Hacker News new | past | comments | ask | show | jobs | submit login

The history misses several salient steps, like Lorentz and Poincare describing special relativity before 1905. I guess if you want to describe the controversy of general relativity you might not want to get distracted by the controversy over special relativity... but a pattern starts to emerge.

Then the history leaves out Nordstrom's contributions to the theory of gravity which are really important if you are going to state that "It is indisputable that Hilbert, like all of his other colleagues, acknowledged Einstein as the sole creator of relativity theory," it seems Hilbert was simply willing to drop it. Almost all practitioners I am aware of are at the very least aware of the contributions of Marcel Grossman even if nobody knows about Nordstrom and others. It is a huge overstatement to say that Einstein was the sole creator.

Reading about the history of Nordstrom's theory of gravity is far more illuminating on the actual active research attempting to find a relativistic theory of gravity. In fact a student of Lorentz, Fokker, working with Einstein was able to show that Nordstrom's theory was equivalent to an expression involving the ricci scalar and a trace of the stress energy tensor. Unlike Einstein's proposal around this time, it was diffeomorphism invariant. It is likely this development, by Fokker, lead Einstein to propose the R_ij = 8\pi T_ij formulation he was pushing before the controversial period with Hilbert.

Why might this be important? Well people have a tendency to be interested in history. The extended history involving Hilbert, Nordstrom, Grossman and more is important because it is more illuminating to the reality of how physical theories are actually developed. It turns out that maybe Einstein doesn't deserve the level of hero worship he gets, which certain types of people may find invigorating. Also, this episode shows that petty squabbles and politics exist in "modern" science.




I would love to read a deep dive into the history of (semi) modern science. In particular, one that would avoid pop-science analogies and that is unafraid of technical detail. Ideally one that is more focused on the intricacies and lineage of research than the particulars of researchers' personal lives.

I feel that academics often have significant insight into the history of their own fields and personalities therein, but that knowledge is rarely condensed/disseminated. Perhaps it's because matters can get quite political and subjective at that level.


In physics there is a rich literature documenting its history. And as you mentioned, it is scattered around books, periodicals and scientific journals. A very small selection for you and any people interested:

From Emilio Segre (Nobel Laureate) there is this 2 book series.Recommended. - From Falling Bodies to Radio Waves [1] - From X-rays to Quarks [2]

Then, you have Abraham Pais, all his books are highly recommended, I will include 2 here

[3] Subtle is the lord (Best Einstein Scientific Biography) [4] Inward Bound (Superb Scientific history of XX century physics)

Jagdish Mehra wrote a gargantuan history of Quantum Mechanics in six volumes. He studied and /or interviewed most of the big hitters who developed the theory. The science content is high so you need a good foundation.

[5] The Historical Development of Quantum Theory

Richard Westfall wrote the best Newton biography that I know.

[6] Never at Rest: A Biography of Isaac Newton

Finally the issue number 2 of the 72nd volume of Review of Modern Physics [7] is a gem; packed with historical reviews of the development of all fields in physics during the 20th century, written by eminent people.

Of course this is just a minuscule sample of an extraordinary bibliography. Sadly life is so short to make it justice.

[1] https://www.amazon.com/gp/product/0486458083 [2] https://www.amazon.com/gp/product/0486457834 [3] https://www.amazon.com/dp/019853907X [4] https://www.amazon.com/dp/0198519974 [5]https://www.amazon.com/Historical-Development-Quantum-Theory... [6] https://www.amazon.com/dp/0521274354 [7] https://journals.aps.org/rmp/issues/71/2


Thanks for the recommendations! Found a couple of books I'd like to purchase, read the first twenty or so pages at one go, then relegate it to the pile along with the 50 others that have been replaced by new ones.


Academics really do not have significant insight into the history of their own field. The problem is really, that history makes for a nice introductory paragraph and by the time you can call yourself an academic you are very confident in your knowledge of history of the field because you read hundreds of times the introductory paragraph.

To make matters worse, the introductory paragraph was written by someone who has no training as an historian, and therefore also just paraphrases some tradition of introductory paragraphs.

As a concrete example, Newton's laws are actually not in the principia, but instead only appear a hundred years or so after his death. ( It is actually not unreasonable to still call them Newton's laws, but the argument is a lot more complicated than "Newton wrote Newton's laws.")


Oxford's Newton Project text of the book has the Newton's Laws in the 1687 version here: [http://www.newtonproject.ox.ac.uk/view/texts/normalized/NATP...]

What makes you think they were not?


I had a professor briefly that studied the history of mathematics. During lectures while he was developing out proofs he would often describe the historical context in which the proof was first developed. Galois has a particularly interesting (tragic) story.

There are a lot of interesting historical figures. Probably most people stick to the figures most familiar to them in their own field. The histories of Dijkstra and Turing come to mind... it strikes me that I don't know a lot about Knuth.


Speaking of Knuth, he is extraordinary in this regard apparently as the rigor of his expositions matches his attention to the historical details. Add to that his meticulousness for properly (and aesthetically) spelling people's names for good measure.


"What is real?" by Adam Becker. It doesn't touch much on general relativity if you are interested specifically in that, but it's a great read and gives you a bit of perspective how "modern" science works (I mean before the Internet, and back when we were actually moving forward..)

> Perhaps it's because matters can get quite political and subjective at that level.

Much more than I would have thought.

https://www.goodreads.com/book/show/35604796-what-is-real


Yes, I remember being surprised when reading in Poincaré's "La science et l'hypothèse" a variant of the famous formula, probably m = E/c^2 ... but in a different context than Einstein's (maybe the 3-body problem?).


Einstein gets the hero worship mostly because of the Annus mirabilis papers on the photoelectric effect, Brownian motion, and Special Relativity. General Relativity was more a curiosity of sorts until the 1970s, and even today people connecting astrophysical observations with Einstein gravitation tend to work with a (often linearized) approximation for practical reasons and with a range of theoretical justifications and mechanisms to connect back to the Einstein Field Equations as the underlying more-fundamental theory.

Practically every grad student encountering Kaluza-Klein (and other supergravities) or the PPN formalism will spend some time with Nordström's gravitation (and of course it is discussed in §27.6 and §38.2 of Misner Thorne & Wheeler, the gold standard textbook). In the latter case it is straightforward to see in modern terms where Nordström's 1913 theory would clash with solar-system observations (it was ruled out by the Eddington eclipse observation) and spectacularly falls apart for compact massive objects like the Hulse-Taylor binary. The theory inverts the spatial curvature generated by matter (PPN \gamma parameters for Newton, GR, and Nordstöm 1913 are 0, 1, and -1 respectively) and is only half-right for gravitational nonlinearity (\beta parameters 0, 1, and 1/2). Einstein-Fokker 1914 fails under PPN analysis in the same way, except that they properly fix the \zeta_{4} dynamical conservation parameter as 0, unlike Nordström 1913.

In my view, early 20th century alternatives to GR are interesting for how they failed. I accept the part of your argument that it is more interesting to some than to others about how much cross-fertilization there was among the developers of these theories. However, only GR has survived contact with all experimental tests, and it has proven extremely difficult for alternatives to match over multiple solid angle and wavelength scales what astronomers observe. The ways in which various extra-field approaches have failed are frankly more useful for theoretical physicists than the ways in which they were developed in the first place. And for astrophysicists, the only thing that matters is what has not -- so far -- failed.

So however he got there, at least with respect to gravitation Einstein deserves recognition for having produced pretty much the only viable physical theory. There's really only Jordan/Brans-Dicke in the limit of a vanishing \frac{1}{\omega} parameter, but at zero the theory reduces to GR itself; Einstein-Cartan-Sciama-Kibble if one does headstands to suppress spacetime torsion; early-decay bimetric theories, where the second metric field decouples or vanishes before nucleosynthesis; supergravity, if one figures out how to hide the gravitino; and string theory (M-theory), if one figures out the landscape problem and/or works out a general emergence of a PPN match. Notably every one of these evolved from General Relativity mostly because of its unique successes -- however, I think it is more interesting and better to say that these remain viable because they can in certain circumstances fully reproduce General Relativity.


> In fact a student of Lorentz, Fokker, working with Einstein was able to show that Nordstrom's theory was equivalent to an expression involving the ricci scalar and a trace of the stress energy tensor. Unlike Einstein's proposal around this time, it was diffeomorphism invariant.

"I know some of these words."


Ricci scalar: number assigned to each point of a space describing curvature at that point

Stress-energy tensor (field): multidimensional quantity assigned to each point of a space describing how matter and energy are concentrated and moving at that point.

Trace: coordinate invariant transformation that turns a tensor field into a scalar field.

Diffeomorphism invariant: the object in question has the same description regardless of arbitrary coordinate changes


IIRC from the book Chaos it was often referred to as Poincaré’s relativity theory in France.





Applications are open for YC Winter 2020

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: