Science Fiction Studies

# 17 = Volume 6, Part 1 = March 1979

Darko Suvin

The State of the Art in Science Fiction Theory: Determining and Delimiting the Genre*

... a distinction is drawn by arranging a boundary with separate sides so that a point on one side cannot reach the other side without crossing the boundary . . . . There can be no distinction without motive, and there can be no motive unless contents [on each side of the boundary] are seen to differ in value. -If a content is a value, a name can be taken to indicate this value. -Thus the calling of a name can be identified with the value of the content. -- G. Spencer Brown, Laws of Form (1969)

The problem of any cultural domain as a whole can be envisaged as the problem of limits to that domain. -- Mikhail Bakhtin, Voprosy literatury i èstetiki (1975)

It is often thought that the concept of a literary genre (here science fiction, further SF) can be found directly in the objects investigated, that the scholar in such a genre has no need to turn to literary theory since he/she will find the concepts in the texts themselves. True, the concept of SF is in a way inherent in the literary works — the scholar does not invent it out of whole cloth -but its specific nature and the limits of its use can be grasped only by means of theoretical methods. The concept of SF cannot be extracted intuitively or empirically from the works called thus, as was usually tried until the 1960's and is still often tried today by positivistic critics (especially frequent in Anglo-American criticism, vast stretches of which are therefore unrepresented in the following discussion). In such cases, unfortunately, the concept arrived at is primitive, subjective, and unstable. In order to determine it more pertinently and delimit it more precisely, it is necessary to educe and formulate, (1) positively, its specific domain, and (2) negatively, its relationships with other literary genres and cultural determinants among which it develops.                

The following select list of criticism on the theory of SF is intended to suggest the state of the art in this field and to permit a discussion of its achievements. The list includes items which were available to me by the middle of 1976. In order to concentrate on the crucial questions of a field in rapid development, it has had to be restricted in several ways beyond the usual selectiveness. First of all, it does not encompass general theoretical works on literature and culture useful (even supremely useful) for formulating SF theory, from Aristotle and Curtius to Frye and Goldmann, Bakhtin and Brecht: it encompasses only texts dealing explicitly - at least in part - with SF. This also means that such fundamental works about literary utopia as those by Barthes, Elliott, Frye, Marin, Ruyer or Walsh, which tell us much about SF, are absent from the list.1 Second and perhaps more important, the crucial questions restrictively — but I hope not arbitrarily — chosen for discussion are: (1) What are the necessary and sufficient conditions or characteristics whose presence identifies a fictional story as SF; in particular, this leads to the question: what are the relationships between science and SF? (2) In consequence, what are the limits of SF as literary genre which is to be understood by differentiating it from the mimetic or mundane ("naturalistic") as well as supernatural or metaphysical ("fantastic") genres? These two main questions can be called respectively (1) determination, and (2) delimitation, though the very etymology of these terms shows that they are but the internal and external approaches to the same theoretical problem.                

Third, the above means that this essay will not be dealing with many general aspects of SF, and in particular not with criticism contributing to the (certainly useful) knowledge of the motifs, conventions, and sub-forms into which SF can be subdivided or which go to make it up. Thus some extremely salutary propaedeutic articles do not find a place in the list.2 Fourth, there are as a rule special semantic contexts to discussions of SF theory in the USSR, and some very interesting Soviet texts have been reluctantly omitted from the following list, since their use of terms such as realism, romantic, utopia, fantasy, etc., would necessitate separate discussion for which this is not the place.3 Fifth. my list is not homogeneous either by quality or by degree, and the inclusion of an item does not indicate that I approve or disapprove of it, but simply that it seemed necessary for a full overview of achievements in the field (including some dead ends or negative experiments which were sufficiently consistent and significant). Finally, after a reconsideration of the subject-object relation in scholarship, it has seemed to me that to leave out the Suvin texts would be misleading and therefore less informative than to include them; considerations under my immediately preceding point hold for them too.                

The list is alphabetical, and the entries are as a rule by short title and to the most available edition. A number of items were first published in small magazines, including privately printed fanzines, so that a proper chronology would be difficult to establish; besides, this essay proceeds as a rapid overview of main problem-clusters rather than of genesis (or of the finally irrelevant scholarly "precedence"). Nonetheless, an overall chronological breakdown is possible. Two marginal and more or less symbolically representative items date from before World War 2 (Trotsky and Gove), three still interesting but fast receding ones from the 1950's (Brown, Schwonke, and Heinlein); thus, even allowing for a more rapid obsolescence of the texts before ca. 1960, it is clear that really sustained work on SF theory began to be published in the 60's (9 items: Caillois, "Atheling," Nudelman, Ostrowski, Zgorzelski, Delany, Handke, Klein — all are pioneering but by the same token mostly tentative or partial) and even more clearly in the 70's (24 items, beginning with Lem's book with which one can say that SF theory came of age). The most numerous contributions come from the USA, Poland (some of them in English), and Canada; this proportion in significant theory is rather different from that of SF criticism in general, the bulk of which is, no doubt, published in the USA, then in the USSR, and then in the UK. The reasons for the smaller number of contributions from USA and USSR have already been touched upon; accordingly, in the following list there is a larger number of items from Europe (inclusive of Russia) than from North America.   

If one wanted to do a meticulous genetic survey of the first questions -rather than answers — about SF, it would be necessary to begin with its prominent practitioners, from Poe's notes to Hans Pfaall about the necessity of verisimilitude, through Lasswitz’s writings on "scientific fairy tales"4 and a number of Wells's articles (such as the Preface to his Seven Famous Novels in 1934) to the Soviet writers Beliaev and Dolgushin, as well as Stapledon in Britain and a number of US writers (of which I have retained only Brown and Heinlein). It is interesting to note that out of the list's 38 items 8 are by practicing SF writers (including Klein and Lem), which indicates a very considerable self-consciousness and laudable articulativeness on their part. Rather less euphoric is the state of affairs concerning theoretical questions among the critics of the genre, academic or otherwise, before (say) the mid-60's. As Professor Philmus remarked of three books (and could have remarked of practically the whole period), these early critics "did not examine seriously the meaning of SF or ask whether there may be meanings the genre is particularly and peculiarly qualified to express."5 As late as 1971, the first academic collection of essays on the genre contained only two which could be said to possess serious theoretical interest, and both were by SF writers.6 I have retained Trotsky and Gove from such a prehistory of SF theory as examples of path-breaking discussion — albeit on the margin of SF — of, in the first case, a historical genre (the imaginary voyage), and in the second case, social perception in literature (Russian "cosmist" poetry). Conveniently, they can also serve as examples of academic and non-academic - one might perhaps say of formalist and activist - criticism, two kinds or modes which will run parallel until the 1970's and show some tendency of fusing during these last years.

1. Atheling, William Jr. [James Blish]. "Science-Fantasy and Translation."More Issues at Hand. Chicago, 1970, pp. 98-116 (first publ. 1960 and 1963).

2. Bellemin-Noël, Jean. "Des formes fantastiques aux thèmes fantasmatiques."  Littérature No. 2 (1971), 103-18.

3. Britikov, A[natolii]. "Sovetskaia nauchnaia fantastika," in L. Poliak and V. Kovskii, eds. Zhanrovo-stilevye iskania sovremennoi sovetskoi prozy. Moskva, 1971, pp. 308-50.

4. Brown, Frederic. "Introduction." Angels and Spaceships. London, 1955, pp. 9-13 (first publ. 1954).

5. Caillois, Roger. "De la féerie à la science fiction." Images, images. . . . Paris, 1966, pp. 11-59 (first publ. 1960).

6.Delany, Samuel R. "About Five Thousand One Hundred and Seventy Five Words," in Thomas D. Clareson, ed. SF: The Other Side of Realism. Bowling Green OH, 1971, pp. 130-46 (first publ. 1969).

7. Eizykman, Boris. "On Science Fiction." Science-Fiction Studies 2, No. 2, (1976), 164-66 (untitled original in Les Nouvelles Littéraires 52, No. 2427 [19741, 7).

8. Foht, Ivan. "Slika covjeka i kosmosa." Radio Beograd: Treci program (prolece 1974), 523-60.

9. Gove, Philip Babcock. The Imaginary Voyage in Prose Fiction. New York, 1975 (first publ. 1941).

10. Handke, Ryszard. Polska proza fantastyczno-naukowa. Wroclaw, 1969.

11. Heinlein, Robert A. "Science Fiction: Its Nature, Faults and Virtues," in Basil Davenport, ed. The Science Fiction Novel: Imagination and Social Criticism. Chicago, 1964, pp. 17-63 (first publ. 1959).

12. Hienger, Jörg. Literarische Zukunftsphantastik. Göttingen, 1972.

13. Jameson, Fredric. "Generic Discontinuities in SF," in R.D. Mullen and Darko Suvin, eds. Science-Fiction Studies: Selected Articles on Science Fiction 1973-1975. Boston, 1976, pp. 28-39 (first publ. 1973).

14. Jameson, Fredric. "World Reduction in Le Guin: The Emergence of Utopian Narrative," in R.D. Mullen and Darko Suvin, eds. Science-Fiction Studies: Selected Articles on Science-Fiction 1973-1975. Boston, 1976, pp. 251-60 (first publ. 1975).

15. Ketterer, David. New Worlds for Old: The Apocalyptic Imagination, Science Fiction, and American Literature. New York, 1974 (first publ. 1971 to 1973).

16. Ketterer, David. "Science Fiction and Allied Literature." Science-Fiction Studies 3, No. 1 (1976), 64-75.

17. Klein, Gérard. "Entre le Fantastique et la Science Fiction, Lovecraft." Cahiers de l’Herne No. 12 (1969), 47-74.

18. Lem, Stanislaw. Fantastyka i futurologia. 2 vols. Kraköw, 1970; rev.edn. 1973.

18a. Lem, Stanislaw. "On the Structural Analysis of Science-Fiction," in R.D. Mullen and Darko Suvin, eds. Science-Fiction Studies. . . 1973-1975. Boston 1976, pp. 1-7.

l8b. Lem, Stanislaw. "The Time Travel Story and Related Matters of SF Structuring," in R.D. Mullen and Darko Suvin, eds. Science-Fiction Studies ... 1973-1975. Boston, 1976, pp. 16-27.

19. Lem, Stanislaw. "Philip K. Dick: A Visionary among the Charlatans," in R.D. Mullen and Darko Suvin, eds. Science-Fiction Studies . 1973-1975. Boston, 1976, pp. 210-22 (first publ. 1975).

20. Lem, Stanislaw. "Todorov's Fantastic Theory of Literature." Science Fiction Studies 1, No. 4 (1974), 227-37.

21. Nudelman, Rafail. "An Approach to the Structure of Le Guin's SF," in R.D. Mullen and Darko Suvin, eds. Science-Fiction Studies ... 1973-1975. Boston, 1976, pp. 240-50 (first publ. 1975).

22. Nudel'man, R[afaill. "Fantastika i nauchno-tekhnicheskii progress." Angara38, No. 4 (1968), 62-67.

23. Nudelman, Rafaill. "Conversation in a Railway Compartment." Science Fiction Studies 5, No. 2 (1978), 118-30 (first publ.. 1964).

24. Ostrowski, Witold. "The Fantastic and the Realistic in Literature." Zagadnienia Rodzajów Literackich 19, No. 1 (1966), 54-71.

25. Philmus, Robert M. Into the Unknown. Berkeley and Los Angeles, 1970.

26. Philmus, Robert M. "Science Fiction: From its Beginning to 1870," in Neil Barron, ed. The Anatomy of Wonder. New York, 1976, pp. 3-16.

27. Price, Derek de Solla. "Science Fiction as Science: Why Sci-Fi Zaps," The New Republic (30 Oct., 1976), 40-41.

28. Rabkin, Eric S. The Fantastic in Literature. Princeton, 1976.

29. Russ, Joanna. "Towards an Aesthetic of Science Fiction," in R.D. Mullen and Darko Suvin, eds. Science-Fiction Studies . . . 1973-1975. Boston, 1976, pp. 8-15 (first publ. 1975).

30. Scholes, Robert. Structural Fabulation. South Bend and London, 1975.

31. Schwonke, Martin. Vom Staatsroman zur Science Fiction. Stuttgart, 1957.

32. Suvin, Darko. "On the Poetics of the Science Fiction Genre," in Mark Rose, ed. Science Fiction. Englewood Cliffs NJ, 1976, pp. 57-71 (first publ. 1972).

33. Suvin, Darko. "Science Fiction and the Genological Jungle," Genre 6, No. 3 (1973), 2 S 1-73.

34. A, B, and C [Darko Suvin]. "The Significant Context of SF," in R.D. Mullen and Darko Suvin, eds. Science-Fiction Studies . . . 1973-1975. Boston, 1976, pp. xiii-xix (first publ. 1973).

35. Todorov, Tzvetan. The Fantastic. Cleveland & London, 1973 (Introduction à la littérature fantastique. Paris, 1970).

36. Trotsky, Leon. Literature and Revolution. Ann Arbor, 1960, pp. 210-12 (Literatura i revoliutsiia, Moskva, 1924).

37. Trzynadlowski, Jan. "Próba poetyki science fiction," in K. Budzyk, ed. Z teorii i historii literatury. Warszawa, 1963, pp. 258-80.

38. Zgorzelski, Andrzej. "The Types of a Presented World in Fantastic Literature." Zagadnienia Rodzaiòw Literackich 10, No. 2 (1968), 116-27.

1. How can the domain of SF be determined, on what does it hinge as its theoretical axis? The answering is clouded by the present wave of irrationalism, engendered by the deep structures of the irrational capitalist way of life which has reduced the dominant forms of rationality itself to something narrow, dogmatic, and sterile inasmuch as they are the forms of reasoning of the dominant class. Nonetheless, I do not see any tenable internal determination of SF which would not hinge on the category of the novum (Suvin No. 32, borrowing the term from Ernst Bloch7). A novum or cognitive innovation is an important difference super-added to or infused into the author's empirically "known" — i.e., culturally defined — world (Brown No. 4, Klein No. 17); or, more usefully, it is an important deviation from the authors norm of reality. As a consequence, the essential tension of SF is one between the reader, representing a certain type of Man of our times, and the Unknown introduced by the novum (Nudelman No. 23). The postulated innovation may be of quite different degrees of relevance and magnitude - the latter runs from the minimum of one discrete new "invention" (gadget, technique, relationship) to the maximum of a scene (spatiotemporal locus), agent (main characters), and/or relations basically new and unknown in the author's and the implied reader's environment, which is testified to and identifiable by the SF text's historical semantics, what Rabkin No. 28 calls a "grapholect" marking "the writing 'voice' as coming from a particular time, place and social group." The postulation of the novum is based on and validated by the post-Cartesian and post-Baconian scientific method. It would follow that the opposition between science and the science-fictional hypothesis or innovation is (pace Philmus No. 26) only epiphenomenal and cannot serve to disjoin SF and science — though it can serve as a warning that a proper analysis of SF as a literary genre cannot base itself on its "explanatory scientific content" (italics mine). Indeed, Philmus No. 26 himself very usefully distinguishes naturalistic fiction which does not require scientific explanation, and fantasy which does not allow it, from SF which both requires and allows it.     

If the novum is the necessary condition of SF (differentiating it from “naturalistic" fiction), the validation of the novelty by scientifically methodical cognition into which the reader is inexorably led is the sufficient condition for SF. Though such cognition obviously cannot, in a work of verbal fiction, be empirically tested either in vitro or in vivo - in the laboratory or by observation in nature — it can be methodically developed both against the background of a body of already existing cognitions and as a "mental experiment" following accepted scientific, i.e. cognitive, logic. The final reliance of SF on the basic axioms of articulated and clear cognition leads to the crucial necessity of distinguishing between the "really possible" — that which is possible in the author's reality and/or according to the scientific paradigm accepted in it — and the "ideally possible" (Foht No. 8). Only in "hard" or "near-future" SF does the story's thesis have to conform to a "real possibility"; on the contrary, any SF thesis has to conform to an "ideal possibility" in the sense (as Foht No. 8 again notes) of a conceptual or thinkable possibility the premises of which are not in themselves or in their consequences internally contradictory (as in, e.g., time-travel, or omnipotence and similar metaphysical wish-dreams). It is intrinsically or by definition impossible for SF to acknowledge any meta-physical agency, in the literal sense of an agency going beyond physics (nature), beyond the ideal possibilities of physics or any other science. Whenever it does so it is not SF but a metaphysical or (to translate this from Greek to Latin) a supernatural story. The presence of scientific cognition — not only and not even primarily in the guise of facts or hypotheses but as the manifestation and sign of a method identical to that of the philosophy of science - differentiates thus SF from the "supernatural" genres or fantasy in the wider sense, which include fairy tales, mythical tales, moral allegories, etc., over and above horror or heroic "fantasy" in the narrower sense.

Thus, it is not sufficient to say that the narrative world of SF is "at least somewhat different from our own", and that the difference is "at least apparent against an organized body of knowledge" (Rabkin No. 28): for magic too is an organized body of knowledge, and a fairy tale is often more at variance with the author's empirical world than SF. And of course the one paragraph in Todorov No. 35 is quite insufficient for a serious placing of SF, even within a system that is open to, as grave doubt, as the other systems it so convincingly demolishes (cf. the critiques in Bellemin-NM No. 2 and Lem No. 20)8. With much justice, science has been called the basic world-view of SF (Russ No. 29, Lem No. 18, Nudelman No. 23), its "initiating and dynamizing motivation" (Trzynadlowski No. 37). However, this emphatically does not mean that SF is "scientific fiction" in the literal or crass sense popular between the World Wars and still found in Heinlein No. 11. Indeed, a number of important provisos and elucidations ought immediately to be attached: I'll mention three. First, "world view" (probably "horizon" would be a less ambiguous term) is not identical to ideology. Our world view or conceptual horizon is, willy-nilly, determined by the fact that our existence is based on the application of science(s), and I do not believe we can imaginatively go beyond such a horizon; even a machineless Arcadia is today simply a microcosmos with zero-degree industrialization and a lore standing in for zero-degree science. On the other hand, within such a scientific paradigm and horizon, ideologies can stand anywhere between the full support and full denial of this one and only imaginable state of affairs. Thus, anti-scientific SF is just as much within the scientific horizon (namely a misguided reaction to the repressive - whether capitalistic or bureaucratic - abuse of science) as, say, literary utopia and anti-utopia are both within the perfectibilist horizon. In other words, the so-called "speculative fiction," e.g. Ballard's, clearly began as and has mostly remained an ideological inversion of SF. No doubt, in SF the "locus of credibility . . . must be extended from the scientific rationale to [the significance of] the entire fictive situation — . . . ultimately to the perception . . . of the reality that it displaces, and thereby interprets," as Philmus No. 25 argued; but I would add that the key to the interpretation is exactly the fact of its being conducted within the scientific horizon. Second, "sciences humaines" such as anthropology-ethnology, sociology or linguistics (that is, the mainly non-mathematical sciences) are equally based on scientific methods such as: the necessity and possibility of explicit, coherent, and immanent or non-supernatural explanation of realities — Occam's razor — methodical doubt — hypothesis-construction — falsifiable physical or imaginary (thought) experiments — dialectical causality and statistical probability — progressively more embracing cognitive paradigms — and similar; these "soft" sciences can therefore probably better serve as basis for SF than the "hard" natural sciences; and they have in fact been the basis of all better opuses in SF (partly through the characteristic subterfuge of the science in which hard nature and soft humanities fuse — cybernetics). Third, science is since Marx and Einstein an open-ended corpus of knowledge, and all imaginable new corpuses which do not contravene the philosophical basis of the scientific method in the author's times but are continuous with existing science (e.g., the simulsequentialist physics in Le Guin's The Dispossessed) can play the role of scientific content and validation in SF.

Cognitive novelty as the necessary and sufficient kernel or "conceptual premise" (Lem No. 18B) of any SF tale implies a number of things. Again, I shall mention only three which seem to me to stand out, or to be deducible, from the writings in the list. First, the novelty has to be cognitively explained in each tale or group of tales in concrete (even if imaginary) terms, i.e. in terms of the specific time, place, cosmic and social totality within which it is acting, and especially in terms of its effects on the (overtly or covertly) human relationships upon which it impinges. Only thus can a novum become esthetically valid in a fictional narrative (see already Heinlein No. 11 on a cue from Reginald Bretnor). This means that, in principle, SF has to be judged - in some ways like naturalistic fiction and quite unlike the supernatural genres - by the density, objects, and characters described in the microcosm of the text (Bellemin-Noël No. 2). One could easily set up a Hegelian triad, where the thesis would be naturalistic fiction, which has an empirically validated effect of reality, the antithesis would be supernatural genres, which lack such an effect, and the synthesis would be SF, in which the effect of reality is validated by a cognitive innovation. Obversely, the particular "essential innovation" of any SF tale has in its turn to be judged by how much new insight into imaginary but coherent and this-worldly, i.e. historical relationships it can provide. Defining SF by means of an irreversible and significant change for better or worse in its world and people implies that the simple addition of adventures (see also section 2 below), where plus ça change plus c’est la même chose, is an abuse of SF for purposes of trivial sensationalism and degrades it to clichétized fixed topoi; e.g., Lem No. 18B has shown this for time-travel paradoxes per se (when they are not put to ethico-cognitive uses such as in The Time Machine, I would add). No doubt, the easiest and dominant way of driving a significant change home is to have the hero grow into or with it, and most valid SF uses the device of the "educational novel" with its protagonist who has to understand the novum for himself and for the readers.

Second, an imaginary history each time to be re-imagined afresh in its human significance and values may perhaps - though it does not have to -borrow some narrative patterns from mythography (mythological stories), but the "novelty" of gods validated by unexplained super-sciences at the beck of the Cambridge School's or von Däniken's super-mortals is a pseudo-novelty, old meat rehashed with a new sauce. SF's analogical historicity may or may not be mythomorphic, but it cannot be mythopoetic in any sense except the most trivial one of possessing "a vast sweep" or "a sense of strangeness/wonder" (cf. Suvin No. 33). 1 think the conciliatory solution of Philmus No. 26, postulating among other possibilities a "mythic" or "metamythic" classification of SF with mythomorphic, mythopoetic, and demythologizing subclasses (although a useful device in comparison to the undifferentiated use of the term "myth" in Philmus No. 25), is not clear enough — witness its proliferation of largely incompatible terms. As for Ketterer No. 15, its refusal to come to any grips with the novum and science invalidates the book's thesis theoretically and insofar as it is applied to texts irreducible to eschatological catastrophe (Bellamy, London, Lem, Le Guin). However unclear the postulate of a "mutation of desire" away from the existing class values may still be (Eizykman No. 7), it seems to point in the right direction.9               

Third, the novum can narratively be either a new spatiotemporal locus, or a figure (character) with new powers, or both. Furthermore, since the effect of the innovation is to estrange the implied reader's familiar conventions, and SF thus always reflects back to his world (Nudelman No. 22, Suvin No. 32, No. 33), the new "chronotopes" 10 and the new protagonists will in all significant cases, in direct proportion to the narrative potency of the tale at hand, imply and reinforce each other — as do Wells's Time Traveller and the sequence of his visions, or Le Guin's Shevek, his physics, and the planetary system of Urras-Anarres (see Nudel'man No. 22, Britikov No. 3, Jameson No. 14, Klein No. 17, Nudelman. No. 21). Elucidating such relationships could probably function as a via magistra to a literary analysis of SF which would be neither purely ideological nor purely formalistic.                

Mutation of desire; epistemological functionality equal to new insight into historical human relationships; correlation of chronotope and hero — all of these implications of the cognitive novum as a kernel of any significant SF lead to the conclusion that such SF is in fact a specifically roundabout way of commenting on the author's collective context — often by a surprisingly concrete and sharp-sighted historical comment at that - even where it (sometimes strongly) suggests a flight from that context. The escape is, in all such significant SF, one to a better vantage point from which to comprehend the human relations around the author. It is an escape from constrictive old norms, a device for estrangement, and an at least initial readiness for new norms.

2. The theoretical discussion so far seems to lead to the conclusion that the scientifically validated, although sometimes anti-scientific, novum is, within the admittedly vague limits of fictional literature, the necessary and sufficient condition for an SF tale. If this is so, it becomes easier to heed Spencer Brown's stern epigraph to this survey and delimit SF against other types of writings. No doubt, as Philmus No. 26 rightly reminded us, one should always talk of historical genera as of classes with "identifiable, if not absolutely definite boundaries," but such identification is quite enough for a generalizing theoretical approach.

The first boundaries to be drawn were those most immediately necessary - toward horror fantasy, naturalistic fiction, and fairy tale. There was wide agreement here that both SF and fantasy deal "with things that are not," but fantasy then deals with the subclass of things "that cannot be" and SF with that of things "that can be, that someday may be" (Brown No. 4). As Heinlein put it, "[all fiction] is storytelling about imaginary things and people"; fantasy fiction is "imaginary-and-not possible." Though Caillois No. 5 does not testify to a good acquaintance with overmuch SF, he first suggested some fundamental distinctions in regard to horror fantasy and fairy tale; using them, Klein No. 17 and Lem No. 18 elaborated on the duality (nature vs. super-nature) and black intentionality of the "fantasy" universe towards its figures, in stark opposition to the singleness and the lack of intention in the universe of SF as well as of naturalistic fiction. This interaction of physics and ethics was then in Lem No. 18-18B enriched with a discussion of the mythological tale, and systematized with some additions in Suvin Nos. 32-33. Another interpretation of the by now almost stiflingly canonic trichotomy science-fiction - merveilleux - fantastique in No. 2 brings out some further stimulating points, but suffers from a very narrow empirical basis for its generalizations.                

Other boundaries have been less clearly marked. Very few students have followed the timely warning of Blish No. 1 against hybrids such as "science-fantasy" - in his article exemplified by some works of Merritt, Bradbury or Aldiss, but by now a large pathological growth devaluing much of the field into what Brecht called a branch of the bourgeois dope trade. In "science-fantasy" - as Blish noted - "plausibility is specifically invoked for most of the story, but may be cast aside in patches at the author's whim and according to no visible system or principle," in "a blind and grateful abandonment of the life of the mind." A further warning in the same place that the hybrid of SF and detective tale leads, because of the incompatibility between the detective tale's contract of informative closure with the reader and the manifold surprises inherent in the SF novum system, to a trivial lower common denominator of the resulting tale, has been developed only by Nudelman No. 21. His article has convincingly demonstrated the incompatibility between the plot structures of the cyclical detective tale, whose conclusion returns the universe "to its equilibrium and order," the linear structures of simple additive adventure tale, and the spiral structures of valid SF, whose plot alters the universe of the tale. Further, Suvin in No. 32 has tried to clarify some relationships of SF to imaginary voyage and pastoral, in No. 33 to adventure tale and popular science articles, Lem No. 18-18A to fantastic allegories à la Kafka, Delany No. 6 to reportage, Russ No. 29 and Scholes No. 30 to didactic romance, and Price No. 27 — very interestingly — to the scientific paper and monograph. Nonetheless, much more work remains to be done on these and a number of other genres. The latest survey on "SF and allied literature," Ketterer No. 16, lists, beyond the above, problems of relations to legend, historical fiction, other visionary worlds from Dante to the Romantics, the "natural sublime," the hoax, surrealism, and the nouveau roman.11 All of these juxtapositions are apposite, but I find that the survey's incompatible mixture of genres, aesthetic modes, literary strategies, and movements would have to be supplemented by a poly-parametric classificatory rigor of the Philmus No. 26 kind in order to be theoretically enlightening. Possibly it will be found that all generic contaminations of SF with fairy tale, detective mystery, adventure thriller, and mythological tale happen at the expense of developing the problematics of the novelty specific to SF, as Lem No. 19 argues. But at present, except for the mythological and the detective tale, we must, with some help from the notion of generic discontinuity in Jameson No. 13, return the verdict of not proven. In particular, nobody has dealt with the pressing problem of relationships between SF and "high fantasy" of the Tolkien-Dunsany-Le Guin type. We await further systematic work, for which we have so far seen two kinds of very fruitful approaches developing: the Lemian blend of cognitive, ethical, and sociological analysis, and the Bakhtinian blend of ethical and spatiotemporal analysis. This is provocatively supplemented in Delany No. 6 by an only too brief excursion into the semantic subjunctivity (relation of physics and semantics) differentiating naturalistic fiction, fantasy, and SF in a way that - together with Lem No. 18, Suvin No. 33, and Philmus No. 26 - helps to overcome the clumsy defining of SF (cf. Caillois No. 5 and Heinlein No. 11) simply as naturalistic or believable fiction laid in the future. I close this section with such a reminder that SF is first of all a "word-beast."

3. After all this, I can be mercifully brief on the vexed and by now rather irrelevant subject of a precise definition of SF; all the more easily as a number of points have been elegantly, if sometimes a shade too agnostically, summarized and clarified in Philmus No. 26. He goes through the definitions of Bretnor, Bailey, Amis, Moskowitz, and Aldiss (and could have also gone through Heinlein No. 11 and many others, not excluding Suvin No. 32) in order to show that they are too wide or too narrow, usually both simultaneously. Most importantly, these definitions all use one parameter, e.g. Heinlein's imaginary possibility (only Suvin No. 32 has two parameters, cognition and estrangement). No doubt, the discussions in these and many other items on my list have historically been very useful, in particular the rich and stimulating if often puzzlingly associational rather than systematic insights of Lem No. 18. Yet after Philmus No. 26 it ought to be clear that a pretence at fully explanatory definitions should be restricted to popularizing handbooks, and that on the theoretical level we should focus on discussing the necessary and sufficient conditions for SF which have then in each case to be blended with historico-sociological analysis in order to educe specific realities from the generic potentiality. One ambitious try at an 8-parameter definition in Ostrowski No. 24 has to be mentioned here only to regretfully state that its 8 generating matrices for all fiction (body and consciousness of characters, matter and space of things, action, causation, purpose, time) seem to me invalidated by the underlying confusion between the ontology of empirical life and the ontology of fiction; nor does a second grid of "authors’ attitudes to the fantastic" help to overcome this mechanistic naiveté (which is to a lesser extent present also in Zgorzelski No. 38). To return to Philmus No. 26, it is not necessary to believe that its fourfold meta-classification is any more sacred than Dante's fourfold allegoresis in order to accept it as an important step towards further clarification. Of its four meta-categorizations — topical, structural, modal, and meta-mythic - I've already expressed some reservations about the clarity as well as the justification of the "meta-mythic" one. The "modal" meta-classification (serious vs. satirical key), though unexceptionable, would seem to me properly located in a general theory of literature, in which case it would not have to be repeated separately for SF; with Lem No. 18 I would much prefer to concentrate on the problem-solving or committed vs. game-playing or ludic aspect, mode or pole coexisting in SF works. About the "structural" meta-categorization (the title of which seems confusing to me) I'd say that I do not find the attempt in Philmus No. 26 to supplement the Suvin Nos. 32-33 division of the SF heuristic models into extrapolation and analogy with a third model - nor the attempts in Jameson Nos. 13-14 - convincing. The trouble with the original dichotomy was not that it contained too few but too many models, since it was itself only an insufficiently radical reaction to then unchallenged definition of SF as extrapolation (cf. Caillois No. 5, Heinlein No. 11, Zgorzelski No. 38, and even Delany No. 6); such a definition, to which the title of a critical journal devoted to SF still witnesses, should by now be decently and deeply buried. The problem then is one of differentiation within the concept of analogy: if SF tales are always some kind of analogy, how does the implied reader respond to and deal with such an inverted, reverted, converted, everted, averted, etc. Other to his Self (our conventional world, dramatis personae, chronotopic relationships, etc., which, as Hegel says, are clouded by their very illusory proximity — bekannt but not erkannt)? And what mutations does the scientific horizon bring about within the historical topoi of alternative versions of reality, from Gilgamesh through Cockayne to Lewis Carroll? For one challenging and crucial example, what are its language and semantic specificities (cf. the discussion of grapholects in Rabkin No. 28 and neologisms in Handke No. 10)?

4. No clear conclusion is possible in this survey of a still rapidly developing field of studies, except for two provisional indications. At the present moment, aided by a more generous, non-elitist conception of literature and by the achievements of the genre in the last 40 but especially the last 15 years, SF theory is in full cry and seems to be flushing out some mysteries such as the purposes, limits and devices of the genre. However, a further development of the theory itself has come up against the necessity of integrating socio-historical knowledge into the formally aesthetic and generic one, and diachrony into synchrony (cf. Suvin No. 34). Much can and ought to be learned from the new tools of post-Goldmannian sociology of literature, such as the implicit reader or reception aesthetics, as well as from post-Proppian narrative analysis 12 — not forgetting the old tools of both these approaches. In particular, a crucial investigative locus would seem to be the rise of "lower" or non-canonic genres into a special "para-literary" formation, opposed but also twin to "high literature."13


This article is one of the results of a research project, in which I gratefully acknowledge the financial aid of a Québec Ministry of Education FCAC grant.

1. For a brief account of the place of literary utopia according to SF theory, with a number of further references, see my "SF Theory: Internal and External Delimitation, and Utopia," Extrapolation 19 (Dec. 1977). Both that article and this one arose out of a paper given at the 1976 MLA session. I am grateful to Professor Alexis Aldridge, chairman of the special session on Utopian-Dystopian Literature, for her encouragement, and to Professor Irena Bellert at McGill Univ. for helping me to avoid some logical fallacies.                

2. For example, C.S. Lewis's "On Science Fiction," in Mark Rose ed., Science Fiction (Englewood Cliffs, NJ: Prentice-Hall, 1976), first publ. in 1966; Joanna Russ, "Dream Literature and Science Fiction," Extrapolation, 11 (Dec. 1969).               

3. See the works of Chernysheva, Fainburg, Gromova, Kagarlitskii, Neelov, Smelkov, and Zhuravleva in the annotated checklist of criticism in my Russian Science Fiction 1956-1974: A Bibliography (Elizabethtown, NY: Dragon Press, 1976).               

4. See William B. Fischer, "German Theories of Science Fiction," Science-Fiction Studies, 3 (Nov. 1976).               

5. Robert M. Philmus, "The Shape of Science Fiction," Science-Fiction Studies, 1 (Spring 1973): 41.               

6. I am referring to Thomas D. Clareson, ed., SF: The Other Side of Realism (Bowling Green, OH: Bowling Green University Popular Press, 1971); the two essays are those by Delany (No. 6 on my list) and by Lem (superseded by No. 18 on my list).

7. The novum is a fundamental concept of this greatest philosopher of open horizons and radical possibilities for humanist change in our age, probably best explained in his magnum opus Das Prinzip Hoffinung 1-11 (Frankfurt: Suhrkarnp, 1959).               

8. "It must be noted here that the best science fiction texts are organized analogously [as Gogol’s Nose or Kafka's Metamorphosis, i.e. their events are as real as any other literary event, D.S.]. The initial data are supernatural: robots, extraterrestrial beings, the whole interplanetary context. The narrative movement consists in obliging us to see how close these apparently marvelous elements are to us, to what degree are they present in our life. "The Body," a story by Robert Scheckley [sic, D.S.], begins with the extraordinary operation of grafting an animal's body to a human brain. At the end, it shows us all that the most normal man has in common with the animal. Another story begins with the description of an incredible organization which provides a service for eliminating undesirable persons. When the narrative ends, we realize that such an idea is quite familiar. Here it is the reader who undergoes the process of adaptation: at first confronted with a supernatural event, he ends by acknowledging its 'naturalness' " (No. 35, p. 172). - It is difficult to know where not to begin faulting this skimpy and intellectually infelicitous paragraph. For one thing, in the stories mentioned - though one is cavalierly unnamed - there are clearly no supernatural data (as described, they are barely imaginary); and only inferior SF carries the message that everything is essentially everywhere and always the same as in our empirical normality, i.e. "natural." But then, two stories can hardly tell us much about a genre lasting at least one century in a dozen national literatures; even American SF in the last half a century must be estimated as having produced several thousand books.                

9. I am not discussing in this article Eizykman's turbid and prolix Science-fiction et capitalisme (Paris: Marne, 1974), as I believe all of its relevant points are subsumed in the auto-résumé cited as No. 7.               

10. In his essay "The Forms of Time and Chronotopes in the Novel," in Voprosy literatury i èstetiki (Moskva: Khudozhestvennaia lit., 1975), Bakhtin takes over the term chronotope from Einstein and redefines it as "the essential connection of temporal and spatial relationships, as shaped in literary art." In it, "the characteristics of time are unfolded in space, while space is given meaning and measured by time"; it is therefore a given type of chronotope that determines a literary genre (pp. 234-35, translations mine).                

11. The suggestive article by Ulrich Broich, "Robinsonade und Science Fiction," Anglia, 94, No. 1/2 (1976), reached me too late to do anything more than note that the desert-island tale à la Robinson Crusoe is certainly another form whose relationships to SF deserve clarification.                

12. For a first introduction in English to narratology see New Literary History, 6, No. 2 (1975). Conversely — be it noted — SF theory might be a privileged helper in the study of narrative, since it is necessarily engaged in explicating the notion of a "possible world" implicit in the very fundaments of narratology — cf. the essay by Teun A. van Dijk, ibidem.                

13. For a first approach to paraliterature see Marc Angenot, Le Roman populaire (Montréal: Les Presses de l'Université du Québec, 1975), with a rich bibliography.

moonbut.gif (4466 bytes) Back to Home