In Kerstin Uvnas-Moberg’s new book on oxytocin, she talks about the role of the hormone within human and human-animal relationships and its importance for bonding, health and in therapeutic treatments.
“For a relationship to occur, you need in some way to get closer to the other individual and to do this you need to not be afraid of that person.“
“During closeness in various kinds of relationships, the body’s reward system is activated and stress and tension are decreased. After some time, these reactions are automatically linked together with the other person and that makes it sufficient for their mere presence to trigger positive reactions.”
( Uvnas-Moberg: The Hormone of closeness: The role of oxytocin in relationships  – paraphrased slightly for ease of reading, emphasis mine)
This applies not just to human relationships but also to our relationships with dogs and cats. When you understand this, the futility of using training or behaviour modification techniques based on fear or pain are obvious. You cannot build a relationship with a dog if there is fear or mistrust (on either side).
What does this have to do with behaviour modification?
“Behaviour modification involves encouraging certain behaviours in a safe, non-threatening context, using desensitisation and counter conditioning and replacing the rules that encourage the dog to react with new rules that allow the dog to relax and take cues from the environment”
“For this to happen there must be clear signalling and learned trust from both parties and reliability from the humans“
(paraphrased from Karen Overall, Manual of Clinical Behavior, p 66).
When you use fear or pain to train, you may get an animal to perform a behaviour, but can you build or sustain a relationship with that animal?
When you can get both through kind, reward-based training, then why do anything else?