Recently, after a long spell of receiving almost no submissions of novels for possible serialization in Analog, I found on my doorstep a large box containing not one, but four. All were complete manuscripts; all were by authors well known and highly regarded by me and many of you; and I expected to have a hard time deciding which one to buy. As it turned out, I did in fact like all of them–but decided not to serialize any of them.
Why would I do such a thing?
That question might be viewed as a corollary of a more general question that comes up from time to time, right here in "Brass Tacks," at conventions, or wherever else a reader with a strong opinion on the subject happens to catch me. It has long been true that many readers have had strong opinions on serials, either loving them or hating them; but it occurred to me while writing my last apologetic note about the four mentioned above that this might be a good time to consider the subject at greater length than a "Brass Tacks" reply permits–if only because some of the factors to be considered in deciding whether and when to serialize have changed over the years.
The old familiar arguments are not likely to surprise you. In favor of serials (from a reader's point of view) there is the fact that many good science-fictional ideas are so meaty, so full of implications, that only a novel gives ample room to explore them in a satisfying way. There is evidence for this in the readers' own opinions as expressed in the earlier days of this magazine. For many years "The Analytical Laboratory," our reader preference poll, was conducted on a monthly rather than an annual basis. Instead of ranking their favorites among all the stories published in a particular length range during a given year, readers were asked to rank, in order of their preference, all stories in each issue, without regard to length. The results showed a suspiciously strong correlation between length and popularity. Long stories usually beat out shorter ones, and if a serial installment was present, it tended to beat out everything else. That fact strongly suggested that readers liked serials–or at least the kinds of stories that were serialized.
Whether they particularly liked getting them in serial form might be another question. There have always been readers who complained about having to wait a month between one part of a story and the next or, alternatively, to save issues until all had arrived to be able to read the story straight through. (Some readers have even claimed that collecting all the issues is difficult because they tend to disappear, but this seems to me not so much a problem with serials as with those readers' personal filing habits.) Readers have also objected, quite justifiably, to getting "watered-down" serial versions of novels which later appeared as books. I'm told that this has actually been common practice in some other publications, but not in Analog. Our serials have almost always been complete and identical to their book counterparts, where book counterparts existed–except that the serials have come with far more illustrations, some of them memorable in themselves, than the books.
There is another consideration which seems to me more important than any of these, and more important now than ever before. Let me approach it obliquely, via something which might at first glance seem unrelated–but isn't.
Recently I had lunch with a talented young writer who has been spending much of his recent time in an academic environment. Since I've spent a lot of time there myself, in a variety of capacities, we got to comparing notes. We both admitted to having been somewhat disillusioned by the nature of some of the research we saw being done in academic circles–partly because it seemed to be done more because the researcher needed more publications for his resumé than because the work was intrinsically highly worthwhile. He pointed out that even when the research topic was particularly interesting and valuable, some researchers tended to "milk" it by publishing slightly different reports of the same work in many different journals. I remarked that this was not only a waste of good trees and library space, but not particularly useful from the standpoint of advancing the author's field because any competent researcher would be keeping an eye on all likely sources, and would probably catch any one of the multiple reports and therefore not need the others.
There is a somewhat analogous situation with books and serials. Why, readers have asked, should we serialize a story which will soon be available in book form? Wouldn't the space be better used to publish shorter stories which might not otherwise be published at all?
Quite often, I agree that it would–and that's why I suspect you'll find, if you bother to count, that Analog has in recent years been tending to use rather fewer serials and more issues containing no serial installments. In the not too distant past, very few science fiction novels were published except as magazine serials. In those days, it didn't really matter if readers would have preferred to get their novels in book form; they weren't available that way, so if they wanted to read them at all, they had to read them as serials. So they did, and were grateful for the opportunity.
In recent years, that situation has changed dramatically. Now there are far more books than magazines being published in this field, and most of those books are novels. Therefore magazine space is a precious commodity as an outlet for work which might not otherwise have one. In science fiction, as much as in any scholarly field of research, I think it is more important to readers and to the general health of the field to make as many worthwhile pieces of work as possible available in at least one form, than to provide multiple outlets for any one piece. (There are, of course, other groups within the field which have slightly different motivations than readers. Writers, for example, always appreciate the opportunity to be paid twice for the same story–but they also appreciate the opportunity to be paid even once. A writer getting extra pay for serial rights to a novel already sold as a book may be making it impossible for some other writer to be paid even once for an equally impressive story. Publishers sometimes like serials because they think that if readers are wild about installment 1 they'll be sure to buy installments 2, 3, and 4–but if enough readers find the whole concept of serials distasteful, that motivation can backfire.)
So why publish serials at all, given the new shape of the field? As I've already said, I think I'm doing it less, largely because of considerations like those I've been talking about. Of the four potential serials that led me to bring the subject up, all four already had book contracts. Some were already scheduled for publication so early that it would have been difficult or impossible to serialize them before the books came out. One was a fine story, but even the author doubted that it was the right kind of story for Analog. Another was something that I hadn't even realized existed until I started editing: a good book that would not be as good a serial. (A serial must break into pieces in such a way that a reader who reads installments as they come receives a certain satisfaction from each one, yet remains eager to dive into the next a month later, and not all good books do that.) But the most important consideration of all was that those novels came in at a time when I was getting lots of good novelettes and novellas. Since I can buy only up to the capacity of the magazine, I can't always buy everything I see that I like. In this case, it seemed more important to provide a primary (and possibly only) outlet for the novelettes and novellas than to provide a secondary outlet for the novels.
Nevertheless, I do still sometimes buy and publish serials, and will probably continue to do so for the foreseeable future. They aren't the right choice as often as they used to be, but sometimes, I think, they are, for several possible reasons. For instance:
1. While many serials will eventually appear as books, and often arrive on my desk with book publication dates already attached, not all do. My own first novel, The Sins of the Fathers, for instance, eventually appeared as a book, but at the time Ben Bova bought it as an Analog serial, I had no assurance that it would. It was his expressed interest in the idea, and his willingness to take a chance on it, that for the first time gave me the nerve to tackle a writing project as big as a novel. In the years I've been editing, we have serialized several novels which were well received by our readers but for one reason or another never (as of my latest knowledge) appeared in book form. Many of you wrote to tell us how much you enjoyed Bob Buckley's World in the Clouds, Dean McLaughlin's Dawn, and Andrew Offutt and Richard Lyon's Rails Across the Galaxy–but the evidence to date is that if we hadn't serialized them, you wouldn't have got to read them.
2. Both books and magazines have their own practical (and occasionally impractical) constraints on what lengths of stories they want or are able to publish. Short novels, say around 40,000 words, make book publishers uncomfortable because they don't think readers will buy something that looks "skimpy." But stories of that length are not necessarily skimpy in content, and they work just fine as two-part serials. Last year we ran Michael F. Flynn's In the Country of the Blind, and our mail made it clear that a large percentage of you loved it. But I suspect he's going to have trouble selling it as a book (unless he expands it), not because of any lack of quality, but simply because it's of a size that's hard to market.
3. Sometimes a story comes along which seems so uncommonly well attuned to the Analog audience that I think you'll want to see it here even if I do know you'll soon be able to buy it elsewhere. These include things like Donald Kingsbury's Courtship Rite, Robert L. Forward's Rocheworld, Larry Niven's The Integral Trees, and Charles Sheffield's Between the Strokes of Night. Those I'll buy even if they do arrive already in possession of a book contract–provided I don't find them in competition with something I like even better that doesn't.
4. Since most of you are reading science fiction primarily for fun rather than as part of your job, you're not as likely as a professional researcher to maintain a thorough literature search of everything that appears in the field. Therefore you may not always see everything that comes out in places other than the ones you read regularly–and even if you do, you may be limited in how much money you can or want to spend on extras. In some cases, this fact adds extra support to Point 3 as a reason for buying serials.
5. Good manuscripts of any particular type do not flow in at a constant, never-changing rate. Just as there are periods when I get more good short stories, novelettes, and novellas than I have room for, there are also occasional "dry spells" in which I see very few that meet the standards both you and I expect of Analog. In such a case, I'd rather give you a serialized novel that you'll like, even if you'll later be able to get it elsewhere, than a lot of short pieces that you won't.
The basic, overriding principle in all cases is that I'll try to give you the best stories I can find–that is, the ones I think most of you will like best–regardless of length. Other things being equal, I will choose a piece that you are unlikely to see otherwise over one that is sure to be available elsewhere. This principle will tend, in general, to lead to fewer serials and more issues without serials. But other things are not always equal, and serials will probably continue to be good things to do under some conditions, for one or more of the reasons above. If you happen to be one of those readers who doesn't like them, your best chance to get me to cut back even further on them is to write me a shorter piece that's so irresistible that I want to use it instead.
As for longer ones. . . .
It's ironic that writers sometimes ask me, "Do I have to have a book contract before you'll consider my book for serialization?" Not only do you not have to, it's a minor strike against you if you do. Having a book contract doesn't necessarily mean I won't use your story–but if I'm offered another that doesn't have one, and that I like at least as well, I'd rather use that. Good stories come in all lengths, and not all of them will appeal to all editors. But in the long run it's best for all of us, readers and writers alike, to get the maximum possible number of the best possible pieces before the public.
