Ten Authors on the ‘Hard’ vs. ‘Soft’ Science Fiction Debate | Tor.com

Ten Authors on the ‘Hard’ vs. ‘Soft’ Science Fiction Debate

Please enjoy this encore from January 2016, in which ten authors discuss the broad categories of science fiction.

In the wake of big-screen success stories like The Martian and Star Wars: The Force Awakens, debates about whether one movie or another is scientific enough have been cropping up in various corners of the internet. Is a deeper, harder line being drawn in the sand about “hard” science fiction than usual? Or are we discovering that perhaps there’s a whole lot more sand available with regards to how imaginative and future-looking fiction can develop, and even entertaining the possibility that these developments could become blueprints for future-fact?

I asked ten science fiction authors about their definitions of “hard” and “soft” science fiction, and how they see science fiction (hard, soft, and otherwise) in today’s terms. They returned with ten fascinating—and not surprisingly, entirely different—answers.

Have a read and then maybe jump in the comments to discuss!


Nancy Kress

The Best of Nancy Kress is available from Subterranean Press

“Hard SF” and “soft SF” are really both misnomers (although useful in their way). Hard SF has several varieties, starting with really hard, which does not deviate in any way from known scientific principles in inventing the future; this is also called by some “mundane SF.” However, even the hardest SF involves some speculation or else it would not be science fiction.

High-viscosity SF takes some guesses about where current science might go IF certain discoveries are made (such as, for instance, identifying exactly which genes control things like intelligence, plus the ability to manipulate them). Or, alternately, it starts with one implausibility but develops everything else realistically from there (as in Andy Weir’s The Martian, with its huge-velocity windstorm on Mars). From there you go along a continuum toward things that, with our current level of knowledge, do not seem possible, such as faster-than-light travel. At some point along that continuum, high-viscosity SF becomes science fantasy, and then fantasy, when magic is involved. But the critical point is that it IS a continuum, and where a given innovation belongs on it is always a matter of dispute. This is good, because otherwise half the panels at SF cons would have nothing to argue about.

I would define “soft SF” as stories in which SF tropes are used as metaphors rather than literals. For example, aliens that don’t differ from us much in what they can breathe, drink, eat, or how their tech functions. They have no delineated alien planet in the story, because they are meant to represent “the other,” not a specific scientifically plausible creature from an exosolar environment. This seems to me a perfectly valid form of science fiction (see my story “People Like Us”), but it is definitely not “hard SF,” no matter how much fanciful handwaving the author does. Nor are clones who are telepathic or evil just because they’re clones (it’s delayed twinning, is all) or nanotech that can create magical effects (as in the dreadful movie Transcendence).


Tade Thompson

Author of the sci-fi novel Rosewater (Apex Books) and the horror novella The Murders of Molly Southborne (forthcoming from Tor.com Publishing)

First, a working definition of SF: fiction that has, at its core, at least one science and/or extrapolation of same to what could be possible.

Second, a (messy) working definition of a science: a field of knowledge that has at its core the scientific method, meaning systematic analyses of observed phenomena including objective observations, hypothesis/null hypothesis, statistical analysis, experimentation, peer review with duplication of findings. I am aware that this definition is a mess.

Defining ‘Hard’ SF is a bit difficult. If we use the Millerian definition (scientific or technical accuracy and detail), it won’t hold water. The reason is not all sciences are equal in SF. In my experience, fictional works that focus on physics, astronomy, mathematics, engineering and (to a lesser extent) chemistry tend to be filed as ‘Hard,’ especially if there is an exploratory or militaristic aspect. The further the extrapolation of the science from what is known, the more likely the story will be classed as ‘soft.’ On the other hand, those that Jeff VanderMeer jokingly refers to as ‘squishy’ sciences like botany, mycology, zoology, etc. tend to be classed as soft SF along with the social sciences like anthropology, psychology, etc. Medicine can fall either way, depending on the actual narrative.

That the definitions are problematic becomes obvious immediately. I find the terms intellectually uninteresting because they assume that social sciences use less rigor, which I know to be untrue. My background is in medicine and anthropology, and I have seen both sides.

There may be other elements to the definitions. There may be a pejorative flavor to being designated ‘soft’. There may be some gender bias, although I have seen this in discussions, and not in print. Take a lot of the work of Ursula Le Guin. Many would not class her SF as ‘Hard’ despite her clear understanding of anthropology and psychology. The exploration of cultures should not take a back seat to the exploration of the solar system. Take Frankenstein, which is often regarded as the first science fiction novel. Few would regard it as Hard SF, yet it used contemporary scientific beliefs. At the time the novel was set, galvanism was a big thing. Reanimation was not thought to be impossible. The Royal Humane Society in England started with reanimation of the dead at its core, and its motto is a small spark may perhaps lie hid.

At the root of the Hard-Soft divide is a kind of “I scienced more than you” attitude, which is unnecessary. There are fans of all flavours of SF and the last thing we need is to focus on divisions that were introduced in the late 1950s.


Elizabeth Bear

Author of Karen Memory (Tor)

I feel like the purported hard/soft SF divide is one of those false dichotomies that humans love so much—like white/black, male/female, and so forth. The thing is, it’s really arbitrary. I write everything from fairy tales to fairly crunchy sciency SF, and I think the habit of shoving all of this stuff into increasingly tiny boxes that really amount to marketing categories is kind of a waste of time. There’s no intrinsic moral element that makes a rigorously extrapolated near-future cascading disaster story (like The Martian) “better” than an equally critically hailed and popular sociological extrapolation. Is anybody going to argue, for example, that 1984 or The Handmaid’s Tale aren’t worthy books because they are about societies in crisis rather than technology?

I love hard—or rigorously extrapolated physical—science fiction, for what it’s worth. My list of favorite books includes Peter Watts, Tricia Sullivan, and Robert L. Forward. But it’s not new, and it’s not dying out. It’s always been a percentage of the field (though Analog still has the biggest readership of any English-language SF magazine, I believe) and it’s still a vibrant presence in our midst, given writers like Kim Stanley Robinson and James L. Cambias, for example. It’s hard to write, and hard to write well, mind, and Andy Weir kind of knocked it out of the park.

My own pocket definition of SF is that it’s the literature of testing concepts to destruction: space travel, societies, ideologies. At its best, that’s what science fiction does that most other literary forms do not. (Most of them—the ones with a literary bent, at least–are about testing people (in the form of people-shaped objects called “characters”) to destruction. Science fiction does it on a scale up to and including entire galaxies, which is kind of cool. Drawing little boxes around one bit of it and saying, “This is the real thing here,” is both basically pointless and basically a kind of classism. It’s the Apollonian/Dionysian divide again, just like the obsession of certain aspects of SF with separating the mind from the meat.

(Spoiler: you can’t: you are your mind, and your mind is a bunch of physical and chemical and electrical processes in some meat. You might be able to SIMULATE some of those processes elsewhere, but it seems to me entirely unlikely that anybody will ever “upload a person,” excepting the unlikely proposition that we somehow find an actual soul somewhere and figure out how to stick it in a soul bottle for later use.)

Anyway, I kind of think it’s a boring and contrived argument, is what I’m saying here.


Max Gladstone

Author of the Craft Sequence Novels (Tor)

Hard SF is, in theory, SF where the math works. Of course, our knowledge of the universe is limited, so hard SF ends up being “SF where the math works, according to our current understanding of math,” or even “according to the author’s understanding of math,” and often ends up feeling weirdly dated over time. In very early SF you see a lot of “sub-ether” devices, from back when we still thought there might be a luminiferous ether; more recent SF that depends on a “Big Crunch” singularity collapse end of the universe seems very unlikely these days, since observations suggest the universe’s expansion is accelerating. Often you find stories in which the orbital dynamics are exactly right, but everyone’s using computers the size of a house, because of course 33rd century computers will still be made with vacuum tubes, or stories that have decent rocketry but a lousy understanding of genetics, or stories that get both rocketry and genetics right, but don’t have a clue how human societies or beings function.

I don’t think there’s a dichotomy, really. “Hardness” is a graph where the X axis starts at zero, and that’s, say, Star Wars—SF that doesn’t even mention math or orbital dynamics, but is still recognizably SF—and proceeds to, say, Apollo 13, which is so hard it’s not even fiction. On the y axis you have “quality.” You can place every SF text somewhere within that space, but no curve exists. Neal Stephenson’s Cryptonomicon is SF so hard that it borders on a technothriller, but that hardness doesn’t determine its quality when set against, say, The Left Hand of Darkness (where the plot hinges on FTL comms), or Childhood’s End (force fields, psychic storm omega point gestalts, etc.).

But if we really want something to pose against “hard,” how about “sharp SF”? Sharp SF acknowledges that our understanding of the universe is a moving target, and believes the point of SF is to show how human beings, relationships, and societies transform or endure under different conditions. Sharp SF takes math, physics, sociology, economics, political science, anthropology, psychology, etc. into account when posing its hypothetical worlds—but cares more about the human consequences of those hypotheticals than it cares about the hypothetical’s underlying architecture. I’d include 1984, The Handmaid’s Tale, Parable of the Sower, Nova, Dune, and Lord of Light as canonical examples of good sharp SF.


Aliette de Bodard

Author of The House of Shattered Wings and its forthcoming sequel The House of Binding Thorns (Roc/Gollancz)

I think they’re labels, and as labels they’re useful because they tell you what kind of story you’re going to get, and what it’s going to focus on (in the case of hard SF, hard sciences such as maths, physics, computer science, and an emphasis on the nitty-gritty of science and engineering as core to the plot. Soft SF is going to focus more on sociology, societies and the interaction between characters). The issue with labels is twofold: first, they can be used dismissively, i.e., “it’s not real SF if it’s not hard SF,” or “hard SF is the best kind of SF and everything else is of little worth,” which is unfortunately something I see happening all too often. And it’s doubly problematic, because this dismissal is disproportionately used to single out women/POCs/marginalised people as not writing “proper SF.” (I should add that I’ve got nothing whatsoever against hard SF and will quite happily enjoy an Alastair Reynolds or a Hannu Rajaniemi when I’m in the mood for it).

The second issue is that like any labels, they can be restrictive: they can create an impression in the author’s mind that “real SF” should have such and such; and particularly the emphasis on the nitty-gritty of science makes a lot of people feel like they shouldn’t be writing hard SF, that you should have several PhDs and degrees and everyday practice of physics, etc., to even consider writing something. It’s not that it doesn’t help (as someone with a degree in science, I can certainly attest that it helps make things go down more smoothly with only minimal amounts of research), but I worry that it raises a barrier to entry that doesn’t really have a reason to be there. My personal testimony is that I held off from writing SF because I didn’t think I had the chops for it (and that’s in spite of the actual maths/computer science degree…); and also that it took me a long time to write what I actually wanted to write because I was afraid that taking bits and pieces from every subgenre I liked was somehow an unspeakable crime…


Walter Jon Williams

Author of Impersonations (Tor.com Publishing)

I’d define Hard SF as a subdivision of Geek Fiction. I’m currently at work on a General Theory of Geek Fiction, and while my ideas are still in flux, I can define Geek Fiction as that fiction in which the greatest emphasis is given to process. The story becomes not one of plot or character or setting—although ideally those are present as well—but a story in which the action is broken down into a series of technical problems to be solved.

Thus The Martian is a book about all the technical problems that need to be surmounted in order to survive on Mars. C.S. Forester’s Hornblower books are about the technical issues involved in commanding square-rigged sailing ships in wartime. Police procedurals are about the process of police procedure. These sorts of books can be about other things as well, but if the emphasis isn’t on process, it’s not Geek Fiction.

As for Soft SF, it’s better to define it by what it is instead of by what it isn’t. After all, Soft SF includes space opera, science fantasy, dystopia, near-future works, alternative history, time travel stories, satirical and comic SF, and great big unclassifiable tours-de-force like Dhalgren. Just call the thing what it is.


Ellen Klages

Winner (with co-author Andy Duncan) of the 2014 World Fantasy Award for the novella “Wakulla Springs” (Tor.com)

Attempting to differentiate hard and soft science fiction implies that “science” has gradations on some sort of undefined, Mohs-like scale. Talc science vs. diamond science. But that seems to me a misunderstanding of what science is. Science is not an established body of knowledge as much as it is an attempt to explain things that we don’t yet know, and to organize what we do know in a systematic way. It is the manual that the world ought to have come with, but was somehow left out of the box.

Things We Don’t Know is a rather large category to begin with, and is also quite fluid, because everything we do know is continually shifting and changing—our understanding of reality is a work in progress. When most people say “this is hard science fiction” they mean the plot depends on demonstrable, provable, known facts about the physical world. Hard, like concrete, not fluid and mutable like water.

I sometimes think they also mean it in the same sense as when Mac users were looked down on by PC users 30 years ago: if you didn’t know how to program your computer, you didn’t really know how to use one. If it’s not hard (as in difficult to do or to understand), it has less value.

Historically, hard science fiction has been more about how inanimate objects work than how human beings live. More about plot than about character. Go figure. Humans—or at the very least, biological beings—are part of any world, and there’s so, so much we don’t know about them. So studying what makes humans tick—the sciences of sociology, economics, linguistics, psychology, etc.—must surely be as much part of that missing world manual as physics and chemistry. A person is more complex than any machine I can think of, and when we start aggregating into groups and societies and nations, that complexity grows exponentially.

I prefer my science fiction to be well-rounded, exploring and explaining the people as well as the furniture and the landscape.


Maurice Broaddus

Author of the short story “Super Duper Fly” (Apex Magazine) and the forthcoming novella Buffalo Soldier (Tor.com Publishing)

The thing is, my background is as a hard science guy. I have a B.S. in biology and I can still remember the grumbling during our graduation when those who received degrees in psychology were introduced as fellow graduates of the School of Science. Ironically, even after a 20-year career in environmental toxicology, the science of my SF writing tends to lean to the “soft” side of things.

There is an imagined line in the sand that doesn’t need to be there. In fact, hard and soft SF go hand-in-hand. Much of the SF I’m drawn to turns on the soft science of sociology. The impact of technology in a culture’s development, how people organize, and how people interact with the technology and each other because of it. (Think of how prescient 1984 seems now.) And for all of the hard science of The Martian, it would all be science porn if we also didn’t have the soft science of psychology in play also. A story is ultimately driven by the psychology of its characters.


Linda Nagata

Author of The Red: First Light (Saga Press), a Publishers Weekly best book of 2015

My definition of hard SF is pretty simple and inclusive. It’s science fiction that extrapolates future technologies while trying to adhere to rules of known or plausible science. “Plausible,” of course, being a squishy term and subject to opinion. For me, the science and technology, while interesting in itself, is the background. The story comes from the way that technology affects the lives of the characters.

I don’t use the term “soft science fiction.” It’s one of those words whose meaning is hard to pin down, and likely to change with circumstances. Instead I think about science fiction as a continuum between hard science fiction and space fantasy, with no clear dividing line—although when you’ve wandered well into one or the other, you know it. And besides, just because we’ve split out the hard stuff, that doesn’t mean that everything that’s left can be dumped into the same “not hard” category. So there is science fiction, and within it there is hard science fiction, planetary stories, retro science fiction, space opera, military science fiction, and a lot more—but I don’t have an all-encompassing term for the non-hard stuff.


Michael Swanwick

Author of Chasing the Phoenix (Tor)

I go with what Algis Budrys said, that hard science fiction is not a subgenre but a flavor, and that that flavor is toughness. It doesn’t matter how good your science is, if you don’t understand this you’ll never get street cred for your hard SF story. You not only have to have a problem, but your main character must strive to solve it in the right way—with determination, a touch of stoicism, and the consciousness that the universe is not on his or her side. You can throw in a little speech about the universe wanting to kill your protagonist, if you like, but only Larry Niven has been able to pull that off and make the reader like it.

Fran Wilde’s work includes the novels Updraft (Tor, 2015) and Cloudbound (Tor, 2016). Her short stories appear in Asimov’s, Tor.com, and Nature. Her novella “The Jewel and Her Lapidary,” will be published by Tor.com publishing in May 2016. She writes for publications including The Washington Post, SFSignal, Tor.com, Clarkesworld, iO9.com, and GeekMom.com. You can find her on twitter @fran_wilde, and at franwilde.net.


Back to the top of the page


This post is closed for comments.

Our Privacy Notice has been updated to explain how we use cookies, which you accept by continuing to use this website. To withdraw your consent, see Your Choices.