@James
Don’t know if you are still looking for an answer, but…
Those are anchors, something we NLP practitioners (and Speed Seductioners) love to play with. To remove one, anchor it to its opposite and it will at some point get cancelled out.
For example, if there is a song that played during my dad’s funeral and now always makes me sad (UB40 comes to mind), I should deliberately play that song and then start focusing on things that make me happy while it is playing. That either removes or re-purposes the anchor, depending on how often I do it.
It could be that certain modules can help with un-linking an anchor. Rebirth comes to mind, since we’re essentially re-framing the anchor. But it will in all likelihood require conscious effort in addition to the sub.
For the animal rights activists among you, removing the anchor was part of Pavlov’s experiment. In that regard, Pavlov was quite kind, although possibly overfeeding the dog in a time before obesity was “hip”. Seligman’s learned helplessness experiment on dogs however, is one of the most cruel experiments I have ever heard of (and variations are often used on human beings by terrorists, human traffickers, “bad” cults and intelligence agencies). Learned helplessness is one of the most “evil” yet effective ways to control people.
And for those wondering… yes, if you get the average training manual for dogs and replace all instances of “dog” with “human” it is highly likely to work equally well, if not better. Just don’t tell your subject what you’re doing, just like you wouldn’t tell the dog. Ignorance is bliss. 
Maybe we should write a sub as if it is a dog training sub, but with a human target instead. 
Anyways…