I have argued that the legitimization argument for the autonomy thesis and the corresponding bracketing strategy fails. In doing so, I have tried to establish, moreover, that the autonomy thesis is false, and that the bracketing strategy is not an option. I am arguing, that is, that if we wish to achieve the aims we have as philosophers
of science, which include developing satisfying accounts of scientific representation, modeling, explanation, idealization, confirmation, and the like, and we seek to do so by engaging in talk of models, mathematical structures, or fictional objects, as many do, then we will have to address metaphysical questions about the existence and nature of such creatures.
Let me close with an analogy to underscore the point, and to up the polemical ante on the way out. Suppose we are philosophers of mind trying to come up with an account of the relationship between the mental and the physical which explains the correlations we observe between mental and physical events. Suppose then that one of us says, “Here’s my account: God is constantly intervening in our mental lives to produce pain sensations in us when we’re kicked, sensations of blueness when we’re standing in front of blue objects, and so on; God is also constantly intervening in the physical world to make sure that my leg moves when I decide to kick something, and so on.” A little old-fashioned, but fair enough. Suppose, however, that our colleague then goes on to say, “I take no stand on whether there is any such thing as God [the existence question], nor on whether, if there is such a thing, it’s the kind of thing that would take any interest in human affairs [the nature question]. Those are questions for metaphysics and the philosophy of religion; I’m just going to stick to the philosophy of mind. And my account of the question before us as philosophers of mind is the account I gave a moment ago.” Such a philosopher would be employing the bracketing strategy; and it clearly wouldn’t do.
-  See section 2 for more about what I do and do not mean here by “address.”