The conventional hypothesis of gametophytic self-incompatibility in the angiosperms involves one to four multiallelic incompatibility loci and the positive inhibition of incompatible pollen tubes. However, this concept does not accommodate recent experimental data indicating that there may be many loci. An alternative hypothesis which incorporates many loci and complementary pollen-style interactions suggests that there may be no S gene, as previously thought, and that gametophytic self-incompatibility is perhaps merely one aspect of extensive pollen-style interactions.