Most commentators on vagueness accept the Verdict Exclusion Principle: applying a predicate to one of its borderline cases yields an unknowable proposition. If three million dollars was a borderline case of 'excessive bail' for Michael Jackson in 2004, then no one can know that a three million dollar bail was excessive and no one can know that it was not. Since you should assert only what you know, you can only shrug your shoulders. Adjudicators lack this agnostic option. Their professional obligation to be decisive conflicts with the Verdict Exclusion Principle.

The second way vagueness makes judges lie is through the underspecificity of their knowledge. To prevent oppression, disjunctive indictments are forbidden. Knowing that the defendant is guilty of either A or B is not good enough. If neither alternative can be proved individually, then the defendant prevails. This standard of proof, reminiscent of intuitionism in mathematics, violates an attractive moral principle 'Known criminals ought to be punished for their crimes'. Consequently, a judge is under pressure to make assertions that are more specific than the propositions he believes. So are others involved in the process: witnesses, police, experts, and prosecutors.

All of these dutiful liars have my sympathy. My thesis is that their justice-hungry lies are part of a systemic moral dilemma created by the inevitable penetration of vagueness into law. The ingenuity with which legal scholars and vagueness theorists have sought to escape the dilemma is a testament to its poignancy. Perhaps we can ameliorate some of the difficulty with reforms that make a better trade-off between competing desiderata for a legal system. But there is no improvement that stops vagueness from systematically motivating judicial lies.