When I talk about measurement as “uncertainty reduction” I imply that there is some prior state of uncertainty to be reduced. And since this uncertainty can change as a result of observations, we treat uncertainty as a feature of the observer, not necessarily the thing being observed.
