Science Fiction Criticism

by Rob Latham;

  Hoping eventually to arrive at fundamental hypotheses relevant to a semiotic definition of the genre, I shall begin with some remarks on what distinguishes “realistic” from SF discourse. By way of illustration, I have chosen more or less arbitrarily the following fragments:

  1.“A young lady with a blue merino dress trimmed with three rows of frills came to the doorstep. She introduced Mr. Bovary in the kitchen, where a big fire had been kindled.” (Gustave Flaubert, Madame Bovary [1856], 1: 2)

  2.“Around the swimming pool were strangely shaped chairs made of blentox . . . . Over her driscoll she was wearing an iridescent gown fashioned of vovax and adorned with bernital inlays.” (B.-R. Bruss, Complot Vénus Terre [1963], chapt. 1).2

  Obviously the second fragment belongs to SF. The presence of four “neologisms” enables the reader to come to that conclusion unhesitatingly. But can these made-up words really be considered as neologisms? If not, what are they? How is the sentence intelligible through their opacity and “meaninglessness”? What type of message does the text convey, if it is not totally intelligible? What type of reading pleasure is invested here that is different from the one experienced with Flaubert?

  Underlying these questions is another more basic one. The quotation from Bruss, it is true, exemplifies the most mediocre SF (more complex examples will be discussed in the pages to follow). But with its clumsy attempt at estrangement (“strangely shaped”), it serves for the purpose, of comparison. The question, then, is how to differentiate a “blue merino dress trimmed with three rows of frills” from “an iridescent gown fashioned of vovax and adorned with bernital inlays.”

  A reader not versed in the frivolities of fashion in the first half of the nineteenth century would have some trouble in making any distinction. He might find words like “merino” and “frills” just as queer as “vovax” and “bernital.” But this does not mean that an apprehension of the realistic character of a narrative is totally subjective. It only shows that reading requires a code. Which brings me to the crux of my argument.

  1 Sign/referent/paradigm

  The fact that an SF story is by definition void of referent—e.g., that the planets Gethen, Fomalhaut, Hain, Urras, and Anarres (to mention a few from Ursula Le Guin’s opus) do not exist—this fact does not in any way help us to characterize SF. After all, “Madame Bovary, “ ”Monsieur Homais,” and “Yonville I’Abbaye” are also signs without a referent. The co-presence of such “fictional” signs with others, likewise supposedly deictical or referential, is at the very core of the problem of specifying the meaning of “mimesis” and “credibility” in realistic narratives.

  In contrast to realistic fiction, SF is a conjectural genre in two respects. Its aesthetic goal consists in creating a remote, estranged, and yet intelligible “world.” The narrative about such a world itself requires a conjectural reading. It does not call for the reader to apply the norms, rules, conventions, and so forth of his empirical world, but instead assumes a paradigmatic intelligibility that is both delusive and necessary. The reader, in the act of cognitively coming to terms with the text, shifts from the unfolding (syntagmatic) sequence of the plot to an “elsewhere”—to the semantic paradigms, and hence to the immanent practical or theoretical models, which are supposed to confer meaning on the discourse. From a semiotic point of view, then, SF characteristically is fictional discourse based on intelligible syntagmatic rules which also govern, and are governed by, delusive missing paradigms. This is what I shall try to clarify and demonstrate.

  But before going on, I should stress that the concept of paradigm twinned with syntagm in European functional semiotics, does not carry the meaning usually given to that word in English. In order to clarify this concept as it is understood in Saussurean general linguistics, it might prove useful to illustrate the basic axioms of the “structuralist” school.

  The sign “horse,” for example, does not in itself immediately refer to any empirical being. As a sign, it forms a class in a semantic field; and its extension and comprehension varies according to other complementary classes with which it is associated. We can thus construct the following paradigm:

  In this paradigm, six signs are disposed on several axes with six “semes” (units of signification) in binary opposition: male vs. female, adult vs. non-adult, castrated vs non-castrated. Each sign is meaningful only because it is different from all the others. As a sign, “filly,” while it contains the semic core of “horseness” common to the whole paradigm, is defined in contrast to “horse,” “mare,” and “colt.” It is worth noting that in any paradigm one element can be neutralized, i.e. it can subsume the whole set of oppositions: “horse” may mean /horseness plus male plus adult/, or by neutralization/either male or female, either adult or not/ - /either castrated or not/.

  This amounts to saying that in the linguistic code there are no contents but only contrasts. Each sign is constituted out of a common semantic articulation. Meaning is therefore always “in relation to” other elements. The value of a term depends on the presence or the absence of a contiguous element. The English word “mutton,” for instance, does not have the extension of the French “mouton,” since it is contrasted on a given axis to “sheep,” an articulation which French does not possess.

  “The sign itself (the link between the signifier and the signified) is nothing but the counterpart of all the other signs in language.”3 Therefore—and this axiom is essential to our discussion – paradigmatic and syntagmatic structures imply one another: syntagmatic structures are defined as the rules that combine signs in praesentia in a linear sequence of utterance. These two types of functions (syntagmatic and paradigmatic) are interdependent. The linguistic system which is by definition paradigmatic, exists only through syntagmatic conjunctions.

  If on an unfamiliar beach I notice a yellow flag, a little dirty and faded, and if I identify it as a signal, I do not know offhand its meaning. In order to assign meaning to it, I have to associate the concrete object with an abstract class of “Yellow Flags” (and from then on, the dirtiness or ragginess of the flag becomes irrelevant). I must learn afterwards that the class of Red Flag is contrasted in the paradigm with the classes of Yellow Flags and Green Flags, and that each and every one of them corresponds to a class of messages: “Swimming forbidden” (Red Flag), “Unsupervised swimming” (Yellow Flag), “Swimming supervised and permitted” (Green Flag). I can now make something out of this—namely, that I can swim today, although it might be dangerous since no lifeguards are at their posts. The collection of possible signifieds, the semantic paradigms, determine the extent of possible messages. Nothing in the messages transmitted by these flags (in their “universe of discourse”) will tell me whether the water is warm or cold, deep or shallow. Semiotic phenomena, of course, do not exist in a vacuum. In everyday life they are transmitted together with other symbols, indexes, and clues of a different nature. If huge dark clouds are collecting or if a harsh wind blows, I do not need a red flag (that the wind has anyway probably torn off) to warn me of the danger. If the flag is up, and if everyone is swimming just the same, I may think that the sender of the message was overly cautious and decide to follow the crowd.4

  The reader of SF is in the same situation as the swimmer on an unfamiliar beach. He can only reconstitute the meaning of some signs in the texts if he knows the paradigm, i.e. the full array of classes complementary to a given sign. In any fictional text, some signals are lures that do not refer to an empirical phenomenon. Moreover, in SF the semantic paradigms, hypothesized as being much more complex than in my example of the flags, are always partially missing and therefore delusive since they do not integrally exist in the text.

  My aim, then, is to show how SF functions in its semiotic axioms as a paradigmatic “phantasm,” shifting the reader’s attention from the syntagmatic structure of the text to a delusion which is an important element of the reader’s pleasure. In this regard, SF is a u-topia (no-place) both through its ideological influence and i
n its mode of decipherment: As in a utopia, the reader is transported from a locus or place—the actual syntagmatic sequence to a non-locus or no-place, the paradigmatic “mirage” which presumably regulates the message. I therefore hold that SF is not determined by any direct relationship between its fictional given and the empirical world, but by the relations inherent to the fictional discourse between syntagm and paradigm.

  2 Neologisms and fictive words

  Words created by SF writers to convey a feeling of estrangement are not at all neologisms but “fictive words.” Such words represent one of the most obvious characteristics of the genre. They can be divided into two categories: a) words that supposedly anticipate forms of language from the future or from “parallel” universes; b) words that are supposedly taken from extraterrestrial languages—related to what Myra Barnes has termed exolinguistics.5

  Despite these differences in chimerical status, the semiotic impact of these utterances is much the same. Every “fictive word,” no matter what its “etymology,” will be read in a particular context. Surrounding elements irradiate virtual meanings on the opaque signs. The use of the same fictive word in different contexts enables the reader to form a more and more “accurate” idea of its meaning.

  Hiero had thought he was familiar with many types of leemutes, the Man-rats and Hairy Howlers, the werebears (which were not bears at all), the stimers and several others besides. (Sterling E. Lanier, Hiero’s Journey, chapt. 1).

  Lewis Carroll’s lexical creations may cast light on the similar practice in SF:

  ‘Just the place for a Snark,’ the Bellman cried,

  As he landed his crew with care;

  Supporting each man on the top of the tide

  by a finger entwined in his hair. (The Hunting of the Snark).

  For Carroll, it is not a matter of describing the Snark as an “extralinguistic being,” but of attributing to it a set of predicates intended to produce a stable paradigm. Of course, Carroll is amusing himself at our expense: he launches into a maelstrom of associations which, although they are not totally incompatible, lead nowhere. The SF writer does not lead the reader to a semantic dead end as Carroll does: he leads the reader to believe in the possibility of reconstituting consistent paradigms—whose semantic structures are supposedly homologous to those in the fictive textual “world.”

  It is however the syntagm that actualizes the virtual meaning of the signs. Words are not stable “entities”: only their syntagmatic surroundings imprint a meaning on them. The word “duck.” for instance, is out of context so ambiguous as to be virtually meaningless. In one sentence, it may signify “a swimming bird,” in another “a darling,” in another “an instance of duckling,” or “a durable closely woven cotton fabric,” or “an amphibious truck.”6 Nevertheless, the user has the impression that the meaning of the word precedes its use: syntagmatic consistency leads to a belief in the pre-existence of paradigmatic structures.

  “An iridescent gown fashioned of vovax, and adorned with bernital inlays”: if the reader understands to a certain extent that fragment, he tends to perceive “vovax” and “bernital” as nouns—which in the trivial conception means substitutes for empirical objects. But in that case, he is ignoring the semantic irradiation conveyed by the syntagm. The syntagmatic insertion of the fictive word creates the illusion of a paradigmatic structure, which supposedly corresponds to empirical constructions.

  The coined word may sometimes appear to be a creation ex nihilo. Often it will seem to have an ascertainable etymology—which both permits an attempt at a definition and gives the reader the impression that the word is a sociolinguistic clue, a symptom of the epoch described. Here again, the reader’s attention is shifted from the sign to a diachronic model of lexical creation. Neology implies a number of rules which govern word formation. An example of this can be found in Albert Robida’s Le Vingtième Siècle (1883) where the author invents the “Tele,” short for “telephonoscope”—something later known as “television”. “Tele,” with its apocope (i.e., omission of syllables at the end of the word), is a clue indicating a society in which advanced technology (implied by the Greek prefixes) is part of everyday life (as indicated by the “colloquiality” of the shortened verbal form).

  The reader perceives the contextual meaning of the coined word but he also gets an idea of the type of society he is dealing with through such a word’s semantic formation and use. In the nineteenth century, Greco-Latin derivations gave a text a scientific look. Nowadays, other modes of formation give an impression of “futurism” to the contemporary reader. This is true of portmanteau words, acronyms and heterogeneous compounds regarded by purists as barbaric (e.g. the techno-bureaucratic vocabulary: vertiport, warphan, swimmando, stagflation, transistor, radar, etc.).

  SF has, little by little, created its own vocabulary which is to a large degree used in common by different writers and has penetrated everyday language: android, cyborg, robotics, chronolysis. It is quite possible that words such as airship, aeronautics, cosmonaut, and televison were first employed in fiction and only later entered common usage.

  3 Exolinguistics

  One of the first French space operas is Les Posthumes (The Posthumous Letters) written by Restif de la Bretonne in 1802. In a voyage through the galaxy, the Duke of Multipliandre comes across gigantic fleas, intelligent forms of life on a comet. A flea declares his love to his demoiselle in poignant terms: “Amagilego! syni dllia psouheh sgyngllu bouun evintage”: which means (says Restif) “Lovely Lady! Do grant me the possession of the ravishing charms that I contemplate on the surface of your seductive and provocative body.” (Vol 111, p. 256) Here, early in the history of SF, appears one of the first examples of extraterrestrial languages. This is only one example, since the author also describes the Rondin language of Selenite origin, a language “made of 24 masculine words and 24 feminine, that are sufficient for every circumstance” (Letter 208). The hero is assaulted in these terms by a Martian swamp hippopotamus: “This animal called on Mars Nüsüsümü (with the ü pronounced the French way), jumped on him screaming Mümüarümü,—which means: I want to fuck you.”

  What is relevant and significant in these strange passages is not so much the Martian linguistic “performance” or utterance as the paradigm it presupposes—the absent paradigm—i.e. the whole Martian language, without which no utterance would be possible. The verisimilitude is invested in the presuppositions of the text, in what the text itself implies without attempting to show it extensively. A great many things can be evoked in a text set in an alien world, but only its words can be directly reported. The existence of a language, even if represented only by a few queer “quotations,” implies the existence of intelligent life and a fictive “referential” world. Georges Mounin has suggested that someone ought to make a study of what—for a given language—represents the utmost strangeness: namely (in phonetic terms), the unpronounceable.7 The unpronounceable in a paradoxical element that confers authenticity on even the most mediocre SF.

  Exolinguistics is as old as conjectural fiction. Make-believe languages are to be found in John Mandeville and Thomas More. The most famous examples of these fabricated languages—George Orwell’s Newspeak, Lyon S. De Camp’s Uchronical English (The Wheels of If, 1948), Anthony Burgess’ Nadsat (A Clockwork Orange, 1963), Tolkien’s Hobbit languages, etc.—have been investigated by Myra Barnes, who made their study the subject of her dissertation. By means of fragments “cited” by these authors, she tried to isolate the characteristics of these forged languages and test their coherence. Whether this attempt was linguistically legitimate or not, it illustrates my point: the reader’s cognitive activity, the very possibility of understanding SF, is necessarily interwoven with conjecture. The reader is transported to an illusive “elsewhere” of a semiotic nature, to the paradigms both suggested in and yet absent from the textual message.

  4 From the actual syntagm to the missing paradigm

  Coined words and forged extraterrestrial l
anguages are only two obvious manifestations on a phenomenon peculiar to SF semiotics as a whole. The SF narrative always assumes a “not-said” that regulates the message. The rhetoric of credibility aims at having the reader believe not so much in what is literally said as in what is assumed or presupposed. Emphasis is not placed primarily on the characters and events of the future or extraterrestrial world but on types, models, norms, and institutions that are only summarily and allusively represented by these characters and events.

  In any linguistic performance, a rule is implied, and in any utterance, a semantic paradigm. Any action supposes a model (to which it conforms or not); any category, an inventory; any sample, an overarching structure; any value, an axiology; any group, roles and criteria of conformity. Any events (and narratives of such events) are related to a more global structure.

  While reading SF, one slowly drifts from the narrative sequence as such (syntagm) to these illusory general systems (paradigms). The notion of paradigm must here be enlarged to encompass a full array of semiotic configurations, not limited to linguistic structures. For example: SF characters, as Kingsley Amis has rightly observed in his New Maps of Hell (New York, 1975), are never individuals but representatives of a group or species.

  While the realistic novel should lead the reader to believe in the events it narrates, the SF novel must also have him believe in what it does not and can not show: the complex universe within which such events are supposed to take place. The reader of a realistic novel proceeds from the general (the commonplace, the ideological topos) to the particular (the specific plot governed by this ideological structure). The SF reader follows the reverse path: he induces from the particular some imagined, general rules that prolong the author’s fantasies and confer on them plausibility. The reader engages in a conjectural reconstruction which “materializes” the fictional universe.