The other option was Wilson's multiple-model agnosticism, where neither "the sun will rise tomorrow" nor "the sun will not rise tomorrow" would be confused with reality, the thing-in-itself, but both would be seen as models that could be assessed to see which was preferable in the current situation. In this example, the "sun will rise tomorrow" model appears to be pretty useful, while the alternative appears to be rubbish and should probably be put into storage.
