My lecturer says "Special relativity is absolutely wrong"

[Jonathan Scott]

Note that FTL communication in a specific frame of reference does not necessary directly lead to causality problems, although it would directly contradicts SR because it requires a preferred frame. Some interpretations of non-locality for QM entanglement effectively rely on this concept. The communication path transforms between frames like a spacelike separation, and for fast moving frames the effective speed in one direction may be only fractionally greater than the speed of light.

If it were possible to communicate faster than light by at least a minimum fixed amount in all frames, that would lead directly to causality problems, as it would make it possible to receive a reply to something before it was sent.

 

[Orodruin]

My point exactly,... I should have specified that his lecturer should confirm his belief that neutrinos were FTL via a repeatable experiment...lol That was my attempt at humor and sarcasm.

If you read the thread, this is not his lecturer's belief.

 

[gjonesy]

In an assignment I wrote for university I was penalised for claiming that FTL neutrinos would violate special relativity.

The OP's opening statement was that his lecture believe that the 2011 OPERA experiment was correct for it's observation wrong in some other unspecified way and that SR was absolutely wrong? Isn't that what the thread is about? Basing his own personal conclusion that SR is wrong because of that one OPERA experiment even though it was later retracted because the experiment was flawed? what am I missing?

If you read the thread, this is not his lecturer's belief.

what am I missing?

 

[jbriggs444]

The OP's opening statement was that his lecture believe that the 2011 OPERA experiment was correct for it's observation wrong in some other unspecified way and that SR was absolutely wrong? Isn't that what the thread is about? Basing his own personal conclusion that SR is wrong because of that one OPERA experiment even though it was later retracted because the experiment was flawed? what am I missing?

The attachment on the original post makes the professor's viewpoint clear -- that SR is "absolutely wrong", having been superseded by general relativity, that general relativity does not forbid faster than light travel and that the Opera result would have violated SR, but that does not matter since SR is "absolutely wrong". The professor's contention is that any objection to the plausibility of the Opera result that is grounded in SR is a non-starter precisely because SR is known to be "absolutely wrong" -- that any such objection must be couched in terms of GR for it to have validity.

Of course, that's hogwash.

 

[gjonesy]

Of course, that's hogwash.

Agreed

I misunderstood the reason for the penalization and thought it was for specifically sighting the OPERA experiment. How can you give someone a bad grade for being right?

Just to clarify before I get schooled again regardless of the lecturer's personal belief that (SR is wrong) according to SR FTL neutrinos would be a violation and therefore it would be a correct textbook answer.

 

[Orodruin]

Agreed

I misunderstood the reason for the penalization and thought it was for specifically sighting the OPERA experiment. How can you give someone a bad grade for being right?

It is easy, you just have to be wrong.

 

[ChrisVer]

Well, note that massive particles traveling FTL clearly violates SR

you mean particles with positive mass-squared traveling FTL could violate SR... because as far as I know SR doesn't necessary tell you that the mass squared has to be positive or negative for it to work. In fact Lorentz transformations can be used for both spacelike and timelike 4vectors (I guess).

Thus, kinematic SR states emphatically: massive particles travel < c, massless particles travel at c, and imaginary mass particles travel >c, always.

imaginary mass wouldn't make much sense, that's why I'd prefer the negative mass squared term.
Even in the kinematics of SR if you try to include imaginary masses you will end up with real energies and momenta (indistinguishable to what you had in the slower than light region)...and well the equation:
E^2 - p^2 = m^2
won't be affected at all...(cause invariants are not affected by the region you are looking).

If imaginary mass particles can interact with massive matter, then messages can be sent to your past.

well I am OK with this explanation, but... this is just an interpretation based on the common notion of causality.. it is not SR (neither excluded thanks to it).
Causality arises because in "our region" of SR you can define events that happened before and led to later events...In a tachyonic region, you wouldn't have that effect (however you would have a similar one)...

Negative mass squared effects can be found there and here...but SR is not violated by them.

 

[PAllen]

you mean particles with positive mass-squared traveling FTL could violate SR... because as far as I know SR doesn't necessary tell you that the mass squared has to be positive or negative for it to work. In fact Lorentz transformations can be used for both spacelike and timelike 4vectors (I guess).

It depends on what formulas you consider to be SR. The norm of a 4-vector involves a square root. The norm of the 4-momentum of an ftl particle is imaginary.

imaginary mass wouldn't make much sense, that's why I'd prefer the negative mass squared term.

Well, if you look at historic the literature on tachyons, it is presented as imaginary mass.

Even in the kinematics of SR if you try to include imaginary masses you will end up with real energies and momenta (indistinguishable to what you had in the slower than light region)...and well the equation:
E^2 - p^2 = m^2
won't be affected at all...(cause invariants are not affected by the region you are looking).

Yes, of course imaginary mass traveling FTL produces real energies.

well I am OK with this explanation, but... this is just an interpretation based on the common notion of causality.. it is not SR (neither excluded thanks to it).
Causality arises because in "our region" of SR you can define events that happened before and led to later events...In a tachyonic region, you wouldn't have that effect (however you would have a similar one)...

As I said, it depends on your definition of SR. MANY have been used over the years. Please note that what I said was: if tachyons could interact with ordinary matter and were consistent with SR (e.g. followed the POR), THEN you could send a message to your own past. This much is inarguable. Whether you take this as violating SR depends on whether you include causality assumptions in your definition of SR.

Negative mass squared effects can be found there and here...but SR is not violated by them.

Not per se. And if you take it that tachyons cannot interact with ordinary matter, then you don't violate causality either. On the other hand, then Occam's razor would lead to reject them.

 

[PAllen]

Let me clarify what I mean about different formulations of SR. One alternative to traditional axiomatic frameworks is to start from causal structure. One type of causal structure leads to Galilean space+time. A different causal structure leads to SR (or GR depending on what additional assumptions you add). In the GR case, you can (at this level of abstraction, without adding something more to rule them out) end up with CTCs. However in SR, starting from causal structure as an axiom, both FTL and CTC are deductively precluded. It happens that I prefer this conceptual approach - that causal structure is the essence of SR, and everything else flows from this.

 

[PeterDonis]

Lorentz transformations can be used for both spacelike and timelike 4vectors (I guess).

Certainly they can. But the way Lorentz transformations act on spacelike vectors has a key difference from the way they act on timelike vectors: a Lorentz transformation can change a spacelike vector that points "into the future" to one that points "into the past", or vice versa. (I have expressed this heuristically, hopefully it is clear what I mean--but see below for a more precise formulation.) By contrast, a Lorentz transformation can't change a future-directed timelike vector into a past-directed one, or vice versa.

A more precise way of stating the above is: all spacelike vectors form one subgroup under Lorentz transformations; a Lorentz transformation can take any spacelike vector into any other spacelike vector. But timelike vectors form two disconnected subgroups under Lorentz transformations: the future-directed ones and the past-directed ones. A LT can take a future-directed timelike vector into any other future-directed timelike vector, but not into a past-directed one; and vice versa. (Note that ordinary 3-rotations are included in "Lorentz transformations" here.)

The above means that tachyons, which have spacelike 4-momentum vectors, must be fundamentally different, physically, from ordinary particles with timelike 4-momentum vectors. Furthermore, it means that if all inertial frames are equivalent, as SR postulates, allowing tachyons to propagate causal influences inevitably leads to CTCs. The only way to have tachyons propagating causal influences without CTCs is to have a preferred frame that defines which tachyonic 4-momentum vectors are "future-directed" and which are "past-directed"; there is no way to do this without a preferred frame (because of the above).

In a tachyonic region, you wouldn't have that effect (however you would have a similar one)...

I don't see how this is possible in view of the above. Can you elaborate?

 

[rootone]

Maybe the teacher meant that special relativity was a good idea, but general relativity was even better.

 

[haushofer]

My two cents:

[*] SR is a special case of GR: if you take flat spacetime in GR, you obtain SR. So the domain of validity of GR is bigger than the one of SR. Saying that SR is therefore wrong is silly. In physics a theory always comes with a domain of validity, which is pretty much always limited. So in that sense all theories would then be "wrong", which is an empty statement.
[*] FTL speeds are not forbidden by SR or GR. We only don't really know how to interpret these so-called "tachyons" in a physical way, (in field theory they are seen as picking an unstable vacuum for perturbation theory, but that's a technicality). In SR, one cannot accelerate from v<c to v>c, but that's a different statement. In GR, one can only talk about local velocities.

 

[e.bar.goum]

A philosophy of physics class at one of the top universities in Australia. Kind of disappointing. (btw guys please don't go trying to message my lecturer saying he is wrong, I get the feeling that would screw me over pretty bad).

I'm hoping that this philosophy of physics course isn't the one at my "one of the top universities in Australia"!

 

[peety]

Relativity absolutely wrong sounds like a play on words.

 

[PAllen]

My two cents:

[*] SR is a special case of GR: if you take flat spacetime in GR, you obtain SR. So the domain of validity of GR is bigger than the one of SR. Saying that SR is therefore wrong is silly. In physics a theory always comes with a domain of validity, which is pretty much always limited. So in that sense all theories would then be "wrong", which is an empty statement.
[*] FTL speeds are not forbidden by SR or GR. We only don't really know how to interpret these so-called "tachyons" in a physical way, (in field theory they are seen as picking an unstable vacuum for perturbation theory, but that's a technicality). In SR, one cannot accelerate from v<c to v>c, but that's a different statement. In GR, one can only talk about local velocities.

It depends how you define motion and momentum, in either theory. If you define motion via 4-velocity, itself defined as derivative with respect to proper time, then tachyonic motion is ruled out by definition (as are massless classical particles). If you define motion via 4-momentum that is not necessarily connected to 4-velocity, and allow 4-momentum to be null or spacelike, then you can say null 4-momentum is associated with massless particles, and spacelike 4-momentum with imaginary mass tachyons. IMO, original SR/GR definitionally ruled out tachyons, and it required modifying definitions to allow them.

[edit] Note that the re-definition of motion for tachyons is more radical than for null momentum motion. Null motion is still past/future determinate. Tachyonic motion is characterized by inherent inability to define which part of the motion comes before another part of the motion; it is past/future indeterminate. This is not motion in any historic sense of the term. No one normally says Newtonian mechanics includes the concept of 'simultaneous motion', and modifies definitions to treat this as a 'motion', but you could; you would be led to say the mass needs to be zero, etc. This would be taken as an extension, not a natural part Newtonian mechanics. Similarly, viewed historically, tachyonic motion is extension via redefinition of traditional SR/GR.

 

[Charles Kottler]

I think there are two points worth considering.

First, SR is 'absolutely' the wrong theory to use in the context of claimed FTL measurement in this experiment. It is specifically only applicable far from the influence of gravity. In the case of this experiment it is possible that greater density close to the two end points could locally slow time at the ends more than in the middle leading to a (very small) decrease in the expected travel time of a body moving at light speed between them.

The second, and more relevant, point is that this was a 'Philosophy of Physics' course. As such the emphasis of the course should lean more towards the technicality of making a valid argument than agreeing with 'mainstream' physics. The purpose of the course should be to help you critically analyze earlier papers and discussions to find flaws in the arguments which can help lead you towards new understanding of the subject. If this lecture followed on from previous lectures and tutorials on the matter you would need to show where in his earlier arguments your professor had made a mistake, or if this specific point has not been raised before then you need to discuss with him why he believes SR does not form a valid framework for physics.

 

[atyy]

In case it hasn't been said clearly enough, one reason the lecturer's correction is wrong is that all our present theories are known to be absolutely wrong.

Secondly, it is misleading to say that GR allows superluminal travel, because the notion of velocity used in such a statement is different from the one used in the claim of superluminal neutrinos. If we use a consistent concept of velocity, neither SR nor GR permit superluminal travel.

Orodruin (post #6) and jbriggs444 (post #10) made these points earlier in the thread.

 

[Dale]

First, SR is 'absolutely' the wrong theory to use in the context of claimed FTL measurement in this experiment. It is specifically only applicable far from the influence of gravity.

Not really. It is specifically tidal gravity which requires something beyond SR. For this experiment the gravitational field is largely irrelevant, and tidal effects even less so. SR is a perfectly legitimate theory to use in analyzing the opera experiment.

The characterization of scientific theories as "right" or "wrong" is a very naïve approach to science. Theories are better characterized by their domain of applicability. The opera experiment is well within the domain of applicability of SR.

 

[PeterDonis]

SR is 'absolutely' the wrong theory to use in the context of claimed FTL measurement in this experiment. It is specifically only applicable far from the influence of gravity.

This is not correct. SR is applicable in any region of spacetime in which the effects of spacetime curvature are negligible. This is a more permissive condition than "far from the influence of gravity"; the presence of gravity restricts the size of a region of spacetime in which SR can be applied, but it does not prevent it from being applied in a region of small enough size. The term used to describe the application of SR in a small enough spacetime region is "local inertial frame". So you can't just make a blanket statement that SR isn't applicable when gravity is present; you have to actually look at the size of the spacetime region under consideration to see if it's small enough to be considered a local inertial frame.

In the case of the neutrino experiments, the time of flight was several milliseconds and the distance traveled was about 731 km. Since the neutrinos were traveling at essentially $c$, the time of flight gives essentially the same "size" as the distance traveled for the region of spacetime under consideration. The test of whether a given region of spacetime can be considered a local inertial frame is the magnitude of the corrections to the metric coefficients due to spacetime curvature. For the neutrino experiments, these corrections are given by $L \sqrt{GM / c^2 R^3}$, where $L$ is the size of the region of spacetime in question (here the distance traveled by the neutrinos), $G$ is Newton's gravitational constant, $M$ is the mass of the Earth, $c$ is the speed of light, and $R$ is the radius of the Earth. This gives a result of about $3 \times 10^{-6}$, which is quite small--several orders of magnitude smaller than the margin by which the neutrinos were originally thought to exceed the speed of light (which was about 1 part in 1000). So the region of spacetime occupied by a run of the neutrino experiments is small enough to be considered a local inertial frame for the purpose of analyzing the experiments, and SR can be applied.

 

[vanhees71]

I would talk to a physics professor at your university about it. You are unlikely to win an argument like this on your own, regardless of what sources you provide.

I'd even go further and appeal in some official way against the statement by the professor posted in the OP. Is this even a physics professor? If so, I'd say, there's a big problem in the physics faculty, which is perhaps very difficult to solve!

 

[vanhees71]

I'm hoping that this philosophy of physics course isn't the one at my "one of the top universities in Australia"!

Well, before you study the philosophy of physics you should study physics. That's more or less trivial, but unfortunately many philosophers try to do philosophy of physics without having a clue about physics.

The apparent "faster-than-light neutrinos" measured by the OPERA colaboration are gone for some time now. It turned out to be an error in the setup of the measurement (some issue with the connection of glass fibers and some time-measuring oscillator, if I remember right). So there's not the slightest hint for tachyons so far, and if there were, theorists have a huge problem to solve, because today there's no consistent theory of interacting tachyons. Free tachyons can be theoretically described to a certain point, but they are not observable, because they don't interact with any detector by definition. So it's useless to study them too.

 

[haushofer]

It depends how you define motion and momentum, in either theory. If you define motion via 4-velocity, itself defined as derivative with respect to proper time, then tachyonic motion is ruled out by definition (as are massless classical particles).

I'm not seeing this directly. Could you elaborate? Massless classical particles are described by an action using an auxiliary field, and I don't see why this action cannot also be used for tachyons. :)

 

[PAllen]

I'm not seeing this directly. Could you elaborate? Massless classical particles are described by an action using an auxiliary field, and I don't see why this action cannot also be used for tachyons. :)

That's why I said it depends on definitions you adopt. I gave one definition, you immediately changed definitions. For 4-velocity defined as derivative by proper time, it fails to exist for null paths because the differential interval is zero. For spacetime paths, it fails to exist because the differential interval is proper distance rather than proper time.

 

[Charles Kottler]

Well, before you study the philosophy of physics you should study physics. That's more or less trivial, but unfortunately many philosophers try to do philosophy of physics without having a clue about physics.

It goes without saying that a professor teaching philosophy of physics must have a thorough understanding of the maths. However, while I agree that it would be useless to study the philosophy of physics without also studying physics, I don't see the necessity of learning the physics first. Often these days students are drilled to 'internalize the maths' to avoid getting sucked into the much harder challenge of understanding what it actually means. I would argue that studying the maths without also studying the reasoning behind it would be a waste of time. The following article outlines the need for philosophy far more eloquently than I ever could:

http://www.pbs.org/wgbh/nova/blogs/physics/2015/04/physics-needs-philosophy/

Every course on relativity should include a discussion of the implications and the weaknesses of the theory. Students need to be taught to look for the purpose of what they are learning before being taught the detail.

 

[PeterDonis]

The following article outlines the need for philosophy far more eloquently than I ever could

Unfortunately, I think this article unintentionally illustrates why philosophy contributes so little to physics. The writer says that "physicists strongly discourage questions about the nature of reality". Yet he never considers the possibility that the reason physicists discourage such questions is that "reality" is too vague a term to be used when formulating questions that can actually be answered. If philosophy were really doing its job, it would either give a precise meaning to the term "reality"--a meaning that could actually be used in understanding physical theories--or admit that the term is too vague and drop it, and find better terms.

Another example of unintentionally illustrating the limits of philosophy is the article's discussion of the twin paradox. The writer asserts that the explanation given in the Feynman lectures is incorrect. But he never says what the "correct" explanation is. He also asserts that "it is easy to describe cases where the opposite is true, and even cases where neither twin accelerates but they end up different ages". This is simply incorrect when working within the limits of special relativity--which is what the chapter of the Feynman lectures that the article refers to was discussing. Scenarios where the unaccelerated twin ages less, or where two unaccelerated twins can end up aging differently, can only be set up in curved spacetime, which is beyond the scope of SR. Within the scope of SR, the rule Feynman gives works fine, so calling it "incorrect" is, well, incorrect.

I could go on, but I think you get the idea.

Every course on relativity should include a discussion of the implications and the weaknesses of the theory. Students need to be taught to look for the purpose of what they are learning before being taught the detail.

I agree with these statements, but they are not arguments for learning philosophy before learning physics. They are arguments for learning physics. Physics includes the implications and weaknesses of theories, and understanding the overall purpose of a theory--what domain it is supposed to cover--before learning its details.

 

[haushofer]

I like Feynman's quote on this issue:

"We can't define anything precisely. If we attempt to, we get into that paralysis of thought that comes to philosophers, one saying to the other: 'you don't know what you are talking about!'. The second one says: 'what do you mean by talking? What do you mean by you? What do you mean by know?"

:P

 

[haushofer]

Another example of unintentionally illustrating the limits of philosophy is the article's discussion of the twin paradox. The writer asserts that the explanation given in the Feynman lectures is incorrect. But he never says what the "correct" explanation is.

He also includes this in his "Philosophy of physics: space and time". The corrext explanation Maudlin means is the fact that according to him only the ratio of both lengths of the worldlines should enter the explanation of the twin 'paradox', and not the acceleration. (page 83) But I consider that a bit of nitpicking: I think a lot of physicists would understand bending of the worldline as being caused by acceleration. Maudlin likes to rephrase SR in purely geometric terms.

The cleavage between physicists and philosophers of physics is also something that fascinates me.

 

[atyy]

I can't agree that physics and philosophy don't go together. The concept of reality is absolutely fundamental in physics. The best physicists have always been interested in philosophy - Landau and Lifshitz, Dirac and Weinberg. And of course, Feynman made so many comments on philosophy, he must have thought it important. Late in his career, Feynman made a statement that is equivalent to saying that the violation of Bell's inequality sums up quantum weirdness.

So I would contend that the teacher is doing bad physics and bad philosophy.

 

[vanhees71]

Well, concerning Weinberg I'd cite

http://www.pitt.edu/~mem208/courses/phph_s15/documents/weinberg_against_philosophy.pdf

 

[atyy]

Well, concerning Weinberg I'd cite

http://www.pitt.edu/~mem208/courses/phph_s15/documents/weinberg_against_philosophy.pdf

He wrote a really long philosophical essay :)