suspicion: modeling pain through the lens of pain-reduction is a good way to end up with a model of a lasting conversation between pain and relief
suspicion: modeling pain alongside the specific nature of the pain-experiencer's felt desires *around* that pain is a good way to let two user-generated forces (so to speak) cancel each other out? and just .. recede from experience?
this itself is a diagnostic; if a conversation *just won't end*, where is it self-propagating? or auto-leveling? and, before you undo it, observe that you've got a metastable surface, and those tend to *find* metastability because they're loadbearing for something else. if you check your orthogonals, what do you see?