Science in science fiction, or how realistic should and can we be?
By Aliette de Bodard
A large subset of science fiction concerns itself with extrapolating the future, and most particularly with imagining where science could take us . A question that naturally follows is: how realistic should the science be in those stories/novels?
A common answer (and indeed, the first that comes to my mind) is that, if you want to be rigorous, it should be realistic: after all, science fiction is a prediction of what futures might come to pass, and there doesn’t seem to be much point in having something we know can’t possibly exist–say, spaceships that violate the laws of physics or of thermodynamics.
So far, so good. Or is it? The problem here is the “we know can’t possibly exist”. I have deliberately forgotten a very important word in my argument: it’s not that the faster-than-light spaceships would violate the laws of physics–but that they would violate the known laws of physics.
Fine–but that doesn’t seem like a big difference. The basic laws of, say, thermodynamics, are immovable. Unfortunately, that is far from the case. Science is not etched in stone once and for all, but it keeps being updated as we make new discoveries. To give you an idea: the last significant time when humanity thought it understood everything about the world; that we’d hammered into stone the limits of the possible was near the end of the 19th Century, right after we tied Maxwell’s laws to optical phenomena. It seemed that we’d achieved total comprehension of the universe with electromagnetics and mechanics. There remained just two-three niggling little experiments that no one could explain, surely blips on the radar…
Or perhaps not. Within a couple decades, those two-three niggling doubts had given rise to quantum physics, and then general relativity, which pretty much turned upside-down our perception of the universe. Before that, it seemed impossible to have a dual wave-particle nature; and preposterous to think systems could be in several states at the same time. After quantum physics, it was not only possible, but natural–at least within the scientific community.
So science is not a monolithic bloc: it keeps evolving and contradicting itself, and it’s possible (actually, highly probable) that the laws we take as unbreakable today will end up being inaccurate tomorrow. With that in mind, when we write science fiction set more than 20-30 years in future, what we’re writing is not realistic–rather, it’s what is plausible based on today’s corpus of research. And I’m being optimistic here, as not every SF writer will be able to keep abreast of research in every field (though some are really impressive at it). So at best we’re plausible, but not really accurate.
Does that matter? To me, not really. The main value of SF is not scientific prediction, or we would all be writing research articles for Nature, Science or Physical Review Letters, and working in labs. SF can be entertainment; it can be a warning about the effect of some technologies; it can be amazing feats of worldbuilding that impress us and transport us to another kind of reality; it can and has been a thousand things, and realistic science is ultimately a very small part of it.
What matters is writing plausible stories about the future; and “plausible” matters very much indeed. “Plausible” determines resonance with the audience. “Plausible” is what gives meaning and context to SF stories.
And, even supposing we could predict future science perfectly, it likely would be so alien after a few decades that explaining it to the reader would be as hard as explaining quantum physics to a 19th-Century scientist (God knows it’s enough trouble explaining it to undergraduate students). So we fake it.
But then again, isn’t that what all writers do?
 I’ll leave aside the subgenres of SF where scientific extrapolation isn’t a big part, like alternate history or science fantasy. There’s lots of good things in Ann McCaffrey’s Pern stories, but they’re not exactly scientifically accurate.