There is in fact another paradox that exists within the rules of Diplomacy, which is also known as the Convoy Paradox (or also as Pandin's Paradox). It looks something like this:
Fleet Wales - English Channel
Fleet London supports Wales - English Channel
Army Brest - London
Fleet English Channel convoys Brest - London
Army Yorkshire supports Brest - London
Fleet Belgium - English Channel
Fleet North Sea supports Belgium - English Channel
The big question is - is the army convoyed? At first glance, the English and German fleets bounce, and the French fleet can still convoy thanks to the beleaguered garrison rule. However this would mean that the French attack on Wales succeeds, cutting support to the move on the English Channel. Thus the German fleet (with one support) can move in unopposed, dislodging the French fleet and preventing the convoy. The English fleet in Wales can still support London, causing a bounce, which means the army gets convoyed...and so on.
This paradox was reviewed by Simon Szykman and Manus Hand in the Winter 1999 Retreats publication of The Diplomatic Pouch's e-zine1, and it was decided that one of two rules had to be introduced in order to solve this problem:
"If a convoyed army attacks a fleet that is supporting an action in or into a body of water that contains a convoying fleet, that support is not cut by the convoyed army under any circumstance (but a convoying army does cut all other supports normally)."
This basically means that the French army does not cut support, the English fleet in London still supports Wales, and the English and Germans butt heads in the Channel with no resolution. However the attack does dislodge the English fleet in London, which is disbanded as it has nowhere to retreat to.
The big argument against this is that it breaks one of the sacred rules of Diplomacy - that support offered by a dislodged unit is always cut. The big argument for this is that this is the only rule to be broken - once we've resolved this issue, we just apply all rules as normal.
"If a situation arises in which an army's convoy order results in a paradoxical adjudication, the turn is adjudicated as if the convoying army had been ordered to hold."
This basically means that the French army doesn't get convoyed, the French fleet holds in the English Channel, the English fleet in London continues to support as normal, and nothing much changes.
The argument for this ruling is that if it takes effect, a destroyed unit does not offer support as happens in the first ruling. However, the ruling still breaks one of the rules of Diplomacy by assuming that the army will refuse to convoy despite being given orders to do so.
Which is correct?
Opinion is divided on which ruling should be used, but the chances of the paradox coming up in a game are extremely small. Manus' solution is currently in effect on the diplom.org servers, but Simon's ruling can be chosen by using the game option SAFE_CONVOYS.
1 - http://www.diplom.org/Zine/F1999R/Debate/paradox.html