On Wikipedia the d'Alembert operator is defined as
$$\square = \partial ^\alpha \partial_\alpha = \frac{1}{c^2} \frac{\partial^2}{\partial t^2}-\nabla^2 $$
However, my professor uses the notation:
$$ \square = \partial _\alpha \partial^\alpha$$
with $\partial_\alpha = \frac{\partial}{\partial x^\alpha}$ and $\partial^\alpha = \frac{\partial}{\partial x_\alpha}$
Is there a difference or are both notations equivalent
-
2$\begingroup$ In Minkowski space, what's the difference? $\endgroup$– Qmechanic ♦Commented Feb 11, 2021 at 10:01
-
1$\begingroup$ Hi Alessio Popovic. Suggestion: Instead of asking which is better, try to ask into the difference in definitions. $\endgroup$– Qmechanic ♦Commented Feb 11, 2021 at 10:12
-
$\begingroup$ More relevant is which metric you use: (-1,1,1,1) or (1,-1,-1,-1). This will add a sign to the notation. $\endgroup$– my2ctsCommented Feb 11, 2021 at 10:42
2 Answers
As it was pointed out in the comments, both expressions are the same: Note that $$\partial^\alpha = \eta^{\alpha\beta}\, \partial_{\beta}$$ and hence
$$\square = \partial_\alpha\partial^\alpha = \partial_\alpha \eta^{\alpha\beta}\partial_\beta = \partial^\beta \partial_\beta \quad .$$
We have used the Minkowski space metric $(+1,-1,-1,-1)$.
-
4$\begingroup$ Why so complicated? $\partial_\alpha$ and $\partial^\alpha$ commute. $\endgroup$– my2ctsCommented Feb 11, 2021 at 10:44
-
$\begingroup$ Interesting, my professor used the metric (+,-,-,-). Should there be a - sign? $\endgroup$ Commented Feb 11, 2021 at 10:46
-
1$\begingroup$ @AlessioPopov It is just a difference in convention. It should always be clear which convention is used. For (-1,1,1,1) there should be a sign. $\endgroup$– my2ctsCommented Feb 11, 2021 at 10:47
-
2$\begingroup$ @my2cts This answer is the proof that $\partial_\alpha$ and $\partial^\alpha$ commute. And it has to be proven, because they don't commute if the metric isn't constant. $\endgroup$– JavierCommented Feb 11, 2021 at 15:41
If operators commute, namely if commutator : $$ [ \hat {x},\hat {y}] = \hat x \hat y - \hat y \hat x = 0$$, then there's no difference in order of operators applied, as in this case, because we know Clairaut's theorem : $$ {\frac {\partial }{\partial x_{i}}}\left({\frac {\partial f}{\partial x_{j}}}\right)\ =\ {\frac {\partial }{\partial x_{j}}}\left({\frac {\partial f}{\partial x_{i}}}\right) $$
So these two expressions : $$ \square = \partial _\alpha \partial^\alpha = \partial ^\alpha \partial_\alpha $$ are equivalent.