First, science fiction might be regarded as any imaginative story which, in contrast to fantasy, might actually be true, today or in the future. This is the implicit basis for occasional critical efforts to extend the history of science fiction back to ancient mythologies and classical literature: since no one in Homer's day knew that giant, one-eyed humans were impossible, the argument would go, readers could regard The Odyssey (c. 750 BC) as a potentially accurate account of conditions beyond their ken, making it a sort of science fiction. Because of the later prominence of space travel in science fiction, however, more attention has been focused on Lucian's True History (c. 160 AD), a satirical account (which its author admits to be untrue) of various extravagant adventures on Earth and in outer space.

However, noting that science fiction might necessarily be regarded as fiction about science, other commentators would insist that science fiction must reflect some awareness of science, so that true science fiction could only come into existence during the Renaissance, when pioneers like Galileo, William Gilbert, and Johannes Kepler helped to establish the scientific method and became the world's first true scientists. By these standards, the earliest works of science fiction might include Kepler's Somnium (1634), which includes a scientifically accurate account of conditions on the Moon; Francis Bacon's New Atlantis (1627), which describes the many scientific inventions of the scientists of a distant utopia; and Francis Godwin's The Man in the Moone (1638), another story about a flight to the Moon. In applying this standard to evaluate later works, Hugo Gernsback insisted that true science fiction had to include scientific terminology and scientific explanations, while stories that lacked these features could be dismissed as "fairy tales." A variant approach along these lines would be to locate in science fiction a characteristic form of perception related to the cold, objective eye of the scientist—which would be one way to describe Darko Suvin's definition of science fiction as "the literature of cognitive estrangement," a formula which has excited generations of critics but has had little impact beyond academic circles.

From a third perspective, science fiction did not emerge until somewhat later—the start of the nineteenth century—at a time when the effects of science first began to have an impact on people's everyday lives, often inspiring a negative or fearful reaction. As most famously argued by author and critic Brian W. Aldiss, then, science fiction could be then characterized as an offshoot of the Gothic novel focusing on the horrific impact of scientific innovations, and Mary Shelley's Frankenstein (1818) could be plausibly regarded as the first science fiction novel, with Edgar Allan Poe serving as another progenitor of the genre. Not simply a vehicle for scientific information and speculation, science fiction would instead be defined as a largely critical response to science, warning against reckless scientists whose work would only lead to monsters and mayhem.

A fourth opinion is that science fiction is best regarded as the literature of a community of writers, publishers, and readers who are united by their common knowledge of, and commitment to, the genre of science fiction. Works of science fiction, then, would be straightforwardly defined as works written by authors who would identify themselves as science fiction writers, works published by publishers who would identify themselves as science fiction publishers, and works read by readers who would identify themselves as science fiction readers. By this definition, science fiction could not exist until the 1920s, when Gernsback first promulgated the term "science fiction" and, through critical commentaries, the publication of readers' letters, and the formation of the first national society devoted to science fiction, began to create the vibrant science fiction community which endures to this day.

As noted, there are no logical grounds for privileging any of these definitions, since all of them continue to have their vocal adherents. But it is helpful to recognize that most debates about whether a given work is "science fiction" or not involve differing perceptions of the genre. There is often some controversy, for example, when an apparent work of science fiction like Margaret Atwood's The Handmaid's Tale (1986) is published as "mainstream" fiction and described by its author as not being part of the genre. Atwood is obviously not denying that her story involves a scientifically plausible future in the manner of other works of science fiction, but she is attempting to avoid any association with the works generated by active members of the science fiction community which, in her view, generally fail to match the standards of genuine literature. Yet science fiction readers might complain that the exclusion of works like The Handmaid's Tale, which clearly fulfill other definitions of science fiction, is unjustifiably contributing to false negative stereotypes about the quality of their texts. Members of the science fiction community might also object strongly to science fiction films that unreasonably portray all scientific progress as basically unwise, conveying the warning against scientific research described by The Invisible Man (1933): "I meddled in things that man must leave alone." These films, they might say, are "anti-science fiction," not genuine science fiction; for while they easily fit the definition of science fiction as a form of Gothic literature, they may lack the sort of scientific knowledge, and almost inevitable respect and admiration for science, demanded by other definitions.

One might attempt to avoid these sorts of arguments with an ameliorative formula, following a suggestion from Damon Knight, that any work described as science fiction can be accepted as a work of science fiction, preventing anyone from taking a work others regard as science fiction and attempting to argue that it is not really science fiction. Yet this stipulation simply transforms debates about whether works are science fiction or not into debates about whether works represent superior forms of science fiction or inferior forms of science fiction. Thus, one might concede that Star Wars (1977) is a sort of science fiction, but since it is largely inattentive to scientific facts and seems instead much like a fantasy, one would proceed to note that it is not as good as films like Blade Runner (1982) which take their science a bit more seriously. As others have claimed, sometimes farcically and sometimes seriously, one little-noted defining characteristic of science fiction is that its works inevitably inspire contentious debates among writers, critics, and readers. There is more than sufficient evidence available to support that view; indeed, a special issue of the scholarly journal Extrapolation was devoted to the topic of "Contentiousness in Science Ficton."