Reasoning with Agent Preferences in Normative Multi-agent Systems (Extended Abstract)