How Much Science Do You Need to Write Science Fiction?
Here is the short answer: you need enough science to keep your reader from tripping over a mistake, and not one fact more than the story uses. The science in science fiction writing is there to serve the story. It is not a final exam, and nobody hands out a degree before you are allowed to write about spaceships.
I get this question from newer writers a lot, usually phrased as a confession. “I want to write SF, but I am not a scientist.” Neither are most people who read it. The trick is knowing where your story sits on the hard vs soft sci-fi writing spectrum, then researching science fiction the smart way: targeted, limited, and in service of the scene in front of you.
Science in science fiction writing: what readers actually expect
Readers do not expect you to be right about everything. They expect you to be consistent and to not insult them.
That breaks down into three promises:
- Do not break the real-world science your story leans on. If your plot depends on a slow leak of air, readers will notice if the air lasts forever.
- Follow your own rules. If your faster-than-light drive takes three days to spin up in chapter two, it cannot fire instantly in chapter twenty because the plot needs it to.
- Give the important stuff weight. The science that drives the plot deserves attention. The science that is just furniture can stay in the background.
Notice what is missing from that list: equations, citations, and a lecture. Readers forgive invented technology all day long. What they do not forgive is a story that ignores its own logic.
Hard vs soft sci-fi writing: pick your spot on the dial
Hard vs soft sci-fi writing is less a pair of boxes and more a dial.
At the hard end, the story treats real physics, chemistry, biology, and engineering as the main source of conflict. Andy Weir’s The Martian is a well-known modern example: much of the tension comes from a stranded astronaut solving one technical problem after another. Readers of that kind of book want the numbers to work, and they will check.
At the soft end, the science is a backdrop, and the real focus is people, societies, politics, or ideas. Star Wars sits way out here. Nobody watches it for orbital mechanics. Ursula K. Le Guin’s The Left Hand of Darkness is a classic example of science fiction where the “science” that matters most is anthropology and society rather than hardware.
Most stories land somewhere in the middle. A space opera might keep its starship combat loose but get the medical details right. A cyberpunk story might hand-wave the neural implants but get the networking and corporate politics grounded enough to feel real.
Here is the practical part. Decide where your story sits before you start researching. That one decision tells you how much research you need. A story near the soft end needs plausibility and consistency. A story near the hard end needs real homework on the specific systems your plot depends on.
You do not need permission to write soft SF. It is not a lesser form. It is a different promise to the reader.
Researching science fiction without drowning in it
Research is a tar pit. It feels productive, it is fun, and it can eat months that should have gone into pages. When you are researching science fiction, use a few guardrails.
Research the plot, not the universe
Start with the scene. What does the character need to do, and what could go wrong? Research that. If your engineer is fixing a cooling system on a space station, you need a working idea of how heat is dumped in a vacuum. You do not need the history of every refrigerant ever made.
Ask “what would break?”
The most useful research question in SF is not “how does this work?” It is “how does this fail?” Failure modes give you conflict. A ship’s reactor that hums along forever is set dressing. A reactor that overheats when the radiators get damaged is a plot.
Use good general sources first
You do not need journal access to get started. Some reliable places to begin:
- NASA’s public websites and educational material for space, planets, and spaceflight basics.
- University and museum outreach pages for clear explanations written for non-specialists.
- Introductory textbooks or popular science books on the field your story touches.
- Simulation and worldbuilding tools that let you test an idea. I keep a list of these in my post on world building tools for authors and game masters, including star system generators and planet simulators.
When something matters to the plot, check it in more than one place. Popular articles sometimes oversimplify.
Ask an expert one good question
If you know a nurse, a mechanic, an engineer, or a pilot, buy them a coffee and ask one specific question. “What would really go wrong if…?” is a great opener. People who do a job every day know the details that make fiction feel lived-in, and they usually enjoy being asked.
Set a time limit
Give yourself a fixed block of research time per scene or chapter, then go write. If you hit a gap mid-draft, drop a note in brackets, something like [CHECK: how long does a suit battery last?], and keep moving. Fix it in revision. The draft does not care that you did not know yet.
How my Navy and IT background shapes my science fiction
I came to fiction through a side door. I spent eleven years in the U.S. Navy working in electronics, helicopter avionics, and advanced computer systems. After the Navy, I worked in information technology as a technical writer and teacher, including a year in Thailand teaching component-level repair of computerized systems. All told, that was about 35 years in technology and related fields before I finished my MFA in Creative Writing in 2018.
What did that background teach me about writing SF? Not formulas. It taught me how systems behave.
Machines fail in specific, boring, expensive ways. A connector corrodes. A cable chafes. A fix that worked in the shop does not work in the field. Somebody skipped a step in the manual because they were tired. That is the texture that makes technology feel real on the page, and you can get it without a technical career. I wrote more about this in writing science fiction with a technical background. Talk to people who maintain things. Read accident reports and post-mortems written for general audiences. Watch how real troubleshooting goes: guess, test, fail, guess again.
That same instinct carried over into my game writing on the Interface Zero 2.0 core rules team and into Solar Singularity, the Interface Zero 2.0 novel I co-wrote with Peter J. Wacks and Josh Vogt. Cyberpunk lives on believable tech, and believable tech is mostly about how it breaks and who profits when it does.
If you have a technical background, use it. If you do not, borrow the habit of thinking that way: every system has limits, costs, and failure points.
Getting the science right on the page
Having done the research, you still have to put it in the story without stopping the story. A few rules I lean on:
Show the science through action. A character checking an oxygen gauge every few minutes tells the reader more than a paragraph about partial pressure. My post on show, don’t tell with active voice in speculative fiction goes deeper into how to do this.
Let characters know things without explaining them to each other. Two engineers do not explain how a reactor works to each other. They argue about the part that is broken. The reader picks up the rest from context.
Turn constraints into tension. Physical limits are free conflict. Air runs out. Fuel runs low. Light-speed delays mean nobody back home can help in time. That is exactly how a ticking clock works, and I break it down in creating tension with a ticking clock.
Keep the iceberg under water. You might research ten facts to use one. That is fine. The other nine still help, because they keep you from writing something wrong. Do not dump them all into the text to prove you did the work.
Build the world only as far as the story needs. This is the same principle I apply to worldbuilding in general, which I cover in why sci-fi worldbuilding should serve the story.
An illustration: two versions of the same scene
This is an invented example for illustration only, not from any published story.
Version one: “The ship’s fusion reactor used magnetic confinement to contain plasma heated to millions of degrees, and the coolant system was failing. Captain Reyes knew that without cooling, the reactor would have to shut down.”
Version two: “The coolant alarm had been chirping for an hour. Reyes stopped hearing it after the first ten minutes. She noticed it again when it stopped, and the engineer looked up from the panel with an expression she had seen once before, right before they lost the Tamsin.”
The first version is accurate and dead on the page. The second version uses the same science (cooling failure, reactor shutdown), but it shows the reader a person under pressure. The research is still there. It is just doing its job quietly.
A quick research checklist for new SF writers
Before you start, or when you get stuck, run through this list:
- Where does my story sit on the hard vs soft dial?
- What science does the plot actually depend on? List it. Keep the list short.
- For each item: how does it work in broad strokes, and how does it fail?
- What are my invented rules (FTL, implants, aliens), and are they written down so I stay consistent?
- Which one or two facts would an informed reader check first? Verify those.
- Did I set a research time limit, and am I sticking to it?
If you can answer those, you know enough to write the draft.
FAQ
Do I need a science degree to write science fiction?
No. Many SF writers come from non-science backgrounds. What you need is curiosity, a willingness to look things up, and the discipline to keep your own rules consistent. Targeted research can cover the specific science your story uses.
What is the difference between hard and soft science fiction?
Hard science fiction keeps close to real, known science and often makes technical problems the core of the plot. Soft science fiction uses science more loosely and focuses on characters, societies, or ideas. Most stories fall somewhere between the two.
How much research should I do before I start writing?
Enough to understand the science your plot depends on, and no more. Research the key systems, learn how they fail, then start drafting. Flag smaller gaps with a bracketed note and fill them during revision.
What if I get the science wrong?
It happens to everyone. Get the plot-critical details checked by a knowledgeable reader if you can. For minor errors, consistency matters more than perfection. Readers will forgive an invented technology much faster than a story that breaks its own rules.
Write the story first
The science is a tool. It is there to make your world believable and to create pressure your characters have to push against. Pick your spot on the hard-soft dial, research what the plot needs, and get back to the page. The story is the part only you can write.