Introduction to the Special Issue on Automated Writing Evaluation
Jim Ranalli, Volker Hegelheimer · Language learning & technology · 2022
Technologies that allow the automated evaluation of machine-readable text—a process which has become known as Automated Writing Evaluation (AWE)—have been around since the 1960s, and for almost as long, there has been interest in the ways such technologies can support the work of second-language (L2) learners, teachers, and practitioners in the field of computer-assisted language learning (CALL). The mainframe computers and punch cards used in those early forays into AWE have long given way to much more accessible, affordable, and user-friendly systems, such that most learners writing today on computers with internet access will be able to avail themselves of some form of AWE. The speed of this development means users’ understanding of these technologies and their broader implications for L2 writing and learning has always lagged behind. While it has only been six years since our involvement in the production of a previous special issue on AWE (Hegelheimer et al., 2016), the abundance of more recent research seemed to require an updated consideration of the field. In addition, the global COVID-19 pandemic temporarily shifted much L2 language and writing instruction online, necessitating more dependence on written forms of communication and thus pushing AWE presumably even further into L2 writers’ common experience. These were the motivations for this special issue on AWE. Given the need for an updated perspective on AWE as described above, we were surprised by the rather lackluster response to our call for papers. We received 38 abstracts, and of these, we selected 13 from which to invite submissions of a full article. Following peer review, we ended up with the four empirical articles presented here, all of which were based on research conducted in East Asia in tertiary-level Chinese L1 contexts in which the language of the focal AWE system was English. Despite our appreciation for the fine work of our contributors, we could not help but be somewhat disappointed at the lack of range of research that this special issue represents, given the emphasis in our original call for papers for work based in nontertiary contexts and AWE tools addressing languages other than English. On the positive side, we were pleased to see a variety of AWE systems represented, including those that are commercially produced (e.g., Pigai) versus those developed by the study’s authors (CyWrite and its engineering abstracts module). There were also a variety of target features for the AWE systems in question, such as grammatical forms (Write & Improve with Cambridge), content/argumentation (Virtual Writing Tutor), and genre conventions (again, CyWrite and its specialist engineering-discourse module). There is no doubt that the pandemic, which began a few short months before our call for papers went out, has had an inhibiting effect on the conduct and dissemination of research.