“…For example, this study of dialect classification contains inner-circle (Australia, Canada, United Kingdom, Ireland, New Zealand, United States), outer-circle (India, Malaysia, Nigeria, Philippines, Pakistan, South Africa), and expanding-circle (Switzerland, Portugual) varieties together in a single model. The problem is that these more recent approaches, while they consider more varieties of English, have arbitrarily limited the scope of variation by focusing on a relatively small number of features (Grafmiller & Szmrecsanyi, 2018;Kruger & van Rooy, 2018;Schilk & Schaub, 2016;Collins, 2012). In practical terms, such work uses a smaller range of syntactic representations than comparable work in authorship analysis (c.f., Grieve, 2007;Hirst & Feiguina, 2007;Argamon & Koppel, 2013).…”