Fenton, These are rhetorical questions. You begin a sentence with "Who is to say?" or "What is to say?" in order to cast doubt on the statement that follows. No answer is expected The meaning of "Who is to say?" is "Who could possiby say?" Who could possibly say that we are to blame? No one could. What is to say? = What could possibly say? Here "say" means "show or indicate" because it applies to a thing not a human being. What could possibly show that we are to blame? Nothing could. Grammar The infinitive "to say" modifies the subject "who" or "what". You could rewrite them like this: Who, to say that we are to blame, is there? What, to say that we are to blame, is there?

Who's to say = who has the authority to declare.... What's to say = who cares if....

In English, the two are often used interchangeably. It's not necessarily correct, but that's just how we use them. Both of them are generally used as rhetoric questions. "What's to say" is more often used as a sort of challenge question and if you substitute it with "maybe" or "how do you know", then it might make more sense: "What's to say I won't be able to learn another language?" -- ("How do you know that I won't be able to learn another language" OR "Maybe I'll be able to learn another language.") I hope that makes some sort of sense.