John Allison O'Neil replies

I have argued that Grunbaum's philosophy is profoundly misleading, when not actually in logical error Hanly's discussion illustrates this: he actually makes the logical errors which Grunbaum invites. Grunbaum misleads, and Hanly is misled Both embrace “eliminative induction” formally, but where Grunbaum embraces generative induction only rhetorically, Hanly embraces it outright Further, Hanly confounds generative and eliminative induction throughout, and adds other meanings of induction as well, e.g., a “correspondence” theory of induction: “the logical process as a whole was inductive because of the way it links theories to the facts of observation as a means for testing them”

“Correspondence” may be the main reason Hanly fails to grasp Popper's argument He conceives of two major theories of truth. correspondence and coherence.1 He links correspondence with realism and science, and coherence with idealism and hermeneutics He believes that correspondence requires induction, and so fears that an anti-inductivist stance would argue against a correspondence theory of truth and thus against both realism and science and for hermeneutics Popper's logic rules out induction, but he is nevertheless resolutely correspondentist and realist Popper cannot be classified by Hanly's correspondence-coherence scheme

Hanly does not accept the argument that “eliminative induction” is the serial application of modus tollens, but fails to specify what else happens, logically, in hypothesis testing Testing will argue that a “surviving” theory is superior relative to the refuted theories, there is nothing “nondeductive” in this process. Hanly alludes to such deductive selection, but insists on calling it inductive e.g., Einstein's theory relative to Newton's, or the structural theory to the topographic, or Harvey's theory to Galen's Surely he would not wish to conclude that Einstein's theory, the structural theory, and Harvey's theory are positively true Newton's theory was “truer” than Gali