In general relativity, applying this concept has led to numerous technical and conceptual issues, not only when attempting to quantize it, but also when interpreting the theory. In this talk I will examine this concept for theories with symmetries, and introduce an important distinction between global and local gauge symmetries. I will argue that the diffeomorphism invariance of general relativity cannot be interpreted as a local gauge symmetry, and that this has an impact on the way we should think about the observables of the theory. In particular, this analysis shows that some “timeless” interpretations of general relativity, popular in the quantum gravity literature, are profoundly misguided. After all, the physical content of general relativity is closer to the textbook version of the theory than to some timeless reinterpretation.