5.4.2 How to account for learning from actions with negative outcomes (avoidance)

We start by assuming that we are only dealing with 'raw' information, i.e. as it comes from perception and as it is required for bodily action.

We need to account for how action might be avoided in order not to experience pain.

The remembered feeling of pain presumably has associated feelings such as fear.

There are several ways not just to 'freeze', i.e. to stop acting, but to select an appropriate action sequence.

• We search through the memory to find a sequence where the feeling matches the pain associated with the fear.
• We then go further back in memory to see whether this action sequence had an earlier perception matching our current perception
• If not so, we can splice in (play) this action sequence newly found in memory.
• If so, we continue searching memory.

• We search backward through the memory to see whether we had a perception matching our current perception
• We then look forward through the memory to see whether the action sequence after this perception resulted in a feeling of pain.
• If not, we can splice in (play) this action sequence newly found in memory.
• If so, we continue searching memory.

The problem is that we want to avoid an outcome. Avoiding the action is most helpful when the action always leads to the negative outcome to be avoided, and when other action-sequences never lead to this negative outcome.