Ellipsis is one of the most pervasive and mysterious aspects of the grammars of natural languages -- theoretically challenging especially in the way that it combines the properties of a discourse-anaphoric process with a sensitivity to the fine grain of morphosyntactic representations. With that much as the larger context, this talk has three particular and interlinked goals:
(1) to provide an overview of the work of the Santa Cruz Ellipsis Project and the research resources that it is beginning to make available.
(2) in that context to consider the role and the potential of corpus and annotation work as a research methodology and its articulation with the work of theory construction
(3) to review some of the principal empirical findings to have emerged from the project to date and to consider what the theoretical implications of those findings might be -- especially with regard to the question of what kind of morphosyntactic parallelism must hold between antecedent phrase and elided phrase in an ellipsis interaction.