SOURCE: Stockwell, Peter. “Language, Knowledge, and the Stylistics of Science Fiction.” In Subjectivity and Literature from the Romantics to the Present Day, edited by Philip Shaw and Peter Stockwell, pp. 101-12. London: Pinter Publishers, 1991.

[In the following essay, Stockwell provides a stylistic analysis of “The Night,” focusing on Bradbury's utilization of language and the story's place within the conventions of science fiction.]

This [essay] will be concerned with language in relation to the genre of Science Fiction (SF), and will deal with the practice of stylistics and the reading of texts. In the spirit of SF itself, the discussion will draw on work from a variety of disciplines which will be brought to bear on a short story ‘The Night’, by the American SF writer, Ray Bradbury.¹ Firstly, however, to facilitate this discussion I would remind the reader of the story of the roadrunner. In those cartoons, first shown in the 1960s but regularly re-run, the basic plot centred around the efforts of a wily coyote to chase and catch a small, fast bird—the roadrunner. The animators of that cartoon imposed rules on themselves in its production: there was to be no dialogue; the roadrunner was to remain on the road; the bird never directly harmed the coyote; and so on. The coyote employed a variety of tools and techniques to catch the roadrunner, but the roadrunner always escaped, and the coyote was always destroyed by his own plan. Each episode that made up the programme was completely independent of the others, so the coyote could miraculously be brought back to life and start again. But the point to emphasize is that the coyote always began again from scratch. Even if only a small part of a particular plan didn't work, instead of learning from the mistake and improving the plan, the coyote always abandoned it completely in favour of something different. He never learned from his mistakes, and his method of proceeding was to suppose that if a plan of attack did not work perfectly the first time, then it was totally useless. The principle that underlies this procedure is that the failure of part of a theory in action leads to the complete abandonment of the whole theory. I will call the tendency in people to do this the ‘coyote complex’.

This rather frivolous example is used here to make two serious points. One is concerned with pragmatics, about which more later. The other is to suggest that what I've jokingly called the ‘coyote complex’ is a pervasive practice in the methods of most disciplines of inquiry into knowledge, but especially in literary studies. In almost any book on literary theory, the author will spend a substantial part of the book orientating him or herself in relation to other theories or ways of reading a text. Previous theories will be demolished on the basis of a contradiction that is characterized as a ‘fatal flaw’ in the theory, leaving the way open for the author's own views to be presented. This practice of theory-abandonment, on the basis of theoretical discussion without any regard for the practical circumstances of application, is so widespread in literary theory as to have attained the status of a consensus methodology. A good, concentrated example can be seen in surveys of literary theory such as Eagleton (1983). The ‘coyote complex’ is the product of a sort of academic purism and a failure to understand the nature of theory.

One of the best discussions of this for the present purposes is an article by the neo-classical economist Milton Friedman (1971), originally written in 1935. Friedman begins by asking whether a theory, hypothesis, or model can be tested by a judgement of the realism of its assumptions. He decides that the assumptions are irrelevant to whether a theory works or not—in other words, whether it has predictive or usable power. Assumptions allow a theory to be abstract and complete—E doesn't equal mc² more or less, or only on Wednesdays, but always and exactly, in theory. But when the theory is applied in practice, it changes its nature. It is up to the observer—economist, scientist, linguist or critic—to assess whether the affecting circumstances in practice have any material relevance or significance in altering the results. A theory can be abstract and ideal because it does not have to take account of circumstances in the real world. Friedman uses an example from Euclidian geometry: a line in theory connects two points in the shortest distance. It is a measure only of length, in one dimension. Of course, an actualized Euclidian line is not possible. The theory assumes a line with no width and no depth. But no matter how fine the pencil or how faint the line, a real line drawn on a surface will have volume, will be three-dimensional. This reality does not invalidate the theory. The idea of a Euclidian line is still useful in physics and mathematics, but actual lines must be thought of differently. This is because the theory is a metaphorical representation of an idea from reality. A theory cannot therefore be right or wrong, only applicable or inapplicable in different specific circumstances. A metaphor such as ‘Juliet is the sun’ (to use the favourite example of the academic literature) cannot be said to be true or false; it is, of course, literally false, but this would be an inappropriate response to the metaphor. Theories are useful, says Friedman, only if the economist (or anyone) remembers the significance of circumstances in the actualization of the theory.

There are three points to emerge out of this. Firstly, that theory is only useful when actualized and the circumstances taken into account. Secondly, that theory is abstract and complete, but the actualization does not have to be, and can still be valid. Thirdly, that some circumstances will vary so far from the assumptions of the theory as to render the theory invalid in those circumstances. For example, the significance of ‘error’ is small when a Euclidian line for a building design is actualized with a fine pencil, but the ‘error’ is significant if actualized by a thick marker.

The inherent assumptions of a model often implicitly specify the circumstances under which it will work—in other words, a model only works within its own frame of reference. For example, in physics Newtonian mechanics was the prevailing theory of motion for almost two centuries. Around the turn of this century, it was discovered that Newtonian mechanics only applies in the circumstances of low velocities and short distances, and that strange things happen in the actual circumstances of velocities approaching the velocity of light, which can only be accounted for by relativity theory. Of course it would be silly to abandon Newtonian mechanics completely on these grounds, since it is still applicable in the frame of reference of our every-day world in which we drop objects and drive cars around and so on. If the ‘coyote complex’ held in physics, then perhaps cars would be designed using relativity theory, which could probably be done but which would be extremely circumlocutionary and redundant. No single frame can be applied to all frames of reference. In literary terms, no reading theory has a monopoly on truth. The critic who does not adapt theory to the practice of analysis will end up with an unrepresentative analysis. Chuck Jones, creator of the roadrunner, said: ‘People who look through keyholes are apt to get the idea that most things are keyhole shaped’ (quoted in Hall 1968: 28).

The method of the following discussion will proceed in accordance with the implications of Friedman's guidelines. The assumptions (or, at least, some of them) underlying this paper will be presented as articles of faith that the reader is asked to accept simply so that the discussion can move beyond them. I hope to justify them, not by theoretical argument, but by showing how they apply to the practice of reading; not proof, therefore, but demonstration of the frame theory in its appropriate frame of reference and circumstances. The first article of faith partly underlies what has been said so far.

ACTUALIZATION OF A MODEL DEMANDS CONTEXT CONSIDERATION

It is this principle that forms the basis of the descriptive model of language functions in De Beaugrande (1980). De Beaugrande sees the virtual (abstract) system of grammar in language as a default system which texts can override in utilization. In other words, a text can construct its own grammar in reading, and can be found meaningful. Text is an actual system which is derived from the virtual system of language. Any discussion of a text in terms of the virtual system is therefore meaningless. De Beaugrande's orientating frame, also followed in this paper, is based on functionalism and pragmatics. Pragmatics here holds both the everyday meaning of being concerned with the circumstances of specific application, as well as being that branch of linguistics which is concerned with utterances (language in a social context, eg Levinson (1983)). Pragmatics avoids the ‘coyote complex’.

De Beaugrande differentiates between features of textuality (which make a text processible by the human brain), and design which is where he locates qualitative judgement of the text. Design criteria include efficiency (greatest returns for least effort) which gives effectiveness (relevance) and appropriateness (of the text in a particular situation). Qualitatively ‘bad’ texts are thus those which are inefficient, ineffective, or inappropriate. All of these are evidently inherently concerned with a specified reader in a specified circumstance. Evidence from the textual aspect can therefore provide a basis for a judgement of value (ie validate) but it cannot on its own assign value (ie evaluate). Value is a contextual feature of texts that is culture- and subject-based.

De Beaugrande also dismisses discussion of texts that look at potential ambiguities as products of the abstract system of langue rather than the actualized system of parole. Although ambiguities occur in the context of real utterances, these are seldom misunderstood because the brain matches and predicts the probable specific meaning in the context of utterance. However, De Beaugrande does not elaborate the importance of probability, and in later articles he even seems to view it as being less significant. The importance of an understanding of probability can be seen when the process of actualization is applied to the theory. The basic traditional model of probability says that in tossing a coin there are two possible outcomes, so each side has a 50 per cent chance of selection—in theory. But in actualization, this theory must change its nature and a sort of ‘bookmaker’ probability will take over (formalized as that branch of mathematics known as Bayesian statistics). In a race with five horses, simple probability would say that in theory each has a 20 per cent (1 in 5) chance of winning. But no bookie would ever give that as a starting price. This is because bookmaker probability—or, more properly, Bayesian statistics—takes account of circumstances, of previous performance, of expectations; it is cumulative and context-dependent. A coin that comes down 90 times heads and 10 times tails still, in theory, has 50 per cent probability either way. But, in practice, the observant gambler would choose heads. The same is true of the reading process, what De Beaugrande calls ‘utilization’. Readings are thus disambiguated not absolutely, but to an acceptable level of probability in order to be processed under a principle of efficiency—maximum processing returns for least processing effort. A sentence ‘means’ not only in isolation, but based on everything that has gone before it in the text, itself interpreted probabilistically. This is how a text generates its own grammar, rules for its own reading. How else could the reader make sense of a sequence like ‘The king was pregnant’? Ungrammatical in the default system, but perfectly grammatical in Le Guin's The Left Hand Of Darkness (1981:89) since the context has been established of the alien winter-world of Gethen, with its single-sex humans capable of becoming male or female for a few days once a month for the purposes of reproduction. So, a second article of faith.

A TEXT-READING IS ACCEPTABLY DETERMINATE BECAUSE OF CUMULATIVE PROBABILITY

The appropriateness of this framing theory for the frame of reading interpretation can be demonstrated by looking at the alternative. If cumulative determination did not hold, then the brain, in the primary stage of reading perception, would have to run through every possible entry of a lexical item in its neural network. For such an apparently simple sentence as the opening of ‘The Night’—‘You are a child in a small town’—I calculate 432 possibilities (without metaphorical connections). Longer, more complex sequences would produce hundreds of thousands of potential alternatives. Clearly the brain has neither the size nor the speed to run through this many possibilities. The brain reads pre-selectively, basing its selection on its accumulated rules in the text.

Cumulative interpretation is qualitative not merely quantitative, depending not only on the frequency of occurrence of previous items but also on a judgement of their significance (foregrounding or weighting). And it holds at every level of interpretation, from word-interpretation as shown above, to the history of genres. Hans Robert Jauss (1982) seems to imply this in seeing the reader as a cultural consensus, in the second thesis in Toward An Aesthetic Of Reception:

The analysis of the literary experience of the reader avoids the threatening pitfalls of psychology if it describes the reception and the influence of a work within the objectifiable system of expectations that arises for each work in the historical moment of its appearance, from a pre-understanding of the genre, [and] … from the form and themes of already familiar works.

(Jauss 1982: 22)

The consequence of the two circumstances of cumulative interpretation process and the specificity of context is that every reading is a new and different one, even a second reading of one book by one reader. Doris Lessing articulates this in Briefing For A Descent Into Hell:

Sometimes when you read a book or story, the words are dead, you struggle to end it or put it down, your attention is distracted. Another time, with exactly the same book or story, it is full of meaning, every sentence or phrase or even word seems to vibrate with messages and ideas, reading is like being pumped full of adrenalin.

(Lessing 1972: 155)

Cumulative probability must work right down to the most basic level of decision-making in the brain: the bit (binary digit), which is defined as the operation needed to make a decision between two equally probable events. This leads to the third article of faith.

FUNCTION AND IDENTITY ARE CO-DEFINABLE

The idea that language is what it does underlies functional and descriptive linguistics. In terms of inquiry into processes, it is the function that is primary, and deeper levels—whether they are called deep structure, langue, competence, universal grammar or whatever—are inferred from the surface utterance. Michael Johnson, in Mind, Language, Machine (1988), has pointed to the co-incidence of function and identity at every stage of interpretation—from the primary code of syntactic and semantic processing in the language centres in the brain, to the micro-code of conceptual-semantic networks in the neural networks, down to the neurochemical code in basic neuroanatomy. It would seem that this regression of codes has no ending, going down to the DNA code and beyond, with the actual uncoded, untransformed real meaning (S) in the deep structure never attained, ever receding infinitely. Johnson ‘explains’ this by a mystical invocation of Derrida, saying the meaning is continually deferred. This seems to me inadequate as an explanation, and evidence for a more appropriate model can be found in the brain and in evolution, and embodied in a fourth article of faith.

ENVIRONMENTAL AND TEXTUAL INTERPRETATION ARE CO-TERMINOUS

That is, they have common boundaries. The language centres in the brain are predominantly Broca's area and Wernicke's area in the left hemisphere, dealing with syntagmatic structure and paradigmatic structure respectively. These areas are adjacent to the visual areas around Exner's centre and the auditory cortex, which analyses sound. Fossil remains show Broca's area (and thus the origin of language capacity), as enlarged about two million years ago, which, in evolutionary terms, is not long enough for these areas to become totally discontinuously specialized. Writing and reading use existing brain structures wholly, beginning only eleven thousand years ago. So the parts of the brain that deal with environmental orientation are the parts that we use to interpret texts. There is a biological analogy here as well: the area in the brain that reconstructs two-dimensional visual data into three-dimensional space-concepts occupies the same position in the right hemisphere as the language area in the left, suggesting that environmental orientation and textual orientation work in the same way.

This analogy indicates that the grammar of neural connections lies in spatial relations. Different qualities of perception, cognition and emotion activate different nerve cells and fibres, all operating at once (in parallel). So, brain activity must be measured as a dynamic map, as spatial relations in progress. Meaning depends on a configuration of the map; a meaning, in other words, is a frozen mapping—an experience at a point in time. This culminates in, and metaphorically supports, the final article of faith.

STYLE IS MEANING

Information is form, materially true in the brain. Grammar is topological.

This is how the transformation of encoded meaning proceeds—not forever deferred in infinite and abstract regression, but into a configuration which can be imaged—in much the same way as a computer can image complex fractal repetition structure from relatively simple equations (cf. Peitgen and Saupe (1988) and Gleick (1988: 90–103)). Perhaps one day there will be the technology to map the status of every brain cell and synapse in this complex process. Simmons (1973), in an article on semantic networks, suggested that:

The meaning of any node is an ordering of the rest of the nodes of network with which it is related. Assuming a richly interconnected network, the complete meaning of any particular node may involve every other node in the system.

(Simmons 1973: 78)

But complexity is even less of a reason for abandoning a theory than the ‘coyote complex’.

The roadrunner cartoon was used at the beginning of this paper, not only to make a serious point, but also to provide a link into a discussion of SF. In the April 1968 edition of Psychology Today, Mary Harrington Hall interviews Chuck Jones, the creator of the roadrunner, and Ray Bradbury, the American SF and fantasy author. In the interview, among other things, they discuss what makes a narrative science fiction. Chuck Jones says:

People think there is one set of rules for every form of literature and another set for fantasy, and that's where most mistakes in analyzation are made. The rules are exactly the same.

(Hall 1968: 29)

It is the difference in emphasis of the discourse rules that makes SF. SF is a form of fantasy, but one of the things that differentiates it from pure fantasy is the degree to which it literalizes metaphors—actualizes them, to use consistent terminology. If you can imagine a story in which Juliet really is the sun, and the sun was a sentient being capable of generating solar and magnetic storms to influence the climate in Verona, that story would be a SF story; indeed, the idea of stars having consciousness appears in Olaf Stapledon's Star Maker (1972). Many of Isaac Asimov's stories take metaphysical problems and actualize them in technological form to work them out. Chuck Jones says:

You must build an entire world that is believable. Everything about this world must ring true, and the facts of the imagination must become as acceptable as the facts of reality.

(Hall 1968: 29)

He talks about imaginative facts being actualized as text-facts. Ray Bradbury adds: ‘What you must do is take one simple, fantastic idea and implement it on every sensual level’ (Hall 1968: 29). On every sensual level in a text, means at every linguistic and cognitive level. In creating belief-worlds of this sort, SF has been variously characterized as ‘possibility fiction’, or ‘context manipulation fiction’, or ‘speculative fiction’. I would suggest ‘epistemic fiction’ as a good descriptive label, conveying the idea of a fiction to do with knowledge, the impact of knowledge on humans and how we deal with it. (I mean here something different from Dolezel's (1976) modal categories, under which SF would be part of an alethic system.) A stylistic analysis of an SF text must therefore take account of the knowledge structures that the reader brings to the text. This can be demonstrated by what Chuck Jones calls an ‘analyzation’ of a text—‘The Night’ by Ray Bradbury.

An exhaustive analysis is precluded here because of considerations of space and because only two interrelated aspects of the text are relevant to the previous discussion. The construction of the text-world and the orientation of the reader towards that world are matters of the epistemic features of the text. I have already said—it is the fourth article of faith—that environmental and textual orientation are almost the same process. This assumption underlies the theory of scripts or frames developed in artificial intelligence. In that field it is mainly text-production that is of interest, but Erving Goffman (1974) has given frame theory a textually descriptive and analytic slant. When people frame an event or text in order to make sense of it, Goffman says, they assign the frame a particular operational status. Two of these are keying and fabrication. Keying is the process by which play is derived from ‘real’ activity; reality and game are distinguishable more as a consequence of their assigned status than by inherent properties. A fabrication is an attempt to transform the frame-belief of someone else—it is intentional deception. These seem to me to correspond to the difference between realism and fantasy. Realism in fiction is marked by a process of keying: minimal fictiveness. Fantasy such as SF is based on fabrication: a marked fictiveness that is significantly different from the reader's knowledge of the real world. Fabrication is achieved by building up a new belief-world, and strategies used to do this in SF generally can be seen at work in ‘The Night’.

The construction of the belief-world of ‘The Night’ involves orientating the reader into a new identity, that of the eight-year-old boy, Doug. The narrator does this by orientational sequences that are outside the narrative progression of events in the story, ie the plot. The first paragraph of the story is the most obvious example of these sequences:

You are a child in a small town. You are, to be exact, eight years old, and it is growing late at night. Late, for you, accustomed to bedding in at nine or nine-thirty; once in a while perhaps begging Mom or Dad to let you stay up later to hear Sam and Henry on that strange radio that is popular in this year of 1927. But most of the time you are in bed and snug at this time of night.

(Bradbury 1976: 154)

This establishes the age and location of Doug, the year, and a particular social environment. Throughout the story, these orientational sequences form the knowledge-base accumulated by the reader. Textually, they are characterized by several features. Firstly, there is much direct propositional content such as ‘You are a child in a small town’, or ‘Skipper is your brother’ (p.155), or ‘There are a million small towns like this all over the world’ (p.160). Secondly, there is a high degree of additional informativity, elaborating on details such as ‘Skipper is your brother. He is your older brother. He's twelve and healthy, red-faced, hawk-nosed, tawny-haired, broad-shouldered for his years, and always running’ (p.155). Thirdly, there is a high occurrence of generic propositions. These are propositions with a proverbial flavour which are presented as timeless truths; the classic example, though ironic, is the opening of Jane Austen's Pride And Prejudice (1813): ‘It is a truth universally acknowledged that a single man in possession of a good fortune, must be in want of a wife’. In ‘The Night’, a sentence like ‘The reedy playing of minor-key violins is the small-towns' music’ is a generic sentence. Finally, these orientational epistemic sequences are characterized by high certainty modality, which is the attitude expressed by the narrator to the subject. This can be seen in many of the modal auxiliaries: ‘You should feel encouraged’ (p.158), ‘You must accept being alone’ (p.159–60)—my emphasis; and also in modal and evaluative adverbs, adjectives and phrases: ‘to be exact’ (p.154), ‘which is natural’ (p.155), ‘certainly’ (p.159).

The certainty of the belief-world thus constructed is then undermined as the story progresses. Uncertainty is expressed by questions in the narration: by low value modality ‘Blackness could come swiftly’ (p. 159), my emphasis; by confusion in collocational clashes such as ‘thick green odour’ (p.157); by the animation of the unknown and inanimate ‘the whole ravine is tensing’ (p.160) and ‘Doubts flush you’ (p.159); and especially by underlexicalization ‘the dark dark dark’ (p.155) and ‘All of the nameless things are there’ (p.158). The certainty of the frame is restored at the end, with the high value modality and directness of the last paragraph (p.162), and the last sentence: ‘It is’. The victory over the unknown, and the reassertion of the timeless certainty of the family united at the end, is what the story is about. It is, then, an epistemic story, a story about knowledge and certainty.

Intertwined with all this is the textual aspect by which the point of view of the narrator is realized. A precise way of examining textual point of view has been developed by Roger Fowler (1986) from work by Boris Uspensky (1973). On the psychological or perceptual plane of point of view,² Fowler delineates four categories: A, B, C and D. Type A is narrative from within the text, as if from within a character's consciousness. Type B is narrative as if the narrator has knowledge of the character's thoughts and feelings and can move around them at will. Type C is external to the characters, as if an objective reportage of events, with minimal interference (typical of Hemingway). Type D is narrative in which the persona of the narrator is highlighted, trying to reconstruct an event from outside (as in Fielding or Kafka). I have summarized these categories A to D as mimetic, omniscient, objective and interpretative narratives respectively.

The problem with ‘The Night’ is that the narration does not seem to fit any of these categories very well. Starting from the beginning, the first two paragraphs would seem to indicate a type C narrative point of view. It is very factual and apparently objective—it seems to evaluate as little as possible; in fact it seems similar to the ‘invisible’ narration of an instruction manual. However, it is not precisely that. The sense of a narrative presence is strong, realized by evaluative modalities (‘to be exact’ and ‘perhaps’), as if there is a narrator constructing a character, building him up by presenting propositions of his external behaviour; so this is type D. But perhaps it is not an external narration at all, neither C nor D. After all, this narrator seems to know the character, internally. The reader has to be informed of his/her fictive role in the narration, providing the narrator knows more about the character than the character/reader himself (‘him’ because it is Doug). Later, for example, the narration runs: ‘He'll be here, you say, knowing very well that he will be’ (p.156). The narrator here has knowledge of Doug's consciousness, and presents Doug's feelings and evaluations throughout the story as he thinks them; so is it a type A narrative? At times—most notably in the orientation sequences already mentioned—the narration seems more like type B: ‘Is there, then, no strength in growing up? no solace in being an adult? no sanctuary in life? no flesh citadel strong enough to withstand the scrabbling assault of midnights?’ (p. 159). The lexical range here is not that of an eight-year-old boy. The narrative is not even wholly felicitous to the point of view of Doug. Occasionally it moves into the mind of his mother: ‘she cannot look anywhere, in this very instant, save into her heart, and there she'll find nothing but uncontrollable repugnance and a will to fear’ (p. 159). In the third person, this is like the omniscient narrator of a type B narrative point of view.

It would seem, then, that if a story as short as ‘The Night’ can contain all of these narrative points of view, then Fowler's categories are not much use. When this first occurred to me, I tried to save the model by rewriting the categories along the lines of the grammatical person in which the narrative was realized. So, between Fowler's type A (first person) mimetic narrative and his type B (third person) omniscient narrative I inserted a third type of narrative point of view in the second person, which I called the instructive or ghost narrative. It is this ghost narrator that appears in ‘The Night’. A ghost or instructive narrative seems to alternate between being internal and external, present and disengaged: present in the deixis such as ‘that strange radio’ and ‘this year of 1927’ (p. 154); present in the internal knowledge of the character; present in assimilating reported speech into the narrative ‘full with “chocolate on top, yes!”’ (p. 154). Yet at the same time it is disengaged in external comments such as ‘to be exact’ (p. 154) and ‘which is natural’ (p. 155); disengaged in shifting to the future tense ‘Later, when you have grown you'll be given names to label them with’ (p. 158). The shift between the two occurs from sentence to sentence and even within sentences (a phenomenon known as slipping): ‘It is a wide ravine that cuts and twists across the town, a jungle by day, a place to let alone at night, Mother has often declared’ (p. 157)—here a narrative description slips into free direct speech. My point is that the level of narrative presence corresponds with the level of uncertainty at various points in the text. The linguistic and orientational cognitive levels of the text all work together in the thematic progression of the narrative.

My reformulation of Fowler's categories seems, therefore, to work in this instance. However, from a general overview, there are not many narratives that would fall into the category of instructive point of view. It can therefore be borne in mind more as an exception to the rule. What I hope I've done is to demonstrate how a theoretical model has to undergo alteration when it is actualized by application to a text. Fowler's categories remain useful, since my necessary modification of them was derived from components of the theory and selected according to the circumstances of the text. In being aware of this, I avoid the ‘coyote complex’ and retain a useful theory of narrative point of view.

To draw to an accumulated conclusion, I will make some outrageously generalized comments about SF, derived from the previous discussion, in the hope that they will provoke interest and further research. I claim ‘The Night’ as archetypal SF, though it might not seem to be. There are no bug-eyed monsters, no flying saucers, no aliens, no technological hardware. It is not set in the future, there is no time travel besides tense shifting. There are not even any impressive field equations. And yet ‘The Night’ can be read as SF, for the following reasons. Firstly, it is SF by association (metonymically SF?) since it is found in a collection of fantastical stories that are more recognizable as SF, written by a famous writer of SF. The reader will probably therefore come to the story with the expectation of reading in the convention of SF. Secondly, SF deals not with the impact of technology on humans, but with the impact of science on humans (where science is ‘knowledge’, from the Latin scientia). Doug's story is about the impact of knowledge. Thirdly, and very generally, many critics have noted the twentieth-century tendency in literature towards internalization. In mainstream fiction, this tendency has often taken the form of psychological exploration where a character usually corresponds with an individual. In the SF genre, the idea of character often corresponds with humanity itself, and so the cultural tendency towards internalization here takes the form of philosophical and epistemological exploration, since philosophy can be said to be the thought of the mind of humanity. ‘The Night’ is therefore typical of SF in that it is an epistemic story; it is about the state of knowledge of typical eight-year-old boys—in ‘a million small towns like this all over the world’ (p. 160)—when faced with the unknown. Even more modern SF, that problematizes the concept and status of knowledge—in the work, for example, of Philip K. Dick or Brian Aldiss—still concerns itself with epistemological exploration.

Finally, the critic Mark Rose (1981) has identified the area that SF explores as being the relationship of the human to the non-human—reinstating concerns of Romantic literature in the dialogue between humanity and nature, and of previous literatures including medieval romance. The human and the non-human is also the subject and the other, the known and the unknown, the finite and the infinite, which is what confronts Doug at the edge of the ravine. SF, as in ‘The Night’, is an attempt to name the infinite; to bring, through language, as much knowledge as possible within human understanding. In other words, the aim of SF is to transform scientia (knowledge) into sapientia (wisdom—which is knowledge in use, applied knowledge). It is this therapeutic, cathartic, broadly political, non-escapist, practical aspect of art that Bradbury refers to when, in the Psychology Today article, he grandly declares:

The so-called realists are trying to drive us insane, and I refuse to be driven insane. I go with Nietzsche who said: ‘We have art that we do not perish in the truth’.

(Hall 1968: 29)

It is with this kind of overblown statement, which is also typical of SF, that this paper should end.

Notes

1. The short story is ‘The Night’, which is stylistically interesting in that it is written in the second person. It is most easily found in the collection The Small Assassin (Bradbury 1976: 154–162). SF has been, from its ‘pulp’ magazine origins, a genre of the popular paperback, and so page references to this story and to the other SF works mentioned in this paper are from the paperback editions of the books.

2. For ease of reference I have here summarized the categories of perceptual point of view in narrative, adapted from Fowler (1986: 127–146). For a new approach to point of view in terms of modality, see Simpson (1990).

   Internal

   Type A: The mimetic narrative, inside a character's mind. Usually in the first person, with high evaluative modality.

   Type B: The omniscient narrative, with knowledge of character's thoughts. Often in the third person.

   External

   Type C: The objective narrative, outside characters' thoughts. Usually in the third person, with no evaluative modality.

   Type D: The interpretative narrative, with persona of narrator highlighted. Typically in the first person.

Bibliography

De Beaugrande, R. (1980), Text, Discourse, and Process: Toward a Multidisciplinary Science of Texts, Norwood (N.J.), Ablex Publishing Corp.

Blakemore, C. (1977), Mechanics of the Mind, Cambridge, Cambridge University Press.

Bradbury, R. (1976), The Small Assassin, St. Albans, Granada.

Dolezel, L. (1976), ‘Narrative Modalities’, Journal of Literary Semantics, 5: 5–14.

Eagleton, T. (1983), Literary Theory: An Introduction, Oxford, Basil Blackwell.

Fowler, R. (1986), Linguistic Criticism, Oxford, Oxford University Press.

Friedman, M. (1971), ‘The Methodology of Positive Economics’, in Breit, W. and Hochman, H. N. (eds) (1971), Readings in Microeconomics (second edition), Hinsdale (Illinois), Dryden Press, 23–47.

Gleick, J. (1988), Chaos: Making a New Science, London, Cardinal.

Goffman, E. (1974), Frame Analysis, New York, Harper & Row.

Le Guin, U. K. (1981), The Left Hand of Darkness, London, Futura.

Hall, M. H. (1968), ‘A Conversation with Ray Bradbury and Chuck Jones: The Fantasy Makers’, Psychology Today, 1(11): 28–37 & 70.

Jauss, H. R. (1982), Toward an Aesthetic of Reception, Brighton, Harvester Press.

Johnson, M. L. (1988), Mind, Language, Machine: Artificial Intelligence in the Post-Structuralist Age, Basingstoke, MacMillan.

Lessing, D. (1972), Briefing for a Descent into Hell, London, Grafton.

Levinson, S. C. (1983), Pragmatics, Cambridge, Cambridge University Press.

Peitgen, H-O and Saupe, D. (eds) (1988), The Science of Fractal Images, New York, Springer-Verlag.

Rose, M. (1981), Alien Encounters: Anatomy of Science Fiction, Cambridge (Mass.), Harvard University Press.

Simmons, R. F. (1973), ‘Semantic Networks: Their Computation and Use for Understanding English Sentences’, in Schank, R. C. and Colby, K. M. (eds) (1973), Computer Models of Thought And Language, San Francisco, Freeman, 63–113.

Simpson, P. W. (1990), ‘Towards a Modal Grammar of Point of View’, Liverpool Papers in Language and Discourse, 3: 40–80.

Stapledon, O. (1972), Star Maker, Harmondsworth, Penguin.

Uspensky, B. (1973), A Poetics of Composition, (trn. Valentina Zavarin and Susan Wittig) Berkeley, University of California Press.