New open access paper of ours published today by It’s the first large scale comparison of Twitter lexical dialect maps and traditional survey data