Algorithmic Grammar

Google Docs and I do not have the same understanding of grammar. I know nothing about the skills of whoever programmed the Docs algorithm, but since I learned grammar from my mother, who was a great copy editor, I have a great deal more faith in my opinion than I do in theirs. 

Here is an example of a sentence where Docs wants me to change the words “is more”: 

Why do we believe
the solution to violence
is more violence?

Maybe the reason is simply that I’m using the haiku form to write that sentence, because the blue line under “is more” disappears when I write it without line breaks. 

I should note that the blue lines appear whenever Google Docs finds something “wrong” even though I have “suggest changes” turned off. Apparently you can’t really turn off “suggest changes.”

Here’s another line from one of my daily senryu:

Build your healthy life.

Docs doesn’t like my use of “your.” Since that is a perfectly good imperative sentence, I do not understand it. This one isn’t related to the formatting; Docs marks something wrong either way.

Docs also doesn’t like my use of “doesn’t” as a verb in the first sentence of that last paragraph, since Docs looks like a plural even though I am using it as the name of a program, which makes it singular. Docs is also inconsistent in its application of this rule, because it has no objection to the “is” in this sentence. 

I had thought that capitalization of Docs alerted the algorithm to the fact that I was using it as a proper name (for the very program I’m writing about). When it objected to “doesn’t,” I thought the problem was that it was the first word in a sentence and the algorithm couldn’t tell whether I meant docs in general or Google Docs. 

But the lack of objection to “is” destroys that theory. Now I don’t know what the hell it’s doing. 

Here’s another line where it made a correction.:

Pick the one that makes you happy.

Docs objects to “that.” It would allow “which.” I don’t care about the difference between that and which, but I once had an editor who would have screamed at me for using “which” in that sentence. (Her basic rule, as near as I can remember, was to only use “which” if you could put a comma in front of it. I don’t think that’s the actual rule, but if I did that she didn’t scream at me. She was a decent editor, but a terrible person.)

(Docs also objected to “Docs objects” in the above paragraph. Shrug.)

And a last example:

I wake up disturbed.

I admit to being completely flummoxed by this one. I tried changing wake up to “awake,” but Docs didn’t like that either. 

I’m beginning to think “suggest changes” is  just a random underlining program intended to screw with English majors. That’s probably why you can’t turn it off.

I really hope that no one is using the grammar algorithms of Google Docs or, for that matter, Microsoft Word, to learn English grammar. It’s not just that they will learn to write stilted sentences — I had English teachers in school who tried to teach me to write using similar prescriptive rules, fortunately without a lot of success.

But many of the so-called errors they mark aren’t. That means people who don’t know much about English grammar will assume something that is perfectly acceptable is wrong.

English is a notoriously flexible language. Once you develop a good grasp of the basics of grammar, you can do a lot of things around the edges. Some of those bend or even break the rules, but many are legit, just unusual.  

I don’t like any corrections while I’m writing, so I always try to turn those things off. I do run spell check, because I know I have some spelling blind spots and I also make typos.

But I don’t want to know what some algorithm thinks is good grammar. 


5 thoughts on “Algorithmic Grammar

  1. There was a point in my career where my writing was ruled by one of those “readability” indices–I forget which one. In order to shoehorn a simple sentence into what the index considered an appropriate readability level for my audience (assumed to be 12 year olds) I had to dumb it down to the point of imbecility.

    The bots are simply stupid some times.

    1. In this technological age, it is important for us to figure out when an algorithm is inaccurate so that we don’t rely on it too much. That’s what worries me, that the badly programmed algorithm will dictate language use.

      I know a lot of research has gone into the readability programs to which you refer, but I don’t think they’ve mastered it yet. Also, there are concepts that cannot be expressed in a simple sentence. Some of those concepts can still be understood by 12-year-olds.

  2. What bothered me about the Readability Indices is that they left no room for the flavor of language. Satisfying their limited sensibilities meant reducing everything (Shakespeare, for God’s sake) to the level of how-to instructions. Even a beginning reader should have a little something to aspire to, surely?

    1. Yes to aspiration. I also recall that one of the best parts about reading Shakespeare or, for that matter, Greek comedies, were the notes that explained the jokes. I mean, those things are full of dirty jokes and if you leave the language intact but explain it, the kids will love it. Although I also had the experience of reading Greek comedies in very good modern translations that updated the jokes so that people would get them. (I recommend Douglass Parker’s translations of Aristophanes.) Granted I read them in college, but I don’t think they’d have been beyond the grasp of younger readers. People should productions of Parker’s translation of Lysistrata in high schools.

  3. My sister sent me an example of a grammatical error that does create a problem: a dangling participle:
    “ As the matriarch of the Jack Kerouac School at Naropa, we often turn to Anne for inspiration.”
    The phrase “As the matriarch …” modifies “we,” though it is intended to refer to “Anne.”This error can be very confusing in some sentences.
    Google Docs does not mark this error.

Leave a Reply

Your email address will not be published. Required fields are marked *