More Great Words

Someone named "Josh S" at the Coordination Problem blog makes this rather insightful comment that echoes much of what I have been thinking for the last couple of weeks:
DK [well-known econ-blog reader Daniel Kuehn], then is Keynesianism a theory? If Keynesians like you and Krugman can admit that Harding's spending cuts didn't cause the never-ending death deflationary spiral you say it should (and, similarly, that Canada also didn't experience this doom scenario more recently), if you can admit that decades of stimulus haven't brought Japan's economy screaming back, if you can admit that Bernanke's ZIRP hasn't created full employment, if you can admit that rising inflation and rising unemployment can coexist, if you can admit that huge postwar spending cuts and layoffs of government employees (aka the military) didn't cause a crippling depression, if you can admit that printing and spending didn't lift Zimbabwe's economy to the heavens, then under what set of conditions could your theory possibly be falsified? 
It seems like you've always got a mental Band-Aid ready when events don't go the way your theory says they should. If there's no conceivable set of conditions under which your theory could be wrong, it's not a theory at all. It's more like a philosophical system.
Josh S speaks of economic theories, but I would extend the criticism to leftist and interventionist political concepts in general. Under what set of conditions do we second-guess our pet theories? When something becomes a closed loop it is no longer intellectually powerful. It is simply a religion.