When I did my BA in creative writing in RMIT we discussed famous writers’ writing speeds. I believe it was to help allay any fears for those that felt they were writing too slowly. In any case, this information was a surprise to me because one writer said she only wrote about 500 words a day. I can’t remember who that was (and I think it was quite a few writers that said that) but that idea was so wildly different to my 2,000-3,000 words a day that it had the opposite effect. i. e. What am I doing wrong(ly)?
After writing, editing and publishing 2 books of about 40,000 words each in 8 weeks for the popular genre of space opera, and looking back over the million or so words I’ve already published in the space fiction category alone, I couldn’t imagine writing just 500 words a day.
Impossible, I thought.
Two years later, and I can’t write more than 500 words a day! Academic writing is research heavy and, unless conjecturing, must be absolutely correct as all resources and references used to make the point will be checked by peer reviewers.
This came into stark relief recently when I tried to edit together all the research I’d done for chapter 3 – the chapter that would a) show that there hasn’t been much progress in creating an Automatic Popular English Fiction Genre Classification (APEFGC) program since work began around 1994, and b) how I had created a simpler way to find 19th and early 20th century stories in popular fiction genres categories in large text databases, with the creation of my keycloud idea. For the chapter, I read hundreds of scholarly articles on automatic genre identification and related works (in mostly the data science disciplines but also in digital humanities), hoping to find some compelling results that I could build on. What I found was that those who did investigate this area, mainly focused on non-fiction. The thing is, I decided that all these automatic genre identification papers had failed as they hadn’t created a useful fiction genre classification system. I am grateful to my supervisor for carefully and kindly pointing out to me that the point I was making about the data I’d spent months collecting and a few weeks writing about amounted to a straw man’s argument. i. e. Their papers weren’t focused on creating an APEFGC, only testing programs to see if certain genres could be detected with them. I had been so disheartened by hundreds and hundreds of papers not giving the results I wanted when actually that wasn’t really the focus of most of those papers. My argument had no real foundation.
This is why we have supervisors when writing a PhD thesis. We get too close to our writing and can’t see the all encompassing error in the direction we’re heading. Thankfully, this was my first draft and a certain amount of ‘composting’ is expected. As it turns out, instead of those 5000 words describing the ‘failures’ of all these papers, I can simply say something like “Of the 150 papers I looked at, most of them did not focus on finding fiction genres, with many of the papers either saying or implying that the programs used were not successful in distinguising between genre fiction classifications.” Or something like that. This sentence effectively reduced my chapter to about 500 words.
Hmm, I guess I’m now a 500 words a month writer. Oh well. Still, the writing plan for chapter 3 can still work. I just need to find papers where people 100% focused on creating something akin to an APEFGC and were relatively successful. I’ve since found a few so I’ll rewrite chapter 3 from that perspective.
Til next month.
