Let $X:=(X_1,...,X_n)$ be a random vector (we can assume that the $X_i$'s are exchangeable) with joint density $(x_1,...,x_n)\mapsto f_X(x_1,...,x_n)$.
No independence assumption is made. Then, what can we tell about the joint law of the order statistics $(X_{(1)},...,X_{(n)})$ ?