Information vs. knowledge in the philosophy of science

Must we appeal to the notion of knowledge, in the subjective sense typically discussed by epistemologists, in the philosophy of science? Many scientific realists appear to think so, in so far as they assert that we can achieve knowledge of (and not merely true beliefs about) unobservable things, and of theories concerning them. As a natural result, perhaps, this has recently led Bird to suggest that scientific progress should be understood in terms of knowledge, rather than merely truth. But I would instead suggest that making scientific progress consists in acquiring useful information. I will argue for this by pointing out the scientific importance of collecting information that (a) is not and will not ever be believed, (b) is false (as it may be on some views of information), and (c) is truly believed without being justifiably believed. I will also say something about the different ways that ‘useful’ may be understood, with special reference to the property of truth and its role in positions such as constructive empiricism and structural realism. My own view is that even fictions may be both useful and informative; and this gels nicely with some recent work, e.g. by Frigg, on scientific models