September 26, 2020, 03:39:17 PM

Author Topic: Big data study on narrative structures  (Read 262 times)

Offline Magnus Hedén

  • High Lord of commas and Grand Master of semicolons
  • Ambient Mage
  • ****
  • Posts: 440
  • Total likes: 352
  • Gender: Male
  • My name is Magnus. I make stuff up.
Big data study on narrative structures
« on: August 11, 2020, 08:46:16 AM »
Some researchers have been using big data to analyse stories and have found underlying narrative structures that permeate storytelling through time and culture. It's been a long time since I had any doubt it was so due to a lot of supporting finds in neuropsychology in the past few decades. But his study has a different approach, and while it is by no means the end of that research -- it's more of a door opened towards further studies -- it's nonetheless very interesting.

Here's a short article that summarises it:

https://www.psychologytoday.com/intl/blog/the-athletes-way/202008/the-narrative-arc-what-big-data-tells-us-about-storytelling

Here's a youtube video by one of the researchers: https://www.youtube.com/watch?v=Cs8OTGJR0V0

Here's the study itself (from what I can tell, it's the whole thing, but I haven't read it yet -- though I certainly will): https://advances.sciencemag.org/content/6/32/eaba2196

[edit] And I found their official homepage as well, though haven't had time to look at it yet: https://www.arcofnarrative.com/ [edit 2](Oh, you can actually put in your own texts into their algorithm there... interesting!)
« Last Edit: August 11, 2020, 08:53:03 AM by Magnus Hedén »

Offline Magnus Hedén

  • High Lord of commas and Grand Master of semicolons
  • Ambient Mage
  • ****
  • Posts: 440
  • Total likes: 352
  • Gender: Male
  • My name is Magnus. I make stuff up.
Re: Big data study on narrative structures
« Reply #1 on: August 11, 2020, 09:06:12 AM »
I put in my published story (found here) into their algorithm and not surprisingly it went against the norm in all three aspects. I was well aware of its non-standard structure, but it's interesting to see it quantified.



I'd say their method will need a fair bit of further refining (and they say so themselves) as I don't quite agree with the cognitive tension graph. The plot progression one is dead on, though. The plot doesn't progress enough in the middle, a problem I ended up not being able to solve without a complete rewrite, for which there was never any time.

To be clear: the researchers don't claim that the norm is the only way to write a good story, just that it's the most common structure.
« Last Edit: August 11, 2020, 09:22:18 AM by Magnus Hedén »