Why The Theory of Relativity Doesn't Add Up (In Einstein's Own Words)

Relativity is as successful a theory as it is mind-bending - yet Einstein himself did not believe it was complete, and in a 1914 paper he critiqued its internal consistency at some length. Indeed, at one time or another we have all found ourselves in a state of healthy skepticism about the tenets of relativity, seemingly confronted by a mysticism of warping space and time that is nigh impossible to wrap one's head around -- and so here we find ourselves compelled to ask the same question Einstein did over a century ago: is the theory of relativity truly consistent, and if not, what does this mean for its future?

Please help support us on Patreon!

Einstein's 1914 Paper "On the Relativity Problem"

Henry Lindner's "On The Philosophical Inadequacy of Modern Physics"

Contents:
00:00 - Intro
01:04 - Of Axioms & Absolutes
04:22 - Einstein Calls Out His Own Theory
05:54 - Defining "Absolute" Acceleration
07:35 - What are We Accelerating Relative to?
10:00 - Einstein's Mistake
11:57 - Where Do We Go From Here?
16:14 - Acknowledgments

[ScienceClicEN]

Yet again a fascinating video! I am wondering however what prevents us from defining "acceleration" as the anisotropy of the laws of motion? For instance: in an accelerating frame, a spring changes length if we rotate it, and objects which are initially at rest start moving in a specific direction, which indicates that there is an anisotrpy in the laws of physics for this observer. In this sense, an "inertial frame" can be defined as a frame in which the laws of motion are isotropic. Or more generally, an inertial frame would be a frame in which the laws of physics have a maximum amount of symmetry. I am probably missing something but I don't see why this would fail?

In a way this is related to the fact that - even though time and space are relative - spacetime itself is absolute (in the "orthodox" interpretation of relativity) : we can tell whether or not a line through spacetime is straight or curved, using this anisotropy approach. And acceleration is the fact of turning / changing direction through spacetime.

[buckeyefan]

I'm not a mathematician, physicist, or engineer. I'm just a finance and accounting guy.
BUT I LOVE THIS CHANNEL. THANK YOU !!!

[jenda]

Great, thought-provoking video. I would just like to point out that constant speed of light is not merely an axiom. It is one for special relativity; however, the fact that the speed of electromagnetic waves (light being one type thereof) is constant and independent of observers comes as a clear result of Maxwell's equations and their solution for space without point charges and macroscopic currents. This result was measured and verified countless times since then. This is in stark contrast to Galilean relativity whereby velocities add and this problem puzzled physicists for more than 40 years.

The reason Einstein's 1905 paper is called "On the Electrodynamics of Moving Bodies" is, that he attempted to resolve this apparent contradiction between classical mechanics and electromagnetism.

[foroyalty]

To say that time dilates in proportion to distance travelled by light is to say that time avoids contracting in proportion to the distance travelled by light which is to day that the proportion of time taken to distance travelled by light is maintained when light is reflected by a surface that is moving. However, these are merely definitions of the statement that light does not speed up as a result of its reflection by a moving surface. Why not just say that light, unlike everything else, does not increase its speed as a result of hitting a moving surface? Why say instead that the time taken for light to travel a certain distance avoids contracting and that this means that the speed of the moving surface by which it has been reflected is reduced in addition to the rate of all events that have been in contact with that surface? If the avoidance of a contradiction of the time taken for a certain distance to be travelled would mean a maintaining of the same speed of light reflected by moving surface, the same avoidance of time contracting relative to distance travelled would apply to the movement of the moving surface and the reference frame of which it formed a part meaning it and all events in connection with it would simply not speed up. Further, if a tennis ball, for example, hit a moving wall and experienced a contraction of the time it took to travel away from the wall relative to the distance it travelled this increase in speed of the tennis ball due to the time contraction mentioned would not mean that the moving wall would also increase in speed due to the mentioned time contraction and even if this were the case this would mean that when light hit the moving wall and avoided a contraction of the time it took to travel a certain distance away from the moving wall this would mean that unlike the tennis ball speeding up the wall by time contraction the light would simply avlid speeding up the wall by avoiding time contraction. Time dilation or the avoidance of time contraction relative to distance travelled does not slow down a moving surface or the reference frame of which it forms a part because it is a proportion-oriented term that is also a statement of the definition of no increase in speed which is equivalent to saying that distance travelled is not increasing in proportion to the time taken. Therefore, time is not dilating at all for either light in a moving reference frame or for the reference frame but rather the distance travelled by light reflected in a moving reference frame is what is not dilating in proportion to the time taken which is to say that light is the exception to the rule that whatever comes into contact with a moving reference frame will speed up as a result; light not speeding up does not require that time slows down in its moving reference frame. Even if this did occur, light hitting the ceiling and travelling diagonally down would increase its speed less than light hitting the wall of a train at 90° and travelling away from the wall in the direction of train travel which means the time dilation slowdown required for the wall would be greater than for the ceiling and the slower wall would eventually disingegrate from the ceiling. Alternatively, even if light would be prevented from speeding up by time dilation that would cause a less measurable decrease in the speed of the reference frame in which the light was reflected the slowed down surfaces of the reference frame would be required to experience a repeating loop of time dilation after time dilation for the surfaces of the moving reference frame because each time new light in a continuous stream of light hit a moving surface there would be a renewed requirement for further time dilation further and further slowing down the surface which would mean a much more observable slowing down of time in a moving reference frame. Finally, if time dilation can occur for different rates of the passage of time within different, separate reference frames, why not just argue that light can experience time dilation applying only to the light separately from everything else thereby avoiding the assertion that events within a moving reference frame occur at a slower rate?

[ytcollin]

My impatience makes me wish I had discovered this channel after conclusions are presented, not before!

[aniksamiurrahman]

Calibration is only required to accurately measure how much acceleration there is. Detecting the presence of acceleration doesn't need calibration. And this is a fundamental flaw in this video that puts a lot into question.

[aangulog]

The quality of this videos is amazing, as a math major I really like the approach they take, not easy neither super theorical. It's really impresive the way this man can address this topics. I'm eager for more videos (specially the logic related ones), have a nice day and thanks for the upload!

[marccawood]

Time is not relative- it just looks like it is. Clocks or other measuring devices run slower under acceleration or gravity does not mean Time is flowing differently. It means we measure time differently. Measurements are not reality, they are representations of reality based on assumptions. Measured time is relative.

[ThoughtsAreReal]

You keep interchanging velocity and acceleration. A person standing on the surface of Earth is "accelerating" at the speed of the gravitational force, like 32 meters per second per second. But their velocity, relative to the patch of ground on which they stand, is zero.

[raiangw]

Hi, I really like your videos, you are one of the few channels I what every video as soon as it is available :)
But I have to disagree with you regarding the issue with calibrating the accelerometer.
Taking as exemple the box with springs described in 6.35min, it is completelly possible to completelly build this accelerometer inside a accelerated referential frame, calibrate it there and still have it displaying the correctly that you are not inertial.

For that you just have to check if the 6 springs inside the box are identical. To do that you just have to put the 6 of them side by side and check if:
1- They all have the same length L
2- They all have the same deformation DeltaL for the same external stimulus you apply on them. (They dont even have to be a "linear spring"). Of course, it is important to make these measurements with all springs pointing in the same direction/being side by side.
Or you could also make sure to build them from the same material and that they have exacly same shape and measurements.
But anyway, once you are sure that the springs are identical, when you assemble the accelerometer, it will "point" in the direction of acceleration, even if you checked their "identicalness" in a accelerated frame.

What I will agree with you, is that the actual value of acceleration will indeed depend on the calibration (if you measure your DeltaL of check 2 in the direction of the acceleration it could be different than at an angle for example). But such accelerometer would only show 0 for a inertial observer. Also, as other people said, you can also rotate it in our hands once assembled as an extra check/test. Such accelerometer would always measure the "proper acceleration" (scalar value, independent of reference frame), and it is also a local measurement, as you can build it as small as you want,
One of my university teachers would call it an "honest accelerometer" and he would define inertail reference frame as a frame where such accelerometer measures 0. Likewise he also would define time as "what an "honest watch" measures" with similar checklist to identify an honest watch.

[isbestlizard]

It's easy to detect absolute acceleration, just rotate your accelerometer. If you get the same reading, you're not accelerating. If it changes, e.g. from being positive, to being negative, you are. Since we can assume there is no evil demon that is watching you, and causing the whole universe to change its relative acceleration exactly in sync with your rotation.

[haluk]

6:50 No it is not. Any uncalibrated measurement unit still gives you the true absolute "measurement", not output. If it is 2 times smaller than m/s², then every possible movement is (according to that information you receive). But it still measures without a reference.

[rogerfroud]

Isn't rotation absolute? I'm not a phsyicist, i'm an Engineer. it seems to me that a sphere tries to stretch perpendicular to the axis of rotation. This is always positive, and could be measured. Rotating something slower in that axis reduces the stretch until you reach a minimum at which point you know you're reached absolute lack or rotation.
I'm sure this must be wrong, but in what way?

[justaguy]

On the spring example, I was thinking no calibration is needed for you to detect the presence of an aceleration as calibration helps you to measure the value of said acceleration, but the value doesn't matter if you're only interested in determining whether or not there is some aceleration.

But then something crosses my mind: when everything is accelerating together, there's no way to detect the acceleration. If the spring is also acelerating with us, we won't be able to use it to detect our acceleration.

So I think I got your point: we can only detect our own aceleration by comparing ourselves to something that's not acelerating and, thus, aceleration is relative.

Many people here is arguing about acceleration being a measurable property - and, therefore, absolute - but I think it's because it's generally very easy to find an inertial frame, so we assume we'll always find one. But what if we are all under a force making every single particle accelerate together? How would you find an inertial frame to compare to?

I think Dialect really has got a point here.

[mhirasuna]

At 6:30, the accelerometer is calibrated so it can measure acceleration apart from gravity. But in general relativity, acceleration is equivalent to gravity. You can assume that you are not moving and measure gravity by just dropping something; no calibration needed.

[frede]

I think the key to this issue can be found when you define spacetime and its structure. At first, you define spacetime as an abstract real smooth 4-manifold. So far, so good. There's currently not much we can do in this spacetime. But then you add another assumption, namely that spacetime can hold a smooth Lorentzian metric. This is the key. For given this metric, you then find the coordinate chart that corresponds to a Minkowski metric. That then defines what you call an inertial frame (well, in general, you can only find a chart that is locally inertial, ie. there's a point where the metric is the Minkowski one, and the derivatives of the metric are zero. But the idea is the same). If there are many smooth Lorentziab metrics on the spacetime, then you simply choose one of them. That then defines a convention for what you call an inertial frame. You could choose any of these metrics, and define an inertial frame relative to it. It doesn't matter which one you choose.

[LeeCarlson]

And having "done away with Aether" we needed to create the concepts of Dark Matter, and Dark Energy when we recognized that the universe was not behaving congruently with Einstein's math.

[juliavixen]

There's a third possibly, which Einstein and Poincaré already explained 120 years ago when creating the "Theory of Relativity" : Inertia comes from confinement. Remember all that E=mc² stuff? Ask yourself: Why does _anything_ travel *slower* than light?

I'd elaborate, but I'm going to bed now. So, I'll write more when I wake up. Basically, proper acceleration of a rigid object is like the front of the object moving relative to the back of the object.

[MatthewDickau]

To everyone here, I suggest reading Tim Maudlin's book on the Philosophy of Physics: Space and Time. Brings some clarity and needed understanding to the issues presented here.
Also John Bell's Lorentzian Route to Special Relativity is a must-read.

[thibautklinger]

Hey guys. I have thought about your example with the boat and the car. It reminded me of bells paradox. Even if both spaceships accelerate at the same rate( according to the resting observer) the see each other accelerate too because the string breaks. I will at times try to use rindler coordinates to plot the cars worldline from the pov of the shore. I love how you guys can make me think deeper about things.