Fun with Grammarly

As you may recall, I’ve been using Grammarly on occasion to check over my musings. I even went so far as to purchase the Premium version, only to find that I disagreed with some of its advice on repeated words and punctuation [1].

One interesting aspect of having a Grammarly account is that they send me a weekly report. Let’s look at the data from this week’s report.

1229709 words checked. You wrote more words than 99% of Grammarly users did.

Yes, I’m sure that I had actually written over one million words, I would have written more words than 99% of Grammarly users. But I’m pretty sure that I wrote significantly fewer words. It’s just that with the letter to the incoming chair, it feels like I wrote over one million words.

1687 unique words used. You have a larger vocabulary than 95% of Grammarly users.

Given that the first piece of data was wrong, should I trust this one? Let’s see … what did it say the prior weeks? 1985 unique words last week. 1606 about a month ago when I was using it more regularly. So, yes, that seems reasonable [2].

Does the number of words I use when writing indicate the size of my vocabulary? I don’t think so. There are words in my vocabulary, like ennui and defenestrate, that I rarely, if ever, use in these musings [3]. But perhaps they’ve done careful studies about the ratio of words used in Grammarly-reviewed writing to actual vocabulary size [4]. It’s a bit sad that 1700 or so words is a larger range than most of Grammarly writers use [5]. I wonder what that says about their user community?

Dear Grammarly Developers: If you really want to provide data that I would use and perhaps even pay for, you might consider counting the number of snarky remarks I make.

[1] I don’t disagree with the claim that some variety is good. And I don’t disagree with the claim that in the US, punctuation belongs inside quotation marks. I do, however, disagree with the text and examples that they used to support those assertions.

[2] It is a bit depressing to think that I used fewer than two-thousand unique words in writing over a million words. Maybe Grammarly is right about repetition.

[3] This may be the first time I’ve used ennui. Does anyone remember when I last used defenestrate?

[4] No, I don’t think so.

[5] Grammarly says that I should write 1700 or so words are a larger group than …. That feels wrong to me. Why? There’s an implicit a group of 1700 or so words, which makes the subject singular. I suppose it’s unreasonable to expect their AI to make such fine distinctions.

Version 1.0 released 2017-06-05.

Version 1.0.1 of 2019-03-12.