The Perfect Theory

Cosmologist Pedro Ferreira has a new book about to come out, entitled The Perfect Theory. The author accurately describes the book as a “biography of general relativity”, and it’s quite a good one, of the short and breezy variety (as opposed to the detailed and exhaustive sort).

The theory’s parentage (Einstein), conception and birth are covered, in particular the way in which mathematics played a crucial role, with Einstein getting important help on this from his friend Marcel Grossman, as well as David Hilbert, who found the right dynamical equations at the same time. This material goes by fairly quickly though compared to many other sources for this history, in order to get to the main topic: the life story of the theory so far, nearly 100 years on. After the birth of the theory, it soon started to get wide acceptance, with Eddington helping to provide both the experimental confirmation in 1919 of the theory’s distinctive prediction about deflection of light by the sun, as well as ensuing publicity.

Some implications of GR were found immediately (e.g. the Schwarzschild solution in 1916), and the 1920s saw early work on applying the theory to cosmology (de Sitter, Friedmann, Lemaitre). By 1929 Hubble’s observations of an expanding universe had shown the way forward in this area. Ferreira goes on to follow several different strands of how the theory has developed. These include: black holes (Oppenheimer, Wheeler, singularity theorems from Penrose, Hawking, the information paradox), cosmology (Hoyle and steady state models, the CMB, Peebles and the now standard model, with dark matter and dark energy), relation to quantum theory (DeWitt, supergravity), gravitational waves (Weber, LIGO, proposals like LISA), and quite a few others. A wealth of different topics and interesting pieces of the history of the subject are covered, although none in great detail. The emphasis is on this history, along with the present state and prospects of the subject, with not much attempt to try and explain the intricacies of the physics (which would take a much longer book).

On the hot-button issues of string theory and the multiverse, Ferreira does a good job of giving an even-handed description of the arguments. For instance, he counsels reader to pay attention to George Ellis as well as multiverse proponents.

For more about the book, there’s a very good review by Graham Farmelo here. Oddly though, just like with his excellent biography of Dirac, which ended with a weird attempt to claim Dirac as a string theorist, here Farmelo ends by trying to enlist Ferreira and GR in the cause of string theory:

50 years later, the mathematical aesthetic of relativity has been enhanced by the beautiful demonstrations of its veracity that Ferreira describes. These would probably have made Born ponder why he and his peers did not spend more time developing a deeper appreciation of the theory soon after Einstein first presented it. Maybe there’s a lesson here for some of today’s string-theory sceptics?

I’ve never seen anyone else try and claim that the history of GR is analogous to the history of string theory. As Ferreira’s book explains, unlike string theory, GR is a classic example of a testable scientific theory, coming with one impressive post-diction (precession of the perihelion of Mercury) and followed up by an impressive test of a distinctive prediction (bending of light at the 1919 eclipse). As for mathematical aesthetics, GR uses beautiful mathematics (Riemannian geometry) and its dynamics is determined by the simplest possible Lagrangian density (the scalar curvature and nothing else). Whatever the equations of string theory might be, they remain unknown. Rather than a lesson for string theory sceptics, this book provides some good lessons about what a successful fundamental theory looks like, ones that string theory proponents would be well-advised to ponder.

“The Perfect Theory” is a good title for the book, with GR remaining our best example of a beautiful, powerful fundamental physical theory, based on the deepest mathematical ideas, with almost no free parameters. Ferreira does a great job of leading readers through the story so far of this amazing theory.

Update: For another review of the book, see Ashutosh Jogalekar at The Curious Wavefunction.

: Nature now has a podcast with an interview of Ferreira.

This entry was posted in Book Reviews. Bookmark the permalink.

21 Responses to The Perfect Theory

  1. Bob says:

    missing a “z” in Schwarzschild

  2. Peter Woit says:

    Thanks! fixed.

  3. Chris W. says:

    Scientific American has another review (also limited to subscribers).

  4. Nathalie says:

    I didn’t find the book by Graham Farmelo excellent at all. What Richard Dalitz used to tell about Dirac was far more interesting. Dalitz had known Dirac personally, but unfortunately didn’t write a book about him.

  5. Bernhard says:

    “Maybe there’s a lesson here for some of today’s string-theory sceptics?”

    It doesn’t matter how empty an argument is, you can always count with someone comparing it to Einstein and relativity. That’s a rule that works both with cranks and people with reputation.

  6. Jose Natario says:

    Typo in the third paragraph: “Fereirra” should be “Ferreira”.

  7. Peter Woit says:

    Jose Natario,
    Thanks! fixed.

  8. Giotis says:

    We should not mislead people. GR is far from perfect; in fact it’s another effective filed theory with the Einstein-Hilbert term being just the first term in an infinity series of irrelevant terms.

    The correct view is that GR is the effective filed theory of massless spin 2 particles which is valid only up to certain scale. Beyond that scale you need new degrees of freedom to describe the theory.

  9. Peter Woit says:

    Given that we don’t know what the theory is that GR is supposed to be an effective field theory for, I don’t think “the correct view” is the right terminology.

    Of course, GR is a classical theory, presumably the classical limit of something we don’t understand. That given, please, that’s no excuse for another tedious ideological argument about quantum gravity.

  10. Art says:

    Does it include refs to the literature?

  11. Peter Woit says:

    Yes, the book has extensive references to the literature. Since he’s interested in history, these are mostly to primary sources, the original papers, rather than later expository materials.

    For example, for the “string wars” he recommends looking at blog entries from that period from this blog and from a couple others…

  12. Pingback: The Perfect Theory | Dr. Goulu |

  13. Art says:

    Cool. Thx!

  14. Neil says:

    There are indeed some parallels between GR and string theory. Many experimentalist physicists considered GR a waste of time, as experimentalists do today with string theory. Some were quite prominent (and not nazis)–Millikan, Essen, Rutherford to name a few. Rutherford called it rubbish I believe. And although GR did make the deflection prediction and precession retrodiction you mentioned, most of its predictions and the ability to test them were decades in the future.

    Another interesting parallel. Einstein’s “happiest thought” was that he could transform away gravity by moving to the accelerated frame. The string theorist’s happiest thought is that he can transform away gravity by moving to the AdS boundary.

  15. Chris W. says:

    Somewhat later in the 20th century, Percy Bridgman was another contemporary of Einstein who regarded general relativity with considerable skepticism as I recall. (I made a brief attempt to locate a decent online reference for this, without success.)

    Bridgman was also primarily an experimentalist, whose writings on the philosophy of science were influential, but arguably more strongly in psychology than physics. There is a certain parallel with Ernst Mach here. Indeed, Mach himself lived to see the advent of general relativity, but did not receive it well.

  16. Peter Woit says:

    Chris W.,
    I’ve never heard of anything about “considerable skepticism” towards general relativity from Bridgman. His interest in the philosophy of science and “operationalism” made the question of how to think about special and general relativity a major concern for him, but that’s quite different.

    Mach died at the age of 78 three months after Einstein and Hilbert published the field equations, long before the 1919 experimental confirmation, so I don’t think one can reasonably try and sell him as an example of a skeptic about GR.

  17. nige cook says:

    Mach in 1913 (after Einstein’s nascent and incorrect pre-GR paper of 1911, which predicted only half the correct deflection of light by gravity) was apparently skeptical:

    “I can accept the theory of relativity as little as I can accept the existence of atoms and other such dogmas.” – E. Mach, 1913.

  18. Bernhard says:


    The history of science is full of examples of correct theories that found skepticism from from contemporaries, that parallel would not be exclusive of GR. More importantly though, is that history of science is also full of examples of incorrect theories that were correctly dealt with skepticism from contemporaries, and this is most likely the right parallel to be applied to string theory, given that the theory, unlike GR, was unable to provide predictions up to now.

    I agree that testing all predictions from GR is not easy and this is actually still work in progress (e.g. direct observation of gravitational waves). But in any case the deflection of light by the Sun prediction found almost immediate confirmation. Even if you consider the 1919 result to be biased you can still use the 1922 confirmation by the Lick Observatory. It is a joke to compare string theory to this.

    Furthermore, the examples you give are not all correct. I did not go to the trouble of checking your claims about Millikan and Essen but at least Rutherford I know changed his mind about it. On December 1919 Rutherford co-authored a Nature paper called “Radioactivity and Gravitation” (see “Studies in the History of General Relativity” by J.M. Sanchez-Ron), where he reported results on experiments to test whether the rate of radioactive substances is affected by subjecting them to high-centrifugal acceleration (given the equivalence principle). This is not coming from someone who thinks the theory is rubbish.

    Your ““happiest thought” parallel is not really worth a comment.

  19. Neil says:


    Yes, I agree that GR has proven itself capable of generating testable hypotheses, including frame dragging and gravitational lensing, in a way that string theory has not to date, and may never will. Reading the history of GR, however, it was certainly the case that GR was considered controversial and young physicists were counseled to avoid it as a dead-end. It is significant, I think, that Einstein’s 1921 Nobel citation expressly stated that his “services to Theoretical Physics” was not a reference to his relativity work.

    I thought my “happiest thought” comment is at least worthy as an amusing joke, as it was intended.

  20. Anonyrat says:

    Regarding Millikan, this is what Wiki says:

    Since Millikan’s work formed some of the basis for modern particle physics, it is ironic that he was rather conservative in his opinions about 20th century developments in physics, as in the case of the photon theory. Another example is that his textbook, as late as the 1927 version, unambiguously states the existence of the ether, and mentions Einstein’s theory of relativity only in a noncommittal note at the end of the caption under Einstein’s portrait, stating as the last in a list of accomplishments that he was “author of the special theory of relativity in 1905 and of the general theory of relativity in 1914, both of which have had great success in explaining otherwise unexplained phenomena and in predicting new ones.”

    So Millikan acknowledged that relativity had a tie to phenomena.

Comments are closed.