Find Einstein's GR Papers: Read His Own Words

[Jonnyb42]

I asked this in one of my previous topics, but I just cannot find where I asked it, so here I am asking it again.

Can someone give me a link to a web version of Einstein's GR. I want the one with Einstein's OWN WORDS, not reworded (other than simple language translation, if he did not write them in English.) Many books and webpages I find on GR are just summaries of what Einstein said, but where can I get/find his own published papers? (copies of course)

Thank you to anyone that can show me.
Einstein was brilliant and I want to learn GR from his words only.

 

[Dale]

I want to learn GR from his words only.

I think that is a bad idea. GR has progressed in the last >90 years.

 

[atyy]

http://www.Alberteinstein.info/gallery/gtext3.html

The major conceptual error in Einstein's work is the relationship between "general covariance" and "no prior geometry".

Another tricky point, which Einstein was aware of, but many new to the subject are not, is that the equivalence principle neither specifies general relativity completely (there are other theories compatible with some form of equivalence principle), nor holds more than "locally" (defined in a very technical way). The texts by Misner Thorne and Wheeler, and by Rindler cover these.

 

[HallsofIvy]

A quick google search turned up this:
http://www.Alberteinstein.info/gallery/gtext3.html

 

[Jonnyb42]

I have found an interesting part, I have read it before but never asked, and thank you again for the replies. This has to do with the principle of equivalence.

That to explain the accelerated motion of other masses, instead of claiming the coordinate system is accelerated, those masses can be said to be in a gravitational field, my question is, that would mean there is a linear gravitational field everywhere, which has never existed before, and where is the mass that corresponds to that gravitational field?

 

[PAllen]

I have found an interesting part, I have read it before but never asked, and thank you again for the replies. This has to do with the principle of equivalence.

That to explain the accelerated motion of other masses, instead of claiming the coordinate system is accelerated, those masses can be said to be in a gravitational field, my question is, that would mean there is a linear gravitational field everywhere, which has never existed before, and where is the mass that corresponds to that gravitational field?

That's exactly why you need to talk about the equivalence principle locally. Sufficiently locally, any real gravitational field is uniform.

 

[Jonnyb42]

Alright, but just a bit earlier from where he said that, he talks about two fluid bodies, where either is rotating about their common axis, each one sees the other, but one sees the other an ellipsoid and the other sees the one as a sphere. The question is what is the reason for the difference in the two bodies? Where is this answered? I hoped the equivalence principle answers this, but please could someone tell me what is the answer or show me where is the answer. If it is the EP that explains it, then I do not understand and would appreciate it if someone explained it to me.

 

[atyy]

Skip that part. Einstein was confused.

 

[Jonnyb42]

Skip that part. Einstein was confused.

What!? I am confused! What solved Einstein's confusion then? That is a very big problem I have been trying to solve. Is that problem ever mentioned at all in his paper? Are you saying it had nothing to do with anything!??

 

[Dale]

Which is one reason why trying to learn this way is a bad idea. Learn from the advances that have been made in the last century.

 

[Jonnyb42]

No, I have had this problem for a long time before I read anything by Einstein. Could anyone else tell me how he was confused?

 

[PAllen]

No, I have had this problem for a long time before I read anything by Einstein. Could anyone else tell me how he was confused?

Can you clearly state the actual issue you are confused about? Then I'm sure someone here can help you.

 

[bcrowell]

Alright, but just a bit earlier from where he said that, he talks about two fluid bodies, where either is rotating about their common axis, each one sees the other, but one sees the other an ellipsoid and the other sees the one as a sphere. The question is what is the reason for the difference in the two bodies? Where is this answered? I hoped the equivalence principle answers this, but please could someone tell me what is the answer or show me where is the answer. If it is the EP that explains it, then I do not understand and would appreciate it if someone explained it to me.

Skip that part. Einstein was confused.

I disagree with atyy's take on this. IMO Einstein was not confused. Einstein was stating what properties he thought a theory of gravity should have. He thought it should embody Mach's principle. GR doesn't turn out to embody Mach's principle as completely as Einstein thought it would. The question then becomes whether the *universe* actually embodies Mach's principle in the way Einstein thought it would. The answer, based on experiments from the 1970's, is basically no. There is a good popular-level book on this called "Was Einstein Right?" by Clifford Will.

 

[atyy]

What!? I am confused! What solved Einstein's confusion then? That is a very big problem I have been trying to solve. Is that problem ever mentioned at all in his paper? Are you saying it had nothing to do with anything!??

No, I have had this problem for a long time before I read anything by Einstein. Could anyone else tell me how he was confused?

Einstein's contributions to special relativity include an axiomatizing (Fredrik, keep out of this :tongue2:) of the theory. The whole theory does more or less fall out of his two principles, so he kind of "cleaned up" (and quite a bit more) what Lorentz, FitzGerald, Poincare etc had already contributed messily, the only modern thing he didn't have is Minkowski's contribution.

In the case of GR though, he more or less blundered his way to the correct equations. So he got the correct equations in spite of being confused, and he was still confused after he obtained the correct equations (amazingly, he still managed to derive the correct perihelion correction). Schwarzschild was very confused about coordinate and curvature singularities when he discovered his famous solution. The modern textbook presentation of Einstein's and Schwazschild's discoveries differs from the original presentations in many details.

 

[bcrowell]

I'm a big fan of Einstein's scientific writing. He did a great job of writing clearly about concepts. Both the intro to "The Foundation of the General Theory of Relativity" and "Relativity : the Special and General Theory" are wonderful. Anyone who reads and understands both of those things will understand more about relativity than 75% of the people who post about relativity on PF.

That's not to say that one should make an affection out of taking a "great books" approach to physics. As DaleSpam points out, we know a lot more about relativity now, 100 years later. This is a case where "the best book on any subject is two books." I.e., read Einstein, but also read other treatments. Two introductory books I like are Gardner's "Relativity Simply Explained" and Taylor and Wheeler's "Spacetime Physics."

 

[atyy]

I disagree with atyy's take on this. IMO Einstein was not confused. Einstein was stating what properties he thought a theory of gravity should have. He thought it should embody Mach's principle. GR doesn't turn out to embody Mach's principle as completely as Einstein thought it would. The question then becomes whether the *universe* actually embodies Mach's principle in the way Einstein thought it would. The answer, based on experiments from the 1970's, is basically no. There is a good popular-level book on this called "Was Einstein Right?" by Clifford Will.

OK, here is the correct :tongue2: version. Wiki Newton's bucket. That demonstrates that the bucket rotates relative to absolute space. What is absolute space? The metric, the gravitational field. (I learned this from Rovelli.)

 

[Jonnyb42]

Ok, I have another question, which is ultimately the main question. In the paper, Einstein says:

"The laws of physics must be of such a nature that they apply to systems of reference in any kind of motion."

Does GR satisfy that nature?

Or was Einstein confused here as well... :(

 

[atyy]

Ok, I have another question, which is ultimately the main question. In the paper, Einstein says:

"The laws of physics must be of such a nature that they apply to systems of reference in any kind of motion."

Does GR satisfy that nature?

Or was Einstein confused here as well... :(

Einstein was correct, but it has nothing to do with general relativity specifically. This is true even of special relativity and Newtonian physics. This property is called "general covariance".

The distinguishing property of GR is "no prior geometry", not "general covariance".

 

[Jonnyb42]

Um, how can it be true for Newtonian mechanics? the laws there aren't applicable in accelerated frames, right?

 

[aashay]

General covariance does not hold for Newtonian Physics as well as special relativity. It is only applicable for general relativity. In fact the very nature of curved space-time requires the laws to be generally covariant.

 

[atyy]

See, till this day people are confused due to reading Einstein!

Newton's theory in generally covariant form is described here as Newton-Cartan theory. http://arxiv.org/abs/gr-qc/0506065: "(1) It shows that several features of relativity theory once thought to be uniquely characteristic of it do not distinguish it from (a suitably reformulated version of) Newtonian gravitation theory. The latter too can be cast as a “generally covariant” theory in which (a) gravity emerges as a manifestation of spacetime curvature, and (b) spacetime structure is “dynamical”, i.e., participates in the unfolding of physics rather than being a fixed backdrop against which it unfolds."

 

[aashay]

See, till this day people are confused due to reading Einstein!

Newton's theory in generally covariant form is described here as Newton-Cartan theory. http://arxiv.org/abs/gr-qc/0506065: "(1) It shows that several features of relativity theory once thought to be uniquely characteristic of it do not distinguish it from (a suitably reformulated version of) Newtonian gravitation theory. The latter too can be cast as a “generally covariant” theory in which (a) gravity emerges as a manifestation of spacetime curvature, and (b) spacetime structure is “dynamical”, i.e., participates in the unfolding of physics rather than being a fixed backdrop against which it unfolds."

But that is done by Cartan and not by Newton! Its a re-formulation of the theory. And there is no concept of curved space in Newton's theory. So no need of General covariance.

 

[atyy]

But that is done by Cartan and not by Newton! Its a re-formulation of the theory. And there is no concept of curved space in Newton's theory. So no need of General covariance.

We can use any coordinate system we like in Newtonian physics.

 

[Jonnyb42]

So does anyone have an answer to the fluid bodies question?
All I have so far is that Einstein was confused. Whether he was confused or not, that is still a question I haven't seen an answer to yet. Thanks again for all the replies.

 

[atyy]

http://arxiv.org/abs/1007.3368

Provocative stuff from a careful thinker.

 

[PAllen]

http://arxiv.org/abs/1007.3368

Provocative stuff from a careful thinker.

And he is the rare exception to the rule that if you work a long time in isolation from academia, you will likely produce crank work. Though, after taking the time he needed, he now publishes regularly in peer reviewed journals.

 

[aashay]

We can use any coordinate system we like in Newtonian physics.

I still don't understand how can you say that. Clearly Newton's laws are not valid in accelerated frames of reference and there is no concept of curved space-time either. Can you please clarify?

 

[PAllen]

I still don't understand how can you say that. Clearly Newton's laws are not valid in accelerated frames of reference and there is no concept of curved space-time either. Can you please clarify?

This is actually all covered in Misner, Thorne, Wheeler book Gravitation, including a synopsis of the history of confution on general covariance versus 'no prior geometry'. A slightly different attempt to make this distinction is one pursued by James L. Anderson over the last 40+ years, of defining the idea of the symmetry group of a theory as distinct from its covariance group; and the GR unlike alternative theories has the manifold mapping group as its symmetry group. The Anderson approach is, I believe, considered a bit controversial, and as recently as 2010, there are papers arguing pro and con. MTW praises this approach, but does not follow it their text.

 

[atyy]

I still don't understand how can you say that. Clearly Newton's laws are not valid in accelerated frames of reference and there is no concept of curved space-time either. Can you please clarify?

We start with Newton's laws in an inertial frame. When we transform to a noninertial frame, we pick up "Christoffel symbols". If instead of defining "same form" without Christoffel symbols, we define it as including the Christoffel symbols, then Newton's laws are valid in any frame.

It is the same with special relativity, which is capable of handling accelerated frames (eg. Rindler coordinates).

I like http://arxiv.org/abs/gr-qc/0603087 .

 

[atyy]

And he is the rare exception to the rule that if you work a long time in isolation from academia, you will likely produce crank work. Though, after taking the time he needed, he now publishes regularly in peer reviewed journals.

Ahh, too bad, what a sell out

 

[aashay]

We start with Newton's laws in an inertial frame. When we transform to a noninertial frame, we pick up "Christoffel symbols". If instead of defining "same form" without Christoffel symbols, we define it as including the Christoffel symbols, then Newton's laws are valid in any frame.

It is the same with special relativity, which is capable of handling accelerated frames (eg. Rindler coordinates).

I like http://arxiv.org/abs/gr-qc/0603087 .

But the moment you bring Christoffel symbols into the picture, don't you march away from Newtonian Physics to General Relativity by taking into account the effects of curvature? And all this has been done quite recently after GR was developed and not at the time of Newton.vCorrect me if I am wrong as I am quite new to General Relativity.

 

[atyy]

But the moment you bring Christoffel symbols into the picture, don't you march away from Newtonian Physics to General Relativity by taking into account the effects of curvature? And all this has been done quite recently after GR was developed and not at the time of Newton.vCorrect me if I am wrong as I am quite new to General Relativity.

http://www.mth.uct.ac.za/omei/gr/chap6/node4.html

You can have Christoffel symbols in flat space, eg. when you use polar coordinates.

The essential idea is that the Christoffel symbols are first derivatives of the metric, where the Riemann curvature tensor which indicates the difference between flatness and curvedness is made of second derivatives of the metric.

In GR, fake gravity (from acceleration) makes first derivatives of the metric, whereas true gravity (more properly, tidal gravity) is due to curvature. (I'm sure bcrowell is going to disagree with my terminology here!)