> If you want to argue something is a religion, you need to define “religion”. Wittgenstein famously argued that the word "game" cannot be given an intensional

> If you want to argue something is a religion, you need to define “religion”. Wittgenstein famously argued that the word "game" cannot be given an intensional definition – you cannot produce a list of features that all games have and which only games have. For, he argued, "game" is not a category defined in terms of singular essence, rather it is a collection of things which all have a lot in common but there is no one thing which they all have in common. Like members of a family, which all resemble each other, but all in different ways – hence he called this family resemblance (Familienähnlichkeit in German). Well, I would say the exact same thing is true of "religion" – just like "game", the word can't be defined, because there is no one thing all "religions" have in common. But, our inability to clearly define "game" doesn't make the concept useless, and isn't an inherent obstacle to using the concept. Well, the same is true of "religion". If you are going to call something a "religion" which isn't widely considered to be one, you need to identify which particular features you think it shares with those phenomena which are widely considered to be "religions". And I think Grothendieck has done that here.