There are a number of points of interest in Wittgenstein’s On Certainty but for the purposes of this module, many of the points are negative.
Knowledge is not certain. Certainty is a feature of the background, both the inherited world-picture and animal behaviour, which is presupposed. Knowledge and doubt are symmetric and part of a game of asking for and giving reasons.
The background is partly instinctive and partly inherited. Given that it makes the game of knowledge possible, there is a connection between what we can know and what we inherit. This connection puts the connection between knowledge and what we are responsible for, what is not a matter of luck, under threat. If we are not responsible for the background, and the background conditions what we can know, can we be responsible for what we know?
The answer to that might be that, firm though the boundary between what is know and what is certain in any particular case, it is not fixed for all time. Thus what can be certain at one time, can be recontextualised as a claim for knowledge or doubt given enough other changes in the rest of the background.
If knowledge is not certain – if Wittgenstein is right to reject the connection between knowledge and certainty that Descartes seems to assume – then that might make knowledge seem less valuable. Perhaps we might have thought that we value knowledge because it is certain. So if it is not, perhaps that lessens its value? Two reactions:
Even if knowledge is not certain, if something (a belief, say) is knowledge then it has to be true. You cannot have false knowledge. So knowledge still has the value that truth has. (You can think you have knowledge when you do not in fact have it: that’s the genuine problem of a lack of certainty.)
Even if knowledge is not certain, there might still be an argument against scepticism. It would go something like this.
But overall the picture that Wittgenstein gives is of a communal, naturally based (rather than intellectually based ‘all the way down’) and practical picture of knowledge based on an unquestioned background. So he provides an argument against the idea that knowledge is certain, individually verified and a matter wholly within our control.