Writing advice as a science

by look i have opinions

I’m convinced that most writing advice is bullshit. Part of the trouble is that I suspect that writing well could be reduced to a science—not an exact science, but a sound one nonetheless. Here is an experiment that, as far as I know, has never been done.

Gather some aspiring writers for a sort of seminar. Have each writer submit for evaluation what they consider to be their most accomplished work.

Divide your writers at random into multiple groups. The control group will simply write, without guidelines or feedback. Other groups will get the benefit of general guidelines (a different guide for each group) and/or individual feedback (a different set of readers for each group).

At the end of the experiment, the writers will once again submit for evaluation what they consider to be their most accomplished work. Each “before” and “after” writing sample will be reviewed blind by a panel of professionals who are comfortable with the genre and style, and assigned quality ratings. The writers will complete a survey on the experience.

We can all agree that good guidelines and feedback will (on average) cause improvement in writing quality, as determined by ratings. (Yes, the ratings will be highly subjective, but on average they should be good enough for our purposes.) Here are some additional hypotheses:

  • The most significant factor in improvement will be the writer’s initial writing quality and experience level. Beginners will improve the most, even in the control group.
  • The next most significant factors of improvement, even in the control group, will be (in no order) individual talent, motivation, time spent writing, and number of words written.
  • Bad guidelines/feedback will actually cause some groups to get worse. The worst guidelines/feedback will be bad because they’re confusing or genre-inappropriate.
  • The quality of individual feedback will have more effect than the quality of general guidelines, because guidelines tend to be harder to apply and easier to ignore.
  • The guidelines/feedback do not need to be reflected in the final writing sample in order to have caused improvement or worsening. It is possible for a writer to be inspired by a piece of advice, or rendered painfully self-conscious by it, without following it.

Regardless of whether any of those hypotheses are correct, it would be tremendously valuable to have solid evidence to work with, instead of anecdata and pseudoscience. Then we could confidently answer questions like these:

  • Which guidelines are the best?
  • What type of feedback is the best?
  • For the average beginning writer, which approach causes the most improvement on sentence-level writing quality: memorizing nearly mindless grammatical rules, studying actual grammar, reading one’s own work aloud during revision, reading others’ work aloud, or writing prolifically? (This isn’t meant as a rhetorical question. I can see potential value in each.)
  • For the average beginning novelist, what’s the best way to get better at novel writing: working on short stories first, following an outline, writing as many chapters as possible, or working on the first chapter first? (Again, not rhetorical.)
  • How often should the average writer get individual feedback? Presumably the benefits of good feedback reach a plateau or even reverse at some point, but when?
  • Is it helpful to get feedback from readers who aren’t familiar with the genre? If so, what should be the proportion of familiar readers to unfamiliar ones?