Let us take for granted that we are subject to massive illusions regarding the world, that our senses fail to deliver the world to us veridically, that we commit many inferential errors, that we are often misinformed, and that as a consequence, we are wrong about quite a lot. In the context of pervasive illusion and error, is knowledge possible? Does knowledge require some kind of foundations? If what we take to be knowledge is nothing more than what we happen to agree about, can we even distinguish between truth and error? How can we make progress if we begin in ignorance and have only defective epistemic instruments at our disposal? These questions were at the heart of Indo-Tibetan debates in epistemology. We will explore one of those debates, which may be useful to us as we think about the edifice of knowledge and how to construct it.