My sister has a food blog and an Instagram related to it that I thought of as I read this article. I don’t use Instagram regularly, or check any blogs other than hers—neither do my parents, who were, when they first looked at the Instagram she set up for them to follow her and like all her pictures, confused by the endless stream of tags at the end. What are these for, they asked when they saw them, I wish I could remember exactly what she said—I think it was something along the lines of those are all the words people could search to find me. I remember her trying to come up with more to add to her list—I was worse than useless, knowing nothing about Instagram food accounts. Her more recent post there is a picture of lemon bars—there are twenty three tags. Gallagher says that metadata “asks students to consider how their readers would find their content and what kinds of steps they need to take to increase their videos circulation” (27). I realize that this is what my sister was doing, even if the phrase “algorithmic audience” was never used. As she photographed what felt like every piece of food before she ate it, and posted frequently to the Instagram, she told us that was important for people noticing. The tags were also important. And when she set up Instagram accounts for both of my parents and for me, that was three automatic, guaranteed likes for every picture, which helps a little, too.
Like Emily said, it feels wrong to me as a writer to come up with content that an algorithm might sort as important, or to “sacrifice elements of personal style” (31) to write highly-ranked content. But it reminded me of things I’ve heard in workshop classes that I already to in my own writing—general rules like there’s little you can do with a 40 or 50 page story that falls between short story and novella, and there’s little you can do with a book that falls between novella and novel in length. There are some limitations, a little like the algorithms, on writing that isn’t geared towards the web. Gallagher’s point on page 32, where he asked his students to generate metadata to help drive up their views, was easier to take than the ideas about driving up style—this seems like a kind of marketing, not unlike how a book might be advertised. Changing syntax and sentence structure to make the page rank more highly seems like a problem.
Do you think this could ever lead to content that works better, or that writers like more? (I think of learning how to write short stories, or how to make better section breaks in a piece, or the advice in novel writing workshop, and wonder if this is possible if the audience is an algorithm and not a style of writing taught by a program.) Or is it just about writing for the existing power structure? Would just changing the marketing/metadata be a better place to start instead, before changing content, and could this be enough?
Word Count 507
Gallagher, John R. “Writing for Algorithmic Audiences.” Computers and Composition, 2017, p. 25-35.