We present a method for automatically annotating treebank resources with functional structures. The method defines systematic patterns of correspondence between partial PS configurations and functional structures. These are applied to PS rules extracted from treebanks. The set of techniques which we have developed constitute a methodology for corpus-guided grammar development. Despite the widespread belief that treebank representations are not very useful in grammar development, we show that systematic patterns of c-structure to f-structure correspondence can be simply and successfully stated over such rules. The method is partial in that it requires manual correction of the annotated grammar rules.