To give a little bit more context: "On Certainty" is about the foundations of empiricism and how they relate to the foundations of logic. It takes as a startin

To give a little bit more context: "On Certainty" is about the foundations of empiricism and how they relate to the foundations of logic. It takes as a starting point that logical deduction is the one truly certain mode of inference we humans have, then asks "how does this apply to physics?". But it does this while being extremely indirect, and presupposing knowledge of every major philosophical/scientific debate of its time, particularly the centuries-long debate over the nature of empirical inductive reasoning. Wittgenstein sometimes seems to be talking only about, say, how a certain type of sentence works, but is actually making extremely broad statements about how all of language (and linguistic communities (like scientists)) work. As with all of Wittgenstein's writings, it only seems simple on the surface. The reason this stuff is still read and thought about is because it has a lot of implications. Here's my software engineering metaphor: Wittgenstein's method is to point out particularly poignant problems in the implementation of a class, in an object-oriented system, that are meant to demonstrate problems in the underlying architecture. According to him, the way most of us think about language is wrong, and that includes science, politics, and philosophy (even though he is unwilling to say much about those subjects directly).