필사 모드: Science Fiction Does Not Predict the Future — Reading SF as Thought Experiment

Opening — What Neuromancer Got Right and What It Never Saw Coming

William Gibson's Neuromancer came out in 1984. Consciousness jacked straight into the network, corporations more powerful than states, a data space where intrusion and defense play out. Gibson had used the word cyberspace first in a 1982 short story, and this novel carried it to the world. Judged on hit rate alone, it is uncanny.

And yet there are no mobile phones anywhere in the book. The characters still run through the streets hunting for pay phones. The actual internet grew not as a three-dimensional space but as text and advertising on a flat screen, and what came to rule it was not the novel's hackers but ad agencies and platforms.

Even so, Neuromancer has not aged a day. The value of this novel never lay in its hit rate to begin with. What SF does is not guess the future but push a world with one condition changed all the way to the end, so that we see ourselves again. This piece lays out how that works and where it came from. No twist or ending of any work is revealed.

Le Guin's Declaration — SF Is Not Predictive, It Is Descriptive

Ursula K. Le Guin settled the matter outright in the introduction she wrote for The Left Hand of Darkness (1969). Science fiction does not predict; it describes. Her English runs, "Science fiction is not predictive; it is descriptive." Short, and it changed the genre's understanding of itself wholesale.

What Le Guin warned against was the writer playing prophet. A prophecy is either right or wrong. If it is right the author becomes a fortune teller, and if it is wrong the novel is thrown out. Either way the work turns into a scorecard with time. Description, by contrast, aims not at the future but at the here and now. You set up one unfamiliar condition and write down in detail what becomes of people inside it.

What The Left Hand of Darkness actually does makes this plain. The novel depicts Gethen, a planet whose people have no fixed sex. They spend most of their time without gender and become male or female only in cycles. This is not a forecast that humanity will someday evolve that way. It is an experiment to see, if you subtract the single condition of gender, what parts of a society collapse along with it and what parts stay standing. Le Guin herself called the work a thought experiment.

So the question you should bring to good SF changes too. Not "will this really happen" but "what becomes of a person under this condition." The first question turns a novel into a product with an expiry date; the second turns a forty-year-old novel into a problem of today.

Hard and Soft — An Overrated Ruler

The first axis anyone reaches for when sorting SF is hard versus soft. Hard SF respects the known constraints of physics, chemistry, and engineering. Whether the orbital math checks out, whether the communication lag is honest, is part of the experience. Soft SF shifts the weight toward sociology, anthropology, and psychology, or focuses on people and societies rather than on scientific rigor.

It is a useful distinction. The trouble is that this axis is so often mistaken for a hierarchy. A quiet ranking circulates in which hard is the real SF and soft a side branch. Look at it for a moment, though, and the ranking will not hold. Faster-than-light travel collides head-on with known physics, and works called hard use it without blinking. Perfectly hard SF effectively does not exist; everyone takes a liberty somewhere.

What matters is not which rules you keep but whether you keep the rules you set for yourself. Andy Weir's The Martian (2011) feels hard even though critics pointed out that the dust storm in its opening scene could not do that much damage in the real Martian atmosphere, because once the conditions are set every problem inside them is solved by calculation. The reader comes to trust that the author will not fudge, and that trust generates tension.

By the same standard, Le Guin's Gethen is hard on anthropology rather than on physics. The kinship structures, the tax system, the insults, and the myths of a genderless society are all woven consistently. Hard and soft are less a grade of rigor than a difference in which discipline you staked that rigor on.

The Novum — Change One Thing and Push It All the Way

The person who gave this discipline its exact name was the literary scholar Darko Suvin. In a 1979 book he proposed the two concepts that define SF: the novum and cognitive estrangement.

Novum is Latin for new thing, and Suvin borrowed the word from the philosopher Ernst Bloch. In principle, one novum per work is enough. One positronic brain, one genderless humanity, one star system with three suns. And the writer's job is to follow all the way out how that one thing spreads through society, economy, language, and daily life.

Cognitive estrangement names the effect. It is the business of taking something too familiar to be visible, such as labor or family or borders, and setting it in a world with one different rule so that it becomes visible again. The attached condition is that it must be estranged and still cognitively graspable. That is why SF asks the reader to work, and why readers read by inferring the rules of the world.

Failed SF is mostly a case of inserting a novum and never calculating the fallout. A world where teleportation is commercialized but commutes and property prices are unchanged; a world where aging has stopped but inheritance and pensions run on twentieth-century rules. If you put in one new thing, you have to move everything that one thing touches.

As you may have noticed, this discipline is exactly the one used to build a magic system in designing a fantasy world. Whether the material is a faster-than-light engine or a spell, setting the rules and then playing honestly inside them is one and the same job. The wall between genres is thinner than it looks.

Six Branches of the Lineage — From the Golden Age to Korean SF

The Golden Age — Rules Are Not There to Be Kept

Isaac Asimov's Three Laws of Robotics were first spelled out in the 1942 short story "Runaround" and collected in I, Robot (1950). A robot may not harm a human, must obey human orders, and protects itself within those limits. They are still cited in debates about the ethics of artificial intelligence today.

Read Asimov's robot stories, though, and something odd turns up. Accidents happen even though the Three Laws are obeyed. Or rather, they happen because the Three Laws are obeyed. Two laws collide, or the scope of the word harm is ambiguous, or the robot interprets a law so faithfully that it arrives at a conclusion nobody wanted.

The Three Laws, in other words, are not an ethical guideline but a logical device. Asimov was not proposing rules; he was exhibiting the ways rules fail. Stated rules, an apparently impossible phenomenon, a dismantling carried out by logic alone. This structure is a sibling of fair play in detective fiction. That is why Golden Age SF so often reads like a puzzle.

The New Wave — From Outer Space to Inner Space

In 1964 the current split, when Michael Moorcock took over the editorship of the magazine New Worlds. J. G. Ballard argued that the space in need of exploration was not outer but inner. In place of rockets and aliens came psychology, sex, catastrophe, and experiments in prose.

Samuel R. Delany's Babel-17 (1966) took as its novum the very hypothesis that language shapes thought, and Dangerous Visions (1967), the anthology edited by Harlan Ellison, gathered up the material the magazines of the day would not print and issued it like a manifesto. Philip K. Dick spent the whole period repeating one question, what reality is, from a dozen angles.

What the New Wave left behind is not subject matter but an attitude. It was the period when SF redefined itself from a magazine genre into a mode of literature, and after it no branch was ever entirely free of demands about sentences and form.

Cyberpunk — High Tech and Low Life

In the 1980s the direction turned again. Not space but the city, not the state but the corporation, not the mind but the body and its interfaces. Mirrorshades (1986), the anthology edited by Bruce Sterling, became the label for the current; Ridley Scott's Blade Runner (1982) built its visual grammar; and Masamune Shirow's manga Ghost in the Shell (1989) together with Mamoru Oshii's animated film (1995) laid that grammar back over an Asian city.

Look again at the report card on Neuromancer and it gets interesting. What Gibson got right was the structure of a society: corporations outgrowing states, access to information becoming a class marker, the body becoming an object of modification and commerce. What he missed was the shape of the devices. Mobile phones, search, social media. The structure right and the gadgets wrong — that is exactly the character of what a thought experiment does.

Hard SF — When the Math Becomes the Story

Arthur C. Clarke's Rendezvous with Rama (1973) is the story of humans investigating an enormous artificial structure as it passes through the solar system, and Larry Niven's Ringworld (1970) seriously calculates a ring structure wrapped around a star. Kim Stanley Robinson's Red Mars (1992) pushes the remaking of Mars from the political and the engineering side at once.

East Asia's standard-bearer is Liu Cixin. The Three-Body Problem ran as a magazine serial in 2006, appeared as a book in 2008, reached the world through Ken Liu's English translation (2014), and won the Hugo Award in 2015. What impresses about the work is the size of its novum. It takes one computational difficulty from celestial mechanics, the three-body problem, and pushes it up into a condition for the survival of a civilization, then up again into game theory between civilizations.

The pleasure of hard SF lies not in accuracy itself but in the necessity that accuracy produces. When the laws of physics rather than the author drive a character into a corner, the reader feels that the world, and not coincidence, is pushing the story.

Feminist SF — Making the Obvious Strange

After The Left Hand of Darkness, Le Guin set a society without property beside a society that is nothing but property in The Dispossessed (1974). There is a reason the subtitle calls it an ambiguous utopia. Neither side is drawn as the right answer.

Joanna Russ's The Female Man (1975) broke form itself, layering several possible worlds of the same person, and Octavia Butler took another road. Kindred (1979) is the story of a Black woman in contemporary America dragged repeatedly back to a nineteenth-century slave plantation. It uses the common device of time travel but converts it into the sense that history is not yet over. Parable of the Sower (1993) depicts an America where climate and society have collapsed together.

Margaret Atwood's The Handmaid's Tale (1985) comes with a principle that is often quoted. She put in no form of oppression that lacked a precedent in actual history. Thanks to that principle the novel is not an imagined horror but a rearrangement of things that have already happened somewhere, and that makes it far more frightening.

Korean SF — From the Position of Those Left Behind

Since the late 2010s Korean SF has visibly thickened. Kim Choyeop's If We Cannot Go at the Speed of Light (2019) and Greenhouse at the End of the Earth (2021), Bae Myung-hoon's Tower (2009), Cheon Seon-ran's A Thousand Blues (2020), and Bora Chung's Cursed Bunny (2017) branch off in different directions. Cursed Bunny was shortlisted for the International Booker Prize in 2022 and found readers outside Korea as well.

The difference in grain is unmistakable. Where the long-standing default of Western SF was expansion, conquest, and pioneering, recent Korean SF points the camera at the ones left behind rather than the ones who left. The places technology never reached, the settlement after the migration is over, care and disability and loss. The spaceship is the background; in the foreground there are relationships.

Bora Chung is another axis again. She erases the borders between SF, horror, and the fantastic, and takes aim at the workings of capital and patriarchy in the form of a fable. It is still early to bind this current into a single movement, but at the very least the taste in choosing a novum is clearly different.

Dystopia Is a Warning, Not a Prophecy

No form is misread as prediction more often than dystopia. Every time surveillance technology makes the news, George Orwell's 1984 (1949) is summoned and people say Orwell got it right. But this way of marking the paper misses two things.

First, the lineage is longer. Yevgeny Zamyatin's We (1924) is closer to the archetype of the form, and Aldous Huxley's Brave New World (1932) precedes 1984 by seventeen years.

Second, and more importantly, the two flagship works aim at different dangers. Orwell's domination works through fear, scarcity, and surveillance. It shrinks language until thought is impossible and rewrites history until nothing can be checked. Huxley's domination is the exact opposite. It works through abundance, pleasure, and conditioning. It forbids nothing and gets people to love their own oppression. The media theorist Neil Postman set out this contrast in a 1985 book, saying in effect that Orwell feared the people who would ban books while Huxley feared a state in which nobody would want to read one.

Read as prophecy, the difference is erased. The book becomes a scorecard about having called the telescreen, and the diagnosis coarsens. Read as warning, the question changes. Is it fear or is it pleasure that has hold of us now. Instead of counting surveillance cameras you start asking which mechanism is running, and only then do the two novels work as two different medical charts.

Where to Start — Entry Paths by Taste

• If you want the pleasure of ideas: Asimov, I, Robot → Ted Chiang, Story of Your Life → Liu Cixin, The Three-Body Problem
• If you want the fun of calculation: Andy Weir, The Martian → Arthur C. Clarke, Rendezvous with Rama → Kim Stanley Robinson, Red Mars
• If you want to look at society and people: Le Guin, The Left Hand of Darkness → The Dispossessed → Octavia Butler, Kindred
• If you enjoy atmosphere and prose: Philip K. Dick, Do Androids Dream of Electric Sheep → Gibson, Neuromancer → J. G. Ballard
• If you would rather start in Korean: Kim Choyeop, If We Cannot Go at the Speed of Light → Cheon Seon-ran, A Thousand Blues → Bora Chung, Cursed Bunny
• If you would rather get the feel on screen first: Blade Runner (1982) → Arrival (2016) → Ghost in the Shell (1995)

Closing — Good SF Leaves You a Question, Not an Answer

The commonest misunderstanding about SF is that the genre is about the future. In fact the tense SF works in is always the present. A world with one condition changed is a mirror held up to the world we live in, and a mirror is not a device for showing what lies ahead but for showing what is here.

So the next time you open an old SF novel, you need not be disappointed that there are no mobile phones in it. The thing to ask is not what the author got right, but how honestly he followed the one condition he changed as it spread outward. A hit is luck, but honesty is skill. And the one still being read forty years later is always the latter.