It may seem to be the perfect solution: it allows us to say we know lots of stuff, without having to go down the route of proving the skeptic is wrong. But the problem is that if the contextualist is right, knowledge has shifting standards, and that simply doesn’t seem to be reflected in the way that the word works.
Consider how, intuitively, whether someone has knowledge depends on how justified his beliefs are. And this seems to be all that truly matters. So if two people hold true beliefs and they both have equally valid and sufficient reasons for believing what they do, we must surely conclude that both people know the fact in question.
And yet contextualists must deny this! They must say that for two people with the same reasons for believing something, one may know and the other may not depending on, yes, the context. And what’s more, a single person may go from knowing something to not knowing something once they change from chatting in a classroom to entering the magistrates. And it’s undeniably the case that this seems damn weird. I repeat: the reasons for believing something stay the same, but the presence of knowledge may alter.
Secondly, recall the analogy to a conversation about flatness. It seems that what’s noteworthy here is a real disconnect between the two examples: when discussing flatness, it’s quite easy to see our mistake and tidy up our definitions so that we clarify what we mean. Does anything similar happen in conversations with the skeptic? On the contrary! We dig in our heels and insist we either do or do not know. The last thing we do is to decide we never contradicted one another in the first place. So the contextualist must say we are completely conned and misled by our own language.
These are some truly awkward implications, which you can only judge the importance of for yourself.