In 1994, critical thinking within the SF field belatedly caught up with reality. Credit for this goes to David Hartwell and Cathryn Cramer, whose analysis in the anthology The Ascent of Wonder finally acknowledged what should have been obvious all along. Hard SF is the vital heart of the field, the radiant core from which ideas and prototype worlds diffuse outwards to be appropriated by writers of lesser world-building skill but perhaps greater stylistic and literary sophistication. While there are other modes of SF that have their place, they remain essentially derivations of or reactions against hard SF, and cannot even be properly understood without reference to its tropes, conventions, and imagery.
Furthermore, Gregory Benford’s essay in The Ascent of Wonder on the meaning of SF offered a characterization of the genre which may well prove final. He located the core of SF in the experience of “sense of wonder”, not merely as a thalamic thrill but as the affirmation that the universe has a knowable order that is discoverable through reason and science.
I think I can go further than Hartwell or Cramer or Benford in defining the relationship between hard SF and the rest of the field. To do this, I need to introduce the concept linguist George Lakoff calls “radial category”, one that is not defined by any one logical predicate, but by a central prototype and a set of permissible or customary variations. As a simple example, in English the category “fruit” does not correspond to any uniformity of structure that a botanist could recognize. Rather, the category has a prototype “apple”, and things are recognized as fruits to the extent that they are either (a) like an apple, or (b) like something that has already been sorted into the “like an apple” category.
Radial categories have central members (“apple”, “pear”, “orange”) whose membership is certain, and peripheral members (“coconut”, “avocado”) whose membership is tenuous. Membership is graded by the distance from the central prototype — roughly, the number of traits that have to mutate to get one from being like the prototype to like the instance in question. Some traits are important and tend to be conserved across the entire radial category (strong flavor including sweetness) while some are only weakly bound (color).
In most radial categories, it is possible to point out members that are counterexamples to any single intensional (“logical”) definition, but traits that are common to the core prototypes nevertheless tend to be strongly bound. Thus, “coconut” is a counterexample to the strongly-bound trait that fruits have soft skins, but it is sorted as “fruit” because (like the prototype members) it has an easily-chewable interior with a sweet flavor.
SF is a radial category in which the prototypes are certain classics of hard SF. This is true whether you are mapping individual works by affinity or subgenres like space opera, technology-of-magic story, eutopian/dystopian extrapolation, etc. So in discussing the traits of SF as a whole, the relevant question is not “which traits are universal” but “which traits are strongly bound” — or, almost equivalently, “what are the shared traits of the core (hard-SF) prototypes”.
The strong binding between hard SF and libertarian politics continues to be a fact of life in the field. It it is telling that the only form of politically-inspired award presented annually at the World Science Fiction Convention is the Libertarian Futurist Society’s “Prometheus”. There is no socialist, liberal, moderate, conservative or fascist equivalent of the class of libertarian SF writers including L. Neil Smith, F. Paul Wilson, Brad Linaweaver, or J. Neil Schulman; their books, even when they are shrill and indifferently-written political tracts, actually sell — and sell astonishingly well — to SF fans.
Eric S. Raymond, “Libertarianism and the Hard SF Renaissance”, Armed and Dangerous, 2002-11-09.