The stories of Cloud Atlas are accessibly written. They are, to use a pejorative term, readable. Each is a complete, albeit interrupted, tale and can be read as a relatively simple plot-driven narrative. A man is slowly poisoned by his deceitful doctor. A composer leaves home and is destroyed by the birth of his own masterwork. A young reporter uncovers a murderous conspiracy. And so on. Even at this level, each story is explicitly something told. It’s a diary, or a set of letters, a memoir, a deposition, or a fireside yarn. The one omniscient third person narrative is carefully situated as a story written within the frame of the novel, and ascribed a fictional author. Even reading it ‘straight’ we are aware that this is a tale told by Hilary V. Hush. The layer of mediation, the teller of the tale, remains. This is always the case, of course.

This is a novel in which telling is a problem. Adam Ewing worries that his diary will be discovered, Frobisher worries too about his letter’s discovery, and a nosy postmaster. Elsewhere Sonmi~451 is introduced to her orison and the rules of discourse are negotiated. Timothy Cavendish inks his memoir in longhand. He disapproves of ‘backflashes, foreshadowings and tricksy devices’ None of these narratives allow the reader simply to bathe in an evoked ‘reality’ since she is constantly reminded of the tales’ fictional tellers.

The stories also exist in explicit relation to popular and historical genres including dystopian science fiction, airport crime novel, the 19th century sea adventure, and so on. These relationships run deeper than glancing reference. They shade somewhere between pastiche and homage. Many specific novels are also referenced as we shall see.

‘Half Lives — The First Luisa Rey Mystery’, is written in the third person in a sharp present tense. The narrator hovers close to a point of view character’s perspective for most scenes, but wavers occasionally, hopping heads, or sitting like a camera on the wrong side of a door about to open. The chapters are short. Characters tend to fall into types: the sassy protagonist, the match-making mother, gruff rough diamonds, and black-hearted villains. The descriptions are terse and efficient.

Wednesday morning is smog-scorched and heat-hammered, like the last hundred mornings and the next fifty.

This is writing with a put-on accent. It owes debts to Hammett and Chandler via Sara Paretsky and James Patterson.

‘An Orison of Sonmi~451′ also references genre sources. Its themes include corporatist government, ecological disaster, cloning, and social control through leisure and entertainment. These explicitly conjure science fiction tropes epitomised by various movies and books including Brave New World by Aldous Huxley, 1984 by George Orwell, Blade Runner (which in turn is based on Do Androids Dream of Electric Sheep by Philip K Dick), Jennifer Government by Max Barry, The Sheep Look Up by John Brunner, Soylent Green (based on Make Room! Make Room! by Harry Harrison) and many more. In particular, of course, the story’s name references Fahrenheit 451 by Ray Bradbury, another dystopian fantasy which portrays a world in which reading has been outlawed and moronic mass entertainment helps control a drugged population. Even a reader with no background in science fiction will recognise a few of these influences, and cannot but read the story with reference to them.

So there is a quoted element to the stories in Cloud Atlas. We must read them with an awareness that these texts are more than simply windows onto a world, or a set of worlds. Each story is constrained by, or at least relates to, sets of literary conventions that we recognise, and know determine the nature of the worlds portrayed.

Cloud Atlas is structured so that each tale wraps another, giving way to, and only concluding after, its successor. Each parent tale wraps its child. Then, within the child story, part of the parent text is found and read. Here is another point at which form is foregrounded, made explicit. Each new reader provides an assessment of the previous text, calling upon us to re-engage with each preceding narrative. So Frobisher is unsure about Adam Ewing’s narrative:

Something shitfy about the journal’s authenticity–seems too structured for a genuine diary, and its language doesn’t ring quite true…

Then, the doubly-fictional Luisa Rey discovers Frobisher’s letters, and is troubled by the “dizzying vividness of the images of places and people that the letters have unlocked.”

Then Timothy Cavendish receives an unsolicited manuscript of Half Lives which, he says “would be a better book if it weren’t so artsily-fartsily Clever”.

‘Nutcase ahoy’, he says when he reads the covering letter by a ‘lady author, one dubiously named Hilary V. Hush’. Both the comic overtones of the name with its implication of pseudonym (and sexual ambiguity) and Cavendish’s doubts underline the constructed nature of the story. The fictional author of a fiction within a fiction, is herself doubted, and even further fictionalised when Cavendish fantasises about her.
And so on…

So the role of these stories as stories is highlighted on at least three levels. Firstly within each story by emphasising the mode of telling. Secondly, in the way that the stories identify themselves as stories in relation to genre conventions. Thirdly, in the way the stories reference one one another as texts.

The referentiality within Cloud Atlas is so layered that it is hard to stay on topic when following any one thread of reference. This is, in part, because the nature of the novel makes internal, as well external, references intertextual. Cavendish’s single cry “Soylent Green is people!” ties in a cannibalism theme, a reference to genre, a reference to a novel, a reference to a movie, and a reference between stories — all on its own. The fact that Cavendish’s tale is ultimately consumed not in its presented form but as a movie also resonates here.

Cloud Atlas emphasises the fact that it exists as a collection of texts by first splitting them up and then joining them back together. Much of the referentiality within Cloud Atlas, therefore is internal. This creates a mechanism of the story, a system, reminiscent of the closed system of Catch 22, and of the figure 8 in Tom McCarthy’s Remainder.

Within the narratives, the obvious points of relation between the stories can be found at the moments at which the characters consume the tales of their predecessors. As we have seen, these moments are often accompanied by doubts or criticisms that call authorship into question. These are moments of transmission. Whatever the truth of the comet birthmark each character hands off both doubt and knowledge to the next. The uneasy resonance of each story for the consumer binds the structure together. At the same time the character’s explicit doubt emphasises artifice. By questioning the provenance of the preceding text, for example, Frobisher, challenges the frame and he ushers in Mitchell the author.

There’s much fun to be had with explicit references between the tales. Ayrs dreams of Somni’s servery– a ‘nighmarish cafe’ where the ‘waitresses all had the same face’. Cavendish, is mis-sold tickets by ‘stem-cell twins’ and is served slop by a ‘sexless automaton’. He also, of course, sings out his line about Soylent Green, and encounters a Sikh taxi driver who says: “Sick teen-squid Zachary”.

This last, apart from being an absolutely appalling pun, is interesting because it illustrates yet another instance of double think or double-focus. On the one hand these references across stories bring the novel together, they support a thesis of interconnectedness. We are tempted to use them to construct a theory, a plausible model of the universe. On the other hand, Mitchell-the-author is clearly winking at the reader over the heads of his characters here (as he does with the title for Luisa Rey’s story: Half Lives and any number of other nudges).

Less flippantly, the central themes of Cloud Atlas appear everywhere–in symbols, asides, tableaux of plot. Everywhere we see power and submission. Alberto Grimaldi, Somni~451, and Morty Dhondt all ruminate on the will to power, as does Doctor Goose in his own way. Kapuka and the Maori teach the same lesson with murder and rape. Nea So Copros subjugates its clones and its ‘downstrata’ citizens.

The powerful feast, sometimes literally, on the weak. “Soylent Green is People!”, wails Timothy Cavendish once again. And of course, it turns out in Somni’s tale that Soap, the food of clones is in fact made of people. Goose, who is found in the opening scene collecting teeth in a ‘cannibal’s banqueting hall’ is a vampire, thinks Frobisher. His first Law of Survival is ‘The Weak are Meat the Strong do Eat’

Despoilation too is everywhere. Western hunters deplete the seal population of Rēkohu 14. The Swannekke B plant threatens ecological disaster, The corpocracy of Nea So Copros is fighting a losing battle against the encroaching badlands, and Zachry lives in a world on the other side of ruin.

And always, there is circularity.

The characters are bound up in a complex and strange machine, one in which the parts don’t quite fit. ‘We cross, crisscross and recross our old tracks like figure skaters’ says Timothy Cavendish. The sense that everything will repeat, or that souls reincarnate pervades. There is a conceit that the story’s protagonists are somehow instances of the same spirit. They are, most obviously bound together by the presence of the comet birthmark, but also by their premonitions of one another’s lives, or maybe of future possible versions of their own lives.

Frobisher reads Nietzche’s Thus Spake Zarathustra, and Ayrs plans a symphony to be named Eternal Recurrence. The minor-league rock musician in Luisa’s story has a guru on his last reincarnation. Stoned and stuck on a roundabout, Cavendish hears a song ‘about how everything that dies someday comes back’, and Frobisher finally comforts Sixsmith with the prospect of rebirth: ‘We do not stay dead long’.

Of course the nature of the circularity in Cloud Atlas is predictably unclear. In one sense the structure of the novel is itself a cycle. It almost ends where it begins, with the strong preying on the weak, and an ineffectual representative of a complicit ‘civilisation’ on the sidelines. But the cycle is also embedded in images of reincarnation, and the replaying of history.

In fact Cloud Atlas leads us almost all the way to grand theory, the ultimate realist alibi of a plausible explainable system. But the novel remains disconnected, and un-join-uppable (to the evident disappointment of some readers). The largest clue to this is Luisa’s story which is fictional within the novel and so sits at an odd angle to any clean theory of transmission. The probable temporal overlap between Luisa and Cavendish is also a problem here, as is the mystical appearance of the comet on Somni’s birthmark-resistant body.

Although Cloud Atlas resists joining up, and even though no single material world can be seen through the lenses of its broken stories, it still offers a remarkably singular moral vision of the world, and in this it’s very much in the realist tradition. There is no doubt about where the good lies in these tales, or the bad, and on whose side we’re meant to be, despite Ewing’s cry that there are ‘as many truths as men’). And in his appeal at the end, we move from a fractured amoral world, to a more comfortable ordered moral universe.

In Cloud Atlas it seems that experimentalist elements win out over the realist. Despite its many hints, it resists explanation, it uses its referentiality to emphasise its own artifice. In their liberal referencing of, even pastiche of, popular genre, the stories distance the reader from any sense of a reflected world. In these senses, Cloud Atlas is unresolvable.

For all its playfulness, though, for all its resistance to unquestioned mimesis, Cloud Atlas presents, and arguably pretends to reflect, an ordered moral universe, and a vision of continuity. The comet-self persists from story to story, and the struggle plays out in the same patterns. While Italo Calvino’s If on a Winter’s Night a Traveller (an acknowledged influence) created an open system, a system of starts without ends, Mitchell smooths things over. He closes the wounds.

Last week I happened on an interesting Guardian blog piece by Jenny Rohn that discussed speculative fiction that ‘gets it wrong’. So I reflex-posted to Facebook (and automatically reposted to Twitter thanks to iftt.com): “When SF gets it wrong. Interesting, but who said speculative fiction is primarily about prediction?“.

Well, quite a few people say that, presumably, and this was duly and cleverly pointed out to me.

But it’s an interesting question. Any story set in a teller’s future may one day be compared to the future itself, and well before that future arrives since a gulf can quickly open up. What? No Internet? Everyone smoking? Really? So there clearly is a measure by which a writer gets it ‘right’ or ‘wrong’.

We can go further. Some SF writers, including Robert Heinlein, Isaac Asimov, and Arthur C Clarke made prediction an important part of their mission and might be judged to some extent for their prescience or lack of it. Modern writers such as Kim Stanley Robinson have taken up this mantle. The Mars Trilogy and 2312 are crucibles in which the social and political implications of climate change, corporate domination, space colonisation and life extension are all intelligently played out. And already the early novels in the sequence reveal some of the kinds of anachronism Rohn enjoys.

The predictive function is a topic within SF narratives too, manifesting most famously in the psychohistory of Asimov’s Foundation novels. Hari Seldon applies a magic formula to future history and leaves guidance messages for generations to come. Until it all begins to go wrong.

But SF has other roles to play. Presumably if we count these up and quantify them, we might arrive at a working definition of the genre? Unlikely, says James Gunn in his introduction to Speculations on Speculation: Theories of Science Fiction.

The difficulty with identifying science fiction–and proceeding from that to definition–is that science fiction isn’t just one thing. It has no recognizable action, like the murder mystery, or recognizable relationship, like the romance. it is about the future–except when it’s about the past or the present… It is best characterized… by an attitude, and even that is hard to define. It is the literature of change, the literature of anticipation, the literature of the human species, the literature of speculation, and more. And because it is the literature change it is continually changing.

Whew. Now I wished I’d just stuck with prediction. This thing’s a MONSTER. Looking at a couple of critical anthologies, I find a daunting scope, both in terms of themes within science fiction and critical approaches to it. There are essays on feminist SF, psychoanalytical approaches, postmodern theory. On marxist, postcolonialist, posthuman and cyborg themes. Essays on queer theory, utopian studies, on virtuality. And that’s not including the subgenres.

So let’s just think a little about strange worlds of the future. Any projected future owes most of its shape to the ground of its production. By which I mean it is like or unlike the writer’s present. Sometimes, maybe more often than not, it’s the interplay of difference and similarity between the time of writing and the portrayed world of the future that forms the project of a science fiction story. By creating a similar but different world, a writer can satirise or advocate or theorise, can warn or urge. The mission is–can be–cultural and political, rather than predictive. And in that sense it is a mission of the present (the present of writing) and not of the future. So reading after the fact, sometimes long after the fact, we accept a work’s cultural context, we find resonance in it, either new or enduring, and relating either to our own time or the moment of production. We don’t condemn it for failing to predict the fall of Soviet communism, for not anticipating Angry Birds or Celebrity Big Brother.

Speaking of Big Brother, when we reread Orwell’s 1984, we don’t judge it a failure for its divergence from the actual 1980s. In fact, we read it as an exploration of a particular kind of ideological totalitarianism, and we find nuances in it we can apply to our own time (the way that opposition is subverted to shore up rather than undermine the dominant ideology). In fact, by setting a date, Orwell’s novel neatly places itself outside the realm of prediction. Its very specificity gives it a pass. (UPDATE: a commenter on Jenny Rohn’s article claims that an interviewee on BBC Radio 4′s Today Programme accused Orwell of ‘getting it wrong’(!))

1984, of course, follows on from Huxley’s Brave New World in its dystopian vision (another vision that still resonates — a wilfully ignorant elite drugs itself while others are left to rot). And we can trace this mode of writing back and back, utopias and dystopias scattered through the ages, until we arrive at Plato’s Republic.

In his essay The Roots of Science Fiction Robert Scholes argued that SF is a ‘special case of romance’:

…this tradition always insists upon a radical discontinuity between its world and the world of ordinary human experience. In its simplest and most ancient form this disconuity is objectified as another world, a different place–Heaven, Hell, Eden, Fairyland, Utopia, the Moon, Atlantis, Lilliput.

The future is just another of these worlds, and even then it may be more Hell (1984, Brave New World) or Heaven (Iain Banks’ Culture) than prediction. Not all SF is satirical or didactic, but any strange world invites comparison with its originating culture. This difference tends towards breach (the brutal meritocracy of The Republic) or exaggeration (Swift’s mad scientists extracting sunbeams from cucumbers).

Even though the main project of a future-set SF novel may not be predictive, it’s stubbornly true that anachronisms are often disconcerting and work against immersion. Although alienation can be an effective literary trick, unintentional alienation often seriously backfires. It’s the difference between a deliberate and an accidental pratfall. Ironically, the more plausible an SF story’s future world appears, the more incongruous any divergence from the actual future can seem.

In fact SF/F faces similar risks when it creates new worlds set in the past. Alternative history (most recently, perhaps CJ Sansom’s Domination) must be meticulously researched so that a divergence from actual history can seem plausible. Time travel novels require similar effort. Connie Willis’s book Black Out about the adventures of some very dull time travellers during the Blitz was criticised by British critics for the accidental inclusion of the Jubilee Line (opened 1979) in the wartime tube network. This kind of error can puncture a story. It’s like seeing a coke can at the Last Supper. Unfortunately fact-checking alone would not have rescued Black Out.

So even if SF is rarely ‘primarily about prediction’, I think Rohn is on to something in her article. We do find certain kinds of predictive inaccuracy disconcerting no matter the mode of the story. Why is it, though, that these ‘errors’ are more problematic than other, glaring, divergences from reality (explosions in space, sounds in vacuum)? I think that might need a separate article.