The Chess Mind

That's the question GM Genna Sosonko raises in the newest New In Chess Yearbook (#90). What does he mean, and what is his conclusion?

He doesn't offer a definition, but from his examples one gets a clear enough picture of what he has in mind. Theorem-chess, we might say, is a way of approaching the problems in a game as a series of formulae to solve: by playing this or that opening (for Black) and exchanging pieces x, y and z, even at the cost of a pawn, I achieve such-and-such an ending (e.g. with opposite-colored bishops), force the White pawns on the right squares, and draw in my sleep. It's an approach I sometimes call "recipe chess", and it's one I've used at certain times, and have seen in some of Kramnik's and Anand's draws with Black (examples: Kramnik's 6th game vs. Topalov in their match and Anand's draw with Gelfand in the second cycle of Mexico City).

There's something very seductive about this approach. Much of theory - in junk openings especially, but in serious openings too - attempts to structure series of moves into wholes - recipes (or formulas, or "theorems") - allowing Black to achieve a technically drawn or White a technically won position. And given many of Sosonko's examples (very good ones, at that) you might think he's advocating it as well. At the end of the day, though, he's not - or at least not without very heavy reservations. Here are the final paragraphs of his essay (p. 26):

I am not without sin in this respect. [DM: Playing lines where Black is worse and has to suffer a bit trying to hold one of these formula draws.] For a very long time I considered the position after 1.d4 d5 2.Nf3 Nf6 3.c4 c5 to be almost equal and I analysed it time and again, trying to keep slightly inferior endgames upright. In my results I alternated a string of draws with the occasional defeat.

I quit playing the line when I read a comment by the great Aaron Nimzowitsch. In a complex position he was trying to find the best move, calculating several lines over and over again until he said to himself: 'Quit! A chess game is a struggle, not a theorem'.

This looks like a repudiation of especially the modern theoretical approach, but Sosonko takes a step back in his final comment.

Although the ever-present computer has made certain corrections to this definition, I strongly believe these words should not be forgotten, even in our days.

I believe this strikes the right balance. Qua chess players, we should know as much as possible, but we must know that there are things we don't (and maybe can't) know, too. Chess is a struggle!