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

About Curvature

  1. Mar 10, 2007 #1
    This may be a simple question for some of you but it has baffled me for a long time.

    When we say that spacetime is curved, do we mean that from a flat space of a higher dimension, our spacetime would appear curved, in the same way that the surface of a sphere looks curved when viewed from flat 3-space?

    And if so, then what is the stuff of this spacetime? In the case of the sphere, we know that the surface is made of some material which can be curved. But in spacetime, what is it that is curved? Is the curvature only observable through the effects it has on the motion of photons and the like? Or do we really mean that clocks run at different rates in the field and that rods measure dfferent lengths in the field and we call that curvature?

    And a related question that is bugging me is: Do we know of any mechanism by which matter/energy/pressure actually causes the curvature - besides the recipe for calculating that using Einstein's field equations? What I am trying to understand is - how does matter and energy interact with spacetime causing it to bend?

    I suppose I am getting to the following - is spacetime actually flat but endowed with a gravitational field which effects clocks and rods in such a way as to give us the impression that spacetime is curved?

    Thanks a lot,

  2. jcsd
  3. Mar 10, 2007 #2


    User Avatar
    Science Advisor

    Google for the mathematical difference between extrinsic and intrinsic curvature. Spacetime is *not* as far as we know embedded in anything.. Instead the curvature is intrinsic, it is what it is. So, differential coordinate vectors of length can receive nonunity contributions thus warping and curving our space relative to what we would expect from a euclidean treatment.

    "Do we know of any mechanism by which matter/energy/pressure actually causes the curvature - besides the recipe for calculating that using Einstein's field equations? "

    No, we only know the field equations, which give us a macroscopic definition of what happens. I think maybe you're asking about the microphysics, and that is something that is simply not known yet
  4. Mar 10, 2007 #3


    User Avatar
    Gold Member

    Whether space-time actually is curved cannot be determined for sure. In some ways it's just a name for a mathematical property that gives more than one definition for a straight line. There is a formulation of GR which reduces to a flat 3-D space, old fashioned time, and a field to represent gravity. It isn't GR but allows dynamical calculations to be made with GR metrics. It's called ADM after the people who invented it.

    There's another thread called 'a simple question about curvature' you might want to look at.
    Last edited: Mar 10, 2007
  5. Mar 10, 2007 #4


    User Avatar

    Staff: Mentor

    At the present time, asking "Why are Einstein's field equations the way they are?" in the context of general relativity, is like asking "Why are Newton's laws of motion the way they are?" in the context of classical mechanics, or "Why are Maxwell's equations the way they are?" in the context of classical electrodynamics.
  6. Mar 10, 2007 #5
    Thanks a lot for those answers. I will check out ADM. I was aware we are dealing with intrinsic curvature. I used the flat n>4 space to get to the real question which was about the essence of spacetime itself and how it is effected by matter and energy - although it is also interesting that the 4-velocity, which we would probably consider to be a physical entity, lives in the tangent plane outside the 4-d manifold. I appreciate the frank and honest answer "that is something that is simply not known yet" when it comes to how mass interacts with spacetime (and affects clock rates etc).

    If you have a large spherical mass in an otherwise empty space, and you have a test mass in the vicinity, we know that there is a relative acceleration between the two masses.

    Einstein tells us through the field equations that the large mass curves the spacetime near it and that the test mass follows a straight line (geodesic) through that curved spacetime.

    So the first part of what I am trying to understand is - does he literally mean that there is a *thing* - a physical entity - called spacetime that gets moulded and twisted and curved by energy and mass?

    Or does he simply mean that measurements would show that clocks run slower near the larger mass than farther from it and that this time gradient shows up in the metric making it differ from the metric of the flat spacetime of special relativity and so we call it a curved spacetime?

    A follow on question: Is this difference in clock rates at different heights somehow causing the relative acceleration?

    Basically I'd genuinely like to better understand why an apple falls to the ground! I am fascinateld by it! I've understood so far that it is curved time in the weak field near the earth that is the cause, and I can picture the apple happily following it's geodesic - at each moment sitting in a local lorenz frame - and then suddenly getting hit by the earth's surface.

    And that it is really the earth's surface which is actually doing the accelerating, not the apple, right? Or at least the earth's surface is resisting its own motion along the geodesics - and so it is in some sense accelerating, right?

    I'd like to better understand if (and how) this time gradient actually causes the relative acceleration between the earth's surface and the apple.

    Does anyone know of any good visualizations of the curvature of time near the earth?

    Thanks a lot.
    Last edited: Mar 10, 2007
  7. Mar 10, 2007 #6


    User Avatar
    Gold Member

    In my opinion Einstein did not believe that space-time has substance. That would take us back to aether. GR has no dynamical content, so we can't say that the difference in clock speeds 'causes' the accelerations. There are no forces in GR. ADM add dynamical content and have definitions of energy etc.

    This is a big topic, but I have to leave it there right now.
  8. Mar 10, 2007 #7
    No offence, but the highlighted parts are, frankly, garbage. The first part shows that you don't understand the experimental support for GR, and the second part (about the ADM formalism) is completely incorrect also.
  9. Mar 11, 2007 #8


    User Avatar
    Gold Member

    No offence taken. I do believe in the experimental support for GR and I don't know how you can deduce from what I've written that I don't.

    Can you tell me about any experiment that actually measures space-time curvature, as opposed to its effect ?

    I yield on ADM, not having studied it in detail.
  10. Mar 11, 2007 #9


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    There is a theorem about n-dimensional riemannian manifolds that can always (if desired) be embedded in a 3n-dimensional Euclidean space if I'm not mistaking. Now, I don't know if this also applies to a pseudo-riemanian manifold (and normally, this is not the way people look upon GR solutions).
  11. Mar 11, 2007 #10
    As an aside, I read Dirac's book on GR and he used a flat n-space (n>4) when explaining parallel transport. I liked that approach.

    It is repeated here and might be of interest to some of you: http://en.wikipedia.org/wiki/Introduction_to_mathematics_of_general_relativity#Parallel_transport

    I've also just read elsewhere in these forums that, at least in 1920, Einstein himself was arguing that spacetime was a physical entity, capable of elastic deformation.

    http://www.tu-harburg.de/rzt/rzt/it/Ether.html [Broken]

    "Recapitulating, we may say that according to the general theory of relativity space is endowed with physical qualities; in this sense, therefore, there exists an ether. According to the general theory of relativity space without ether is unthinkable; "

    I had read elsewhere however that he in later life moved away from these ideas. Can anyone tell me what the consensus is today among physicists?

    Is spacetime a *something* capable of being bent by matter and energy?

    Last edited by a moderator: May 2, 2017
  12. Mar 11, 2007 #11
    John Baez, wrote a few years ago on the Sci.Physics.Research forum:
    But I am not sure if the 91 dimensional space remains Lorentzian or is now supposed to be Euclidean. And two time dimensions, how interesting!

    I would not be surprised it is impossible to embed any Lorentzian manifold in some higher and not infinite dimensional Euclidean manifold. :wink:
    Any takers who actually know?
    Last edited: Mar 11, 2007
  13. Mar 11, 2007 #12
    I recall that Kip Thorne also briefly mentioned this (91-dimensional) complexity while lecturing on GR - http://elmer.tapir.caltech.edu/ph237/CourseOutlineA.html - see lectures 4 and 5.

    Its also interesting that none of that complexity arises in Dirac's work to a synopsis of which I provided a link above. He sails through it beautifully in a few lines. Dirac simply asserts "Physical spacetime forms a four dimensional "surface" in the flat N-dimensional space. " and continues from there. Are his assumptions incorrect?

    (And yes two time dimensions are indeed interesting - three would be even funkier! Nice symmetry with the 3 spacial dimensions... does anyone know if the physical consequences of a 3+3 space-time have been studied?)
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Similar Discussions: About Curvature