When I wrote the first part of What Rough Beast, I wasn’t altogether sure that a second part would follow. True, the more topics I touched upon the more that seemed to demand my attention, and I certainly hadn’t covered everything I wanted to write about. But at the same time, I assumed interest in the debate would recede, or at least move on, and I didn’t want to flog a dead horse. Moreover, the first part had proven quite time-consuming to write, and I did have other things to do; it seemed perfectly possible that I simply wouldn’t have the chance to write part two.

But then, people like Christopher Priest started asking when the second part would appear. And the ripples from my original essay never quite seem to have gone away.

I have tended to ignore responses of the ‘reviewer heal thyself’ variety (there were quite a lot of them), if only because anything that starts with the assumption that the problem lies entirely in my perception of science fiction doesn’t actually leave any way that I, at least, can take forward the discussion. Nevertheless, there continue to be interesting pieces that pick up on or reference the exhaustion debate. These are just some of those I’ve noticed since the first part of What Rough Beast:

Eleanor Arnason

Brian Baker

James Bloomer

Chris, King of Elfland’s 2nd Cousin

Will Ellwood

Paul McAuley (again)

Tobias Reber

Alastair Reynolds (again)

Adam Roberts at the Guardian

Maureen Kincaid Speller

Dario Tonani, interviewed by Paul Di Filippo

If, in what follows, I pick up on these articles specifically, they are still raising issues that would have emerged during the ever-expanding growth of part 1.


Let me start with what is perhaps the most difficult issue, because it is the one that is hardest to pin down.

Chris, at The King of Elfland’s 2nd Cousin, says, quite correctly, that my argument is based upon a disappointed perception of what science fiction should be. He says that I ‘suggest an indistinct and idealized vision of science fiction’ but ‘leave the purpose of science fiction implicit and unarticulated’.

This makes a lot of sense, and yet it still feels to me as if it is approaching the argument from the wrong end. In other words, Chris wants us to know clearly what science fiction is, and from that solid ground build up an understanding of our disappointments in the genre. To put it rather glibly, my position is the opposite: I know my disappointments, and from that solid ground want to try to build an understanding of what science fiction is. What he sees as being a necessary starting point for the debate is, to my mind, closer to being the end point of the exercise.

I don’t know what science fiction is. Or rather, every time I think I know what it is, it changes, becomes something different. This pursuit of the protean nature of the genre has been at the heart of most of my writing about science fiction. (Since Chris references Damon Knight’s ostensive definition of science fiction, I should perhaps point to my own development of Knight’s position.)

In other words, my original LARB essay, and the various pieces that have ricocheted off from it, was precisely an attempt to work towards the sort of articulation of the purpose of science fiction that he wants.

And yet there has to be some solid ground upon which to stand. The trouble is, when it comes to science fiction the ground cannot be too solid. Every statement I make about the genre is subject to change without notice because the genre itself is in a constant state of flux. That is why I believe any ‘definition’ of science fiction has got to be relativistic, there can be no absolutist set of conditions to which all examples of the genre must conform. I touched on this in the first part of What Rough Beast, for example, when I argued that science has no necessary role in science fiction. The trouble with definitions that are based on the characteristics of the fiction is that for any set of characteristics you care to name, it is all too easy to identify at least one generally recognised work of science fiction that does not share those characteristics. In the past there have been moves to dismiss such anomalies by saying that they are therefore not science fiction, unfortunately there are so many anomalies that we would be left with very very little that might genuinely be called science fiction. And in the process we would lose a lot of works that other people would consider absolutely central to their idea of the genre.

Nevertheless, I approach my readings of science fiction with certain expectations. Over the years, those expectations may have become woollier, but they are still a part of my reading. It was the failure of so many stories to meet those expectations that led to a sense of disappointment that in turn underlay my arguments about the exhaustion of the genre. If I attempt to lay out some of those expectations, you must understand that the list is far from exhaustive and is in no way intended to be taxonomic. I am just trying to explain some of what I expect to get out of a science fiction story. I should also say that there is no hierarchy in this list, none of these expectations takes precedence over any other, and it is certainly not the case that all of them will necessarily be met by any one story. More likely, you get a muddle of responses that vary with each work.

One thing I think is important in my response to science fiction is astonishment. I expect to encounter some thing (or things) that I have not encountered before. (I realise that there is some congruence between this expectation and Darko Suvin’s defining characteristic of the ‘novum’, but I mean something much broader and looser than Suvin does, and I do not intend this as a defining characteristic.) I am looking for surprise, for freshness, for difference, for novelty. I am not looking for comfort, for familiarity, for repetition. A science fiction should not leave you feeling that you’ve read it before, which is not the same as saying that it should not employ familiar tropes. In the last year both M. John Harrison (with Empty Space) and Adam Roberts (with Jack Glass) have written novels that are stuffed with conventional genre tropes, but both have dealt with these tropes in ways that make them seem fresh. My problem with too many of the stories in those Best of the Year collections was that they recapitulated stock genre elements that I had seen too many times before, and did nothing fresh with them. Indeed, several of them tried to make a virtue of repeating what was old and familiar, as if writing a 1950s sf story now was somehow transgressive. There is, of course, a generational aspect to this. It is much easier to be astonished when you are 20 and encountering a trope for the first time than it is when you are 60 and have already encountered hundreds if not thousands of iterations of that particular trope. (Is this what they mean by reviewer heal thyself? Make yourself young and innocent once more?) But I still read novels and stories aplenty every year that do astonish me, that do make me feel like I am encountering the genre again for the very first time (the Harrison and Roberts novels, the Molly Gloss story that Maureen writes about). Indeed, if I didn’t continue to get that astonishment from the genre, there would seem to me to be little point in continuing to read it.

A related but subtly different expectation is challenge. I do not believe that science fiction should be comfortable reading, by which I do not mean that it should invariably be full of tragedy, disaster, unhappy endings or the like. Rather it should make you think, it should challenge your expectations, it should suggest different ways of looking at the world. Anything that questions our assumptions, undermines all the things we do not normally question, is inherently uncomfortable. Any science fiction that makes me feel safe, that tells me yes of course all of my prejudices are perfectly correct, is not doing its job. One of the things that upset me about the stories I was reviewing was the number of them that led me straight to the ending I first thought of.

Related to this (I am talking about an interconnected web of expectations, each feeding into the next; that’s why I specified that these are non-hierarchical) is an expectation of engagement. That is: I like my fiction to be about something. They may be big questions: what it is to be human? what is the nature of our world?; or they may be small: how do you survive the break-up of a relationship? But fiction should not be hermetically sealed off from the world. The purpose of fiction, to my mind, is to tell us things, to make us feel, to test the way we look at the world. Any fiction which is only about other fictions fails to have any point as far as I am concerned.

Built upon all of these expectations is the expectation of conviction. Conviction is a word that loomed large in my original LARB piece, perhaps larger than exhaustion, though here I’m approaching it from a different angle. Conviction is like Coleridge’s willing suspension of disbelief: I want to believe in the story for at least as long as I am reading it. When I begin to read a story, I am making a conscious investment in that story, I am (like, I would suppose, any other reader) working on behalf of the writer, more than half prepared to take what the story tells me and accept it. It actually takes a lot to throw me out of a story, to undermine my trust in the storyteller. But if I begin to detect that the author is not fully committed to the world of their fiction, has not got it fully worked out even in their own head, then my faith in the fiction is rattled. Once the even tenor of my acceptance of the story is disturbed, it is very easy for my disbelief to come back. In those stories I found, again and again, that I could not believe what I was being told. I felt that not only was I not convinced by the story, but all too often I felt that the author was not convinced.

Perhaps tangential to these other expectations is the expectation of quality. If I am going to invest time and trust in a story, then I want that time and trust to be repaid. I want prose that delights me, characters that convince me, a sense that the author has worked hard to turn out the very best piece of work within their capability. If the story appears in an anthology labelled ‘best’, then this expectation is reinforced, because the editor has presumably predetermined that the story in question does indeed represent quality in some degree or other. So when I encounter stories that are structurally weak, poorly written, clearly lacking in some way or other, then I feel short changed, my investment in the fiction is not being rewarded.

Astonishment, challenge, engagement, conviction, quality: these are obviously broad and imprecise terms. I am not talking here about a scientific engagement with the text, a quantifiable set of qualities that can be measured dispassionately. These, and too many other expectations to list here, represent a personal and emotional response to the text. This is art, after all, its effects are supposed to be personal and emotional. But I have been reading long enough, and writing about that reading for long enough, to know what I am looking for in a fiction, and to recognise when I don’t find it. And what generates those effects, what triggers those responses, is something actual and identifiable in the text. In other words, I am not simply saying that these stories did not speak to me; but rather that there was something specific lacking in the stories which meant that they could not speak to me.

Extruded genre product

In a recent TLS review, Stefan Collini quotes a comment about Queenie Leavis’s ‘trick of comparing today’s average with yesterday’s best’. It is an inherently bad way of approaching literature for several reasons. It means relying on some pre-determined notion of quality; it means you miss what might be good in the literature you are decrying; and it also means missing what might be good in the literature you are meant to be extolling. Yet it is something we all do.

One variation on this is to pick one or two works of science fiction and say these are really exciting, therefore the whole of science fiction is exciting. I have seen several iterations of this position in response to my exhaustion article. One book or one author, usually one that is a particular favourite of the writer, is held up as an example of the good things that science fiction is capable of achieving. By implication, therefore, the genre cannot be exhausted.

Well, I can play that game too. In fact I have done. There were several stories in those best of the year volumes that I praised highly. In the Coode Street podcast, the Nerds of a Feather interview and the first part of ‘What Rough Beast’, I picked out novels that I thought were fresh, innovative, exciting. In the last few months I have read exceptionally good novels like Empty Space by M. John Harrison and Jack Glass by Adam Roberts; I have read extraordinary collections of stories by Kathleen Ann Goonan and Kij Johnson; I have read thrillingly original yet hard to define works like Communion Town by Sam Thompson and Hawthorn and Child by Keith Ridgway. These are, without exception, innovative, challenging works that just take the breath away.

So does that mean that obviously, by demonstration, my original claim was wrong?

Of course not, none of these nor any other single work can represent the state of the entire genre. Even if we assume that all the works that have been so cited are really very good (and some of them actually strike me as prime examples of a self-regarding and unadventurous literature), all they tell us is that there are individual examples within the genre that achieve what we hope for science fiction.

For all the books that can be held up as examples of innovation in the genre, how many more are precisely the opposite?

One test of this is to look at how many works of science fiction are deliberately not innovations. Book 5 in the series, the latest exploits of such-and-such a character, author A has revisited the world of author X’s much-loved classic, return to the world of … Before I got thoroughly bored with the game, I used to read through the ‘books delivered’ column in Locus every month and count up the number of stand-alone novels that appeared there. One month there wasn’t a single book that was not part of a series. These are books that are being written and sold on the strength of their familiarity. More of the same, please. More of the same.

Yet there is a cost to this. I suspect that the later Culture novels by Iain M. Banks are selling as well or possibly even better than the earlier one. But their critical reception is in sharp decline even as the books themselves grow ever fatter. It is because there are certain tropes that the avid reader expects to find in a Culture novel. The list will grow longer with each successive volume, so the volume will grow fatter in order to fit them in, until the books become shaped less by the author’s vision and more by the readers’ expectations. Famously, Sir Arthur Conan Doyle tried to kill off Sherlock Holmes for precisely that reason. He had to bring him back, of course, because of the demands of his readers; but can anyone honestly say that the post-Reichenbach Falls stories are better than the pre-Reichenbach Falls stories? Familiarity can be a trap.

There is, of course, nothing inherently wrong with a trilogy or a series. I remain convinced that the four-volume Book of the New Sun is one of the greatest literary achievements in the entire history of the genre (though I am less convinced by the various pendants to the series). Empty Space, which I have praised here and elsewhere, is the third volume of a trilogy. But I am equally convinced that the works that sustain themselves across such length are the exception rather than the rule, but in current sf publishing the trilogy or multi-volume series is the rule rather than the exception.

You have to ask yourself, when Douglas Adams, for instance, announced Book 4 in the Hitch-hiker’s Trilogy, whether that suggests an author who set out on his task confident that he had the material and the interest to sustain his work beyond the first three volumes. A little while ago I reviewed In the Lion’s Mouth by Michael Flynn, the latest volume in his Spiral Arm series. It is competently written (I have enjoyed other books by Flynn) and there’s lots of violence masquerading as story, but when you look at the book dispassionately, nothing happens. No event of any consequence within the world of the novel actually occurs. All he is doing is moving the players around the board so that they are in the right place for the start of the next volume. Flynn’s book is far from alone in this; books are now being written to sustain a series rather than because there is anything actually to be said.

One of the things I am referring to when I talk about exhaustion is precisely this bloat. If you have nothing to say, say it repeatedly, so that the thing is sustained by sheer volume. Look at Blackout/All Clear by Connie Willis, a work deemed so good that it won the Hugo Award, and yet its 1,400 pages contain remarkably little story, while scenes, descriptions, the angsting of the characters, are repeated over and over again. If it takes so long to say so little, all we can assume is that there is little else to say.

Thereof one must remain silent

I suppose one of the things I haven’t done here is lay out exactly what I mean by exhaustion. That is partly because that was what I thought I had been doing all the way along the line. But sometimes it helps to be explicit.

So, I would consider a genre to be exhausted if:

1)      It has nothing to say. As I suggested in the previous section, there is too much science fiction being written today that says nothing at great length.

2)      It repeats itself. So much of what we see presented as science fiction today is a return to characters or situations we have seen before. Or else it is an old form of sf presented in modern clothes.

3)      It talks about itself. There has always been a sense (which is perhaps one of the defining characteristics of genre fiction) that science fiction is in conversation with itself, writers pick up on each other’s ideas and develop them in new and interesting ways. But now sf is its own subject: we get science fiction stories about science fiction. This isn’t so much taking the conversation forwards as the participants in the conversation repeating their own words back again and again.

4)      It disguises itself without fully understanding the disguise it has taken on. A lot of today’s science fiction pretends to be something it is not, it takes on the protective colouration of another genre, usually fantasy, though sometimes crime or horror. And yet there is the sense that, for instance, the devices of fantasy may be incorporated into an sf story without trying to use them as they might work within fantasy; a case of having one’s cake and eating it. I’ll give, as an example, a writer whose work I like and enjoy tremendously, Paul McAuley. I consider The Quiet War to be one of the very best works of science fiction to have appeared in the last decade. It sits comfortably within the most traditional of sf forms (it is a planetary adventure, there’s lots of hard sf technology, there are space battles galore) yet it approaches those tropes in a questioning and innovative way. In other words, it is a very fine example of a genre that is not exhausted. Yet, for me, the third volume, In the Mouth of the Whale, feels tired, feels as if it is picking up the devices of sf but can’t quite work out what to do with them. And part of the problem is that one third of the novel is cast in the language of fantasy, but every time the story starts to follow the logic of the language in which it is told, McAuley has to stop and say, no it’s not fantasy after all, and change the grounds on which the story is operating.

5)      It is incoherent. In the LARB essay I blamed this on the notion of the Singularity, but to be honest I think that is just one among many causes. Another part of the problem may well be the repeating loop of conversation that is the nature of so much sf these days. By incoherent, what I mean is that the stories use familiar sf tropes, often very many of them in the same story, but drained of any external meaning. Bits of sf hand waving or incomprehensible technological marvel or magical transformation (often dignified with the term ‘digital’) will be thrown into the mix simply as a way of highlighting the sf-ness of the story but without having any external reference, any metaphorical or structural link with anything outside the story. Paul Cornell’s story in the Dozois collection, for example, is one of a number he has written set in a world with twists in the fabric of reality that he calls ‘folds’. Fair enough, but when he describes something as being the colour of a fold, what on earth are we to make of that? It is so self-referential, so forgetting of anything outside itself, that it is incoherent.

The Blame Game

James Bloomer asks a very reasonable question: if sf is exhausted, who is to blame?

It is not a question I really want to answer, because I don’t think, in the end, it serves a very useful purpose. But also because I’m not sure there is any easy answer.

The authors are to blame because they write the stories. But they write the stories because editors ask for them, and they have to live. So editors are to blame. But the editors want these stories because they know they will sell, and selling books to the public is their whole raison d’etre. So the reading public is to blame. And so we keep going on, round and round. No-one is to blame, and everyone is to blame.

Yet there are authors who insist on producing work that is different, that challenges the notion of what science fiction can achieve. Yet there are editors who buy these challenging books and stories. Yet there are readers who read them, who seek them out.

Is it a matter of willpower, a readiness to surrender to the easy option?

Honestly, I don’t know. I don’t know the seat of the sickness (if we are to call it a sickness), and I don’t have a remedy for it.

One thing I have noticed, though, often in pieces by people who are ostensibly supporting or extending my original argument, is a repeated reference to ‘gatekeepers’. This seems to imply that there is a wall of some sort between ‘science fiction’ on the one side and ‘readers’ on the other. There are gates in this wall, but someone, a gatekeeper, stands there controlling what goes in or out.

The more I think about this, the more puzzled I am by the notion. Who are the gatekeepers, and what movement is it that they are controlling?

In a sense, I suppose, publishers are gatekeepers. After all, they control what does or does not get into print, and therefore the flow of science fiction that comes out to the reader. But for all the publishers that are producing the sort of works that make sf seem, to me, to be a stultifying inertia, there are others, very often the same people, who are producing the books that I seek out as a sign of the continuing energy and excitement of the genre.

In a sense, critics are the gatekeepers. After all, they tell us what is and isn’t worth reading. But I’m a critic, and I don’t feel like I’m controlling the flow through any gateway.

In a sense, readers are the gatekeepers, or at least the book buyers, the ones whose economic power tells the editors and the publishers and the writers that this sort of thing sells and this sort of thing doesn’t. But this makes a nonsense of the notion of gatekeepers: the person standing between the reader and science fiction is the reader.

No, let’s just accept that somehow, at the moment, science fiction has got itself into a place where it feels trapped, not going anywhere. This doesn’t apply to all writers or all science fictions, but it applies to enough to make the genre as a whole feel dull and restricted, exhausted in other words. And as in the original John Barth essay, exhaustion is not meant as a term of despair, but as a call for people (and that means all of us, writers, editors, publishers, critics, readers, gatekeepers, whatever) to think a little more seriously about what we want science fiction to be. If you want science fiction to be more of the same, what M. John Harrison refers to as ‘the intense commodification of ideas & styles evacuated of their original meaning & impact, an apparently deliberate industrialisation of the commonplace & worn out’, then as far as I am concerned, you’ve got that. If you think science fiction should be innovative, challenging, engaged and engaging, then where do we go?

And that, I think, is it for now. I’ve still not exhausted the thoughts stirred up by this debate, but I suspect there will be no third part. Counting the Nerds of a Feather interview and the two parts of What Rough Beast I’ve done over 17,000 words of commentary on a 5,000 word review. I think that is enough.