The job of the science fiction writer, like that of writers of any other sort of fiction, is telling stories, that is, inventing characters and placing them in dramatic opposition to one another. The special task of the SF writer, though, is to supply not only the drama but also the stage: to build the entire set upon which one’s characters act out their conflicting purposes. And so we must create not merely characters and plots but entire worlds.

 That sounds like a god-sized assignment, and in a sense it is. Of course, our worlds are merely things set down on paper, and that spares us the considerable trouble of producing actual mountains and seas, skies and deserts, and all the other tangibilities, down to microbes and algae, that real gods must traffic in. We deal in the illusion of creating worlds, not in the worlds themselves. Even so, the job has to be done right or the illusion won’t hold.

I can name any number of examples of the job done right: Hal Clement’s Mesk-lin, Frank Herbert’s Arrakis, Brian Aldiss’ Helliconia, James Blish’s Lithia, Ursula K. Le Guin’s Gethen, Anne McCaffrey’s Pern, Harry Harrison’s Pyrrus, Larry Niven’s Ringworld, Stanislaw Lem’s Solaris, and on and on, an infinity of fictional worlds having little in common except plausibility and unforgettability. But how is it done? What factors must you consider, what knowledge must you have?

The place to start, I think, must be the physical characteristics of the world to be created: its size and mass, the nature of its sun and the imaginary world’s distance from it, its gravitational pull, its period of revolution and axial rotation, its orbital tilt, the makeup of its atmosphere, its biochemistry, and so forth. From these things all else inevitably follows.

Two of our greatest practitioners of this area of the art of creating imaginary worlds were Poul Anderson and Hal Clement. These two grand masters had the playfully speculative cast of mind that any science fiction writer must have, but also the benefit of scientific educations that were both deep and broad, particularly in the areas of physics and chemistry, and their work demonstrated a degree of accomplishment in those aspects of world-building that most of us can only hope to approach. Both are now gone from us, but it’s our great fortune that each of them left valuable essays on their working methods: Anderson in “The Creation of Imaginary Worlds” and Clement in “The Creation of Imaginary Beings,” both published in Science Fiction, Today and Tomorrow (1974), edited by Reginald Bretnor. I urge anyone interested in knowing how worlds are built to seek out these magisterial texts. The Bretnor book is long out of print but readily available via second-hand channels. Most of its fifteen chapters are still relevant to modern SF writers, but two, the Anderson and the Clement, remain essential reading today.

Anderson focuses quickly on the importance of consistency of scientific logic: writers who arbitrarily cobble together worlds built out of incompatible factors risk forfeiting, at least to the scientifically informed reader, their plausibility in all other sectors of their stories, those involving such things as character, emotional texture, plot. Giving the example of a planet that circles a blue-white sun and has an atmosphere of hydrogen and fluorine, Anderson says, “This is simply a chemical impossibility. Those two substances, under the impetus of that radiation, would unite promptly and explosively.” Blue-white stars are too hot to be surrounded by inhabited planets of any sort: they burn so fiercely that they don’t last long enough for planets to develop around them, let alone for life to evolve on those planets. He deals similarly with red giants, white dwarfs, variable stars, and other sorts of stellar bodies unsuitable for the creation of habitable worlds, pointing out the problems inherent in choosing such familiar stars as Sirius, Vega, Antares, or Mira as settings for stories. Those stars are familiar to us because they shine brightly in our own sky, but for half a dozen different reasons it is not the most luminous stars that will be found to have inhabited worlds. By running through the stations of the stellar temperature-luminosity chart, Anderson explains how to select (or invent) a star that would be likely to provide worlds useful to the storyteller. He does point out that any sufficiently ingenious writer can make use of any kind of star for story purposes: the thing to avoid is unknowingly to place your fictional world around a star that can’t possibly have planets, or to postulate intelligent life on a world where no life of any sort could exist.

With these limitations having been invoked, Poul goes on to remind us that a planet’s distance from its sun affects its climate, that its mass and size and density determine its gravitational pull, that the presence or absence of moons will control its tides, that the degree of axial tilt will shape its seasonal variations. “By bringing in this detail and that, tightly linked,” Anderson says, “the writer makes his imaginary globe seem real. Furthermore, the details are interesting in their own right. . . . They may reveal something of the possibilities in these light-years that surround us, thereby awakening the much-desired sense of wonder.”

He doesn’t insist that the conception of a fictional world must be preceded by months of preliminary study, but only that a reasonable understanding of the laws of astrophysics will allow writers to convince their readers that they actually know what they’re talking about, and will also help in the process of inventing the story itself: “Whatever value the writer chooses [for a planet’s axial tilt], let him ponder how it will determine the course of the year, the size and character of climatic zones, the development of life and civilizations. If Earth did travel upright, thus having no seasons, we would probably never see migratory birds across the sky. One suspects there would be no clear cycle of the birth and death of vegetation either. Then what form would agriculture have taken? Society? Religion?” He illustrates the Andersonian planning methods with diagrams and mathematical calculation, but there’s nothing there, intimidating though it may look at a quick glance, that a would-be science fiction writer with at least a high school degree can’t follow. (And if you aren’t willing to think through a little bit of high school astronomy, what are you doing setting up shop as a science fiction writer?)

Though the Anderson essay is, in and of itself, a splendid little handbook for planetary creation, he recommends two classic reference books that remain invaluable to science fiction writers to this day: Intelligent Life in the Universe (1966) by I.S. Shklovskii and Carl Sagan and Habitable Planets for Man (1970) by Stephen H. Dole. I second the recommendations—those two have stood me in good stead for decades—and add to it Cycles of Fire (1987) by William K. Hartmann, Red Giants and White Dwarfs (1967) by Robert Jastrow, and The Planetary System (1988) by David Morrison and Tobias Owen, which, while dealing entirely with our own solar system, provides a wealth of fundamental information about why that solar system has the form it does, so that any writer can readily generalize new worlds from the data supplied.

The Clement essay that follows Poul Anderson’s in the Bretnor book likewise stresses the merit of using rigorous logic, or at least common sense, in populating one’s invented worlds with living creatures. One should know something about how earthly creatures work, Clement says, before dreaming up extraterrestrial ones, since the basic biological rules of our planet very likely will hold true for any planet that is capable of bringing forth life. (“The trick of magnifying a normal creature to menacing size is all too common. The giant amoeba is a familiar example; monster insects, or whole populations of them, even more so. It might pay an author with this particular urge to ask himself why we don’t actually have such creatures around. There is likely to be a good reason, and if he doesn’t know it perhaps he should do some research.”) He goes on to explain why Pegasus wouldn’t be able to fly, why six-foot-long ants don’t infest our gardens, and why birds don’t travel at supersonic speeds; and, having done that, he then proceeds to demonstrate how, given a deep enough knowledge of planet-building, a writer can conjure up worlds and creatures capable of operating in contravention to all our own planet’s rules. (The indispensable Shklovskii-Sagan volume, Intelligent Life in the Universe, has a particularly valuable chapter on the possibilities for life of a non-terrestrial sort in other solar systems.)

But, as I’ve said, working out the physical and biochemical characteristics of your invented world is just the starting point. Those characteristics will provide you with the setting for your story, but not the story itself, though in large measure they will govern the essential nature of that story. (The concerns of the thirsty desert-dwellers of Frank Herbert’s parched Arrakis in Dune are quite different, for example, from those of the inhabitants of the completely aquatic Hydros in my own The Face of the Waters). Once you’ve determined your world’s gravity, climate, geography, geology, natural history, etc., you need to work out the details of its culture, which will involve you in such matters as economics, politics, religion, and urban planning. Ideally all of this will grow out of the special physical characteristics of your invented world.

I’ll talk next time about how I went about this phase of the process in creating my own best-known imaginary world, the planet Majipoor of Lord Valentine’s Castle and its various companion volumes.