I'm not sure labelling pragmatism as 'anti-epistemological' is the best approach here. Pragmatism is anti-idealist or anti-Platonic; or perhaps better put it's Bayesian, not normative. It certainly has a theory of knowledge and method (epistemology), it merely has no use for absolute knowledge or perfect method.
At any rate, pragmatists will cede the argument the moment they allow the phrase 'true fact' to enter the discussion; that phrase disrupts the core feature of pragmatism. A pragmatist might accept that there are 'truths' somewhere out there in reality, but wouldn't think they are necessarily (or even commonly) knowable or important. At best a pragmatist might suggest that a theory is a functional approximation of some unknown and unknowable truth. More likely, though, they'll say: "this theory works well enough, so we'll use it until something better comes along."
If you want to argue against this:
Well, if we take a 'true fact' to be a useful fact, then doesn't that force us to define some absolute ground for truth, usefulness, which in turn requires an epistemic grounding?
There are a few things to do:
- Ask what the phrase 'epistemic grounding' means. As best I can tell, that is word-salad pointing at the idea of some Platonic ideal.
- Point out that 'usefulness' is not an abstract concept that needs grounding. 'Usefulness' means that a theory can be used to produce a desired outcome within the constraints of a system. We don't need to go further than that.
- Clarify that within a pragmatist worldview the term 'fact' is more or less synonymous with 'observable event' or 'data point'. Facts are neither true nor false ; they merely are . The epistemological work comes in building a model which organizes or accounts for those facts. But such models (theories) are never themselves 'facts', since they can't be observed.
That should liven up the discussion a bit…