[Retros] Actualization of Retro-active states

[Guus Rol]

Dear retro friends,

The most important aspect of the retro-active field - from the viewpoint of
composer and solver - is the manifestation or actualization of retro-active
states. At such a time an uncenrtainty regarding a retro-active state
converts to a certainty. Example, the uncertainty of having "no castling
rights" converts to the certainty of "having no castling rights". If you
read the Codex plus its explications you would get the idea there is only
one way in which a retro-active condition becomes an actual state in play
and that is by executing the corresponding action move, i.e. castling or en
passant. On closer inspection this appears quite a long way from the truth
on actualizations.

The Codex has a small vocubulary to support its view which centers around
"permissions". The term "permission" is highly suitable to describe the
basic conventions, like the "permission to castle" or "no permission to
play e.p.", or even "no permission to draw on 3R unless proven". In
their native forms, these "permissions" or "no permissions" take on a
permanent and untouchable status in relation to certain isolated events on
a chess board. The Codex has attempted to extend that idea of permanence in
"permissions" to the "Retro Logics" but such hides the true nature of
retro-active states once they become entangled. And so the Codex reads on
mutually exclusive castling in RS (retro-strategy) "*whichever castling is
executed first is deemed to be permissible". *So I ask you, what was,
according to the Codex, the *state* of the "castling variable" before the
castling move was executed? Its authors would probably answer: such doesn't
matter a bit until the question of the castling action is raised. That is
like saying: the law on gravity doen't matter until you drop a stone in a
gravity field! There is and always was an enormous emphasis in the Codex on
the *visible actions* and little attention for the invisible *states*,
either "rights" or "no rights" alike. Not zero attention since even with
the current Codex some state analysis is required for pRA variants.

As I have argued in a previous chapter, the true nature of retro-active
states is that they are variable, uncertain or even "superpositioned". The
latter term is taken from quantum mechanics where several possible states
are often considered to exist simultaneously. The state we experience only
"actualizes" once measured and may be destroyed when measured directly. And
therefore rather than going by the Codex term *"permissions"* which
radiates an aura of permanence, I introduced the term *"license"* as an
uncertain right revokable at any time of measurement. The license is a
"state descriptor" which says (1) that a state is uncertain (2) what will
happen to the state when one attempts to measure it directly. Example.
where there is a license to castle then the state of the "castling
variable" can go either way but will for an instance change to "castling
right" when one attempts to castle.

The focus on states and low visibility brings many more options for
actualization in view than existing in the current Codex, even in or near
Orthodox chess. Some of my favorite actualizations come from AP-logic.
Rather than going with the move-oriented approach of the Codex - e.p. move
justified by castling move - one can go one level up to the state oriented
approach - e.p. actualization justified by castling actualization. Did you
know one can construct this even in orthodox chess without ever playing an
e.p. move or playing a castling move? Or you can mix them up in 4 different
combinations of moves and indirect state changes. That result is only to be
expected. Would not the law of gravity be more powerful than the action of
a stone falling through it? Examples will be given when AP-logic is treated
later.

In the common RS field there are also pretty obvious examples of state
actualizations with a low profile. Andrew at one time stated that one need
not care about the state of an e.p. move since it could not be executed
anyway (unless 100% certain). However in a reflex mate problem, not playing
e.p. on the first move (which would have mated) results in a certainty of
the state of "no e.p. right". The state actualization of "no e.p. right"
may lead in turn to a further state actualization of an entangled "no
castling right" which affects the remaining part of the solution. Another
example is "temporary e.p. rights elevation" (better term than "rights
promotion" which I used previously). If denying all e.p. rights in a
diagram would result in "no proof game", then at least one e.p. move must
be allowed. Actually, for lack of preference rules, all possible e.p. moves
must be permitted at that point. An example on castling actualization
without castling move:is by twice repeating a position with King and Rook
on home squares, e.g. Ke1-e2,x,Ke1,-x,Ke2,x,Ke1,-x. Even though a castling
right is destroyed, this sequence also proves a former state of "castling
right". By  inference another castling right may have been lost with impact
on the solution.

These are not all the examples in orthodox chess and certainly not in fairy
chess. The intent of this message is to indicate that there is a more
generic way of strategically employing state actualizations than suggested
by the "move permission" angle of the Codex. This will pave the way for
understanding the Retro-logics in the fairy domain which could not never be
captured by superficial case (by case) laws.

Best wishes, Guus Rol.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://one.pairlist.net/pipermail/retros/attachments/20140708/54311db0/attachment.html>