[Retros] Article 16 and Ceriani's example

[Rol, Guus]

[see in the Retro Corner: "Partial Retrograde Analysis" and "Retro
Strategy" in the modified Codex by Werner Keym, Meisenheim]

Werner Keym presents one of Ceriani's compositions as example 4 in his
clarification to the new article 16 in the Codex. He makes the point
that it is self-evident that this problem should be understood as a PRA
problem rather than an RS problem since the PRA solution includes and
exceeds the RS solution. I analyzed this problem a few years ago and
concluded that neither the RS nor the PRA interpretation is
satisfactory.


Looking at Ceriani from a PRA standpoint we can readily observe that the
on/off state of the white castling right 0-0 plays no relevant role in
controlling the "PRA parts". The move 0-0 only serves to avoid the
knight check in the variation where such is required. This becomes clear
when you change wPc2 to c4. In spite of the fact that the white and
black castlings are no longer mutually exclusive, the 2 partial
solutions remain precisely the same. Instead of changing the Pawn
position one might also place wKe1 on e2. Eliminating the white castling
right altogether while simultaneously disabling the potential check
still delivers the same partial solutions but this time Rf1 replaces 0-0
in one solution part.


In RS, the opposite is true. Both suggested modifications to the PRA
setting would make the RS solution fail. The e.p. state, so important in
the PRA version, cannot be capitalized on (by convention) and therefore
disappears from the picture. Removing bNa1 and moving wPb3 to a2 (and
possibly a few other changes) creates a position which illustrates that
the white castling choice is primarily motivated by the need to apply
retro-strategy. The check escape is only the duplicate argument here.


The comparison of the RS and RV approach tells us that the PRA solution
only seemingly overlaps the RS solution. As seen from the previous
analysis, the vital argument in the RS solution - mutual exclusivity of
white and black castling rights - has zero significance in the PRA
solution. The content of the white castling variants in the RS and the
PRA approach can therefore not be considered "the same" and it becomes
dubious to justify that Ceriani's problem is "obviously" PRA and not RS.
Which in turn indicates that Cerinani's example may not be the most
appropriate elucidation to article 16, clause(3).


Another way to look at it is by conceiving Ceriani's creation as neither
RS nor PRA. Possibly a twin of both logics, but not a very successful
one considering the duplicate arguments and one identical key move.
There are other ways to classify problems like these, and Valery
Liskovets has made an earlier attempt at that. In due time I will give
my own view but such is outside the current scope of evaluating the
match between article 16 and Ceriani's 1960 composition.


Guus Rol.