[shash]

I disagree entirely, especially about history. You form hypotheses on the basis of evidence (literary, archaeological, documentary, etc), make predictions about the kinds of effects you’d expect to find in the historical record, and then modify those hypotheses based on what you find later.

How is that not a science?

[yk]

> make predictions about the kinds of effects you’d expect to find in the historical record,

You want to make predictions about the past?

[shash]

No, predictions about what other undiscovered evidence you’d expect to find about the past. In a way this is like astrophysics in that you can’t conduct experiments, but you can predict that X is correlated with Y. Come to think of it, paleontology...

For example, take the Shakespeare authorship controversy; you could create a hypothesis that say “Shakespeare was indeed the author of Hamlet”. A prediction from this might be, “if a manuscript of Hamlet were ever found it would be in Shakespeare’s handwriting”. Not a really good example to be sure, but just off the top of my head...