In yesterday's NY Times, Robert Shiller writes about mood swings ("An Echo Chamber of Boom and Bust"). In particular, "... the business cycle is tied to feedback loops involving speculative price movements and other economic activity -- and the talk that these movements incite." Yes, there are mood swings among investors and consumers. No one can model them. And, Shiller suggests, we do not know how to bring them into economic models anyway.
If particles had mood swings (not original), one can wonder at the progress that physics could ever make.
Mark Thoma at Economist's View and his commenters discuss all of this, but I cannot see that we are any further along than this discussion by Guy Sorman: He cites Gerard Debreu saying, "that the only thing economists cannot do is predict. This is not quite true: Economists can predict that a bad policy will necessarily lead to a catstrophe." (p. 5).