This question already has an answer here:

In special relativity, the proper time is defined as $$d\tau^2 = c^2t^2-(x^2+y^2+z^2).$$ One usually introduces a matrix $\eta$ to represent it. I have seen two sign conventions. One has three minuses: $$\eta_3=\left(\begin{array}{cccc} 1 & 0 & 0 & 0 \\ 0 & -1 & 0 & 0 \\ 0 & 0 & -1 & 0 \\ 0 & 0 & 0 & -1 \end{array}\right),$$ and the other one has only one, so $\eta_1=-\eta_3$. Thus, $\tau^2=x^\mu\eta_{3,\mu\nu}x^\nu=-x^\mu\eta_{1,\mu\nu}x^\nu$, summing over repeated indices.

I have been told this is an issue in the physics community, and that relativity typically uses the $(-,+,+,+)$ East coast sign convention while particle physics/quantum field theory often uses the $(+,-,-,-)$ West coast sign convention. Why is that?


marked as duplicate by Qmechanic Jul 15 '18 at 8:20

This question has been asked before and already has an answer. If those answers do not fully address your question, please ask a new question.

  • $\begingroup$ I like the metric (+,-,-,-) since it leads to particles with mass going through real space time intervals $\endgroup$ – Iván Mauricio Burbano Mar 25 '15 at 17:24

Relativists tend to use the proper time, $d\tau$, and the proper distance, $ds$, interchangably. If you're working with proper time you'd expect the equation for it to look like:

$$ d\tau^2 = dt^2 + \text{other terms} $$

while if you're working with proper distance you expect:

$$ ds^2 = dx^2 + dy^2 + dz^2 + \text{other terms} $$

The sign problem comes about because the spacetime signature requires that $ds^2 = -c^2d\tau^2 $. So you end up with:

$$ d\tau^2 = -\frac{ds^2}{c^2} = dt^2 - \tfrac{1}{c^2}\left( dx^2 + dy^2 + dz^2 \right) $$


$$ ds^2 = -c^2d\tau^2 = -c^2dt^2 + dx^2 + dy^2 + dz^2 $$

Both describe identical physics, so which you use is just personal preference.

  • 3
    $\begingroup$ I find it interesting the Thomas Moore uses the (+---) convention in his Traveler's Guide to Spacetime and the (-+++) convention in A Workbook for General Relativity. Maybe we should ask him why he changed. $\endgroup$ – Bill N Mar 25 '15 at 14:03
  • $\begingroup$ Would you say that particle physicists tend to use +--- because they are dealing with proper time along a particle's trajectory (which is timelike or null) and relativists tend to use -+++ because they deal more with cosmology than particles and they would be interested in spacelike intervals (such as the radius of an AdS, for example)? $\endgroup$ – Dvij Mankad Jul 14 '18 at 11:24
  • $\begingroup$ @DvijMankad: I think the particle physics convention stems from the fact that energy & mass are much more important quantities for them than time. For the $(+ - - -)$ convention, you have $m^2 = p^\mu m_\mu$, which makes calculations easier than having to carry around a bunch of minus signs whenever you contract a 4-vector. $\endgroup$ – Michael Seifert Jul 14 '18 at 12:11
  • $\begingroup$ @MichaelSeifert Oh, I see. That makes sense. I thought that since the Lagrangian involves the $\int d\tau$ term for a particle so that might be the reason. $\endgroup$ – Dvij Mankad Jul 14 '18 at 12:46

Could you provide a simple reason for these two conventions?

The reason behind the (-,+,+,+) convention (the "mostly plus metric") is that a positive length in 3 dimensional space (e.g., the distance from my head to my toes) should still be a positive length in 4 dimensional space-time. Why should the distance from my head to my toes all of a sudden become negative just because I begin to consider special relativity?!

The reason behind the (+,-,-,-) convention (the "mostly minus metric") is that... well... There is no good reason. Clearly, you can see what side of the debate I am on.

Also, an argument from authority: Schwinger preferred the mostly-plus metric. Everybody cool uses the mostly-plus metric. For example Wald says: "We choose to use the metric signature - + + + because it is generally much more convenient than the alternative choice + - - - in that it induces a positive definite (rather than negative definite) metric on spacelike hypersurfaces."

Also, when you decide to look at thermal field theory or when you decide to wick-rotate to a euclidean field theory, if you are using the mostly-plus metric everything works out more smoothly. Whereas if you are using the mostly-minus metric everything ends up having a stupid minus sign introduced for no reason.

  • 2
    $\begingroup$ Perhaps the mostly minus metric is because this way $\tau^2$ doesn't have a minus in front of it, that is $\tau^2=x^\mu\eta_{\mu\nu}x^\nu$ without a minus? $\endgroup$ – MickG Mar 25 '15 at 8:34
  • $\begingroup$ Just a thought :). $\endgroup$ – MickG Mar 25 '15 at 8:35
  • 2
    $\begingroup$ It's psychological with me; I just can't bear to see $-dt^2 + \dots$. I've always been a $(1,-1, -1, -1)$ physicist. $\endgroup$ – JamalS Mar 25 '15 at 8:57
  • 7
    $\begingroup$ The (or, at least, one) reason for $(+---)$ is so that $p^\mu p_\mu = m^2$ instead of $-m^2$. Which is just as good of a reason as the argument about distance - why should particles have imaginary mass just because you begin to consider special relativity? ;-) Fix that and you get my upvote. $\endgroup$ – David Z Mar 25 '15 at 9:36
  • $\begingroup$ Nooo... unlike "m" the thing denoted by "p" has internal structure so $p^2=-m^2$ doesn't imply anything like "imaginary p"... In fact, that is nonsensical. To recap: $p^2\equiv p^\mu p_\mu=-E^2+|\vec p|^2=-m^2$. The mass $m$ is positive like usual, as is the energy $E$ and the square of the 3-momentum $|\vec p|^2$ and the magnitude of the 3-momentum. $\endgroup$ – hft Mar 25 '15 at 23:58

I think it depends on what you are interested in. For example, in my work considering time dependent fields in field theory coupled with gravity like inflation or quintessence, in the +--- signature you obtain mostly positive terms, intead of having to work with negative kinetic terms all the time. So do as you wish.


Not the answer you're looking for? Browse other questions tagged or ask your own question.