|You know it makes sense|
This is a discussion of a long-standing dispute within the science fiction field, about boundaries and categories. It has to be said that in recent writing such boundaries have become increasingly permeable, and in some ways this discussion may be outdated or at least old-fashioned. But in much of the field the debate is still current. Even for those who see the boundaries as outmoded, it may be useful to think about them as background to what will come next.
Science Fiction is often divided into "Hard" and "Soft". Proponents of "Hard SF" sometimes tend to imply that it alone is real SF. This is not, in my view, a tenable position.
- What are "Hard" and "Soft" Science Fiction?
- Defining "Science Fiction" and "Fantasy"
- Imaginary Science vs. the Supernatural
- What is Hard SF good for?
- Star Trek
- Concluding thoughts
What are "Hard" and "Soft" Science Fiction?
"Hard science fiction" means the category of science fiction which is within the range of what is believed to be scientifically possible (or at least not yet known to be impossible). The technology may be way beyond anything we can do or have any idea how it would be done, but what the technology in the story does is not contrary to any scientific laws we know about. The term "soft SF" is sometimes used for science fiction which does not adhere to this limitation. For example, the forms of time travel normally appearing in SF seem to be impossible. There is also the question of where "fantasy" fits in.
This page started out as a note to the review of Arthur C. Clarke's novel The Songs of Distant Earth. As I noted in the review, Arthur Clarke explained the inspiration partly in terms of a reaction to what he saw as the "fantasy" elements of Star Trek (he also mentioned Star Wars, but that is avowedly Space Opera and doesn't seem relevant). Writing a book within the constraints of "hard SF" is fine. But if this is the definition of science fiction as against fantasy, then a significant part of Clarke's own writing comes under fantasy. Against the Fall of Night (1948–53) (and its later, but in my view less satisfactory version The City and the Stars) involves a spaceship able to travel between stars in short periods. Some of his best short stories involve time travel and other dimensional adventures (e.g. "All the Time in the World", 1951). 2001: A Space Odyssey, both in film and novel versions, involves some sort of short-cut through space. Childhood's End (1953), often regarded as his masterpiece, involves several things that are pretty hard to fit into known science. But to be fair most of his later novels take a fairly hard-SF approach, including his notable Rendezvous with Rama (1972).
I think we can reasonably add that the test of possibility is that of the time a work was written. Stories do not move out of the SF category because of advances in science, though we would probably draw readers' attention to the change.
Clarke wrote in 2000 that defining science fiction was so difficult, he would accept "Damon Knight's magisterial: 'Science fiction is what I point to and say 'That's science fiction'." (I.e., "I know it when I see it.") He went on:
Much blood has ... been spilled on the carpet in attempts to distinguish between science fiction and fantasy. I have suggested an operation definition: science fiction is something that could happen—but usually you wouldn't want it to. Fantasy is something that couldn't happen—though often you only wish that it could.
(Foreword, The Collected Short Stories of Arthur Clarke, 2000)
This is essentially a definition of "hard SF". However, despite the claims sometimes made by its writers and readers, hard SF has never been the only form. "Soft SF", SF that does not limit itself to known science, is harder to define. It allows some imaginary science but works within it. Probably the most common convention is the assumption that faster-than-light travel is possible, as in Star Trek—this makes all sorts of stories possible. Time travel is another common convention. Interestingly, soft SF is often more suited to social commentary, since hard SF may limit you too much. (To confuse the issue further, "soft SF" is also used to refer to SF about the "soft" or social sciences, but I think this usage leads to confusion.) Isaac Asimov wrote all sorts of fascinating social SF, but it was often made possible by the assumption that you could travel between stars in reasonably short times. This applies to Earth and the Spacer colonies in The Caves of Steel and on the grand scale of the Foundation books. This is not universal, of course. See the classic of social SF "The Snowball Effect" by Kathleen MacLean (1952) which doesn't involve any technology beyond a pencil and paper.
Although the terms are relatively recent, the division goes back to Jules Verne and H. G. Wells. Verne sent his voyagers to the moon by a giant cannon; Wells by an imaginary substance ("cavorite") which blocks gravity (which violates both Newtonian physics and general relativity). (On the other hand, Verne's passengers would in reality be turned into jelly. So is it really less imaginary?) Wells also invented the Time Machine, enabling him to write about seeing future societies. The technology in Wells's stories was there to enable thinking about humanity outside its current environment, and it was of secondary importance whether it was possible.
Defining "Science Fiction" and "Fantasy"
The complete OED defines "science fiction" as "Fiction in which the setting and story feature hypothetical scientific or technological advances, the existence of alien life, space or time travel, etc., esp. such fiction set in the future, or an imagined alternative universe." "Hard science fiction" is "...science fiction dealing with technological advances which do not contravene currently accepted scientific laws or principles; of or characteristic of such science fiction." ("Soft SF" doesn't seem to be in OED yet.) Note that "science fiction" is defined in terms of examples (with an "etc." get-out). Off-hand it seems to be a fairly good fit.
For "fantasy", however, the OED gives (in this sense) "A genre of literary compositions". This is unhelpful, to say the least. OED does give, as one of the examples, a quotation from 1955 (F. Brown, Angels and Spaceships) that "Fantasy deals with things that are not and cannot be. Science fiction deals with things that can be, that some day may be." This is essentially the same as Clarke's definition. But, like many definitions, this does not correspond to actual usage. Clarke himself did not ask for Against the Fall of Night or 2001 to be reclassified as "fantasy". Isaac Asimov's novels involve faster-than-light travel and (in one case) time travel. But his novels are not described as "fantasy". Rather, he is listed as a famous writer of science fiction.
Imaginary Science vs. the Supernatural
Science fiction and fantasy are essentially literary categories. We have hard SF, soft SF, and fantasy; where are the lines drawn? I think it is helpful to note that there are two different sorts of dividing line being used. Hard and soft SF are distinguished by scientific possibility. SF and fantasy are distinguished by being different literary types.[1a] Although both soft SF and fantasy involve things which are not scientifically plausible, hard and soft SF are artistically closer to each other. (For readers, how much difference does it make that the spaceships in Clarke's The Songs of Distant Earth travel by the (possibly possible) vacuum energy rather than by some undefined "antigravity" or "hyperdrive"?) OED's main definition of science fiction, which simply describes the typical forms, is more accurate: time travel is mentioned along with space travel. I would suggest, as a first approximation, that we can distinguish imaginary science from the supernatural. Asimov's ships make jumps in hyperspace; there may be no such thing but in the story it is just the advanced science of the time. It is based, in the story, on logical mathematical models and physical machines. Scotty famously declares that he can't change the laws of physics ("The Naked Time", TOS). On the other hand, when Lucy finds a wood at the back of the wardrobe, there is, even in the story, no such logical system. It isn't mathematical, or replicable. The borders are inevitably fuzzy and have become fuzzier. Many now see no reason to draw a line. But apart from that, is Out of the Silent Planet science fiction or fantasy? It involves space travel by a supposed technological means, though C.S. Lewis said that the explanation was, to use the later phrase, pure technobabble. With the sequels, however, Perelandra involves magical space travel, and That Hideous Strength is clearly fantasy. E. Nesbit's masterpiece The Story of the Amulet (1906) involves time travel, but by magic, and is fairly clearly fantasy.
Mind you, it was Arthur Clarke who said that "Any sufficiently advanced technology is indistinguishable from magic", whch complicates the issue a bit.
When we talk of "imaginary science", however, there are further ambiguities. Isaac Asimov's jumps in hyperspace are not meant to be closely related to any real science, though we do learn a few of the limitations that supposedly govern it. However, science does change. You may recall that a while ago it seemed briefly that neutrinos might be travelling faster than light. This seemed unlikely, and turned out as expected to be a mistake, but it would have changed things. It also gave rise to one very good joke:
Suppose, as a thought experiment, that someone had written a science fiction story just before Einstein, in which travel between stars seems short to the astronauts though many years pass on Earth—the time dilation effect. This would have been contrary to the laws of physics as then understood, and so "soft SF" using "imaginary science". But it would be in accordance with the laws of physics as understood now. Thus, there is nothing impossible in principle with the proposition that future discoveries may make possible what is not thought possible now. However, some developments are much more plausible than others. In the classic SF short story, "The Big Bounce", by Walter S. Tevis, someone invents a rubbery material which bounces higher than it falls. Where does the energy come from? From heat—the ball ends up colder than it was. This violates the Second Law of Thermodynamics, i.e. it reverses entropy. The Second Law of Thermodynamics seems to be something very basic about how the universe works. So in fact, despite not seeming so dramatic as hyperdrive or whatever, this is one of the least plausible developments. In fact the story illustrates neatly what it would mean to violate this law. Curiously, it is about real science in a way that much hard SF isn't.
Although I think the distinction above (imaginary science versus the supernatural) is broadly accurate, the distinction between the two is not always 100% clear. In some stories, "supernatural" elements are presented as actually having a scientific (usually imaginary-science) basis. For example, Larry Niven's 1969 story "Not Long Before the End" posits that, in the remote past, there was a sort of non-renewable energy source which powered magic. This energy source was gradually used up by magic, and it is implied that eventually it was completely or almost completely exhausted, leaving the present magic-less world. The story, which Niven expanded into a series, is usually described as "fantasy" (e.g. in the Wikipedia page). Its classification as fantasy brings us to an important point: the classification is justifiable on the basis of literary characteristics; the story is Niven's take on a well-established motif in "sword-and-sorcery" fantasy writing. For someone who isn't familiar with the genre a headnote would be a good idea. This indicates that the attempt to define these categories by straightforward rules about science, possibility, etc., are overlooking the role of literary style. Hard and soft science fiction are printed together in the same SF magazines and anthologies because they are very similar sorts of stories.
Harry Potter is another ambiguous case. It has been frequently pointed out that the magic in J.K. Rowling's world is an allegory for scientific and technological power. Arguably, it is a form of technology. The children learn techniques from expert teachers, and there are theoretical laws that govern what is possible magic. In some ways the premise is that there exists, apart from the laws of nature familiar to us Muggles, an extra or parallel set of laws; Muggle knowledge of this was suppressed at a point in the past, leaving behind distorted traces. However, you need to possess some inborn magical quality to use these forces, and it seems to depend on a somewhat less methodical magic in the background. But the stories, while different in feel from, say, Ursula Le Guin A Wizard of Earthsea, are clearly perceived by readers as "fantasy"; and especially in the later books we start to see connection of the magic to things like the soul.
What is Hard SF good for?
Hard SF is (in my view) most interesting where the insistence on actual science contributes to the story. Arthur Clarke's short story "Jupiter Five", which turns on some science which one of the characters (and quite probably the reader) did not understand, is an example. Another example is his novel The Fountains of Paradise, about a space elevator: a technology which is not yet possible for Earth conditions but is discussed as a genuine possibility for future space travel. Also, in principle the limitations of hard SF could be a source of artistic shape, like any limited artistic form; but I can't think of a very good example off-hand. In many cases, though, it is hardly relevant; as I noted, The Songs of Distant Earth would be hardly changed if Clarke had used an imaginary space drive. The story does require that the spaceships take a long time to reach the other stars, rather than jumping there instantly, but that's all. Hard SF can also be a way of thinking about newly important real-world science and technology; this is often compelling when new but quickly becomes out of date.
Saying that Star Trek is fantasy (rather than SF) only makes sense if we are using a technical sense of the word different from that in normal use. Since the normal use is itself a technical term, for distinguishing a particular genre of writing, this seems unhelpful. Star Trek is in some ways a category by itself. It is variable, and includes things which are definitely science fiction, like "This Side of Paradise" (TOS), but also things which merely use a SF setting, such as "Balance of Terror" (TOS).
Perhaps the idea of "hard vs. soft" is misleading. One possible approach would be to take "science fiction" as the main category (fuzzy round the edges but reasonably recognizable) with "hard science fiction" as a special type, which is at the same level as other special types. If you are especially interested in "hard SF" you could describe all SF other than "hard SF" as "soft SF". In this sense "soft SF" would be a residual category, but unless you have a special interest in hard SF it isn't a useful category. A film enthusiast might divide things into "film" and "non-film", the latter including novels, poetry, TV, and oral sagas, but other people would probably not agree these are categories at the same level. Another approach might be to see it as a spectrum.
Hard SF has never been the only, or even the main, form of science fiction. I usually find it less interesting, because it is more often a hindrance than a help for the sorts of stories that most interest me. Isaac Asimov, despite being (until he became a full-time writer) a professional scientist, wrote mainly "soft" SF because he was interested in ideas. How does a "positronic brain" work? Who cares? Asimov uses his robots to make us think about human beings and society. Could you really zoom around the Galaxy in hyperdrive ships? Probably not, but it enables us to think, for example, about alternative ideas about the best sort of human future (Foundation's Edge, 1982). Science fiction has a lot on offer: take your pick.
 The term is best explained by Kingsley Amis's classic study of SF, New Maps of Hell (1960): "In space-opera, Mars takes the place of Arizona with a few physical alterations, the hero totes a blaster instead of a six-gun." That is, the paraphernalia and ambience of science fiction are borrowed for a story of a more conventional sort. The classic modern example is Star Wars. In the original film (later retitled A New Hope) the story has classic martial arts and political aspects. But the technology, the new environments, the creatures, do not seem to have changed anything fundamental in life. As Star Wars shows, space opera can be superb, but, despite some overlap, I think it should be distinguished from mainstream SF. [Return]
[1a] It's rather as if you asked "what is the dividing line between the French army, the British army, and the British police?" If you are imagining some sort of continuum with dividing lines, the question is unanswerable; in fact you are asking about two different sorts of distinction.) [Return]
 It's worth noting however that Lewis, in these books, deliberately challenges the distinction between science and the supernatural. "Space" is reinterpreted in terms of the premodern "Heavens". Lewis had argued in his academic book The Discarded Image that the medieval world-view was a single integrated view that included both science and religion; Out of the Silent Planet tries to restore this. Lewis was very interested in SF when it was still quite unfashionable. (See also the page on C.S. Lewis and Star Trek.) [Return]
 In Asimov's novel Fantastic Voyage II: Destination Brain (1987) he thought out in some detail what the physics of miniaturization might look like—I don't suppose it's really plausible but certainly he treats it in a different way from the loose description of hyperspace jumps. By the way this novel is different from his earlier Fantastic Voyage (1966) which was a novelization of the film and which he felt was unsatisfactory. [Return]