Google has announced the launch of a new tool aimed at detecting when students submit work that is not their own, thrusting itself into long-running debates over how to root out plagiarism while also protecting students' privacy and teaching them how to responsibly cite others' work.
"We realized we had a real opportunity to approach this problem differently," Zach Yeskel, a group program manager for Google for Education, said in an interview.
"Our goal is not to detect plagiarism. It's to detect similarity and originality."
For years, other companies have offered services aimed at the same problems. The biggest, Turnitin, now says that it is used by more than 15,000 education institutions in 150 countries, including many U.S. high schools. The general idea is to leverage the power of big data and digital technology to automatically review student work, scanning to see if anything was lifted from another paper, scholarly work, or other source without proper citation.
Google is pinning its hopes of capturing a chunk of the K-12 market for such services on a few factors. The new feature, which the company has dubbed "originality reports," will be part of its uber-popular Classroom learning management system. The originality reports will also leverage the unrivaled power and ubiquity of Google's core service, Search. And company officials said Google will not be building and maintaining a "global repository" of student work-the approach used by Turnitin, which has prompted lawsuits and privacy complaints.
What do experts on writing instruction, media literacy, and education technology think?
Most hope Google's new tool might help foster much-needed dialogue between teachers and students about citation, academic writing, and the sometimes-fuzzy lines between one's own ideas and the ideas of others.
But they fear the opposite will happen.
"If people just rely on Google's algorithm, conversations about what it means to make creative use of other people's work will be [replaced] with a superficial understanding of 'originality,'" said Renee Hobbs, a media-literacy expert and professor of communication studies at the University of Rhode Island.
"We need to recognize the great limitations of a tool like this."
Like any tool, the biggest question is how it will actually be used.
Yeskel described Google's vision for originality reports as being a way of "teeing up teaching moments."
Once the reporting feature moves out of testing, he said, teachers who use Classroom will be able to enable the new feature when they give students writing assignments. Before turning their work in, students will be able to use the feature to check their own writing up to three times, to see if it's been flagged for unoriginal content. Once students submit their work, the Google service will automatically generate a new originality report for teachers to use as they grade.
"If a student clearly copied somebody else's work from the internet and doesn't cite it properly, that's the kind of stuff we'll catch," Yeskel said.
It's not hard to imagine how that process could present an opportunity for teachers and students to have a discussion about academic writing and proper citation. But it's not at all clear how Google's originality reporting tool might actually facilitate such discussions, or improve their quality. Yeskel said additional features that might help, such as a feedback tool, are still in development.
Such limitations highlight a broader reality, said Danielle Nicole Devoss, a writing professor at Michigan State University.
"Although they claim their purpose is to help, the majority of these services are 'detect-and-punish,'" Devoss said.
Other experts said that dynamic reflects deeper problems.
Ed-tech historian and critic Audrey Watters, for example, said plagiarism-detection software in general frames all writers as potential cheaters, undermining the trust that is essential to strong student-teacher relationships. She said the companies making the software tend to accept as given that most writing assignments are so cookie-cutter that students can reasonably consider copying someone else's work a viable strategy.
When algorithms are elevated into the role of a primary audience for both students and teachers, Watters added, it inevitably leads humans to conform to machines' logic and rules.
"We should be thinking about how to have pedagogies of care, not of surveillance," she said. "I'm not sure we can ask Google to be a part of that."
And what about the reality that much of student writing is no longer solely text-based?
Narrowly casting "cutting and pasting" as a problem misses the reality of many modern classrooms, where student work regularly includes a mix of words, images, video clips, memes, and other multimedia-much of which is recycled and repurposed from other sources.
In the media-literacy world, the line between originality and plagiarism when creating such work revolves around the legal concept of "transformative use," said Hobbs, of the University of Rhode Island. Are students merely re-transmitting someone else's work? Or are they adding value to the existing work, perhaps by using it for a new purpose or audience?
That's not something an algorithm can detect, she said.
Yeskel from Google agreed.
"Ideas come from everywhere, and a big part of the education process is about synthesizing and growing them," he said. "There's a lot of nuance there, and that's why there still always needs to be an instructor involved in the process."