Emotional Shape of Stories Emerges in Math
Andrew J. Reagan, an applied mathematician at the University of Vermont, thought about all the extraordinary new the knowledge about genes that has been generated by data from the Human Genome Project. And that made him wonder what data could teach us about stories.
Dr. Reagan and his team analyzed more than 1,300 works of fiction in the digitized Project Gutenberg collection. They traced the “emotional arc” of the stories by graphing the happiness and sadness of the words as they appeared in the text. The emotion arc of a story doesn’t provide direct information about plot or characters, they explain in a recently published paper. It reflects changes in sentiment as the story progresses. They discovered that about 85 percent of the stories they analyzed—across different cultures and time periods—fall into one of six emotional patterns. The team notes that Kurt Vonnegut observed years ago that stories have the shapes, and the shape of the Cinderella story was similar to the Biblical description of the origins of Christianity as well as the creation story of nearly every human society.
The six emotional trajectories illustrated in the different story arcs are:
Rags to riches – Rise
Tragedy, or riches to rags – Fall
Man in a hole – Fall, rise
Icarus – Rise, fall
Cinderella – Rise, fall, rise
Oedipus – Fall, rise, fall
The researchers believe these arcs are the building locks of stories. They created mathematical graphs, depicting the shapes of the six patterns as exemplified in classic and popular literature based on their computational analysis. The graphs are shown in a Scientific American story by Mark Fischetti. This story also describes a study by Poland’s Institute of Nuclear Physics showing that the lengths of sentences in books often form a fractal pattern—one in which shapes are replicated in small and large scale, similar to tree branches. Scientific American asked the Vermont team to analyze two of the books in the fractal study, Fischetti writes, and the mathematicians found those books did have two of the common emotional arcs. Do books with the same emotional arcs have the same fractal patterns? No one knows, though further investigation might answer that. Read more about the research in Reagan’s University of Vermont Story Lab blog.
Why analyze the mathematics of literature? Dr. Reagan and his colleagues write that we are driven to find and tell stories to communicate our ideas, needs and beliefs and describe our observations of the world.
“In science, we formalize the ideas that best describe our experience with principles such as Occam'sRazor: The simplest story is the one we should trust,” the authors write. “We tend to prefer stories that fit into the molds which are familiar, and reject narratives that do not align with our experience.” Advances in computing, natural language processing and digitization of text are providing quantities of data that can be mined for new insights on the evolution of culture and communication.