The Graviton as Aether
Well, first: Does any collapse theory have any experimental support? No.
With that out of the way...
If collapse actually worked the way its adherents say it does, it would be:
The only non-linear evolution in all of quantum mechanics.
The only non-unitary evolution in all of quantum mechanics.
The only non-differentiable (in fact, discontinuous) phenomenon in all of quantum mechanics.
The only phenomenon in all of quantum mechanics that is non-local in the configuration space.
The only phenomenon in all of physics that violates CPT symmetry.
The only phenomenon in all of physics that violates Liouville’s Theorem (has a many-to-one mapping from initial conditions to outcomes).
The only phenomenon in all of physics that is acausal / non-deterministic / inherently random.
The only phenomenon in all of physics that is non-local in spacetime and propagates an influence faster than light.
WHAT DOES THE GOD-DAMNED COLLAPSE POSTULATE HAVE TO DO FOR PHYSICISTS TO REJECT IT? KILL A GOD-DAMNED PUPPY?
- Eliezer Yudkowsky, Collapse Postulates
In the olden days of physics, circa 1900, many prominent physicists believed in a substance known as aether. The principle was simple: Maxwell’s equations of electromagnetism had shown that light was a wave, and light followed many of the same equations as sound waves and water waves. However, every other kind of wave- sound waves, water waves, waves in springs- needs some sort of medium for its transmission. A “wave” is not really a physical object; it is just a disturbance of some other substance. For instance, if you throw a rock into a pond, you cannot pluck the waves out of the pond and take them home with you in your backpack, because the “waves” are just peaks and troughs in the puddle of water (the medium). Hence, there should be some sort of medium for light waves, and the physicists named this medium “aether”.
However, difficulties soon developed. If you have a jar, you can pump the air out of the jar, and then the jar will no longer transmit sound, demonstrating that the wave medium (the air) has been removed. But, there was no way to remove the aether from a jar; no matter what the experimentalists did, you could still shine light through it. There was, in fact, no way of detecting, altering, or experimenting with aether at all. Physicists knew that aether must be unlike all other matter, because it could apparently pass through closed containers made of any substance. And finally, the Michelson-Morely experiment showed that the “aether” was always stationary relative to Earth, even though the Earth changed direction every six months as it moved about in its orbit! Shortly thereafter, the inconsistencies were resolved with Albert Einstein’s Theory of Special Relativity, and everyone realized that aether was imaginary.
Shortly thereafter, during the 20th century, physicists discovered two new forces of nature: the strong nuclear force and the weak nuclear force. These two forces, as well as electromagnetism, could be described very well on the quantum level: they were created by the influence of mediator particles called (respectively) gluons, W and Z bosons, and photons, and these particles obeyed the laws of quantum mechanics just like electrons and mesons did. The description of these three forces, as well as the particles they act upon, has been neatly unified in a theory of physics known as the Standard Model, which has been our best known description of the universe for thirty years now.
However, gravity is not a part of this model. Making an analogy to the other forces, physicists have proposed a mediator particle known as the “graviton”. The graviton is thought to be similar to the photon, the gluon, and the W and Z bosons, except that it is massless and has spin 2. I posit that the “graviton” is essentially the same theory as the “aether”: a misguided attempt to explain something by reference to similar-seeming things that were explained in the same way. Consider the following facts:
Some of the brightest minds in the world have been working non-stop on how to to extend the Standard Model/quantum mechanics theory of forces to gravity, for more than forty years, and we have yet to find a satisfying way of doing it. Theories of physics have a known tendency towards elegance and simplicity, and nothing we have come up with in the past forty years is either simple or elegant, let alone both.
Of the explanations that have been proposed, string theory is currently the most popular. The idea that a theory of physics can have “popularity” should, in and of itself, be a warning sign. There can only ever be one reality, and so there can only ever be one correct theory of reality. The physics community, of course, cannot all just switch to a new theory as soon as its correctness has been proven; there are lots of inevitable time lags. However, we have had ongoing disputes between people in the string theory camp, people in the loop quantum gravity camp, and people in other camps for more than thirty years now, with zero sign of a winner or of the matter resolving itself anytime soon. This, to me anyway, is a sign that no one really knows what is going on, and lots of people are just following their monkey intuitions and forming Blue and Green teams.
String theory, of course, has zero testable predictions. Loop quantum gravity, another graviton-based theory, also has zero testable predictions. In fact, to my knowledge, all of the use-gravitons-to-explain-gravity-on-a-quantum-scale theories have zero testable predictions.
String theory has additional problems: it’s not a satisfying explanation in the same way that relativity, or Newtonian gravity, or Keplerian astronomy, or Maxwell’s laws are, because there are so many different versions of it. You can’t just go read a textbook on string theory, and say “Aha! So that’s how the world really works”, because there are a ton of different kinds of string theory. There are (as of this writing) bosonic string theory, type I string theory, type IIA string theory, type IIB string theory, and heterotic string theory. (This list does not include the different ways of describing the universe within each of these theories, which are practically infinite.) It also requires numerous additional dimensions, which are unobservable and untestable, but which are required to make the math work right.
And, with reference to the graviton itself:
Gravitons have, of course, never been detected, and (so far as I know) there aren’t even any serious proposals on the table for how to detect them. In fact, it is widely thought that gravitons have such low interaction cross-sections that detecting them with any physically possible detector is a hopeless proposition.
The classical theory of gravity that we know works well, General Relativity, makes no mention whatsoever of gravitons, and describes gravity in terms of the curvature of space-time. None of the other three forces (weak, strong, electromagnetic) have ever been described in this way, but gravity has, with a precision of more than twelve decimal places, in addition to many other unique predictions (such as the geodetic effect, which has recently been confirmed experimentally).
Gravity is much, much, much weaker than all of the other forces. In fact, it is more than thirty-five orders of magnitude weaker than both the electromagnetic force and the strong nuclear force.
Gravity is nonrenormalizable, which means that, if you try to calculate the strength of the gravitational interactions using a quantum-mechanical, quantized-graviton model, there’s no simple way to make infinities go away, like there is for electromagnetism, the strong force and the weak force.
So, what’s really going on here? I don’t know. I’m not Albert Einstein. But I suspect it will take someone like him- someone brilliant, very good at physics, yet largely outside the academic system- to resolve this mess, and tell us what’s really happening.
As I get closer to really understanding the theory (that is, knowing all the mathematical detail), I find myself becoming more and more of a string-theory fundamentalist, so I should attempt a response to this.
First, some basic quantum field theory. In quantum field theory, particles and fields are generally regarded as complementary. Particles are energy quanta from field modes, and field states are equivalent to certain superpositions of particle states. Every particle in the Standard Model corresponds to a field, and vice versa. But if you have to pick one description as more fundamental, it’s probably the field, since the fundamental variables in the basic equation are field variables.
The classical theory of gravity is general relativity, and it’s a field theory, so as a quantum field theory there should be gravitons. But as I just explained, a graviton state is a superposition of field states and vice versa.
The principal reason for the conceptual churn in quantum gravity, at least before string theory, was that quantum general relativity is nonrenormalizable, i.e. you cannot perform calculations with it. Renormalization (in its simplest, original forms) is a procedure for dealing with the divergent (infinite) integrals showing up in “perturbative quantum field theory”, which is where the particle states show up. An elementary particle state may be regarded as a certain superposition of states of the noninteracting field. Perturbative approaches to interacting field theory are possible if the interaction is weak enough that the actual dynamics may be regarded as a correction to the free field dynamics. Feynman diagrams are a notation for these corrections which also symbolize some of the space-time processes appearing in the superpositions. But even in perturbative field theory, you still get infinite values for some of these corrections. You will be summing amplitudes over a variety of possible processes, and sometimes these sums just don’t converge. Renormalization involves the assumption of fictitious processes whose integrals sum to infinities which cancel out the divergent integrals.
This is one of those concepts which is ubiquitous in popular exposition of physics but which sounds bizarre and which the expositors even describe as bizarre. However, the simplest way to understand the meaning of renormalization is to suppose that the theory you are working with (e.g. quantum electrodynamics) is incomplete, and that the complete theory (the complete set of interactions, which you do not know, and which may include completely unknown fields) is not divergent and undefined. The renormalization procedure is then a way of incorporating this hypothesis that a better theory has no divergences because other interactions cancel the infinities arising from the incomplete theory.
This is a practical process only because, in a renormalizable theory, only a finite number of parameters are required to define the divergence-cancelling “counterterms” at every order of perturbation theory (order corresponds to the complexity of the virtual processes entering into the amplitude sum over histories). When quantum gravity is said to be nonrenormalizable, what is meant is that an infinite number of parameters are needed to construct all the counterterms. In a renormalizable theory, the finite number of parameters are actually set by empirical values like observed mass and charge. But the infinity of parameters for quantum gravity renormalization render it unpredictive.
However, renormalization arose within a particular approach to making QFT calculations—the approximations of perturbation theory. One can instead aim to solve an interacting quantum field theory exactly. Just because perturbation theory breaks down, doesn’t mean that the full unapproximated theory is undefined—just that you can’t approximate the solution using that method.
So, back to the Standard Model, gravity, and fundamental theory. That gravity is not part of the Standard Model is an arguable proposition. The Standard Model is a codification of what is observed. The Standard Model without gravity consists of various matter fields and various force fields in interaction (and a Higgs field to provide mass; this is a hack and the Higgs has not been seen). All these interactions are renormalizable and so you can calculate with them. But gravity is also observed, of course, and we can write down an equation expressing the idea that the other Standard Model fields interact with a gravitational field as Einstein described it. We just can’t do much with it as a quantum field theory, because gravity is nonrenormalizable. The Standard Model including gravity, however, as codified in that equation, really ought to be regarded as our empirically best theory of physics.
Beyond this, there have really been two directions of inquiry. One is unification of everything but gravity in various field theories with large symmetries which are then broken in some way; the other direction of inquiry focuses on quantum gravity, and has seen a great deal of formal invention and philosophical thrashing about, and even one solid result (Hawking radiation). The first path is the one pursued by particle physicists and quantum field theorists, and eventually they constructed unified field theories which did include gravity in a master symmetry; first supergravity, and then finally the string theories. The second path has been pursued by a lot of other people and does not have the methodological unity of the first path. The most substantial criticism of the first-path philosophy from the second path is that quantum field theory generally assumes flat Minkowski space and does not represent many effects from general relativity, such as black holes. However, this criticism has lost a lot of its impact across the years as physicists have learned to apply quantum field theory in curved spacetime backgrounds. There is still the objection that the philosophy of general relativity, in which matter and geometry react to each other, is lost in a philosophy which treats geometry as a classical “background” to quantum fields, but again, I see a slow remediation of this deficiency on the quantum side.
Now to string theory. Mathematically string theory is a work in progress. Tremendous progress has been made, especially when the various string theories of the 1980s were discovered in the 1990s to be intertransformable, but we do not yet know the fundamental variables of the unified string theory, in the way that the undamental variables of quantum field theory are the field degrees of freedom.
It is clear that string theory can produce something qualitatively like the Standard Model—though intriguingly, no-one has yet found a string “vacuum” which has exactly all the qualitative features of the Standard Model (dimensionality, number of generations, particular symmetries and so forth). The problem is that string theory can also describe a large number of other physical outcomes. It depends on what you do with all those dimensions. Until the previous decade (2000s) the mainstream of string research was always hoping that one geometric configuration would end up preferred, with everything else dynamically unstable, and that this would finally produce a falsifiable prediction. However, the anthropic approach crept into string theory in recent years, producing controversy. It seems clear that the answer to the question of string theory’s ground-state geometry—is there just one, or are there many, and is the anthropic approach viable—depends on cosmology. The evolution from unstable geometries to stable ones occurs on cosmological scales, and if there are radically different geometries at different places in the universe, that too will only be understood once string cosmology is more advanced.
I think I’ll stop there, since there are so many different issues which get raised in a discussion like this, and just field specific questions if there are any. I know I certainly haven’t responded to everything in Tom’s post, but I hope this clarifies a few things.
Thank you for this excellent summary of some of the technical details. I disagree on a few points, but most of this is a great explanation of how some of this works.
There’s been a proposal by Verlinde of treating gravity as an entropic force which is pretty neat.
There might even be some “predictions” of things that we already know from it—see It from Bit.
I’m surprised that nobody’s mentioned this yet—it was a bit of a blogosphere kerfuffle quite recently.
I wanted to mention Verlinde’s paper but stopped because I don’t quite understand it. It seems to require all kinds of physics background knowledge that I don’t have. Maybe you could give a purely mathematical explanation of the underlying datastructure? Something along the lines of “R^4 with Lorentz transformations” or “Hilbert space with Hermitian operators”?
The fragments that I understand don’t come from the paper, which is over my head, but from Verlinde’s blog.
In particular, he says: “The starting point is a microscopic theory that knows about time, energy and number of states. That is all, nothing more. This is sufficient to introduce thermodynamics. From the number of states one can construct a canonical partition function, and the 1st law of thermodynamics can be derived. No other input is needed, certainly not Newtonian mechanics. TIme translation symmetry gives by Noether’s theorem a conserved quantity. This defines energy. Hence, the notion of energy is already there when there is just time, no space is needed.
Temperature is defined as the conjugate variable to energy. Geometrically it can be identified with the periodicity of euclidean time that is obtained after analytic continuation. Again there is nothing needed about space. Temperature exists if there is only time.”
So I think the underlying structure might be something like Kauffman’s random boolean networks, or even simpler—a set of states and a transition function from states to states—the “microscopic dynamics”.
Woah, it really seems that Verlinde’s insight is gaining momentum (and citations) in the academia. There may be a full-blown paradigm shift in the making...
This is a standard semiclassical motivation as to why gravitons most probably exist (I think from Steven Weinberg “gravitation and cosmology” but I have since long lost the book so I am not sure): In the limit of weak gravitation GR looks similar to the Maxwell equations. In particular there should exist gravitational waves.. (Have not yet been detected experimentally but if GR is (at least approximately) correct they should exist.) This means that you could in principle build a gravitational wave microscope. Say you want to measure the position of a test particle using this microscope. Now if gravitational waves were actually classical you could use arbitrarily feeble waves and thus arbitrarily small recoil on the test particle. And thus measuring position and momentum of the test particle with lower unaccuracy of position times momentum (along a given direction) than allowed by the Heisenberg uncertainty relation. But if gravitational waves are quantized in gravitons of energy = h times oscillation frequency Heisenberg uncertainty relation will be satisfied (Heisenberg’s original semiclassical derivation goes through for any wave quantised like this).
I do agree with you that GR cannot be a fully correct description of reality on a quantum level (because of this and other issues, especially many worlds). I was saying that we should take the structure of GR into account when building a unified theory, rather than just starting off with the assumption that gravity works exactly like all the other forces.
This is what the proponents of the loop approach usually stress—demand that the theory be background independent, or general covariant (I am not sure whether there is some important difference between the use of these terms).
Posts like this shouldn’t appear on this blog. Adequately considering the question requires a truckload of special knowledge that most visitors don’t have and that is not associated with the topic of the blog.
The rhetoric that is visible on the surface (comparison with something ridiculous, etc.) strongly privileges a hypothesis (of the teacher’s password opaque kind, no less—to most readers), which in absence of understanding of the subject may as well be pure Dark Art.
Being contrarian doesn’t mean that we should be interested in entertaining contrarian hypotheses any more seriously just because they are contrarian. This way lies madness.
I am quite curious indeed as to how my post is any less topical than the quantum physics post by Eliezer that I linked.
I deliberately didn’t write about anything that would have been foreign to any not-reasonably-scientifically-literate reader. If you disagree, please do point out where.
I do not think that the aether hypothesis was “ridiculous”, and never said that I did. It was wrong, sure, but lots and lots of very smart people believed in it for quite a while. Similarly, I don’t think gravitons are “ridiculous” and never said that they were.
The post being off-topic is not part of my argument. (That being said, if I were arguing that it is off-topic, then being no more offtopic than a given other post is hardly an argument against this post being off-topic in the absolute sense.)
The difference between this post and the quantum mechanics sequence is that this post doesn’t get the readers up to speed, and so if they are not already up to speed, they can’t understand what’s going on (also, you won’t be able to get the readers up to speed on this post via merely a blog sequence—it’s too much work). I said:
If you teach such knowledge on the blog, the argument no longer applies (though if it’s too off-topic, the attempt might not be appreciated). If the knowledge is associated with topic of the blog, readers are expected to have it or seek it (e.g. standard biases). If most of the readers happen to already have the knowledge for whatever reason, then it can be assumed as well (e.g. calculus, because most readers happen to be well-educated).
You are either expecting too much, or demanding too little. It takes way more than familiarity with the terms to discuss the nature of gravity...
And I never said you did. It wasn’t ridiculous in its time, but it is now, given our state of knowledge, which is the relevant factor in formation of nontechnical connotations.
The many-world interpretation of quantum mechanics was discussed here several times. People write about artificial intelligence here often. Evolutionary psychology is a standard topic. All these require a lot of specialised knowledge, which is not necessarily associated with rationality. Why not discuss the reality of graviton? Physics is probably more scary than evolution, since everyday intuition seems useless there, but that doesn’t mean that everyday intuition isn’t actually as useless in evo-psych.
I agree that the topic of the OP is a difficult and specialised question, but there are some aspects of it which can be accessible to non-experts, e.g. the question of testability of string theory or loop gravity, use of anthropic reasoning in physics, or how much popularity of theories influences research. I do not want to have here a strict policy of not discussing specialised topics.
My estimate is that most (>90%) of the opinions about these topics expressed on LW are wrong. After all, we are not all specialists, and even specialists are often wrong. But as long as we are trying to evaluate the questions rationally, albeit with limited knowledge, and as long as we are aware about our limits, I don’t see why we ought to stop.
Because there are better things to do.
The point of the post seems to be that the physics establishment may be in danger of falling into an egregious (and maybe even obvious, from an outside perspective) group irrationality trap. That point is highly topical—it shows that even mighty physics is not immune to rationality breakdowns. String theory is probably popular because lots of senior physicists, who based their careers on string theory, take on students who want to study string theory, and reject students who want to study alternatives to it.
Do you seriously consider saving the experts from error, without yourself understanding the subject? (Doesn’t matter whether this limitation applies to you in particular—the important thing is that it does apply to the blog as a whole).
I find it hard to believe that posts about the collapse postulate are on-topic but this one is not. Is there a substantial difference between the two that I’m missing?
Taking two seconds to click on the Collapse Postulate link it appears that the article was originally posted on Overcoming Bias. Also, it appears to be part of a larger sequence on quantum mechanics.
I haven’t read that sequence or that article so I cannot compare them to yours, but all of those links in the block you quoted presumably enhance the discussion to make the conclusion more obvious. Your article has one link.
I disagree with this. A theory is basically a model (or pertains to one). Models by necessity leave out details of the thing they’re modelling (if you disagree, then the best model of reality is simply reality, and we already have that). So depending on which features of reality you think are relevant, you can have multiple models of reality bringing out each of those features. The theories based on those models will sometimes make different predictions, but if they’re good models they’ll agree most of the time, and you make predictions using the model that makes those sorts of predictions correctly most of the time.
As an illustration, look at maps of the surface of the Earth. There are multiple projections based on preserving different relevant information. If you want to plan a sea journey, you use one; if you want to plan a land journey, you use a different one. A globe represents both more accurately, but is harder to carry around / print out of a computer.
Surely our models of physics can have the same sorts of properties; what allows you to make predictions about gravity may not be the same model that allows you to make predictions about electromagnetism.
I would say the same in simpler words:
We only have one territory but we can have an infinitude of maps each serving different purposes.
I have to disagree with tommccabe—the fact is that model and theory are as different as technical manual and source code. There is only one perfect theory of reality—the theory that is reality—but innumerable models for different domains and purposes. Yes, many of our models began as theories, but that is because a good theory is by necessity a good model.
You have a peculiar use of words reality and theory. I feel that theory and model are more or less the same—our ways to describe the reality. The word model is used when the description is not elegant and visibly incomplete. As for the analogy, you can have multiple source codes that do the same work. How do you establish which is the only perfect one, having access to the user interface only?
You have a peculiar use of the word model—up my alley, that word is used when you are analyzing a particular problem, however few approximations you make.
Well, maybe I am too much influenced by the Standard model of elementary particles. There are no important reasons why not call it a theory, IMO.
None of these terms have rigid definitions, it seems.
There can be many different parametrizations of the same surface, but they will each have the same topological and geometric properties.
So we could have several correct models—i.e. parametrizations which have the correct characteristics—and a single theory—i.e. set of characteristics with which to choose parametrizations. But each model will predict different yet-undiscovered characteristics. Using these new characteristics to make predictions would constitute different theories which are consistent-as-of-yet. Given the current state of quantum physics, different theories seem to be able to remain consistent for a while.
My use of model and theory have been prescriptive based on my interpretations of the thread below. I didn’t intend this but all of the confusion and disagreement seems to have been in our minds.
We certainly seem to be arguing about words—you’re probably right that the confusion is just in our minds.
This doesn’t seem right. A good model necessarily leaves things out; if you didn’t need to leave out some details, then you’d use the object itself, not a model of it. But if a good theory is necessarily a good model, then a good theory also necessarily leaves something out. But then the theory can’t be reality, since reality can’t leave any of itself out of itself.
Unless it was more expensive to build the object than to build the model. Or if the design process required information in the model that is not obvious in the object.
But the true answer is “I meant ‘good’ in the sense of accurate to reality, not in the sense of ‘computationally tractable’.”
(Er, you might want to check the author, I did not write the above post.)
I did—I disagreed with your response to thomblake, but what I wanted to say was to thomblake, so I replied to him. If we were in a room, I would be facing him* and gesturing at you only to indicate that I was not supporting your argument.
I apologize for not being clear.
Edit: And further for not reading—you said model. I do not disagree, it appears.
* “Him”, right? “Thom” sounds like a masculine.
Sure why not
Wait—is that “whatever pronoun you like, I don’t mind” or “that’s not the right deduction, but I am a man, so whatever”?
It took me a while to figure out how to answer this. Probably, “I don’t care and won’t object to you using that particular pronoun, and I generally don’t see how my gender is anybody else’s business.”
Thank you for telling me.
Or does that mean thomblake is Clippy?
Wait, don’t you also have a “I don’t care what pronoun you use for me” thing too?
“So depending on which features of reality you think are relevant, you can have multiple models of reality bringing out each of those features.”
Of course, but then, these theories don’t compete with each other. Everyone simply believes in both. I believe in both the atomic model of chemistry and the quantum model of electrons, and both can be used to describe the behavior of, say, a sugar molecule (but on different levels).
Loop quantum gravity actually does have at least one testable conclusion: non-constancy the speed of light at high energies. There’s even some support for this as of a couple of weeks ago, with measurements from the Fermi telescope.
A quick Internet search reveals that there is research being conducted to test this testable conclusion. (Full disclosure: I never understood special relativity very well, much less general relativity; I do not feel qualified to comment on the results described in the link, attractive though they are.)
Interesting. However, I think it’s reasonably plausible that this is only a “test” in the same sense that the Brans-Dicke theory is “testable”: namely, if the test doesn’t work out (as indeed it didn’t), you can just cross out “X >= 10^13” and write in “X >= 10^15″ (or whatever). I don’t know LQG well enough to know if this is the case, though.
If nothing else, such alterations are likely to make the theory weak as a meme—much like Arthur Eddington’s theories about the fine-structure constant. (By these arguments, it must be exactly 1/136! What, it’s closer to 1/137? Well, by these arguments it must be exactly 1/137!)
The particles (and the specific number of particles for each force) are a consequence of how the different forces are modelled. The local gauge groups for the electromagnetic, weak and strong forces are U(1), SU(2) and SU(3) (this is a simplification, especially for the weak force).
The one dimensional U(1) generates a single force carrier, the photon, the three dimensional SU(2) generates three (W+, W- and Z), while the eight dimensional SU(3) generates eight (superpositions of gluons and anti-gluons with three different colours).
The existence of the graviton is a consequence of assuming that gravity can be modelled, in any way (some people have suggested E_8 as the gauge group) similarly to the other forces. So the graviton is not aether; its existence is predicted, perfectly sensibly, by a lot of theories, and is vital to the working of those theories.
In contrast, physicist came to realise you could discard aether without affecting their theories. So the non-existence of aether had no consequences, while the non-existence fo the graviton would have.
Tom, I upvoted this because I’m a fan of the LW “physics for rationalists” series that Eliezer started two summers ago. Interesting stuff.
My uneducated intuition is that you’re right, and the graviton is the same naive-hypothesis-extension failure mode as aether.
But it’s interesting to note that these two meta instances of reasoning-by-similarity are apparently acceptable:
“Making a hypothesis about a different physical phenomenon by naively extending common properties of a group of well-understood physical phenomena is bad.” (paraphrase)
“Theories of physics have a known tendency towards elegance and simplicity.”
I just found this: http://fqxi.org/community/articles/display/129
You correctly decry popularity as a non-rational measure of veracity, but to the extent that it expresses a sort of straw poll, it may be a good indicator anyway. The idea of expert futures markets comes to mind.
My point is related: is it not also a fallacy to assert it’s GOT to be simple? That’s awful close to demanding (even believing?) something’s true because it ought to be, because we want it so bad. Occam’s razor has worked like a champ all these years but inference is risky and maybe now, we find ourselves confronted with some hard digging. I too hope some crystalline simplification will make everything make sense, but I don’t think we’ve a right to expect that, or should. What you and I want doesn’t matter.
The analogy between æther and graviton has some appeal. Photons, W and Z are more or less directly detectable. There is no hope of doing that with graviton. Hence, speaking about graviton brings probably no good: such language has no direct correspondence to observed reality and there is a danger that it can enforce some incorrect intuitions. As far as it goes, I agree. But note also that neither gluons are directly observed.
What I disagree is that the new theories are “graviton-based”. They are rather “gravitational-field-based”. The gravitational field is an observed entity. For other three forces, a field is associated to particles (well, W and Z and gluon fields are also not directly observed, but at least they are very useful concepts for making theoretical predictions, field theory is more successful than once popular analytic S-matrix theory which worked without the idea of fields). So, we may reasonably assume that the gravitational field also behaves like particles in quantum regime. Maybe it is not the case, but the particle-like behaviour isn’t the cornerstone of quantum gravity.
Also, why do you think that somebody has to be largely outside the academic system to solve the quantum gravity puzzle? Are there, in your opinion, some specific biases widespread in academia blocking the abilities to find a new successful theory?
Actually, E&M has been described as a warping of a 4+1-dimensional space-time (in combination with gravity). http://en.wikipedia.org/wiki/Kaluza–Klein_theory
It had some problems.
You know something impressive is happening when ‘4+1’ doesn’t equal ‘5’.
It’s common shorthand for 4 spatial dimensions, and 1 time-dimension, stressing that there is indeed a split, and that not all dimensions are equivalent. Perhaps, given the compactification of 1, it should be 3+1+1, but I’ve never seen that used.
Thanks, I wasn’t sure whether the +1 was the time or the compact dimension.
Generally, it means that one of the signs is reversed in the ‘inner product’.
Interesting. I think there are fundamental things we don’t know about gravity yet.
A basic physics question about light waves: I understand how water waves are a property of the medium. My analogy for how it works mechanically is potential energy and kinetic energy swapping in a (mostly) self-sustaining rhythm. I don’t understand light waves though. What causes them to oscillate?
Changes in electric fields induce magnetic fields. Changes in magnetic fields induce electric fields. If you combine the equations describing these effects, you get a wave equation. The wave described is light.
ETA: The section in the linked wikipedia article with the heading “In vacuum” has the equations in convenient form, and some links that look promising for describing this in more detail.
I really am grateful for JGWeissman for helping me click on the fact that light isn’t something that obeys the wave described by the Maxwell equation, but is that wave. The difference is imagining light as a type of substance compelled to oscillate with the wave pattern, and there being a wave pattern, resulting naturally from causal interactions, that is interpreted by our vision as “light”.
Thus this is the explanation I would give my past self for strikethrough(why light oscillates) what light is:
A charge creates an electromagnetic field. If the charge moves, the electromagnetic field will have to change. However, while the field is defined over infinite space, the field cannot update instantaneously over all of space. Instead, the field updates at the speed of light from the new position of the charge. At a small, fixed moment in time after the point charge has moved, the field has updated within a sphere of a certain radius, but has not yet updated outside this radius. What we call ‘light’ is the defect radiating outward though space like a ripple. When our eyes intercept this defect, we gain information about the point charge’s displacement and -- in some way I don’t understand, and don’t need to for the immediate explanation—the field no longer needs to keep updating and the ripple stops propagating (the waves collapses to an intercepted particle / photon).
So I no longer see light as a thing traveling though space, but as information about an updated field traveling in finite time.
Does this make sense? I suppose it could be completely wrong, but it is what I mean by a ‘mechanical’ explanation.
Oh, and I’ll add that light oscillates because the electric and magnetic fields update each other in finite time, and there is a slight lag, so that the wave has an amplitude. I see this as analogous to predator-prey oscillations in a Lotka-Volterra model; if the fields responded instantaneously there would be no oscillation.
This is a nit-pick, but the oscillation is not because there is any direct delay in the interaction between the electric and magnetic portions, it’s because the electric and magnetic portions effect each other through derivatives. This is similar to how the the acceleration (second time derivative of position) is directly related to position in any number of mechanical oscillators, such as springs, pendulums, and even circular orbits, when viewed right. For light, while there are still two time derivatives, they are coupled so that one time-derivative arises between magnetic and electric, and the other arises between electric and magnetic.
I don’t see where Byrnema claimed there was such a direct delay.
Ok, now I am wondering how I completely missed that last paragraph. I agree with your nit-pick.
It’s OK—it’s a matter of language, and not being very precise. Very loosely, in the case of a pendulum, you could say that in the upswing of the pendulum, it takes finite time (a delay) for the pendulum to respond to the downward force of gravity and start moving to 0. By the time it gets to 0, it already has momentum in the other direction and overshoots the equilibrium again. I see how this is the result of the dynamics being described by changes in the derivative of the motion, rather than—say—in the direction of motion itself.
Right. I’d describe that as a delay for gravity to finish overcoming the motion, rather than a delay in response.
There is no delay for the pendulum to respond to gravity, it starts accelerating immediately. There could be a delay before it achieves a velocity large enough to be perceived.
Most excellent. Now, glasshoppah, you are ready to lift the bowl of very hot red coals. Try this
That is broadly correct. The details of the propagation (and even of the electric field before the charge moves) can be derived from the local laws of how how charges and fields interact.
I know you said you don’t need to understand why the field doesn’t need to keep updating, but the way that you detect the light is that it moves a charged electron within a molecule in your eye, and that movement of the charged electron causes a light wave that (along the blocked path) approximately cancels the original wave.
Interesting. That has a nice symmetry, that to intercept light you need to move a charge in your eye that counters the original wave.
What I was referring to particularly was the quantum mechanical aspect: the wave propagates from the source in three dimensions—an expanding sphere of information. Yet as soon as your eye detects the light, the entire wave collapses into a particle. And this is instantaneous, with no delay.
But that’s QM, outside my pay grade.
Well, that is one of the many reasons that points to Many Worlds being superior to Collapse theories.
But if we are talking about light you see with your eye, that actually registers in your brain, there is way more than one photon emitted, and classical electromagnetic theory is a good enough approximation. Photons are roughly a discrete unit of amplitude of the light wave. For a high enough amplitude wave, you can ignore that it is discrete.
Many Worlds made sense to me as a solution when I considered the case of an apparently random choice. Instead of the world collapsing on an arbitrary choice, each world gets one choice. In the case of interaction with a propagating wave of light, though, I don’t see how it would work. Perhaps something is incorrect with my fledgling model of light.
Let’s consider a single photon. That would still propagate as a spherical wave from the source. The wave expands uniformly from the source and I suppose that according to a classical theory (?), that wave could be perceived simultaneously by different people in different places around the source. Even if I interact with the wave by moving an electron that approximately cancels that wave, then my cancellation would propagate only at the speed of light, not instantaneously.
So how would Many Worlds work in this case?
That is indeed what the classical theory says. It is wrong. This is where the assumption that domain of the amplitude is continuous is a bad approximation.
Quantum amplitude flows into separate configurations. For each detector, there is a configuration such that that detector was the only one to detect the photons. There are also configurations where no detector detected it. So, if in some configuration, a detector detects the photon, and goes to check on another detector, it will find that the other detector has not detected the photon, not because some instantaneous space spanning signal collapsed the wave function, but because in that configuration the photon did not go that way.
I see, so the photon left the source as a particle and the wave picture represents the idea that the particle could have been anywhere, until you know which world you’re in.
But the mechanical-model-that-made-me-so-happy was that the photon was actually just the electromagnetic field trying to update. The electromagnetic field would have to update isotropically … it couldn’t just update along the route to a given detector.
Well, there is a similar mechanical model to the evolution of the Schrodinger wave function, which is to particles (including photons) as the electric and magnetic fields are to light in the classical model. This wave function is fundamental, the particles, and the configurations, or “worlds” are derived consequences.
Well, it you believe in a collapse postulate. Which I don’t think many people around here do.
You know, I didn’t even know that was the same thing as the collapse postulate. So when people talk about the ‘collapse of the wave function’, they’re talking about—for example—the perception of light. OK, sure, that makes sense.
So our solution to that was Many Worlds…
Nothing basic about that question. It may not even be a meaningful thing to ask.
I agree it may not be so basic—I added that it in because I wasn’t sure.
But I feel I understand things when I understand them mechanically and locally. If the universe doesn’t actually work that way, I’m at a loss.
(And if it’s not understood that way—though I would bet someone does—I would think that to some extent, we don’t really fully understand it yet.)
I think understanding the universe in terms of concepts and intuitions that humans developed innately is unlikely to be possible. No reason the ways of thinking about the world helpful in the ancestral environment have to be helpful for describing the fundamental nature of reality.
Wow. You and I have had this type of discussion at least once before here on Less Wrong, about whether we ‘really’ understand something (for example, gravity) or if we understand it ‘well enough’. I suspected an underlying difference in the way we were thinking about things.
I don’t believe this: that there are physically realized things that we can’t understand.
I think that our concepts and intuitions are flexible enough to accommodate any possible reality. Quantum mechanics, and even light, are really weird. But there’s still hope for an aether—or whatever is required for this mechanical/local understanding I’m talking about—to bring it back down to human comprehension. The fact that these things are mathematically coherent (explicitly and fully described by equations) is especially compelling, since you can interrogate the equations to build structures in your mind that would model it.
For me, so far, the Maxwell equations are just floating in the air with no physical structural basis. However, if you spent time with them, wouldn’t you start building a physical intuition about them?
Let me see if I understand your claim:
Let physics be described by whatever math is necessary. You predict that human general intelligence will be flexible enough to understand it. If someone is consciously “doing math” rather than “applying intuition”, then they don’t understand it yet. But the solution to that may involve growing an intuition, rather than changing the math.
That… sounds reasonable enough. But I question whether the intuition always comes in the form of a mechanism, or even any additional concepts at all.
I would only qualify my earlier statement: while human intelligence is flexible enough to understand anything that is possible, it might not be large enough. If there’s too much going on, the brain may simply not be able to compute it. In which case, the non-understanding doesn’t feel non-intuitive, it just feels too complicated.
Even correct intuition? I guess I don’t mind putting forth a more definitive assertion that intuition must be based on a mechanical understanding. (While it’s likely I’m wildly guilty of the typical mind fallacy, that’s nevertheless my view.)
I’ve been considering the hypothesis that mathematical intuition (especially intuition about highly abstract, non-physical things) comes from an ability to model that math physically in the brain. When we interrogate our ‘intuition’, we’re actually interrogating these (mechanical) models. Modeling is a high-intelligence activity, and a model ‘correct enough’ to yield intuition may be hardly recognizable as such, if we were forced to explain in detail how we knew.
(If we have a correct intuition about mathematics outside our experience, how else could we have it?)
New post?
This is correct, but there is a useful layer of abstraction to consider. There are a set of operations the brain does that we can be conscious of doing, and inspect the structure of how they interact within our own brains. These operations are, of course, implemented by physics, they come from the structure of neurons and other supporting biological components. And therefore, the structures that are built out of these operations are also, ultimately, implemented by physics, though a lot can be learned by looking at the introspectively observable structure. These operations can be used to build a model of arithmetic. This does give us some power to “explain in detail how we knew”.
Perhaps you can elaborate? My first instinct was to reply as JGWeissman did, but I felt a slight sense of disquiet about doing so—it would be interesting to find my sense of disquiet justified.
So JGWeissman explained why we describe light as a wave. The behavior of light can be accurately modeled with a certain kind of equation, this is the same kind of equation we use to describe traditional waves. But that doesn’t seem to answer byrnema’s question “What causes them to oscillate?”. Some interpretations won’t even leave us with waves and I don’t know if there can be causal histories of wave oscillation when the wave doesn’t have a medium. I guess you could generate an EM wave and go “Look!” but it seemed like byrnema was asking something like “why does light move like a wave” which is more or less unanswerable right now and probably will be for some time.
You’re right. I was looking for a mechanical/local answer. (I have been since high school physics when I completely rejected the way magnetism was explained, only to find that was the way it is always explained).
If I found such an answer here—and I wouldn’t be totally surprised if I did—I would use that answer about light to try and maybe finally understand electromagnetism better.
What do you mean “mechanical/local”? Maxwell’s equation’s (in what I consider their most fundamental form) describe how the electric and magnetic fields change at any point based on derivatives of those fields, the charge, and the movement of charges at that point. That is certainly local. And, together with the laws describing the forces experienced by charges in the fields (also local), they mechanically describe how a system will evolve.
OK, I’ll update again that it is just me that doesn’t understand the light wave mechanically. (I didn’t mean this facetiously—I know light is really strange for everyone, I just meant that I can’t seem to possibly understand it.)
Do you mean that you do not understand Maxwell’s equations? That you don’t understand how the wave mechanics are derived from Maxwell’s equations? Or that some part of this is not “mechanical”?
I’m beginning to suspect that my insistence on understanding things a certain way is peculiar and possibly overly narrow.
But suppose that I look outside my window and see a light wobbling across the horizon. I think that I don’t understand the trajectory of the light. Then someone finds the equation that describes the motion of the light: it turns out to be perfectly regular and periodic. I still feel I don’t understand it. For me, the path of the light isn’t understood until you discover that the light is a reflector attached to the wheel of a car, and the trajectory you see is a combination of the car’s linear movement and wheel’s rotation. This is what I mean by a mechanical understanding.
I could study Maxwell’s equations, but I know they wouldn’t help. Do we have a ‘mechanical’ understanding of the motion of light, or just the equation description?
Are you saying you want to understand light as being made out of components that behave like the macroscopic objects you are used to interacting with? That isn’t going to happen, because light does not work that way.
I don’t mind if light behaves in different ways than I’m used to, but I still expect that these ways are causally dependent upon other things. Especially with a spatial pattern, I expect that any pattern produced by certain geometric rules can be reproduced by a model of those rules.
Even at the macroscopic scale—if a model is possible (physically realizable) at that scale. If it is not possible, I would have to spend a lot of mental energy modeling those rules mentally, but that would still lead to mechanical understanding. My problem is that I haven’t heard (or don’t believe I’ve heard) exactly what rules should be modeled.
Earlier I said (emphasis added):
Would it make more sense if I said:
Um. I’m having one of those I-can’t-believe-I’ve-been-this-stupid-over-the-last-ten-years moments.
I went back and reread what you wrote and the part I missed before was this:
So it isn’t that light “happens to follow” this wave equation. That wave equation IS light—that is, that specific interaction between the electric and magnetic fields is light.
Honestly, I’d never thought of it that way before. I can go back to that chapter in electromagnetism and see if I understand things differently now.
I look at the light bulb on my desk and I wouldn’t even call it ‘light’ anymore. It is electromagnetic interaction.
I photographically recall the poster over an exhibit at a science museum, “Light Is Electromagnetic Radiation’. I thought that meant that light was radiation (obviously, it radiates) that was associated in some way with electromagnetic theory and I remember thinking it was a decidedly unpleasant verbal construction.
I’m thankful, and sorry...
You know, this really calls for a cartoon-y cliche “light bulb turning on” appearing over byrnema’s head.
It’s interesting the little connections that are so hard to make but seem simple in retrospect. I give it a day or so before you start having trouble remembering what it was like to not see that idea, and a week or so until it seems like the most obvious, natural concept in the world (which you’ll be unable to explain clearly to anyone who doesn’t get it, of course).
Seriously. Apparently, I wrote the key insight she needed (not knowing that it was the missing insight), but she didn’t click on it the first time, and then, as I am asking questions to try to narrow down what the confusion is, something I said, as a side effect, prompted her to read that insight again and she got it. Now, how can one systematically replicate a win like that?
Be polite and patient when people are confused?
Well, that is important. But that is more part of not automatically failing, than actually making progress towards dispelling the confusion.
Lacking understanding, that’s the best advice I can give.
I remember you and I also discussed what it means to understand something, and I definitely sympathize and largely agree with your standard for what counts as “understanding”. (I’ll find the link to that discussion when I get a chance.)
My standard is that you understand something if and to the extent that:
1) You have a mathematical model that generates the observations with good success. (Not necessary here what labels you use—this part can be “Chinese room”-ish.)
2) That model is deeply connected (via the entities it shares, quantities it uses, mutual interaction, etc.) to your model for everything else, and thus connected, ultimately, to your intuitive (raw, qualia-laden) model of the world.
Correct me if I’m wrong, but I think this is where you are: for light, you understand it in the sense of meeting 1), but don’t meet it with respect to 2). Would you say that is accurate?
I am not certain if this is going to be helpful, but I’m going to try re-expressing what the equations describe, in non-mathy words. It seems to me that the local understanding you are looking for actually is contained in Maxwell’s equations, and that you are blocking on something in that description. Of course I could be quite mistaken in this; it’s hard to understand a multi-year confusion based on a few forum posts. So if I’m not helpful here, sorry!
Let me start with an antenna; that is, a straight, electricity-conducting piece of metal. I run an electric current through it, or more accurately, I start running a current. As I’m doing this, the electric field within the antenna is changing; electrons are moving around, spitting out photons, and generally changing their state, heading towards the steady movement they’ll have when the current is fully established. Before they can get there, however, I perversely and maliciously reverse the current’s direction, and the electrons scramble to attain a completely different steady state! In fact, at no time in the following is the antenna going to be in an equilibrium state; I, the experimenter, am constantly changing its condition and keeping the electrons hopping.
Now, when I change the electric field in this manner, that change causes a magnetic field to arise. It seems possible to me that this step is the cause of your confusion, so I’ll digress a bit: Why does a changing electric field cause a magnetic field? Maxwell’s equations do not say anything about the causation; they merely quantify the observed fact. Studying Maxwell does not give you any greater understanding of the causation than you would have from the good old 1830s experiment of running a current through a wire and seeing a nearby compass needle deflected; all it does is to allow you to calculate how much deflection to expect. I often see this sort of confusion in the way basic physics is taught; because the equations are the full description of what’s going on, people expect them also to contain the full understanding at the causal level. So there is confusion about Maxwell, and also about special relativity; people ask “What does it mean that the time-direction’s sign is reversed in the inner product?”, and the answer is that it describes the way matter behaves, but the causality is much deeper. It may be a mistake to teach Newton before anything else in physics, just because F=ma is so intuitively clear; we all see that this is just a formalisation of the way rocks behave. Throw the rock harder and it hits the other monkey faster, causing more damage: Our brains are well adapted to this piece of physics! But it does us a disservice in studying other equations, because we expect them to be similarly clear and contain a similar causal-level explanation, and they just don’t.
At any rate, then, the crucial point is that when I move the electrons, they cause a magnetic field to exist. If you look deeply enough into QED, you can find an explanation of this in the way the force-carrier photons are moving, but personally I am quite unable to visualise this; all I can do is go through the math that shows Maxwell’s equations coming out as the classical limit of QED. (Well, anyway, I could do it for an exam some years ago.) However, it may be helpful to visualise it like so: When I accelerate the electrons, the virtual photons that they spat out a few nanoseconds ago (messengers for their electric field) are ‘unable to return home’ (home having moved) and must find something else to do; the something else is to interact with other particles, which we measure as a magnetic field.
Now, because I’m varying the acceleration of the electrons, the size of the magnetic field caused by their acceleration is also changing. And a varying magnetic field… causes an electric field. It is inaccurate, but possibly helpful, to view this as being caused by the aforementioned no-longer-so-virtual photons interacting with electrons in the quantum foam and moving them around, creating a momentary polarisation of space.
So now there is a changing electric field not just at the antenna where I’m doing my thing with the electric current, but also some distance away where the resulting magnetic field is causing a reflection of that process. Rinse and repeat: This electric field’s changing causes a magnetic field over here, which causes… You will get a chain of electric/magnetic fields running across the entire universe. This is what is meant by a ‘light wave’.
Maxwell allows us to describe in numbers what I just described in words, but the causal understanding is all in the observed fact that a changing (not a steady) electric current causes a compass needle to deflect. You can take this as unadorned, experimental observation, “We don’t know why that happens”, or you can try to visualise it in terms of messenger photons as I outlined above. I hope that helps.
Thank you for this explanation, it includes exactly the sort of mechanical explanations I was looking for. Not necessarily simple or easy to understand, but about entities with certain properties interacting in certain ways.
For example, this is a mechanical explanation. It’s probably somewhat inaccurate due to being expressed verbally, but a person can then turn to the equations to get the detailed, accurate picture.
I think I would have done so much better in physics if they had explained the causality with the equations. For me, now, understanding that light is the interaction of the electric and magnetic fields has been a wonderful paradigm shift in way I view things. I had already assimilated that most things I experience are electrostatic—that a table is non-compressible and the chair supports me because of electrons and their interactions. And now my idea of light has changed—I no longer see it as an independent substance reaching me, but as information about the change in an electric field propagating towards me at a fixed speed from whatever sources are emitting or reflecting the light.
My ideas are topsy-turvy at the moment, it will take some time and reading for them to settle in any accurate way. But I’ll share some of my first-day thoughts . For example: A room full of light is ‘bright’ because the light contains so much information. And: it seems amazing that visible light is so faithful (non-noisy) when you think of it as a wave propagating in all directions; only a particle at the moment of observation.
I find that your bright room-observation and an analogy with low-light photography almost helps me grasp how come the world seems classical despite its quantum “underwear”:
Think of a (digital) photograph taken in very low light. Something like this.
Think of being in a pitch black room, and imagine your eyes are perfectly sensitive. You have a flashlight pointed away from you.
Now imagine the flashlight is very dim: it only sends one photon every few seconds, by moving one of its electrons; the electric field propagates at c as a spherical wavefront to “notify the universe” of the change. Whenever this front passes through another electron (say, of the wall before you), it may be absorbed. If it’s not absorbed, the electron does nothing (in a way, it doesn’t care that the first electron changed position). However, if it is absorbed (the two electrons exchanged a photon), then the absorbing electron now changes position to conform to the new information; doing so starts another wave-front communicating this information. Note that this second wave-front is not just going back, it’s still a spherical front starting from the wall electron. Again it starts propagating through space until it hits another electron. Suppose this last electron is used by your eye as a detector. Then you just noticed a tiny “flash” of light somewhere in your vision field.
Note that the two absorptions happen (until now) randomly. If you use just that flashlight you’ll only see random flashes in the room. More precisely, in the many-worlds interpretations, from a single train of emission-absorption-reemission-detection “events”, each you in every world will see the tiny flash in random parts of their vision field. This is both because different wall electrons would have done the absorption-reemission, and because of different electrons in your eyes would do the detection. You can extend this metaphor a bit more: in worlds where the re-emitted photon didn’t hit your eye, but (say) another electron deeper in the wall, that world’s version of you won’t notice anything. (The absorption-reemission chain will just bounce electrons and nuclei randomly throughout the wall, which just means heat.)
OK, here’s the “brightness” part: even though the absorption-reemission electron is randomly chosen by each world, not all of them are equally likely to be chosen. The wave-functions of the particles, and the interactions of those wave-functions as given by QM equations, cause the distribution of “picks” to have a certain “shape”.
Imagine that you take your flash-light and you increase its brightness; say it sends 10, then 100, then a million photons at the same time. You’ll start seeing several flashes, from random positions in your field of vision. (Each of you in the many-worlds will see the flashes coming from different places.) But the wave-function says how likely it is that you’ll see flashes from different places: even if each version of you sees different flashes, each of them will see more flashes coming from bright (white) objects than from dark (black) objects. As a result, you’ll see a grainy image of room in all worlds, even though a different one (that is, with the grain positioned differently) in each. The more photons your flashlight sends, the more “smooth” your image will be, converging on the “shape” of the wave-function. The image above shows a noisy image obtained in low light (although for normal cameras the source of noise is different).
Imagine you take several consecutive photos in those conditions; each image will be very grainy and dim, and the position of grain will vary among your many-world alternates. However, if you combine your successive photos in one, you’ll accumulate a brighter, clearer image (the more so as you add more photos). Each many-world version of you will get a different one, but they’ll converge to the same: the shape of the world’s wave-function. (Of course, the different worlds will eventually diverge in shape, too.)
You can stretch this analogy to visualize all sorts of interactions. For bright objects the re-emission is more likely to occur “towards” the outside of the object (or, inversely, electrons within bright objects tend to not communicate between themselves the news about outside). For dark objects it’s the other way around: photons are more likely to be passed among the object’s electrons rather than towards you.
Or take diffraction: A photon is emitted, passes through a screen with two holes (thus, it’s not absorbed by its electrons), and hits a wall. Even if you send photons one-by-one, they’ll still form a diffraction pattern on the wall. You can imagine it this way: the initial emission (the wave-front carrying the message) is spherical; when it hits the screen, either it’s absorbed (the message was passed to the screen), and you see nothing), or not: the message passed through both holes. But from each hole the message continues to propagate in a sphere centered on that hole. When the “message” hits the detector area, it is both these spheres that hit; depending on the difference in distance of the two paths, in some areas the two “copies” of the message can contradict each other or agree; you’ll have successful detection (a movement of an electron in the detector wall) only where the two copies agree, thus forming the interference pattern—but, for a single photon, exactly what point of the interference pattern will be hit is random.
Nitpick: s/similar causal-level explanation/similar intuitive appearance of a causal-level explanation/
I actually think this is a misleading characterization. Those inconsistencies with resolved by positing that physical bodies contract as they move through the aether. Lorentz even gave us the math to show us how this worked. Then Einstein posited a theory empirically equivalent to Lorentz’s. People liked Einstein’s theory more for various reasons (good reasons, I think) but aether has never been ruled out in the way, say, phlogiston has been ruled out.
I don’t think the fact that all our false theories have been elegant and simple could possibly count as evidence that reality is, in fact, elegant and simple. And certainly those concerns aren’t reasons to reject an approach that has been mind-blowingly accurate in it’s predictions.
Is it really? That is depressing. Though I don’t at all have any problem with theories being popular or unpopular so long as the empirically invalid theories are always really unpopular. Why would there be something wrong with this?
Try jumping off the earth. :-)
Sure if you don’t mind having two totally different and wildly incompatible theories to describe reality.
While I think string theory is a mess too I don’t think I see evidence here that all quantum gravity approaches are flawed.
“People liked Einstein’s theory more for various reasons (good reasons, I think) but aether has never been ruled out in the way, say, phlogiston has been ruled out.”
It wasn’t ruled out in 1905, but it was as soon as quantum mechanics was invented. We really do know how light works nowadays, and it’s very, very definitely not disturbances in some sort of material substance.
“I don’t think the fact that all our false theories have been elegant and simple could possibly count as evidence that reality is, in fact, elegant and simple.”
Why not? You know that all those “false” theories were accurate up to the seventh or eighth decimal place, right?
“Why would there be something wrong with this?”
Because coupling theories about how the universe works to political factions is a HUGE impediment to discovering the truth. See http://lesswrong.com/lw/gw/politics_is_the_mindkiller/.
“Try jumping off the earth. :-)”
If you don’t understand the difference between observing that gravity exists, and observing that there are particles which obey the same laws of quantum mechanics as photons and electrons, which are massless and have spin 2, and are responsible for gravity, you should go back and read some of the earlier stuff on this blog.
“Sure if you don’t mind having two totally different and wildly incompatible theories to describe reality.”
The whole point is that saying “suppose gravity is transmitted by carrier particles called gravitons” starts off by completely ignoring all of GR, and then tries to patch it up afterwards (with little if any success). Newer theories should supersede older theories, not ignore them.
“” < >
Wait, a theory was falsified by the invention of another theory?
Why put false in quotes? And, well, thats some shady induction. So you had a simple theory that explained some class of phenomena. Now there is some other phenomena that your theory fails to explain, why should simplicity be conserved here?
Well yes, people deciding which approach to work on for political reasons is bad. But the existence of competing approaches with varying degrees of popularity does not entail that. Indeed, if people are deciding which approaches to work on for political reasons only having one approach would be a lot more troubling.
The smiley face was supposed to indicate that I was joking.
This does sound like a good heuristic to work with.
“Wait, a theory was falsified by the invention of another theory?”
One that’s been confirmed against hundreds of predictions to over a dozen decimal places? Yes.
“Why put false in quotes?”
Because there’s a HUGE distinction between a theory like Newtonian gravitation and a “theory” like phlogiston, even if they’re both “false”.
“So you had a simple theory that explained some class of phenomena. Now there is some other phenomena that your theory fails to explain, why should simplicity be conserved here?”
I’m saying, “Successful theory of physics A was simple, and theory B was simple, and C was simple, and D, and E, and …. , but there’s a new class of phenomena which needs a new theory, so this theory will probably also be simple.”
“But the existence of competing approaches with varying degrees of popularity does not entail that.”
“Approach” here is a HUGE misnomer. “Approach” is a term commonly used in engineering to mean “different ways of accomplishing goal X”. You can build a machine in manner A to do X, or manner B. This ABSOLUTELY DOES NOT generalize to science, because there’s only ever one reality. If you have theory A and theory B both purporting to explain some phenomenon X, either A or B must be wrong, while in engineering sometimes there are ten different ways of attacking a problem, depending on what your goals are.
(Or equivalent in a way you haven’t understood yet.)
Is there a single example of this that you can think of? There are the different ways of computing classical mechanics (Newtonian, Lagrangian, Hamiltonian), but these were known to be just different ways of doing the same math at the time of discovery.
The original formulations of QM were famously shown to be equivalent, though I’m not sure they were ever expected to be incompatible. QFT and S-matrix theory were originally politically opposed (though QFT produces an S-matrix), but I have heard that recently people advocate widening QFT to the point that it appears to cover all of S-matrix theory.
No, it’s just a theoretical property of the ‘A, B, X’ abstraction that you mention. In fact, it would not surprise me if the general problem of proving whether or not two theories are exactly equivalent is intractable in a similar way to the halting problem.
I was just clarifying, you said that it was ruled out “as soon as quantum mechanics was invented” which clearly isn’t right. Anyway, I had just meant LET versus SR, but what exactly is the experimental evidence against an aether? Obviously QM gets along pretty well without it and the vocabulary is incompatible… but the vocabulary of QM is incompatible with the vocabulary relativity as well.
Maybe, but they’re still both false! What exactly is the distinction you have in mind?
This would be a lot more convincing if the most recent and most successful theory of physics weren’t such a glaring counter-example. We count on induction because we have no other option but this kind of thing is meta-induction (that changes in our understanding of regularities has regularities) hasn’t been justified enough to make it a tool in eliminating approaches.
I disagree. Even under the naive theory of truth that is popular here we don’t know in advance which theoretical apparatus will yield the right explanation for phenomenon X. While that is still an open question people might come to different conclusions about which way is most promising. As agreement here seems unlikely it makes sense to just have physicists work in the areas they think will be most fruitful. And then once you notice that theories of physics have this nasty habit of turning out false… well then I don’t even know what you’re using to declare A right and B wrong. Yeah “A or B must be wrong” but that is a seriously inclusive or. When we don’t know which theory is right, when they’re both probably wrong or if that turns out to be a nonsense question since they’re empirically equivalent it then makes a lot of sense to think about the benefits working under different sets of theoretical assumptions (i.e. approaches). That isn’t a prescription for anything goes. Some approaches are stupid and laughable. Others are powerful and clever. So I think “approach” nomes just fine.
(fyi, the downvote isn’t mine)
“Maybe, but they’re still both false! What exactly is the distinction you have in mind?”
Yes, and me and the Pacific Ocean are both more than 50% water by mass. Newtonian gravitation successfully explained a huge number of phenomena. Phlogiston did not.
“Anyway, I had just meant LET versus SR, but what exactly is the experimental evidence against an aether?”
With quantum mechanics (and modern experimental technology), we can actually look down below the level of individual particles, and we have found that photons are actually their own particles, not patterns of vibration (or whatever) within other particles.
“This would be a lot more convincing if the most recent and most successful theory of physics weren’t such a glaring counter-example.”
You mean quantum mechanics? Quantum mechanics is very elegant, it’s just usually explained badly. See http://www.amazon.com/QED-Strange-Theory-Light-Matter/dp/0691024170.
“I disagree. Even under the naive theory of truth that is popular here”
Would you care to propose some alternative theory of truth?
“While that is still an open question people might come to different conclusions about which way is most promising.”
Obviously, but that disagreement should then be resolved by reference to experiment. There is no room for persistent disagreement. In engineering, you can have five different methods, each with their own advantages and disadvantages, and this is a stable state. In science, having five different theories is not a stable state; it needs to be resolved, rather than harden into different factions.
“And then once you notice that theories of physics have this nasty habit of turning out false… well then I don’t even know what you’re using to declare A right and B wrong.”
Experimental evidence?
“it then makes a lot of sense to think about the benefits working under different sets of theoretical assumptions (i.e. approaches).”
What does that even mean? How would you apply that to a theory of physics (past or present)? What “theoretical assumptions” are involved in, say, Special Relativity? Special Relativity makes the assertion that the speed of light is constant regardless of reference frame, but this isn’t just a mathematical axiom that you can pick up and discard at will; it is based on a huge pile of experimental evidence.
I wouldn’t be so sure. (http://www.jimloy.com/physics/phlogstn.htm) But it certainly had other problems.
We knew this before quantum mechanics. Lorentz’s aether wasn’t matter.
I’ve read (indeed, I own) this book. I don’t know how to evaluate elegance, but the Standard Model particle zoo isn’t simple, at least not in the way Newton or special relativity is simple. I wish you would assume I have some idea what I’m talking about and that my concerns and questions might be well motivated, however alien they seem to you.
I would like too but I haven’t really figured out what I think. I’m not particularly radical, but I think the epistemology Eliezer has laid out on Less Wrong has some holes, leaves important questions unanswered etc. Maybe one day I’ll write something.
But they aren’t always resolved that way! Special relativity beat out Lorentzian Ether Theory even though they are empirically equivalent. Obviously if you have two theories that predict different outcomes of some feasible experiment you can run that experiment and resolve the difference. The question is, what we do when there is no experiment to run? Well we do some math and try to come up with testable hypotheses. Using different theories or different vocabulary seems to affect how easy it is to do the math and generate hypotheses.
As the sentence immediately following this one was supposed to indicate, usually both theories are wrong in the long run.
It isn’t based on any experimental evidence that distinguishes it from a theory that says mass contracts in the direction it moves in. But it turns out that if you start from a theory which says’s light’s speed is constant you can come up with things like the theory of General Relativity. Alternately, you might have two theories that describe totally different phenomena without error but have the potential to describe things about other phenomena and eventually one might be subsumed under the other. But it won’t always be obvious which theory is the more fundamental one. I suspect one reason there is a lot of work done trying to incorporate gravity into quantum mechanics rather than the other three fundamental forces into General Relativity is that SR/GR just doesn’t have the vocabulary to make hypotheses about particle physics. The former is sort of obvious though “Oh there is another force, there must be this other wavicle: a graviton.” That doesn’t mean it will be an successful approach but that is part of the reason it is the popular one.
These two theories are not equivalent at all; they predict different outcomes for the Michelson-Morley experiment, unless you patch up the ether by requiring it to be at rest relative to the Earth at all times.
Could you be missing the “Lorentzian” there?
What FAWS said. The problem is solved with length contraction.
Would the person or persons down voting my comments here mind at least explaining this one? Afaict I’m just correcting someone on a factual error and providing a citation. If wikipedia is wrong on this matter I’d like to know.
The problem with that solution is that you end up with a theory that makes the same predictions as SR, but contains an extra concept that plays no role in making predictions: a distinguished but unobservable reference frame of absolute rest, with respect to which all the other reference frames involve time dilations and length contractions. When you cut off that useless spinning cog you are left with SR.
Which one is that? Do you mean naive in the sense that it is ‘unsophisticated’ or do you mean ‘actually wrong because it is too simple’? If the latter you could well be right but I’d like more information.
Naive as in “This essay is meant to restore a naive view of truth.”.
I do think “A true theory is a map that corresponds to the territory” is right as far as it goes. But there are going to be people who will ask things like “What the hell do you mean by territory?” and “How do you have any idea if your map corresponds to the territory?”. I don’t think those are wrong questions and I think answering them might require a little more work.
That said I might be missing where people are at on this because this line, from the wiki, is exactly right:
I think stopping there is about right but “reality” tends to get loaded with a bunch of additional properties. I think attributing features of your theory to reality other than it’s experimental predictions is a kind of map-territory confusion that nearly everyone still falls for, so I think looking at two empirically equivalent theories and saying one is true and the other is false can’t just mean “one matches the territory and the other doesn’t”. So either we say of these theories that they are both true but one is better for reasons other than truth or we say that truth involves something other than just corresponding to the territory.
A related issue is that it is unlikely the core concepts we need to state any theory themselves correspond exactly to an external world. A mathematical description is fine by itself but we always feel that merely stating the math is some how insufficient so we end up reifying our variables. At least when the object of the theory is significantly divorced from the environment that gave rise to the concepts used to state the theory it is extremely unlikely that these concepts map exactly to things in the the world. This doesn’t always change our predictions but it suggests that of two empirically identical theories neither is likely to conceptually correspond to an external world (the external world, if there is one, isn’t made up of concepts) but both will correspond empirically. So then how can we contrast these theories on the basis of their correspondence to reality?
… well, maybe...
Well I don’t know if “the external world is made of concepts” is a meaningful thing to say. It strikes me as a kind of category error. But yes, I can always be wrong. I actually had a hedge in there original but took it out for stylistic reasons since it was already a parenthetical.
It probably depends on what we mean by ‘concept’. The laws of physics can probably be described as concepts and if we knew all the laws of physics the concepts would then be quite likely to include the entire universe as artifacts. But yes, parenthetical to the extreme.
Eliezer Yudkowsky stated that he “meant to restore a naive view of truth” in his essay The Simple Truth—I believe the more technical term is “correspondence theory of truth”.
As far as I can tell, people like SR because Einstein produced GR. That is not a terrible reason, but it seems to me that they rewrote history to exaggerate the differences, even though contemporaries could barely tell the difference and tended to like them all or dislike them all, and more the latter.
I was very surprised when the article h-H found yesterday spent so much time on Minkowski.
ETA: what I meant was that people praise SR because the same guy produced GR (ad hominem!) and not for any object-level reason. That is very weak evidence that SR was a better theory and a necessary stepping stone. History has been rewritten to claim that, but I am skeptical.
This is definitely a big reason. As I understand it the other thing is that descriptions of electric current generation look silly with Lorentz because you need different vocabulary depending on whether it is the coil or magnet that you’re moving. Identical descriptions for identical phenomena is a nice thing to have.
I guess faint praise is not well conveyed in writing. Hero worship is a bad reason. And I am nervous to draw any conclusions from hagiography.
Ohhh. I see what you’re saying. Maybe there was a kind of hero worship… But a big reason, and the reason I took you to be giving, for preferring SR over Lorentzian aether is that SR makes GR possible. And there is no equivalent theory based off of aether theory. So the thinking (as I understand it) isn’t Einstein produced SR and GR, GR is brilliant and true therefore so is SR. Rather, SR is pretty good and better than Lorentzian aether because from it Einstein produced GR which is brilliant and true.
Einstein’s SR and GR work were actually looked on with huge suspicion for a while. It was his work on Brownian motion and quantization evidence from the photoelectric effect that were originally so warmly welcomed.
Can you say more? My understanding was that SR was picked up pretty quickly for more or less the reasons Einstein preferred it. But I’m not that familiar with the history.
Not without actually doing research—I’m trying to speak at the level of generalities. It’s perhaps wrong to conflate the reactions to SR and GR, and I’m possibly overstating how welcoming the community was to the idea of photons being quantized, but I think there is a reason his Nobel prize was given “for his services to Theoretical Physics, and especially for his discovery of the law of the photoelectric effect”, with no explicit mention for either SR or GR.
I thought that was more about the committee being cautious and favoring experimental evidence over abstract theorizing. You don’t want to give someone a Nobel prize and have them quickly turn out wrong.
If it were true that SR were necessary for GR, then SR would be a tremendous improvement on its competitors. really! I get that! But I don’t see much reason to believe it. It looks like hagiography (=history rewritten because of hero-worship; probably FALSE) to me. I have not read Poincare, but I am skeptical his take was inadequate. Minkowski seems a big improvement over everyone else. If it took SR to produce Minkowski, then that’s good for SR. Minkowski claims to be in Einstein’s tradition. But maybe that’s just because they’re both German. (nationality was a pretty good predictor of people’s views on the matter; I think nationalism was a big part of it.)
I’ll think about this some and come back. The whole history is fascinating because the way SR and GR are usually explained (on television and in other popular treatments) it is as if Einstein was just sitting around and used his genius to invent it all whole cloth. So we have all this mythology about what a genius the guy was but the actual story looks like it is more about the collective power of science, of the process, than of one man’s brilliance. Given what Lorentz and Poincare accomplished it seems pretty plausible that it wouldn’t have taken more than 10-15 years for someone else to come up with SR- it probably would have been even less. Maybe the retrospective view doesn’t convey scientific revolutions well but I don’t even get the sense the Einstein was especially smarter than his contemporaries.
It seems to be a common view among phycisists that SR someone else would have come up with sooner or later (probably sooner), but GR required a critical insight so rare that had Einstein not existed, we might still not have an adequate theory of gravitation.
Once David Hilbert became aware of the problem, he almost beat Einstein to the punch.
ETA: Actually, looking at the history, it seems that Hilbert’s interest in physics and mathematical prowess is not evidence that he could have come up with the necessary physical insight. He didn’t become interested in GR until well after Einstein had laid the groundwork.