The article examines four areas within ergativity that merit further consideration, including with respect to child language acquisition data: (1) syntactic alignment, including in particular the delimitation of semantic alignment from other kinds of alignment with lexical exceptions, and symmetric voice systems; (2) alignment splits, in particular the interpretation of exceptions to generalizations relating to animacy and definiteness; (3) syntactic alignment biases, especially the interpretation of exceptions; and (4) diachronic issues, in particular the decay of ergativity giving rise to idiosyncrasies like horizontal alignment or to conventionalized use of erstwhile ergative morphology.