I'm skeptical of "explainable A.I." in many cases and I use the curse words as an example. You really don't want to tease out the thought process that got there, you just want the behavior to stop.