Killtech

Firstly i do believe that i (and we in general) don't know enough and so i wouldn't trust in any believe, especially my own. Instead i prefer to switch between believes in figuring out what works best rather then dig myself into one and disregard all else as false. flexibility and open mindedness works better when looking for answers. I also found that this method of thought is significantly more effective to get a deeper understanding of each believe/interpretation, its weaknesses and advantages. In particular you learn so much more by failing starting from a wrong assumption then never straying from the safe path... but it gets you in trouble when you figure out that some alternative assumptions cannot be made to fail. Don't take this approach as being blasé about it. Its quite the opposite, as this is my way to approach such problems. I did mention it in my original post and somewhere later as well here, but true, not detailed enough. It was discussed in detail in my conversation with ChatGPT on the topic which link was deleted. The original question was to take the assumption of a variable c to what it means for atomic physics - but assuming it remains effectively constant over the region covered by an atom. One way of approaching this is doing quantum mechanics with generalized LET assumptions. Without loss of generality one can do all this with the simple hydrogen atom. And whatever we assume for c (including a ether wind effect), we can find coordinates that undo it to reduce the problem to the original hydrogen, so we don't have that much to solve. Effectively whatever we do, it scales all atomic quantities, including transition frequencies which serve to define the unit of time and the older definitions of length. The latter is because this approach continues to solid state physics affecting the distances between atoms. if we measure time or length by these standards we can conclude all variations to c must cancel out. Btw. this is also an interesting showcasing the mechanism by which Michelson-Morley must still yield a null result in LET. Now, the last statement depends on you choice of a metric. i know this is not an easy one to understand, especially given out intuition, but with math we can do a lot of trickery that defies intuition. Let's recall what Lorentz transformation originally are: coordinate transformations. so let's forget all interpretation and look at reality entirely from the perspective coordinates give us. Let's start with sound waves in a frame x at rest to the medium and pick some other frame x' which moves with a velocity v. What would happen if we apply a Lorentz coordinate trafo from x to x' but using c_s, the speed of sound, instead in the transformation? how does the sound wave equation look like in the new coordinates x'? Alternatively we can get the same answers when we start with the sound wave equation and ask ourselves under what kind of coordinate transformations this equation will remain invariant under? The acoustic metric i mentioned adopts these coordinates and their transformations between frames as the standard - which will be at odds with out intuitive way of looking at these situations but you cannot argue it to be logically wrong. and that metric (or these coordinates if you want) makes it so, that the (coordinate) speed of sound will be constant with respect to the speed of the observer. And by doing so, it enables us to make a fully analogy between light in general relativity and sound - so that we can study (acoustic) black holes in a lab and find the analogue observations astronomy yields. But there are bit better sources on the topic then me, as there has been some research on this. it would be very helpful if you could read up on it so we can discuss this further on a more equal ground. Within the acoustic metric, the formulas become identical. This shows that we can interpret special relativity as a general mathematical methodology for removing the background dependence for any wave equations. Of course the big difference is that for general relativity we have clocks that directly correspond to time coordinate as produced by the Lorentz trafos, but for acoustics, while we could construct such devices that behave analogous, they won't be considered natural in any way... although maybe a for a bat, they will be as it sees the world though acoustics instead of light.

well, what is are one and only "true" coordinates? for me conventions like these have no fundamental meaning of right or wrong, but can be exchanged if deemed more practical in a given context. The geometry of a manifold is mathematically similar and can be changed by redefining geometric objects (you just have to make sure to stay in the same differential topology) - it does not affect the predictions you do on it, only the notation of how you describe them. So i do not understand, why you would stick to a singular geometric description of spacetime, if there is no true one. Instead, we could look at all possibilities and check out what advantages they may have. same as we do with coordinates. Yeah, and it is the same two experiments also with the old standard, as the old one standard of length is still based on a derivative of the speed of light and therefore has no potential of deviation. This is my what bothers me. Maybe you should go through the history of how the speed of light was measured and Poincaré critique of it. Because it turns out astronomers actually did the mistake of using the same experiment to determine distance as they used to determine the speed of light, mistaking that such an arrangement cannot tell if it is constant or not. All later attempts can be reduced to the same problem, even if it is less obvious. Your line of thought only works if the alternative standard of length used in an experiments can be considered sufficiently independent of c. In the absence of such an option we can however ask another question: what would happen, if the length of a meter or second changed depending on the location? the answer is simple: this produces geometric curvature. We can also ask the question what would happen, if we defined the standard for length and time implicitly or explicitly both via the propagation of some type of wave. it turns out all relativistic effects can be reduced to this relation between the wave propagation and the definitions of time and length. and this idea works for any wave, not just light. the acoustic metric is a perfect example of that. it shows that we can treat sound waves identical to light in a vacuum with curvature and using that special definitions of time and space we get all the familiar framework. yes, and it is the same with the acoustic metric - that is if we change geometry of spacetime just so sound is described analog to light and becomes perfectly constant. you should explore that, if you are not familiar with it. it is a nice learning experience leaving the question how does an acoustic black hole differ from a gravitational one in terms describing the signal propagation around it? but we can also go the other way around, as we do know a description where the speed of sound is variable. Normally you would start looking for a spacetime definition which is perfectly flat but it that case it is given. However, that alone enforces the propagation speed of your waves to become variable, if spacetime wasn't flat before. of course that produces an issue with the original metric, so you do not want it to have any dependence on it. and if you drop it, you end up with the old Euclidean metric. You lose the entire complexity of a variable geometry but of course this comes at the price of moving the physics previously described by the geometry to other physical entities instead. You get something like a Gravitomagnetic field with the same number of degrees of freedom your geometry had before. How would that look like? Actually, space agencies like NASA uses effectively that when calculating trajectories of objects in the solar system. they do all calculation in BCRS coordinates. if the geometry becomes more complicated than in the simplistic Schwarzschild case, GR natural representation becomes unsuited for practical calculations. so for many body system we do choose specific coordinates to do calculations or simulations in, but really, these coordinates align with how you would define a flat spacetime framework like LET and hence the coordinate speed of light in such coordinates is never constant.

No, it's not what i think. i used the quotes for a reason. so i think we are in agreement here. i was just confused because you highlighted that your instrument would have to measure these constants - which made little sense to me, once they are defined. i suppose you meant to measure them according to previous definitions in order to establish an accurate redefinition? My concern lies not with the definition of time itself but with the miscommunication of the nature of light as represented by the constancy of c. We look for new physics by making hypothesis extending our current laws of physics and can put those new theories to the experimental test. GR was discovered this way. but when it comes to the speed of light, the situation is fundamentally different. we cannot even hypothesize it to vary due to how it is tied to both the definitions of time and space. It should be highlighted that the problem remain the same, even when we go back to the previous definitions, simply because how the behavior of the atom is tied to c. My concern is that we do not distinguish between laws of physics which are implied by the definitions of time and space to laws of physics which behavior shows some degree of independence from them. The latter can be approached making new assumptions and theories, while it is not possible for the prior because such assumptions result in contradiction to these definitions. I understand that looking for contradicting assumptions in a theory instead of experimental verification is not how physicist think - yet it is a step that cannot be avoided when discussing the nature of c. Looking back at the definition of T, L and M that you proposed, it of course makes perfect sense if we could get G with high accuracy. But it also defines a spacetime where c is perfectly fixed since you use it as a basis for your definitions, right? Yet unlike laws of physics, definitions are conventions that are judged by practicality. I argue that there might be some purposes for which a different definition of spacetime which allows for a varying c is more advantageous. But this does not mean that we have to pick one over the other. I propose to treat the definition of spacetime similar to how we treat coordinates: accept that there are multiple possibilities and we choose the one most suited for the problem.

In your example you have only 3 constants left, hence you eliminated the 4th \( \Delta \nu_{Cs} \) constant which currently servers to define the second. If that was your intension, that i misunderstood your statement. But yeah, come to think of it, this one is indeed lacking fundamental physical meaning, hence it is a very fair point to wanting to eliminate it. i did not have that in mind when i answered your post, hence did not consider to contest it. So fair point, i do agree with you. No, there is no such problem as the fundamental constant cannot be measured in this situation and therefore must be defined. This is fine because the constants - or more precisely the core physics they originate from - serve as the rulers to measure everything with. for example the speed of light together with the other constant allows you to construct a real photon emission which wavelength is fully determined by the constants and this base wave lengths servers as a ruler to compare any other real length to. For this we only need to know the defined values of the constants instead of measuring them. The definition via constants has the nice advantage that we have some freedom in constructing the ruler we use - yet all the different rulers constructed in this way will still agree (if the physical laws underlying them use are indeed correct). In this situation the "true" value of the constants remains unmeasurable, or if the constants are even constant at all. As all measurement is fundamentally relative, that is a comparison of e.g. a length vs a reference length, hence absolute length independent of the reference cannot be determined. It is a bit like asking for my absolute velocity - but we can only meaningfully measure my velocity relative to some point of reference. The definitions of the constants are instead always chosen to ensure maximum continuity and consistency with the previous older definitions of the SI system rather then for physical reasons. So this is why we first need to measure G to high accuracy in the old system before we can define it anew in a new SI iteration.

exactly Yes and this is a fundamental aspect of the math used to describe wave physics in general. for example it holds to for acoustics just the same, if you use the acoustic metric (https://en.wikipedia.org/wiki/Acoustic_metric), that is \[g_{\mu \nu}^{acoustic} = \frac \rho c_s \begin{pmatrix} -(c_s^2-v^2) & -v^j \\ -v^i & \delta_{ij} \end{pmatrix} \] with that the difference between acoustics and light physics almost vanishes. they do really only differ in interpretation. apparently this is know and very useful to study astronomical situations, like the situation around black holes in a lab... yes, apparently sonic black holes can be build and are so similar that even the sonic analog of Hawking radiation can be experimentally observed. So much for talking to an AI. I figured previously the similarity between light and acoustics is astounding and the AI immediately understands what i meant and pointed me to sources that do apply that very idea and showed me the exact method how the math between both is made to agree. The atom physics is described by quantum mechanics. As you rightly say, the choice of atom does not matter in theory, so is sufficient to constrain to the simple hydrogen model. And hydrogen is a bound state of positive and negative electric charge that gets all its properties from the electromagnetic interaction modelled at quantum level - the very field which core characteristic is c. From here we can deduct it from the math that anything that would affect the speed of light would identically affect the atom. That is for any change of the speed of light there exists a coordinate transformation that undoes it and returns everything to the original situation. So an atom is entirely unable to observe any change in the speed of light at a theoretical level. Any deviations lead to immediate contradictions in the theory. Hence no measure of time or length defined by atoms is suited to measure the speed of light. This argument becomes weaker if we would chose a transition between bound states of the strong force, since that would at least measure the speed of light in units of the speed of the strong force and therefore have physical meaning as there would at least be a logical possibility of a deviation. Hmm, it could be but it would require some other constant to lose its definition. The unit of G is composed of units of time, length and mass. It makes no sense to compete for the first two, so its would compete with Plancks constant to define the Kilogramm by its value. As only the value of one of them can be defined, the value of the other becomes a derivative as a consequence.

True, up to the point of being a tautology and hence misleading. measuring c locally is leading measurement ad absurdum, since it means measuring it in units of proper length and time, which definition tracks back to the atom, which again is a bound state of the field characterized by c. so it means measuring c in units of itself which always yields exactly 1c, and that is a tautology. LET interpretation therefore is that c cannot be measured entirely by local means (or at least not utilizing electromagnetic means only). hmm, the metric i started with has a component for both time dilation and spatial curvature. but with the weak field first order approximation you drop terms here and there, hopefully i didn't drop one too many. but it look fine to me, when i check the result against sources for gravitational lensing.

You do learn so much when asking an AI. Figures, that what i suggest is almost what we do practically do anyway to make any predictions, for example on how light is bend in the solar system. For such predictions we do use BCRS coordinates (with TCB time), which correct for gravity effects, just like i suggested. When looking at these coordinates in context of LET, we can also identify them as the native spacetime of this interpretation, that where the metric LET chooses becomes trivial ( \( g'^{\mu \nu} = \eta^{\mu \nu} \neq g^{\mu \nu} \) in BCRS) and thus partial and covariant derivatives become the same. Now, how do we calculate gravitation bending in the solar system? At first the metric in first order approximation (for weak gravity fields) in BCRS coordinates is given as \[ g_{\mu \nu} = \delta_{\mu \nu} (\pm 1+\frac {2 U(x)} {c^2}) \] with \( U(x) \) the gravity potential. plugging that into Maxwell and translating the covariant derivative via \[\nabla_\mu F^{\mu \nu} = \frac {1} {\sqrt {-g}} \partial_\mu (\sqrt {-g} F^{\mu \nu} ) \] we get the Maxwell equations for the solar system in BCRS, or we can interpret them as Maxwell in LET spacetime. In contrast to the usual Maxwell equation, it has now additional terms. When looking at how light is bended, we start with a plain wave ansatz and apply it to the new equation. From there we can obtain the Eikonal equation for the plane wave \[ -(\partial_t \phi(x,t))^2 + |\nabla \phi(x,t)|^2 = 1 + \frac {2 U(x)))} {c^2} = n_{eff}(x) \] that is an optical equation with an effective refractive index \( n_{eff}(x) \) and hence a local varying coordinate speed of light \( c(x) = \frac c {n_{eff}(x)} \). From there we can calculate the gravitational lensing effect like the bending angle depending on the trajectory. In the LET interpretation the calculation is the same except that we can simply call it the actual speed of light. The thing is that for any realistic case with a slightly more complicated geometry then Schwarzschild like a many body problem, the calculations are performed in specific coordinates that either are or very closely coincide with the spacetime of LET. So it turns out the pink uniform i was chasing is a simplified interpretation of what we have do in our calculations anyway. Thus the question if c is a constant, is in fact down to the choice of interpretation. But all that has no consequence for predictions made. Where it starts to matter is when we look at the observed expansion of the universe and the arbitrarily added cosmological constant to fix GR. For LET interpretation, it is very natural that an ether densely packed into a small region of otherwise empty space will start to expand. It turn out reinterpreting GR in terms of LET, it assumes the ether has a constant energy density defined by the cosmological constant and a constant pressure equal to the negative of that density - despite its expansion. in this picture dark energy corresponds to an uniform tension in the ether. Hence, it appears very natural why such a weird attempt at a fix ultimately produces more problems then it solves. I think the gist of my problem is that we got locked into a singular interpretation, that is becomes almost impossible to talk with people about it. The constancy of the speed of light seems at times more like churches dogma then a practical view. Specifically what bothers me is that most don't seem to understand where it even comes from unaware of the intrinsic circularity in measuring it. Well at least the AIs are smarter...

Argh, but it does not. A coordinate time is not the same as time, i hope you understand that distinction. a pendulums position can be measured in spherical coordinates, but you wouldn't mistake \(phi\) for a measure of length? coordinate times may be similar or even numerically equal to a clocks reading in some instances, specifically when defined via theoretical clock in a special location. But they are conceptionally something else and interpreted differently. However, TCB is very close to Newtons idea of absolute time since as coordinates are not dependent on the frame. In that sense it can be considered as an alternative to clocks. in theory it is just as consistent and reproducible means record events as a clocks reading. in practice, it is usually far simpler to measure proper time then a coordinate time. but when it comes to calculating collisions of celestial bodies in the solar system, we do prefer to do them in TCB coordinate time rather then any proper time involved.

but we are discussing the general case of a clock which can be in any frame or location, hence this specific scenario is not helping. yes, and it is different for every location and frame, so the general conversion is a complicated formula that depends on the trajectory taken over which you need to integrate, hence the result will differ depending on the circumstance. Along some trajectories TCB will measure the identical time passed as a clock would, along others the TCB time interval will be shorter then what the clock reads and along others again it will be longer. In every case a complicated conversion exists, which if you want to make it entirely correct, not just approximately like in 10.2 section of the article i posted before. where did you get the idea from that i don't think they are clocks? a device reading TCB will not count as a clock though since the time it reads is very different from clocks.

these are conversion between various coordinate times, not between TCB and clocks. the difference you wrote is between TT and TCB, that is it is the conversion valid only for the situation of a clock on earth. it will yield wrong results in space where TT does not align with the reading of clocks. far off from the sun the tick rate between TCB and a clock vanishes as they tick at the same rate. the difference depends on the location and frame. Here for a few more details with proper formulas https://iers-conventions.obspm.fr/content/chapter10/tn36_c10.pdf (though these are lower order approximation; actual formulas are significantly more complicated).

TCB, implemented as a device, behaves close to a radio clock (with a specifically arranged source signal). These differ from other clocks in how they respond to conditions where time dilation becomes relevant. This is what makes it interesting for consideration. most constants, as expressed in units of TCB have slightly different values. Moreover, c specifically varies in value depending on the region because the time dilation correction in the time standard moves into c. This isn't anything novel though, as normally it is treated as a coordinate speed for which this is absolutely fine. However, if we declare such time measurements to count as clocks, we elevate what is normally a coordinate time to count as proper time - therefore we change the interpretation of what is proper. This subtlety is why it is relevant how a clock is defined exactly in terms of the theory. In this instances this would come with the consequence that the covariant form of physical laws have to take the form of how they look specifically in these special coordinates, that is how the coordinate speeds, coordinate forces and so on behave there. It's not a problem either way, if we know what we are doing. It does however impact the calculus without changing the predictions it makes.

yes, that's a good summary. the question is however, what if a regular and reproducible pattern does not uniquely specify a clock? For example, if someone build a device that was able to calculate the Barycentric Coordinate Time (TCB) at any instance, it would provide a perfectly consistent method of measuring time. But its rate of change differs relative to an atomic clock depending on the frame and location. in particular, it ticks faster then a atomic clock on earth, but far out of planetary influence, the tick rates are be much closer. The consideration of alternative devices that behave differently from regular clocks (those that conform to SI standards), yet still provide a consistent mechanism to measure time is the reason for this thread. wasn't the time in office of a Britisch prime minister not so long ago counted in cabbage rots?

there is no evidence that they have any mass either. the only reason why it is speculated it to have a mass is due to the discovery of their oscillation. but so far all experiments trying to determine it came empty handed. Also no experiment found any evidence of them moving any slower then photons. So we are coming to a point where the matter is getting more complex and alternative explanations are looked into for a reason.

for a massless neutrino, this becomes interesting when we compare that frequency to a photon traveling parallel to it with the same frequency in a starting frame. if we measure the both frequencies from another frame moving with velocity v relative to the base frame in the direction of both particles, the photon frequency becomes red shifted, but the neutrinos either won't be or it will be by another value. Such would be the nature of CPT violation. Sure, in the base frame, both time measurements would agree, in the other frame they wouldn't. so what now? Not an ontological discussion, but about our model of spacetime. Because for every given ruler which measures a curved geometry you can derive and construct another ruler which will measure the same space to be flat (at least locally. but it holds globally for most standard topologies). This is why you have to be specific about the choice of rulers you use. to put it in other words: is spacetime curved or is it flat and it is just our rulers that change in length (and clocks) depending on region due to some local physical interactions / conditions? There is a class of physical interactions which makes it impossible to distinguish between these two possibilities and so it boils down to a different choice of clocks and rulers. So for a theoretical purpose, the specification matters a lot.

Not true, maybe that is so in colloquial language but not in science, or at least not in math. The distance between two points is one value given a metric, but the length between them varies by the chosen path. The distance is always the minimal length between two points. At least according to the definitions i know. Now back to your objection, how does your definition of time and clocks even prevent some nonsense like your example “take a mass and suspend it from a string” from being a clock? if your definition does not even demand clocks to provide their measurements as numeric values, then it becomes so weak anything would count. that is my objection to the way you would define those terms. a definition is meant to give rules to recognize if something meets these criteria or not, so you can test things against this definition same way as you test the assumed laws of nature. usable, yes. disagreeing, also yes. if you have different and disagreeing definitions of time it requires to distinguish between them carefully and establish transformation behavior. the neutrino oscillations, if we had more efficient ways to observe it would also provide usable means to measure time. If they were subject to CPT violations (hence not entirely Lorentz invariant), such a time standard would drastically differ from other clock based on an atomic standard.

and the many conflicts around it. solar or moon year? a common or a leap year? there are many variants of this term, so we need different prefixes to distinguish between then when it isn't clear from the context. so when it comes to accounting the passing of years, it turned out highly contentious topic in politics and religion... would you really mistake a introductionary explanation for a random reader with a formal definitions that follow afterwards?? length is defined via a distance, which math calls a metric d as a mapping to the real numbers and it is defined by the follwing 4 properties: 1. \(d(x,x) = 0 \) 2. \(d(x,y) > 0\) if \(x \neq y\) 3. \(d(x,y) = d(y,x) \) 4. \(d(x,z) \le d(x,y) d(y, z)\) So it references itself to declare the rules that specify its characteristic properties. But it gives us something to work, enough to establish basic geometric concepts. Note that this definition is deliberately open for various possible implementations, so it knows that the concept of length cannot be uniquely determined this way and there are multiple non-equivalent possibilities this allows. This is why you will sometime encounter a term like d-length and d'-length to distinguish between different measures of length. There is the concept of isometry to compare different concept of length and establish if they yield an equivalent measure.

i do very much have a problem with this, hence check my answer to @KJW about his suggestion. it is a topic of geometry amongst others and mathematicians had quite some trouble to figure our a usable ways to do it. These are the definitions at the very core of Riemann geometry and more general, metric spaces in analysis. They explore the concept of length by the properties it has. Here is how much you can dwell of this definition, starting with simple ideas to the generalization for more general geometries: https://en.wikipedia.org/wiki/Arc_length

so if you define time as “that which is measured by a clock” and a clock "that which measures time" you explained nothing. if you do that, you end up with definitions which are so devoid of any meaning that from them you cannot even deduct that time is measured in numbers and that it establishes a compare relation, like which time interval is longer. Sorry, you have to put more work into that. look at how such definitions are done in math so you come up with something that can be worked with at all. For example look at how math defines a metric in order to be able to use the concept of distances. You need a bare minimum like that. Basically you have to define either time or clocks at least by the properties / rules they are subject to. something that characterizes them. That's how mathematitians and geometers always approached this - and to their surprise they figured this rarely leads to a unique construct but rather classes of such. In germany, there is an "Eichbehörde". if you construct a clock that you cannot show that it obliges by the SI specifications of the second, you won't be allowed to call it as such nor use it for any business - i.e. a taxi would not be allowed to use such a clock since it could measure whatever the creator desired to rip off the customer.

Hehe, yeah, been there. At which point i figured these instructions can be more or less reduced to the definition of the SI second. Maybe you understand why i use that a bit synonymous with the definition of a clock - for the lack of an alternative as physics books tend to skip on defining on what counts as a clock. And you have another issue. Those instructions will still be somewhat arbitrarily chosen, a convention that is. They cannot be deducted as a unique choice from the laws of nature. Thinking as mathematicians would, we can have a look at the entire set of possible instructions and start to distinguish and classify them by what effects they have. We would have to constrain this set by additional rules to ensure the resulting device yields reproducible results regardless how it got there (no kind of hysteresis effects). and if i give you some other set of instructions, say for a specific coordinate time, will that be a clock as well? This leads to finding the set of all possible interpretations of spacetime compatible with our reality.

the OP mentioned the scenario of a a massless particle oscillating and mentioned the conflict this causes with time himself - and thereafter made some speculations about the speed of light vs "time" i did not fully follow.

and what would a definition of a "clock" be? Any oscillation can be used as reference to define some kind of time. the neutrino oscillation is an active research topic that is not yet fully understood. as of now, assuming the neutrino having a mass is just as much speculation given our current experimental results. There are even some conflicting data which are partially at odds with invariant nature, though there may still be other explanations. there is a reason why Lorentz violation is considered, including experiments that could test for it.

within the standard interpretation, a massless neutrino would cause a problem as our definition of time simply has a singularity there. for as long as we don't have any evidence of an neutrino mass, the other alternative would be that some laws of physics of neutrinos is not Lorentz invariant: https://en.wikipedia.org/wiki/Lorentz-violating_neutrino_oscillations that would allow a particle to oscillate even when travelling exactly at the speed of light. our current model wouldn't be ideal to describe it, though it is still possible. There is no unique way to define time, so a Lorentz-violating neutrino would specify its very own concept of it, practically defining its own clock that perceives the flow of time differently from our electro-magnetic based ones. and if such clocks existed and were used to measure the speed at which light moves, it would turn out it changes depending on the frame.

here you go: https://en.wikipedia.org/wiki/Flat_manifold e.g. \(S^1 \times R \) is a flat cylinder

who said something about a circle? here you go, twin paradox in compact spaces: https://arxiv.org/pdf/gr-qc/0101014

And there is another effect which may offer a new possibility to measure the one-way speed of light. For now we were not able to measure the mass of neutrinos which is needed to explain their oscillation and this is what the interpretation of SR locks us into. But the neutrino is not entirely electro-magnetic in origin. This offers a possibility that the laws it is subject to may be to some degree independent from the laws we assume to define constant for the SI system. Hence there is the hypothetical possibility that it is subject to some physical law which is not Loretz invariant. Any such violation would imply that is has its own idea of proper time and may exhibit oscillation even when traveling at the speed of light. In such a case the comparison of the neutrino oscillation frequency with the frequency of proper time in a given frame provides a tool to deduct the one way speed of light.