Editing by Algorithm?
In recent years, advertising for Grammarly has blanketed the airwaves, including astronomically expensive Superbowl commercial slots and carefully targeted online ads. The clever marketing leads many people to feel they are remiss in not subjecting their written communications to Grammarly’s automated, yet somehow expert, review. After all, the claims go, Grammarly goes well beyond the standard word processor grammar and spelling checks. Instead of simply flagging glaring errors, Grammarly can also offer style and usage suggestions. Everything from college essays to business emails can be dramatically improved, they maintain. The no-commitment free version helps lure people in, while constantly reminding users of the myriad features their $29.99 monthly fee will unlock if they will only upgrade to Grammarly Premium. At first glance this is all quite enticing.
But is all the buzz borne out by the actual user experience? And more to the point, should academics feel confident that they can replace their editors with the latest incarnation of algorithms and automation? I decided to go ahead and explore the research about Grammarly. It should be noted that online user reviews are generally positive: people appreciate the effortless and automated review Grammarly provides, especially when it is free.
What the Academic Research Community Thinks of the Software
However, academic research into Grammarly’s efficacy is somewhat less enthusiastic. While writing instructors appreciate that it saves them time correcting student errors, their investigations into the use of the tool have exposed its limitations and disadvantages. Svetlana Koltovskaia, for example, found that students with limited English proficiency over-rely on its automated writing feedback. Ruth O’Neill and Alex M.T. Russell report that academic staff find Grammarly helps their students, but that these academic staff remained essential to prevent students from accepting improper “corrections.”
Putting Grammarly to the Test
To further explore this question, I signed up for my own Grammarly account and tried it out for myself. Indeed, I found the need to assess whether Grammarly’s suggestions were valid or not the greatest pitfall for non-expert writers, particularly English language learners. When I uploaded a document to the Grammarly website—there are multiple options for letting Grammarly read one’s work, including document upload, website cut-and-paste, and browser extensions—I received speedy feedback on changes that would improve readability and address some grammatical issues. For example, “My main interest were the buildings” was correctly flagged for improper verb agreement: “The verb were does not seem to agree with the subject. Consider changing the verb form.” Sometimes, though, the feedback resulted in new tangles of words. For the sentence that began, “These faiths, contrarily to Christianism and Judaism,” Grammarly suggested “contrarily” be changed to “contrary.” But this does not really solve the problem; “in contrast to” would be more appropriate, not “contrary to.” And the word “Christianism” should have been flagged as an error—even Google asks me, “Did you mean Christianity?” when I type in that word. Grammarly was silent on the matter.
Suggested Corrections that Introduce New Errors
Grammarly sometimes reminded me of an overzealous peer editor, whose eagerness to demonstrate knowledge results in a completely marked-up essay draft that leaves its author more confused than confident. For example, a sentence beginning, “This change in social structure…” (a perfectly appropriate phrasing) received the unhelpful suggestion to add an article: “The noun phrase social structure seems to be missing a determiner before it. Consider adding an article.” Elsewhere in the article, Grammarly suggested that the specialized anthropological term matricultures be changed to mariculture, a word with a completely different meaning (and also, a singular form instead of the plural original). Flagging phrases and sentences that are correctly written is not only not helpful but can lead to the creation of new mistakes that didn’t exist prior.
Furthermore, Grammarly was a bit lost with bibliographic entries and foreign words, recommending punctuation marks and spellings that would have introduced errors, not corrected them.
How Private is the Research You are Sharing?
A final note: readers should be cautious of online reviews that appear objective, like this review (which is littered with issues of grammar) on the Become a Writer Today website. Grammarly offers powerful incentives for people to become affiliates. They currently pay $20 each time someone signs up for a premium account using an affiliate-provided link. This policy cleverly creates a legion of writing professionals pitching another company’s products, with very little cost on the part of Grammarly.
The Benefits of Finding a Professional Academic Editor
Grammarly certainly has a role to play in the ecosystem of writing improvement. It offers far greater assistance and more detailed feedback than the older generation of grammar checkers within word processing programs. In many cases, it can uncover errors that weaken writing. However, for a thorough review that combines sentence-level correction with more substantive feedback, and that will not leave an author wondering which pieces of feedback are valid and which should be ignored, finding a professional editor with relevant expertise remains the best option—particularly for writers whose English language writing skills are still developing.Receive an individualized quote!