This paper argues for reducing the inventory of grammatical functions by eliminating the function COMP, standardly assumed in LFG to be assigned exclusively to clausal categories. Using evidence from Catalan, Spanish, Malayalam, and English, we show that this move is desirable not only because it results in a simpler framework (a framework with fewer constructs), but also because it yields simpler and more revealing analyses. As the position we are arguing for is in fact the null hypothesis, our arguments for the elimination of the function COMP are necessarily framed in the context of the arguments that have been proposed in favor of COMP.