For starters I may be mistaken bringing this up as a notation problem, maybe for some reason unknown to me this is not about notation at all and I simply do not understand the topic well enough yet, this uncertainty tho is precisely why I am asking this in the first place.

In Modern Quantum Mechanics Sakurai calls the eigenvectors of an operator $A$:


But then he goes on writing things like this one:

$$\sum _{a'}c_{a'}|a'\rangle$$

Now I get that he means to cycle over all the eigenvectors of $A$, but this seems to me a really confusing way to write it. Written like this it seems that $a'$ should cycle from 1 to $N$, but of course this is not what it means. A much better notation would be calling the eigenvectors $$|a^{(1)}\rangle,|a^{(2)}\rangle,...$$ and then to write: $$\sum _{n=1}^N c_{n}|a^{(n)}\rangle$$

This can seem a detail, but for me it gets very confusing when we start dealing with things like this: $$\sum _{a'}\sum _{a''}\langle a''|X|a'\rangle$$ in this expression seems that we are only dealing with $|a'\rangle$ and $|a''\rangle$ when in fact we are cycling on all the eigenvectors.

Is there something that I am missing? Why do we use such confusing notation?

  • $\begingroup$ Maybe a bit confusing because of the first equation but things like $\sum_{mm'}$ are very common in mathematics. $\endgroup$
    – JamalS
    Aug 19, 2020 at 21:18
  • 1
    $\begingroup$ You're not missing anything; Sakurai just has bad notation. $\endgroup$
    – knzhou
    Aug 19, 2020 at 21:26
  • $\begingroup$ He was heavily inspired by Dirac's notation in his QM book. $\endgroup$
    – DanielC
    Aug 19, 2020 at 22:02

1 Answer 1


Short answer: no you're not missing anything. The notation is indeed a bit vague.

However, for completeness:

As @JamalS noted in a comment, the notation is not that uncommon. However, in mathematics the "good" notation would be $\sum_{a\in A}$ where $A$ is the set over whose elements we are summing (in your case this would be the spectrum of a given operator). Your example of natural number is then the case where $A=\{1,...,N\}$. But when it is clear from the context, people (physicists) tend to drop the set A and just write the summation index.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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