I have been battling the worst writers block I have known for a while. The past few months or so, I have been following Pete's suggestion, slightly modified, of writing 2-3 pages every other day. The ideas and examples usually bombard me such that by the time I finish typing them I lose some of them. The last week, the maximum writing I have done is 3 sentences for Wagner not including Levy's exam which was about 5 pages double spaced. The result is that I missed presenting my paper today. I had a few sentences on path dependency of expectations and I wonder if that renders it easy or difficult to model in the walrasian framework. Certainly, if expectations are assumed to be constant and exogenous then we can model them in a comparative static model. What if as is observed some expectations are not just spontaneous, but implicit? If lets say we try to model them in the neo classical or a game theoretic framework what happens? Of course I do not want to model it, but is it feasible in an alogorithm or something like Potts' formulation? I know that at some point we have to concede defeat to ignorance in itself and due to uncertainty. In the absence of the latter there is no action. My mind is churning but my muse has abandoned me!! Hopefully she will grace me soon!!