I'm pretty sure we're going to be able to have nesting/hierarchy both in our feature structure declarations and in our feature structures inside entries. I have one possible simple instance relating to clitic particles which will make a good test case, if ECH confirms what I believe to be the intended analysis based on her text.