We explore a relation we call 'anticipation' between formulas, where A anti
cipates B (according to some logic) just in case B is a consequence (accord
ing to that logic, presumed to support some distinguished implicational con
nective -->) of the formula A -->B. We are especially interested in the cas
e in which the logic is intuitionistic (propositional) logic and are much c
oncerned with an extension of that logic with a new connective, written as
'a', governed by rules which guarantee that for any formula B, aB is the (l
ogically) strongest formula anticipating B. The investigation of this new l
ogic, which we call ILa, will confront us on several occasions with some of
the finer points in the theory of rules and with issues in the philosophy
of logic arising from the proposed explication of the existence of a connec
tive (with prescribed logical behaviour) in terms of the conservative exten
sion of a favoured logic by the addition of such a connective. Other points
of interest include the provision of a Kripke semantics with respect to wh
ich ILa is demonstrably sound, deployed to establish certain unprovability
results as well as to forge connections with C. Rauszer's logic of dual int
uitionistic negation and dual intuitionistic implication, and the isolation
of two relations (between formulas), head-implication and head-linkage, wh
ich, though trivial in the setting of classical logic, are of considerable
significance in the intuitionistic context.