I present a novel interpretation of Frege's attempt at
Grundgesetze I §§29--31 to prove that every expression of
his language has a unique reference. I argue that Frege's proof is
based on a contextual account of reference, similar to but more
sophisticated than that enshrined in his famous Context Principle. Although Frege's proof is incorrect, I argue that the account of
reference on which it is based is of potential philosophical
value, and I analyze the class of cases to which it may
successfully be applied.