(I assume you meant natural numbers, as "-1" is not a useful cardinality for a set) The construction you refer to is due to Frege and Russell but it has its ow

(I assume you meant natural numbers, as "-1" is not a useful cardinality for a set) The construction you refer to is due to Frege and Russell but it has its own kinds of subtleties. In particular it does not work in ZFC which we often base upon, so we usually use Peano's axioms instead to define them (depending on your context of course). While they are not all too surprising, I wouldn't call them absolutely obvious.