This paper offers an analysis of Nominative and Ergative Case systems in
the spirit of Chomsky's (1995, 1998) Minimalist Program. In particular.
Chom$ky's elimination of AGR from the inventory of admissible
functional heads leads to a re-discussion of most generative analyses on
Ergativity, which were crucially based on the role played by AGR-heads.
In this work I will show that the two Case systems can merely be seen
as the outcome of a principle of Economy which establishes that only
one Case in a transitive sentence needs to be marked. In other words,
the crucial point is not which Case is checked in an unaccusative clause,
as traditionally assumed, but rather which of the two Cases is marked in
a transitive clause. To test the validity of my proposal, I will discuss most
of the phenomena commonly linked with Ergativity, while bringing
together data from a wide inventory of different languages.