Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

A complete course in relativity.

  1. Jan 30, 2007 #1
    I was wondering what book would be useful in learning special relativity and basic general relativity. I know vector calculus, multivariable calculus, and a bit of variational calculus. I'm looking for something that takes an informal approach to the underlying mathematics(but still rigorous). I'd liek the book to start from special relativity.

    I obviously know basic, mechanics, basic relativity, and a bit of celestial mechanics.
    Last edited: Jan 30, 2007
  2. jcsd
  3. Jan 30, 2007 #2
    If you're planning on buying a book, I highly recommend you look through university libraries/bookstores first. You'll also probably need at least 2 sources.

    I put a few free resources here. I'll be adding to them as time (& interest) allows.
    Last edited: Jan 30, 2007
  4. Jan 30, 2007 #3


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    My suggestion is that you start with "A first course in general relativity" by Bernard Schutz. That book would have been my first recommendation even if you had said that you are only interested in SR, because the chapters about SR are what really makes that book worth reading. It will also give you an introduction to GR, but in my opinion it doesn't contain enough information about the mathematics of GR. So I suggest that you also get "General Relativity" by Robert Wald. This is a great text on GR, with lots of information about the mathematics, but it says very little about SR.

    You may however find Wald's book too mathematical. Perhaps you should just start with Schutz and then decide if you want more.
  5. Jan 30, 2007 #4
    I've read some of the mathematical portions of shcutuz and siliked it for it's lack of rigour.
  6. Jan 30, 2007 #5

    Chris Hillman

    User Avatar
    Science Advisor

    Beware of "simple physics" websites

    You could try the textbook by D'Inverno. See http://www.math.ucr.edu/home/baez/RelWWW/reading.html#gtr for some other recommendations (including the books by Schutz and by Wald which have also been suggested).

    I wish to avoid debunking at PF, having already amply paid my dues in that regard elsewhere. However, I see that another poster (possibly Eugene Schubert?) mentioned a website, everythingimportant.org, which is operated by Eugene Schubert; unfortunately, this website is a typical example of "simple physics" crankery. Such sites claim that maintream textbooks "make physics look too hard". I remind serious students of Einstein's dictum: a theory should be as simple as possible--- but no simpler. I decline to discuss specifics because that is simply not worthwhile.
    Last edited: Jan 30, 2007
  7. Jan 30, 2007 #6


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    I agree. That's why I also suggested Wald for the GR stuff. It's very rigorous for a physics book. For example, the definitions of a manifold and its tangent space is done as meticulously as it would be done in a book on differential geometry. Schutz does this very poorly I think.

    However, I still think Schutz's book has excellent stuff about SR. It's the best text on SR I've seen, but there may of course be others that I haven't seen.

    One thing I like in particular about Schutz's book is its very careful explanation of tensors in the context of SR. That's by far the best introduction to tensors I've ever seen. When I read the definition of a tensor field in Wald's book, I was very happy that I had already learned the elementary stuff about tensors in Schutz's book.
  8. Jan 30, 2007 #7
    There is not a single book I can call very good. If there is such a thing as a perfect and best book, it will be the one that you are going to write yourself. Start with a blank notebook or blank papers, start writing down notes and it will turn into a book with your own approach, taste and even originality which you can use to teach others. The sources can be from anywhere. In fact some posts in this forum can be considered quite good!
    I do not agree with Chris (again!) that D'Inverno's book is good.
    I think you can try to get hold of Kip Thorne's lecture notes, under his "gravitational waves" lectures and the "application of classical physics" lectures, availiable free in the internet. (Thanks Kip!) Caltech people like Kip Thorne and Richard Feynmann are reputed to be good teachers, teachers who are able to churn out physicists.
    There is another problem with "good" books: they are very expensive! Try buying the Wheeler's "Phone Book". In fact I wanted to buy Schutz's book but ended up buying a Chinese book published in PRC at 10% the price. Not good? But it is cheap and at least serve some purpose.
    Be careful about books that teach maths to physicist though. They are not very clear and sort of like betray the mathematicians.
    Last edited: Jan 30, 2007
  9. Jan 31, 2007 #8


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    I second that. It is a great read! Of course it is only a stepping stone, but at least it gets you going. Don't look as of now for what you might read after that: you'll be able to make a better choice once you are there.

    As to those objecting to a certain lack of rigor, that's not the point. First of all you have to get yourself used to the material on an intuitive level, and that's what D'Inverno does very well. As I said, it is just a stepping stone.
    Last edited: Jan 31, 2007
  10. Jan 31, 2007 #9
    Schutz is quite good, but he may leave you a little unsatisfied in some areas (hard not too on a first pass). He does a good job on SR, but you still may want to fill in details with, say, Rindler's SR book.

    Wald is quite a difficult book IMO. I think he has a tendency to overformalize. I would try Carroll before Wald, and his book is fun to read.
  11. Jan 31, 2007 #10
    So I've narrowed it down to sean caroll's and d'inverno's. How much do each cover?
  12. Jan 31, 2007 #11
    I used Ohanian's book for an undergrad course in general relativity and thought it was one of the best textbooks I've ever used. There's a newer edition now.
  13. Jan 31, 2007 #12
    Our Chris Hillman has a handy dandy chart. Carroll assumes some knowledge of SR.
  14. Jan 31, 2007 #13
    You're studying a subject, not a book.
  15. Jan 31, 2007 #14


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    I'm answering a private message here, so I can check and edit the LaTeX stuff.

    I'm not sure what you're asking really. Isn't it the notational stuff more of a struggle when the book isn't very formal? Anyway, a big part of GR is differential geometry, and no matter what book you read about that subject, you will find that it involves a lot of struggle with the notation.

    A good thing about Wald is that it uses the abstract index notation for tensors. That makes a lot of mathematical expressions much easier to write, compared to the index free notation (which is the most common in pure math books, I think).

    Here's an example. Given a tensor T, of rank (1,2), i.e. a map

    [tex]T:V^*\times V\times V\rightarrow\mathbb{R}[/tex]

    and a vector v, we define a tensor S of rank (1,1) by

    [tex](\omega,u)\mapsto T(\omega,v,u)[/tex]

    for all

    [tex](\omega,u)\in V^*\times V[/tex].

    Now that's the index free notation. Here's the same thing (all of it, including what I described verbally) in the abstract index notation:

    Last edited: Jan 31, 2007
  16. Feb 1, 2007 #15

    I would advise "Field theory" by Landau-Lifchitz.
    Reading "Mechanics" first by the same authors would help.
    It looks a bit hard at first sight, but it goes always to the point with the most useful approach.

    Reading some more elementary book before could be useful.
    However, personally, I have found the long literary discussions quite useless and counter producting.
    You know ... these discussion with light rays and signals ...

    The central point in special relativity is the invariance of the 4-distance ds².
    This is a direct consequence of the Maxwell's equations, or the constancy of the speed of light.
    Replacing the Galilean transformation by the Lorentz transformation is a consequence, while the changes in mechanics are quite logical and natural.

    You could also read the lessons on special relativity by Einstein.

  17. Feb 1, 2007 #16
    "The central point in special relativity is the invariance of the 4-distance ds"

    I agree!!!
  18. Feb 1, 2007 #17
    Come on, I first learnt my relativity at the age of ten with this book : The large scale structure of space-time, Hawking.
  19. Feb 1, 2007 #18
    I disagree, invariance of 4-distance is not something special. What is special is the way that distance is measured.
  20. Feb 1, 2007 #19


    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    In my experience, I learned the formalism first from portions of numerous sources (course textbooks: Skinner, Lawden, Landau, Wald; various other textbooks: Taylor-Wheeler, MTW, Weinberg) ... but that's all it was.. formalism.

    It didn't start to all "click" until I was introduced to the "more physical" operational definitions using light-rays and signals and interpreting spacetime diagrams and the geometry of Minkowski and more general spacetimes. I would recommend (in addition to some of the above suggestions) the deceptively simple book "General Relativity from A to B" and the enlightening chapter on Minkowski space in "Mathematical Physics" (both by Bob Geroch), "Flat and Curved Space-Times" (by Ellis and Williams), and possibly "Relativity: The Special Theory" (by J.L. Synge). Now, the formalism makes more sense to me.

    Furthermore, it leads some down a road which asks WHY does Minkowski spacetime (or a more general Lorentzian-signature spacetime) have the various structures we now take for granted [manifold, metric (and its signature), connection, curvature, topology, and causal structure].
  21. Feb 1, 2007 #20

    I understand your point of view, but my personnal conclusion is really very much the opposite.

    For me, questions like "why the Minkowski's geometry" are meaningless.
    Analysing signal experiments and assuming c is a universal constant is nothing more than a special case of the Maxwel's equations. And these thought experiments are not better at getting the basic ideas.

    It is impossible to explain "why the Minkowski's geometry".
    It is only possible to make plausible the idea that geometry is part of physics.

    However, I also feel that theoretical physics can be a black hole if it is not supported by a real link with experience. That's why I feel important to learn about this aspect. The classical examples are the clock slowing down, the GPS, the MMX, the gravitational redshift and its observation (like mossbauer), ...

    Strong technical training in this respect is more useful than endless discussions of the first steps.
    The first step must fast and determinate.
    Going back to the basics is my motto, but the basics are mathematical.


    When I was very young I read a tale about a bug living in a bowl ...
    The idea that geometry can be explored was obvious from this tale.
    The space-time aspect doesn't change anything.
    And it is useful even for thinking about special relativity.
    Last edited: Feb 1, 2007
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?