Bonjour"s A Priori Justification of Induction

Justifications of induction, and certainly a priori justifications
of induction, are out of fashion these days. In a
chapter of his recent book, In Defense of Pure Reason
(1998)1, however, Lawrence Bonjour, the respected American
epistemologist, bucks the trend and makes a valiant
attempt to revive the latter. What he claims can be justified
a priori is that if the premise of a standard inductive argument
obtains, then it is likely or probable that the conclusion
will hold. A standard inductive premise, for Bonjour,
will state that a certain proportion m/n of observed cases of
A have been cases of B, as well as specify that there has
been "suitable variation of the collateral circumstances�
and that the "observed proportion ... converges over time
to the fraction m/n� (Bonjour, 206-07). The standard
inductive conclusion will state that there is "a corresponding
objective regularity in the world� (212), in other
words an objective regularity of the form: m/n of all As are
Bs.