Back to: And another thing ... | Forward to: AFK

The death of genre

This is not the blog entry you are expecting.

Science Fiction literature is unusual in that much of the work within the field exists in constant dialog with other works. Author A writes something; Author B reads it and writes something else by way of an oblique rejoinder. (For example: you won't get all the jokes and references in "Saturn's Children" unless you've read Heinlein's "Friday", to which it is in part a response. Again: Jo Walton's Among Others—on the Hugo shortlist this year, and I'm voting for it—contains numerous references and discussions of the sort of SF/F that a bright, bookish child growing up in the UK in the 1970s would be familiar with: and indeed, it spoke to me, because I was reading many of those works at the same age and time ...)

Authors responding to one another isn't unusual. But in SF/F it's particularly visible. It got started in the pages of the pulp magazines of the 1920s and 1930s and continues today, both in short fiction (we're unusual insofar as we still have a vibrant short fiction ecosystem) and at novel length.

So you probably began reading this blog essay expecting a cunning reference to Elizabeth Bear's essay in Clarkesworld, Dear speculative fiction, I'm glad we had this talk ... or to Abi Sutherland's response in "Making Light", on talking it over. Both well worth reading, I should add: Bear's conceit is that SF is reified into personhood and is of course having one of those annoying mid-life crises, wanting to be taken seriously and consequently dressing in black and reading too much bad goth poetry while hanging out outside the doors of literary award bashes thrown by that cool kid, Mainstream.

Well, that's not what I'm here to talk about. I'm here to talk about something much more concrete: the likelihood that within another decade, two at most, science fiction as a literary genre category may well die.

What, you ask, is the problem?

Well, the process has already begun (indeed, is well under way) in some other media: in film, for example, around 30% of the big budget movies to come out of Hollywood each year are recognizably science fiction. I mean, aliens: that's a pretty obvious signifier, isn't it? And Hollywood feels no need to market these movies as SF; they just are, big budget glossy special-effects beanfests featuring aliens. They're grown-up, quite capable of finding their own audiences. But something is missing upstairs. They're the sixty-foot-tall armoured cyborg idiot children of our genre. All fire and tantrums and no cerebral context whatsoever. There's no internal genre dialog going on, and precious little introspection. (Yes, you can name exceptions like "GATACA"; the fact that you have to note the exceptions is itself a warning sign.)

I am not sure it is possible to write introspective, complex SF as a screen medium. The natural length of a feature movie is around 120 minutes; the traditional movie script runs at one page per minute, with 250 words per page—that buys you, in literary terms, a novella. Add in the expectations of studio executives and the dumbing-down effects of editing by committee you end up with huge pressure to make the script commercial rather than complex. Some director/scriptwriters have the clout to get what they want: but then you end up, as often as now, with George Lucas. Nor is there much scope for a dialog in which directors build on someone else's ideas. So a large chunk of cinematic SF is stuck, spinning its wheels, mistaking ever better special effects and ever bigger first weekend box-office draws for progress.

Written SF harbours a much more complex ecosystem in part because the works are potentially bigger (big enough to encompass big ideas) and in part because it's still, to some extent, ghettoised.


Well, let's look for a moment at the semiotics of book cover design, and what it says about the contents, and the effect it has on what we choose to read.

A book cover is a promotional vehicle intended to achieve two goals:

a) To make a reader who is unfamiliar with the author and/or the book pick the book up in a bookstore (because retail psychology teaches us that customers who handle the produce in a shop are more likely to buy it),

b) To tell the book store staff where in their curated produce display they shold place the item, for best sales impact.

Point (a) eludes many readers (and authors). The cover isn't meant to accurately depict the content of the novel; it's meant to make someone who doesn't know the author of their work handle the product. The fans already know what they want; you could market the book in a brown paper bag and they'd buy it. So the goal is to reach out to the vast majority of potential customers who don't know they're customers yet.

Point (b) is less obvious, and it is a function of the economics of retail. Shops cost money to run. In particular, floor space costs. You have to rent it. And books, physical books, are bulky. I reckon you can cram about 50 paperbacks onto a one meter wide shelf. And you can have as many as ten shelves stacked above one another. You need another eight centimetres in front of the shelves for the poor bewildered customer to stand in, so that's a square metre of retail floorspace that you can't use for any other purpose, eaten by a scant 500 paperbacks, some of which will be duplicates (because the top-selling titles need to be available to multiple customers coming in on the same day). A typical bookstore can only carry single or low-double digit thousands of titles; this is why the long tail effect works so well for Amazon. Regular bookstores have to rely on churn, to attempt to provide a customer who returns every month to buy a couple of books with a fresh selection, to provide the illusion of something wider than the choice dictated by the rent they pay on floor space.

But suppose you're a reader looking for a new novel by your favourite author in a shop with thousands or tens of thousands of titles! You need some sort of indexing system. Consequently, books are filed by category—which in fiction means by genre—and then, hopefully, alphabetically within their category.

The book store clerk, then, has to be able to rapidly identify the category to which a book (coming in one of several cartons, along with hundreds of other books) belongs. And that's where the rocket ship logo on the spine, or the headless woman with a stake (back turned to reveal the tramp stamp) comes in. It tells the store clerk that this is a work of SF, or a work of paranormal romance. Which in turn tells them where to shelve the book.

And this is where our genre ghetto comes from.

Why is it going away?

The answer is both simple and non-obvious: ebooks.

Back in 2007, ebooks accounted for less than 1% of sales of fiction. By the end of 2012 they'll be up to 40%, and they're on course to hit 60% in the next few years—probably by 2015.

Now, the people who write ebooks are the same people who write p-books. The ebook is just an alternative distribution channel, like the mass market paperback or the hardcover or the cuneiform clay tablet. And for the most part, the people who read ebooks with a given type of genre content are the same as those who read p-books in that genre. But there is a key difference: the distribution channel has changed.

You do not buy ebooks in a physical bookstore. You buy them online. It doesn't matter whether you buy them directly from a publisher like O'Reilly & Associates or Baen or a huge retailer like Amazon; it is still an online purchase.

And the curation of soft goods delivered online is fundamentally different from the curation of physical lumps of paper on shelves in rented retail premises. By "curation" I'm talking about how the digital goods—the ebooks—are organized and made accessible to the customers. No longer do we have harried clerks unpacking cartons of stuff and shoveling them onto shelves, looking for visual cues to remind them which particular category the book goes under. Instead, we have tags—metadata identifying the work as being by a given author, part of a series, of interest to readers who want SF, police procedural, near-future, virtual reality, dragons, MMOs. (That's a plausible tag set for "Halting State".) We still have a pictorial cover, but it has to be legible when shrunk to roughly 160 pixels: and this includes the author's name and the title. (Look at your recent book acquisitions. Have you noticed anything about the title length, or the typeface, or the font size on the cover? Have they changed in the past year, relative to five years ago?)

Ebooks are going to be simultaneously easier to find and buy—and much harder.

We are in the position, as readers, of being stranded in an infinitely large bookstore. There are millions of items on the shelves. Many of them are junk, the incoherent ramblings of schizophrenics with hypergraphia, who hitherto merely clogged up publishers' slushpiles but which are now flowed through into virtual print because of the ease of access to the virtual storefront. Many more are generated by spambots: this, for example, or this (don't buy them! They're overpriced rip-offs of my wikipedia entry, assembled and published by web scraping robots). There are translations of books you have already read in foreign languages, often with confusingly changed titles ("Halting State" in Germany is "Du Bist Todt"—a great title, but likely to excite and then disappoint English-reading readers searching for my name on

Getting readers to tag their recent purchases and rate them is a great way of collecting data, and it permits new types of marketing: Amazon's recommendation system is eerily prescient, except when it isn't. (Ask anyone who has bought a book as a present for a friend what it did to their reader recommendations!)

The infinite bookshelf is already a problem for us. To add to the fun, once we enter the world of ebooks, nothing ever goes out of print. So works going back many years or decades are presented with equal priority to the latest new titles.

Upshot: we badly need better curation. Amazon and their competitors could present the results of author searches pre-sorted by time since publication and by language and by series. But that's barely a start.

Genre, in the ebook space, is a ball and chain. It stops you reaching new audiences who might like your work. You are an editor, presented with "Rule 34": do you choose to market it as SF, as crime/police procedural, or as mainstream literary fiction? Wouldn't it be better to market it as all three, with different cover designs and cover blurbs and marketing pitches and reader recommendations and reviews for each bookstore category? We've seen this in microcosm with Harry Potter: the use of adult-friendly covers allowed parents to buy the books and read them during their commute to work, for example.

On paper, that's very expensive/hard to organize: in electronic media it is simply a matter of commissioning as many cover designs as your book design budget will stretch to, and then convincing the big retailers to associate a different cover image with the results of each search by genre category.

We already see ebooks being tagged as multiple categories. It's only a matter of time before publishers and authors develop more sophisticated electronic marketing strategies that either micro-target a specific audience, or that target multiple readerships in parallel.

(Length is also a ball and chain. I've previously blogged about how the length of a story may be dictated by physical printing constraints—the cost of binding a big fat book turns big fat books into the domain of best-sellers (at least in the USA), while the exigencies of selling mass market paperbacks to fill supermarket wire racks during the 1970s and 1980s forced publishes to increase their page counts (to justify price increases during a period of inflation). In electronic form there's nothing stopping us from selling novelettes and million-word blockbusters on an equal logistical footing.)

But anyway, to summarize: my point is that our genre sits uneasily within boundaries delineated by the machinery of sales. And that creaking steam-age machinery is currently in the process of being swapped out for some kind of irridescent, gleaming post-modern intrusion from the planet internet. New marketing strategies become possible, indeed, become essential. And the utility of the old signifiers—the rocket ship logo on the spine of the paperback—diminish in the face of the new (tagging, reader recommendations, "if you liked X you'll love Y" cross-product correlations by sales engines, custom genre-specific cover illustrations, and so on).

This is going to drastically affect the quality and content of the internal dialog within our genre—the subject matter of the imaginary conversations by Elizabeth Bear and Abi Sutherland to which I linked up top.

I don't know what it means, yet. Jo Walton opined that the conversation will go on, and I'm sure it will ... among those of us who are already aware of it.

But there was so much less SF in the 1970s that it was quite possible for those of us who grew up reading the field back then to acquire a comprehensive coverage of it. Today, there's far more stuff out there: but without the clear signifiers, the tags saying "queue here to join the ongoing conversation", it may become increasingly hard for new readers to recognize what's going on and join in.

What is to be done?



Maybe there is nothing to be done. Maybe its just a matter of 'keep writing good books' ... and the rest will work itself out. Perhaps its not essential that that people are widely read in any particular genre.

Certainly, as someone deeply schooled in that 1970s SF genre, I no longer read exclusively SF, hell, or even fiction in general (due to an academic research career which requires deep specialisation in a particular canon). As a consequence, I can only read two or three SF titles a year. I will never keep up with the genre. Does it matter?


Devil's advocate: is this a bad thing? One thing I don't like about many science fiction discussions is the insularity of the tastes, and this especially goes for books "in conversation with each other". But it's always the same conversation.

Do you remember those early book recommendation programs? "If you like X, you might like Y?" And how the same names would come up over and over?

"We've been in prison so long, we all know the same jokes. So we yell out the number instead of saying the whole joke."

And when the die-hard science fiction reader timidly steps outside the bounds of the genre, it's to a very narrow range of genre-like books: Patrick O'Brian, for instance, but not Andrea Barrett.

The dissolution of the genre's boundaries should help reduce that narrow-mindedness, and revitalize the conversation -- and hopefully, revitalize the genre, which is dying for other reasons too.


I'll have to chew on that for a while. Interesting analysis.

As to book covers: do people who buy e-books even bother looking at them? They're notoriously information-poor and (at least with some publishers) the cover art has precious little to do with the story. The back blurb is replicated on the storefront page and supplemented with tags &c, so no need for that either. IMO about all the cover art is good for is distinguishing a professionally-published book from slush and rip-offs.


About ten years ago, Amazon's recommendations were brilliant. Now they annoyingly recommend alternate editions of books I already own - and pointedly, bought from Amazon. If I say I'm not interested in Vol 1 of someone's new Paranormal Romance Military saga, then it will recommend Vol 2, 3 etc until I want to tear my eyes out lest Vol XXXIX Doc Savage's Pony Brotherhood vs Perry Rhodan's Unicorn Brigade be offered up to me.


We need more SF reviews by good reviewers, ie NOT capsule reviews or those 3 line, miniblurb, consumer-aid jobbies. If Amazon were less - how to put it - commercially jackbootesque (and simultaneously negligent towards what readers would genuinely find helpful), you might have a situation where an Amazon entry would have links to real, actual propah reviews which were informative.

Ah, but ... a bad book might not get the kind of reviews that would maximise Amazon's turnover...huh, problem there.

But as to your proposition, the death of genre, eh, I wouldnt say so. SF the genre has won the War Against The Mainstream, so in time SF's forms will multiply and mature and hybridise and bloom forth with as yet unimagined subsubclades (asteroid colony gene pet stories?...), while recognisable SF forms become the mainstream - against which the periphery will be force to rebel, of course...


" They're notoriously information-poor "

Example: the unabridged direct to English translation of Stanislaw lem's Solaris is now out as an e-Book. Determining that it is that edition and not the old one requires that you download the preview chapter and find that information in the front matter.


The other problem with the "infinite bookstore" is that the old classics are harder to dislodge to make way for the new.

How do new books get the attention of readers when there are so many amazing existing books that are available that you probably haven't read?

These days I follow most of my favourite authors on social media. When one like Alastair Reynolds tweets "reading China Mountain Zhang, one of the best novels I've read in many many years. So good it almost makes me want to give up or get a lot better" - you bet I go and buy it. But that's also a novel that's almost 20 years old.

But if you were starting out from scratch and you knew you liked Sci-Fi - should you work through one of the 100 best SF novel lists or start with something newer that hasn't had time to develop a reputation? It only takes reading a few crappy books to appreciate that there is an amazing amount of stuff out there and that you could spend a very long time reading what today's authors consider to be the classics without actually dipping into what those authors actually write.


Devil's advocate: is this a bad thing?

That's a question which, I think, is the subtext running beneath the surface of Elizabeth's and Abi's essays.

There's also a perception among genre readers that mainstream lit-fic is very much "eat your greens" -- something you might do dutifully but don't actually enjoy -- but I can see folks on the other side of the fence (yes, I know you have grave doubts about the author in question) arguing for the opposite, kinda-sorta. Or rather, arguing that lit-fic should be fun and, er, encompass the fantastic as well as the dismal mid-life crises of creative writing academics.

As I said, I don't have an answer to this (other than "write good books").


I'd love an online bookstore with an interface ripped wholesale from Apple's Time Machine backup utility (the scale at the right is a zoomable timeline; slide the pointer up or down it to see different points in your backup history), applying the time line to publication era. Is that too much to ask from Web 2.0?


"What is to be done?"

Explicit linking between books? A book will continue to stand as its own work of course, but that doesn't mean you need to keep 100% tight seals between books by different authors. Sometimes the same author "leaks" references between their own (otherwise unconnected) works after all.

When books are permanently and equally available across time and space is there a reason you could not be more explicit about the thread of conversation between authors by referencing each other in a more open way than now?

"...or the cuneiform clay tablet."

A limited collectors edition of Accellerando in the works perhaps?



You're a 'glass half empty' kind of person, aren't you?

Glad you realised some of the changes that eBooks can bring (cover design, who needs one). However you are still trying to say how can we adapt what we know to a new world - never a great idea. There are many more opportunities now to redefine how things that operate very different approaches that simply were not possible before.

That an attitude thing.

But to limit things just to your exam question, who says we need to put stories into categories? Why can't we understand the personality and interests of the individual and present them with stories they might be interested in - based on how the story works, how it's put together, where it goes?

"You might be interested in this story, because it has an intelligent plot, several twists, but comes to an uplifting ending." The fact that it might have been classified under romance in the past is unimportant.

Something that AI might actually be good for.

If that means authors refer to another author's work - well that can mean that someone who's just finished the first can be suggested the second as a good related work. Does it really matter?

In the end the most important person - the reader, not the author - can get a better experience by having things focused around them, classifiers be damned. And that's a win.


My brother-in-law the film buff tells me there's the same situation with movies, with various scenes and references to past works, and how I can't properly appreciate X without having seen A, B, or Q, so it's not just SF that's in conversation with itself.

On the ebook front, a thumbnail could and should still be mostly art; the place for author and title are in real text, not rendered into an image. After all, presented with a huge list of book offerings, you don't want someone to hit ^F stross ENTER and not find the merchandise! Meanwhile, you're not tied to just the cover of the actual book anymore: the icon might be just Liz's face, the full image is the Rule34 deadtree cover, and you could include the un-covered-by-words original artwork in large format as a bonus.


Why can't we understand the personality and interests of the individual and present them with stories they might be interested in - based on how the story works, how it's put together, where it goes?

Because that's an AI complete problem.

For example: I just read William Gibson's Bigend trilogy. What was the next thing I read? Why, it was a hokey mil-SF novel. Followed by a werewolf novel. Then I got my teeth into a non-fiction book by Seth Lloyd, on the cosmological limits to computation. Readers are inconsistent. Their taste varies over time, as does their level of motivation for stretching their cognitive muscles. Sometimes I want a challenge, and sometimes I just want the equivalent of a lightweight soap opera. An AI that can second-guess what I want is actually deeply scary, insofar as it's gotten way inside my decision loop and knows more stuff about me than I'm happy about anyone/anything else knowing.

Finally, if you spend even half an hour surfing the Amazon reader reviews of books you're familiar with, you might downgrade your expectations of reader reviews and crowdsourcing as a platform for deriving such information.


Two things:

  • I think criticism is going to undergo a "new new" renaissance, and serve (much of) the job that the old "publisher slush file" did in terms of filtration. Things like Metacritic and Rotten Tomatoes -- while still only being alpha versions of what I'm pointing to -- are now primary sources of information for me in the (equally crowded) game and movie markets.
  • This is a dramatic change from my habits 5+ years ago, where I bought games after browsing in game shops, and "bought" movies after seeing trailers in other movies (i.e. being "in the store").

  • IMO, the vast majority of publishers (there are notable exceptions) still have no fucking clue about even the most basic e-book concepts. An example? I wanted to scratch an itch and purchase "Dune" last week as an e-book. Except that I can't. Because at the moment the publisher has an "anniversary edition" in the stores, and doesn't want to "undercut" it, and so the only e-book version of Dune available, right now, is exactly the same as the old e-book version, except $$$$ more expensive, and being deceptively sold as an "anniversary edition" with supposed "additional content" that you don't, in fact, get when you download it (thanks ever so to the burnt reviewers who pointed this out)!
  • I spend thousands of dollars on books per year, mostly p-books. I am your dream customer, oh publisher, and yet you do not seem to understand the basic maths that if you allow me to buy more books (by pricing e-books reasonably) then you will make more money even though I am spending exactly the same amount!

    I'd love an online bookstore with an interface ripped wholesale from Apple's Time Machine [snip] Is that too much to ask from Web 2.0?

    The answer is clearly no. Time Machine as an interface would be a bit of a cow to implement quite that nicely using HTML 4 and JavaScript but just about doable I think. Once HTML 5 is standardised, or if you're prepared for patchy implementation across different browsers, you might not manage a complete rip-off but you should be able to get something close. Similar alternatives with time-lines have existed for teaching history (for example) for a while if you can't quite manage the Time Machine look.

    Which leads to why it's not there already? I wonder if I went to the bank with a mad idea for an internet book company that was pleasant for the consumer. It enabled them to find the book they wanted, previous books by the same author. Ideally approximate future release dates in a series, firming them up as more and more is made available about them etc. if they'd give me the money? There's already one gorilla, or maybe a large South American rather than African creature - Anaconda say, in the room to compete with. There's a fair-sized competitor too from the fruity one.

    The iTunes store seems to be a commercial success, as do both App stores based roughly on it. The iBooks store seems less so - that Anaconda in the room again - so maybe they could be persuaded to redesign that way, avoid the patent lawyer issues since it's their interface already, and attract customers by providing a different, more friendly to some customers, interface. I know I'd kill on occasion for the ability to track a series, order all the old titles, be notified without extra spam about publication dates for new ones and the like. And as an iPad-iBooks consumer they could probably get even more custom from me this way.


    Well, apart from concurring with the sentiments of #1 and #2 (I've never cared for anything more than the story and the interesting idea - the concept of dialogue via books across years is, to me, just silly), I have to say that this infinite bookshelf has always been there for me.

    I'm in my thirties. I've always been an avid reader. And right from the start I've been swamped by the sheer volume of what is out there. We're talking about a few hundred years of literature, growing in bulk as both reading and publishing became more wide-spread.

    Even limiting to a sub-group like a genre doesn't make much of a difference. I read my father's Doc Smith, Wyndham, Wells, Clarke and others, and then started looking for more and quickly found far more than I could ever hope to read (and this was in my teens, with plenty of free time). That was the late 80s, early 90s.

    Tags, as long as they are accurate, are just what I've wanted. I long ago stopped bothering to seej just fantasy or just sci-fi - there are plenty of sub-groups within both that I just don't care for, but there's no quick way to spot them, even with the book in my hands. Tags could do that for me. I don't need a genre, I need a range of subject matter.

    And genre is a rather subtle limiter too. I would mentally classify quite a number of favourite 18th, 19th, and early 20th century classics as fantasy, but according to the world at large they are simply "literature" (and owing to their age and establishment as classics, people I speak to just can't seem to wrap their heads around the idea of, for example, the works of Dumas or Haggard being fantasy). If I went looking by genre I might never find them. Remove genre and just use tags like "adventure", "swash-buckling", "futurism", or even "bodice-ripper", and finding books to match your interests becomes so much easier.

    I've long thought that bookshops ought to have computers for public use with databases of books tagged like this, as it would be so much easier (and allow on-the-spot identification and ordering of books that would be of interest but not stocked - a great benefit for both customer and shop), and could never understand why publishers wouldn't be all for it - right up until I read your writings about the nature of the industry :) With luck all this e-book business might actually change things for the better.


    Your apple and snake story is eerily familiar..

    "The iTunes store seems to be a commercial success, as do both App stores based roughly on it."

    Amazon may raise Apple's game since iTunes has a lowest common denominator approach to music that devalues the metadata that might make it easy to find works by a particular artist or composer. Apple does file artists into musical genres and if they're filed in multiple categories it can be impossible to browse from one to the next. Composers are barely worth worrying about so good luck finding tracks based on that factor.

    Apple have also ossified the old market divisions based on geography into their new country by country stores, so you may find yourself less able to buy content simply because there is no foreign CD you can import.


    I have to read a novella worth of text every day just to keep my feed reader manageable. There is just too much stuff out there in too accessible a form and the fact that 90% of it is still crud is completely overwhelmed by the sheer volume of crud. What hasn't changed is the number of hours in a day. Sooner rather than later we are going to drown in a sea of information that is completely beyond any manageable human scale.


    Wholeheartedly agree on the curation aspect, though arguably it will extend to a lot further than just online bookstores. Working on something along those lines now, you've got my email address if you want to be included in the initial release of our project! ;)


    Perhaps improved curation is a way that publishers could distinguish their direct-to-consumer ebook storefronts from Amazon...if they indeed choose to build such storefronts.

    If I could buy Tor's much-ballyhooed DRM-free books directly from Tor, in an online store that incorporates the ideas Charlie has started to spitball above, it would be wonderful.


    What's to be done?

    It's back to the 1930s when vibe was more important than genre sub/genre. The sub genres will be sorted by their secondary tag, e.g. The "Mil" will be more important than the "Hist" or "SF" in "MilSF" and "MilHist".


    Janne @ 10 suggested the following as a possible solution to the loss of "genre" as a workable sales categorization schema:

    Explicit linking between books?

    I have to admit, I'd be interested in seeing whether it's possible to do this. Maybe something along the lines of a reference list, or a "books the author was inspired by" list would be a start. Certainly one of the things I enjoyed about OGH's first couple of "Laundry" novels (beside the bonus novellas at the end) were the afterwords, where Charlie would speak about what inspired that particular book, and what he was working towards. Somewhere in the back of my mind, I have plans to look for copies of Len Deighton's "The Ipcress File", as well as digging up some of the original Bond books. And probably John LeCarre (I think I read some of his stuff years ago, but it's been so long I've forgotten it). All I need now is either the time to find the books in the library, or the money to purchase them myself.

    I can think of other ways of playing around with it, too. For example, one of my favourite Georgette Heyer books is "Venetia" - but it's full of what are now considered scholarly references to obscure bits of the British literary canon. So maybe a list of the poems being referred to would be a good help for more modern readers of the book. Or a list of authors contemporary to the characters of the book. Or a few references to the history of the Regency period. Just little things to provide a little extra depth, and possible further directions for reading. There could also be some linkage of Ms Heyer's mini-series within her books (for example, "These Old Shades", "Devil's Cub" and "An Infamous Army" are a series - and "An Infamous Army" also shares characters with "Regency Buck" - so why not provide explicit information pointing new readers toward those other stories?). If you're looking for books around the same historical era, then other authors could be pointed to (for example, for an alternative perspective on the Peninsular Wars, they could point to Bernard Cornwell's "Sharpe" books; for naval battles, why not "Horatio Hornblower"?).

    (Disclaimer: I'm coming to this out of about a month solid of doing research for various psychology unit items for university, and one thing I've discovered is that psych reference lists are a bit like randomly browsing the internet. I start by looking at one article, and they've got references to other articles, so I go and look at one of those, and then there's an interesting reference in that one, and before I know it I've wandered off into the equivalent of the backblocks of TV Tropes and there went the day. Good reference lists can be scary!)

    Basically, I suppose the point I'm trying to make is that the end aim of all booksellers is to keep people reading, and reading stuff they might be interested in. So why not use the text we're starting from as a jumping-off point to bigger and better things?


    A very interesting post, put my head in a spin. As a bookseller i completely agree with '(b)'.

    When you contrast your statements with the ideas of Stephenson's Hieroglyph project this post is even more interesting. Stephenson's argument that SF has lost its soul and your argument that SF has been diluted and is in serious danger of being further diluted as e-books expand - two sides of the same coin?

    Perhaps SF needs its philosophical soul and purpose back?

    But how can you market hard core SF? I'm thinking that the algorithms need to change. The problem is that changing them is unprofitable. It's simpler to create a formula which takes basic measurements like total purchases and customer 'five-out-of-five' votes. You guarantee to get the median perspective. The long-tail just gets lost amongst the median.

    I think relying on 'expert panels' or even experienced customer recommendations, with a sticker and label visible at the point of purchase almost saying 'This book is really good - So says the SF Council'. I guess moving away from corporate brands to expert-controlled or experienced-crowd brands.


    ...the concept of dialogue via books across years is, to me, just silly...

    Perhaps it is. However I remember Michael Swanwick saying that the dialogue is something along the lines of reading that month's Asimov and going "No, no, that's not what the future is going to be like, it'll more like this" and writing a story. Which I take to mean that it's about going "That's an interesting idea, but it would be even more interesting if..." or "He's not thought that through, a better story would take that and..."

    So the dialogue is more about improving your own work by considering and reacting to others and of course showing that you can do it better.

    I don't think I've disproved that it's silly.

    Whether this dialogue would be a useful tool for curation of books, or if the curation would make the dialogue explicit is an interesting question I don't have an answer to.


    I'd be interested to know just how many SF book purchases are influenced by the internal dialogues of the genre, and how many are impulse purchases at point of sale based on the cover. Certainly some, but I'd bet many more sales are made on the basis of meatspace and cyberspace word of mouth. As a reader I trust other readers to have good judgement about what I'd like more than I trust writers. Online reader reviews and ratings are of no use to me because I have no idea what the person's context and priorities are. So trust enters into it as well.


    Well, I'm quite reassurred by your piece.

    I don't do e-books; I've never got into the habit of reading on-screen for pleasure, I won't have a Kindle (partly because Amazon is so resolutely anti-trade union), and I have quite enough p-books in my To Be Read pile to keep me going until my dying day even if everything transferred to e-format overnight. So the demise of the p-book as a mass market item holds out some hope to me of catching up!

    And given that a large segment of the reading population get by in life without ever reading any serious SF, it won't be an appalling personal tragedy if I miss some future gem of classic sf because I don't currently touch e-books.

    (You just watch: now I'll probably come into posession of an e-reader by some unexpected route sometime in the next three months and become a Convert...)


    I'm not sure if this is a comfort or a source of frustration; but this is a link to the 1970s cover from one of my favorite (can we call it speculative fiction?) books by Kurt Vonnegut under the Pseudonym Kilgore Trout >>

    He got this then hormonally excitable teenager to read a better class of fiction, science fiction or whatever you want to call it. Maybe cheesy marketing that has nothing to do with the content is not such a bad thing if it gets us (by necessary means) to depart from our comfort zone long enough consider new ideas.


    Sorry for the double posting - I got an "Internal Server Error" message when I tried to post the first one...


    The cover of Rhetorics of Fantasy was selected with e-sales in mind. Look at a hard copy, a nice enough cover. On line it glows.

    I think you are right but not in the sense that the genre will die, but the fuzzy set is going to get a lot fuzzier.


    That wasn't Kurt Vonnegut; that was Philip Jose Farmer.


    That's usually a sign that the spammers are getting frisky (hitting the server hard). Groan.


    "the concept of dialogue via books across years is, to me, just silly"

    Well, whether it's silly or not to you, it's something that exists in all forms of writing, at every level of the high/low art spectrum. Whether it's Joyce patterning Ulysses on Shakespeare and Homer or Grant Morrison patterning his superhero comics on Pilgrim's Progress, our host nodding to Deighton and Heinlein, or Shakespeare seeing a play called King Leir and thinking it'd work better with a downbeat ending, it's a fundamental way that writing works.

    And that means that anything that changes that -- anything that alters the presumed shared cultural context of reader and writer -- is potentially either dangeroud or interesting for both writers and readers, depending on how well they can adapt to the change.


    E-Book genres are already mostly dead, you only have to wade through Amazon's god-awful genre listings to see that.

    And Amazon's are clean compared to Sony's new ebook store / pile of steaming.

    Multiple genre listings is fine - it's a reasonably natural way to go. Computer games have been doing it for years.

    Maybe the end of DRM will allow more specialised shops to appear, with curators that care about their products. Maybe when the dust settles a new breed of trustworthy reviewer will have become established. Maybe one day I will have that moon on a stick.

    Currently, (apart from personal recommendations and well known authors) I search on Amazon and read the negative reviewers.

    If they are erudite (rather than pompous), or at least intentionally amusing I don't buy the book.

    If many would be more comfortable with Harry Potter:

    "This book is a confused and confusing mess about characters...",

    "...just too difficult to read"


    I buy.

    It's not a perfect system, but it's the best I've found yet.


    While I've been buying online rather voraciously for 15 years I have a point of always buying titles I discover in a bookshop at that bookshop. Booksellers who curate so well that they can surprise a person drowning in book feeds need to be supported. Music, ditto.

    However it must be said that I can't remember the last time I discovered a SF genre title in a bookstore - even in a genre specialist like Forbidden Planet. I have gone to a store to buy titles explicitly because it's convenient or I know they discount new titles, but I don't pick up anything extra (in genre). I might see some lit or nonfic title and pick that up, but the genre shelves tend to be so thoroughly overgrown with indistinguishable fantasy series, game & TV tie-ins that it's not worth my time to browse.

    In Australia, AFAICT the genre shelves are filled with whatever the distributors send out. I would have had better luck at the local newsagent back in the 70s when their genre half-shelf was filled with Aldiss, Harrison and Heinlein.

    Much of the internal genre dialogue about hard-core SF seems to point towards the output of small-press publishers. However I've been burned so many times by their fulfilment people (titles years late, wrong items sent, partial orders sent) that I can't be bothered much now.


    I don't know what it means, yet. Jo Walton opined that the conversation will go on, and I'm sure it will ... among those of us who are already aware of it.

    But there was so much less SF in the 1970s that it was quite possible for those of us who grew up reading the field back then to acquire a comprehensive coverage of it. Today, there's far more stuff out there: but without the clear signifiers, the tags saying "queue here to join the ongoing conversation", it may become increasingly hard for new readers to recognize what's going on and join in.

    What is to be done?

    I think you may be overstating the case a bit here. Sticking something on a shelf marked sci-fi/fantasy is one way of defining the boundaries of sci-fi as a genre, but there's a bit surely there's a bit more to it than that. It seems to me that a large part of what makes sci-fi a discrete genre is that sci-fi fans are quite a discrete group of people. If a newcomer wants to know what counts as relevant sci-fi, all she (we can hope, right?) needs to do is find someone in the relevant community and ask.

    I guess that this has always been the case, but it's something that technology makes much easier. As a relative newcomer to sci-fi myself, a big part of the way I oriented myself around the genre was by getting on blogs like this and finding out what I should be reading if I was going to properly understand Saturn's Children. You don't really need to be picking up on anything so subtle as a signifier when people will actually just tell you where to queue to join the ongoing conversation.

    I realise that this process assumes a bit of active work on the part of the new fan. But insofar as we're discussing the sort of sci-fi fan who, in the 1970s, would have been expected to develop a comprehensive coverage of the genre, I think that's a reasonable assumption.


    It seems obvious that the solution to this problem is an open source crowd-sourced database with reviews and reputations and a shitload of metadata about each book. Then your time machine UI becomes simple, once Apple's patents have expired, and people can indicate that Du Bist Todt is a translation of Halting State.

    Actually, this information ought to be maintained by the publishers, who presumably would have pretty good reputations.

    It seems to me that Delicious Library is poised to own this space, except of course that they aren't open source, and so all the network effects of having an open exchange don't exist for them—they're more like that old CD track database that Apple closed. I can't even remember what it's called anymore.

    Anyway, in order for this to happen, publishers would have to get on board. It seems to me that it would be strongly in their interest to do so. Title pollution is a really big problem.


    While in the main agreeing...

    In the 70s & 80s it was relatively easy to cover a genre for free, because we had libraries. As a child & then teenager I'd read a book a day - if a book was not to my taste, I took it back the next day and got something else.

    Perhaps pirated (Arrr!) ebooks will fill the same role now.


    The new filter for SF will be review sites, blogs, podcasts and online free collections like:


    "Determining that it is that edition and not the old one requires that you download the preview chapter"

    Amazon, at least, lets you enter up to 4,000 characters for the book description (somewhere in the range of 600-700 words, depending on how you count them), so the problem here lies with the publisher.


    The classics are, indeed, classics for good reason. They are also usually well and truly dated. Look for computers in Asimov's Foundation or E.E. Smith's Lensman. Look for the internet. They just aren't there. Mission of Gravity sort of holds together, though the world-building seems less reasonable these days.

    FWIW, I find older fantasy to be much more readable than older Science Fiction. The basics of interpersonal interactions haven't changed. Yet.

    That said, there are benefits of genre that stretch well beyond marketing, and the limitations on time mean that one can't even follow a small genre (Science Fiction) comprehensively. So "interest tags" will become important. And "like tags"...but "like tags" will be especially subject to junk pollution, so identified reviewers will become more important if things are to turn out well.


    You need another eight centimetres in front of the shelves for the poor bewildered customer to stand in

    Charlie, I think you mean eighty centimeters.

    (I should also note that my eyes are getting worse as I age. I also need to be eighty centimeters from the text while I browse in a bookstore.)


    It seems obvious that the solution to this problem is an open source crowd-sourced database with reviews and reputations and a shitload of metadata about each book

    That's "obvious" in the same way that it's obvious that all you need by way of an ecosystem for a space colony is blue-green algae and soy beans.

    Which is to say, it's both "obvious" and subtly wrong.


    "Spoof reviews posted on Amazon have sent demand for the most unlikely merchandise soaring and are playing havoc with the online giant's famed algorithms, the complex formulas that recommend what else buyers of a particular product may be interested in purchasing. The current must-have item heavily lauded by the spoof reviewers is a £50 canvas print of the TV presenter Paul Ross. Mugs with the same image costing a more modest £8.99 are selling well, with Amazon suggesting that customers also buy a particular bottle of methylated spirit attracting the sort of reviews normally associated with Château Latour."


    Seen from another angle (that I totally not stole from Warren Ellis) the genre definitions might grow ever more fuzzy till we come to an amalgamation event when the Literati have to incorporate ever more elements of modern life to stay relevant and thus become more science-fictional.


    The field is too large, and will have to be split up. What the new organising principle will be I don't know.

    Some current organising principles : individual blogs. Personal reccomendation. Story and plot summaries (e.g. on wikipedia).

    Some fields might be:

    Near future digital environment, correct positive and negative moral philosophies positive and negative ethics.

    Near future ai, positive and negative moral philosophies, positive and negative ethics.

    Both of these would be cyberpunk to a degree.

    Approaches to economics.

    Space flight in general.

    The philosophy of being in a modern society.

    Fantasies of various sorts - the freeing of the mind by flight, raw intelligence or similar.



    Banks losing his M. could be an example of the process. Or something else entirely by all means.


    What I want from a book shop that I don't always get:

    If your book is in a series, where is it in that series? I don't care if it's book #26 necessarily, but if it starts somewhere else I want to know that. I usually start with book 1 when possible. (example: JD Robb's X In Death series, each almost all entirely self-contained, but with little bits of story that run in series. The publishers stopped mentioning which book was which about four books in. You have to read the front matter to determine where in the series the book falls.) Amazon in particular fails quite spectacularly at this, especially when a set of books is published together.

    I want this from a bookshop because while other sites exist that I've found through reading here and other authorial blogs for reviewing, many of them are so scattershot (and require a login and input of initial books, which when you have hundreds of books in your library is a non-trivial effort)

    What I'm never going to get: a handy device/smartphone app that I can scan a UPC/copyright page with and get an e-book version (preferably at a reduced rate, ideally free, ha!).

    In cases where it's possible, the amount of effort I have to go through to get what I want is more effort than sending a sample from Amazon to make sure I want to actually buy the book in question. (The one exception so far: the hardcover edition of Cryoburn included a CD of all the previous Vorkosigan books I already had, so good on you Baen).


    What is to be done?

    Easy: people who are passionate about niches need to create specialized distribution channels for them. That's what Instagram did.

    You sir, sound passionate about sci fi. Create the world's best sci-fi ebook website and return value to the universe.

  • We will see the formulation of new genres.
  • An interested kid's taste will be guided by relevent blogs (or whatever) rather than relevant shelves.
  • The entire culture's becoming more SF/F anyway, so it will bleed into all other genres, thus simultaneously lowering the level of the SF/F discussion in most, but also creating enough demand that some truly great SF/F gets written. As in greater than all pre-existing.
  • 50:

    Charlie, I took Bear's essay to be a reply to Christopher Priest's rant from last month (Note that the essay did appear in Clarkesworld.) and I read it in that spirit.

    That being said, I do think you're right that the genre (and genre as such) is going away in a cloud of meta-tags. And yes, for everyone who said so, we definitely need better curation.


    Anecdata: Last week I was passing through Montreal Dorval airport en route for a transatlantic flight. I went to the airport bookstore, straight to the science fiction shelves and... they weren't here anymore.

    There are maybe three rows of crime novels, ten of "Fiction" but no SF. I asked an employee and he told me that science fiction books had been moved and blended in the "Fiction" section.

    Problem: "Fiction" in an airport is 90% chick lit. Nothing wrong with that, safe that SF books had become needles in a haystack. I ended up buying one of Brockman's Edge collection of essays (in the Non-Fiction section), which I read on my way back.

    Once home, I went to my Amazon Wish List and ordered a couple of SF paperbacks. This post made me realize that this Wish List is the record of my discovery process. It works. Too bad for the occasional discovery in airport bookstores.


    While I'm not arguing it's a pain with the country by country stores, my understanding is that this was the compromise they reached to be allowed to sell the products. It wasn't so much led by Apple as the dinosaurs in the recording industry. Dinosaurs who managed to have the courts block UK access to The Pirate Bay... which won't slow people who want to torrent at all.

    Which, of course, might be an issue with doing it for books. As OGH knows all too well he has different publishers, contracts etc. for different bits of the English speaking world.

    But detecting your location automatically is pretty easy too and altering the data based on that not too hard.


    But who should read "the genre" and why?

    There was talk among some academics of science fiction STEMulating interest in science among children in the late 50s and even before Sputnik. They mentioned books that demonstrated basic scientific principles in their stories. But what science is there in Star Wars and Hyperion.

    The genres need to be refined to be more descriptive so readers can find what they like but maybe grade school kids need to be encouraged to read certain things rather than letting the randomness of market forces control people so easy to influence and sabotage. And now some of that olde SF is free.

    I stumbled across it accidentally and had to buy it but anyone with a wireless tablet can get it for nothing, and in audio book form.


    As the de facto book medium becomes digital, curation becomes a necessity. The same way that google filters search results so that your not given a lists of pages that just have your query, verbatim, pasted 1000000 times and absolutely no real content.

    There are plenty of reading communities out there, and many of them have massive amounts of readers who give valid reviews and ratings every day, but they are no replacement for the curated shelves of our local bookstore. There needs to be a place where curation is the prime goal—where any book chosen or submitted must have a meaningful story and a worthy cover, and if it doesn't, it can be left to be forgotten.

    There are so many markets and business models that haven't even started to evolve with the Internet. In this case, it's the the book seller. There is no true digital bookstore.


    When books are permanently and equally available across time and space

    Something I wrote five or six gears ago, which might be appropriate:

    "I saw the Epic of Gilgamesh going by on alt.binaries.audiobooks. Just this morning I was reading an archeology book telling how a single schoolteacher in England managed to decode Akkadian cuneiform writing, and spent his nights deciphering the story... only to find the last tablets were missing. So he managed to organize an expedition to the diggings in Persia, and actually found the missing tablets. And so we have the story of Gilgamesh.

    The Akkadians predated the Egyptians; the epic was written somewhere around five thousand years ago, at the dawn of known human civilization. [...]

    Five thousand years. Close enough to forever... using technology they could not have imagined, I can pluck it from the networked aether, and have the voices of the djinni read it to me.

    And who claims there is no such thing as progress?"


    Huh, seems like my post was eaten. Quick thoughts on the issue: part of the problem with Amazon's recommendations is that it works on the assumption that everything you browse and buy for is for you and that you will always want it. It doesn't take into account that once I've bought my mother some jewellery for her birthday that I don't want variations of that jewellery clogging up my homepage all year round. Perhaps a better system would be for a selection of your library to be used for searches.

    For example; having just finished a book a reader may wish to read stories with similar themes, technologies etc. They could search by tags, search by what others bought, or perhaps sort by selecting multiple ebooks from their library and use them as a basis of a search for both.


    Although I can't generalise for everyone, I don't think I've ever bought a book on the expectation of it being a response to something else. I enjoy catching such replies when I catch them. I'm sure I've missed many though... and probably enjoyed those books too.

    Tonight I'm going to see The Avengers. I know there's other movies around the ensemble - Iron Man, Thor etc. I've seen some and not seen others. It might affect what bits of the movie I get occasionally but if it's well made (I have my doubts after some reviews) it should work anyway. Don't believe me? Just look at the notionally children's movies with a dual layer. Something that appeals to the kids (who don't get all the references) and something that appeals to the adults too, smart references to other movies and the like. If you miss a few... there's still good stuff there to enjoy.


    But how can you market hard core SF?

    Just put it out there, we'll find it. I liked EE Smith's Skylark stuff, and John Campbell's "Arcot, Wade, & Morey" books, and Harry Stine's "Star Driver." But there never was much of that to start with, it's much simpler to write about feeeeelings, oh-woe-woe, feeeeleeengs...

    Stuff your socially-relevant stories with their dysfunctional characters. You can have those in any genre. Now, something like Eric Frank Russell's "Sinister Barrier" from 1939...


    Charlie, what about all the literary awards we have? They are a way of curation.


    Apple restrict access by credit-card address. Amazon, eMusic and others restrict by IP-address. Some manage to sell the same digital stock with neither restriction, but unless the artist is "connected" they just leave it all to iTunes, and usually just tick the box for their home market.

    It's not clear what address restriction the vendors want to enforce, but as we all know their reluctance to engage with the matter just drives the audience look for the "free and easy" solution.

    I don't think Apple or Amazon are exactly pushing to remove existing restrictions.


    " part of the problem with Amazon's recommendations is that it works on the assumption that everything you browse and buy for is for you and that you will always want it. It doesn't take into account that once I've bought my mother some jewellery for her birthday that I don't want variations of that jewellery clogging up my homepage all year round. "

    Actually you can easily remove any item you've bought from the recommendation seed list. As soon as I see anything out of field being recommended, I look at the "Fix this recommendation" link under it, and then find the seed item, and click "Don't use for recommendations".

    That's been there for several years.


    reference lists

    There's a book on the design of internal combustion engines that has reached near-cult status, now being used as a major reference work as well as a university text. It's widely accepted without criticism as the Final Word for its subject.

    The guy who wrote it was a university professor, and likely used grad students to do the scut work. But his name is on the cover, so it's his responsibility that many of the references he's using to make various points actually say something different, or occasionally the opposite, of what he implies in the text.

    Fortunately, the references are extensive, and I've managed to track many of them down, as in "locate some place that has a copy, then pay their reference librarian to make a copy and mail it to me."

    I'm starting to see a lot of younger engineers who simply aren't interested any any material they can't access via a web browser. Which limits their resources more than they understand...


    I believe genre will survive, but that it will become significantly harder for a new author to make a living. This might actually be a good thing for publishers.*

    Good, thoughtful science fiction will probably never be mainstream - it doesn't match the interests of the general public.

    The infinite bookshelf will result in clear differentiation of 2 classes of customer. (1) Casual reader: these people will never, ever finish reading the classics in the genre and will be best served by purchasing only those classics. Once copyright is reformed to something reasonable**, like 20 years, they will rarely pay for ebooks. (2) 'Voracious' customer: those people will continue consuming new titles that strike their fancy. They are a substantial portion of the market, but, if I remember an old post here, less than half of it. Those people will continue buying thoughtful science fiction.

    Search is a solvable technical problem for the internet in general. I believe that the eventual solutions will be superior to the current model, at least for me. I am quite tired of walking into bookstores, looking at the new items, and realizing that 8/10 of them are simply awful. (or, at least, not to my taste) I'd much prefer browsing through ebooks from the 1970s - if they were available.

    Obvious immediate suggestions involve curated genre-specific sites linking directly to various ebook sellers. Possible improvements might include filtering by different classes of reviewer. (publisher, author, friend, long-time, well-thought of customer, idiot)

    --Erwin With the current status in copyright, the long tails on classic works may make it somewhat feasible for publishers to finance excellent works in literature even though a single author would starve to death long before he'd published enough books to make a living. *After 20 years, the remaining profits don't provide much of an incentive.


    It's a nice coincidence that your blog post appeared just an hour after I'd written my own LJ post along similar lines to Bear's article. And then you post additional links to related content. Something for me to think about; thanks!

    However, to your main point... genre won't disappear. It will, however, change in how it is visible.

    In the past I have bought books because of the cover. I figured a book with a Josh Kirby cover meant the publisher had enough confidence to splash out and that the content would be interesting to me. It mostly worked. In the world of tomorrow, covers will be less interesting but tags will become so. A book with a "Strossian" tag would be more interesting to than "Tolkeinesque" and infinitely more interesting than "Bodice Ripper". Amazon have already made a start on this, encouraging users to apply and vote on tags.

    The result is possibly going to result in "strange attractor" structures. You'll have primary tags which will end up defining the genre, and then you'll have secondary tags which people will find themselves focusing on. It'll take an effort to find books outside of the local attractor.

    Amazon are trying to "crowd source" the solution. Cheap, scalable, unreliable. And prone to bot attacks, robo-publishing scams and the like. A more curated solution with oversight on tag management and content inclusion is more expensive but likely to be more useful for discovering content.

    Will such "boutique" stores survive? My gut thinks that this is where genre might start to thrive. But it will be more expensive, and if the same content is available on Amazon then the boutique will be undercut and fail. So the curated solution will be hard. We'll need smarter algorithms. I see a parallel between this and bank trading systems; that 3% more accurate system will give a sufficient edge. Bots taking and evaluating Amazon content rather than just providing content. Expert Systems will provide recommendations based on this.

    Genre will survive, as tags, rather than cover art. An ecosystem of tag analysis will grow. Amazon will remain the low-end base-line bulk haven and will grow as a result. Special recommendation systems will grow up around it. (Google will compete, here).


    "Add in the expectations of studio executives and the dumbing-down effects of editing by committee you end up with huge pressure to make the script commercial rather than complex."

    If Amanda Palmer can crowdfund her new album on Kickstarter, I see no reason in principle why SF films couldn't be funded that way. If they were, there would probably be a lot less dumbing down.


    " see no reason in principle why SF films couldn't be funded that way. "

    -> Iron Sky


    Some very good recent SF movies: Limitless Mr Nobody Inception


    Ok in my shitty German Du bist tod means YOu are dead?


    I'm with the crowd who don't see the death of genre, especially science fiction, just yet.

    Here's my simplistic analysis: when the world wide web first took off, there was some talk of how chaotic it would be, when everyone was talking to everyone else. In practice, today's web is almost infinitely balkanized. I can talk to many more people than the people I choose to talk to on a routine basis, and that's fine. I'm pretty sure literature works this way too.

    As Charlie noted, science fiction has always been about authors and fans talking to each other, with a large penumbra of passive readers consuming the results of the conversation. Many art fields are like this. Only two things have changed: the speed of the conversation (no longer necessarily limited by the speed of print production), and the slush pile coming into public view. Otherwise, the web simply mediates the conversation, the way publishers, newsletters, BBS, and conventions did (and to some extent still do).

    I suspect that, so long as writers can afford to write for a living, we won't see the SFF field go away.

    I'll also point out that the idea of a publishing bell-curve centering on the midlist may have always been a bit of a myth (cf this NPR story). It's likely that in SFF, as in most fields, a few writers have always been disproportionally influential, and most writers have been on the sidelines. In the past, we might have assumed that the published work followed a bell-curve (few stars, a large chunk of midlisters struggling to get by, and a few who somehow got out of the slushpile for their a single book before disappearing). In fact, the large bottom of the publishing curve was always kept out of sight as the slush pile. Now the slush pile is simply visible. The field is still powered by the few stars, and it's not clear what else has changed.

    Finally, I'd point out that the major successes in the self-published eBook field so far seem to be straight-up genre work, rather than mainstream literature. Based on extrapolation of current trends (always reliable...), we could equally argue that SFF will continue indefinitely, while mainstream literature will only continue to exist by self-consciously recycling older works of the mainstream, or adopting the conventions of various genres. Actually, isn't this what SFF does already?

    What will change publishing is if the big multinational corporations with their high profit margins get out of the game. If the publishing world becomes less driven by the need to have a set number of blockbusters year in and year out, it might become a bit more livable for everyone.


    The number of people and amount of equipment involved in producing a low-to-mid-budget movie (i.e. one with no $TAR$ on the billing) is an order of magnitude or two higher than the prerequisited for producing a high budget music album with associated promo videos.

    Yes, it's possible to crowdsource-fund movies. It's how IRON SKY got made. But it's difficult and currently it puts a cap on the production -- and more to the point, it doesn't handle the distribution problem (although that might be getting easier with wider adoption of internet downloads for movies).


    My book-finding mechanisms have been online for many years - long before the more recent move to ebooks. The destruction of the bookshop, outside of cities, has pretty much removed the utility of the "look what's on the shelf" approach as far as I'm concerned since I live out in the sticks. Our local library does a better job of having a SF shelf than any bookshop that I can get to.

    I find books through places like this blog, my online social networks, and amazon et al recommendations. Those first two options make me think that the SF genre will survive because SF fandom will survive as an online community.

    Jo Walton's "Among Others" has just gone on my wish list for example :-)

    Maybe that will make SF a bit more niche. Maybe it'll fragment a bit. I can't imagine it disappearing.


    Because it threw me off completely as a native german speaker whilst reading this blog entry: it's "Du bist tot" - "Todt" is a not-unusual last name.


    I don't think that the industry has figured out how to design covers for ebooks yet - which is why they're mostly lousy.

    You need visual designers with a very different skill set from normal cover design. The kind of folk who design icons rather than posters.


    You also, ideally, need authors whose names are monosyllables and whose book titles are short.

    I lucked out in the name draw, but I need to work more on the titles. (Alas, the series books aren't amenable to shorter titles because there's already an established format to, for example, the Laundry books ...)


    It could be that a return to curation is in order. For example, I recently read the wikipedia page on cyberpunk books and rediscovered Walter Jon William's Hardwired. While Hardwired is one of my biggest influences, I hadn't been able to remember the title or author in many years, since I had read the book in my teens. So a curated list of books in a subgenre turned out to be of big help in a way that a recommendation engine is not.

    If a series of books are related somehow, I think a curator can discover that relationship and document it.

    I write about the emergence of artificial intelligence as a response to the many great scifi books that incorporate AI but gloss over the point of emergence. If I do a decent job, then maybe some future curated list of AI books includes my stuff for the contribution they make to a particular subtopic of a particular subgenre.

    On the other hand, as recommendation engines get better, they can expose this information as well. Netflix will tell you why they recommended a particular movie, and Pandora tells you why they are playing a particular song.

    (It's interesting that they take very different approaches to the problem: Pandora songs are analyzed by experts who listen for about 400 different attributes and encode each song according to the presence or lack of those attributes. Netflix uses algorithms based on clustering of ratings that doesn't depend on analyzing the specific attributes - at least at the time of the Netflix Prize - , but I believe they reverse engineer the attributes of the clusters at the end to give you an idea of why a given movie is recommended.)

    If someone develops an open, shared alternative to the Amazon recommendation engine, then third parties could probably come up with new ways to explore and visualize the recommendations. That's something Amazon probably could do, but doesn't care to make the investment.


    Who knows - you might not see the author name or the book title on the "cover" of ebooks in a few years. In the same way you don't see "Adobe Photoshop" or "BBEdit" on their application icons.

    You might have a common Stross "brand" though... maybe some kind of young dog theme... :-)


    There's a proper Solaris translation available? Let me at it!


    I think that part of the problem is that e-library are perhaps selling e-books at the same price as a physical bookshop but when your in a real bookstore you can get an advice about what to read (and to buy).

    In an electronic book-store, nobody is paid to counsel you: there is only customers which have reviewed the book and meta-data tags. In the case of the reviews, you don't know if the person like the genre or if she has only been lost and bought the book by mistake. (The mater of trust which has already been made) As for the meta-data, i think it was a good idea at the beginning. But there is no filling system for them to classify by genre, the year it has been published, which period it is about, the subject... Moreover, there are tags meaning the same thing polluting e-library ( S-F, S - F, Science - Fiction, S- Fiction...). For example, in my calibre library, i can spend quite a lot of time editing redundant tags, renaming authors tags (because of course sometime there is only one first name, sometime several, or the spacing is different).

    To conclude, for e-book to work better, what they really need are an agreed filling system like the dewey for p-book and professional reseller and not people which sell books in the exact same way they sell clothes (without respect for the product, making it no difference wether it is a book, clothes, video-games... by the most common denominator: tha they are reselling products and nothing else!)


    That's actually a more convincing mechanism - new readers can't orient themselves around the genre because its just too damn expensive to read enough!

    I really just wanted to take issue with Charlie's doom and gloom scenario of sci-fi losing any sense of identity as a result of losing its section in Waterstones. I think Heteromeles @67 did a pretty good job of explaining why that's probably wrong.


    My small contribution is to provide small press book reviews in the genre, in Abyss & Apex ( People are hungry for information, and informed opinions about good new books will become increasingly important as more and more sludge gets self-published online. It has significantly raised our stats on unique viewers.

    We need more trusted reviewers on blogs and such. It is increasingly the way for writers to get word-of-mouth advertising.


    One thing that could happen is that the people who like the same things, end up following each other's tumblrs and twitters and what have you, and as a consequence are prompted to notice the same books. So then we have genre assembling itself by association rather than by total knowledge.


    "What is to be done?"

    Well, my money would be on curated access to the lists of the big ePublishers and sellers through referral from third party websites...


    Without a good browsing capacity it's harder and harder to find new books and new authors. In ye olde library and bookstore shelves, one literally browsed, finding all sorts of treasures, so many unexpected and unsuspected treasures. Digital browsing doesn't work that way at all. You have to know what you're looking for to find something. Browsing is about finding what you're not really looking for but knowing it when you see it! :)

    As for the present state of genre fiction book covers, never have they seemed so unattractive and so cookie-cuttered as now. I don't even open 99.9 per cent of the books that come my way, much less buy and read them. I take one look at those ugly colors of rotting algae-green and blue mud -- or their pretentious non-info all white -- and their stupid nekkid but for my weapons gals and those stupid over-thewed fellas with their weapons on steroids and like any smart horse run away run away run away!

    Love, C.


    I am quite tired of walking into bookstores,

    I don't know what the attrition rate of bookstores is, but where I live and in five (5) nearby cities, the count is zero. None. Even the wire racks are gone from the grocery and convenience stores in most places.

    Any business model that relies on "first find a book store" has some problems.

    looking at the new items, and realizing that 8/10 of them are simply awful. (or, at least, not to my taste)

    The last store, when it went away, was carrying the same "best sellers" you could get at Wal-Mart, lots of travel and cook books, and on the SF shelf, mostly shared-fantasy-world, swords and dragons, and reprints of old out of copyright Edgar Rice Burroughs. Back then, there were still shelves of "How To Run Your Pirate Copy of INSERT BLANK" computer books filling a few shelves.

    The bulk of the customers, or at least people hanging around the stores, always seemed to be slightly damp young women with long dresses and elderly women with elaborate hairstyles.

    Any book that had a cat on the cover always seemed to get premium placement, though.

    Anyway, their demise always seemed to be more due to refusing to carry anything a potential customer wanted to buy than anything else. By the time the last one died I was only going by once a year or so, if I got caught somewhere without a book in the car and it was lunchtime. Given a choice between a book and a fork, I'd take the book...


    As a reviewer and reader I see some validity in your statements, but I think the genre is more robust than you give it credit for. Yes Hollywood is creating blockbuster movies, rebooting and rehashing old stuff ad infinitum. While some people are 'not into SF' but add to the box office take on those movies, others, the truly SF-oriented, want new and fresh material, with intelligence added even if this means shock, horror READING BOOKS.

    Reading your comments as both reader and reviewer, I think I should probably begin every review with a 'recommended' 'highly recommended' etc, and hashtag for key categories #spaceopera #dystopian #fantasy etc.

    I believe the genre is more robust than you're giving it credit for, and we'll find a way to filter through the noise and find the gold amongst the crap.

    NB I read Friday when I was about 18, but I don't think I've read Saturn's Children. :)


    Even assuming that deliberate or malicious corruption of your tags doesn't take place, how many recommendations are you willing to accept from some reviewer/bot before you can tell how closely his opinions match your own?

    Heck, we already have a curated tag list. The New York Times has been rating books since forever. And their list provides a perfect example of what I just mentioned.

    I've known some people for decades, and their recommendations for a good book (and one they've read all the way through, not just the blurb and a dozen pages) is no better than 50%.

    I can do better much than that with the cover blurbs.


    I just have to add this: I read my first science fiction story, Johnny and the Space-o-tron, when I was 6, in 1950. Throughout the 50's and the early 60's, finding SF/F books was basically looking for hen's teeth. And, I was blessed to have a professional librarian for a mother, who was willing to indulge my "peculiar" reading tastes.

    In fact, to my recollection, I didn't meet anyone else who read SF/F until 1964, when I tried my hand at professional folk singing in Greenwich Village. (Unsuccessfully, I must add). Which was kind of a geek pursuit at the time.

    Thus, the irony of this discussion is huge for me. The ubiquity of "genre" literature today is one of the joys of my dotage.

    So, Charlie et al. I, for one, do not believe genre is dead. I do believe that lines are blurring. Margaret Atwood or David Mitchell anyone?

    And folks, of course there is a vast amount of time wasting drek out there. But, please remember, Sturgeon's Law ALWAYS applies.


    I'm well aware but the problem is that you then have to go through and click off everything not yours. It's as laborious as going through and saying "I already own this" on the recommendations of things you've bought elsewhere.


    I tend to get a lot of good recommendations from LibraryThing - not open source, but a good and large community of readers, many of whom tag. Charlie's "author page there" is, which contains what seems to me a plausible tag-cloud. I gets lots of "worth a look" books from the discussions there, and then the book-tags and reviews give me a next-stage filter.


    I know I'm being pedantic, but the movie's title is spelled "GATTACA".

    I recently recommended the movie to my parents, who are voracious couch potatoes. After they watched it, they were surprised when I mentioned the movie is 15 years old. Apparently they didn't notice all the clunkily outdated (but then-futuristic) computers and technology in the movie.


    The screens are big CRT boxes? A real giveaway


    Hmm ... this is interesting (and dismaying to a longtime fan) stuff.

    I'm not quite so pessimistic about the future of SF in films.

    I am not sure it is possible to write introspective, complex SF as a screen medium.
    It's certainly possible; it's been done. The remake of Lem's Solaris tried, Primer succeeded, I think. And I'd argue that John Carter would have been the start of an interesting dialog on translating pulp SF to film if it had been marketed by people with enough brains to tempt a zombie. Of course it's easier in a longer format: TV series and mini-series programs (e.g. Dollhouse, FlashForward, Lost Girl). Of course, just because it's possible doesn't mean it will be common.

    On the other hand, there is one genre which film has done well by: comics. There have been at least 8 or 10 movies in the last 10 years which have been faithful to their comic originals and have been marketed and reviewed specifically as movies of comic book stories. And now that some of stories are being remade (Spiderman, for instance), I expect to see some dialog about origin stories and characters' motivations among the different movies.

    But there was so much less SF in the 1970s that it was quite possible for those of us who grew up reading the field back then to acquire a comprehensive coverage of it. Today, there's far more stuff out there: but without the clear signifiers, the tags saying "queue here to join the ongoing conversation", it may become increasingly hard for new readers to recognize what's going on and join in.
    What is to be done?

    It may be that the problem will provide its own solution, the key is "tags". Each ebook is going to be associated with a set of metadata tags. If the set of tags is left open, rather than fixed by the vendor or the publisher (if such exists), and that it's possible to click a link in your ereader to update your local metadata store for a book from the curating site. Then a large set of possible tags and an unbounded set of tags for each book would allow searching for other books based on various subsets of the book's tags, and there could be as many useful associations with other books as the curators (or the author, or the publisher, or whoever) would like to have, and they could be updated (primarily added to, I would think; removing tags would break searches) at any time, including long after publication. So if at some point there's a general recognition of a new subgenre called "penguin-punk" an appropriate tag could be added to the seminal book in the style, Antipodes Magic, so it would come up in a search by someone who'd just read a much later example of the form and wants to find out where it came from.


    I am not sure it is possible to write introspective, complex SF as a screen medium.

    Harlan Ellison's two Outer Limits episodes come to mind, especially Demon. And there are too many Serling and Matheson and Beaumont Twilight Zones to list. No, there are not movies but they pack more IQ than all the mainstream amusement park rides -- which is what they are meant to be -- you decry.

    It's always been possible to write the SF you want. It's hardly ever been possible to get them backed and done.


    Because that's an AI complete problem.

    And not just AI complete-- It's very likely a weakly god-like AI complete problem. In my experience, my own peers are only "hit or miss" accurate about predicting what their fellows might enjoy.


    I'd be very interested, indeed, in seeing figures about the effect first the WWW ( and now ebooks are having on the market for translations.

    Amazon made it easy, or indeed possible at all, for us in non-english-speaking markets to get to most SF&F without a (usually really lousy) translation in the middle, or at least without a steep increase in retail price (think at least 300%) for Special Orders with Ye Olde Bookshop. And upped the potential field from which to choose what to read from by a couple 1,000 % (only stuff which sold really, really well in its original english-speaking market was even considered for translation).

    (Plus, TTBOMK, I've practically never seen an SF/F book translated from anything besides english (well, other than Lem), not even before the times of amazon etc. - is there so little in, say, chinese or russian, or is it that we wouldn't get so much of the underlying cultural assumptions that it just wasn't worth the effort of getting it translated?)


    In the world of tomorrow, covers will be less interesting but tags will become so. A book with a "Strossian" tag would be more interesting to than "Tolkeinesque" and infinitely more interesting than "Bodice Ripper".

    You should check out the way Angry Robot tags their books with their "File Under" tags. It's kind of what you're describing. As an example:

    ctrl-F for "file under" if you don't see it. The same is also on the back of their print versions. I have purchased more than one of their titles on the suggestive power of these tags, and generally not been disappointed.


    the unabridged direct to English translation of Stanislaw lem's Solaris is now out as an e-Book. Determining that it is that edition and not the old one requires that you download the preview chapter and find that information in the front matter.

    There's a reason the item-level descriptions are this poor and it has nothing to do with Amazon, or the publishers. Its the Library of Congress's fault.

    A few years ago, they started laying off all the skilled human catalogers who provided all that information-rich meta data in favor of a "Good Enough" (actual librarian terminology) electronic cataloging based on the bare bones CIP data provided by publishers.

    OCLC, the world leader in automated cataloging, switched over to an OCR system for adding meta data directly from publisher CIP data, then crowd-sourced what little human meta data creation it still deemed necessary to individual libraries. But they in turn followed the LoC business model and started laying off human catalogers in favor of an automated system with "Good Enough" records.

    These days few libraries have human catalogers, and those that are still around perform little to no actual meta data assignment work. It's mostly just the drudgery that can't yet be automated, like putting spine labels on books and spot checking the automated records to see that they at least attempt to describe the book on the shelf. What little original cataloging still happens is a drop in the ocean.

    60% of all new materials coming into libraries (and book stores and retail websites) has its meta data provided by the publisher, which is usually title, author, pub date, and if you're lucky, an edition number. The other 40% has no meta data at all.

    Subject headings? Edition notes? Dream on. That stuff requires humans to physically examine the item and add the meta data. And humans want to be paid cash money for their labor (the ingrates).

    (Why yes, I am an out of work cataloger whose job was outsourced to OCLC's automated system, how can you tell?)


    Charlie, I'm curious what you saw the students producing at Clarion West last summer. I attended Clarion in 2000 and most of what was being written was Fantasy and not Science Fiction. I was also surprised by how little classic Science Fiction my classmates had read. Note that I was the old fart of the class.


    It's already dead to me, alas. There are only about five writers of SF that I'll read at all anymore: You, Scalzi, Banks, Niven, guess that's it.

    It had a good run, and it's been a good life, and here we are in the 2nd decade of the 21st Century, and It Is The Future, and I now read a lot of history. 'Cuz to me there's nothing happening in this now-dead genre.


    I think the most vigorous dialogue that has been carried on in the Science Fiction field is the response to Heinlein's Starship Troopers. I think I could name at least 15 authors off the top of my head who have written differing takes on aspects of the book. I think almost all of the military SF that has been published has been influenced.

    I think the dialogue on the singularity also constitutes a considerable body of work in dialogue.


    Seems like the game theory concept of "common knowledge" provides a useful framework for thinking about this problem. Common knowledge is a strangely recursive I-know-you-know-I-know-you-know form of knowledge that arises in "coordination games".

    See "Rational Ritual: Culture, Coordination, and Common Knowledge" by Michael Suk-Young Chwe.

    Many things discussed here could serve to create "common knowledge": curated bookstores, awards ceremonies, popular blogs. However recommendations systems, being personalized, do not create common knowledge even if everyone is actually receiving similar recommendations, because nobody knows that other people are receiving the same recommendations.


    Charlie, looks like you're being 'glass half empty' again.

    Who cares if it's AI-complete? That isn't the exam question - which is rather "can you do better than the current approaches?" Compared to blindly wandering bookstore shelves (ignoring TV spinoffs), getting recommendations from others (do you trust them), or Amazon's present-unfriendly recommendations - the answer has to be yes. It doesn't have to be perfect, it just needs to have a higher hit rate. Combine actually being slightly smart about what they are reading, how, etc. and slightly smart about the content of the book (tags and text as training for recognition) you can arrive at something that does the job 'well enough' and better, more personalised, than now.

    Regards the implicit article question "can I make money off this into the future" - I'd suggest it's useful to stand back. Books are just an example of entertainment for the masses through imagination and new perspectives. That one might focus on doing that via scientifically plausible world changes - or that it might have has some back and forward with fans - kind of misses the question "what do people want in the way of "entertainment through imagination" into the future?

    Time poor, true educationally deprived, and just plain poor would seem to be the watch words for the future masses. Test-directed 'education' that misses out on building a thirst to learn more - feeding work that's "cog in the machine" disposability, coupled with Dickensian downward forces on disposable wages. It's not a great, fertile playground for a love of books. Vicariously living understandable lives via reality TV seems much more the thing.

    Consequence is obtuse and long-winded flights of unintelligible fantasy will find only small markets. Whereas easily understandable, short and derivative unshacklings of the common from the grind of life, will do well.

    As far as SF is concerned, that means near-term and less dystopian ending stories will probably do well - Stephenson might be on to something.

    It also probably means that the serial and connected short stories/novellas in a wider near-future world make sense as matches both to eBooks strength, and audience desires.

    You should do well, given your strengths.


    Comix stores are still doing well, it seems. Indeed, the ones here in NYC used to generally carry sf/f books too. They were the first to shed the book shelves, even while the chains were still running the book selling block back at the end of the 80's early - mid 90's.

    Today is national Free Comic Book Day, which is a promotional event that

    "has grown by leaps and soft-bounds, as at least a couple-thousand comic shops are expected to participate today in book giveaways that reward comics fans and, it’s hoped, convert new ones — especially the young readers. (The event is now held annually on the first Saturday in May.)"

    Alas, though, I care not at all for comix and only slightly more than that for movies made from comix (and / or games). They're too thin in thought, worldbuilding, story and character for me. Wise cracks do not necessarily equate intelligent writing (though yes, sometimes, it can).

    Love, C.


    "As a child & then teenager I'd read a book a day - if a book was not to my taste, I took it back the next day and got something else.

    Perhaps pirated (Arrr!) ebooks will fill the same role now."

    It's how I've been finding new books for some time now (because libraries just cannot keep up with ever-increasing output of publishers), and music for years before that.

    In fact it is exactly how I was introduced to Charlie's works. TV Tropes kept making reference to these Laundry books (modern Cthulhu-ish with some humour) - sounded great. But how to get them? All my local bookshops have only tiny fantasy / sci-fi sections (never a trace of Stross, or indeed much modern that wasn't already huge). The library was no help.

    Maybe the pirate scanners...

    Yup. And I've devoured the Laundry series and intend to move onto the others in due course. So, from never having heard of Stross to spending money in spite of bookshops and libraries that just can't handle modern publishing's output.

    I expect this to become normal for more and more people, just like downloading music to try it before you buy it (because so many albums are just a hit single drowning in a dozen tracks of dross, and because many music shops here don't allow you to listen to albums in the shop any more) did over the past decade.


    For those of us who weren't reading SF in the 70s, and don't have comprehensive knowledge of the genre: basically what you're saying is "we're no longer going to have lots of injokes that go right over your head".

    I find this difficult to mourn.


    Hmmm - I think you need to read some comics that aren't by DC and / or Marvel (and connected to superheroes). Particularly European ones.


    Thanks for this, Charlie. specifically the bit about Amazon being a gigantic shelf that's unusable. Several years ago Amazon allowed you to basically run SQL queries against their book listings, and someone wrote a Mozilla app that did that. Slice and dice a billion ways.

    A year after I found the app, it broke; they no longer allowed those calls. So I went looking for science fiction books published that year. After all, from the "Year in Science Fiction" omnibuses (omnibi?) I knew. I was around 300-500.

    I found 2000+. Most of which were self-pubs, with 2-3 reviews stating how awesome they were. Good fscking luck finding books on a "real" imprint, read by more than the writer's family, with good reviews and metadata on what particular form of scifi they are. So I'm back at physical and used bookstores, reading book covers, following authors' recommendations, and saving the award lists. Amazon works until it abjectly fails, at which point it never gets better. Goodreads helps a little, but not a ton. Curated lists, at this point, would be welcome.. I love Tor, but they publish a lot that doesn't interest me, though they seem to be the closest to my tastes of the major imprints. (them and Ace, I believe)


    My list was Primer, Moon, District 9.


    The Amazon and iTunes standard web user interfaces are probably good enough that no one who isn't Daddy Warbucks' favorite foster child would be able to compete with them. The mobile interfaces on the other hand (including B&N's as well), have consider suckage, so much so that a really savvy interace designer might be able to sneak in there if she's really quick. Hint: browsing should allow you a lot more precision navigation, not just give you a couple of dozen categories with listing by Author within that (that's you, iBooks).


    I'm surprised no one has mentioned Locus as a source of recommendations, upcoming title listings, reviews, genre book advertising, and all-around interesting reading on the field.


    "I am not sure it is possible to write introspective, complex SF as a screen medium. "

    Maybe, maybe not. But what's important now is that it's possible to write introspective, complex SF as a comic book medium.


    "death of genre" I think I've heard this before. never happened. SF, and boy can I remember the war over calling it that, is for people who want to read for fun. Or was. "literature" is only about the same old fallen state of man. People read it because its good for them, somehow. Can anybody say who won 5 years Nobel prizes for literature starting say 3 years ago? But the small publishers died and things did get worse. The people who pump out movies only want big hits. They also own the publishers and only want cheap sure hits.
    As for covers I read that one famous old time SF editor paid for cover art and put them behind a file cabinet. When he got a book he grabbed the first art and made it the book's cover. No matter what the art or story


    About 85% to 90% of the novels* I've read over the past three years came from recommendations online, either at the Westeros Forums or on several review blogs. Both were vastly more useful than simply browsing a bookstore and looking at back-cover blurbs, particularly since you could have a conversation with the readers and fans on what made the books good to buy. Most of the books I ended up buying were books I probably would never have heard about otherwise, including some non-SFF books.

    • The non-fiction I've read mostly comes from other kinds of blogs, and often after I've read it on a library loan.

    "That buys you, in literary terms, a novella."

    Woah, how did we end up in Nonsense Land? That sentence might as well be "that buys you, in greengrocer's terms, a banana." Or how about "that buys you, in nautical terms, a catamaran."

    Either Mr Stross has no idea what a screenplay is or how one works, or he's pretending not to.

    But I'll forgive him for the hilarious bit where he's like, "Oh, GATACA(sic)? Yeah, I guess that's a pretty introspective, complex screen medium(sic). Next paragraph: There are no introspective, complex SF screen media!"

    In conclusion, uh, something about genre? Sorry, I wasn't listening -- SOURCE CODE's on.


    If SciFi is not to die, perhaps it's time for it to have a spectacularly messy divorce and just walk away from it's marriage with SciFantasy. The dragons are dragging it down.


    Now you know that's going to start a fight... (next, the divorce between physics-driven SF and handwavium).

    I'm now running a mix of "OGH recommendation", "Mike's recommendation" (@Transreal, nearly 30 years now), and "Baen's free library". It's working, and made more interesting by the presence in the house of a ten year old image of an earlier me (loves reading, reading age now approaching adult). Except this one isn't interested if it doesn't have a Dragon in the story (no Pern yet, before you suggest it).

    So, in the spirit of vetting, I have been branching out into "stuff firstborn can read" - that's my excuse for The Hunger Games, and I'm sticking to it. It also means I can make deals like "you can have $NEXT EPISODE if you read Dune", although I got pushback after "Islands in the Sky".

    I'm heartened, not just by the raft of SF films suggested, but by "The Big Bang Theory". A show where it's OK to be a geek? An SF fan? Make in-jokes about SF, and expect the audience to get them? That's a bigger step forward, IMHO.


    It is a different way of telling a story, and the screenplay is only a part of the source code for the story, so I think you're right about it being a weak comparison, but it does look like the only part of the movie business where you can present a story, with the details of what happens, some idea of why, and an overall structure. And then the director has to interpret that, and translate it to the screen.

    In some ways the novella is a good analogy, because it can't have everything in it that a novel can. You have to choose between detailing the characters and detailing the plot, or the setting, and the author, like the author of a screenplay, ends up depending on conventions.

    But a literary novella is the final product, which a screenplay is not. A screenplay needs some slight character desription, but most of what you see in the end product is down to casting and costuming. In Casablanca, the Ferrari in the screenplay could have been anyone. We know what he is, and that might suggest things, but it was Sidney Greenstreet who made him fat.


    Are you in Southern England, somewhere near Oxford? If so keep an eye out on the Blackwell website. I suspect there may be a few positions in the near future. Pay is rubbish though.


    Just realised you're using American spelling so you're probably not within commuting distance. You could always move over here - Oxford is a nice town, if a bit snobby (& pricey).


    In a way, do genres matter to publishers as much as those devoted to one think? For a large number of people I know, they're really interested in whether something is SF, fantasy, police procudural, romance, etc, they're interested in whether the story is any good. I know people who will read whatever the latest big thing is, regardless of the genre, so they'll have read Happy Rotter, The Hunger Games, The Girl with the Noun Noun. If the book filling the popular shelf was hardcore SF, they'd have that as well. In a lot of ways, the genre is unimportant to the people who buy the bulk of the books, surely?
    I'm a big fan of John Wyndham, but I don't think it would make much difference to some of his works if you swapped out Triffids for Mongolian Hordes, which would completely change the genre.


    Dammit, I meant "not really interested in whether something is SF, fantasy, police procudural, romance, etc, they're interested in whether the story is any good."


    I do remember the first of Our Gracious Host's books I bought was Singularity Sky (combined with Accelerando? Or did I grab that with Iron Sunrise?). Anyway, that was a pick it up and read the blurb, then decide I had enough spare cash to afford it kind of deal. The range of quote sources was sufficiently diverse and respectable that I bought grabbed it. I've no idea what attracted me to the cover.

    The big killer for that is that paperbacks here in Oz are running to $20+. That's getting out of of "meh, why not" money for a lot of people, and I probably wouldn't pick it up these days. Interestingly I can order a paperback of Iron Sunrise for $20 from a local bookseller, or the same book with the cover I've got for $13 if I wait 1-2 weeks... Which makes me think they're drop shipping from the UK and pocketing the difference (it's $7 from BD). $7-10 is still "why not" money for a lot of people, being basically a "meal" from McDonalds.


    When it comes to SF there has already been one serious case of attempted murder - "Speculative Fiction"


    I think it's a fascinating point, and one quite well made - Amazon's 'recommendations' and word of mouth from friends is what's lead me to new things to read. I have run into a few 'self published' books - and the most generous thing I can think to say of them, is that I admire the author the courage of their convictions.

    But it's increasingly hard to sort the wheat from the chaff, when you have a vast quantity of stuff out there - I have a list of authors I 'know' and assume that their books will probably be somewhere between 'worth reading' and 'excellent'. New names on the other hand, are quite hit and miss - especially with Amazon, because it's definitely the case that people 'game' the recommendations system, to get sales.

    But it's got me thinking - a 'book curation network' might well be the way of the future. I've picked up enough things recently off the back of recommendations, that it make a lot of sense.

    (And I too, wish that Amazon would have a 'don't show me any more dark fantasy stories' button.)


    How to find good new books?

  • Buy award shortlists. The PKD award keeps turning up trumps here.

  • Read Interzone, then spot mentioned books somewhere.

  • Use the same bookshop as an academic reader of Genre (thank you Farah)

  • Make it clear to your local bookshop that you are quite happy to spend 50-100 every time you wander in, if they do their curation job for you...

  • 126:

    Unfortunately 2-4 in (aggregate) require you live in the UK where you're (2) not paying huge delivery prices as part of an international magazine subscription, (3) have regular access to a bookshop with that stock range, and (4) can impress them because "$50-100" doesn't just equate to two or three paperbacks.

    The only two times I've had access to a good store were Sydney's Galaxy bookshop's first incarnation before prices went crazy and it drowned in Twilight Trek, and the University Bookstore in Seattle where the indefatigable Duane Wilkins curates the SFF section so well.


    There are only about five writers of SF that I'll read at all anymore: You, Scalzi, Banks, Niven, guess that's it.

    Intriguing. One of those authors is not like the others. Quite the contrast.

    Strange however that you omit Ken McLeod from that list. Yet another of the Scottish School of SF writers. Or Hannu Rajaniemi[*]. (Sometimes it seems to me that every other person in Edinburgh must be a famous SF author, and that most of the rest write Detective Fiction... Must be something in the water.)

    [*] I'm sure there's another famous name I should mention here, but I can't quite bring it to mind. Not Paul J. McAuley, although he is an SF author that hails from around those parts.


    The problem with digital books is that you can always find what you are looking for but you need to go to a bookstore to find what you weren’t looking for.

    That was economist Paul Krugman in a Boston Globe interview today. He mentions that he reads a lot of SF, likes Charles Stross and is writing an introduction to a new edition of the Foundation Trilogy.


    Also noted the Krugman quote. Which makes it a little disturbing that even around Boston, which is as literary a city as you can find in the U.S., independent bookstores are getting thinner and more threadbare. We're still doing very well compared to most cities, but many long-standing institutions are gone, and strain lines are showing in some others. (The last time I was in Brookline Booksmith, a many-time Best-of-Boston honoree in the field, it was kind of distressing how many shelves they'd taken out to make room for racks of knick-nacks.)


    How to discover new SF talent in this digital age?

    Read the critics / reviews on the Web sites of SF publications.

    You pay if you want the all the magazine and the short stories it contains but the critics / reviews are usually free.

    I find it's a great way to discover new authors.


    If we're lucky, these conversations you refer to will see more crossover. Why not let John Scalzi write in response to Denis Johnson or Jennifer Egan? We're already toying with more than make-believe within make-believe; the conference I just returned from was a third scientists trying to get fiction writers to engage with some real world concepts. And while SciFi and Fantasy have been responding to the real world since their inception, it would be a happy turn of events if our works more frequently more seriously regarded other authors. I'm not afraid of missing the subtext text because, in our age of Google and Twitter, someone will always alert us to additional context.


    Have you watched Battlestar Galactica (the reboot)?

    It commits the usual litany of cripplingly boneheaded errors that are part and parcel of TV/movie science fiction, but it makes up for it in other ways. In my opinion, at least...


    There is such a profusion on stuff now, including Sci-Fi books, that mere categories were always inadequate; we need tag dictionaries, for tag lists and tag-value pair lists, including item number of season/series attached to items. Retailers have been doing tagging for a very long time without even realising it e.g. for department, brand, age group, size, colour, and style 'flavours' of products.

    e.g. eBook storage/retrieval software like Calibre make it much easier to manage lists of eBooks in multiple formats than a simple directory structure or fixed database format. What Calibre and much other software are missing, including web search engines, are concurrent multiple search views for n-dimensional selection criteria to help narrow down what we are actually looking for.

    Even high end enterprise data warehouses are probably dated now, because they are too inflexible; I bet even the bleeding-edge intelligence community are struggling with search now and it will only get worse as the volume of data and metadata explodes.

    I think that search itself could be/become an Art/Skill.


    I don't know if this analogy is quite valid but frankly massive crossing over between too many titles actually killed comic book reading for me until manga came around and I could read a title without buying ten different series (and attempting to track them). The fact that SF isn't doing that now is because an attempt would be a) a massive undertaking even at one imprint only and b) a huge turn-off to people who read for leisure to take their mind off their own complicated problems (and hunting down references in fiction frequently can't be done at libraries without interlibrary loans to Very Large Cities). Also, the last shared world I read a bunch of? Dragonlance. (I was fifteen and for the most part these were library books.)

    On the other hand, I did read Among Others and Fuzzy Nation and enjoyed them both quite well, having read H Beam Piper and a few Heinlein novels. The fact that there are fewer of these sorts of things now is indicative of prohibitive copyright (Scalzi needed permission from the Piper estate to publish) or a significantly reduced audience (sure the Worldcon people adore Among Others, many of them have read every single book referenced, but it isn't science fiction, it's period fantasy with a lot of sf references).


    I'm going to play devil's advocate here and try to argue against all your points -- not because they aren't justified, but because I've had far too much caffeine today. Hopefully the result will be coherent and not too ranty.

    1) The power-chords of sci-fi are in many ways a bad thing. It doesn't make sense to classify everything with prominently featured space ships or robots as science fiction (as I think you said before in a rant about Star Trek). Action movies with robots are just action movies. While it has historically been useful to have these bits of shorthand, the fact that people could easily turn these things into hieroglyphs and expect a stylized silhouette of a rocket on the spine of a book to be almost universally understood as code for "shelve this in the science fiction section" is an indicator of just how little depth these things have.

    So far as I can tell, the flooding of the market by crass commercializations with all the apparent symbols of the genre but none of the guts is not new. Maybe it's become a bit more common now that action movies in science fiction drag (and romance novels in fantasy-horror drag, and action movies in cyberpunk drag) have become so profitable, but looking through the science fiction section of a used book store yields an enormous quantity of slim volumes by (righteously) unknown authors whose prominent placement of space ships on the cover and use of terms like "groundcar" and "zeerust" instead of "car" and "rust" are the sole saving grace that put them into the science fiction section rather than the slush pile of a much more selective general-audience-fiction section.

    Perhaps this is a foolish and egotistical position to take, but I've always associated science fiction with a kind of intellectual daring and experimentation. Naked Lunch had no space ships, but was classed as science fiction because it was too weird to be placed next to this week's new best-seller in the general section. While a lot of the groundbreaking science fiction of the golden age focused on space exploration (or at least had it as a major background element), both earlier and later science fiction did not. Is it justified to stick a starship on Neuromancer? What about on Odd John?

    Using these power chords as shorthand for the entirety of the genre is occasionally useful, but I would argue that it's gotten to the point where it is far more misleading. Someone who has a strong interest in Halting State may have little to no interest in Hyperion; they are very different books with very different styles set in very different worlds, and it's questionable whether they should even be classified as the same genre.

    This brings me to my next point.

    2) The fragmentation of genres into overlapping tags is good for authors, for readers, for booksellers -- for everybody except the people who are shelving books by hand in brick-and-mortar stores.

    While the science fiction section is a ghetto to be sure, it's a ghetto of ludicrous diversity. Someone who buys science fiction off a list of science fiction in order to stock the shelves of a science fiction section will get a handful of paranormal romance, a handful of action/adventure with a nominally science-fictional setting, a handful of cyberpunk, a couple books like those of Butcher's Dresden Files series (which are a cross between two genres, neither of them science fiction, but often get filed under science fiction anyway), and -- if they're lucky -- a couple books that get filed as science fiction purely for the weirdness factor (like Lethem's Amnesia Moon or anything by Pynchon or Ballard or WSB). Chances are, anyone who goes into the science fiction section because they like a particular kind of science fiction will find nothing of particular interest, unless they are very open-minded or very easily amused.

    Part of this may be because the fragmentation of the genre, when first it gained legs, was extremely successful. Cyberpunk took off and showed science fiction that a book could be successful without a space ship shoehorned into it or a raygun-wielding pinup on the cover. The tendency for science fiction books to put in all the elements as though off a list -- this took a bit of a nosedive. (I am speculating a little here, since I cannot easily perform quantitative analysis of trends in the frequency of unnecessary instances of starships, rayguns, and robots versus properly justified instances of the same!) If fragmentation is encouraged even more, we'll still have the power chords: they just won't be used so often for the sake of classifying it into a particular genre.

    PKD is perhaps an early example of the trend I see as a whole. Early PKD stories invariably had space travel and robots, regardless of whether or not their existence in the story was justified. But, a sequence of what is often considered his best work follows a very noticeable trend. The Three Stigmata of Palmer Eldrich had space travel quite prominently, and much of it was justified. Ubik had space travel prominently in the very beginning of the book, and had no space travel at all for the entire remainder of the book, because bringing it in would not be justifiable. The Man in the High Castle had space travel as a background element, but it was not prominent because it could not be justified. A Scanner Darkly, despite being set in the far-off year of 1992, had no space travel and no robots. It was clearly science fiction, but limited the science fiction elements to those that were necessary for storytelling -- very different from earlier novels, in which robot cab drivers were everywhere and cars flew and things took place on Mars for no clear reason (sometimes breaking fairly well-defined physical laws in the process).

    But, if science fiction as a genre is not cohesive enough to merit being listed in the same section, how is a science fiction fan going to become exposed to authors they have never heard of?

    3) There is a very good reason, from the consumer perspective, to go to a brick-and-mortar bookstore and look through their poorly-curated and half-assedly-assembled science fiction section. It is the same reason, paradoxically, that it is desirable that amazon's recommendation engine not be perfectly accurate.

    I will again point out my assumption that science fiction is the domain of daring neophiles with strange ideas. Perhaps people disagree with me on this and would instead I call this by some other name -- after all, science fiction has a long history of referring to two completely different genres with occasional overlaps between them: the genre of daring neophilic explorations of strange ideas, and the genre of people with robots and space ships and laser guns. I'll call the latter 'raygun adventure' instead, for the sake of avoiding confusion.

    Someone whose primary interest is mystery novels (or whose primary interest is raygun adventure novels) may not have any particular problem with reading many takes on what amounts to the same book. A police procedural can be fairly formulaic, and many popular police procedurals are: the interest comes from the emotional drama, from dramatic tension, from not knowing the exact details of the ending, or from knowing the exact details of the ending. A book that plays primarily with the emotions of its readers can be extremely successful without giving the reader any new information, or even having any kind of consistent internal logic. Porn remains arousing even to people who realize that pizza delivery boys rarely manage to seduce lonely housewives, and The Kite Runner managed to be disgustingly disturbing despite the problems inherent in the idea that a former Nazi could be a small child in the 1970s and want to join the Taliban in the 1990s.

    Science fiction is different. The one defining factor in science fiction is that it attempts to be information-rich. The characters and plot can (and often do) hang on the world-building or the conceit. Dune drags on endlessly, has numerous minor inconsistencies, bases itself around ideas like the existence of a secret sorority of NLP-masters who seed myths and plot to control the universe... but, it's just so damned interesting when the implications are explored that the predictable and recycled plot, the dull language, the all-pervasive humorlessness, and the occasional forays into incomprehensible non-sequitor are excused. Altered Carbon has many instances of weak writing, and a plot that's fairly tangled. It has enormous numbers of scenes of gratuitous sex and violence, and many of them seem entirely out of place. But, as a thought experiment it's incredible. The fact that it features what amounts to body transplants does not make it science fiction. Altered Carbon is science fiction because it points out that the existence of body transplant technology would fundamentally change the insurance industry, make it possible for the extremely wealthy to be essentially completely immortal, almost completely wipe out the Catholic church as a political and religious power, and still wouldn't solve problems like mood swings and menstruation.

    The solution may well not be to make recommendation engines more accurate. While recommendation engines more useful the more accurate they are up to a point, to be sure, within the domain of science fiction recommendation engines are more useful the less accurate they are (up to a point).

    Science fiction as a genre only ever worked because there were hardcore fans who would read books about space ships and read books about automobile accident fetishes. A regular genre reader would not be able to push his or her way through Rucker's Software were he only interested in robot books, or space ship books, because he'd get stuck on the parts about cannibalistic cults and recreational drug use and anarchism.

    While other genres may be syncretic by accident, science fiction is defined by its syncretism. It's defined by its mashups. Anything that looks like science fiction, by definition, isn't.

    If it has space ships and robots, it's probably an action movie. Don't trust anyone over 30.


    I'm sort of shocked that the overall impression of people here is that this matter of genre and classification is one that even needs solving.

    Does anyone truly have a problem finding good books to read? Or is this an outgrowth of a natural tendency to seek a more optimal utilization of our limited reading time by spending it better? Noting of course that "better" is a word whose value of differs widely from person to person, even if both are die-hard devotees of a particular genre.

    I can say with honesty, I've never found a book to read because Amazon recommended it, and I'm not particularly interested in letting the salesman suggest such to me. If we're mourning the passionate, highly literate bookseller who can suggest books because they know you well and have an encyclopedic knowledge of literature, I'll join in, but most of us never had access to that glorious figure anyhow. We made our purchases because our friends suggested we'd like it, or we read about it in a review or best-of list, or because an author we already like made note of it. With the web as it is now, there is absolutely no shortage of books I might find interesting. I have about 70 books in my "to be read" shelves, and another several hundred on various wish- or todo-lists. I couldn't read everything I want to read even if some magical computer could identify and filter out the "chaff".

    I have trouble being sad that I live in a world where, far from wishing I could find something to tickle my fancy (as older genre readers no doubt remember well and Tim Kyger @99 feels now within genre), I am inundated with quality stuff to read.

    I've never held much truck with genre in any medium anyway, I suppose, and my distaste has only grown stronger as I age. As a kid, reading devoutly was likely to ostracize in and of itself and reading SciFi/Fantasy were further markers, so this began, arguably, with a pragmatic desire to dispense with a label that was harmful to me. If fiction were merely fiction, I could better hide my shamefully geeky tendencies a bit.

    But I soon realized that genre was illogical precisely for the reasons some have mentioned here. It's a blunt instrument at best. Genre labels have the effect of shutting out more than they keep in, I think, when a healthy dialogue between creators ought to be as inclusive as possible. The work of "literate" fiction with elements of the fantastic; the superficially fantasy novels founded on concepts that would normally be called sci-fi; sci-fi that's really not sci-fi at all, but something much more classical dressed up in futuristic clothing; and on and on.

    All of us can probably reel off a list of such "cross-over" novels we love, and what that says to me is that the quality of the ideas and the storytelling outweigh adherence to some genre tag. This has always been so, and I see no reason to suppose it'll change... the labels merely give power to the artifice that we can and should toss art into specific buckets.

    So in response to Charlie's initial thoughts, all I can say is that I hope the genre does die, as a label. All the rest of them too. I'll find things I like the same way I always have... the media (these days, inclusive of blogs and social networks) and word of mouth.

    Addendum: I will add to the voices that decry the paucity of information Amazon et al have regarding editions, translations, and so on. That is a problem that needs solving.


    "What is to be done?"

    A great many things, but from my point of view of graphical metadata (which touches marketing but also many other graphic aspects, such as independent referral and personal collection management) there's a total need for alternatives to the current visual presentation of digital books.

    One of these needs is for spine views, as something you can click to from the thumbnail view or as something which you can see in a split screen with a thumbnail view.

    I know it can be done, because I've seen so many artist Web sites which offer one or many spine-shaped banners for refs to their sites and which also use spine-shaped banners to refer to other artist sites.


    " all I can say is that I hope the genre does die, as a label."

    Amen to that. As far as I'm concerned, books come in two "genres": good, and bad.


    [ TROLL DELETED - in future, read the moderation policy before commenting ]


    This is a modern problem requiring a modern answer... yes I know it's a glib statement on the face of it.

    But what I do envisage is some interesting developments in apps that help you select the books you want to read. We'll go through a phase where pattern recognition apps will recommend books for you with by content or by style. Then there'll be the menu selection - main genre with a percentage of one genre and another percentage of a third genre. It doesn't have to be genres, it could be content. Or even by the function of the book... see my blog if you want to know more about what I mean. It would be like putting together your own recipe, but in the choice of reading matter.

    From there it's only a step away to computer generated fiction... did I hear you say I'm being a dumbo? Nah... there will still be room for a few authors at the cutting edge of innovative fiction who will add to the genre content and styles.

    In fact Charles could write a book about it... much like Ben Bova did with his novel Cyberbooks... a novel where time has now caught up with it.

    Hey, maybe I ought to sell this idea to Amazon... or did that just get me banned from Charlie's website?


    Here's a seed of an app. Take a photo of your bookshelf and feed it to the app which then does a lookup and deduces what you liked, and what you would like. Like Amazon but with a much bigger initial database.


    Reply to Dirk Bruere...

    Oh like, like, like... only one problem... it would add in maths texts in my case! Can we work together on developing the requirements?

    Um... this may be a strange coincidence, but there's a computer geek in my novel called Dirk... any relation?


    Due to my own personal incompetence and lack of time, when I recently moved house (by train) I had to leave all my books behind in the old flat.

    So my shelf only has "Lud In The Mist", "The Night Climbers of Cambridge" and the guitar score for John Mayall's Blues Breakers LP.

    It's quite refreshing starting again, but now I'm going to think more than twice before buying any book in paper form.

    Out of interest, given those 3 books what are the likely recommendations for a fourth?

    Would the app be more accurate if it had more complete data than just books - a list of all the physical objects one owns?


    I suppose a dumbed down version would just be entering a list of books and authors the old fashioned way. The problem (for me) is that it would require some sort of script for accessing server side information (PERL?) and I have no expertise in that.

    Apart from that I assume at worst it would be searching for key words in online text or book reviews of the existing list and looking for matches in unread stuff.


    Even better, we will have apps that will not only work out what we like and write books for us, they will read them for us as well, leaving us free to do important things such as decide which apps to download. In the future.


    I like being read to! What I need is a good, free, text to voice program for my PC


    Given the three books in your library, the recommendation would be for the books in Alan Dean Foster's Spellsinger series.




    Reply to Dirk at 145...

    Why bother when they all have ISBNs? Quicker that way.


    And bar codes.

    @Frank, #148: Thanks, I'll give the series a look.


    Charlie - That's the theme of David Weinberger's "Everything is Miscellaneous" : Physical objects must follow the "one place for everything and everything in one place" rule that begets hierarchical treelike classification systems and the Dewey Decimal System, and professional classifiers like Librarians. Digital objects allow many to one mappings and multiple classification schemes. ( I don't recall whether Weinberger delved into the other side effects of the proliferation of tags that has caused the semantic web folks to consider ontologies and restricted vocabularies as a way of ordering the miscellaneous pile again. )

    The video of his you tube talk: misses much of the geeky details of the history of classification schemes.

    He has a blog at:


    French porn fairy tale comix can be fun!

    But I have looked at many different kins of comix -- you must if their editors are your friends -- and it doesn't work for me.


    It sounds like you're not mourning the death of the genre so much as the death of the shibboleth.

    I'm all in favour of books being written in response to each other. It's one of the habits that SF shares with academic writing, and it enriches the entire conversation as a whole. But that conversation only holds meaning for the people who are truly interested in listening closely or deciding to participate. For everyone else, it doesn't matter how many references to another book that you throw into yours. All they want is a good story. And including too many in-jokes results in excluding people from the conversation, rather than including them in it. In the long run, that exclusivity stifles the development and sustainability of the community.

    You have to take the bad with the good. You have to believe that for every ten people who thought The Green Lantern was great, there's one who was moved by District 9, and that some percentage of readers who love the books that you despise might also be willing to give your own works a try. Otherwise, you're just trying to find an ideal reader. And there's no such thing. More importantly, the reader whose experience is limited to a single genre, who has to pick from limited offerings, who struggles to find a community, is not having an ideal experience. I've heard what it was like to be a genre reader in the '70's. I wouldn't wish that on anybody.

    I say this as someone who filled her own robot novel with references to Asimov and Dick, and also Evangelion and The Red Spectacles. (If you must acknowledge your influences, acknowledge all of them -- even the ones some of the audience don't know.) Those jokes are there for everybody to figure out, if they want to. If they don't want to, that's cool. I wrote them before I knew the MS would ever be bought. I'm sure I'll include others in the sequel. That's how my sense of humour works. But I don't demand that each reader understand all of them, or count them as less of a fan because they don't. Their engagement with the community is up to them, not me.

    And readers already know this. They communicate about books (and media of all sorts) via their own channels. Some of them love Amazon. Some love Goodreads. Some love LJ. Some love actual, regular, in-the-flesh book clubs. All of those are valid solutions to the problem of finding good books. And all of those practises build communities -- not of genre readers, but of readers. What's to be done about that? Nothing. At least, nothing more than what you're already doing -- building your own community, right here.


    And we can design the voice to suit the listener's mood as well... talking book newly written just for the listener...

    This really is the stuff of science fiction...


    You don't want to purchase the 'Anniversary Edition' of Dune anyway, because the ebook at least is a really bad OCR job, by far the worst I have ever encountered from an allegedly professional publisher: the first page avers that "Paul was bom on Caladan", and blatant OCR typos continue at the rate of two or three per page, with multiple missing para breaks, big chunks of missing or transposed text, and failures to switch into or out of italics fouling things up even more badly. It is plain that nobody read even the first page of this thing before putting it on sale.

    What grates particularly badly is that if the publisher had just sent an old NEL copy of Dune to Amazon and said 'chop this up and turn it into a Topaz ebook' the result would have been ever so much better. ebookifying old books with no electronically-ready copy is why Topaz exists. But, nooo, they had to do a really bad job of the conversion itself.

    I bought an old NEL physical copy instead.


    Yeah, I saw that Krugman thing today too. Clearly neither Our Host nor I are the only potential book buyers and readers frustrated by this situation.

    Also the obscuring of anything that isn't a best seller by all the self-published stuff, make it a nightmare -- just like reading book reviews at the UK Guardian is getting to be, since they publish more reviews by kids than by mature, professional writers. To read over and over "This book is very good. Everybody should read this book. I loved this book." This isn't helping me.

    So many websites that review materials write like this with the addition of lots internet teenage fanboy/girl speak, jargon and snark. That doesn't help me either.

    I want to read a high recommendation that I spend my money on something to read that is written in organized, professional prose. Otherwise I am not convinced. At least I'm no longer convinced. I've been burned too often.

    So many websites review all the same four titles and squee squee squee over them all the way home. When you yourself are persuaded by everyone's claims This is the Awesomest of the Awesomest Most Original! Different! Better! and buy the book, you then find it is in reality just more of the same and fairly pedestrian. A condition of distrust and cynicism comes to fore, the condition to which I am now habituated.


    Many SF authors belong to the sfwa.

    I expect many authors expect to make money writing books.

    Go to the sfwa web site.

    Look for a link to where you can buy books from sfwa members, or links to their books at their publishers, or even Amazon or walmart-com.

    I'm not seeing any likely-looking links there.

    I didn't bother tracking down groups representing crime, romance, western, or other genre authors, but "buy our books!" is one of the things I'd expect to see there. If it's not coming down to "buy our books!" somewhere along the way, why bother?


    @132: Have you watched Battlestar Galactica (the reboot)?

    It looked like they taped the camera to a basketball, gave it to a former Harlem Globetrotter with cerebral palsy, and let some autistic crack monkey loose with the special defects on what came out.

    Fortunately, I was able to kill it before barfing.

    Perhaps someday someone will code a "software Steadicam" I can pipe .avi files through. Until then, much of the visual media of the last ten or fifteen years isn't something I can watch.


    In some ways the novella is a good analogy, because it can't have everything in it that a novel can.

    "I'm sorry I wrote you a long letter, because I didn't have time to write you a short one."

    With a megabyte or more of bloat room, there's no reason for tight prose, so I keep encountering books that just flail around drowning in their own words.


    Just read Blindsight on Peter Watts website, mainly due to the recommendations I saw in this blog. Since its free online, and exceptionally good, I sent him $2.


    I've got another e-book copy of Dune that has been rather well checked for OCR errors (and I know it was done by OCR, because it ain't, um, "official", but it is much more convenient than the brick of pulp I have been used to).

    One of the many, many things modern publishers need to do better. Do they really not see just how sad it is that they are actually outdone by pirates?


    There are some pretty blatant cases of incompetent search design I have come across, such as not being able to distinguish between "Robert Graves" and "Five Graves to Cairo".

    I suspect some outbreaks of not-invented-here syndrome.

    The trouble is that, where I have come across such things, it's a site-specific engine, and they have locked out Google's search spiders.


    I think one solution is less granularity. As Charlie's noted, therewas a time when it was possible to read and become familiar with everything the field offered. During that time there was no such thing as this, that or the other sub-genre. It was all just science fiction (rocket and planet logo on the spine of everything from Asimov to zelazny) and it was a question of liking or not liking - not a question of pre-determining and then deciding on likeness. We need to dump the sub-genre stuff and get back to liking or not, based in one tag - the SF tag. It will be much easier to sort it out. Taking a higher viewpoint blurs detail, sure, but also allows one to survey more.


    "a couple books that get filed as science fiction purely for the weirdness factor (like Lethem's Amnesia Moon or anything by Pynchon or Ballard or WSB)."

    Lethem wrote purer SF before so Amnesia Moon so that title followed the first books onto the SF shelf. When he broke out with the crime-ish Motherless Brooklyn, I noticed that everything got re-shelved into crime.


    Recommendation systems work by finding a sample of people with very similar taste to yours, and then picking new suggestions for you out of the pile of things that they liked, but which you have not read yet. - They dont need to understand what they are recommending at all, knowing that the list your list of "Best Ever" and "verily; doth suck" is indistinquishable from that of Dorothea Ratheraa from the congo, and that Dorothea also liked "Martians With Great Big Guns and Superior Economic Theory" a bunch is grounds enough to recommend it to you.

    Amazon's algoritim is limited in its power by the fact that the only datapoint they have on most items is a binary : "D bought it". This doesnt let the database distinquish between impulse buys you came to regret, and books you reread until they fell apart. In order to really optimise recommendations, an ebook vendor needs people to volunteer their opinion on every single purchase they make. Honestly, I would suggest offering a 50 cent store credit rebate for rating a purchase. (only on books you have already bought. To avoid people spamming the database with fake ratings for the credit)


    "the utility of the old signifiers—the rocket ship logo on the spine of the paperback—diminish in the face of the new (tagging, reader recommendations, "if you liked X you'll love Y" cross-product correlations by sales engines, custom genre-specific cover illustrations, and so on)."

    This certainly rings true for me. I buy mostly ebooks at this point, and most of my shopping is done by recommendation, mostly by friends, but increasingly by these sales engines. They're hit or miss. For every "Prince of Thorns" there are two Tramp Stamp Paranormal-Romances.


    "...therewas a time when it was possible to read and become familiar with everything the field offered."

    That was also true of computing up to around 1980/85. Any link?


    Assuming that a recommendation system has: (1) Books you've purchased (2) A large set of buyers who've purchased 'similar books'. (taste group) (3) A subset of recommendations from buyers in your taste group covering all books.

    It should be possible (and actually pretty simple) for an online bookseller like Amazon to provide an usefully curated experience.

    As an aside, one method that may substitute for recommendations is checking whether one book from an author or many books from an author are purchased.

    There will be an internet bubble effect, but, other than that, this system will probably be better than bookstores. The limitations of physical space are difficult to overcome.



    Not denying ebooks are the coming thing, but a large part of the point of SF in my youth (70s) was the magnificent cover art employed to shift the stuff.

    WH Smith stacked SF titles on wall racks, cover out, because the stuff was wall art with (hopefully) a great story underneath it. Foss and Pennington were favourites of mine, and their work was usually layered over worthwhile content. Hell, I still prefer paperbacks in part because they are more portable than any other form (yes any other form) and in part because of the artwork.

    And it wasn't all rocketships and blasters in the 70s either (though the Foss "junkpile" spaceships were innovative and much copied - even today - eye-heroin for me): The Panther covers for Asimov's work are a case in point - black with a simple photographic motif.

    The moving away of music and literature from having a cover design has diluted the pleasure for me. To walk through a record store in the 70s was to visit an art gallery. Ditto the sf stacks of WH Smith.

    What do I care if the art was a manipulative move to get me to buy product? I enjoy looking at it above and beyond the enjoyment of whatever it is wrapped around.


    This is a fantastic essay, which raises a lot of interesting questions for SF/F/Speculative fiction/etc.

    I think Charlie is right on the money in a lot of ways, but I disagree with his conclusion that new models of curation spell the end of genre, or that cross-genre marketing and the end of the genre-ghetto are the likeliest outcomes of the changes that the publishing industry is currently undergoing. Rather, we might see intra-genre ghettoization.

    A more detailed response here:


    Jonathan, I'm a farmer, and that bit about statistical analysis to determine genre sure looks like BS to me. I reckon it would be as productive as trying to milk a bull.


    I agree to a certain extent, the changing of p-books to e-books as a main source is changing the face of literature. I also agree with the point of consumption argument. However, Sci-fi as genre, or any genre for that matter, will remain. Most assuredly it will evolve, everything does, it has to and the more imaginative the genre, the easier for it to change. By it's very nature, literature resides in imagination. All literature. Shakespeare's writings were packed for mass release and still are used as templates. Because they followed universal templates hacked out before them. As a race, we respond to some basic motivations. Literature, for one, exploits them. The genre is just the way the story is wrapped up and delivered. People like nice packaging and respond to it. Therefore, "genres" will always exist. Not the same as they were in the 1800's, the 1950's, the turn of the millenium or even now. But they will always be around. We all like the wrapping paper, it's in our nature.


    Please do not feed the narcissist. (Note that his comment has been unpublished.)


    Broadly agree. Humans like to categorize. The ways they categorize may change, but I think it's usually more evolutionary than revolutionary. We may see a broadening of what's considered SF/F, or we may see it shatter into constituent niches, and see less and less conversation within what we, today, would call the "genre." Hard to tell which one wins out. Maybe it's both?


    I'm wondering what his namedrop quotient was this time...


    Re 171, 173, my personality is not the issue. I apologize if my attempt to constructively contribute was misunderstood as mere self-promotion. I was pointing you, since you brought up "genre", to significant scholarship by others.

    THE MECHANIC MUSE What Is Distant Reading? By KATHRYN SCHULZ The New York Times Published: June 24, 2011

    "Franco Moretti has a solution: don’t read them. Moretti is not a satirist. He’s an Italian literary scholar and the founder of the Stanford Literary Lab, which opened last year, published its maiden pamphlet in January and followed up with another last month. The first pamphlet asks whether computers can recognize literary genres, and the second uses network theory to re-envision plots.... The Lit Lab seeks to put this controversial theory into practice (or, more aptly, this practice into practice, since distant reading is less a theory than a method). In its January pamphlet, for instance, the team fed 30 novels identified by genre into two computer programs, which were then asked to recognize the genre of six additional works. Both programs succeeded — one using grammatical and semantic signals, the other using word frequency. At first glance, that’s only medium-interesting, since people can do this, too; computers pass the genre test, but fail the “So what?” test. It turns out, though, that people and computers identify genres via very different features. People recognize, say, Gothic literature based on castles, revenants, brooding atmospheres, and the greater frequency of words like 'tremble' and 'ruin.' Computers recognize Gothic literature based on the greater frequency of words like . . . 'the.' Now, that’s interesting. It suggests that genres “possess distinctive features at every possible scale of analysis.” More important for the Lit Lab, it suggests that there are formal aspects of literature that people, unaided, cannot detect...."


    What? That's almost completely random. Is that a serious recommendation or some kind of joke I'm not getting. If it's a joke, please explain it to me as to a five-year-old child. I've been known to be too serious.

    Lud-in-the-Mist is a thoughtful piece of social commentary that deals with the clash between artistic and mundane life, with coded themes of alternate sexuality (by which think Rosetti's Goblin Market; it's all very metaphorical but then it did come out in the 1920s.) It's a damned good book, too. On the other hand, I read the Spellsinger books when I was a teenager and did everything on the SF shelf in my local library for lack of alternatives, and I really could not recommend them to anyone in good conscience.

    If someone liked Hope Mirrlees, I'd probably send them more towards Octavia Butler rather than Alan Dean Foster.


    I must not post when suffering from insomnia. I must not post when suffering from insomnia. I must not post when suffering from insomnia. I must not post when suffering from insomnia. I must not post when suffering from insomnia. ...


    @Adrian 'I don't think that the industry has figured out how to design covers for ebooks yet - which is why they're mostly lousy.'

    They're mostly lousy because many publishers/authors/sales depts and the cleaning lady seem to think they know what constitutes a decent cover image. Visual communication is another layer to the text. Usually, when the artist/designer is left to do what they know, the cover images work. When left to a committee.. well.. I'll say no more.

    Not to mention indie authors now crowd sourcing artists for $150 to do a cover, or their cousin knows someone who knows a little photoshop. You WILL get what you pay for!

    The current crop of professional artists are more than capable of creating cover images that work in the new e-format.

    Whether anyone cares about cover images is another matter entirely.




    Mr Stross, you make a lot of the role of e-books in the decline of genre and I only have my own anecdote to rebut this.

    I formerly lived in Silicon Valley, went to most of the big bookstores, at least 1 visit week and $20-40/week habit (not all fiction of course). Starting about 8 years ago - but reaching an intolerable level 4-5 years ago - the SF/Fantasy genre in the big chains (esp. B&N, Borders) morphed. The SF/Fantasy section became 95% the following: either established authors (who I don't need nay help finding, thanks), and vampires. Lots and lots of vampires. Some werewolves. I've lived my life browsing shelves looking for something new (my threshold for trying something is low) and in the span of a few year this became annoyingly inefficient (nothing against vampires, just not my taste). I was an addict and gave up general bookstore visits for genre fiction before (no, not far before, but unequivocally before) e-books had any major relevance.

    I don't know what my larger point is, but the idea that genre fiction is inherent in the practicalities of selling p-books in bookstores - I now doubt this. It's not consistent with my own, thus anecdotal, observations. Not sure what I have to offer instead though. But there's something more going on than e-books.


    @Stevie 'What do I care if the art was a manipulative move to get me to buy product? I enjoy looking at it above and beyond the enjoyment of whatever it is wrapped around.'

    A lot of that art wasn't just churned out because that's what the artist was told to do, or simply for a pay check (although that always helps!), it was lovingly created out of a passion for the genre. In books and album art.

    The artist Rudy Gutierrez talks about the artwork on the covers of Santana albums as being the visual accompaniment to the music. "It seemed like the music went through the room, circled the room, messed with everybody and then landed on this album. The visual was so close to the sound".


    Just saw this on the news, maybe using AI to solve the genre classification problem is not that far-fetched.


    A lot of musicians are performance artists. They're used to putting on a full show, complete with visual elements. This means that even if they have relatively small budgets they take the trouble to find decent graphic artists, and the time to reach an understanding on the visuals they want, for the album cover, for the alternate album covers, for decorating their Web site, etc.

    In contrast authors just try to see to it that they won't provoke epileptic fits when they go to a convention. They avoid combining stripes and plaid when they dress. And some dress in black to make things simple. And that's all.

    So, here we have a musician surrounded by some authors:

    There's a lot of work to be done.


    What is to be done? Tor needs to keep paying Jo Walton to review books and get their other authors to join in. For every book our host here writes, I'm sure he reads a few dozen and can tell us which ones were good. Thanks to Jo Walton, I'm reading Debt: the First Five Thousand Years and having my mind expanded accordingly. This is the curation we so badly need, and those wh brought us cons and fanzines are the ones to bring it to us all again.


    Looking at the cover illustrations for SF books and magazines has always been fun for me; there are at least half a dozen illustrators who are or were extremely good artists, as well as good at illustrating the story. I have books and prints of the works of Kelly Freas, Hannes Bok, Michael Whelan, and Leo and Diane Dillon, and I go back and look through them every so often, probably more often than I look through the works of Picasso or Braque (picking from approximately the same century).

    If SF and fantasy cover illustrations go away because of ebooks I will be very sad.


    Thanks for the reference.

    The apparent difference in the patterns the computer uses, and what would be obvious to a human reader, makes me wary of that result. There's quite a few cases of pattern recognition classifying by the wrong differences in a small sample. As I recall, one of the early ones was an attempt to recognise AFVs in photographs. The sample set used US and Soviet vehicles in rather different environments, and the system was classifying on backgrounds.

    So I'm still sceptical.


    I've just found a website that groups similar authors in a sort of map - I thought it might be interesting to people here, and it seems reasonably accurate.


    It certainly "sort of works"; It does come up with some "never heard of" or "read one and under no circumstances..." suggestions though.


    It's rather polluted with semi-duplicates, inverted and otherwise broken names e.g.there's a C S Lewis, and a C.S. Lewis. One of them is apparently a cross between Enid Blyton and Monty Python. I must have read the other one.


    Netflix is the best digital media storefront I have encountered thus far... though it is, of course, not actually a store. It just wants to convince me not to cancel my subscription. Perhaps that is the attitude one should cultivate? Keeping people in the store is a strategy for physical retailers as well. Netflix has me adding stuff to my queue all the time. I can quickly scroll down to get to a dozen or so recommendation lists of different types. It asks me to kick things I have already seen out of the list, and to submit more information for better recommendations which I want to do because the recommendations are good. It mixes in newer stuff as it shows up. It makes me want to watch movies.

    Steam would be second, with good genre support and recommendation of newer items. Sadly they have removed their one bit of rating info, used to have metacritic scores you could sort by.

    Amazon seems fairly bad. I'm not a big user though so it's recommendations aren't powerful for me. I tend to go there knowing exactly what I want... but finding that is still something of a pain.

    The Apple App store is really quite bad at showing me stuff I might be interested in. The crudity of a 'top sellers' category, as one of only four broad search types is a bit unbelievable. I think this really has them stuck in lowest-common-denominator territory.


    Sorry, but that was actually Phillip Jose Farmer having a bit of fun with Vonnegut's character. If you don't believe me, you can check the entry at the Library of Congress authority file for Kilgore Trout. (


    He need only follow the link from Trout's name in the article he himself linked to.


    Some things that I think will happen:

  • conversation will happen in random clusters, often sub-genre but often cross-genre. Yes, the newbie hordes make it harder in some ways to find the people you'd enjoy talking to most, but I think we're in a temporary mass-media-forcefit transition period that is going to keep dis-integrating. There will (continue to) be secret email-lists.

  • genre-specific, or maybe more audience-specific, imprints will become the rule for branding. Tor has a following, Macmillan does not.

  • more authors will run online blogs and forums, so there will be plenty of free-tastes available to aid in discovering new authors/books to buy/love.

  • More people will use social-reading tools, esp automated tools, so you'll find out what your "friends" are reading, which is always a good lead, though not necessarily super predictive... (but certainly more predictive than "who paid for the end-cap" in the current model).

  • 195:

    Genre is real, I would argue, in the following sense: given a matrix of the degree to which each person likes each book, very simple algorithms (k-means, fuzzy k-means, non-negative-matrix factorization, etc etc) can extract a set of categories. I've certainly seen this in the Netflix data set, I doubt books are different.

    This is data heavy, algorithm light. There's so much data that we can say with confidence that one simple algorithm is better than another by how well it explains the data, and we can hope to obtain an objectively best-possible simple algorithm. The algorithm doesn't have personality as an Artificial Intelligence might, or, to say it another way, there's no need for subjective or cultural interpretation. There is a best choice of parameters for the algorithm, based on the data itself.

    (If we want to understand the fine grained structure of the data then we would be getting into creepy AI territory or into people who know each other's likes very well, but "science fiction" is a broad category.)

    An interesting corollory of this is that genre labels or tags apply equally well to people as they do to books. (Or at least people at a given moment in time. Using time did give an increase in accuracy in the Netflix data.)


    The games industry has survived the transition from physical to digital retail. An A+ publisher can put out their $60 game with brand recognition, while an indi developer can put out a $5-$15 game that people can play it without being burned. In the same way an $8 book from an A+ publisher from a supported writer is going to get more attention compared to an $2-$4 'indi' book from Jack using Word98. It also opens up 'episodic' books. Jack can put out a few chapters of his epic story for $0.25 every month, giving people an ongoing story that they can continue to read and enjoy far more regularly than the 1 book a year contract that many writers get. They have the freedom to write what they want, when they want and readers get to read more often (though perhaps less reading). It's not a direct parallel. There are far more books a year than games and with video/screenshots you can see exactly what kind of game you're getting before you buy it. Even getting the first 20 pages of a book (like a game demo) wont narrow down the 100,000 sci-fi books you'd have to go through to find something you like.

    It's going to come down to support from a publisher and/or support from readers with a huge refinement of the genres to move books. By genres I mean Sci-fi>adventure>15 minutes>distopia>world-war or fantasy>romance>vampires>books-for-girls-who-can't-find-a-guy


    With the collapse of a major bookchain, Borders and Angus and Robertson, we found ourselves (in Melbourne, Australia's outer-east) with no retail bookshops within 40 minutes of us. The collapse also took down my favourite boutique/retail bookshop which luckily a wealthy entrepeneur decided to invest in to bring it back from the dead. Unfortunately...I've moved on, and largely buy ebooks.

    A simple classification for Amazon would be to detail which books have been self-published, which have not been properly edited...and promote them in a separate category from the major authors/publishers. The amount of dross I read before I cottoned on to the fact that Amazon was pushing rubbish on me is embarassing! I agree that metadata, by trained and experienced classification/genre people (not people who put Ian Banks, rather than Ian M Banks on the science fiction shelves), combined with your standard reviews is an approach that Amazon should make.

    Something like Metacritic for publishers/authors, which might consolidate reviews of science fiction/fantasy into a single place for ease of use.


    Try finding Banksie in Scotland - Any title, with or without M, can appear in any of "general fiction", "science fiction and fantasy", "Scottish fiction" and "Scottish general". Yes, I have seen "Raw Spirit" in SF&F!!


    Coming in way late (it's finals week for us again), but I suspect the death of the genre is something much more fundamental.

    I sometimes like to refer to the 20th century as "The Century of the Spaceship", if that tells you where I'm coming from.[1] There's also the fact that when I started reading the stuff the average age of the typical reader was probably under 25, whereas now it's over 40.


    [1]Iow, all those tropes that were still reasonably fresh in my extreme youth are now getting very long in the tooth. Publishing trends won't change that essential fact.


    @Alain '..In contrast authors just try to see to it that they won't provoke epileptic fits when they go to a convention'.

    Nail.. meet head!



    @Bruce 'If SF and fantasy cover illustrations go away because of ebooks I will be very sad.'

    Amen brother.. although.. If eBooks are going to strip the artwork from the book, then the art is now free to explore sf in ways it was unable to do before.

    This opens up a new avenue for sf artists to explore visually without the publishers brief and/or committee dictating bad artistic direction. Sf art no longer needs to be shackled to a book as pure packaging. How this translates to the real world, I'm not sure yet, but I am trying! ;-)


    I don't think SF is going to die, it's just going to be called non-fiction.


    Even with a lifetime LibraryThing membership and a bar-code reader in my position for over a year, I still haven't found the time to catalogue the bulk of my library (partly because of the other stuff stashed in there as well physically limiting the access...)

    Also, the problem is the older the edition you have, the less likely is the LibraryThing software to come up with an immediate hit - and this doesn't take into account differing ISBN standards (where there is an ISBN in the first place).


    I found Walton's Among Others boring, trivial and occasionally stupid. Maybe that's just me but I honestly can't see what everyone is so excited about.

    Vote for it if you want, but I can't escape the observation that you're voting for your own childhood, not for the book. It really is pretty lightweight.


    It is just you. Sorry.




    And it's won a Nebula too.



    About this Entry

    This page contains a single entry by Charlie Stross published on May 5, 2012 11:44 AM.

    And another thing ... was the previous entry in this blog.

    AFK is the next entry in this blog.

    Find recent content on the main index or look in the archives to find all content.

    Search this blog