I had heard a while ago about how Google was going to start live-checking the grammar in Google Docs, using the power of machine learning to identify potential errors as a person typed. I set up a document for testing purposes, to see if I could trigger the blue squiggles that were supposed to show up, marking grammar problems, but nothing happened and I forgot about it.
Then, yesterday, I was writing a blog post and the blue squiggles started appearing. Here was one of them:
And here, after I used Google to search for the information about how to translate the blue squiggles in Google Docs into grammar suggestions (I was supposed to “right click” on them), was the suggested correction:
What could be more conducive to writing than having a computer trying to grab your attention to nag you about your grammar? Being nagged about your grammar by a computer that doesn’t understand grammar. It kept going:
Correction:
I’ve long since gotten used to the idea of Google being wrong. But it was a new experience to have Google actively chasing me down to try to enforce wrongness on my own writing.
Maybe it had a stylebook question about hyphenated adjectives? Nope.
The annoyance at the constant interruption gave way to a fleeting, giddy intimation of semantic collapse. What if it was just that my brain couldn’t understand what “this the now-familiar process” was supposed to mean? Maybe my own writing and reading consisted of nothing more than a badly stitched-together series of guesses about how words and ideas fit together. If Google says gibberish is language, then language will have to be gibberish.
