In Marian David's "Truth as the Epistemic Goal" (which is in a collection of papers on normative epistemology and truth, wherein the swamping problem is already central, or in some sense gives rise to the questions [Knowledge, Truth, and Duty: Essays on Epistemic Justification, Responsibility, and Virtue (Oxford 2001)]), she formulates the swamping problem:
Given the truth-goal-oriented approach to justification with justification understood as a means to that goal . . . it is now hard to see how justification could be anything but a constitutive means to that goal, which will make justification collapse into truth. (161)
footnote: More or less related argument are found in Sartwell (1992) (although he handles these issues rather oddly, by my lights); Depaul (1993), chap. 2.4; and Maitzen (1995).
So again, I seem first here, odd though I am. I'm going to have a look at Depaul. Also she discusses Richard Foley around these issues, and if I recall correctly Foley and I corresponded some about this in the early '90s. So I'll have a look there too. I am telling you that all roads lead back, though.
Just for the hell of it: David's wedge is to assert that it's not plausible that truth is always the ultimate goal in believing. My answer to that was in the first chapter of my unpublished book (written in '91 as I was also writing the papers). Truth as a goal is built into the nature of belief. You believe p iff you take p to be true. That's why epistemology is fundamentally 'teleological' or 'instrumentalist.'
So as I catch up, a couple of bald assertions: (1) Anything that could plausibly count as an epistemic virtue will have be truth-conducive; (2) Social factors such as being authorized within a community or engaging its consensus are neither here nor there: irrelevant to truth and to knowledge. Here's my paper "Anti-Social Epistemology" (2015).