[docs] Definition for .O. given in docs is different from Karttunen (1998)’s definition

[bluebear94]

Assuming that the operator documentation is referring to this paper for lenient composition, it gives [X .o. Y] .P. Y as an equivalent expression while Karttunen (1998) gives [X .o. Y] .P. X. The definition in FOMA itself suggests [X .o. Y] .P. X; if so, the documentation should be corrected.