Why the hardest logic puzzle ever cannot be solved in less than three questions (Q427236)

The puzzle \textit{G. Boolos} described in [Harvard Review of Philosophy 6, 62--65 (1996)] as the hardest ever has been surpassed by \textit{B. Rabern} and \textit{L. Rabern} [Analysis, Oxf. 68, No. 2, 105--112 (2008; Zbl 1143.03317)], and theirs in turn by \textit{G. Uzquiano} [Analysis Oxf. 70, No. 1, 39--44 (2010), \url{http://analysis.oxfordjournals.org/content/70/1/39}], and each has provided a two-question solution to his predecessor's puzzle. In this paper the authors show that Uzquiano's puzzle cannot be solved in two questions, and provide a solution in three. The paper concludes with a new puzzle with the respondents True and Random, but with False replaced by the new character Devious.