Gregory Benford has championed hard science fiction as the core of the sf genre, saying that any other type of science fiction is "playing tennis with the nets down" (Spinrad etc., 93). Stanislaw Lem has criticized sf by saying it tends to "domesticate the universe" because it ignores contemporary cosmology and physics while adhering to formulas for popular fiction. Both of these statements seem to be heralding a call for "harder" sf, sf that pays closer attention to the realities of existing science. But to what extent should science fiction adhere to scientific fact when faced with the choice between accuracy and dramatic content? Neither Benford's nor Lem's statements allow room for departing from scientific reality in order to tell a better or more dramatic story. Should writers sacrifice artistic quality in order to stay true to what we currently believe to be fact? If they do not, can what they write still be considered hard science fiction?
In order to answer this question, we must first decide exactly what it is we mean when we say "hard science fiction." It should not be limited to stories focusing on science and technology, nor can it include stories in which the writer departs from known scientific fact. But if writers adhere too closely, "if the science is 100 percent faithful to the best available knowledge, then a piece of fiction isn't science fiction at all, since scientific speculation is then entirely absent, and what we have is mimetic fiction with futuristic technological trappings" (Spinrad, 103). So what we need is a happy medium. Norman Spinrad's definition seems to be the most workable: "fiction that applies all the available literary techniques to preserve the illusion of verisimilitude, while it pushes the edge of the best known scientific world view just far enough to enter terra incognita without actually contradicting known scientific fact" (104).
According to this definition, then, sf writers may extrapolate freely as long as they don't break any laws currently held to be true by science. Thus hard sf cannot have spaceships zooming around at faster-than-light velocities or people hurled backwards in time. As far as we know today, these things are impossible to achieve by any means, not just by current technology. Even if a reasonable explanation is given as to how these things could happen, they still defy what science as we have it today sees as true, and thus it does not fit into the definition we are using. The only way that this can be sidestepped is if theoretical science suggests that current knowledge may be incorrect and this allows the writer to use the current theory in place of what was previously held as fact. In this case the departure from scientific "fact" is not the work of the writer, but science itself. Controversial theories aside, though, hard sf can still write about artificial intelligences or "generation starships." Although we don't currently have the knowledge or technology necessary to accomplish these things, they are not things that depart from scientific fact as we see it today.
Benford believes that all science fiction should be handled in this manner. To disregard known scientific fact is lack of rigor on the part of the writer. But should this necessarily be the case? What if a pseudo-scientific device is necessary to make the story more effective? Say we would like to tell a story about how human evolution will continue into the far future, but we need a time machine to do so. Or what if we would like to examine what might have happened if the axis powers won the Second World War, but we must tell the story in an "alternate universe"? Benford's position implies that the science comes first and must be adhered to even if it makes the story less effective, but in this case stories like The Time Machine and Philip K. Dick's The Man in the High Castle described above could not be told at all if all sf were to play strictly by the rules. Fiction, though, is a form of art and entertainment and should be done in such a way to make it enjoyable or aesthetically pleasing, regardless of whether it conforms to science. This is not to say that science fiction shouldn't comply with scientific fact if the story is not hindered by it; if a departure from reality is made without reasonable justification, then it is laziness on the part of the author not to apply the appropriate scientific rigor to his work. But the purpose of sf, like any other type of fiction, is to tell a story, not to popularize the latest scientific theories; and if we must forego some aspect of science in favor of telling the story, then that's what should be done.
The real question, then, is not whether writers should conform to the laws of science, but rather is it really sf if they do not. It certainly cannot be called hard sf if they break a known scientific fact, but we can still safely classify a story as sf if time travel is used (and it is in fact one of the most common elements of sf stories.) It seems logical that we may call a story sf--although not necessarily of the hard variety-- if it has a certain illusion of verisimilitude, exhibiting plausible explanations of how certain physical "laws" have been sidestepped or foregone (e.g. Dan Simmons' Hyperion books, in which time travel is used but explained in a believable manner in a context in which the author displays a good deal of scientific competence in other respects) rather than presenting overt impossibilities with no attempt at explanation (such as in Star Wars, which is more akin to what we call fantasy).
Stanislaw Lem's critique of sf suggests that sf is too detached from cutting edge research in cosmology and physics. He claims sf is not living up to its full potential because it doesn't "take notice" of current cosmological theories; it has in effect "domesticated the universe." Lem thinks that not only should sf adhere to existing science, but also pay close attention to theories defining the cutting edge of current knowledge. With this logic, Lem takes Benford's assertion one step further. He questions why sf hasn't taken notice of such things as the cosmos of colliding galaxies. But why are colliding galaxies important to stories dealing with other subjects, and why should sf need to take notice of this fact at all? Writers cannot possibly encompass all of existing cosmological theory; not even cosmologists can keep up with every new detail in their theories. Lem is himself a well-respected sf author, if he thinks that sf should take notice of such a thing as this, why doesn't he write the story himself? And even if he did this, must someone else write a story about singularities, about which Lem himself says "What heroic characters, what plot can there be where no body, however strong or hard, could exist longer than a few fractions of a second?" (207). But yet Lem's statement implies such things are what should be written about in sf. He also claims that science fiction "looks so shabby when compared to the background of cosmology" (205). Well, it is certain that the true nature of the cosmos is far stranger and more interesting than anything a mere mortal could dream up, but Lem seems to be implying that sf is somehow flawed because it deals with human concerns, while he feels that cosmological theory is far more interesting and important. Lem is, in fact, asking sf to strip away all irrelevant material (read "human concern") and focus on the more interesting aspects of science. It seems to me, though, that scientific journals do this quite well and sf is not necessary for this function.
While Gregory Benford believes that sf should adhere to the rules of existing science and Stanislaw Lem insists that it pay closer attention to current cosmological theories, it is still imperative to tell a story in the most effective manner. Science should be used as a device to aid in the creation of a sf story, not limit its possibilities. Even if an element of the impossible is used, plausible scientific explanations can be used such that, when given a good thump, the story will ring true, not hollow. So, as for playing tennis with the net down, it ain't so bad if you can use your imagination. You might even tell your playing partner that you don't need a net because an electromagnetic field is present instead, and if the ball comes in contact with this it will be sent flying back in his direction. True? Maybe not, but it'll sure make the game a hell of a lot more fun!