Chatbots and Academic Writing?

Astar the Robot could put their arm back on, but can they provide effective feedback on writing?

Today I came across this article on a university teaching assistant who was actually a computer. “Ms. Watson” managed to fool students and fellow TAs alike, as “she” quickly and accurately answered the barrage of questions on the online forum for a computer science course in Georgia.

Many of the questions this robot, and most of us teaching in a university setting, respond to are requests for information that is easily accessible elsewhere–course syllabus, handouts, lecture notes, textbook, etc. What if we extended this information-gathering function that AI is so good at to the regulative rules of English grammar, spelling and structure? Could a robot like this be useful in an EAP or academic writing class?

There already “robots” of sorts being used to correct writing, such as the ETS Criterion Online Writing Evaluation Service, which provides “immediate, detailed feedback on grammar, spelling, mechanics, usage, and organization and development” so that instructors can “can concentrate on the content and style of students’ work and teach higher level writing skills”. I’d love to hear from anyone who uses this or a similar service at their institution. I’ve only ever been exposed to this program in the context of TOEFL practice exams. I wonder, does this division of labour between grammar/mechanics and style/content actually work out this way? I also wonder about the format and nature of the feedback given by this software; there’s certainly no shortage of debate in the academic literature on the issue of written corrective feedback with far from across-the-board consensus on the most effective treatment of errors. .

My students regularly use software like Grammarly and Ginger Grammar, but in some cases, without the guidance of a teacher or someone more proficient in and knowledgeable about the English language, they often have difficulties in correctly applying the suggestions made or stumble on the gap between grammar and style. Chatbots have also been used as a tool for TESOL, mostly for writing practice to improve fluency, but they have their shortcomings as well.


One thought on “Chatbots and Academic Writing?

  1. Great conundrum you’ve raised, Jennifer, thank you for sharing this.

    I don’t use any of these programmes, but my research is touching on issues of agency and morality, and your post has made me think … Would the Springer hoaxes, generated by SCIgen and much mentioned back in 2014 or the (in)famous Sokal hoax, be detected as hoaxes by the various bot programs you refer to?

    I approach academic writing as a social activity, one that requires ‘talk around’ text (eg Lillis and Curry) and a respect for the writing agent, so I could never see myself relying on a computer algorithm to determine the quality and the meaning of an academic text. Even ‘errors’ in grammar or language could be deliberate manipulations to create meaning (Derrida would fail such a bot test, surely!) or could be spectres actually be hidden meanings that need to be teased out by the combined agencies of teacher and student?

    Mary Scott’s book, which I reviewed for JEAP, explains the idea of seeing errors as ‘ghosts’, as traces or echoes of meaning that need a discussion, not a correction. Academic writing should always be a trigger for such discussion, and not a green light for correction, in my view.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s