Reuniting 'Is' and 'Ought' in Empirical Legal Scholarship