Do Androids Dream of Aesop?

Imagine you’re building a robot (no big stretch, right?), and your robot will need to survive in a human ecosystem. What abilities would you give it, particularly with respect to cognition? You will probably want to load it up with lower order cognitive abilities – enough sense to fully perceive and react to its immediate environment – but what then? When it comes to higher order thinking, would the ability to create and reflect on simulated models of the environment help it survive? If the answer is yes, does that mean you would give it the capacity to create, communicate, and understand basic narratives to ensure its survival? What about stories and literature?

This last question, which artificial intelligence researcher Jerry Hobbs poses in his 1990 paper “Will Robots Ever Have Literature?” is an important one for those who instruct and those who design instruction. As Hobbs says, paraphrasing the Roman dramatist Horace, “we imagine [stories] to instruct and delight ourselves” [emphasis added by me]. The question is: are stories are a pleasant diversion from our daily survival or are they an adaptive function that has historically helped us survive? If stories and fiction are a component of human survival and one of their chief purposes is to instruct, could it be that storytelling is as central to effective instruction as direct modeling, practice, and coaching – not just a pleasing but optional motivational strategy used to reach learners?

I will explore the role of narrative and stories in instruction in the next few blog posts, taking the position that we have developed adaptive brain structures that allow us to create and receive stories as an important means of instructing ourselves. To be fair, Jerry Hobbs concedes that the question of the evolutionary place of fiction and literature is a complex one for AI researchers, and he does not arrive at a firm conclusion on this question. However, in the intervening years since “Will Robots Ever Have Literature?” was published, neurocognitive research has found good evidence that we’re hard-wired to respond to narrative. In particular, a 2008 paper published by researchers at Washington University in St. Louis offers at least one strong piece of evidence in support of this argument, using fMRI imaging to show how brains track a story. Their conclusion: “neural systems track changes in the situation described by a story” and “readers understand a story by simulating events in the story world and updating their simulation when features of that world change.”

So we’re talking about the ability to create instructional simulations at little or no cost using one of the most ancient and established tools known to man. If that description of the power of story doesn’t make instructors sit up and listen, I’m not quite sure what will.