Clauses (1) and (2) guarantee the inclusion of all 'intuitive' natural numbers, and (3) guarantees the exclusion of all other objects. Thus, in particular, no nonstandard numbers, which would follow after the intuitive ones are admitted (nonstandard numbers are found in nonstandard models of Peano arithmetic, in which the standard natural numbers are followed by one or more 'copies' of integers running from minus infinity to infinity)1. What is problematic about this delimitation? I suspect that its hypothetical proponent would see (...) its weakest point in the unexplained concept of successor. However, we logicians know better (or at least some of us are convinced that we do): it is clause (3) which harbours the neuralgic spot, by dint of resisting any reasonable logical formalization to the point of appearing utterly void! There is, of course, no problem with regimenting (1) - we only need an individual constant 0 and a unary predicate constant N (the one whose meaning we are interested in) and we postulate.. (shrink)