Science Fiction Studies

#56 = Volume 19, Part 1 = March 1992



David Ketterer

A Typology of SF

Carl D. Malmgren. Worlds Apart: Narratology of Science Fiction. Bloomington: Indiana University Press, 1991. xii+208. $22.50.

I suspect that the subtitle of Worlds Apart was an afterthought. Even if it were not, the invocation of narratology is misleading. The word "narratology" nowhere appears in Carl Malmgren's text, and none of the names associated with that structuralist offshoot, notably Mieke Bal, Grard Genette, and Tzvetan Todorov, are even mentioned. What Malmgren does provide, in the course of an extended definition of SF, is a typology of SF (and that last phrase would have made a more appropriate subtitle). He is not the first to do so. In New Worlds for Old (1974), I argue for a five-category mapping of the genre: other worlds out of time and space (visionary SF), other worlds in time and space (frequently satiric), and the present world in other terms (philosophical SF, which may be separated into three categories depending upon whether the focus is on a new definition of mankind, a new definition of reality, or some form of outside manipulator). I envision all these categories (all within the larger overlapping context of "apocalyptic" literature, which includes the overlapping categories of science fantasy and other forms of "consequential" fantasy) as overlapping each other. In The Known and the Unknown (1979), Gary K. Wolfe sets up a five-category iconic organization of SF, three images of environment (the spaceship, the city, and the wasteland) and two of humanity (the robot and the monster). A typology or "anatomy of SF" is at least implicit in the intersection of human versus non-human that Mark Rose in Alien Encounters (1981) identifies as the "semantic space" of SF. Most recently, William Sims Bainbridge in Dimensions of Science Fiction(1986) sketches a reader-response typology, the fan perception of three SF constellations: New Wave; a science-fantasy, horror-and-weird, and sword-and-sorcery cluster; and hard science. Although Worlds Apart is distinguished by the extent to which it takes account of, integrates, and builds upon previous SF theory and criticism, Malmgren makes no attempt to compare or relate his own useful and provocative typology to these others. Had he done so, the value of what is, nonetheless, an important book would have been significantly enhanced.

Malmgren maintains that what primarily distinguishes SF is the grounding of its discourse in a scientific epistemology, and the world (rather than the story) that discourse represents. Loosely speaking, all fictional worlds comprise a character, or characters, and one or more settings. Because the characters in SF need not be human (and not because of A.J. Greimas's narratological usage), Malmgren substitutes the term "actants" for characters. In place of settings, Malmgren uses the term "topoi" in a non-narratological sense to mean "not only the settings through which the actants move, and the social order that structures their interactions, but also the implicit time frame (historical, futurological, etc.) of the action and the operative laws that govern the topological domain" (7). On the assumption that a topos is composed of three interdependent systems, Malmgren provides the following diagram of the four interanimating systems that make up any fictional world:

        WORLD                                                  (1) Actants

                                                                       (2) Social Order

                                         Topos                     (3) Topography

                                                                       (4) Natural Laws

What separates SF from other fictional worlds is the introduction of a novelty or innovation into one or more of these systems. In calling the novelty or innovation a "novum," Malmgren is following the Ernst Bloch-inspired terminology of Darko Suvin (the theoretician to whom, up to a point, Malmgren would seem to be most indebted).

What emerges is the typology diagrammed on the facing page and the distinction between SF and science fantasy, the first involving actantial, social order, and/or topographical novums, the latter involving natural-law novums. Malmgren concludes that SF is

characterized by a fictional world whose system of actants and topoi contains at least one factor of estrangement from the basic narrative world of the author, and by a discourse which naturalizes that factor by rooting it in a scientific episteme. The factor of estrangement, or novum, at once defines the genre and determines the range of aesthetic and cognitive functions that it is able to serve. (10)

"Basic narrative world" is Malmgren's term for something like "the world of consensus reality." His definition is essentially Suvinian with the notable exception that, for Malmgren, SF can include supernatural novums provided they are grounded in the appropriate discourse.

The distinction between extrapolative (linear and logical) and speculative (intuitive, quasi-visionary) approaches to the creations of novums corresponds to the distinction between scientific method and analogy, the tropes of metonymy and metaphor, and "the main difference between Verne and Wells, within the grapholect of their time" (15), "grapholect" being a term that Malmgren appropriates from Eric Rabkin's The Fantastic in Literature. However, in the next chapter Malmgren analyzes the paradigm SF text by the speculative Wells, The Time Machine, as an example of extrapolative SF. Malmgren explains: "In comparison to Verne, Wells did rely more on speculative leaps...from a historical perspective. But within any synchronic study of SF, Wells's work must be located within the extrapolative domain" (33). Nevertheless, as we shall see, The Time Machine is a tricky test case for Malmgren's entire system, since from the current scientific viewpoint and hence in terms of the current grapholect (neither being in this respect much different from Wells's), time travel should be considered a novum that violates natural law. If The Time Machine is primarily a time-travel story, it surely belongs to Malmgren's science-fantasy category: it does not qualify as speculative SF and certainly not as extrapolative SF where Malmgren places it. Presumably he has forgotten his distinction between extrapolative metonymy and speculative metaphor when he writes, "The Time Machine's basic novum, the machine itself, acts as the central metaphor in Wells's critique of science" (40). The problem might be skirted by the assertion that devolution, considered as quite distinct from the literal time machine, constitutes the tale's basic novum, but Malmgren does not raise this possibility. I shall return to the explanation Malmgren eventually offers regarding this placement conundrum. In the meantime, the reader is invited to scan the examples in Malmgren's chart of extrapolative and speculative SF. Perhaps the most arguable categorization is that of Frankenstein as predominantly an extrapolative work.

Chapter 2, "SF and the Reader," explains how SF involves the reader in the activities of "concretization" (constructing what Marc Angenot in a 1979 SFS article calls the "absent paradigm") and "interpretation" (which involves the relationship between estrangement and cognition, and the ability to distinguish between noncognitive novums, such as the banal one, and cognitive novums). Following an appropriation of Brian McHale's distinction in Postmodern Fiction (1989) between "zero degree" and "one degree of interpretation" estrangement (in the first case the reader is immediately immersed in the unfamiliar world, in the second at least one "terran actant acts as the reader's ambassador or representative" [31]), Malmgren offers helpful (if a little too familiar) analyses of The Time Machine and Solaris as one-degree-of-interpretation SF, in the first case in the extrapolative mode, in the second in the speculative. Malmgren's essential point is that, since "this first degree of interpretation highlights the process of reading SF in general," some exemplary cases of "worlds with one degree of interpretation become meta-SF, meditations upon, and models for, how to read SF" (31).

It is unfortunate that Malmgren elected to use, and thus give currency to, the unreliable text of The Time Machine edited by Frank D. McConnell (1977). Furthermore, in a footnote Malmgren claims of The Time Machine that "nothing in the text (or the critical canon, for that matter) tends to call into question the validity of the Time Traveler's final reading of the relation between the Morlocks and the Elois" (182, n13). This is not the case. Clearly Malmgren is unaware of David J. Lake's article "Wells's Time Traveler: An Unreliable Narrator" (Extrapolation 22:117-26, Summer 1981) and his revisionary sequel, The Man Who Loved Morlocks (Melbourne, 1981).

In Chapter 3, the heart of Malmgren's book, he fleshes out the SF portion of the theoretical model established in Chapter 1 with analyses of two novels for each of the four types, the first extrapolative and the second speculative. Alien-encounter SF is exemplified by analyses of Robert Silverberg's Dying Inside and Ian Watson's The Martian Inca, alternate-society SF (in which Malmgren claims the barrier image that Gary Wolfe highlights in The Known and the Unknown as a general feature of SF is particularly important) by analyses of Kate Wilhelm's Where Late the Sweet Birds Sang and Joanna Russ's The Female Man, gadget SF (to which Malmgren interestingly links cyberpunk) by analyses of Arthur C. Clarke's Rendezvous with Rama and Arkady and Boris Strugatsky's Roadside Picnic, and alternate-world SF by analyses of Gregory Benford and David Brin's Heart of the Comet and J.G. Ballard's The Drowned World. The ontological questioning of reality that occurs in speculative alternate-world SF, "built as it is on new assumptions about the nature of a world itself" (124), makes it "very much like science fantasy" and, I would add, very like my "present world in other terms" category of SF. This SF/science-fantasy ambiguity once again bedevils Malmgren's system. The most interesting current SF, what is now being called postmodern SF, would appear to fall into the vexed speculative alternate-world category. And although Malmgren does not mention a connection, William Gibson's cyberspace works surely belong as much, if not more, with speculative alternate-world SF as with gadget SF.

Among the detailed analyses of SF texts, those of The Female Man, Rendezvous with Rama, Roadside Picnic, and Heart of the Comet are especially useful. Malmgren's overall argument, however, does become somewhat repetitive. We do not need to be told three times of Clarke's Childhood's End that the Overlords amount to an extrapolative Otherness while the Overmind amounts to a speculative Otherness(17, 57, 60-70). And there are, inevitably, the occasional mistakes. It is presumably because Malmgren has forgotten (or deliberately omitted?) prehistoric SF (and parallel-world SF?) that, early on, he writes: "It happens in SF that narrative motifs or entities which, at the time of their inscription, represent a departure from the author's empirical environment become actualized in a later empirical environment (e.g., submarines, space flight, atomic energy)" (8). Not all SF is set in the future; "a later" should be "another." It is not totally accurate to say that in Algis Budrys's Rogue Moon "those who die are brought back to life" (104), for it is only the doubles of a sacrificed original that survive.

But more importantly, Malmgren makes no mention of the placement of one significant subgenre in his SF typology: the prehistoric tale alluded to above. Perhaps, like Everett F. Bleiler in Science-Fiction: The Early Years (1990), Malmgren classifies the prehistoric story as a form of historical fiction. Malmgren does link SF with the 19th-century application of the "Historic Method" as "the scientific paradigm for all the other social sciences" (5), but the argument for treating the prehistoric tale as SF does not depend on that paradigm. Rather it is because the prehistoric tale involves extrapolation or speculation on the basis of Darwin's scientific theory of evolution that it counts as SF. Although the prehistoric story would appear to embody aspects of all four of Malmgren's SF types, since it frequently involves the confrontation of one or more stages of human and animal development (intelligent dinosaurs, for example), in terms of the dominant novum it should be included with the alien-encounter type.

In Chapter 4, Malmgren elaborates on his earlier definition of science fantasy:

A science fantasy world is predicated on the violation or contravention of five different kinds of "scientific givens": the epistemology of science itself, an accepted scientific theory, an accepted scientific fact, or "natural" actantial possibility. A science fantasy violates the epistemology of science when it presumes that magic is the operative discipline in humanity's relation with the external world; it violates scientific theory when it explicitly ignores basic scientific principles (such as the unidirectionality and irreversibility of time); it reverses a given scientific fact when it presumes, for example, the viability of sentient or humanoid life on Mars, and a historical fact when it posits the existence of alternate time tracks based on such reversals [this last statement is confused if not incoherent; alternate time tracks have to do with scientific theory, not history, and not necessarily with the violation of scientific theory]; finally it violates actantial possibility by introducing a counternatural actant into the system of actants, an entity whose morphology, powers, or existence contravenes scientific possibility. (20-21)

For such fictional worlds to be considered science fantasies rather than straight fantasies, such violations must be viewed as at least potentially explainable, perhaps in terms of a new scientific epistemology. All, however, is not quite as straightforward as Malmgren implies. The scientific theory/ historical fact category confusion that my bracketed insert draws attention to also points to the fact that the four NATURAL LAW categories are not equally cut and dried. The area covered by scientific theory is rather more elastic than the other three areas. Much of today's scientific theory is, in fact, so fantastic that it can hypothetically account for even the most far-out novums. More often than not, then, the scientific theory category has the effect of converting works of science fantasy into speculative SF.

A portion of Chapter 4, notably the analysis of two test cases, Fritz Leiber's Conjure Wife and Stanislaw Lem's The Investigation, will be familiar to readers of SFS, Malmgren's "Towards a Definition of Science Fantasy" having appeared in #46, November 1988. It is in claiming both sublimative and cognitive value for science fantasy that Malmgren differs most radically from Suvin's theoretical position. He is, I believe, right to do so, but it is nevertheless Malmgren's science-fantasy typology that will provoke the most argument. Not only is it often very difficult to distinguish between science fantasy and speculative alternate-world SF, but speculative SF generally tends to segue into science fantasy. What most readers will object to, however, is the placement of alternate-history SF as represented on Malmgren's chart by Philip K. Dick's The Man in the High Castle. Readers have intuitively classified that novel as SF, indeed as a classic of the genre. True, Malmgren does place The Man in the High Castle at the SF end of an alternate-history science-fantasy spectrum, but that does not solve the problem. He does not explain why, in spite of the same branching time-lines assumption in Russ's The Female Man, that book should be classified as speculative SF and Dick's much more extrapolative novel should be classified as science fantasy.

As it happens, Malmgren does suggest a solution to this dilemma in explaining how certain time-travel stories--the key text being The Time Machine--qualify as SF while others (notably stories in which the time-loop novum is dominant) are science fantasies. In the case of Wells's time machine, Malmgren answers (unconvincingly?) "that within the grapholect or writing practice of Wells's time, such a machine did not so clearly violate contemporary scientific possibility," and (more convincingly) "that Wells provided for his machine a (now-unconvincing) scientific rationale" (148). Malmgren goes on to say that

Any SF novum (such as a time machine, FTL travel, and ESP) can become part of the repertoire of SF conventions and therefore a device or tool for other authors. It should be noted that conventionalized novums are indeed devices, that they serve as means to an end, namely the introduction of the dominant or foregrounded novum in the fiction. In other words, the conventionalized novum has in effect lost its status as a novum and now serves simply as a device subtending the "real" or dominant novum (as when an FTL drive is used to stage an alien encounter). In addition...time travel can be engineered in "pure" SF when its scientific rationale accords with the realm of possibility [Malmgren instances Benford's Timescape and Poul Anderson's Tau Zero].... In short, despite the fact that time travel would seem to be the kind of impossibility associated with fantasy, it can be a purely SF motif when it is used as an enabling device or when it is inscribed in a naturalizing and scientific discourse. (148-49)

Malmgren appears not to have noticed that his last clause here subverts his basic distinction and allows for the conversion any science fantasy--including the time-loop stories--into SF! That uncomfortable possibility aside, in addition to being countenanced by the more far-out reaches of scientific theory, the branching time-lines alternate history is the same kind of SF novum as those Malmgren cites above. Although the alternate-history notion is perhaps the dominant novum in The Man in the High Castle, it is surely also a conventionalized and enabling novum. Consequently, Dick's novel, like Russ's, should be classified as SF.

The science-fantasy chapter concludes with a fine analysis of an exemplary work, the "posthistoric world" (160) that Gene Wolfe describes in his Book of the New Sun. As it happens, that analysis is preceded by the most unfortunate typo that I noted in Worlds Apart. I noted it because in occurs in a quote (that Malmgren agrees with) from my 1982 article "Power Fantasy in the `Science Fiction' of Mark Twain":

the intrusion of the fantastic into what appears a science fiction text or a naturalistic text often simply alters the function of the fantastic material. Instead of being encouraged to think about questions of psychology or mortality, the reader is being encouraged to consider matters of epistemology; how do we know what we think we know is accurate? It is the function of epistemology to relate any debate about the "real" and the "unreal" to the relationship between the known and the unknown. (157)

The word "morality" in the text of my article has been mistranscribed as "mortality."

The epistemology issue is taken up in Malmgren's final chapter with the claim that SF "has interrogated the assumptions, procedures, and objectives of the scientific way of knowing the world, an undertaking which grants it a privileged place in modern discourse" (171). SF "is the epistemological genre par excellence" (173). Science fantasy, however, "overlays that epistemological concern with an ontological dominant. That is, it asks readers if the world is indeed everything that is the case, if reality is as monolithic, as given, as science would have it. Its questions include, how many worlds? how ordered? how superordinated? and how can we be sure?" (175).

But can an ontological science fantasy be distinguished from an epistemological SF? As I have indicated, Malmgren himself implicitly deconstructs his distinction between SF and science fantasy. Consequently, his entire science-fantasy typology is somewhat problematic. Simply put, it does not allow for the fact that any fictional world which involves the violation of natural laws can be outer-limit SF if that world is grounded in a scientific or quasi-scientific discourse which dominates or incorporates any competing supernatural or magical discourse. It should be emphasized, on the other hand, that, with some tinkering (and perhaps sleight of hand), Malmgren's SF typology is workable. Even the tinkering, it seems to me, often simply involves applying Malmgren's own insights more rigorously than he does. For the most part, however, his SF typology cogently systemizes, in generally lucid prose, many of the insights of other SF theoreticians. The extent to which Malmgren is able to agree with previous commentators might, in fact, be construed as evidence that the definition of SF has been unnecessarily mystified.

moonbut.gif (4466 bytes) Back to Home