As an afterthought:
If I recall correctly Norman Wheeler has said something like that Gestalt therapy should not focus only on the figure but also on the ground. Somebody else said that this was absurd because if you focus on the ground this former ground becomes figural and you are again focussing on the figure.
Taking into account what I said in my previous message this contradiction is resolved: If you use the Gestalt psychological terminology for what Norman Wheeler is talking about, you would speak not of figure and ground but of the intentionally accented figure and the intentionally neutral ground. And here it is possible to glide between the accentuation of one out of many figures (forming the intentionally neutral ground) to another one by recentering your attention, interest and so on. All these figures remain figural in the phenomenal sense (the phenomenal ground is something else anyway) but they change position in the sense of being phenomenal accented figure or part of the phenomenal ground.