Philip Roth on Politics in Art

From what I consider Roth’s best book, I Married A Communist:

“Politics is the great generalizer,” Leo told me, “and literature the great particularizer, and not only are they in an inverse relationship to each other – they are also in an antagonistic relationship. To politics, literature is decadent, soft, irrelevant, boring, wrongheaded, dull, something that makes no sense and that really oughtn’t to be. Why? Because the particularizing impulse is literature. How can you be a politician and allow the nuance? As an artist the nuance is your task. Your task is not to simplify. Even should you choose to write in the simplest way, a la Hemingway, the task remains to impart the nuance, to elucidate the complication, not to deny the contradiction, but to see where, within the contradiction, lies the tormented human being. To allow the chaos. To let it in. You must let it in. Otherwise you produce propaganda, if not for a political party, a political movement, then stupid propaganda for life itself — for life as it might itself prefer to be publicized.”

Best-of-Three Coin Toss: The Most Likely Way to Win

Always pick whatever side of the coin comes up first, i.e. if heads comes up on the first toss, always choose heads thereafter.

The reasoning: There’s probably some kind of minimal bias one way or another in the coin or in the tosser’s toss. Assuming there’s no data to help you out beforehand for determining the bias, as soon as you get your first piece of data in the form of the first coin toss, chances are the bias is towards whatever side of the coin turns up.

The complication: If the probability distribution is 90% heads and 10% tails and tails happens to come up up first, well, you’re very likely to lose a best-of-three coin toss series by always choosing tails. But if no one actually knows the probability distribution beforehand, you’re better off always choosing the side of the coin that comes up first for a short coin-toss series. Larger coin-toss series, i.e. a best of twenty series, will mean that if the bias is sufficiently skewed, it is indeed better to change your coin-flip choice in light of results.