Paul S. Kemp's defense of shared world fiction

[Desdichado]

Cross-posted from another discussion... I thought the topic interesting enough to me personally that I wanted to have the conversation again with a different crowd.

An oldie but a goodie. What're your thoughts?

Blog of Author Paul S. Kemp - Why Authors Grow on Different Trees

Why Authors Grow on Different Trees Jan. 23rd, 2007 @ 01:16 pm

Consider: Apple A grew on a tree. Apple B also grew on a tree. Therefore Apple C grew on a tree.

Consider further: Raymond E. Feist writes bad non-shared world speculative fiction. Terry Brooks also writes bad non-shared world speculative fiction. Therefore George R.R. Martin writes bad non-shared world speculative fiction.

Obviously neither the the apple-argument nor the author-argument is deductively valid (and for the record, I do NOT think that Raymond E. Feist or Terry Brooks write bad non-shared world speculative fiction; I am just pulling their names out of the air for the sake of an argument). Perhaps, however, one or both is inductively strong.

A strong inductive argument requires a conclusion tightly connected to and drawn from the set of expressed premises. So: Apple A grew from a tree, as did apple B. So, too, did apple C, and ad infinitum. The claim that apple Z also grew from a tree is, therefore, inductively strong. The fact that all of the apples are, in fact, apples, is enough to allow us to draw conclusions about how one grows on the basis of how others grew. In other words, based on our experience with the world, we know there is a strong connection between the quality “appleness” and the way in which apples grow. Other differences that might exist between the apples (e.g., size, appearance, taste, color) are not relevant to the conclusion about growth that we want to draw.

Note, however, that we could not make an inductively strong claim about how a McIntosh tastes based on how a Granny Smith tastes, because the difference in type between the two apples is relevant and significant to the conclusion we want to draw. That difference disconnects the premises from the conclusion. The quality of “appleness” is here not enough because, based on our experience with the world, we know that a McIntosh does not taste like a Granny Smith (except at the most general level).

This is all common sense, I realize, but I have a point. Let’s try it out on another hypothetical:

R. Scott Bakker writes bad non-shared world speculative fiction. Robert Jordan writes bad non-shared world speculative fiction. Gene Wolfe writes bad non-shared world speculative fiction. Therefore J.R.R. Tolkien writes bad non-shared world speculative fiction.

Doesn’t work, does it? Make it a string of fifty names in the premises and it remains a weak argument. The reason it's weak is the same one that prevents us from concluding anything about the taste of a Granny Smith on the basis of the taste of a McIntosh – the premises lack any significant relation to the conclusion. It’s true (solely for purpose of this illustration) that all of the named authors in the premises write non-shared world speculative fiction and that they write it badly. But the only relation the authors named in the premises share with Tolkien is that he, too, writes non-shared world speculative fiction. And the quality of “writing non-shared world speculative fiction” is not enough to allow us to make an inductively strong claim that Tolkien’s non-shared world speculative fiction is bad solely on the basis of the other named authors writing bad non-shared world speculative fiction. After all, based on our experience in the real world, we know that authors vary a great deal in terms of talent, style, and tone. We’re trying to conclude something about the taste of a Tolkien-McIntosh on the basis of the taste of Bakker-Braeburn.

I’ll bet all of this is non-controversial (except to philosophy majors, who are even now critiquing my misuse of various terms like inductively strong and deductively valid; to all of you, understand that I had logic and symbolic logic as a an undergrad almost ten years ago; cut me some slack; I’m operating from memory here ).

This brings me to my point (finally; sheesh!).

Notwithstanding the foregoing, exactly that kind of inductive reasoning is applied with alarming regularity to shared world speculative fiction writers. I frequently hear/read comments that are one variant or another of the following: “I read a few bad shared world fantasy novels back in the 80s. Therefore all shared world writing is rubbish.”

This kind of flawed reasoning is commonplace with respect to shared world speculative fiction. It is also nonsense. The mere fact that a piece of speculative fiction writing is set in a shared world has no relevance to the question of its quality. As with non-shared world speculative fiction, the quality of the author is the determinative factor as to the quality of the work. An example to further highlight the point:

Paul S. Kemp wrote a bad shared-world speculative fiction novel set in the Forgotten Realms. Tracy Hickman wrote a bad shared-world speculative fiction novel set in the Dragonlance universe. Timothy Zahn wrote a bad shared-world speculative fiction novel set in the Star Wars universe. Therefore William King’s shared world speculative fiction novel, set in the Warhammer universe (nay, all speculative fiction set in a shared world) is also bad.

Doesn’t it seem absurd to so generalize, both across lines, across authors, across subject matter? We’re all different apples. Hell, even within the same line (say, the Forgotten Realms) authors vary so much in terms of talent, tone, and style, that concluding anything about the quality of one author’s work on the basis of the work of another author in the line is silly. It’s tantamount to drawing conclusions about all speculative fiction writers who write for Tor on the basis of one speculative fiction writer who writes for Tor.

Again, it is the author’s individual talent that determines the quality of the work. Nothing else. And here's the critical point: There is no more connection between the abundance of an author’s talent and whether or not they write in a shared world, than there is a connection between the abundance of an author’s talent and whether or not they write New Weird stories as opposed to Epic Fantasy, whether they write for Tor as opposed to Baen.

Now, I do not want to venture into the briar patch of why so many readers (and even, I’m sorry to say, so many, many, many other authors) engage in this kind of flawed reasoning. Analyzing human nature is not the purpose of this short essay. My purpose here is simply to expose the underlying weakness in the all too often repeated claim that all shared world speculative fiction is bad. It’s not. Not by a long shot.

I am, of course, not claiming that all shared world speculative fiction is good (any more than I’d claim that all types of apple are good; McIntosh apples stink; curse you, Mcintosh! Cuuuurse yoouu!). It isn’t, any more than all non-shared world speculative fiction is good. The quality of shared world and non-shared world speculative fiction varies by author. By author. By author. And those who dismiss one or the other with a hand wave and unjustified generalization tell us more about their own biases and personal psychology than they do about the category of fiction they purport to be commenting on.

Here’s my plea to those who do not read shared world fiction based on the conviction that it’s all bad – take a bite of the apple, a different apple than you’ve tried before. And if you read a bad shared-world speculative fiction novel, treat it the same way you would a bad novel set in a non-shared world speculative fiction setting – put it to the side and don’t read that author again. But don’t make the mistake of generalizing the quality “bad” to an entire category based on such a small sample size. If you didn’t like the McIntosh, try the Fiji. If not the Fiji, maybe the Gala. There are plenty of good apples out there, believe me.

(Postscript: I am aware that “bad” as I’ve used it above is not self-defining. Further, in the context of art, which is what we’re discussing, “bad” is a slippery concept. But defining “bad” is not necessary to the argument.)

Personally, I've sampled a lot of apples. The idea that "just keep trying more, you're bound to find something you like!" is more masochistic than logically intuitive at some point.

Also, there are constraints put on shared world fiction that Kemp doesn't acknowledge. I read some Eberron novels a while back that I think were held back by their format and the restrictions placed on them by wordcount and whatnot. I.e., they were decent books that could have been much better books yet if not operating under some arbitrary (well, from a writing standpoint; I'm sure they weren't truly arbirtrary from a publishing standpoint) constraints about how they had to be written.

I'd say that I've tried a lot of apples, and I'm still willing to try more, but I've had enough experience to be skeptical and to have very low expectations. It's not that there's (much) about the format of shared world fiction that makes it more likely to suck compared to "regular" fantasy fiction, its just that experience shows, and not just with a handful of novels read in the late 80s and penned by Rose Estes, that most shared world fiction just isn't very good. I've taken a much broader sample than he refers to, and yet I've still come to that conclusion too.

Also, there are some external factors that work differently with shared world fiction than with other fiction, and some of them do, in fact, select for poorer writers. Or, at the very least, they fail to select against them, like "regular" fantasy fiction tends to.

That said, I think there certainly are some writers who at least aren't any worse than others operating in a non-shared world environment. Weis and Hickman made their name in shared world fiction, and since migrated relatively successfully into their proprietary Death Gate setting and Darksword settings, for instance. R. A. Salvatore may not be a great writer, but he's certainly no worse than Terry Brooks or David Eddings. And so on and so forth.


In the interest of full disclosure, I do like a handful of shared world books. Or, well, at least I don't completely dislike them. The original Weis and Hickman Dragonlance books I bought in paperback as a teenager, and I've read them a few times since, and never sold them back like I did hundreds of other books. Same for the original Salvatore trilogy with Driz'zt and Co. I think the original Timothy Zahn trilogy that kicked off the modern Star Wars licensed fiction madness holds up relatively well. The first few Thieves' World anthologies ain't bad.

Other than that, I've read many others that were somewhat forgettable, yet weren't bad per se, just also not really good. Certainly, they were within spittin' distance of most other published science fiction and fantasy I've read in terms of quality. Some of these are RPG settings, like D&D or Warhammer fiction, and some of them are Star Wars or Predator or Aliens based fiction, etc.

I've recently bought a couple books in an Eberron trilogy, but that's because it's about hobgoblins, and I like them. I have no idea if it's any good or not, and I probably won't get around to reading the two I have until the third book in the trilogy comes out later this year anyway. If not even later. I've also recently bought the first in a trilogy written by Kemp himself, and I probably would have bought the entire trilogy, except the bookstore only had the first one in stock.

I keep trying to like shared world fiction, and occasionally I actually do. Most of the time, though, it's well within the realm of mediocre, and most of the really bad novels I've read have been in the shared world realm.

 

[Desdichado]

Well! I cross posted this thread, in part, because I thought it'd have a livelier discussion here than where I originally posted it.

Guess that was a bad call.

 

[Dire Bare]

I read a lot of D&D shared world fiction, in fact I'm pretty sure I've read it ALL with the exception of what I've come to call "Dragonlance Furries" novels (novels about minotaurs, ogres, or whatnot where the characters are indistinguishable from humans).

Obviously, I enjoy them as I keep reading them.

I think there is a higher ratio of mediocre-to-awesome books within the D&D shared world "genre" than in fantasy fiction as a whole, but absolutely not a higher crap-to-awesome ratio. And if the more mediocre books fill certain needs for me, I'm happy while I wait for the next stunner in the line to be released.

I tend to view the medicocre D&D novels as simple pulp fantasy novels akin to watching episodes of Xena or Legend of the Seeker . . . . not "great" by any means but certainly enjoyable.

I read more D&D shared world fiction than non-shared world fiction because it's basically print comfort food. Like going to the Outback Steakhouse, I have a pretty good idea of what I'm going to get before I go there, and I also might get a pleasant surprise of a better than normal experience.

For my non-shared world fiction, I tend to stick with trusted authors like Tad Williams, George R.R. Martin. Jim Butcher, Stephen R. Donaldson, and others. I don't take a lot of chances with authors unknown to me.

 

[Dire Bare]

Personally, I've sampled a lot of apples. The idea that "just keep trying more, you're bound to find something you like!" is more masochistic than logically intuitive at some point.

If you have developed a dislike for shared world fiction, fair or not, randomly picking up some new shared world fiction to "give it another try" would be a bit masochistic, I think. However, if a friend recommends an author or book, don't reject the suggestion because it's shared world fiction, go ahead and give it a try! Unless, of course, your friend has a habit of recommending stinkers . . .

Also, there are constraints put on shared world fiction that Kemp doesn't acknowledge.

Constraints? Yes, writing a novel within a shared world does impose constraints an author wouldn't have with his/her own unique world. But that has nothing to do with quality.

. . . penned by Rose Estes . . .

HAH! Well that's your problem right there!

Rose, bless her heart, wrote some okay Endless Quest books back in the '80s . . . but then somebody gave her the chance at some full length D&D novels that have soured many on shared world fiction.

Also, there are some external factors that work differently with shared world fiction than with other fiction, and some of them do, in fact, select for poorer writers. Or, at the very least, they fail to select against them, like "regular" fantasy fiction tends to.

Totally disagree with this!

It's not that shared world novel lines have factors that select for poorer authors, not at all. Poor authors are selected by poor editors. There are book lines both within and without the shared world genre that have crappy editors and therefore crappy author selections.

D&D shared world fiction started with the Dragonlance Chronicles trilogy, a great novel series from the POV of a teen interested in fantasy, and a mediocre-to-good series from a more well read adult POV. It was an experiment by TSR and was one of those wildly successful products that took the company and the audience somewhat by surprise! (Like the Magic: The Gathering card game)

TSR had no experience prior to running a successful fiction line, and once they got past those initial three books they made some serious missteps. The period in the 80s and early 90s right after Dragonlance debuted saw some good books . . . but also saw a decent amount of mediocre to poor books. Almost every time I hear somebody complain about D&D shared world fiction, it's based on reading some of the stinkers from this period.

This was not due to problems with shared world fiction, but due to problems with an inexperienced company with inexperienced and poor editors (overall, I'm sure they had some good editors back in the day).

Since the dark years, first TSR and now WotC have come a long way. I would argue that NONE of the current crop of novels (since the mid to late 90s till today) are crap. And only SOME are mediocre. Many are GREAT! Future classics of literature, maybe not, but damn fine reads.

In good faith I can't recommend the Dragonlance series right now (although I do think the "Anvil of Time" series is pretty good), and can understand if Eberron isn't your cup of tea . . . but I do think checking out the current crop of Realms novels is a good idea, especially Kemp, Baker, Byers, Jones . . . the "Ed Greenwood Presents Waterdeep" series is filled with new, but good, authors and books (but I'd stay away from anything actually penned by Ed Greenwood) . . . in fact, almost anything Realms published within the past two-to-three years is pretty good (well, ah, maybe except "The Empyrean Odyssey" trilogy, that one was a bit weird) . . . especially older titles republished in omnibus format, such as "House of Serpents" and "Return of the Archwizards" . . . .

 

[Crothian]

Shared fiction worlds are hit and miss and I think there have been far more misses then hits. My issues are rarely with the restraints but with the plots. In Stars it seemed like during the 90's when I read the novels that the authors were trying to outdo each other. Each novel had something more powerful then the last and it was just stupid.

In general the books that tackled something more down to eaerth and not world saving be it Star Wars, Realms, Dragonlance, or whatever were far more interesting and better.

 

[Dannyalcatraz]

There have only been 2 shared-world settings I enjoyed, and both focused primarily on short stories, not novels: Thieves' World and Wild Cards.

The rest? Seeing those setting badges on books- Dragonlance, Warhammer, Star Wars, etc.- is like a big "WARNING: Do Not Read!" sign to me.

IMHO, it has something to do with the shared worlds pre-existing any published RPG associated with them, nor any previous intellectual property. (Yes, I know that Wild Cards were inspired by a superhero campaign played by the writers of Wild Cards 1, but there wasn't a Wild Cards RPG.) As a result, there were more degrees of freedom for the writers. They could really let their creativity have full rein.

 

[Umbran]

Totally disagree with this!

It's not that shared world novel lines have factors that select for poorer authors, not at all. Poor authors are selected by poor editors.

Note that "poor" has two meanings. One is "low quality" the other is "without much money".

If an author is really good right out of the gate, he or she can likely write their own story, in their own world, and get it sold. If an author is perhaps not so good out of the gate, selling them might require a bit of a boost - say, from a known brand identity?

I don't have the data, but I wouldn't be surprised if writing in some of the shared worlds (where the world is also a recognizable brand to the target audience) pays less than non-shared world writing. Simply put, if you're cheap, you'll tend to get lesser writing.

---

There are several shared worlds that hold up well, at least for a while. The first few "Thieves' World" books are good. The first few "Wild Cards" books are solid. Both of those series took a downward turn after a few volumes, however.

It would be interesting to know if in these cases some of the latter books were driven by contractual obligation - if you are required to print something, you sometimes can't be so picky as to what you print.

I'll note that, for most intents and purposes, the original Dragonlance trilogy is not really shared-world writing. It is more like original world writing, as Weis and Hickman were the originators - setting up the scene, on which others had to follow. Things after the original trilogy are shared-world, and are of scattered quality.

I have not re-read them in years, so my thoughts on this are suspect, but some of the first Forgotten Realms books aren't craptastic. I recall a pattern - the better ones are where the author is the first to work in that region of the world (the original Moonshae Trilogy, f'rex) - they are meshing with less established continuity, and probably have to worry less about cramming their works into the formula.

Then, you have things like the Man-Kzin Wars - which to my mind remain good through something like 10 or more volumes! Here, the cause seems pretty obvious - Larry Niven opened the world to sharing because he doesn't consider himself qualified to write war stories. But, he does have a sense for what makes good work. As I understand it he's kept a tight control over who and what gets published in the series. So, here's a case where solid editorial control (and choosing good authors - Niven has the clout to ask for stories from some quality people) has created a shared-world series that maintains quality.



---

So, all in all, I am not surprised that many shared worlds books are crap. But then, as Sturgeon's Law says - 90% of everything is crap.

Here's a thought - when you pick up a book to read, frequently you are doing so on a recommendation, or because you already know the author. There's a selection process prior to the purchase. But, if you are following a shared world, you pick up the next book regardless of who wrote it. You are thus getting exposure to a wider selection of authors - 90% of whom are crap.

The question is this: compare picking up the next book in a shared world with picking a non-shared-world book at random from the same genre section of the bookstore. Which do you expect is more likely to get you a quality read?

 

[Dannyalcatraz]

Thank you, Umbran!

Add the Man-Kzin Wars books to my list- I'd forgotten about them.

 

[AdmundfortGeographer]

Couldn't the world of The Malazan Book of the Fallen series count as a shared world? The world was a collaboration between Steven Erikson and Ian Esselmont. Ian C. Esselmont has written two novels for the shared world in addition to Mr. Erikson's multi-volume set.

How many authors does a shared world make?

 

[Desdichado]

I think there is a higher ratio of mediocre-to-awesome books within the D&D shared world "genre" than in fantasy fiction as a whole, but absolutely not a higher crap-to-awesome ratio. And if the more mediocre books fill certain needs for me, I'm happy while I wait for the next stunner in the line to be released.

Wow, really? I mean, I hate to seem incredulous, but I've never heard anyone make that claim before. I want to make sure that I'm not misunderstanding.

I read more D&D shared world fiction than non-shared world fiction because it's basically print comfort food. Like going to the Outback Steakhouse, I have a pretty good idea of what I'm going to get before I go there, and I also might get a pleasant surprise of a better than normal experience.

Rather, it's more like eating at MacDonalds. You have heard of the movie/documentary Supersize Me, right?

For my non-shared world fiction, I tend to stick with trusted authors like Tad Williams, George R.R. Martin. Jim Butcher, Stephen R. Donaldson, and others. I don't take a lot of chances with authors unknown to me.

You're a lot more likely to encounter authors unknown to you in the shared world universes, where author name isn't as important a draw, and authors are (to a relatively greater extent) much more replaceable.

Constraints? Yes, writing a novel within a shared world does impose constraints an author wouldn't have with his/her own unique world. But that has nothing to do with quality.

Of course it does. If the editing team comes to a writer and tells him all the things he must include in the book, all the things that he can't include, and gives him a very specific word count target to hit, regardless of how clearly he's able to develop the characters and plot he wants to under those strict constraints... all of those are features that don't necessarily drive towards lesser quality, but will trend that direction.

Also, if the writers are write for hire, contract freelancers, basically, they're also much more likely to be less enthused about the project than if they're working on their own material.

Again, not absolute drivers towards poorer quality, but certainly features that will tend to make the odds worse.

HAH! Well that's your problem right there!

Rose, bless her heart, wrote some okay Endless Quest books back in the '80s . . . but then somebody gave her the chance at some full length D&D novels that have soured many on shared world fiction.

Yes, I know quite well who she is. I also think you misread my post; I said that my opinion is not just based on the Rose Estes books. I've read a fair bit of shared world fiction.

It's not that shared world novel lines have factors that select for poorer authors, not at all. Poor authors are selected by poor editors. There are book lines both within and without the shared world genre that have crappy editors and therefore crappy author selections.

I disagree. Line editors have other considerations besides just getting the best authors. They've got brand name to consider. They've got deadlines to consider. They've got a master publishing plan to consider. "Regular" fiction writers don't have those constraints, and dare I say it, distractions.

Again; not a guarantee of higher quality, but certainly a situation that will tend more often to select for higher quality, whereas the shared world situation will tend to create situations in which quality can and sometimes does take a secondary role to production and brand considerations.

D&D shared world fiction started with the Dragonlance Chronicles trilogy, a great novel series from the POV of a teen interested in fantasy, and a mediocre-to-good series from a more well read adult POV. It was an experiment by TSR and was one of those wildly successful products that took the company and the audience somewhat by surprise! (Like the Magic: The Gathering card game)

Yes; I know. You keep telling me things I already know. I have those books. I already referred to them, in fact, as among the few shared world fiction that I considered good enough to retain my old paperback copies of.

And I clearly remember the advent of the D&D fiction line, and the potential inherent in the concept of D&D fiction.

Just like I clearly remember being incredibly disappointed by a lot of it, including stuff like anything written by Ed Greenwood, or the Avatar Trilogy, or the Lingering and Repetitive Adventures of Driz'zt.

Almost every time I hear somebody complain about D&D shared world fiction, it's based on reading some of the stinkers from this period.

This was not due to problems with shared world fiction, but due to problems with an inexperienced company with inexperienced and poor editors (overall, I'm sure they had some good editors back in the day).

Since the dark years, first TSR and now WotC have come a long way. I would argue that NONE of the current crop of novels (since the mid to late 90s till today) are crap. And only SOME are mediocre. Many are GREAT! Future classics of literature, maybe not, but damn fine reads.

Seriously, did you even read my post at all, or just respond to the thread title? I already referred to several new(ish) D&D fiction novels I've read, including several from Eberron. I couldn't penetrate the rambling, poorly structured plot of The Dreaming Dark and dropped it about 2/3 of the way through. The War Torn series started off on an incredibly poor note, although books two and three semi-redeemed them by being decent, if not exactly great.

I've read at least a round dozen Forgotten Realms books... probably closer to two dozen. I've read at least a round dozen each of Greyhawk, Dragonlance, Warhammer, Star Wars, and a few other shared worlds too. Read the first few Thieves World books. Read a few Dark•Matter books.

Again; I'm not some shared world neophyte who doesn't know anything at all about the phenomena. I've read at least 60-70 shared world books---maybe closer to a hundred if I really stop and take inventory, and more than half of them are D&D ones.

Lately, I've found that the Black Library seems to be the most consistently pretty good. I've read a few D&D books that I really like. I've also, hopefully not in vain, gotten myself a bit excited about the prospect of the upcoming Golarion novels.