The Medical Lessons of Science Fiction

What does science fiction have to say about the institution and practitioners of medicine? One might think that ever since 1818, when Mary Shelley's Victor Frankenstein reanimated charnel waste, science fiction has been especially drawn to that branch of science that most intimately holds human bodies, lives, and deaths in its hands. But, in fact, as an institution, science fiction has been unexpectedly reluctant to deal with matters medical, and when it has, it has consistently displayed ambivalence and distrust, views worth examining if one is to understand how society at large conceives of medicine and how medicine must adapt if it is to enjoy the support of society at large.

Since January 1998, the University of Michigan Genre Evolution Project (GEP) has been testing the hypothesis that cultural creations evolve in the same way as do biological organisms--that is, as complex adaptive systems that succeed or fail according to their fitness to their environment and, by their existence and success, modify their environment. As our initial test subject, we chose the twentieth-century American science-fiction short story. While our approach to this subject may seem thoroughly plausible to those trained in biology, it stands against many of the assumptions and methods of traditional literary criticism, based as they typically are on the close analysis of a limited number of unusual works (for example, so-called masterpieces) or a focus on the works of one or a handful of especially valued authors (for example, Keats or Romantic poets). As part of the effort to test our hypothesis, we developed a series of fields for coding stories. Stories are read by teams of student researchers who post their common codings to a shared database. These researchers are assigned readings of whole issues of science-fiction magazines within the period from the first such magazine, April 1926, through the end of 1999. The assignments are made with the aim of generating a representative sample of science fiction. It is already agreed, as Paul A. Carter has noted, that science-fiction short stories are reflective of the field as a whole. ¹

As of 18 December 2000, our database contained records for 1,836 unique stories of which many were reprinted for a total of 3,233 records. Each record contains a field for "dominant science"--that is, the science that dominates the story. Not every science-fiction story has a dominant science, of course, and, as is inevitably the case in large-scale research, not every record is complete. Filtering out records for which dominant science is either "Not applicable" or blank leaves a dataset of 1,094 unique stories.

Over the course of our first year, using a process we call Dialectical Database Design, we converged on a list of eighteen values for the dominant-science field that seem adequate for coding all science fiction and that are sufficiently distinguishable from each other that they can be used comfortably in making forced coding decisions. Those values are anthropology, astronomy, biology, chemistry, computer science, ecology, economics, engineering, geology, history, linguistics, mathematics, medicine, pedagogy, physics, political science, psychology, and sociology. If the stories were randomly distributed among these values, one would expect ~5.56% of the total to be coded for each science, or, for the 1,094 unique stories that were assigned a value, approximately 61 stories. Given how often one hears reference to the "art" rather than to the "science" of teaching, one might expect that pedagogy would rarely be treated as a science, and that is so (6 records among the 1,836 unique titles, or ~0.33%). Given the advertisements offering technical training in many science-fiction magazines, one might expect that physics would often be the dominant science, and that is so (171 records among the 1,836 unique titles, or ~9.31%). Given that medicine is so powerful a field emotionally, as Shelley's Frankenstein makes clear, one might well expect that medicine would often be...