<p>One of the most influential epistemological theories of the 21st century is Ernest Sosa’s telic virtue epistemology. One often-cited criticism of Sosa’s theory is that the phenomenon of testimonial knowledge raises questions about the extent to which such knowledge is the result of cognitive virtues of the individual recipient of testimony, as opposed to, say, virtues of the testifier or, indeed, of an even more distributed group agent. In this paper I argue that the problem posed by testimonial knowledge for Sosa goes far deeper than his extant critics have appreciated. In fact, the case of testimony threatens the entire edifice of Sosa's virtue epistemology, an edifice that rests on a foundation of truth-directed capacities of individual agents. After sketching Sosa's telic virtue epistemology, outlining the AAA and SSS framework on which it rests, I suggest that the notion of unconscious monitoring of testimonial interlocutors represents the best prospect for a “testimonial” competence, and the best way of fitting testimonial knowledge into telic virtue epistemology. The notion of unconscious monitoring, however, stands in tension with the large body of evidence demonstrating that humans lack a relevant competence for assessing the honesty or truth reliability of their interlocutors. Given this, I present an explanation for why humans are so abjectly bad at assessing the honesty or truth reliability of their interlocutors -- namely, that their competence is actually directed at detecting social status, rather than honesty or truth reliability. Sosa’s telic virtue epistemology, however, cannot countenance the possibility of positive epistemic status accruing from competences not aimed at truth. I conclude the paper with a brief discussion of why, contra Sosa, information acquisition driven by status-seeking competences needn't have skeptical implications for the possibility of testimonial knowledge.</p>