<p>The pragmatic encroachment thesis is the claim that pragmatic factors are relevant in determining whether a subject knows that p. Starting from the much-discussed Bank and Train Cases, Wolfgang Barz has argued that the examples used to illustrate, and so support the existence of, pragmatic encroachment, are misdescribed. Barz proposes that we do not have, as has been assumed, two subjects, A and B, who have the same evidence for a proposition, p, where A knows that p, and B, in a situation involving some different pragmatic factor, does not know that p. Rather, A knows one proposition, and B does not know another, different, proposition. Against Barz, I argue that there are serious problems involved with the specification of the pairs of propositions that Barz proposes to deal with the apparent pragmatic encroachment in the example cases he discusses. In addition, Barz’s account would force us to an untenable view of full retractions of knowledge claims made in the face of changed pragmatic conditions. Had Barz been right, pragmatic encroachment would not exist, and, in addition, he would have provided the basis for a very strong argument against contextualism.</p>