The Retrograde Analysis Corner

Mutually Exclusive Castlings

Glossary

Due to changes in the Codex of Chess Composition 2009 this article is outdated; but on the other hand it may still be valid for problems prior to these changes. See also Castling and En-passant capture.

The general convention is that, in a given problem, any side has the right to castle provided it is not possible to prove it has lost this castling right. (See "Castling" in the glossary.)

Now it is sometimes possible to prove that some castlings right are mutually exclusive. M. Havel was the first to notice and use this, in 1922.

Here is an example:

Armand Lapierre
Thèmes 64, Apr. 1959
4th Hon. Mention

[4k2r/p1p1p1p1/1p2P3/8/3R4/6P1/5P1P/R3K3]

7+7. Mate in 2

Here Black and White can't be both allowed to castle. Indeed, consider the Rook on d4. If it is the original QR, then obviously White can't castle anymore. If it is the original KR, then the wh. King had to let it out of the SE corner, and he can't castle anymore.

Another (real) possibility is that the Rd4 is a promoted R. Then it must have left the 8th rank through square d8, or f8, or h8. In any case the bK (or the bKR) must have moved and B can't castle anymore.

In conclusion: it can be proved that the two castlings are mutually exclusive, but none may be proven impossible in itself.

Now, who can castle?

In such a case, the general default convention is that whoever castles first is allowed. Once this is done, the other castling becomes forbidden because, then, it can be proven impossible).

Lapierre's problem is a very pedagogical illustration. The try 1. Rad1?, threatening 2. Rd8 mate fails on 1 ... O-O! The solution is 1. O-O-O! and now 2. Rd8 mate can't be avoided because 1 ... O-O? is illegal.

Sometimes another convention, Retro-Variants, is assumed. In this case, it is required to mention "(RV)" under the diagram.