My proposed research focuses on how language is used on Twitter disinformation accounts to authenticate identity performances. I will focus specifically on the use of metadata (profile information such as usernames), affiliative hashtag use, and other language features associated with certain social groups (e.g. vernacular forms).