Back to: Not long now ... | Forward to: Why your internet experience is slow

Bang, Bucks, and Delivery in Recompense

It's a truism of the writing business that short stories are not like novels. There are any number of novelists who simply can't work effectively in the cramped space of a short story; and there are many writers for whom the short form is their natural métier and the wide vistas of a novel seem impossible to fill, an invitation to agoraphobia.

At first sight, this shouldn't seem very odd. We slice and dice fiction into categories based on length. You, if you're a reasonably fast reader, can probably scan 300-400 words per minute; a page in a typical mass market paperback contains roughly 300-400 words. So now let me define some categories ...

A piece of flash fiction is under 1000 words — a two minute read. Real short stories are typically defined as 1000-7500 words; an experience that lasts three to thirty minutes. Above that end we segue into the novelette, which at 7,500-20,000 words can take anything up to an hour to read. Beyond the novelette we have the novella, the shortest form which is substantial enough to be printed and bound as a slim book, and that runs from 20,000 to 40,000 words in length — an hour or two of distraction. Anything over 40,000 words is described as a novel, although the length of a typical novel is subject to fashion; in the 1950s a typical SF novel was 50-70,000 words (70,000 being at the longer end of the scale), but today a typical SF novel is in the 100-110,000 word category, with an expected reading time of 5-6 hours. (If you've ever wondered why you seem to read fewer books these days, this may be why: the books are longer!)

Various different ingredients go into a work of fiction, and we expect to find them all in a novel — details of characterisation and background, description of the environment, matters of plot and theme, subtle allusions and symbolic references, depth and colour.

But as we go down the length scale towards flash fiction, we find we have to leave out more and more aspects of general fiction, and focus on that which is important. A novella doesn't have room for the complexities of plot and subplot that would fill a novel — not without leaving out something else, and it is a curious fact of fiction that it's difficult to hold a reader's interest for more than a double handful of minutes without some focus on character development. Novellas are the serialised, literary equivalent of a two hour feature movie script: the classic movie script run to 100 pages, with perhaps 100 words of dialogue to the page, and this isn't a bad recipe for a novella — just add framing narrative to taste. (This is one reason, incidentally, why SF novels don't adapt easily to the cinema; it's been said that the only really workable version of David Lynch's Dune was the nine-hour unedited cut which, alas, has apparently been lost: the exigencies of a 150 minute cinema slot butchered the novel almost beyond recognition.)

By the time we narrow the focus down to a novelette, the pretensions at plot or development that go into a novella or novel simply won't fit; we've got room for a character sketch and an incident but no expansive context to anchor it. And the short story itself is even harsher, flensed of all extraneous verbiage: there's room for a single big idea or emotional resonance, and all else must be subordinated to it.

But what happens if we move in the opposite direction?

The novel is an ill-defined form because it conceals a multitude of different forms. It's to some extent a marketing category, defined by the idea of a book which contains a single story. But we know better. Sometimes big books get chopped into two or more segments for commercial reasons relating to printing and production costs or marketing. And sometimes a work of literature won't fit neatly into a single novel — the author has to keep coming back to re-examine the same matter from different angles, or is telling a story in multiple natural segments.

I've just finished the first draft of a segment in a much larger work, the Merchant Princes series. "The Revolution Business" is book #5 in a sequence, and while it's indisputably a novel, it functions much as a chapter does — it picks up threads established earlier in a story, spins them out an embroiders or otherwise develops them, brings them together to a local climax — and leaves them trailing off into the future. (One of my next jobs — after I hand this one in — is to begin work on the sixth book in the sequence, which will hopefully close off a huge hank of these threads for good, concluding a major story line.) It's got me thinking. What is this: a chapter, or a novel in its own right? I'm not entirely sure, or even certain the question makes sense; in a very real way, this multi-book series is self-similar, with sub-elements sharing the same fine structure as the greater whole.

(Or maybe the mega-novel simply isn't my optimum form. It certainly feels like a grind right now, with 527,000 words written and about 100,000-110,000 to go: that's an estimated reading time of 30 hours, give or take, at 350 words per minute, and a length of roughly 1900-2000 pages.)

One thing's sure: our willingness to absorb fiction is geared to our attention span. A book in a series might structurally serve as a mere chapter, but it has to deliver as significant a cognitive reward to the readers as any other 300 page book; otherwise they'll feel cheated, after putting 5 hours of their life into reading the thing, and they won't pick up the next volume. If writing a novel that runs to more than about 120,000 words, it behooves you to put in a climax (be it emotional or conceptual) every 60,000 words that gives the reader some sense that the work is progressing towards a final, haze-shrouded summit in the distance from which they will be able to look back and grasp the distance they have trekked across the landscape of your imagination. Otherwise they're going to spend the last 500 pages of your series screaming "get to the bloody point!" inside the privacy of their own head, then throw the book at the wall when the vast revelation promised over the preceding volumes turns out to be a minor footnote. (As, I regret to say, happened to me, when I finally slogged my way through
The Baroque Cycle.) Indeed, not doing that thing is going to be my obsession for the rest of the year, as I try to nail down "The Trade of Queens". Because? If your novel is six times as long as normal, it needs to deliver six times as loud a climax.



I recently purchased the first four books in the Merchant Princes series and biltzed through them in about just over week (while holding down a job, etc. :-)

They are very enjoyable and I would highly recommend the story, but I would have preferred to discover the whole thing in two years' time when it's finished, and then read the lot and reach the end in a reasonable time. (In fact, if Stephenson had brought The Baroque Cycle out in nine parts, in the way it has since come out in paperback, I would have found it extremely annoying).

I can't help thinking this is a heinous plot on your part to sell hardbacks. You can count on hardback sales (pre-ordered) of parts 5 and 6 to this reader.



John: actually (I don't want to put this on the front page ...) I originally handed in a 600 page book, and the plan was to follow it with an 800-page book. But Tor split the first volume in two, then told me "three hundred pages per book", and in the fullness of time the second Big Fat Book put on weight (and is now looking to run to 1500 pages, in four volumes).

There's a multi-year disconnect between the signals that tell authors what length is in fashion for books these days (i.e. what's on the bookstore shelves) and what the publishers are actually buying, and by sheer bad luck I began work on a Big Fat series just as the New Anorexic came in.


Definitely right about "chapter" books. I've always been resistant to n-ologies for just this reason. (That and the fear the author is going to die on me in the middle.) It's why I never finished Wingorve's Chung Kuo series.

I haven't read any of the Merchant Princes because I want it to be over before I dip in. Then I'll probably read it in a rush. So don't die without finishing it.

OTOH, I loved the rambling doorstops of the Baroque Cycle. For me, it was all about the journey rather than the destination.



JDC: books #1 and #2 work together as a single story, so they're safe to read right now. Book #5 is nearly finished (current draft just needs some polishing before I hand it in) and when Book #6 is done, books #3-#6 will tell another self-contained story as well. So you might as well read #1 and #2 now, and wait until #6 is out to binge on #3 to #6.

(I plan on taking some time off after #6; if/when I pick up the world again after that, it'll be a new story involving mostly-new characters.)


I liked the Baroque cycle, though I felt system of the world was weaker than the first two. I have even reread it.

I wonder if you could invent a fractal theory of writing. Where paragraphs have a certain amount of scene setting and a climax sentence that imparts important information to the reader. You could analyse a book to find the rhythms that each author has to this process of building up.

To be honest I tend to skip the songs in Tolkein, because I know they will have none of the plot forwarding bits in.

For non-SF I tend to dislike short stories, because I like believable relationships between people. With SF it is a little different because the ideas can be more important than the people (stanislaw lems short stories spring to mind) and a short story lets you focus on the idea with being distracted by other things.


"...but it has to deliver as significant a cognitive reward to the readers as any other 300 page book; otherwise they'll feel cheated, after putting 5 hours of their life into reading the thing, and they won't pick up the next volume."

Readers want closure. So they may pick up the next book in the hope of getting it, overriding the cognitive dissonance of being cheated. I don't doubt for as minute that the multi-book novels are working that psychology.

I'm certainly getting very reluctant to read a novel that I know is part of an unfinished series, which is one reason why I have stayed away from your "Merchant Princes" saga - I want to know that it will be finished and that other readers affirm that it is worth reading the series.

"Dune" is a classic series that wore very thin after "Children of Dune" and seemed to be milking the readers. More recently, I felt quite gypped with the Clarke/Baxter "Time Odyssey" series that left me annoyed that closure wasn't achieved after 3 books, with one author expired.


I had the exact same feeling about the Baroque Cycle. In fact I've never made it through the whole thing despite multiple attempts, and I loved Stephenson's other work.

On the other hand I am one book away from finishing Patrick O'Brian's Aubrey-Maturin series and only avoiding the 20th one because I don't want it to end. Of course each of those novels is only about 300 pages and I've read a few of them in a single sitting. I've also noticed that the early books in the series were much more self-contained whereas the later novels feel much more like chapters in a longer work, many of them picking up exactly where the previous one left off.


One of the things I've discerned over 10 years in bookstore retail is that preferences in fiction tend to fall into two generalizations which I think of as "Movie Books" versus "Television Books." People who want Movie Books are those which tend to be self-contained and intended to be read over a relatively short period of time. These are the books people are looking for when they have a long plane flight to Peru. Television books on the other hand are those which are either going to take a while to read or are part of a series, and can have a broader plot diversity over a longer period of time. (Hence why supernatural Romance series have become so popular, they're Buffy surrogates. Neil Gaiman on the other hand can get by with a new novel only every 2-3 years, because it's more like when a cult film director has a new film out.) And yes, there are many, many books that fall outside of these two admitted generalizations. Still, when you spend 40 hours a week in a bookstore watching other people's buying habits, you tend to notice patterns.


A week or two ago, I was at a local bookstore when Lois McMaster Bujold was there to sign books. She said something that was relevant to this: that we have a literary theory of the novel, but not of the series of novels as a literary form. She speculated that this was partly because you couldn't easily assign a literature class to read half a dozen series of novels in a semester and compare all of them. She also commented on definite differences in her approach to her three series of novels, each of which had a different internal structure. This whole topic of the nature of literary works larger, rather than smaller than a novel may be underexplored.

One of the big tradeoffs in series seems to be between having each volume be self-contained and self-explanatory, and not boring the reader of the whole series with repetition of familiar material. The first Vorkosigan novel I read was Mirror Dance, whose opening scenes are from the viewpoint, not of Miles Vorkosigan, but of his cloned brother Mark; not knowing any of the previous story, I was thoroughly confused for about the first third of the book—though not enough to stop me from reading the rest of the series, eventually. But it was a bit of a challenge. The Aubrey-Maturin books were a bigger one: I started out reading a random book in the middle of the series, and just couldn't get into it—going back to the first volume and reading through in order worked much better.


I enjoyed the Baroque Cycle, although I admit, I felt that the journey was its own reward—getting to wallow around in Stephenson's language, historical and scientific digressions, etc was at least as satisfying as the overall story's movement.

Beyond the novel and multi-part series there's another popular form in genre fiction. I'm not sure what the name of it is, but it's what we see in Larry Niven's "Known Space" or Singularity Sky + Iron Sunrise: a consistent universe explored through stories that stand alone. This is more reader-friendly.

With standalone novels inhabiting the same universe, the 5 or so pages of refresher material suffices. Contrast with the Baroque Cycle, which I plowed through uninterruptedly, and would have had a much harder time with it if I read each volume as it was published. Unless I re-read each existing volume before reading the latest, it would have been impossible for me to recall all the characters and events that were in play.

"Cognitive reward" (or what Walt Disney would have called the "weenie") is an interesting way to look at the payoff in each volume of a multi-part series, but I'm not sure I agree with it. For me, it comes down to trusting the author. I remember picking up the SF book Chung Kuo: The Middle Kingdom—a story with lots of world-building and intrigue. It was obvious that it wasn't going to wrap up by the last page, but when it ended there was a notice that it was the first in 7 volumes, and I felt used—like the author was treating my attention as an annuity. I never picked up any of the subsequent volumes. I wound up feeling the same way reading Brin's Uplift Saga. The Book of the New Sun, on the other hand, for all its intricacy and bamboozlement, never left me feeling that way.


And I should mention that I later learned Gene Wolf had written the entire Book of the New Sun before he had even discussed it with his editor, so my response to it is perhaps unsurprising.


I've been trawling/trolling the bookstores of London trying to find the third book of Merchant Princes for the past month and still no dice. Out of stock at the supplier apparently.

A note on Chung Kuo: the first book got me hooked on the series but by the end it was obvious the author was just fulfilling contractual obligations :( I still think the first few are worth reading on their own.

Jordan's WOT is of course the king of epics, and he did indeed die before finishing the last book as JDC feared, although the final book will be published, and I will buy it just like I did all the rest :)


Count me as another reader of the Baroque Cycle who enjoyed the diversions immensely. To me, they were the whole point of the exercise. The plot was more like an excuse to thread them together

OTOH I also thorougly enjoyed all your books so far, so no idea what that tells us. Probably that I like various kinds of things ;)


Note to Jake at #12. You can get "The Clan Corporate" new or used at either or (a great source of secondhand books).

As for the Baroque Cycle, I agree that the journey was its own reward, considering the sheer density of history, science, tech, shenanigans, ... And I think it's obvious that it wasn't the publisher that made Stephenson split it into three books (coincidentally or not, they are each about the same size as Cryptonomicon).

Charlie: excuse my presumption but ... instead of dry-swallowing their pills after world-walking, couldn't the immensely wealthy Clan get their own bottled "jump juice" - a mixture of beta-blocker plus headache pill in liquid form?


Jake @12: "The Clan Corporate" is due out in paperback in the UK in November (from Tor UK, aka Pan Macmillan). Consequently, it's likely that those bookstores who were importing grey-market copies of the US paperback have stopped doing so (because once a UK edition is in the works the UK publisher starts sending nastygrams to people importing too many copies).

John @14: if I didn't have a background in pharmacy I'd probably do what you suggest. Alas, medicines behave interestingly in liquid rather than solid form, and preparing the sort of thing you're asking for would actually be quite difficult. (And I know more than enough about how difficult it would be to not want to go there.)


One technical term for the multi-novel story is "roman fleuve." The most successful one I've read is Anthony Powell's "A Dance to the Music of Time."

Note that there are other kinds of series. Ones in which nothing really changes, so every booklength episode starts from the same place and ends up in the same place. Ones which aren't really related, except that the writer has found out that a particular background sells.

And, of course, the ones where the first book was intended to be a stand-alone. Jack Chalker's Well of Souls books are one example; making the second, etc. book workable required that Nathan Brazil turned out to be lying when he said he was God.


The brutal thing about the one-story-spread-through-multiple-volumes is that one's bookstore never has the next volume. Heck, by the time I hear about such a series, even Amazon frequently has trouble finding the early volumes -- I've three times now been strung along for months by Amazon before they admitted that they just couldn't find a copy of volume 2 of something I'd been getting hooked on (but that the author was already on volume 5 of).

Here as in so many things, Bujold lights the way: Print omnibus editions for us laggards, and then at least we've a chance of catching up to the author.


I'm glad I stumbled across this. I'm a dark fiction (primarily) author who does great with short stories (exhibits A and B, my debut collection and an antho I edited, are excerpted on my website) but has yet to find that one special tale I can have a novel-length relationship with.

I have had numerous comments about the length of my work, ranging from complaints about there not being more story (which to me means I'm doing something right) to the complimentary label of "short, quick jabs to the solar plexus."

I like your breakdown of the different lengths of stories and the amount of work and development a writer can put into each. I might end up linking to this; my readers may find it helpful in understanding me.


Hmm . . . at work, me and my colleagues have a default project: the 60-point course. This is something that (ostensibly) takes the punters about 26 weeks to read, at 20 or so hours a week. It involves writing or selecting about 450,000 words of text, as well as sundry other stuff. So it's a bit like a long series of novels. But the payoffs are split up into blocks (c. 80,000 words), units, (c. 20,000 - 1 week's work) and sections (moveable feast but about 4,000). There's a fractal unity there, I think: also, it's obvious that we are using days, weeks, and months as the default periods for our 'story arcs'.

This isn't much of a precedent for anything, but it makes me think that I agree with Lois M B.


When I told Charlie that my bookgroup was reading The Family Trade, the first of the Merchant Princes Saga (heh, wrote Prices), he warned me that we were reading the first half of a book. I passed that on to the group via email, and although most liked the book but were disappointed with the abrupt end, one was so furious that we had to close both doors between the community room and the rest of the library. We're sure it's SF, btw.

I've recently read a couple of books that would have been better as novelettes. The problem is that authors don't sell those as often or make proportionately as much on them, so even if they wanted to aim at that length, it may be to their benefit to pad it to a novel.


Alex @6: In movie trilogies, the middle installment often sucks. The Indiana Jones series is perhaps a classic example of that. I haven't noticed that nearly as much in sf - most of the series I've read have been of surprisingly uniform quality. And in the case of our host, uniformly high quality.

Dune is the big exception. I loved the first, the second left me cold, I could barely finish the third. I finally forced my way through God Emperor two decades later, and hated it. But to my astonishment, I really, really liked the last two - and wish that Herbert had been around to write more. (rather than having his son type mediocre prequels.)

That said, I still think Herbert's best book is the Dosadi Experiment.


Adam @11: The Book of the New Sun wasn't quite that done in advance. There's actually a neat book of essays and other stuff by Wolfe that includes The Castle of the Otter, a small volume covering the history of writing and publication, an essay about language in Urth, about warfare in the age of the Autarch, and this and that and the other. Wolfe did have an overall plan, but it was originally for a trilogy, with the fourth volume coming in as he realized it just plain would take more space. What he did was get a first draft of the next volume done before final revisions on the last one, so as to accommodate future developments.


Charlie, apropos of this - and I hope you'll forgive the slightly critical observation - my favourite books of yours are Accelerando and The Atrocity Archives, which I believe were both written serially, in short-story-sized chunks. Whereas, fascinating though the concepts are, I can't get along with The Merchant Princes. Is it just me, or do you think there are differences in the way you wrote the former books / how they turned out?



To remind you of a one star review you kindly posted -

"It seems that the author believed that the series would make him rich if he could keep the series going so he trashes the whole concept of the first two books to do it. The first two books were great but this one is so much of a disappointment that I would advise people not to start the series."

I read that just after finishing the second book, and now being half way through the third, I find I agree with it.

With the one caveat - I love where you've gone.

For me, the first two books set up a naive sandbox where the fantasy McGuffin doesn't have world wide consequences, the third book starts to apply common sense to it - how would societies mixing like that really work?

From a narrative point of view, it's a pity the first "chapter" got split in two (as that's possibly how you got disappointed readers like the above, expecting a pat conclusion in the third book rather than seeming almost to veer off into another genre). From a publishing point of view, I guess that pales into insignificance beside getting people to buy the book.

Me, I'd been waiting for the other foot to fall all along :D Good luck with the size 6 conclusion!


Ummm. I think I might be an outlier but... I read about half the speed that this post suggests is normal, and frankly, I prefer books that are at least 500 pages (though I am a philosophy devotee (won't call myself a philosopher simply because I don't have enough people to discuss philosophy with) and I concede that may give me more patience as long as ideas are being developed). Where do you feel this leaves first-timers trying to write their first novels? (Asks the guy who is about a third the way through his novel and its already resting on 78000 words!)


Russ @24: my elevator pitch for the first [big fat] book was, "investigative journalist discovers paratime travel, and that she's a long-lost heiress; says "sod this for a game of soldiers" when expected to do the long-lost heiress thing for real (sans modern plumbing), kicks over the traces, and goes off to make her fortune. Comes up smelling of roses."

And the elevator pitch for the second [big fat] book/story-line was, "the pigeons Miriam unleashed in the first story-line all come home to roost. And they crap everywhere."

(Hope this makes things clearer ...?)

YodasEars: speaking purely from a commercial standpoint, publishers right now will be very reluctant to buy a first novel that runs to 240,000 words, for purely commercial reasons; it's significantly more expensive to print 5000 copies of a single 800-page book block than to print 5000 copies of each of two 400-page books, and at the same time, because of caps the retail chain imposes on cover price, you can sell the two 400 page books for twice as much as the single 800 page book.

So if you're wanting to tell a 250,000 word story, I think you'd do better to restructure it as a trilogy of 80-90,000 word books. Which doesn't necessarily means you need to split the novel you're working on right now, but it'd be a good idea to see if you can put in some kind of major crisis or climax at the 80,000 word mark, and an even bigger one around the 160-180,000 word mark. (Then when you go back and redraft your single novel as a trilogy, it'll fall neatly into three volumes.)

I wouldn't start off by trying to tell a 250,000 word story in the first place. You're working on it now, and you might as well continue if it's the story you want to tell, but a common pitfall new writers make is to start on a series or trilogy and send the first book out. By the time they discover it's not going to sell a couple of years have passed ... and they've written books 2-4, which are now unsalable (because the first volume wasn't as well written, because they were still learning the art).

It's safer to write separate, stand-alone books, each of them a bit different from the one before. By all means make them the first volume of an open-ended series if that's what you want to write, but don't start on the rest of the series until you've sold the first one. Instead start on the first volume of a different series. That's my recipe. It worked for me; I make no claim that it'll work for you, but I think it's worth considering.


I had always viewed Dune Messiah and subsequent stories as an afterthought, rather than a pre-planned series. Consider that Dune was serialised starting in 1963, whereas Dune Messiah (also in serial form) didn't come out until 1969. It doesn't look as if Herbert had Messiah up his sleeve in 1963.


John @24: not true. Herbert planned the trilogy from the start. In his own words: "I conceived of a long novel, the whole trilogy as one book about the messianic convulsions that periodically overtake us ... This grows from my theory that superheroes are disastrous for mankind, that even if we find a real hero (whatever that may be), eventually fallible mortals take over the power structure that comes into being around such a leader."

Also, Charlie: there never was a nine-hour cut of Lynch's Dune. According Raffaella De Laurentiis there was a working print - sans post-production sfx and full of placeholders - of about 4 - 5 hours, but it wasn't a coherent movie and was never intended as such.


Ian @28: well there should have been such a movie! It might actually have worked. (As a series.)


You'll get no disagreement from me there.

Be interesting to see what Peter Berg does with the book. His previous films include Collateral, Cop Land and The Last Seduction...


JDC @ 3

"Definitely right about "chapter" books. I've always been resistant to n-ologies for just this reason. (That and the fear the author is going to die on me in the middle.)"

The classic example of delay between books has to be Jack Vance's "Demon Princes" series, the first three of which came out at reasonable intervals, followed by a twelve year gap....



Torgo - I don't find the Merchant books quite as readable as Charlie's other works either (Still good, though). I've not progressed beyond #2 so far.

Personally I tend towards more to the serial short stories which I hope to string together in a "braided novel" at some points.


Charlie @28: Well the mini-series certainly worked better than the movie, although it had different weaknesses.

Regards Luke


I just count the Merchant series as a good serial...well kinda.

It's definately different from a lot of Charlie's other work, but in saying that - most of his work is different fromt he rest!

I mean you've got Epic-SF, Lovecraft horror/comedy, sf/horror thriller, hard sf...

If there was one thing I'd ever accuse Charlie of, it'd be diversity. If you like a specific genre (in so much as there are any realy genres), then reading all of Charlie's books might flummox you.

Of course I'm easy (or so some of my ex's tell me), so I'll give anything he brings out a go. I'm just putting my pitch in to play Bob Howard in the Laundry mini-series..


I'm completely fascinated by the postings on this subject. Charlie, when you were in Portland, Oregon, I asked you what happens with fantasy writers that they seem to take so long to get to their conclusions. You said you thought that all authors get hung up in different Points of View. I think you're partially correct, but I also think they get hung up in world-building.

I agree with Jake@12 that I will almost certainly buy the final volume of Jordan's WOT. But, I feel a little resentful because a number of the middle volumes were really not that good and did not forward the plot much. I think Rigney agreed, and probably wished he had tied things up sooner.

I loved the Baroque Trilogy because, not in spite of, its many digressions.

Finally to end this long winded post; For those of you who have any interest in history, and particularly British history, Anthony Powell's "Dance to the Music of Time" (Dan Goodman@16) is beautiful, moving, well written and has marvelous characters. It starts indirectly with World War I and traces the decline of the British Upper Classes with grace, beauty and a very keen eye. I highly recommend the whole sequence. Believe me it's very relevant.

And, I love the Merchant Princes. Damn it, why can't you guys write them as fast as I read them!

Rick York


One more thing: Does anyone know where I can get a copy of the last volume of Chung Kuo - "Marriage of the Living Dark" for significantly less than $90?

Rick York


The trade of queens? Why does that remind me of the game of kings?

Hmm, must read it again.


I thought the Dune miniseries had really great hats.


I must say, I love very long books, with lots and lots of digressions, and closure has never scored high on my personal list of things that matter in fiction. The Baroque Cycle hit just that spot for me - I had so much fun while I was in there that I honestly didn't care much where it was going as long as it just kept going. There are plenty of examples of great fiction where digression seems to be more than half of the point: Infinite Jest, A Hundred Years of Solitude, JR, Don Quixote, The Illuminatus Trilogy (I could go on).

Conversely, a lot of short story writing, seems (to me) to end up being all about punch lines (and by extension all about closure) - I really love short stories that avoid that sort of thing, but they're rare.

40: 35, that might be one of the more awesome things about having proper AI authors; speed up their clocks, and they WILL write novels just as fast as we can read them! Imagine Neil Stephenson turning out one Diamond Age-equivalent every week, or Hamilton a Night's Dawn Trilogy once a fortnight!

Charlie, you'll have to software-clone yourself forthwith. I demand one chapter of Accellerando every hour, please.

Of course, my own software-clone will be pissed, because he'll be sped up too.


Russell @40: unfortunately you'll then run into the problem that authors with a unique voice generate new ideas at a decreasing rate -- when they first splurge on the scene with a bunch of interesting stuff to say, what you're seeing is typically a backlog of ideas that took a third of a century to build up while they were concurrently learning the craft of writing and acquiring enough skill to be readable.

Turn up the clock speed and you'll either get more of the same-old, or an author who spends an increasing amount of time brooding about burn-out and angsting because the rest of the world is running so s-l-o-w that they're not getting enough stimulating new ideas.


Dan @ 16

So Riverworld is a roman fleuve?


Apropos Russell and Charlie (@ 40 & 41) have you come across "An Introduction to The History of Bitic Literature" In Imaginary Magnitude by Stanislaw Lem? My memory of it is no longer complete, but I recall that it has a lot to say about the possible futures of AI literature - especially it's inevitable divergence from the sort of stuff us fleshy folks might want to (or indeed be able to) read.


Rick York @ 36:

Don't try too hard to get hold of Marriage of the Living Dark. It's pretty dire.....


I love the Baroque Cycle, but sometimes itch for a bold editor. I feel the same way about the books between the nicely rewritten and polished Books #1, 2, and 7 of the Harry Potter cycle, versus the ones in the middle which sometimes need another draft. I love what Stephen King has done for the Science Fiction/Fantasy/Horror roman-fleuve, where at least half of his novels interconnect in a way requiring hyperlinks, and not onbvious to the nonobsessive reader.

Robert Heinlein's juveniles are mostly a roman-fleuve, as are his stories and books in the Future History roman-fleuve. Isaac Asimov had his juvenile cycle (Lucky Starr), and two or thee main adult cycles: Robots, Foundation, and others, which he stitched together later to make his dekology, which had the 3 authorized addendum novels by Bear, Benford, and Brin.

Likewise, Heinlein's late novels interconnect, and present a theory of the meta-novel and the multiverse.

Besides the Dune example given here, and Bujold, I think that Jack Vance's and Roger Zelazny's cycles deserve mention, and Stephen Baxter, and the Culture novels of Iain Banks, as well as Caroline J. Cherryh, and, Ben Bova's juveniles, and, oh, too many other great examples. But back to the theory.

Re #16:

roman-fleuve [Encyclopædia Britannica]
(French“novel stream��? or “novel cycle��?) series of novels, each one complete in itself, that deals with one central character, an era of national life, or successive generations of a family.

Inspired by successful 19th-century cycles such as Honoré de Balzac's Comédie humaine and Émile Zola's Rougon-Macquart, the roman-fleuve was a popular literary genre in France during the first half of the 20th century.

[end Britannica excerpt]

Sir Walter Scott did this in English, concurrently with inventing the Historical Novel. He was definitely writing for money, to pay off a debt that he could have escaped by bankruptcy, but dealt with honorably. At the same time, he was putting the History of Scotland, and its folklore and culture, on the world stage forever. Extremely readable today, as I've done, with a nice complete edition picked up for a song in Edinburgh.

I adore the Patrick O'Brian's Aubrey-Maturin series, which is arguably the best example of what Sir Walter Scott did, but with multiple countries, and the essentially science fictional feel of the naval jargon and technology which, even if outside your initial knowledge, has that hard SF feel, with rivets aplenty.

Balzac was doing something even more interesting. Honore de Balzac (1799-1850): major author of France. Interesting essay on his Fantasy in "The Encyclopedia of Fantasy", John Clute & John Grant, New York: St.Martin's Press, 1997, pp.83-84.

Balzac planned on writing 100 novels in the roman-fleuve. He visualized these as modules of a vast intelligent agent simulation.

After the 100 novels were published, he would publish 10 nonfiction books on his theories of Psychology, Sociology, and History, using the characters and interactions of the novels as experimental or illustrative data. Then he would write one incredible Philosophy book, drawing on the 10 Psychology, Sociology, and History books as meta-experimental data.

This bottom-up design had two main flaws.

First, the editing of galleys. Printers drew straws to avoid handling his manuscripts, because his notes in the margins of the galleys often exceeded, in word count, the printed draft text. This was a proto-word processing.

Second, his life style was warped to meet the design. He'd sleep until afternoon in his Paris garret, get up, drink copious coffee, go to parties with the aristocracy and glitterati, making mental notes on their costumes, conversation, gossip, flirting, and the like. Then back to the garret, drink more coffee, and write new stuff plus mark up galleys. Lather, rinse, repeat.

Bottom line: his novels are an amazing read even today (albeit I'm stuck with English translation). But his heart gave out, essentially from caffeine poisoning.

So please, Mr. Stross, use your Pharm wisdom to stay healthy. Ideally, forever.


Jonathan Vos Post wrote that Robert Heinlein's juveniles are mostly a roman-fleuve, as are his stories and books in the Future History roman-fleuve.

I just completely fail to get that. There are very faint hints of a common history in Between Planets and Starman Jones; other than that, no two of the juveniles have a common background. The six later juveniles starting with Starman Jones have incompatible FTL technologies (notably Tunnel in the Sky, with no starships in sight) and often conflicting galactographic backgrounds (the "Five Galaxies" vigilantes in Have Space Suit—Will Travel are nowhere to be seen in the interstellar diplomatic crises of The Star Beast). Perhaps the thirteen books from Rocket Ship Galileo to Starship Troopers could be called a thematic triskaidecalogy (in the sense that James Blish said Black Easter/The Day After Judgment, A Case of Conscience, and Doctor Mirabilis were a thematic trilogy), but they are not a "series" in any conventional sense.

Heinlein's published letters show that he was originally envisioning Rocket Ship Galileo as the first of a series of "young atomic engineers" books, with future books focused on Mars, the asteroids, and corporate enterprise. It's rather fun to envision obtaining volumes of that series from the alternate timeline where it was written; it would have been a genuine roman fleuve, if I understand the term right. But it would also have been much more "juvenile" than what Heinlein actually gave us. His books for teenage boys have some very adult concerns, philosophical (the riff on Plato's Republic in Space Cadet, where the three space services clearly reflect Plato's three classes of citizens), political (the discussion of the inevitability of war in Farmer in the Sky), and even familial (the blended family in Farmer in the Sky, at a time when such families and their issues were hardly acknowledge even in adult realistic fiction). I think on the whole we're fortunate that Heinlein did not feel obliged to create any kind of roman fleuve, but used the "series" concept to write whatever story he felt like writing.


I have to agree with the people who said they enjoyed Baroque Cycle. I never understood this stuff about Neal Stephenson's endings not being good, IMO the work stands as a unit, you don't need to pack all the interesting and fun stuff at the end. Consider Glasshouse: I wouldn't say the ending is terribly groundbreaking, but it's still a wonderful book.


I enjoyed Baroque Cycle, especially the digressions. Apart from anything else, there was the game of spotting the forward links to Cryptonomicon (and rereading that afterwards in quite a different light).

I will admit that I'm still not sure what it was all about... but in a fun way.



I think of this as similar to soap operas vs. serials. A soap opera never ends: it brings new story lines in before others finish, so that viewers and readers stay hooked. It is primarily a vehicle to sell advertising. "Lost" is pretty much a soap opera. By contrast, a serial has a beginning, a middle (or two) and an end - it exists to tell a single story. Think "I Claudius" (which was, of course, based on a pair of novels) or "Edge of Darkness".

The more that a book in a serial lacks closure, the more it is akin to a soap opera, although of course the aim is to sell books rather than advertisement slots. Personally, I prefer stories with endings (unless you can produce something as complex as "Dhalgren", which doesn't end as such, but is complete none the less). I'm happy with series as long as each book stands alone, or even if the duology or trilogy was clearly conceived as a single entity and marketed as such. But you're right; the longer the book, the better the pay-off has to be.

As for "The Baroque Cycle", well, after persisting to the end of the execrable "Cryptonomicon", I gave up on Stephenson. That taught me the lesson that it's OK not to finish crap books.

I don't like the Aubrey-Maturin books either - O'Brian clearly rejoiced in his naval spoddery and this doesn't help his writing at all. Thompson's more recent "This Thing Of Darkness" was much better - and a single book too.


Wait... you threw a Neal Stephenson book at the wall? I hope the damage was not too expensive.


Great article. Just to be a pedantic completionist, I'd note that narrative poems can be a story-telling format that's even shorter than the typical short story.

Indeed, a good narrative poem is a story that's flensed of everything except the very barest of bones. I've ream good poems that manage to tell (a very, very simple) story in under forty words.


Andrew, here's my shortest story ever, in seven words:


(Picture it as a newspaper headline.)

But seriously, the shorter the form the harder it is to make a living at it. It may be shorter and desner but it ain't necessarily easier to write. And ultimately, we tend to be paid by the kilogram of dead tree ...


Charlie @26: "you can sell the two 400 page books for twice as much as the single 800 page book"

Knowing that, and knowing from reading this blog that the first Merchant Princes book had been split in two, has seen me get the series from the library rather than buying the hardcovers - as I have with nearly all your other books. I just couldn't shake the feeling that I was being ripped off by the publisher. (As an aside, there's clearly a Stross fan working in the Glasgow public library system - they had half a dozen copies of each of the first two books when they were only available on import.)

It's just a shame that the spectrum of story-lengths you describe above isn't better correlated with price - I suppose the fixed costs per volume overshadow the increased production cost for fatter books. It should be transparent for ebooks though - perhaps these should be sold by "weight", so many pence per kiloword?


YodasEars @25: I don't know if you're an outlier or not, but I also read much more slowly than the conversion Charlie used. A 500 page novel is more like 12-15 hours of entertainment for me.


Apropos @53, some ebook vendors do gear their pricing to the length of the ebook, at least for short stories. Trouble is, it's nonlinear; if you charge proportionately for a 5000 word short and a 100,000 word novel, and the novel goes for $5, then the short story goes for 25 cents. And once the distribution/retail chain takes its 70% cut ...



Thanks Charlie!

Times like this, I'm glad that my mind is a miasma of ideas. ;)

Ralph Giles @54 Thanks man. Feels much better knowing there are others out there - my mum is one of those people who can nail a 600+ pager in a single 2 and a half hour sitting, so I've always felt kinda slow...


I thought HP book 5, at 870 pages, and #1 best seller on preorders 6 months before release was a sign to publishers that big books can work.

I'm a bit surprised that longer books are in fashion, given the ever shrinking attention span. But perhaps the non-book readers are all watching their 7 hours of TV, so the market is those, select few, who can cope.


Being a fan of Lynch's Dune (more as a visual interpretation of the universe rather than an accurate transfer of the book) I looked into the mythical 9 hour version for some time. Basically all the information I found points to the fact that there isnt one. Nine hours of film may have been shot, but could never have been assembled into a coherent whole. So rather than being 'lost' it actually never existed to begin with. Some extra scenes may exist, such as those clumsily added to the Alan Smithee TV version, but it is arguable wethter they add anything worthwhile to the film.

After pining for it for a long time I was happy to let it go. I know you're just using it as an illustration for a different topic, but I'd hate to think it set some other Lynch or Herbert fan off hoping.


This strikes me as why TV series tend to peter out rather than go boom a the end. Once you've reached season six or seven, there's no way in hell you can have a climax big enough, unless the earth explodes. And gods help you if you get to season 18 (the Simpson's) or Season 42 (Doctor Who).

Which is the long way of saying I hope no one thinks there will be any satisfactory climax at the end of Lost. That ship sailed off, circled the three toed statue and then fell into a time warp.



About this Entry

This page contains a single entry by Charlie Stross published on May 8, 2008 12:14 PM.

Not long now ... was the previous entry in this blog.

Why your internet experience is slow 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