Way back when, in ancient Greek, "meta" just meant "alongside" or "next to" or "after". Then Aristotle wrote a book called "Physics" (though its subject was br

Way back when, in ancient Greek, "meta" just meant "alongside" or "next to" or "after". Then Aristotle wrote a book called "Physics" (though its subject was broader than what we now call physics) and another book whose title has been lost, about more abstruse things. It was typically shelved next to the "Physics" and listed after it in catalogues, and it was alleged that Aristotle thought its contents should be taught later than those of the "Physics", and it therefore came to be known as "Metaphysics": the thing that comes after the "Physics". Because the stuff in the "Metaphysics" was felt to be deeper and more advanced than the stuff in the "Physics", "meta-" came to mean something like "beyond" -- with indeed a connotation of "X about X", but by no means always. "Metamathematics" is mathematics about mathematics, but "metapolitics" is theorizing about politics, not doing politics about politics. (The above may sound ridiculous, but it really is where this usage of "meta-" comes from.) I agree that it's weird to take "metaprogramming" to mean "discussion of things like office furniture, big monitors and text editors", and that "metaprogramming" already has a meaning and that that isn't it. But the problem isn't that "meta" can only mean "X about X"; that simply isn't true.