A couple of quick comments that can be followed up later if you want to,
Steve.

1) Check Steve Stigler's book on the history of statistics before 1900,
which I think is subtitled, "The Measurement of Uncertainty"

2) The other source of uncertainty is in the social sciences, especially
Weber and continental progressivists. See James Kloppenberg's Uncertain
Victory.

3) As far as I know, Knight is the first to link the statistical literature
with the subjectivist literature, which creates both the strength and the
weakness of his distinction.

Ross Emmett