In Chapter 14 of The Copyeditor’s Handbook Amy Einsohn casually presents her readers with what she considers to be a well-known fact regarding grammar check for word processing programs and it’s place in the work of an editor.
If you’re hoping that your word processing program’s grammar checker will save you any time or spare you any errors – forget it. Even for the shortest of texts, these checkers are time consuming and frustrating. They routinely ignore simple errors, repeatedly question unmistakably correct constructions, and suggest substitutions that are flat-out wrong. Most copyeditors simply disable the grammar checker.[i]
Einsohn’s suggestion that a grammar checker was a) useless, and b) easily turned off, gave me pause when reading this chapter, as I admit I was unaware that grammar check came with an off switch. I immediately opened my own word processor (Microsoft Word 2011 for Mac) and highlighted the “Tools” feature and clicked on “Spelling and Grammar.” Sure enough, there, in the bottom left-hand corner of the pop-up window, was a little box labeled “Check Grammar” that could easily be checked on or off with a simple click of the mouse. If turning off grammar check was as easy as Einsohn claimed, then I wondered what else about her statement might have validity. I decided to break Einsohn’s assertion into two separate pieces, and investigate further. Firstly, I plan discover if there is any truth to Einsohn’s claim that grammar checkers ignore errors, question correct sentences, and provide incorrect substitutions, and if this is true, what ramifications might there be for the average user. Secondly, if grammar check does prove to be an imperfect system as Einsohn suggests, should it simply be shut off or does it still have any value as a tool or an aid to editors?
The grammar check FAQ page on the Microsoft Support website seemed to be a logical place to begin my search, but unfortunately the webpage provided little help in finding answers regarding grammar check’s ability to catch errors. I found only the following general disclaimer.
In general, the grammar proofing tool incorrectly marks words or proposes incorrect suggestions when the parser (that is, the grammar proofing component that analyzes the linguistic structure of a sentence) cannot determine the correct structure of the analyzed sentence. Although state-of-the-art in its category, the grammar proofing tool (just like any other commercially available grammar proofing program) is not perfect. Therefore, when you use the grammar proofing tool, you may experience some amount of “false” or “suspect” flagging and subsequent wrong suggestions.[ii]
Note: As I typed the previous passage into Microsoft Word, the program has already found various grammatical errors that it does not like. Ironic that Microsoft’s own program doesn’t like the sentence structure written within its Support Page. Perhaps this is a foreshadowing of the results of my search?
The first sentence in Microsoft’s paragraph is the one that struck me the most when reading. The particular wording seemed to suggest that Microsoft was willing to admit that its program came with errors, but remained vague about exactly those errors were, and if they felt the fault lay truly with the program itself or with the user. Realizing that I wasn’t going to find anything here, I began a new search that eventually led me to a webpage created by Professor Daniel Kies of the College of DuPage, a two-year technical institute located in Glen Ellyn, Illinois. The webpage detailed a ten-year study devoted to determining the effectiveness of different grammar checkers in a variety of difference word processing programs. In his study Professor Kies used 20 different common grammatical errors such as run-on sentences, pronoun agreement, verb tenses, sentence fragments, and many others in order to test the success of various grammar checkers in catching these mistakes. Using this criteria and a point scale of two for every error that a program caught, one if the program occasionally caught an error, and zero if the program never caught the error, Professor Kies discovered the following results. Microsoft Word versions 2000, 2002, 2003, and 2007 each scored a 0 on a point scale of 80, catching none of the hypothetical grammatical errors that were posed to these programs. (I tested Professor Kies’ findings by typing a few of the sentences into my own version of Word 2011 and my results were consistent with his. Since my version is four years newer than the last one Professor Kies tested, this is not particularly encouraging.) Professor Kies did see more success with an older version of Word 97, which scored 19 out of 80, Open Office with Language Tool with 6 out of 80, and Grammarian Pro X, which scored highest at 40 out of 80. Additionally, Professor Kies ran a secondary test to determine that once an error was caught, how the grammar check proposed to fix it. Professor Kies found that each program offered differing ways of “fixing” his introduced errors, but across the board the suggestions were either stylistically too formal or were “flat-out wrong” as Einsohn previously warned. [iii]
Facing proof that Einsohn was correct, and grammar checker not only misses errors, but can also actually introduce them into a text, I wondered how often this fault in the program may have affected the work of the average student, writer, and even the average editor who continues to use a grammar checker, unquestioningly trusting in its so-called “expertise.” In this age of reliance on technology, what harm can come from trusting solely in a program that can only obey the rules set to it by a program manager? And who decides what rules are relevant enough to be included in the grammar check? Are the parameters set by a thorough study of the rules of grammar that are in current usage, or are they an accumulation of the program manager’s own preferences and partialities? At the time that this article is being posted I do have an email query sent to Microsoft’s general inbox to see about finding the answers to these and other questions, but so far I have not yet received a reply. However, I do feel that these issues need to be highlighted and properly addressed. As Professor Kies’ study has demonstrated, grammar check, through its incorrect suggestions designed to “fix” the writing of the user, is thus capable of changing the way a writer communicates with others. If the user is unaware that its suggestions are faulty, and blindly accepts the suggested changes, they may unknowingly be changing the entirety of the message they are attempting to convey to an audience.
So, with all the flaws and faults that grammar check has been proven to posses, one question remains – is there any value in using grammar check at all? While interviewing several of my friends and family – current students, teachers, and a professional editor included – to find out if they had known of grammar check’s off switch, I came across an interesting response from a friend who works as a corporate editor. Instead of shutting his program’s grammar check off completely, he told me that he preferred to use the customization option. Intrigued, I went back to the original pop-up window that I found originally in my Microsoft Word program regarding the options of grammar check. By digging deeper into the different settings I discovered that grammar check could be set up in several different writing styles – Casual, Standard, Formal, Technical and Custom – with each of these further broken down into 26 different highlighted rules such as “Capitalization,” “Punctuation,” or Subject-Verb Agreement” that grammar check would look for depending on which writing style was selected. (Just a FYI, “Standard” is set as the default for most programs and comes with about half the options pre-selected.) Investigating the “Custom” setting, I found that the program defaulted to selecting every option, but each one could be easily turned on or off and saved to any user’s specifications. While Einsohn may be correct that using this customization feature may not save an editor time while “checking the shortest of pieces,”[iv] I would argue that when attempting to edit a 50, 100, or even 300 page manuscript on a short deadline, a second set of “eyes” serving as a back up to the editor to check the consistency of capitalization, comma usage, or hyphenation placement could definitely prove useful. As the tools of publishing increasingly expand to include the technological, I believe it behooves any editor to not simply turn a tool off because it presents frustrations, but to find a way to turn on a more useful feature such as customization to make that tool the most efficient it can be for them.
Perhaps the reason that Einsohn takes such a harsh stance against the use of grammar check is because it violates one of her four commandments of copyediting, “Thou shalt not introduce an error into a text that is correct.”[v] But perhaps Einsohn, like me when I first began this paper, did not fully understand the potential behind grammar check. Perhaps she, like me, failed to look deeper to discover customization rather than deactivation because it was easier that way. And who is to blame for this? Is it Microsoft? Or is it perhaps each of ours, for blindly trusting in a machine to do a job like editing that at its core relies on human nuance and judgment? I suppose in the end it doesn’t matter. What matters is what we do from here. As Joseph M. Williams and Joseph Bizup state in their text, “The alternative to blind obedience is selective observance. [Y]ou have to decide which rules to observe and which to ignore.”[vi] I hope that this blog posting is a start. I hope that reading it has motivated you to go into your own word processing program and begin customizing the grammar check, so you can take charge of what exactly it does to your writing, as I plan to do. I hope that it doesn’t stop there, that you share what you now know about Word with others, as I plan to do. Because in the end I feel that an editor should never stop learning and evolving, just like the very language we so often are in charge of revising.
[i] Amy Einsohn, The Copyeditor’s Handbook: A Guide for Book Publishing and Corporate Communications Third Edition (Berkeley and Los Angeles: University of California Press, 2011), 337.
[iii] Daniel Kies, “Evaluating Grammar Checkers: A Comparative Ten-Year Study.” Last Modified November 28, 2012. http://papyr.com/hypertextbooks/grammar/gramchek.htm
[iv] Einsohn, The Copyeditor’s Handbook, 337.
[v] Einsohn, The Copyeditor’s Handbook, 4.
[vi] Joseph M. Williams and Joseph Bizup, Style: Lessons in Clarity and Grace Eleventh Edition (Chicago: Pearson Education, Inc., 2014), 13.