I love to hear Schiller speak, even though I don't subscribe to his particular behaviorist ideas (they are quite outside the mainstream of academic research). He is very measured in his criticism of the mainstream, and his article is a very fair summary (in my opinion) of how scientific mainstream economics is. Schiller is a moderate dissenter in that he thinks the current ration-agent-model approach is flawed, bout doesn't advocate abandoning mathematical models.

Like Schiller, I believe that it is possible to test and validate theories, but at the same time we cannot have the same certainty as in physics, and the "irreducibly human element" will also prevent the same deep mathematical theories that we see in physics, ever being developed for economics.