"Historical Sociolinguistics" revisited: Drawing further evidence from digital editions of historical correspondence