Is there room for reference borrowing in Donnellan’s historical explanation theory?