We adopt in the following the ’frequency interpretation’ of probability . In contrast to the Bayesian interpretation - which
is a ’measure’ of an observer’s uncertainty - the frequency interpretation can be directly tested experimentally (of course, only with a
finite accuracy, as is the case in all other real-world measurements) and assigns an objective meaning to the notion of probability.
For a more philosophical, Bayesian approach, see e.g. Fuchs  (arXiv:quant-ph/9601020).