiment. We are headed back toward Plato now! Call it the present-modern science approach, and with it the search for a theory of everything, applying string theories with higher dimensions and brane worlds and whatnot-- Plato would smile in his grave once again: "you see, it's as I told you-- the real world is one of pure beauty, symmetry, and unification. The human mind has cornered the reality, and it may take some time for our measurements of these phenomena to catch up." But just as it did before, new observations and a new Galileo will emerge that will once again shatter this view. Once again reality will prove to be sublimely more surprising and profound than we understood before. Once again, we will find that our philosophy failed to "dream of" all the things that are under heaven and Earth. And so the wheel will turn, over and over, until we realize that science is a process, not a destination-- we climb the hills we encounter, but like the bear, what we see from the top is... more hills.

[Tensor]

We've unified all the forces, we have plumbed the full depths of how reality works by identifying a single process that beats at its core, and governs the behavior of all its particles in all energy regimes." That's what they'll say, and it might be just right around the corner. But the next thing that will happen is a big letdown-- because people will soon realize that this "theory of everything" is indeed of purely academic interest, an intellectual game like Fermat's theorem that maximizes our sense of power and importance but alters our true power and importance only minimally. And don't expect your dentist or your lawyer or your barber to quickly begin utilizing the theory of everything in their array of services, either.

I can't disagree more with this. Only 50 years after the discovery of a completely unknown process and only 15 years after the putting the final touches, on only one portion of the explanation of that process, the citizens of Hiroshima and Nagasaki were certainly affected by the process that was only of academic interest.
While it may appear that items are of only academic interests, I can't think of any major (and a bunch of minor) discoveries that were initially of only academic interest that, in one way or another haven't found an applied side. Even GR, which, due to the math involved, is the epitome of academic interest has found applications.
If history is any guide, knowledge, even esoteric knowledge, will find an application somewhere, somehow.

[Ken G]

I can't disagree more with this. Only 50 years after the discovery of a completely unknown process and only 15 years after the putting the final touches, on only one portion of the explanation of that process, the citizens of Hiroshima and Nagasaki were certainly affected by the process that was only of academic interest.

No, that's not my point. The process you refer to occurs at an energy regime that is, tragically for those citizens, entirely accessible to science. I am not claiming that anything I can't anticipate a use for is necessarily academic, I'm saying that the extrapolation of science into energy regimes we can only conceive of, and then looking at the projection of those concepts onto our miniscule technological capabilities or what morsels we can extract from the Big Bang (which as a thermodynamic equilibrium, covers its tracks quite effectively), is an academic exercise that is of great aesthetic value but should not be taken too seriously as science progresses toward an ability to probe higher and higher energies and more and more profound mental constructs, just as it has always done. Indeed, I'm predicting that unification of QM and GR will have essentially nil tangible impact on humanity (other than possibly, and regrettably, the tangible impact of a philosophical clash for control of the minds of humans) for a very, very, very long time. Likely, not ever, as it's hard to predict how long we'll even be around (and the theory of everything won't extend our longevity a day.)

Even GR, which, due to the math involved, is the epitome of academic interest has found applications. If history is any guide, knowledge, even esoteric knowledge, will find an application somewhere, somehow.

But we are talking about the Planck regime, are we not? No GPS satellites will include programming from the "theory of everything", I am confident. Still, I can't deny that it is hard to always anticipate what new outlooks may come from a new theory, and those new outlooks may lead to the next discovery that does have applications. I'm not against unification, I just don't believe it will have much to do with the "truth" of reality, and more to the point, science really can't claim that it does in any way that transcends science itself. The relevance of this point emerges as scientific descriptions deviate farther and farther from the familiar nonscientific descriptions used for things that continue to rest well beyond our observational capabilities. (Take, for example, the very hubris of the phrase "theory of everything"-- does that not make my point?)

[Tensor]

No, that's not my point. The process you refer to occurs at an energy regime that is, tragically for those citizens, entirely accessible to science.

But that energy regime I pointed out for 1945 wasn't accessible at the time of the original discovery in the late 1800s and the early moves toward a theory of how it worked, was it?

I am not claiming that anything I can't anticipate a use for is necessarily academic, I'm saying that the extrapolation of science into energy regimes we can only conceive of,

But that was my point. The energy regimes that were used in 1945 and currently in use were not thought of when radioactivity was first discovered.

and then looking at the projection of those concepts onto our miniscule technological capabilities or what morsels we can extract from the Big Bang (which as a thermodynamic equilibrium, covers its tracks quite effectively), is an academic exercise that is of great aesthetic value but should not be taken too seriously as science progresses toward an ability to probe higher and higher energies and more and more profound mental constructs, just as it has always done.

In this (and the previous post) you make it sound as if no applications will ever come out of a discovery of a TOE. The projection of the concepts of radioactivity were useless to the miniscule technological capabilities of the time, in terms of applications, when it was discovered.

But we are talking about the Planck regime, are we not? No GPS satellites will include programming from the "theory of everything", I assure you.

Yeah, and were talking about satelites in orbit, being used to locate things on the earth. Not exactly something that was anticipated when GR first came out. Of course GPS won't use anything from TOE, GR handles that. But what will come out of it? We won't know until a TOE is found. And our previous history indicates that some application will come out of it. It won't stay an interest of only academics.

[Ken G]

But that energy regime I pointed out for 1945 wasn't accessible at the time of the original discovery in the late 1800s and the early moves toward a theory of how it worked, was it?

Yes, it was-- Uranium decay. The energies involved were discovered right along with radioactivity. What they didn't know is that there was a critical mass where a runaway occurs.

In this (and the previous post) you make it sound as if no applications will ever come out of a discovery of a TOE.

I've certainly never heard one suggested from anyone but a crank, and I continue to expect that this is almost certainly true. Let's just say that I'm not buying stock in a company that is R&Ding a TOE, unless I'm getting the book rights.

Of course GPS won't use anything from TOE, GR handles that.

Exactly, GR handles it-- with its inconsistencies with QM and all. Why? Because those inconsistencies don't matter at all to anything that humans have access to. I'm sure neither of us think scientists will really be making mini black holes or universes in other dimensions as a result of this exercise. It's scientific hubris taken to the level of science fiction.

But what will come out of it? We won't know until a TOE is found. And our previous history indicates that some application will come out of it. It won't stay an interest of only academics.

Then I hope to see it, because my money is still on the "big letdown" interpretation of the TOE!

[jlhredshift]

But what will come out of it? We won't know until a TOE is found. And our previous history indicates that some application will come out of it. It won't stay an interest of only academics.

Technology will come of it; spinoffs, unknown and unforeseen.

Then I hope to see it, because my money is still on the "big letdown" interpretation of the TOE!

No, unification will not be the end, it will just cause more questions.

It is not about the destination, it is about the journey. Unification and the search for it is only a psychological goal that gives us a target to shoot at.

I also suggest that as we do discover more and more at smaller and smaller scales that the goal post of how a TOE would be defined will keep shifting.

[Grey]

Now, try to come up with a model that can accurately describe and predict not only that, but the quantization of atomic weights, the pattern of the periodic table and chemicacl interactions, the quantization of charge on ions, various types of nuclear decay, as well as fission and fusion, the results of scattering experiments, including ones where we've used opposed beams of protons and anti-protons (or at least, what we think are protons and anti-protons ), and a whole host of other phenomena. Frankly, I don't think it would be possible.

The reason that I think it would be possible is because in my opinion, reality does it.

I think that's assuming your conclusion. That is, you're claiming that it must be possible to create such a description without anything like protons because nature itself does so. But that's only the case if nature does not, in fact, include anything like what we call protons, which is the very question under contention.

Take the entire particle concept. Wave functions showed us that particles are not what we think they are, they don't actually have "existence" in the normal sense, instead they have more to do with the way measurements get done, and yield quantized results.

Sorry, no good. Even in a quantum wave mechanic description, there are still entities which are quite recognizable as fundamental particles, even if they now have properties rather different than we once thought. So, if we take it at face value, we've learned more about the details, but we have not invalidated the earlier broad description.

So we have a new word, "quantum". We still use "particle" because it's a useful concept for organizing these quantized results of measurements, but we now have a much more sophisticated understanding of the limitations of that word. I would say, the limitations are to the point that reality itself must use a different concept altogether, one without those limitations. That's my basic point-- we generate concepts that work, but every concept has limitations. Reality is not allowed to have those limitations, it has to work perfectly because it is what it is.

This would again seem to be a somewhat weaker statement than your original one. Something like "our models and theories probably don't have all the details correct, and may also include concepts that may be completely wrong". That's again a statement that I would quickly agree with. But that statement is a long way from "our models obvious don't correspond with reality at all".

Perhaps a statement that we both agree on is that science is the study of the projection of reality onto our ability to make measurements and observations of reality. We must fit reality into boxes that make sense to us, and of course in the process we leave bits out because they don't fit. What you are saying is that as the boxes get better and better and more and more useful, we asymptotically approach a complete understanding of what reality is. Or, if our intellect never reaches convergence on that "full" truth, at least what we do is bringing us closer.

I can agree with that, and I think it's a reasonable characterization of my position.

What I'm saying is, the very nature of the boxes, the whole basis of this approach, is guaranteed to leave out very important elements of reality. Those elements simply aren't addressible this way, just as the act of projecting a shadow on the wall will always leave out elements of the object being shadowed.

Why are you guaranteed to leave out important elements of reality? Why is the process analogous to casting a shadow on the wall?

It's an inherent part of science, and leaves us with limitations we should study and understand, rather than overextend our concept of truth into realms that science is ill-equipped for. Like the Planck realm, or the origin of the Big Bang, or what happens when a wave function is collapsed, or how intelligence can be used to study itself (sticking to scientific frontier topics-- obviously the realm of art and aesthetics is another one.) Put succinctly: when the observation is impossible, then so is the science.

Except that, as our experimental capabilities expand, many of these realms that we thought were unobservable become so. For example, there's some speculation that the LHC will have sufficient energy to create tiny black holes during operation. If that actually happens, it probably is going to take a quantum theory of gravity to fully describe what's going on (and the observations themselves will constrain such a theory). It makes perfect sense to have already started to explore what features such a theory might need to have and some of the ways it might work even before the observations are made, rather than insisting on waiting for the observations first. Similarly, just as basic particle physics can make testable predictions about the big bang by looking at element abundances, a theory that merges general relativity and quantum theory may very well make testable predictions about our universe from it's description of what happen when the universe was very young. We won't know whether or not it makes testable predictions until we work it out, and it seems strange to decide ahead of time that it won't so that we shouldn't even make the attempt.

Fair enough, but I'd still like your opinion on the two examples I've mentioned so far, the geocentric universe and Newton's gravity. Note that both of those could certainly be described as "so incredibly accurate and useful" by the measurement standards of their day. Would you then also say that there really are elements of reality that correspond to them? A lot is hidden in the word "elements", and again we come against semantics. Most likely, there is much to be said for the points that we are both making, so the real issue is, what are the various valid ways to interpret the words we're using such that our arguments have something important to say?

Well, I'm not sure the geocentric model was ever all that accurate. And I've commented on Newton already. It's true that we think of the mechanism as something completely different, but there are still some elements that remain the same. Objects still have mass, and that mass acts to change the paths of other objects. In general relativity, it does that by changing the shape of space rather than directly affecting the object, but there are still some common elements. And I agreed that forces are more mysterious than the components of matter.

But you're right, that part of the question may lie in what we mean exactly by a theory having "elements that correspond to reality". So I want to ask a couple questions to determine just how far you'd be inclined to push this. For example, I also think that things like stars, planets, galaxies, and so forth are "real". We may not know everything about how stars work, but I think any meaningful description of the universe will include something that corresponds to "star", and that it makes sense to think of a star as a real thing. What are your thoughts on that? Are stars real, or are they just a concept created by humans that obviously has nothing to do with reality itself?

Perhaps one day this hypothesis will receive its test. Not sure that would be such a good thing for the advancement of humanity's knowledge, or humanity itself for that matter!

Well, we're probably safe from such a test for the forseeable future at least. It would be very interesting, though, to compare notes with someone who had a completely different view on the universe.

The way I would have phrased this is, it's clear that whatever element of reality corresponds to what we now conceptualize as a "proton" is really strange, and is clearly not like the macroscopic things that we use concepts like "trajectory" on.

Certainly true.

But just because measurements achieve quantized results does not prove that reality itself is "made" of quanta-- it just means that if you do a particle number measurement, you get a quantized result, just as if you do an angular momentum measurement, you get a quantized result. Reality can't be bothered by that, because it has to allow the possibilities for different numbers of particles in a given box, in high energy conditions.

It's true that we have to consider virtual particles in our calculations. But even then, we don't have to consider half-protons, or even virtual half-protons. Certain types of measurements always seem to be quantized. Is that just some weird quirk, or is it because certain types of things really are quantized? I don't know for sure, but I think it would be a mistake to simply assume the former.

Science never offers a solution to the problem that the more you know, the more you don't know.

In principle, of course, that might not be true. The amount of potential knowledge about the universe might be finite, but we're so far away from that limit, it seems infinite. Still, you might be quite correct here. But I'm not sure that you can extend that to mean that even the things we know (or think we know) must not really be true.

[Spaceman Spiff]

Exactly, GR handles it-- with its inconsistencies with QM and all. Why? Because those inconsistencies don't matter at all to anything that humans have access to. I'm sure neither of us think scientists will really be making mini black holes or universes in other dimensions as a result of this exercise. It's scientific hubris taken to the level of science fiction.

Then I hope to see it, because my money is still on the "big letdown" interpretation of the TOE!

Putting aside the loaded moniker, "TOE", I suggest that the next revolution in physics is likely to come from understanding the relationship (whatever that is, and it needn't necessarily be "unification") between our two most successful physical theories, GR and QM. Their grating mismatch is screaming at us that something is wrong. And I will bet the bank that the "spinoffs" from that new understanding (scientific, technological, philosophical, ...al) will not be limited to understanding the world at the Planck scale.

Somewhat over 100 years ago, all kinds of people were proclaiming that Physics had come to an end. All that was left were the minute details. And yet staring everyone in the face were the several important holes - (a) what did the electromagnetic wave propagate through? (b) the ultraviolet disaster, to name just two. And then what happened? And who then would have predicted what would happen in ~ half a century's time?

[Ken G]

Why are you guaranteed to leave out important elements of reality? Why is the process analogous to casting a shadow on the wall?

Because of inherent limitations in the scientific approach. Examples might include origins (of the Big Bang) that have no apparent cause within a closed description of the universe (and scientific models must be closed to be complete-- if brane worlds collide to give a Big Bang, that's again "more hills".) Also, we have no theory for the outcome of a measurement, other than statistical. We have proofs within our own system of physics that no scientific theory can ever explain the outcome of a quantum mechanical measurement, yet reality has to have some way to determine that outcome. I believe that is just what Einstein meant when he said "God doesn't roll dice", but he thought that meant we should keep looking for a better theory. I'm saying, he should have accepted that no such theory can exist. Also, we have the fundamental self-referential conundrum that we can try to use science to understand the workings of our own mind, but ultimately we will have to apply those very same workings to understand the answers we get. We'll never be able to use science to break out of that trap, because we'll never know how we are affecting the question by being the one asking it. And those are all science topics-- I'm sure I don't need to tell you that within the human condition we find value and meaning in ways that are inherently subjective, and science can never go there. Identifying mechanisms that produce sentience is not at all the same as experiencing sentience, so that's another box that will always leave something very important, if not paramountly important, out.

If that actually happens, it probably is going to take a quantum theory of gravity to fully describe what's going on (and the observations themselves will constrain such a theory).

But not at the Planck domain. If the only inconsistencies between GR and QM come at the Planck level, we will not be able to probe them with the LHC. The speculations you refer to must be at some far lesser energy level, which would instead point to a problem with the existing applications of GR and QM. It is certainly possible that seeking consistency at the Planck level will lead to a better theory at the level we do observe, following the faith that the universe has some deep order that we missed the first time around. But there's currently no evidence that there is such a missing piece, and no reason to expect that unification at the Planck domain would solve it if there was. Perhaps my understanding of the issues is off target, I am by no means a particle physicist, but this is my take on it.

It makes perfect sense to have already started to explore what features such a theory might need to have and some of the ways it might work even before the observations are made, rather than insisting on waiting for the observations first.

Indeed, as Einstein did. But Einstein's theory is easily tested by solar eclipses, etc., whereas a theory at the Planck domain will never be testable at that level. You'd have to have some faith that unifying at that unobservable level will project in a measurable way onto what we do observe, as in the above example. There may be nothing wrong with that faith, it's certainly one way to go to look for new theories, but it is quite far from being relevant to the Planck domain. I agree with jlhredshift there.

It's not so much that I think unification is a fool's errand, it's that I think it has way oversold expectations. Remember that when Newton's and Maxwell's equations were inconsistent at high speed, we had access to those speeds, to within a few orders of magnitude. Michelson easily conceived of an experiment that would give measurably different answers in the two theories. Imagine instead that we lived in a universe where no speed we ever encountered exceeded 1 cm/s. Would we still have a way to address the inconsistencies between Newton and Maxwell, or would it all just be speculation? To my knowledge, no one has at present suggested a measurable experiment that can distinguish the predictions of GR and QM, such as measuring quanta of gravitational radiation. It's just a guess that new energy levels will reveal new physics, which ironically falls into the argument that I and jlhredshift are suggesting.

Similarly, just as basic particle physics can make testable predictions about the big bang by looking at element abundances, a theory that merges general relativity and quantum theory may very well make testable predictions about our universe from it's description of what happen when the universe was very young.

There's a big difference-- the compositions you are talking about "froze out" at energies that we do have access to, and do observe in our universe. A theory that constrains the Planck domain will never advance beyond the level of consistency and plausibility, not testability. That's my issue with it, I see it as extrapolating science in a way that should not be taken too seriously in terms of the reality. Not taking it too seriously doesn't mean a very pretty and wonderful theory cannot emerge.

We won't know whether or not it makes testable predictions until we work it out, and it seems strange to decide ahead of time that it won't so that we shouldn't even make the attempt.

My basis for this is that we know how scales work. We understand the scale at which a theory must be tested, even if we don't know exactly what to test. There's no way to test relativity, for example, without involving a velocity scale that in some way approaches within a measurable reach of c.

Well, I'm not sure the geocentric model was ever all that accurate.

It agreed with their measurements until Tycho realized that it was possible to learn something by pushing the accuracy further. Indeed, the Copernican model was less accurate than Ptolemy's, because Copernicus used circular orbits. Nevertheless, it was the telescope that killed Ptolemy, not the more accurate measurements of Tycho. The reality was seen to be different than what was thought.

It's true that we think of the mechanism as something completely different, but there are still some elements that remain the same.

Is that because those elements were real, or just that their original usefulness has not vanished? Can we cherry-pick our reality from our theories, or just be honest that our theories weren't the reality? A semantic distinction, perhaps, but both views have value to consider.

Objects still have mass, and that mass acts to change the paths of other objects.

Sort of. But in relativity, it is possible to eliminate the concept of gravitational mass in favor of energy. Newton would never have dreamed that.

And I agreed that forces are more mysterious than the components of matter.

It seems to me the components are pretty mysterious too, if Higgs particles give them mass, or if they are vibrations on strings. (Where by "are" I of course mean, "may be frutifully thougt of as".)

For example, I also think that things like stars, planets, galaxies, and so forth are "real". We may not know everything about how stars work, but I think any meaningful description of the universe will include something that corresponds to "star", and that it makes sense to think of a star as a real thing. What are your thoughts on that?

My thoughts are that the concept of "star" is like the concept of "temperature". Get enough detailed stuff together, and a more statistical ensemble approach becomes better than keeping track of all the details. But this is just what I mean about the difference between our concepts and reality. The more macroscopic the system, the easier we can conceive of it in simple ways, but "reality" doesn't have that luxury. It has to really keep track of everything that is going on! So on that basis, I would say that reality itself has no concept of "star", it's just a higher density region (whatever density is) that has different physics coming to the fore.

Well, we're probably safe from such a test for the forseeable future at least. It would be very interesting, though, to compare notes with someone who had a completely different view on the universe.

Yes, then we'd both get our socks knocked off, and could look back on this discussion with much greater insights. In the mean time, we muddle along wondering! (Yes, I wonder too-- I guess I'll have to retract "overwhelmingly obvious"!)

It's true that we have to consider virtual particles in our calculations. But even then, we don't have to consider half-protons, or even virtual half-protons.

True, but I'd say that means quantization is a useful concept, but not that reality has particles. The basic problem is, if we dig deeply into the very definition of the word "particle", we'll quickly reach all kinds of problems. It will end up being a definition like "particle is whatever concept scientists are using to understand quantization", but that's circular-- reality is defined to be the way we understand it, and the justification is that the definition works. Again I say that science defines truth and usefulness to be the same thing, so cannot really talk about reality in any way that transcends its own definitions. The definitions are those boxes we were discussing-- they give (effectiveness), and they take away (they shave something off). It's like language-- we know that language helps us communicate, but do we ever consider how language limits our communication? That's the curse of the poet-- to express love and anguish with words. Do they ever really think they've accomplished it?

Certain types of measurements always seem to be quantized. Is that just some weird quirk, or is it because certain types of things really are quantized? I don't know for sure, but I think it would be a mistake to simply assume the former.

I think the issue is different if you are talking about before the measurement vs. after the measurement. If you mean they are quantized after the measurement, I agree, but adopting the approach of doing measurements means you are already within the realm of science. Whatever was left out of the box has already happened at that point, you are talking about features of the shadow at that point. That's again the strength and limitation of science in a nutshell-- measurements are found to be objective and quantitative, so science applies to them in a powerful way. But where comes the expectation that all of reality can contain only elements that are objective and quantitative? That's just a prejudice of science. A useful prejudice, yes. But we don't even know why science works, so it is entirely possible (and I feel, quite likely) that the "guts" of reality are not built with concepts that are quantifiable (reality doesn't have numbers, we do) and objective (reality is only one thing, so there's no distinction between objective and subjective) way.

Just think of the unbelievable overhead if reality really had to "make" the same calculations that we do to try and predict it. We get to cherry pick what we want to know, reality would have to do everything all the time. I think reality finds an easier way, but it's not accessible to our intelligence because of what intelligence inherently is.

In principle, of course, that might not be true. The amount of potential knowledge about the universe might be finite, but we're so far away from that limit, it seems infinite.

That is also an interesting question. I think it's infinite, in the same sense that the number of irrational numbers between any two integers is infinite. Science must use decimal expansions of various specified precisions, but reality is fine just with "length". It is what it is.

Still, you might be quite correct here. But I'm not sure that you can extend that to mean that even the things we know (or think we know) must not really be true.

I can't, you're right. The best I can really say is "maybe this is how it is". So I'm not making an argument to prove I'm right, merely to ask: what does it mean for science, and the hubris of a "theory of everything", if I am?

[Spaceman Spiff]

Sean Carroll (the astrophysicist) has an article over at Cosmic Variance with a really nice set of slides on the nature and development of science, which have some application to a branch of the discussion going on here.

[gzhpcu]

Nice links.

So in science, like many other things in life, the less we know, the more we think we know, and vice versa.

[Grey]

Gah. This post was too long, but I've already spent way too much time going over it to try to edit it for length. So it's just going to have to get broken into two pieces, and I'll try to be more concise next time.

Because of inherent limitations in the scientific approach. Examples might include origins (of the Big Bang) that have no apparent cause within a closed description of the universe (and scientific models must be closed to be complete-- if brane worlds collide to give a Big Bang, that's again "more hills".)

I'm not sure that I accept that. Even if we can't explain certain things, it does not follow logically that we must have left elements out of the things we have explained. And who said theories had to be complete in order to reflect reality? Suppose that I'm right, and matter is made up of things that correspond to what we call protons and electrons and so forth (even though they are pretty weird things, and our descriptions of them probably are not perfect). Our theories don't say anything about why that's so, or where they ultimately came from, or what will ultimately happen to them. Yet even without that, they still would accurately reflect that element of reality.

Also, we have no theory for the outcome of a measurement, other than statistical. We have proofs within our own system of physics that no scientific theory can ever explain the outcome of a quantum mechanical measurement, yet reality has to have some way to determine that outcome. I believe that is just what Einstein meant when he said "God doesn't roll dice", but he thought that meant we should keep looking for a better theory. I'm saying, he should have accepted that no such theory can exist.

Wait a minute, why does reality have to be deterministic? You're assuming that there must be a reason why outcome A happened and not outcome B, but I think taht's assuming more about reality than we know. Why would it be a problem for reality to have random behavior at a fundamental level? If you're convinced that our existing models must not reflect reality, it seems incongruous for you to make assumptions about what reality must actually be like.

Also, we have the fundamental self-referential conundrum that we can try to use science to understand the workings of our own mind, but ultimately we will have to apply those very same workings to understand the answers we get. We'll never be able to use science to break out of that trap, because we'll never know how we are affecting the question by being the one asking it.

Self reference need not result in paradox and impossibility. You're assuming at the outset that no system can describe itself, but that's simply not true.

And those are all science topics-- I'm sure I don't need to tell you that within the human condition we find value and meaning in ways that are inherently subjective, and science can never go there. Identifying mechanisms that produce sentience is not at all the same as experiencing sentience, so that's another box that will always leave something very important, if not paramountly important, out.

I'm not certain that being able to experience somethign is a necessary component of being able to describe it well, or for that description to correspond to what it's really like. Do I have to be an atom for my description of an atom to be accurate?

But not at the Planck domain. If the only inconsistencies between GR and QM come at the Planck level, we will not be able to probe them with the LHC. The speculations you refer to must be at some far lesser energy level, which would instead point to a problem with the existing applications of GR and QM. It is certainly possible that seeking consistency at the Planck level will lead to a better theory at the level we do observe, following the faith that the universe has some deep order that we missed the first time around. But there's currently no evidence that there is such a missing piece, and no reason to expect that unification at the Planck domain would solve it if there was. Perhaps my understanding of the issues is off target, I am by no means a particle physicist, but this is my take on it.

I'd say that we're quite certain that our understanding breaks down at the Planck scale, but that doesn't mean that explorations at lower energy scales might not show us discrepancies between general relativity and quantum theory. If we're creating microscopic black holes, that would definitely be a realm where we'd really need both theories to be working together. But as far as that goes, I'd suggest that the Planck scale is not forever beyond our grasp, as you seem to be suggesting. The Planck energy is large on the scale of particles, but it's still a very small amount of energy. It's not a matter of having a huge amount of energy, it's a matter of getting it into a small region of space. Yes, the only time we know of when things were at or near that eneryg density was the very early universe, but at that point, the entire universe was at that energy density. We'll certianly never achieve that, but that does not mean that we won't be able to bring a much smaller part of the universe up to that level. I'm not suggesting anytime soon, but it wouldn't much surprise me if we managed it within the next millennium or so. Not that I'll be here to see it, most likely.

Indeed, as Einstein did. But Einstein's theory is easily tested by solar eclipses, etc., whereas a theory at the Planck domain will never be testable at that level. You'd have to have some faith that unifying at that unobservable level will project in a measurable way onto what we do observe, as in the above example. There may be nothing wrong with that faith, it's certainly one way to go to look for new theories, but it is quite far from being relevant to the Planck domain. I agree with jlhredshift there.

Well, we certainly won't know for sure whether such a theory would lead to testable predictions without working it out. But for me, it actually seems like a reasonably interesting avenue of investigation. That is, if you know that two very successful theories clash, that immediately gives you some clues as to where those theories might need to be refined. It's not a guarantee, but it's as good a place to look for new physics as any.

It's not so much that I think unification is a fool's errand, it's that I think it has way oversold expectations. Remember that when Newton's and Maxwell's equations were inconsistent at high speed, we had access to those speeds, to within a few orders of magnitude. Michelson easily conceived of an experiment that would give measurably different answers in the two theories. Imagine instead that we lived in a universe where no speed we ever encountered exceeded 1 cm/s. Would we still have a way to address the inconsistencies between Newton and Maxwell, or would it all just be speculation? To my knowledge, no one has at present suggested a measurable experiment that can distinguish the predictions of GR and QM, such as measuring quanta of gravitational radiation. It's just a guess that new energy levels will reveal new physics, which ironically falls into the argument that I and jlhredshift are suggesting.

Actually, here's a pretty interesting paper that explores just that, the possibility of detecting individual gravitons. Their conclusion is that it would be theoretically possible, but would require a detector on the order of the mass of Jupiter, which would then have the problem of structural failure because of self-gravitation. And there are issues of noise and so forth. So the conclusion is that it is probably impossible form a practical standpoint. So the question has actually been considered, and you're probably at least correct that there will be no such experiments performed in the forseeable future. But we've often found new physics when moving to higher energy levels, and certainly unifying the other forces have led to some impressive developments in physics, so again, it seems like as good a path as any to explore.

There's a big difference-- the compositions you are talking about "froze out" at energies that we do have access to, and do observe in our universe. A theory that constrains the Planck domain will never advance beyond the level of consistency and plausibility, not testability. That's my issue with it, I see it as extrapolating science in a way that should not be taken too seriously in terms of the reality. Not taking it too seriously doesn't mean a very pretty and wonderful theory cannot emerge.

Perhaps, but there are plenty of domains where consistency with observation is all we've got, and we can't directly perform experiments. A really large portion of astronomy and cosmology fall into the category. I'll agree that when you can't control the experiments that you're always less certain about a theory.

My basis for this is that we know how scales work. We understand the scale at which a theory must be tested, even if we don't know exactly what to test. There's no way to test relativity, for example, without involving a velocity scale that in some way approaches within a measurable reach of c.

I'm not sure that's the case. It just changes how accurate your measurements have to be, and we've always been pretty ingenious coming up with ways to measure things. Test of relativity don't require us to accelerate objects to near light speed, or create extraordinarily strong gravitational fields, yet we're still able to perform them.

It agreed with their measurements until Tycho realized that it was possible to learn something by pushing the accuracy further. Indeed, the Copernican model was less accurate than Ptolemy's, because Copernicus used circular orbits. Nevertheless, it was the telescope that killed Ptolemy, not the more accurate measurements of Tycho. The reality was seen to be different than what was thought.

But ptolemy's system was strictly descriptive. That is, you could work out a plan for a well-observed orbit, but there was no way you could use that model to predict a new object based on existing orbits, or even map out an entire orbit from a handful of observations when something new was observed. Critics of science always like to hold up Ptolemy's model as a great example of how science can seem to be working beautifully and still be hopelessly wrong, but frankly, it was never a very good model.

Is that because those elements were real, or just that their original usefulness has not vanished? Can we cherry-pick our reality from our theories, or just be honest that our theories weren't the reality? A semantic distinction, perhaps, but both views have value to consider.

I don't know the answer to that. Is it true that there have been theories that we no longer think are accurate representations of reality? Absolutely. Should we therefore conclude that it is impossible for a theory to be an accurate representation of reality? That does not logically follow.

Sort of. But in relativity, it is possible to eliminate the concept of gravitational mass in favor of energy. Newton would never have dreamed that.

I wouldn't say "eliminate". I'd just say that under general relativity, mass and energy are more or less the same thing.

It seems to me the components are pretty mysterious too, if Higgs particles give them mass, or if they are vibrations on strings. (Where by "are" I of course mean, "may be frutifully thougt of as".)

There is certainly much we don't know. But even if we don't understand the details of the components themselves, that doesn't mean that it's not meaningful to talk about objects as being made of those components. I can describe a house as being made of bricks, and describe some of the external properties of those bricks, even if I don't understand the details of how bricks are made or their exact composition.

My thoughts are that the concept of "star" is like the concept of "temperature". Get enough detailed stuff together, and a more statistical ensemble approach becomes better than keeping track of all the details. But this is just what I mean about the difference between our concepts and reality. The more macroscopic the system, the easier we can conceive of it in simple ways, but "reality" doesn't have that luxury. It has to really keep track of everything that is going on! So on that basis, I would say that reality itself has no concept of "star", it's just a higher density region (whatever density is) that has different physics coming to the fore.

Wow, you're like the ultimate reductionist! So to you, it doesn't make sense to ascribe any "real" significance to something if it can instead be described at a lower level as being composed of something? I don't agree with that at all. Certainly reality doesn't have a "concept" of a star, because reality doesn't have concepts, But I really think that in any description of the universe, I'd be able to pick out elements that corresponded to our stars. Heck, even just saying a higher density region where certain physical processes are taking place calls it out as something distinct from, say, the intergalactic medium. Or, to use our alien example, I was already willing to wager that any alien's description of the universe would include something analagous to our subatomic particles. But I'd be willing to bet a good deal more that any such description would include elements that correspond to stars and galaxies.

Yes, then we'd both get our socks knocked off, and could look back on this discussion with much greater insights. In the mean time, we muddle along wondering! (Yes, I wonder too-- I guess I'll have to retract "overwhelmingly obvious"!)

Hey, that's all I was hoping for! I'm certainly not trying to defend the opposite stance (that we can somehow be sure that our theories reflect reality). So if you're willing to come join me in the middle ground (maybe our theories could reflect reality), that's good enough for me, even if you remain on the "well, maybe, but probably not" side of the middle.

[Grey]

True, but I'd say that means quantization is a useful concept, but not that reality has particles. The basic problem is, if we dig deeply into the very definition of the word "particle", we'll quickly reach all kinds of problems. It will end up being a definition like "particle is whatever concept scientists are using to understand quantization", but that's circular...--

I don't think that's actually circular. Or perhaps it's just that a better statement might be "reality has elements of quantization", rather than "reality is made up of particles". The point I'm making is, if any description of reality must include certain specific types of quantization in order to be accurate, it doesn't make much sense to me to say "well, reality is surely not really quantized in any way, it just behaves as though it were". Now, are we absolutely certain that it's impossible to correctly describe what we see without recourse to quantization. No, of course not. But it sure is looking that way.

Again I say that science defines truth and usefulness to be the same thing, so cannot really talk about reality in any way that transcends its own definitions. The definitions are those boxes we were discussing-- they give (effectiveness), and they take away (they shave something off).

When you bring this up, I wonder if we may simply disagree about how close something has to be to be considered correct. If I discover that protons seem to be made of quarks, does that mean I was wrong about there being such things as protons? If I later discover that quarks are vibrating strings, does that mean I was wrong about the existence of protons. In both cases, I'd say no. I would quickly agree that my "box" (my understanding or my model of a proton) has changed shape or size. I've added new things and maybe taken away other things. But it looks like you're saying no, that if my description isn't absolutely perfect in all ways, then in fact it doesn't represent reality at all. I see that same attitude above, when you're talking about whether stars exist, saying that reality has to keep track of what's really going on.

It's like language-- we know that language helps us communicate, but do we ever consider how language limits our communication? That's the curse of the poet-- to express love and anguish with words. Do they ever really think they've accomplished it?

We do consider that, actually. What you've stated here is essentially the Sapir-Whorf hypothesis, and linguists and anthropologists have been debating the idea for years.

Just think of the unbelievable overhead if reality really had to "make" the same calculations that we do to try and predict it. We get to cherry pick what we want to know, reality would have to do everything all the time. I think reality finds an easier way, but it's not accessible to our intelligence because of what intelligence inherently is.

I thik I reject your premise here. That is, I'd claim that (say) a mathematical description of an object's trajectory (possibly as affected by various forces, curvature of space, or whatever) could be an accurate description of that trajectory, even if there aren't pixies quickly running through all the same calculations as we are to figure out where everything should be. Yes, the description of some force is not the force itself, but that does not mean that it cannot be an accurate description.

That is also an interesting question. I think it's infinite, in the same sense that the number of irrational numbers between any two integers is infinite.

I have no idea myself, either.

Science must use decimal expansions of various specified precisions, but reality is fine just with "length". It is what it is.

Actually, I'm not sure that you can be sure of even that. For example, some people have speculated that space is quantized at the scale of the Planck length. If that were the case, then there's a limit to the precision needed in any distance measurement (the best you could do would be to count it off in Planck lengths).

I can't, you're right. The best I can really say is "maybe this is how it is". So I'm not making an argument to prove I'm right, merely to ask: what does it mean for science, and the hubris of a "theory of everything", if I am?

Excellent! Because I'm not trying to claim that our theories necessarily do represent reality. As far as what it means, I'd think that we should therefore be cautious about whether we assume theoretical constructs are real. However, I'm not sure I see as much "hubris"* in a "theory of everything". Maybe it's the name. I don't really think that anyone expects that if we can come up with a workable theory of quantum gravity that we'll suddenly understand everything. Maybe in some of the popular accounts, but I think physicists know that even a successful unification of the two could still be wrong, or at the very least still need to modified as more observations are made. Will we reach energy scales at the Planck level in the near future? Probably not. Will we reach energy scales that might require revisions to our current theories? Almost certainly (at the very least, we'll probably need to add new particles into the mix). Might an attempt at resolving the conflict between the two most successful theories give us some ideas about how to go about modifying those theories? I think yes.

* I took an excellent course in Greek literature at one point, where we had an extensive discussion of what hubris was and was not. The short summary is that it was absolutely not as simple a concept as "pride" or "arrogance", and the professor despaired of high school students everywhere being given a far too simplistic view of the matter by teachers who didn't really understand the material. I've tended to avoid the term since then...

[Ken G]

Even if we can't explain certain things, it does not follow logically that we must have left elements out of the things we have explained.

That's true, and there's an important distinction emerging here. An incomplete theory can still reflect reality as far as it goes, so there are two separate ways that the scientific approach might fail to capture the essence of the reality, and I've argued that both are in play. One is that the theories are just always toy models, and you have disputed just what that really means (the semantics are very tough here), and the other is that the theories must necessarily miss part of the story. Call the two issues errors of comission, and errors of omission, in each theory! My example of not being able to use science to see the origin of the universe (since science is causal and every cause requires some other cause) was an example of the error of omission of the Big Bang model, and I agree with you that this is not really an "error", the Big Bang is just moot on that point. But still, I think it is important to recognize that if science is moot on some issues that relate to reality, then we'd better start paying very careful attention to all the possible elements of reality about which science might be moot, before we go around boasting about "theories of everything".

Suppose that I'm right, and matter is made up of things that correspond to what we call protons and electrons and so forth (even though they are pretty weird things, and our descriptions of them probably are not perfect). Our theories don't say anything about why that's so, or where they ultimately came from, or what will ultimately happen to them. Yet even without that, they still would accurately reflect that element of reality.

Yes, we'd only be looking at errors of comission in this regard, not errors of omission, because those concepts aren't responsible for what they leave out. How do these concepts lead us astray? How do we refine them, with string theories and the like? Remember, we must not be talking about the usefulness of the concepts, as that is clear, we must be talking about whether or not they are "real" in a way that transcends scientific utility. I suspect that when it is understood how particles "tick", we will see the entire particle picture as a toy model, but that remains to be seen.

Wait a minute, why does reality have to be deterministic? You're assuming that there must be a reason why outcome A happened and not outcome B, but I think taht's assuming more about reality than we know.

I'm actually not saying that it has to be deterministic, I'm saying that it must have some way to achieve a random result, that we know nothing about at all. Quantum mechanics includes no known mechanism which explains how reality settles on a result-- we can only speak about the probable outcomes. The theory includes dice that have no corresponding mechanism in reality. An error of omission again-- it can't be the whole reality until you know how reality generates random results in quantum measurements (we know how it does it in macroscopic events like rolling a real die-- nonlinear dynamics and sensitivity to initial conditions. But those are fuzzy macro concepts, not good enough for reality. What is an initial condition, anyway? With what precision is it set "in reality?")

Why would it be a problem for reality to have random behavior at a fundamental level? If you're convinced that our existing models must not reflect reality, it seems incongruous for you to make assumptions about what reality must actually be like.

I would say that it is never enough for science to say "reality just does it, that's what we see", because you could say that at square one and never do science at all. The whole point of science is to find the reasons why things happen, so if a random result occurs, then how did it come to pass? I'm not saying predict the result, that would not be random, I just mean understanding the mechanism, like understanding what happens to a coin when you flip it. Without that part, we can never claim to have probed the depths of reality.

Let me put it another way. If I were Douglas Adams or some such, I'd invent a silly story where a great scientist built a robot that was capable of understanding everything that happens. People would pay money to step up to the robot and ask questions like, what will happen if I throw a tennis ball in the air at such and such a speed and such and such an angle? The robot proceeds to pick up a ball, execute just that maneuver, and say "that's what. That'll be one thousand dollars please." We can all see that this isn't science. I feel that by its very nature, science is about replacing reality with something else, something that fits in our heads and performs in a similar manner to reality, to some desired level of precision. So when I say that science cannot capture the essence of reality, I don't mean it as a criticism-- I think this is the strength of science, not its weakness. However, it is a limitation.

Self reference need not result in paradox and impossibility. You're assuming at the outset that no system can describe itself, but that's simply not true.

I don't know the precise requirements on a system that can describe itself in the manner you suggest, but I'm sure Godel's theorem lends insight there. I would expect that something as complex as intelligence could not describe itself. Intelligence would always be able to "surprise" itself, as the very act of learning what intelligence is will embody new capabilities that will then need to be described as part of intelligence. But that's just my sense, I certainly can't prove it! (Maybe Penrose's "The Emperor's New Mind" goes much farther down that path, that was a tough read though.)

I'm not certain that being able to experience somethign is a necessary component of being able to describe it well, or for that description to correspond to what it's really like. Do I have to be an atom for my description of an atom to be accurate?

This is the error of omission issue once again. This is actually a pretty important point-- science is a human invention, it comes along as part of the human condition. Clearly there is much more in that condition than can be expressed in the equations of physics, or the logic of natural selection. Experiencing life and understanding life will never be the same-- I'm just saying that science should never be asked to tell the "whole story" of all that we perceive. These "blind spots" are what I mean by science as a study of a projection of reality, not reality itself. The shadowy projection is objectively measurable, that's how its defined, but there's no reason to expect it to encompass all of reality. There's no reason to even expect it to encompass the most important elements of reality. Mind you, I'm not saying that invisible ghosts occasionally appear to people in sporadic ways that we just can't measure, that requires a model of ghosts and one is basically doing bad science. What I'm saying is that if anyone chooses to believe such ghosts are here, and if they accept that we already know the ghosts do not project onto the reality that we can quantify and measure (i.e., go "bump in the night"), then science is completely moot on the issue. But nevertheless, they can still be a part of reality, if they are needed to make the parts we do observe work the way we observe it to (call it "the ghost in the machine", if you will). We'll just never know in an objective way, since that's what science is properly used for.

I'd say that we're quite certain that our understanding breaks down at the Planck scale, but that doesn't mean that explorations at lower energy scales might not show us discrepancies between general relativity and quantum theory.

Very true, but that would also be true if GR and QM were perfectly consistent at the Planck scale. It's still irrelevant what happens at that scale, unless one simply hopes that seeking unification leads to a better theory on scales we can measure. It's just a guess that it would.

The Planck energy is large on the scale of particles, but it's still a very small amount of energy. It's not a matter of having a huge amount of energy, it's a matter of getting it into a small region of space.

That is a valid point, and perhaps someone thinking along those lines will one day harness that scale and win a Nobel prize. I don't rule it out, but I don't see the evidence. We use (what we call) particles to transport energy into small regions of space, I don't know how else we'd do it, but I'm not going to win any Nobel prizes.

I'm not suggesting anytime soon, but it wouldn't much surprise me if we managed it within the next millennium or so. Not that I'll be here to see it, most likely.

Me either-- a bet that won't pay off one way or the other.

But for me, it actually seems like a reasonably interesting avenue of investigation. That is, if you know that two very successful theories clash, that immediately gives you some clues as to where those theories might need to be refined. It's not a guarantee, but it's as good a place to look for new physics as any.

And I'm completely comfortable with those statements. I think if it was "sold" just that way, it would be honest and true to what science is. The "hype" gets carried away, to the point where science and other forms of asking deep profound questions seem to end up in conflict. This is ultimately the issue that I'm decrying, not the effort to unify QM and GR, it's the sense that science is "on the brink" of figuring everything out, instead of recognizing that the most important things will always slip between the boxes that science uses to quantify and conceptualize. Put succinctly: the TOE will never put poets out of business.

[Ken G]

Actually, here's a pretty interesting paper that explores just that, the possibility of detecting individual gravitons. Their conclusion is that it would be theoretically possible, but would require a detector on the order of the mass of Jupiter, which would then have the problem of structural failure because of self-gravitation. And there are issues of noise and so forth. So the conclusion is that it is probably impossible form a practical standpoint.

And there we at least have a starting point. It would be very interesting to detect the quantization of gravitons, though of course none of that would really be happening on the Planck scale. It would just be a much more modest advancement of physics, the way it's really done.

Perhaps, but there are plenty of domains where consistency with observation is all we've got, and we can't directly perform experiments. A really large portion of astronomy and cosmology fall into the category. I'll agree that when you can't control the experiments that you're always less certain about a theory.

That gets into the very subtle issue of what it means if we say "here's a consistent picture that works after the fact, but does not make predictions that we can test. It was designed to explain everything we see, but we doubt we'll ever get more out than what we put in." Clearly, such a theory would still be quite useful as a way of organizing all the information we've observed into a conceptual framework, but is there a difference between that kind of science and science that actually makes testable predictions that could have been wrong but ended up being right? In other words, when is science a search for the "truth behind the curtain", and when is it just a way to organize the phenomena that emerge from behind that curtain?

But ptolemy's system was strictly descriptive. That is, you could work out a plan for a well-observed orbit, but there was no way you could use that model to predict a new object based on existing orbits, or even map out an entire orbit from a handful of observations when something new was observed. Critics of science always like to hold up Ptolemy's model as a great example of how science can seem to be working beautifully and still be hopelessly wrong, but frankly, it was never a very good model.

That is a point well taken, and speaks to the issue I just raised-- Ptolemy saw nothing "behind the curtain", whereas Kepler and Newton did. Still, it is the heliocentric vs. geocentric models that caused the revolution, moreso than Newton's laws (which indeed caused their own revolution, but for different reasons). People really wanted to know if the Earth was a celestial sphere or a grungy cruddy place for the detritus of the universe to collect, and it was Galileo's telescope that decided it-- which was also purely descriptive. It is one part of science to seek out better descriptions, even if you don't know why-- science by living with the monkeys, if you will. We talk about physics and metaphysics, but perhaps there are actually three levels: descriptive physics, predictive physics, and metaphysics. Will the real reality please stand up?

Is it true that there have been theories that we no longer think are accurate representations of reality? Absolutely. Should we therefore conclude that it is impossible for a theory to be an accurate representation of reality? That does not logically follow.

True, but even if we don't know whether or not our current theories are the reality, it is enough. The hubris is dispelled the instant we admit that our models have uncertain connection to that which is real. I think that approach would go a long way toward mending the chasms between scientists and nonscientists. Yet most scientists visibly blanch even at the suggestion, because they are afraid the nonscientists will take it as a sign of weakness, an opening for counterattack. But what counterattack? Science is, and has always, been about its usefulness, and that doesn't go away when you admit that there may be more under heaven and Earth than dreampt of in your TOE.

But even if we don't understand the details of the components themselves, that doesn't mean that it's not meaningful to talk about objects as being made of those components.

I certainly would not say these concepts are meaningless, I am only saying that their meaning emerges from their usefulness, not their "truth". Science simply doesn't distinguish those concepts, but outside science, there is a distinction. That is my point in a nutshell, though there is much else of interest that we've discussed.

Heck, even just saying a higher density region where certain physical processes are taking place calls it out as something distinct from, say, the intergalactic medium.

Indeed, it is possible to make a definition of a "star", but what use does reality have for such definitions? My point is, the concept of "star" is not part of reality, it is invented by us to make sense of reality because our goal is to replace reality with something simpler. Reality does not need to do that, it is happy just being. Reality's job is to just be, and science's job is to replace it with something similar that has a much higher conceptual value. It's a tradeoff between precision and simplicity, and reality is all the way at one end while science has to play that tradeoff. Perhaps the distinction between descriptive science and predictive science and metascience enters into the dimension of this tradeoff: if I were to order it from most precise to most simplified, I'd put it reality, descriptive science, predictive science, and metascience. Note the sublimeness of the intellectual investment increases the more of the reality that is getting filtered out-- the more you know, the more you don't know.

Or, to use our alien example, I was already willing to wager that any alien's description of the universe would include something analagous to our subatomic particles. But I'd be willing to bet a good deal more that any such description would include elements that correspond to stars and galaxies.

Without doubt, but that's because of the usefulness of the concept. Here I would say we know that a star is purely defined by the intelligence considering it. Reality might have protons, but stars? Why does it need those? Is a star more than the protons and other particles that comprise it? And if not, why would reality need a separate definition for something if its essence was already included in the sum of its particles?

So if you're willing to come join me in the middle ground (maybe our theories could reflect reality), that's good enough for me, even if you remain on the "well, maybe, but probably not" side of the middle.

Oh yes, I'm willing-- my stance was partly polemic, to achieve contrast with the more typical view one finds among scientifically well educated people! Indeed, I don't think our opinions were ever really all that far apart, but understanding comes from the contrast of the chalk and the blackboard, does it not?

[Ken G]

I don't think that's actually circular. Or perhaps it's just that a better statement might be "reality has elements of quantization", rather than "reality is made up of particles". The point I'm making is, if any description of reality must include certain specific types of quantization in order to be accurate, it doesn't make much sense to me to say "well, reality is surely not really quantized in any way, it just behaves as though it were".

It sounds like we are in effect talking about finding a kind of "Turing test" for reality itself. Turing felt that if you couldn't even in principle tell if a computer program or an intelligent human was on the other side of the wall, then you had real intelligence (at least, that's what I think he said!) So what is the Turing test for a theory of reality? Behind one wall, you have the universe's most renowned alien physicist, with all the equations and concepts of a billion years of physics. Behind the other, you have that Douglas-Adams inspired robot I brought up before, who just consults the reality. Now, will you always be able to figure out which is which? I think you always will. If for no other reason, then because reality can be experienced, and will never feel the same as an understanding of reality, will never be replaced in its essence by an intellectual process. The two don't even have the same goals, it's that "projection" business again.

When you bring this up, I wonder if we may simply disagree about how close something has to be to be considered correct. If I discover that protons seem to be made of quarks, does that mean I was wrong about there being such things as protons?

Yes, in my meaning of the word "being", it does, it means there are no such things as protons, there are just quarks masquerading as a single particle. Sure the concept of proton will continue to be used as a simplification, but it isn't the reality, not when what is useful is being distinguished from what is real. The problem is with circularity-- if we define proton as whatever reality is doing that looks like there's a proton there, then we've defined it to be real, we haven't learned anything about reality. It's purely descriptive-- we're not seeing behind the curtain, or we're seeing behind one curtain only to find reality hiding behind yet another one.

I would quickly agree that my "box" (my understanding or my model of a proton) has changed shape or size. I've added new things and maybe taken away other things. But it looks like you're saying no, that if my description isn't absolutely perfect in all ways, then in fact it doesn't represent reality at all.

Oh it represents reality, within the limits of the representation. But it isn't the reality, it is purely a conceptualization. Take away human minds, and what does reality need with protons if it already has quarks? (Of course, reality doesn't have those either...) It is we who need protons, because we don't want to keep track of quarks. That's what I mean by "reality", what's really happening, not the words we choose to tell a story about how we can imagine what's really happening. It's like telling a child that babies come from storks, only on a much much more sophisticated and accurate level-- for the child, this explanation is successful, and I feel that is how advanced civilizations would think of our science, and also how "reality" would think of their science, if it could think about anything but itself.

What you've stated here is essentially the Sapir-Whorf hypothesis, and linguists and anthropologists have been debating the idea for years.

I think I may be even going it a step farther-- they seem to be saying that the structure of language affects how we make sense of our reality, and I'm further saying that there's an inescapable tradeoff when using any such structure-- that same tradeoff between power and truth, the power of science and the truth of reality. The act of making boxes guarantees that some things will not fit, but that's not a weakness of using boxes, or of using language-- it is the entire point. You are seeking a kind of "minimal subset" that will allow you to accomplish what you want to accomplish, and so you acquire the power to communicate as long as you limit yourself to communicating the things that can really be communicated that way. Science, for example.

Yes, the description of some force is not the force itself, but that does not mean that it cannot be an accurate description.

I agree that science can produce accurate descriptions of trajectories, and you can even specify the accuracy you want. I'm just saying that it will always be a description, not the real thing, and what that means is that the description will always throw out the bits that don't fit into the boxes it imposes. Further, I'm saying that the thrown out bits are unrecoverable, it is an inescapable price of the entire exercise, indeed, it is the point of the exercise.

Actually, I'm not sure that you can be sure of even that. For example, some people have speculated that space is quantized at the scale of the Planck length. If that were the case, then there's a limit to the precision needed in any distance measurement (the best you could do would be to count it off in Planck lengths).

It is an interesting question, if quantization implies the universe contains a finite amount of information in any finite volume of spacetime. I think it is infinite, but the boxes accessible to science might actually be finite, that's true.

Because I'm not trying to claim that our theories necessarily do represent reality. As far as what it means, I'd think that we should therefore be cautious about whether we assume theoretical constructs are real.

And this is certainly the place where we achieve convergence of opinion.

However, I'm not sure I see as much "hubris"* in a "theory of everything". Maybe it's the name. I don't really think that anyone expects that if we can come up with a workable theory of quantum gravity that we'll suddenly understand everything.

Sure, they know it won't cure cancer. But some seem to feel that unifying GR and QM will mean there is no more fundamental theory needed, that this is the beating heart of reality, and the rest is application. Maybe difficult application, like multiplying ten-digit numbers in your head, but we'll know the guts of what multiplication is. I think this belies the true relationship between science and reality, the fact that science always asks "OK, what aspects of reality are we going to throw out today so that we may achieve a powerful description of some process and go out and use it to build some bridges, or interstellar travel, or whatever." Just as in the linguistic example we were discussing. It is this fundamental purpose of science that is anathema to the whole idea of a theory of everything. The more you know, the more you don't know-- science is built to cope with this problem, not by solving it, but by learning how to live with it. I think scientists are often forgetting this.

Might an attempt at resolving the conflict between the two most successful theories give us some ideas about how to go about modifying those theories? I think yes.

And I still agree, put that way.

* I took an excellent course in Greek literature at one point, where we had an extensive discussion of what hubris was and was not. The short summary is that it was absolutely not as simple a concept as "pride" or "arrogance", and the professor despaired of high school students everywhere being given a far too simplistic view of the matter by teachers who didn't really understand the material. I've tended to avoid the term since then...

hehe, well, I confess I don't know what the professor was on about, but if "arrogance" is better I can use that instead!

[George]

We talk about physics and metaphysics, but perhaps there are actually three levels: descriptive physics, predictive physics, and metaphysics.

That makes sense to me, but the first two travel together, whereas the latter usually just circles the wagons.

Will the real reality please stand up?

I was wondering when you'd call for an engineer. Proofs are in the puddin'; as long as they taste good, or are flameable, they're really real enough.

Sorry, my existentialistic views are a little too weak to add much to the thread, but the bases look pretty well covered. But this does tie in with a book I'm re-reading entitled Galileo's Mistake which addresses some of the history on the separation caused by science, as well as much of the misunderstanding that surrounds that whole affair. I think you'd enjoy it Ken, though we both would disagree strongly with a number of points presented by the author.

[Ken G]

It would be interesting to collect that perspective, I would think.

[George]

Another thread perhaps. Planck scales aren't mentioned much.

[ngeo]

Nereid wrote at the beginning of this thread, “. . . . all the usual terms we feel we have an intuitive feel for - energy, in particular - require a 'something' in which to make sense. Whether that something is Newtonian or Einsteinian, it is continuous, at least wrt 'space' and 'time'. In the Planck regime (at the Planck scale), 'space' and 'time' can't be continuous ... if we apply one of the key results from quantum mechanics.”

Interestingly to me (possibly uselessly to others), in speculations I have followed through, at the level of particles it seems that quantization could be described as time going backwards.

[Nereid]

Ken G wrote, “. . . science has no answer for the problem that the more you know, the more you don't know.”

Since scientists generally attempt to know, and science in general attempts to know more, would the statement above mean that science is part of “the problem“? If not, then maybe there is no problem in the apparent simultaneous increase in known and unknown. (And is the unknown really a function of the known in this equation?)

This is a very interesting question - why do we do 'science'? 'why' in the sense of 'for what purpose/goal?' and 'what is the motivation?' (they are not, usually, the same thing).

Perhaps we could apply the tools of science to these questions? (more later)

If so, maybe “reality” is a problem only for scientists who believe they are “working” to increase ignorance. How useful is that, and if the purpose of knowledge is strictly utilitarian, what is the purpose for which it is being used? Is it a useful purpose? What is the use, or purpose, of cosmology? Is it just useful for employing smart apes at large salaries?

Several other folk have given their take on at least parts of this; let me say a few words of my own (much is little more than a repeat of what others have said).

A decent review of the history of science will turn up a great many examples of unanticipated consequences of new theories*, of a great many kinds - direct, indirect; postive, negative; ... even where 'practical applications' may have taken a long time to materialise, societal impacts may have happened very swiftly, and vice versa (for example, what was the practical impact of Darwin's theory of evolution? its societal impact? both then and now).

May we then conclude that the consequences of new theories (one's not yet developed) are unknowable?

The OP appears to be interested in defining “space”, “time” and “energy” in a way that “works” or is “useful” in harmonizing two theories that don’t “work” very well together. Would Ken G agree that the words “space”, “time” and “energy” represent something “real”? If so, would the “reality” exist at an “energy” level that cannot now be probed? Or is it a useless exercise? And would those words ultimately be defined only in mathematical terms and put to “use“?

I should go read the Theories thread; what I write here and now may have already been covered ...

The default, for me, is that all terms are theory-based ... if they are part of some scientific description or explanation.

At at very basic level - life and death of an individual (you) - many terms can be linked, closely, with intuitive concepts that we, as Homo sapiens creatures, are hard-wired (or firm-wired) with. Many others cannot be so linked.

At a practical level - including the life and death one - many of the differences between various theories is irrelevant - whether gravity is Newtonian or Einsteinian (or push not pull, or ...), you would still go splat on the footpath if you fell from a 20 storey building (without a parachute, etc).

So 'reality' is only what can bite you; or, more generally, what you can 'see'. Beyond what you can 'see' is extrapolation and belief ... 'reality' is only as deep as the theories you use.

(as an aside, I think this provides a middle way between (some of) the points made by Ken G and Grey. For example, 'light' is 'real' (you can 'see' it) but also theory-based (it was a wave - no a particle - until it became quantum mechanical photons); 'protons' are 'real' (you can 'see' them) but also theory-based (they were fundamental particles ... until they became merely triplets of quarks held together by gluons); the 'life force' (elan vital) is 'real' (you can 'see' if something is alive or not), but also theory-based (it was an innate characteristic of living things ... until biochemistry showed the 'theory' to be inconsistent with observation); and so on.)

Mathematics plays such a key role in modern science because quantifying things (measuring things in a quantitative way) is the most powerful means of testing* we have ... and it 'pays its way' because it gives us effective medicines, computers, aeroplanes, tasty food, ... If you could find a way to form and test theories, without mathematics, then maths would become less important.

Of course this just begs the question 'what is the relationship between maths and 'reality'?'!

If so, why are forum participants using English words to talk about the idea of reality? Maybe words are useful - science! Maybe thought is useful. Maybe what we imagine is as real as what we imagine to be real. Maybe every imaginative creature holds part of what is real, and the parts we don‘t know are less of a problem than the parts we know, and not knowing why we know is the biggest problem, which may be the problem with scientists and the reason they “work” for people who don’t know. Strangely, the universe may show us its purpose, and if we follow it, we may find our own purpose, and start to put our knowledge to good use rather than a use that may or may not be good (which we in our knowledgeable ignorance cannot tell).

Quite.

That, however, takes us a long way from science ... or does it?

I believe “space” and “time” are units we use to measure events, the assumption being that events follow one another.

In other words, they are constructs used to explain causality?

“Energy” (I believe I have said before) ought to be defined as the ability to move (because at its root I believe that is what it is) which then necessitates spacetime as an effect of energy. Where GR cannot go is to measure spacetime by spacetime. If everything is made of spacetime, GR cannot measure everything. It has to stop where spacetime “lives“, which is in material bodies. If material bodies are “made of” spacetime, how do you measure something with itself? It’s like measuring a ruler with a ruler. If they disagree, which is right? It seems to me that somewhere around the Planck region is where spacetime meets spacetime. Then you have to start dealing with energy itself.

I think at this point you need to reach for the math ... without putting these ideas into some kind of mathematical form, I don't see how you can go much further than just words like these.

In particular, you can't form theories, and you certainly can't test them, so you're not doing science.

At a deeper level, aren't you - implicitly perhaps - declaring that there are certain things which you cannot 'see' but which are nonetheless 'real', independent of any theory?

From my point of view (above), such an approach is likely to be quite sterile - no theories, no deeper understanding of 'reality' (whatever it might be), no improvements for your health, well-being, and stock of material goods, ...

(to be continued - in particular, what happens when you apply the tools of science to examine science itself?)

*As a shortcut, I am taking 'theory formation and testing' as the motor that drives science.

[Nereid]

I read an interesting article (Scientific American, in the last year or so?) on some implications of results coming from the world of computer science, when applied to cosmology.

It went something likes (IIRC): the universe is a (giant) computer. What is it computing? Itself.

Assuming that I have not completely destroyed the idea, in my summary, and assuming the idea isn't riddled with inconsistencies (internal, with experimental and observational results, etc), how would Ken G's views, as expressed in this thread, factor in?

It seems to me that it's a different kind of "toy model" than the standard model of particle physics (the two would only be in any kind of potential conflict if it could be shown that, at some general level, 'computing' can't be done/happen with the kinds of mathematical entities that comprise the standard model).

[Ken G]

It went something likes (IIRC): the universe is a (giant) computer. What is it computing? Itself.

I think this idea is the scientific ideal taken to its logical extreme. But this is exactly my criticism of extrapolating science too far: the idea that the universe could "compute" itself is missing what a computation (and science) is. Clearly we are talking about analog computing here, as the limitations of digital computing present an obvious problem. So take a clock, with hands and all, as the quintessential analog computer of time. How does a clock "compute" time? By tapping into some fundamental aspect of the universe that we understand as time, and tapping into it so directly that it is able to measure time. But does a clock recreate time, as it computes it? Or is the computation still just a shadow of the real thing? I am aguing the latter, and saying that the scientific concept of time is a shadow on the wall of something far more sublime-- it's that which makes the clock tick, allowing it to measure something, but the clock is not computing time because it needed the reality already in order to make it work. A computation is never intended to be the real thing-- it is always a kind of substitution or projection. So if the universe were computing itself, it would be shortchanging itself. It would be like a system of real numbers subject to a finite set of axioms-- there would still be truths in the system not provable by the axioms (to borrow from Godel that one concrete glimpse into what I'm talking about.)

[ngeo]

Nereid, it is interesting that you differentiate purpose - end - from motivation - cause. It seems that the utilitarian argument leans toward end use. The universe I see has no end, its purpose is found by looking - backward - at the path it has taken, or what it does, which is to grow, i.e. to create new systems out of existing systems. Indeed maybe the universe can be described as a self-designing system for creating new systems. So if we seek purpose, a good place to find it is in the way the universe works, and the purpose is to grow, but that has no end (if we are lucky).

I believe growth cannot take place with harmony. That is, two or more systems fit together. If they don’t, there is chaos. Harmonic systems have an element of chaos, and harmony arises out of chaos. Continuing growth increases the complexity of harmonies, etc. If humans are lucky, this process can continue forever, but there are no guarantees and as systems become more complex you can go a long way down the wrong path before you know it is wrong.

So the imperative for growth gives a moral compass: seek harmony. I find institutionalized science to be amoral, and being amoral it assumes nature is also amoral. While the “science” of western economics can take evils as goods, and disservices as services, physical science takes no stand on anything. But nature is not amoral. It follows a moral compass - it seeks harmony. Not survival of the fittest, but survival of what fits.

But to get to the point about spacetime and energy, from what I can gather after doing my version of “the math“, one volt of electric potential is equivalent to one second of time. And a volt is itself a frequency - an oscillation! So at a certain level (as I said above) time oscillates (or rather, spacetime). This is from the math. And (also from my own version of “the math”), the gravitational force between proton and electron equals h^2 / 4 - that is, (h / 2) x (h / 2). Wouldn’t that point toward a quantized gravitational force?

Which leads me to ask whether a “quantum theory of gravity” needs to have a graviton. In other words, does there have to be a specific “particle” which is exchanged, if spacetime oscillates?

[Nereid]

Earlier in this thread I mentioned that Aristotle may have had 'bad press' (i.e. that he was, in many respects, much closer to being a 'dedicated follower of (modern scientific method) fashion' than he is normally given credit for.

I tracked down the source of my recollection: The Trouble with Science, Robin Dunbar (ISBN 0-571-17448-5), page 39: "What is especially interesting about the things that Aristotle got right and those he got wrong is that they partition rather neatly into those things that were easy for him to see and those that were not." Dunbar does on to analyse Aristotle's biological works and presents a simple 2 X 2 matrix - columns "Could Aristotle have studied them himself? (Yes/No)", rows "Number of topics that Aristotle: (got right/got wrong)" The results:

Ntopics Ye No
....right 32 2
...wrong 2 10

[Nereid]

We've drifted off the OP - but that's OK, this is a very interesting topic.

There's one angle which we've only just touched on - changes in the nature of science itself.

We can look back and identify - imperfectly - what elements of 'modern science' we see in the work of the ancients; we can look to the more recent past and see how 'modern science' is continuing to change ... but can we project into the future? With what degree of confidence can we say that what we are doing today, as physics, as cosmology, will be regarded as solid 500 years from now? 5 million?? Or the converse - what will a future Robin Dunbar tease out of the dusty records of the 20th/21st century to show correspondence with the then consensus view of 'modern science' (and point out just a few examples of where it falls way short)?

While I am no historian of science (I am an astronomer by profession), let me start this off by looking back.

While it is tempting to compare the cosmological (in the broadest sense of that word) models of Eudoxis, Aristotle, and Ptolemy to that of Copernicus, Galileo and Newton to that of Einstein, Hubble, FRW & modern cosmology (and it appears that some have suggested doing so, above), I hesitate in taking these comparisons too far. While these different major groups of investigators of the natural world had many similar ways in approaching the problem (e.g., the appreciation of simplicity in unifying many observational phenomena, the appreciation of symmetry, amongst others), science as a methodolgy of investigating, predicting and modeling the behavior of nature developed along the way (e.g., the importance of data and experiment in support of a hypothesis, amongst many others). As this methodology matured, the rapidity in scientific discoveries grew enormously - indeed the two likely fed back on one another.

This rapid progression in our ability to predict the behavior of nature (identifying its "laws", deriving general theories, etc) is telling us that we must be doing something right, and not just randomly wapping around in the dark. In fact when measured across all the sciences, one might argue this growth is similar to an exponential function with a shortening time constant. That doesn't mean I believe that the "Theory of Everything" lies just around the corner (if it even exists at all).

That's enough for a start...somebody want to run with (or against) that or project forward into the future?

I think some near-term projections are both relatively easy and safe (in the sense that they are unlikely to be proven seriously wrong, by events):

1) Deeper, wider, sounder application of statistical theory, in all aspects of science, but especially where there are great quantities of (good) data. The recent 'Bayesian revolution' is an example of this.

2) Ditto, wrt data collection, analysis, storage, and access, using techniques from computer science (databases, data visualisation, ...).

3) Ditto, wrt simulation, modelling, etc.

None of these may look all that revolutionary to us, today, but the power of these would have astonished 19th century scientists. They will surely play a (crucial?) role in many of the forthcoming breakthroughs. There's also probably a somewhat related, unheralded trend - the math which underlies so much of this is itself changing in ways that will have deep implications for the future of science too. For example, what impact will the automation of (mathematical) proofs have? or the advent of practical quantum computers?

4) Scope creep: it was not so long ago that the workings of the brain were essentially a mystery, and the realm of dualism huge indeed. Today the last (?) frontier is 'consciousness' and 'free will'. In the time of Newton, the way life changed with (geologic) history was not in the domain of science; today, it can be studied back to a time before the oldest surviving terrestrial rocks; tomorrow maybe the origin of bacteria and 'life' will become accessible. And so on. Of interest to us, in this thread, is the likely advance into the Dark Ages, into baryosynthesis, leptogenesis, the details of how inflation worked, the next step beyond the standard model (of particle physics), the testing of GR in strong field regimes, ...

And maybe much more ...

More speculatively, what are the big, qualitative changes that science will undergo? I think none of us could make a good case for any such, but here are some possible pointers:

5) Reductionism vs holism: for example, is there a way to get at 'emergent phenomena' other than by reductionism? The limitations of the standard approaches are clear, in fields such as economics. Is there a fundamentally different way to tackle these kinds of interacting, highly non-linear systems?

6) Quantum philosophy: 'shut up and calculate' works, brilliantly. Yet thinking deeply about the quantum world can be very unsettling. Is there a (philosophy-based?) approach waiting to be discovered, that will give us tools and techniques to go beyond just calculating?

[Nereid]

[snip]

But to get to the point about spacetime and energy, from what I can gather after doing my version of “the math“, one volt of electric potential is equivalent to one second of time. And a volt is itself a frequency - an oscillation! So at a certain level (as I said above) time oscillates (or rather, spacetime). This is from the math. And (also from my own version of “the math”), the gravitational force between proton and electron equals h^2 / 4 - that is, (h / 2) x (h / 2). Wouldn’t that point toward a quantized gravitational force?

Which leads me to ask whether a “quantum theory of gravity” needs to have a graviton. In other words, does there have to be a specific “particle” which is exchanged, if spacetime oscillates?

I've snipped out the first part, because it takes us far, far beyond not only this thread, but also BAUT.

All the easy options wrt a quantum theory of gravity have, for sure, already been explored (and possibly several times). Why? Because, for the most part, science works as a ratchet - each advance gets locked in, and everyone working in a particular field can (and will) apply any (and all) techniques used (successfully) by (all) predecessors. Of course, this doesn't operate at the level of any particular individual scientist!

However, it does suggest one way to get a jump on a difficult area (such as quantum gravity) - borrow techniques used in a field far from your own ... or, having mastered one field of science, go work in a completely different one.

There are many examples, in the history of science, of the fruitfulness of this 'cross-pollination' - maybe the breakthrough in quantum gravity will come from someone who got accolades for their PhD ... in economics, or paleobotany?

(Another ground that has been, historically, fertile is 'dead ends' of old - maybe hints for a good way forward lie buried in the dusty shelves of physics journals of the 1920s, in a paper that hasn't been cited since 1930?).

[gzhpcu]

Forgive me, if I simplistically just compare all this to the classic "black box" approach in programming:


A programmer/analyst observes a system and selects valid and invalid input and determines the correct output. There is no knowledge of the object's internal structure. He/she comes up with programm code which predicts the output based on the input.

Science seems similar to me: we have theories based on mathematical models and the system in the black box is the universe. As time goes on, the range and scale of the input/output parameters becomes progressively ,more complex, and the programm code (mathetical theory) is modified accordingly. The goal is to be able to predict the inner workings. Any similarity between the code (theory) and the contents of the black box, is purely coincidental. The input/out parameters become progressively smaller, and at some point we are no longer able to specify them, though we know they exist. Then we start speculating... (i.e. M-theory...)

[Nereid]

Forgive me, if I simplistically just compare all this to the classic "black box" approach in programming:


A programmer/analyst observes a system and selects valid and invalid input and determines the correct output. There is no knowledge of the object's internal structure. He/she comes up with programm code which predicts the output based on the input.

Science seems similar to me: we have theories based on mathematical models and the system in the black box is the universe. As time goes on, the range and scale of the input/output parameters becomes progressively ,more complex, and the programm code (mathetical theory) is modified accordingly. The goal is to be able to predict the inner workings. Any similarity between the code (theory) and the contents of the black box, is purely coincidental. The input/out parameters become progressively smaller, and at some point we are no longer able to specify them, though we know they exist. Then we start speculating... (i.e. M-theory...)

While I think there are some similarities between the 'black box' approach you describe here and scientific understanding of the nature of the universe, there seems to be at least one part that gets very short shrift indeed - the power of a theory to predict new tests (and, along the way, connections between 'inputs' that no one had thought of before).

Or, if you prefer, the difference between code which does 'nothing but' re-sort the inputs and print them out (as outputs). In your analogy (if I have understood it correctly, which I probably haven't), a computer code which did a particularly snazzy job of datamining to produce a list of the times of sunrise at {location}, using a huge dataset of historical sunrise times as input, would be equivalent ('just as good', in terms of being science) to one which made the same predictions (outputs) but based on Newtonian gravity (and a quite different set of inputs).

(Maybe Ken G will jump in, with at least a reference to Occam?)