What about Pavlovian conditioning? It seems likely to me that dogs, or any animal, displaying examples of shame and guilt would be caused by interactions with humans and resulting domestication / conditioning. If that behavior is observed in undomesticated animals in their natural environment that might better support the innate/hard-wired theory. However, naturalistic observation is subjective, so how to be sure that the emotions and concepts of shame and guilt aren't being projected by the observer onto the behavior?